Citation
Biofiltration for control of hydrogen sulfide

Material Information

Title:
Biofiltration for control of hydrogen sulfide
Creator:
Yang, Yonghua, 1949-
Publication Date:
Language:
English
Physical Description:
xiv, 199 leaves : ill., col. photos, ; 29 cm.

Subjects

Subjects / Keywords:
Gas samples ( jstor )
Hydrogen ( jstor )
Kinetics ( jstor )
Loading rate ( jstor )
Moisture content ( jstor )
Oxidation ( jstor )
pH ( jstor )
Sulfates ( jstor )
Sulfides ( jstor )
Sulfur ( jstor )
Dissertations, Academic -- Environmental Engineering Sciences -- UF
Environmental Engineering Sciences thesis Ph. D
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1992.
Bibliography:
Includes bibliographical references (leaves 189-198)
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Yonghua Yang.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
The University of Florida George A. Smathers Libraries respect the intellectual property rights of others and do not claim any copyright interest in this item. This item may be protected by copyright but is made available here under a claim of fair use (17 U.S.C. §107) for non-profit research and educational purposes. Users of this work have responsibility for determining copyright status prior to reusing, publishing or reproducing this item for purposes other than what is allowed by fair use or other copyright exemptions. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder. The Smathers Libraries would like to learn more about this item and invite individuals or organizations to contact the RDS coordinator (ufdissertations@uflib.ufl.edu) with any additional information they can provide.
Resource Identifier:
027843571 ( ALEPH )
26625354 ( OCLC )

Downloads

This item has the following downloads:


Full Text













BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE










BY

YONGHUA YANG






















A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1992






UNIVERSITY OF FLORIDA LIBRARIES























To my father for making this possible
and
In memory of my mother













ACKNOWLEDGEMENTS


I would like to express my sincerest gratitude and appreciation to the following people who made this research possible:

To Dr. E. R. Allen, the doctoral committee chairman, for his foresight in support of this research, his encouragement, guidance and invaluable input during the course of this study and my graduate work.

To Drs. D. A. Lundgren, B. Koopman, K. R. Reddy and D. P. Chynoweth for their interests in this research, helpful suggestions and participation on my graduate committee.

To Dr. P. Urone for his friendship, kindness and valuable suggestions.

To Mr. A. White and the Kanapaha Wastewater Treatment Plant engineers for their assistance in the research on the full scale biofilter system.

To Ms. Yu Wang for her help on compost analysis.

To Ms. S. Jordan for her help in construction and set up of the Lab scale biofilter units.

To Mr. R. Vanderpool for his invaluable friendship, his ideas and his help in all aspects of my work that have made my years at the university much easier and so enjoyable.

To my wife, Li, for her continuing support, encouragement, patience and understanding.


iii















TABLE OF CONTENTS


Pace

ACKNOWLEDGEMENTS ....................................... iii

LIST OF TABLES ....................................... vii

LIST OF FIGURES ...................................... ix

ABSTRACT .............................................. xiii

CHAPTERS

1 INTRODUCTION ................-o.............. ... 1

2 BACKGROUND ....................................... 6

Properties of Hydrogen Sulfide and Regulations .. 6
Physical and Chemical Properties ............. 6
Toxicity of HS ........... . . . ............ 9
Sources of H2S Emissions and Regulations .... 11
Biofiltration as an Air Pollution
Control Technology .......................... 12
History and Development....................... 13
Applications .................................. 15
Theoretical Basis ........................... 16
Biological Oxidation of Hydrogen Sulfide ...... 21

3 PROPERTIES OF COMPOSTS AND THEIR DECOMPOSITION .. 25

Introduction ................................... 25
Selection of Filter Materials .................... 26
Decomposition of Composts under Aerobic
Conditions .................................... 28
Materials and Methods ......................... 32
Results and Discussion ......................... 39
Decomposition of composts ................. 39
Effect of H2S on compost decomposition ..... 45
Conclusions .o............................... 50

4 DETERMINATION OF THE DESIGN AND OPERATIONAL
PARAMETERS FOR BIOFILTER SYSTEMS................. 52

Introduction ..................................... 52
System Design and Construction ................ 52


iv








The Dual Tower System .......................... 52
Portable Tower #3 ............................. 58
Column System #4 .............................. 58
Measurement Methods .............................. 61
Temperature .................................... 61
Pressure Drop ................................. 61
Gas Flow Rate ................................. 61
Sampling Methods ............................... 62
Compost Samples ............................... 62
Gas Samples .... .............................. 62
Water Samples ................................. 65
Compost Analysis Methods ......................... 65
Water Content ................................. 65
pH .. ............................................ 66
Total Carbon and Total Nitrogen ............... 66
Water Soluble Phosphorus (WSP) ................. 66
Acid-extractable Cations ..................... 67
Particle Size Distribution .................... 67
Porosity ...................................... 67
Organic Matter ................................ 67
Particle Density .............................. 68
Bulk density .................................. 68
Sulfur Analysis Methods ......................... 68
Sulfur in Compost ............................. 68
Total Sulfur ................................ 73
Water soluble sulfur ....................... 74
Sulfide sulfur.. ............................ . 74
Sulfate sulfur................................ 74
Elemental sulfur ........................... 75
Pyrite sulfur ............................... 75
Organic sulfur................................ . 76
Sulfur in the Aqueous Phase ................... 76
Sulfate sulfur ............................... 76
Sulfide sulfur............................... 77
Total-S ................................... 77
Sulfur in Waste Gas .......................... 77
Results and Discussion .......................... 79
Pressure Drop................................... 79
Effect of Gas Retention Time on H2S Removal ... 88 Effect of Concentration of H2S on Its Removal . 90 Effect of H2S Loading Rate on Its Removal ..... 92
Effect of Compost Water Content
on H2S Removal .............................. 93
Effect of Compost Acidity on H2S Removal ...... 97 Effect of Temperature on H2S Removal .......... 102 Effect of Sulfate on H2S Removal .............. 106
Effect of Nutrient Addition on H2S Removal .... 108 Kinetics of H2S Oxidation in the Biofilter..... 109
Theoretical considerations................... 109
Determination of the kinetics of H2S
Oxidation in a biofilter ....................114


v









5 BIOFILTER PERFORMANCE AND CHANGES OF COMPOST
PROPERTIES ASSOCIATED WITH LONG TERM OPERATION 125

Overall Performance of the biofilters ......... 125
Accumulation of Sulfur in Compost and Its
Effect on System Performance ................. 137
System Upset and Recovery ..................... 146
Selection of Chemical Solutions ........... 148
Effect of Water-Compost Contact Time on SO4'
Leaching Efficiency ................. ..... 153
Effect of Water to Compost Ratio on SO4
Leaching Efficiency ....................... 153

6 FULL SCALE APPLICATION OF BIOFILTRATION TO
CONTROL H2S EMISSIONS AT A WASTEWATER TREATMENT
PLANT ......................................... 158

Introduction .................................... 158
System Design and Construction ................... 160
Sampling and Analysis Methods ................... 163
Results and Discussion ............................ 166
Conclusions .................................... 182

7 SUMMARY AND CONCLUSIONS ......................... 184

REFERENCES ............................................ 189

BIOGRAPHICAL SKETCH ................................... 199



























vi
















LIST OF TABLES


Table Page

2-1 Physical and chemical properties of H2S .......... 10

2-2 Physiological characteristics of sulfur-oxidizing
bacteria ...................................... 23

3-1 Description of compost used for this study ....... 29 3-2 Properties of selected composts before and after
incubation .....................................******** 33

4-1 Retention times, limits of detection and operating
conditions for the Tracor 250H analyzer ....... 78 4-2 Summary of initial compost properties ........... 80

4-3 Particle size range distribution for selected
composts ...................................... 82

4-4 Effect of gas retention time on H2S removal
efficiency .................................... 89

4-5 Effect of H2S concentration on removal efficiency 91 4-6 Models for the kinetics of H2S oxidation in
biofilter .................................................. 124

5-1 Sulfur fractionation of original compost #17A and
compost at different heights in the filter .... 141 5-2 Effect of washing on compost pH and sulfate
content by DI water, NaOH and NaHCO3
solutions ...................................... 150

5-3 Performance of defective compost before and
after treatment ............................... 152

5-4 Effect of water washing on elimination of sulfate
in filter compost .............................. 157

6-1 Summary of Kanapaha biofilter bed design and
operation parameters ......................... 164


vii









6-2 Summary of periodic Kanapaha biofilter bed compost
analyses during operational period from
5/10/88 to 2/5/91 ............................. 168

6-3 Summary of Kanapaha biofilter influent and
effluent gas sample analyses during
three week start-up period .................... 170

6-4 Gas sampling and analysis for Kanapaha
biofilter bed, 2/5/91 ......................... 178

6-5 Sulfur fractionation of a typical compost sample
in Kanapaha biofilter bed ..................... 181











































viii














LIST OF FIGURES


Figure Page

2-1 Solubility of H2S in water at 1 atm . ............. 7

2-2 Effect of pH on H2S Equilibrium .................. 8

2-3 Biophysical model for the biological filter bed.
The concentration profiles shown in the
biofilm refer to: 1) Reaction limitation,
2) Diffusion limitation ....................... 18

2-4 Steps in the oxidation of different compounds by
thiobacilli. The sulfite oxidase pathway is
thought to account for the majority of sulfide
oxidized ...................................... 24

3-1 Schematic drawing of the experimental arrangement
for the study of compost decomposition ........ 34 3-2 Schematic drawing of the experimental arrangement
for the investigation of the effect of H2S
exposure on compost decomposition ............ 37

3-3 Plot of CO2 evolution from composts during the
122 day incubation ............................ 40

3-4 Decomposition stages and reaction rate
coefficients for the four composts studied .... 44 3-5 Effect of H2S exposure on the rate of compost
decomposition as measured by CO2 respiration .. 46 3-6 Plot of CO2 evolution as a function of square
root of H2S concentration ..................... 48

4-1 Schematic drawing of the dual tower system ....... 53 4-2 Sampling and measurement ports on towers.
a. Tower #1 ............ ...................... 55
b. Tower #2 ................................... 56

4-3 Schematic drawing of Tower #3 .................... 59




ix








4-4 Schematic drawing of column system #4 ............ 60

4-5 Schematic drawing of the gas sampling assembly ... 64

4-6 Photograph of the sulfur distillation assembly ... 70

4-7 Flow chart of the sulfur analysis procedures
for compost ......... ......................... 72

4-8 Pressure drop as a function of particle size
range for different gas velocities ............ 84
4-9 Pressure drop as a function of packing height
for different compost particle size range ..... 86

4-10 Pressure drop as a function of gas velocity for
different types of compost ................... 87

4-11 Determination of maximum H2S elimination
capacity of compost ........................... 94
4-12 Effect of compost water content on H2S removal
efficiency ................... ................ 96

4-13 Time required for dried compost to recover
optimum efficiency ............................ 98
4-14 Effect of compost pH on H2S removal effici ncy.
Condition a: H2S loading rate: 10.5 3g/j-hr Gas loading rate: 15 m /m -hr Condition b: H2S loading rate: 35.4 g m3hr
Gas loading rate: 26.1 m /m2-hr... 100

4-15 Schematic drawing of the experimental arrangement
for investigation of the effect of temperature
on H2S removal efficiency ..................... 103

4-16 Effect of temperature on H2S removal efficiency .. 104

4-17 Effect of sulfate on H2S removal efficiency ...... 107

4-18 Effect of nutrient addition on H2S removal.
Total-S content in compost (mg-S/g) A: 17.5;
B: 33.7; C: 20.2; D: 119.7 .................... 110

4-19 Linear least squares regression analysis for zeroorder kinetics of H S o idItion in biofilter.
Gas loading rate: 224 m /m'-hr, compost #17 ... 115

4-20 Linear least squares regression analysis for firstorder kinetics of H S o idjtion in biofilter.
Gas loading rate: 224 m /m -hr, compost #17 ... 116



x








4-21 Determination of the fractional-order reaction rate
coefficient, kf by linear least squ reh
regression. Gas loading rate: 224 m /m -hr,
compost #17 ................................... 117

4-22 Plot showing the fractional-order kinetics of H2S
oxida iog in biofilter. Gas loading rate:
224 m /m -hr, compost #17 ..................... 118

4-23 Concentration profiles for H2S as a function of
packing height within the biofilter. Gas loading
rate: 224 m /m -hr, compost #17................. 119

5-1 Biofilter control of H2S during long term
operation. a) Tower #1, compost #17A .......... 126
b) Tower #2, compost #17 ........... 127
c) Tower #3, compost #16 ............ 128

5-2 Compost water content profile .................... 132

5-3 pH changes of compost in different sections of the
biofilter with operation time.
a) Tower #1, compost #17A ...................... 134
b) Tower #2, compost #17 ....................... 135

5-4 Total-S distribution profile in biofilter,
Tower #1, after exposure to H2S for 100
days .......................................... 143

5-5 H2S removal efficiencies in different regions of
the biofilter, Tower #2. A: 0-0.2 m,
B: 0.2-0.4 m, C: 0.4-0.6 m, D: 0.6-0.8 m,
E: 0.8 - 1.0 m ................................ 145

5-6 Effect of water-compost contact time on sulfate
leaching efficiency ........................... 154

5-7 Effect of water/compost ratio on sulfate leaching 155

6-1 Schematic diagram of the Kanapaha biofilter bed
system ........................................ 161

6-2 Photograph of the grit chamber at Kanapaha
Wastewater Treatment Plant (top view). The
chamber is covered to collect the
malodorous gas ................................ 162

6-3 Biofilter off-gas sampling system ................ 165

6-4 Photograph of the biofilter system at Kanapaha
Wastewater Treatment Plant .................... 169





xi








6-5 Off-gas sampling locations on the biofilter beds
and concentrations of hydrogen sulfide observed
as a function of biofilter operating time ..... 172

6-6 Concentration changes for hydrogen sulfide in gas
samples contained in Tedlar bags as a function
of container holding time ..................... 173

6-7 Effect of varying purging time for sample
collection chamber prior to sampling on
measured hydrogen sulfide concentrations ...... 175

6-8 Compost samples taken from Kanapaha Wastewater
Treatment Plant biofilter beds (2/5/91).
Left: sample taken from west bed. White color
indicates high sulfur accumulation.
Right: sample from east bed. Low sulfur content
compost, color is close to the original
(dark brown) .................................. 180




































xii














Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy

BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE By

Yonghua Yang

May 1992


Chairman: Dr. Eric R. Allen, Professor Major Department: Environmental Engineering Sciences

A laboratory scale biological filter system for control of hydrogen sulfide (H2S) emissions has been developed and optimum design and operating parameters are evaluated. This biofiltration system uses yard waste compost as filter material, and the performance of the system for controlling waste gas containing H2S is evaluated through long term operation.

Extensive tests have been conducted to determine the effect of various filter bed operating parameters such as pH, temperature, pollutant retention time, pressure drop, water content, etc. on H2S removal efficiencies.

A biofilm model is used to characterize the macrokinetics of the biofiltration process. Models for the kinetics of H2S oxidation are developed that allow one to quantitatively predict the performance of the biofilter.



xiii








Decomposition of composts under aerobic conditions and the effect of H2S concentrations on the decomposition rate are quantitatively determined. The half life time for the composts tested is estimated to be between 3.3 and 6.1 years.

Hydrogen sulfide loading rate and maximum H2S elimination capacity of the filter material are emphasized as important design parameters. The maximum H2S elimination capacity of a typical yard waste compost is determined to be 130 g-S/m3-hr under optimized conditions.

Hydrogen sulfide is oxidized to sulfuric acid in the biofilter system, where biological oxidation plays a major role. Acidification of the biofilter system and accumulation of sulfate in the filter material are determined to be natural features of the oxidation process, where the latter is toxic for biological activities of the microorganisms. Appropriate methods have been developed to effectively mitigate this affect.

System 'upset' is identified as being due to compost dry-out and system overloading. Methods have been identified to provide for recovery of the defective filter material.

Operation of a full scale biofilter system at a wastewater treatment plant has been investigated. Both the laboratory and full scale systems have demonstrated excellent performance over substantial operational periods. Hydrogen sulfide removal efficiencies of 99.9+% have been constantly achieved when the H2S inlet concentrations are varied from 5 to 2650 ppmv.



xiv














CHAPTER 1
INTRODUCTION


Hydrogen sulfide (H2S) is a highly toxic air pollutant which has been identified in the list of 190 air toxic substances in Title III of the 1990 Amendments to the Clean Air Act.

Considerable amounts of H2S are produced in association with industrial processes, such as petroleum refining, rendering, waste water treatment, paper and pulp manufacturing, food processing, and in the treatment of "sour" natural gas and other fuels. Hydrogen sulfide is frequently the main component of most observable odorous emissions.

Hydrogen sulfide is an odorous gas, and its presence at low concentrations is easily perceived and recognized due to its characteristic odor of rotten eggs. Hydrogen sulfide is perceptible to most people at concentrations in excess of 0.5 parts per billion (ppb) in air. Control of H2S emissions is essential to protect public health and welfare as well as to mitigate vegetation and material damage problems.

Numerous processes involving physico-chemical principles have been developed in order to effectively remove hydrogen sulfide from air, waste gases and liquids

1






2


(Bethea et al., 1973; Ferguson, 1975; USEPA, 1985; Lalazary et al., 1986; Walker et al., 1986; Lindstrom, 1990). Processes that have been used to remove H2S from waste gas streams involve either physical treatment or chemical oxidation. Some methods require addition of chemicals, and energy expenditure is usually necessary for physical treatment. Additional environmental problems are encountered with chemical additions, where resulting products and by-products require further treatment and disposal.

Biofiltration can provide for a universal, simple, economicly feasible, and efficient pollutant-destructive control technology for a variety of toxic and hazardous substances in waste gas streams. In recent years biological filters have been developed and described which have the potential to simply and effectively control odors, including H2S emissions (Prokop and Bohn, 1985; Allen et al., 1987a; Eitner and Gethke, 1987; Hartenstein, 1987 ). Deodorization methods based upon the activity of microorganisms are beginning to attract increasing attention in the U.S.

Although the biofiltration technique has been shown to be an efficient, practical and simple gas cleaning technology, which is increasingly being used around the world, the design and operation parameters as well as the microbial processes involved have not yet been very well defined. In particular, little research has been directed at the details of the biofiltration control of H2S. A






3

systematic compilation of data from an operational point of view is also lacking. Most designs are conservatively based on blanket 'rule of thumb' criteria (Forster and Wase, 1987). The performance of biofilter systems, therefore, is not readily predictable and sometimes these systems are not operated under suitable conditions. As a result, the desired odor control efficiency is sometimes not achieved (Allen et al., 1987b). It is essential that more work be done to demonstrate the effectiveness of these systems in order to support further progress in the use of biofiltration as well as to develop better biofilters, based on an understanding of the fundamental physical, chemical and biological processes involved.

A major disadvantage of biofiltration technology is the limited degradation capacity represented by the volume of waste gas treated per unit area of filter material per unit time (m3/m2-hr). This limitation restricts the applicability of biofiltration systems to handling dilute waste gas streams and requires the filter bed to be large in order to handle high volumetric gas flows.

In order to overcome the uncertainties and disadvantages encountered in the full scale application of biofiltration technology, an exhaustive study is necessary for the application of biofiltration technology to control the emissions of air pollutants.

The objectives of the proposed research were to develop a quantitative knowledge of the principle and operation of






4

a microbial biofilter system for removal of H2S from waste gas streams and determine the operating parameters necessary to optimize the performance of such a biofilter system.

The objectives were achieved through the following studies:

1. Evaluation of the properties of filter materials and

their decomposition characteristics under aerobic

conditions.

2. Evaluation of the effects of design and operational

parameters on H2S removal efficiencies on laboratory scale biofilter systems. Variables evaluated included temperature, pH, compost water and sulfate content, H2S elimination capacity, pollutant

retention time, etc..

3. Determination of the predictive relationships for

H2S control efficiencies through chemical kinetic

studies.

4. Evaluation of system performance and determination

of optimum maintenance procedures for biofiltration

control of H2S during long term operation.

5. Evaluation of the field performance of a full scale

biofiltration system for control of H2S emissions at

a local waste water treatment plant.

The research reported here focuses mainly on the utilization, improvement and optimization of a compost biofiltration tower system. Optimization of this system has been directed toward the best achievable control of






5

hydrogen sulfide. This research provides a detailed database on the effects of system variables on H2S control efficiency, which provide for optimization of design and operating conditions.














CHAPTER 2
BACKGROUND


Properties of Hydrogen Sulfide and Requlations Physical and Chemical Properties


Hydrogen sulfide is a colorless gas that has a foul rotten egg odor and is slightly heavier than air. Hydrogen sulfide is moderately soluble in water. The solubility of H2S decreases with increasing temperatures. Figure 2-1 shows the solubility of H2S as a function of temperature.

Dissolved H2S dissociates in accordance with the following reversible ionization reactions:


H2S -2 HS- + H+ (2-1) HS- am S2- + H+ (2-2)


The distribution of the above species as a function of pH is shown in Figure 2-2. It is apparent from Figure 2-2 that the concentration of HS- species is insignificant when pH values are less than 6. The latter condition is normal in a biofilter system for control of H2S. S2-, on the other hand, may not occur at all.

Hydrogen sulfide can serve as a reducing agent, reacting with sulfuric acid (H2S04) to form sulfur dioxide (SO2) and elemental sulfur (SO) (Greyson, 1990):



6





7









7

6.5

6

5.5



p 4.5

4

3.5

3

2.5

0 10 20 30 40 Temperature (oC)









Figure 2-1. Solubility of H2S in water at 1 atm.
Data adopted from Piscarcyzyk, 1982.

























100


80- S' HS

60a. 4020 - H2S



0
5 6 7 8 9 10 11 pH



Figure 2-2. Effect of pH on H2S equilibrium.
Source: Sawyer, 1967.






9


H2S + H2SO4 = SO2 + So + 2H20 (2-3)


Hydrogen sulfide also burns in air to form sulfur dioxide and water:


2H2S + 302 = 2SO2 + 2H20 (2-4)


Table 2-1 summarizes the physical and chemical properties and the odor threshould of H2S.


Toxicity of H2S


Hydrogen sulfide is almost as toxic as hydrogen cyanide (HCN), which is used in prison gas chambers (Parker, 1977). Human exposure to small amounts of H2S in air can cause headaches, nausea, and eye irritation, and higher concentrations can cause paralysis of the respiratory system, which results in fainting and possible death. Concentrations of the gas approaching 0.2 percent (2000 ppmv) are fatal to humans after exposure for a few minutes (NRC, 1979).

Hydrogen sulfide has a characteristic rotten egg smell at low concentrations. But as levels of H2S increase, a person's ability to sense dangerous concentrations by smell is quickly lost. If the concentration is high enough, unconsciousness will come suddenly, followed by death if there is not a prompt rescue.

The Occupational Safety and Health Administration (OSHA) has established limits for work place exposure to H2S






10








Table 2-1. Physical and Chemical Properties of H2Sa.


Molecular Weight 34.08 Boiling Point, oC -60.2 Melting Point, oC -83.8 to -85.5 Vapor Pressure, -0.4oC 10 atm
250C 20 atm Specific Gravity (Relative to Air) 1.192 Auto Ignition Temperature, oC 250 Explosive Range in Air, % 4.5 to 45.5 Odor Threshold, ppbv 0.47

a Source: USEPA, 1985.








at 20 ppm (15-minute exposure) for an acceptable ceiling concentration and 50 ppm for a maximum exposure during an 8hour work shift if no other measurable exposure occurs. The National Institutes of Occupational Safety and Health (NIOSH) established an H2S exposure level at 10 ppm (10 minutes) as a maximum permissible limit (once per 8-hours shift), with continuous monitoring required where H2S concentrations could equal or exceed 50 ppm or greater (NIOSH, 1979).

Hydrogen sulfide is an explosive gas. The lower and upper explosive limit are 4.5 and 45 percent in air by volume, respectively.

Hydrogen sulfide can attack materials and cause discoloration and tarnishing. Materials commonly affected are paint, copper, zinc and silver (Painter, 1974).


Sources of H2S Emissions and Regulations


Natural emissions are mainly caused by biological decay of protein materials. The natural global rate of emission is estimated to be about 84 Tg/year (Urone, 1986).

Anthropogenic emission sources include petroleum refining, natural gas plants, sewage treatment facilities, coke ovens, Kraft paper pulp plants, and waste disposal sites. There are no federal U.S. emission standards for H2S at present, nor are there federal ambient air quality standards for this gas, but a number of states have






12


established independent standards for H2S emissions. These states, which include California and New Mexico (10 ppm), and Ohio and Michigan (1670 ppm). California, Kentucky, Minnesota, Montana, New Mexico, New York, North Dakota and Pennsylvania also have air quality standards for H2S. The standards vary from 0.003 ppm for New Mexico to 0.1 ppm for Pennsylvania, whereas and most of the other states specify a standard of 0.03 ppm (Urone, 1986).

Since H2S is a highly toxic air pollutant, H2S has been identified by the USEPA as one of 190 air toxic compounds in Title III of the 1990 Amendments to the Clean Air Act. In view of the wide spread exposure to this pollutant, emission and air quality standards for H2S are going to be set in the near future by EPA.


Biofiltration as a Air Pollution Control Technolory


Biological degradation is widely used for treatment of liquid and, to a lesser extent, solid wastes, but has received little attention as a means of controlling emissions of industrial gaseous wastes. Biofiltration is a relatively new technology for control of air pollutants, in which the air contaminants from off-gas streams are biologically removed in a solid biological reactor. While it is a well established air pollution control technology in European countries, biofiltration as an air pollution control technology has received little attention and application in the United States. Few environmental






13

professionals in this country appear to be aware of the 'biofiltration' process and its applications.

Although there are some applications of biofiltration in the U.S. and some technical papers have been published in the English language, most of the research and development work on biofiltration has been conducted in Europe and the majority of the recent research data have been published in the German language. Excellent reviews of previous biofiltration work have been published by Hartenstein (1987), Leson and Winer (1991), Ergas et al. (1991), and Dharmavaram (1991).


History and Development


The first deodorization method based upon the use of a soil bed in the U.S. was developed and patented by Pomeroy in 1957. Later, Pomeroy (1982) described the deodorization of waste gases emitted from sewer lines by a soil bed system used in Los Angeles in 1957. The microbiological degradation of sulfur-containing gases in the filter bed was observed to be effective in these studies.

Other early applications of biological treatment of odorous gases include a soil bed system built in Nurnberg, West Germany, in 1959 and biofilters built in Geneva, Switzerland, and Mercer Island, Washington, to remove odors from wastewater treatment and compost manufacturing, respectively, in the mid-1960s (Bohn and Bohn, 1987).






14


Additional studies in the US have been carried out by Carlson and Leiser (1966), Bohn and Miyamoto (1973), Bohn (1975, 1976, 1977, 1989), Pomeroy (1982), Prokop and Bohn (1985), Hartenstein and Allen (1986), Bohn and Bohn (1986, 1988), Hartenstein (1987), and Allen et al. (1987a, b, c; 1989).

In spite of the work mentioned above, most of the research and development in biofiltration technology has been carried out in Europe, especially in West Germany and Holland. In the latter countries the principle of biofiltration has been applied to a wide variety of environmental problems. Among the many researchers in the field, Ottengraf and coworkers in Chemical Engineering Department, The Eindhoven University of Technology, Holland, have contributed most of the theoretical research in biofiltration in a series of papers which have been published in the English language (Ottengraf, 1977; Ottengraf and Van Den Oever, 1983; Ottengraf et al., 1984; Ottengraf, 1986; Ottengraf et al. 1986; Ottengraf, 1987). Also, Eitner in West Germany, has made significant contributions to the research and development of biofiltration ( Eitner and Gethke, 1987), although most of his publications are in the German language (see Hartenstein, 1987; Leson and Winer, 1991).

The practice and application of biofiltration has also been reported in other countries such as Japan (Terasawa et al., 1986), New Zealand (Rands et al., 1981) and Canada






15


(Rotman, 1991a).

At present, biofiltration is considered to be a stateof-the art technology for odor removal in West Germany, and it has been estimated that 40% of deodorization facilities at wastewater treatment plants are biofilters (Frechen and Kettern, 1987).


Applications


The first systematic study of odor control using biofiltration in this country was conducted by Carlson and Leiser (1966). They studied the removal efficiencies of sewage odors using a laboratory scale soil bed. Using hydrogen sulfide as the test gas, a 99% removal efficiency was achieved, and biodegradation was reported to be the primary removal mechanism.

Prokop and Bohn (1985) reported that a soil bed system for control of rendering plant odors had been in operation since September, 1983. The soil bed treats 1100 m3/h of cooker non-condensable waste gases using a bed surface area of 420 m2. In this work an odor removal efficiency of 99.9% was obtained.

Rands et al. (1981) reported that a full-scale compost filter system was constructed in 1978 at Moerewa, New Zealand, to treat odors from a rendering plant. The system was designed to treat 900 m3/h of air containing hydrogen sulfide concentrations up to 1000 parts per million (ppm) by volume. An average H2S removal efficiency of 99.9% was






16


observed.

Allen et al. (1987a, b, c) investigated a compost based tower biofilter system used for odor control in a wastewater treatment plant. The odor-causing compounds identified were reduced sulfur compounds such as H2S, methyl mercaptan, dimethyl sulfide and dimethyl disulfide as well as terpene hydrocarbons. Removal efficiency for total reduced sulfur compounds (TRS) was 65 to 72%. The poor performance of this system was determined to be the short residence time in the system, poor gas distribution, and improper maintenance.

Biofiltration control of volatile organic compounds (VOCs) has been reported by Ottengraf (1986), Kampbell et al. (1987), Bohn (1989), Paul and Castelijn (1987), and Hack and Habets (1987).

In recent years, increasing numbers of biological filters are being used around the world for odor control. It has been estimated that more than 500 biofilters are currently operating in Europe (Leson and Winer, 1991). Excellent summaries of recent applications have been provided by Bohn and Bohn (1987) and Rotman (1991b).


Theoretical Basis


The concept of a biological-film or 'biofilm' is frequently used to describe degradation processes in aqueous systems ( Williamson, 1973; Williamson and McCarty, 1976a, b; Jennings et al., 1976; Rittmann and McCarty, 1978). This concept has been adopted and improved to describe the






17

biofiltration processes (Ottengraf, 1986; Hartenstein, 1987; Paul and Castelijn, 1987; Van Lith, 1989). In particular, Ottengraf and coworkers have carried out systematic studies delineating the overall process and have presented sufficient experimental data to support the proposed model.

In biofiltration, evenly distributed waste gases are forced through a biologically active material, such as soil, peat or compost. Many of the pores of the filter material particles are filled with water. Microorganisms are attached to the particle surfaces to form a layer of film. This wet, biologically active layer surrounding the particles is called a biofilm. The biophysical model proposed by Ottengraf for the biofilm is shown in Figure 2-3. The mechanism of the biological process is derived from a combination of physical, chemical and biological processes that occur in the filter material and is related to two processes in particular; sorption and regeneration. As waste gases pass through the countless narrow pores of the filter material, air contaminants as well as oxygen will adsorb on the surfaces of the pores and dissolve in the liquid phase of the wet biofilm. The absorbed and adsorbed gases are quickly degraded by the biofilter's enormous microbial population. In this way a concentration gradient is created in the biofilter, which maintains a continuous mass flow of the component from the gas to the wet biofilm.

Activity of the biofilter depends mainly on the population of the microorganisms. Soil biofilters can






18











Gas Phase




Cg1


Cg2










g2 = Low concentrations of air pollutants.














Figure 2-3. Biophysical model for the biological
filter bed. The concentration profiles
shown in the biofilm refer to: 1) Reaction limitation, 2) Diffusion
limitation. (Source: Ottengraf, 1986,
p. 436).
..... ....
..... ....
..... ....
CIP ........
..... .... .... ....




























p. 436).






19

contain 1 billion bacteria, 10 million actinomycetes and 10,000 fungi per gram of soil (Bohn and Bohn, 1987). The role of these microorganisms is to oxidize combined carbon, nitrogen and sulfur to carbon dioxide, nitrogen and sulfate, respectively, before the compounds leave the bed. The air contaminants are, therefore, effectively removed from the waste gas streams.

For good engineering design and environmental decision making, it is essential to understand the mechanisms involved and to reliably predict the kinetics of the biological reactions taking place in these biofilter systems. Many general kinetic models have been developed to predict the behavior of bioreactions in a biological film, none of these models, however, is specific enough to explain the biodegradation of hydrogen sulfide in a biofilter system.

Jennings et al. (1976) developed a mathematical model to predict the percentage removal of a pure, non-adsorbable, biodegradable substrate in a submerged biological filter using the non-linear Monod expression for the substrate utilization rate. In their model, the authors start from a biological slime layer coating a spherical particle. The slime layer is in turn surrounded by a liquid boundary layer. They concluded that even at relatively high values of influent substrate concentrations, the biological removal of a single substrate follows first order kinetics.






20


Another model developed by Rittmann and McCarty (1978) is a variable-order model of bacterial-film kinetics which incorporates liquid-layer mass transport, biofilm molecular diffusion and Monod kinetics. These investigators concluded that at low substrate concentrations, the reaction follows first order kinetics, whereas at high concentrations the reaction follows one-half order kinetics.

Based on their biophysical model (Figure 2-3), Ottengraf and Van Den Oever (1983) have developed a mathematical model to describe the kinetics of organic compound removal from waste gases for a biofilter system. The model was developed and tested using a soil bed for the removal of toluene, butylacetate, ethylacetate and butanol. From their experimental results, they concluded that all the carbon sources investigated were eliminated according to a zero order reaction, even at very low concentrations of the substrates.

Kampbell et al. (1987) investigated the biodegradation of propane, isobutane and n-butane by soil biofilter beds. They suggested that at low concentrations the rate of biodegradation was proportional to the concentration of the organic compounds (first order reaction), and at higher concentrations the rate becomes independent of the organic compound concentration (zero order reaction). The degradation kinetics appeared to follow a hyperbolic function:





21



1/V = (1/Vmax) +( Km/Vmax)(l/S) (2-5) where:

V = the biodegradation rate, mg hydrocarbon/kg soil-h
Vmax = the maximum possible biodegradation rate,
mg hydrocarbon/kg soil-h
Km = an empirical constant, half saturation value, ppm S = the concentration of organic compound in air, ppm

The values of Km and Vmax can be obtained from a Lineweaver-Burk plot of 1/V against 1/S.


BioloQical Oxidation of Hydrogen Sulfide


Hydrogen sulfide may be utilized by microorganisms in three different ways: assimilation, mineralization and sulfur oxidation (Atlas and Bartha, 1981; Grant and Long, 1981). However, the rates of uptake of hydrogen sulfide based on the assimilation processes are far too low to achieve reasonably high removal efficiencies from a highly loaded waste gas stream. The most important and efficient way for microorganisms to utilize hydrogen sulfide is by the oxidation of sulfur to gain energy. In this process, relatively large quantities of sulfur are oxidized in order for the microbes to receive sufficient energy. The microorganisms living in the biofilter materials are usually mixed cultures. Various groups of microorganisms, therefore, are involved in the energy conversion process under aerobic or anaerobic conditions. However, the






22

colorless sulfur bacteria are believed to play the major role and their ability to oxidize reduced inorganic sulfur compounds has been clearly established (Roy and Trudinger, 1970; Kuenen, 1975; Brock and Madigan, 1988).

The oxidation of inorganic sulfur compounds is carried out by a spectrum of sulfur-oxidizng organisms which include 1) obligately chemolithotrophic organisms, 2) mixotrophs, 3) chemolithotrophic heterotrophs, 4) heterotrophs which do not gain energy from the oxidation of sulfur compounds but benefit in other ways from this reaction, and 5) heterotrophs which do not benefit from the oxidation of sulfur compounds. Physiological characteristics of some sulfur-oxidizing bacteria are summarized in Table 2-2.

Options for microbial metabolism of hydrogen sulfide must employ one or more of the following metabolic pathways: 1) aerobic oxidation, 2) anaerobic oxidation, and 3) photosynthetic dissimilation. Biofiltration of waste gases is a process utilizing aerobic conditions in most cases. In aerobic oxidation, sulfur-oxidizing bacteria oxidize H2S to elemental sulfur or higher oxidation states using oxygen

(02) as an electron acceptor. The biological steps in the oxidation of various sulfur compounds are summarized in Figure 2-4.






23


Table 2-2. Physiological characteristics of sulfuroxidizing bacteria.



Lithotrophic Electron pH Range Donor for Growth Thiobacillus Species Growing Poorly in
Organic Media:


1. T. thioparus H2S, sulfide, So, 203 2- 6-8 2. T. denitrificans H2S, SO, S2032- 6-8 3. T. neapolitanus SO, 2032- 5-8 4. T. thiooxidans So 2-5

5. T. ferrooxidans SO, sulfides, Fe2+ 1.5-4 Thiobacillus Species Growing Well in
Organic Media:


1. T. novellus S2032- 6-8 2. T. intermedius S2032- 3-7 Filamentous Sulfur
lithotrophs


Begqiatoa H2S, S2032- 6-8 Thiothrix H2S 6-8 Other Genera


Thiomicrospira S2032-, H2S 6-8 Thermothrix H2S, S2032-, SO3- 6.5-7.5 Sulfolobusa H2S, So 1-4 a Archaebacterium.
Source: Brock and Madigan, 1988.





24



Sulfide
S2- Elemental Sulfur So

Cell-Bound R - S -- S S - S 02- Thiosulfate Sulfur Complex 2 3

Se--- --w Electro Transport System
Sulfite SO2AMP
ADP
Electron Transport A Phosphorylation ATP

ADP
Adenosine
Sulfite Oxidase Phosphosulfate

2e A ATP(APS)


Substrate Level uvPhosphorylation ADP


SO 2- Sulfate SO24 4







Figure 2-4. Steps in the oxidation of different
compounds by thiobacilli. The sulfite
oxidase pathway is thought to account for the majority of sulfide oxidized.
(Source: Brock and Madigan, 1989, p.
704).














CHAPTER 3
FILTER MATERIALS AND THEIR DECOMPOSITION
UNDER AEROBIC CONDITIONS

Introduction

Biofiltration systems or biofilters employ physical, chemical and biological processes such as adsorption, absorption and microbial digestion and oxidative degradation to remove air pollutants from waste gas streams. Microbial degradation and oxidation of the pollutants, however, appear to be the primary removal mechanisms within a biofilter. In the biodegradation process, pollutants are consumed by the microorganisms, providing an energy source or essential nutrients and are converted usually to, less harmful compounds. The filter materials used, on the other hand, must provide the proper environment for microbial growth and contain materials on which the microbes can feed to ensure that the microbial population can develop and survive.

The effectiveness of a biofilter material depends on its physical, chemical and biological characteristics. The lifetime of a biofilter material mainly depends on its rate of carbon (C) and nitrogen (N) mineralization. When available C and N in the filter material are no longer sufficient to support the microbial population in the system, then the material is no longer suitable as a


25






26
biofilter. The C or N deficient filter material must be replaced by freshly prepared material and the discarded filter material has to be properly disposed of with due caution for environmental impact. One of the most common options is land application. Determination of the decomposition characteristics of the filter material is, therefore, necessary for usage of the biofilter and eventual land disposal applications.

Considerations necessary for selection of appropriate filter materials and the decomposition of such materials under aerobic conditions are discussed in this section.


Selection of Filter Materials

Effective removal of air contaminants using a biofilter relies on the properties of the filter material, especially the nature and activity of the biomass. The filter material provides the necessary environment for microorganisms to survive, generate, function and allows the entire sequence of biofiltration processes to be carried out. The filter material serves as 1) support material for the microbes, 2) supplemental or alternative nutrient source, 3) moisture storage reservoir, 4) surface area for sorption of air pollutants and interaction between the pollutants and the microorganisms, and 5) a buffer volume for variations in water content and gas conditions during operation (Eitner, 1989). In general, the following factors need to be considered when choosing a suitable filter






27

medium:

1). Density: Too dense material may contain a large
fraction of inorganic materials such as stone and
sand which are unsuitable as carbon and energy
sources for microbial growth.

2). Structure: Structure of the medium will affect
the uniformity of the filter load. Too large sized materials should be avoided because the
surface-to-volume ratio will be reduced.

3). Particle Size Distribution: Too small particles
affect the pressure drop by compacting and
restricting the gas flow.

4). Pore Volume (void fraction): This property
determines the total surface area available for
reaction, also it will affect pressure drop.

5). Organic Matter Content: The organic matter
controls the microbial population and the useful
service life of the filter media.

6). pH Value: pH will affect the nature and level of
the microbial population and activity.

7). Water Retention Capacity: This property will
determine the consistency in liquid water content
of the filter material, and

8). Economics: Reasonable Capital and operating
expenditures.

All of these requirements can be met by selecting suitable filter materials. Many kinds of filter materials have been used in biofiltration applications. Examples include field soils, compost, peat, bush, clay, volcanic ash, sand, bark and a combination of such materials (Rands et al., 1981; Prokop and Bohn, 1985; Terasawa et al., 1986 Frechen and Kettern, 1987). The performance of these materials, however, can be very different due to the diversity of their physical and chemical properties. Compost has been considered to be the best choice for filter






28

materials and has been involved in most applications (Don, 1985; Eitner, 1989), since it provides favorable conditions for supporting microbial populations as well as having superior physical and chemical properties.

The properties of individual composts depend on the materials from which they are derived and the composition of the final product. The filter materials used in this research were mainly yard waste compost and sewage sludge compost or a combination of both. These composts were obtained from different sources and used for different purposes. A general description of the types of composts and their sources are summarized in Table 3-1. The physical and chemical analyses data for the composts listed in Table 3-1 are presented in the corresponding chapters where the use of specific composts is discussed.


Decomposition of Composts under Aerobic Conditions


A number of investigations have been carried out to study the decomposition of anaerobically digested sewage sludges in soils (Miller, 1974; Tester et al., 1977; Terry et al., 1979b; Sweeney and Graetz, 1988; Gale, 1988). Decomposition of fresh and anaerobically digested plant biomass in soil is also reported by Moorhead et al. (1987). Only limited information, however, is available concerning the decomposition of compost. Tester et al. (1977, 1979) stated that the decomposition of compost in soil is not only related to the physical and chemical properties of the






29



Table 3-1. Description of composts used for this study.


Compost Source Description
ID#


1 Pompano Beacha Fort Lauderdale sewage sludge compost.
Not completely composted. Seven months old when first used (used for decomposition study).

2 Pompano Beach Two parts yard waste and one part stable cleaning sewage sludge mixed and composted. Seven months old when first used (used for decomposition study).

3 Pompano Beach Yard trash compost. 13 months old when first used (used for decomposition study).

6 Pompano Beach 25% by volume of sewage sludge compost
and 75% of yard trash mixed and composted about 19 months old when first used (used for decomposition study).

12 Kanapahab Pompano Beach compost similar to Compost #6 mixed with tree bark, yard waste and sewage sludge; lime was used to adjust pH before use. Used in Kanapaha filter bed from 11/20/88. Compost obtained from the filter bed in 5/16/90.

13 Kanapaha Same as #12, compost obtained and used in Tower #1 from 12/20/90.

13-1 Kanapaha Same as #12, compost obtained in 2/5/91.

13-2 Kanapaha Same as #12, compost obtained in 3/20/91.

14 Kanapaha Yard trash, grass and sewage sludge were mixed and composted; lime was used to adjust pH; about 2.5 years old when obtained and used in Tower #2, 12/20/90.






30





Table 3-1 -- Continued.



Compost Source Description
ID#



16 WRRc Yard trash compost, 3.5 months old when first used in Tower #3 from 1/27/91.

17 WRR 1:1 by volume of yard trash and grass composted; about 3.5 months old when first used in Tower #2 from 1/27/91.

17A WRR Compost #17 mixed with 2% lime (CaCO3), by dry weight of compost. Used in Tower #1 from 1/27/91.

a Broward County Streets and Highways Division, 1600 NW 30th Avenue, Pompano Beach, FL, 33069. b Kanapaha Wastewater Treatment Plant, Gainesville, FL 32602. c Wood Resource Recovery, Inc., Gainesville, FL.






31

compost but also is a function of the particle size of the compost. The decomposition was observed to be directly related to the carbon content in the compost.

Decomposition is affected by a number of environmental conditions, for instance, pH, moisture content, and the presence or absence of foreign chemicals (Miller and Johnson, 1964; Terry et al., 1979a, b; Delaune et al., 1981). In the application of biofiltration to control H2S emissions, the compost filter material is subjected to conditions that are quite different to that for land applications of compost. In the former case, the compost is exposed to a gas stream which may contain a variety of chemicals, especially H2S, at various concentrations. The presence of xenobiotics in the gas streams and filter materials could change the population and composition of the microorganisms in the compost or significantly affect their metabolic processes. As a result, the decomposition rate of the compost can be altered. Unfortunately, little information can be found in the literature related to this topic.

The objectives of this study were (i) to evaluate the decomposition of four types of compost by determining the CO2 evolution, and (ii) to investigate the effect of H2S at various concentrations on compost decomposition. Such information is valuable for biofilter design and for justifying land disposal applications of the compost after use as a biofilter medium.






32


Materials and Methods

Four types of compost samples were investigated for their decomposition characteristics during the course of this study. All of the compost samples were obtained from Broward County Streets and Highways Nursery Division, Pompano Beach, Florida. The composts were stored in sealed plastic bags at room temperature (23�2 oC) before use. A brief description of the composts used in this study is presented in Table 3-1 ( Composts #1, #2, #3 and #6). The compost samples were analyzed for their physical and chemical properties at the beginning and the end of the investigation. The results of these studies are presented in Table 3-2. Each cured compost was passed through a 10 mm screen to remove larger materials. The compost samples are then placed in 225-mL wide mouth bottles directly for incubation. The experimental arrangement for the decomposition study is shown in Figure 3-1. Four types of compost and one blank, each with three duplicates, were investigated. Compressed air from the laboratory house air supply is controlled to about 4 psig by a regulator. The air stream is passed through a scrubber system consisting of 4N NaOH to remove CO2 and distilled water to saturate the air stream. A dead volume is placed before the 4N NaOH scrubber as a safety precaution in the event that the air system causes a backpressure forcing scrubbing solution against the air system. An empty impinger is placed after the water scrubber to separate larger water droplets from









Table 3-2. Properties of selected composts before and after incubation.


Compost #1 Compost #2 Compost #3 Compost #6
Property Before After Before After Before After Before After
use use use use use use use use


pH 8.67 6.79 8.13 8.44 9.22 8.72 7.26 7.55 Water (Wt%) 60.3 59.5 59.7 57.4 61.0 63.0 60.5 62.0 LOI (Wt%) 79.8 73.7 69.1 64.4 73.8 74.4 65.2 64.6 Water-P (mg/kg) 144 96.9 95.2 97.9 53.3 49.3 231 218 Total-C (Wt%) 36.9 39.6 36.8 35.3 39.5 40.9 39.5 36.2 Total-N (Wt%) 3.00 3.57 2.84 2.78 2.26 2.30 3.45 3.73 C/N 12.3 11.1 12.9 12.7 17.5 17.8 11.5 9.71

Metals (mg/kg)
Ca 37400 40300 57500 60000 66500 65000 47700 49800 Mg 4120 4470 5250 5300 3840 4250 5400 5950 Zn 492 557 651 638 81.0 94.5 897 964 Cu 201 194 125 131 8.50 9.50 114 113 Mn 30.0 36.0 66.5 70.0 23.5 25.0 48.5 52.0 Fe 8820 9760 5220 5130 512 541 6740 7260








Needle
Valve Manifold


Flow
House Regulator
Air
Supply







Dead 4N DI DI Impinger 50g 2X25mi
,a Volume NaOH Water Water Compost 0.5N NaOH






Figure 3-1. Schematic drawing of the experimental arrangement for the
study of compost decomposition.






35

the air stream. Two water scrubbers are used in series to ensure that the air stream is completely free of alkali and to resaturate the air with water vapor in order to keep the water content of the composts constant.

The resaturated, C02-free air stream is then forced to the manifold where it is split into 15 streams. Each stream goes into one incubation-absorption unit. Fifty grams of compost sample is put in each incubation bottle. The CO2 evolved from each of the compost samples is collected in two 25-mL, 0.5N NaOH collectors in series. The total air flow rate is controlled by a needle valve located in front of the manifold. Syringe needles are used as flow regulators to equalize the air flow through the 15 incubation units. The air flow rate through each unit is adjusted to about 15�2 mL/min. The incubation system is continuously operated at constant temperature (23�20C).

After flushing the residual air from the incubation bottles, the outlet tube of each bottle is attached to the CO2 collectors. CO2 collectors are replaced with fresh solutions periodically during the incubation period. The system is leak checked before the incubation. Evolved CO2 is efficiently trapped by two absorption collectors in series. Tests have shown that the first tube absorbed more than 95% of the total CO02 evolved.

CO02 evolution is measured as described by Stotzky (1965) with minor modifications. After CO02 absorption, the solutions in the two collectors of each unit are mixed and






36


titrated with standard IN HC1. The CO2 samples collected from each of the control bottles are concomitantly titrated.

The CO2 evolved for individual samples is calculated as follows (Stotzky, 1965):


CO2 = (B - V)NE (mg) (3-1) where:

B = volume of HC1 used to titrate the NaOH in the
controls to the end point, (mL);
V = volume of HCI used to titrate the NaOH
remaining in the CO collectors after
treatment to the end point, (mL);
N = normality of the HC1, (meq/mL);
E = equivalent weight, (mg/meq), for CO2, E = 22
(mg/meq).


To investigate the effect of H2S concentration on compost decomposition, one hundred grams of Compost #6 was used as the test material. The experimental arrangement for this test is similar to that for the compost decomposition test with some minor modifications (Figure 3-2). Room air is forced through a scrubber containing 4N NaOH to absorb CO2 from the air stream. The CO2 free air is then saturated by bubbling through DI water. Pure H2S is then mixed with the pretreated air stream to obtain the test gas mixture with the desired H2S concentration. The treated gas stream is vented through the manifold, where it is split into four sub-streams: one of these sub-streams is vented to a control column (empty), and the other three streams to duplicates of three compost columns. The gas is forced vertically through the compost from bottom to top at a flow rate of 30 mL/min.








Manifold


Row Meter ator
House
Air
Supply






S Dead 4N DI H2S Gas lOg 200+25ml 2x25ml Dead � Volume NaOH Water Mixer Compost 1N ZnAc 0.5N NaOH Volume








Figure 3-2. Schematic drawing of the experimental arrangement for the investigation of the effect of H2S exposure on compost decomposition.





38


A portion of the hydrogen sulfide is adsorbed and/or oxidized by the compost and the remaining H2S in the effluent gas is absorbed by two H2S scrubbers in series which contain 200 mL and 25 mL IN zinc acetate (ZnAc) solution, respectively. The absorbing reaction used by the H2S collectors is:


ZnAc + H2S - ZnS4 + HAc
t1
H+ + Ac- (3-2)


Total flow in the system is measured by pre-calibrated rotameters. H2S concentrations in the inlet gas to the compost column are controlled by adjusting the flow rates of mixing for the C02-free air and the pure H2S gas. Gas samples from the influent gas stream are taken periodically by gas-tight syringes, diluted with prepurified nitrogen

(N2) and analyzed for H2S content by a Tracor 250H analyzer (See Chapter 4 for details). The effluent gas from the filter columns is first passed through two scrubbers in series containing 1N ZnAc to absorb any H2S remaining in the gas stream. Residual CO2 in the effluent gas stream is subsequently absorbed by 0.5N NaOH solution and titrated as described previously.

Each compost sample is incubated at a desired H2S concentration level for 24 hours. After the incubation period the compost as well as the absorption solutions are replaced by fresh compost samples and absorbing solutions for operation at the next H2S concentration level. The






39


system was previously tested to obtain absorption efficiencies of H2S and CO02 in the ZnAc traps. The results showed that the H2S absorption efficiency was greater than 99% and the CO02 absorption was less than 2% for the ZnAc solutions used.


Results and Discussion

Decomposition of Composts


The decomposed C evolved as CO02 from the four composts studied during the 122 day incubation period is shown in Figure 3-3. The decomposition patterns of composts #2, #3 and #6 are somewhat similar. Decomposition of compost is initially rapid, from 40 to 52% of the total CO02 produced in the 122 days is evolved in the first 42 days of incubation. A total of 9.2, 5.7, 6.1, and 4.4% of the original C was decomposed and released as CO02 for composts #1, #2, #3, and #6, respectively during the total 122 days of incubation.

It appears that decomposition rates of the composts are inversely proportional to their age, in other words, the older the compost, the slower the decomposition. All except compost #1, showed decomposition rates which were similar. Compost #1, however, was not completely composted when used. Also the organic matter content of this compost is higher than that for the others tested. Initial and delayed higher decomposition rates for compost #1 suggest a two stage incubation involving an initial 'conditioning' step followed by a 'conditioned' decomposition. The compost






40










100

90
080

O 70
0
60
0
9 50

40

30

& 20

10

01 "
0 20 40 60 80 100 120 Time (day) Compost #1 + Compost #2 o Compost #3 A Compost #6






Figure 3-3. Plot of CO2 evolution from composts
during the 122 day incubation.






41

decomposition rates measured here are much slower than those found for soils. Tester et al. (1977) reported that approximately 16% of the compost C was evolved as CO2 during 54 days of incubation, when 2 to 6% fresh sewage sludge compost was incubated with soils. Miller (1974) reported that 20% of added organic carbon is evolved as CO2 for a 6 month incubation period under similar conditions. In another investigation carried out by Moorhead et al. (1987) it was observed that about 39 and 19% of the total-C for fresh and digested low-N plant biomass, and 50 and 23% of fresh and digested high-N plant biomass are released as CO2 during 90 days decomposition when these biomasses are added to soils. Fresh plant biomass evolves as much as twice the organic-C as CO2 when compared to corresponding digested biomass sludges. These results of other researchers suggest that the decomposition rate of organic matter strongly depends on the source and the properties of the available organic matter. Miller (1974), Sommers et al. (1976) and Terry et al. (1979a, b) have concluded that sludge composition and incubation conditions, rather than soil properties control sludge decomposition.

Reddy et al. (1980) have shown that decomposition of organic carbon depends on the nature and constitution of the wastes. Low molecular weight (simple) compounds can be more easily degraded by microorganisms than more complex organic compounds. Organic-C components in decreasing order of biodegradability are: (i) readily oxidizable soluble





42


organic-C, (ii) proteins, (iii) hemicellulose, (iv) cellulose, and (v) lignin. In the examples mentioned above, fresh plant biomass releases much more CO2 (especially in the early stages of the incubation) than the digested ones because it contains much more easily decomposable organic-C. In this study, all the composts used were well aged or completely composted. Most of the easily decomposable organic-C such as soluble organic-C, starch and proteins have been decomposed during the composting process. The main organic-C species remaining in the composts studied are the more oxidation resistant residues of the original organic matter (Biddlestone et al., 1987). Also, the four composts studied are either yard waste compost or mixtures of sewage sludge with yard wastes such as wood chips, leaves and tree trimmings, etc.. A high content of cellulose and lignin can be expected in these materials. This feature may explain why the decomposition rates for the composts studied here are relatively low.

Decomposition of a complex substrate C is usually described by a multistage first-order decomposition sequence (Reddy et al., 1980; Gilmour et al., 1985). The mathematical rate equations can be written as follows:


-dCi/dt = kiCi (3-3) where i refers to a particular stage of decomposition.






43

The integrated form of equation 3-3 becomes:


Cti = Ciexp(-kit) (3-4) where:

Ci = organic-C present at the beginning of a
decomposition stage.
Cti = organic-C present at the end of a decomposition
stage at time = t, and
ki = the first-order reaction rate coefficient.

Decomposition stages and the corresponding reaction rate coefficients for the four composts are presented graphically in Figure 3-4. The decomposition of compost #3 is described in one stage and the decompositions of compost #2 and #6 are best described in two stages. It can be seen that the reaction rate coefficient values of ki and k2 for these two composts are very similar. This similarity indicates that these two composts have similar organic C composition.

The behavior of Compost #1 is markedly different from those of the other composts. During the first 3 days of incubation, CO2 evolution is rapid followed by a lag period, lasting for the following 40 days. A second period of high decomposition rate was observed between 42 and 70 days. During the remaining period of incubation (after 70 days)', the CO2 evolution rate for this compost is similar to those for the other composts. The decomposition rate for compost #1 may be described as a 3 stage series of first-order reactions.

Within overal experimental error, reaction rate coefficients for the final stage of decomposition for the






44


0
-0.02 - Compost #1
-0.04 K1 = 0.00042/day
-0.06 - k2= 0.00156/day k3= 0.00057/day
-0.08
-0.1 L

0
-0.01 - = 0.00069/day - Compost #2
-0.02-0.03- k2= 0.00037/day
-0.04-0.05-0.06- -0.07
0
-0.01 - Compost #3 S-0.02-0.03
-0.04 - k = 0.00057/day
-0.05-0.06-0.07
0
-0.01 0.00048/day - Compost #6
-0.02
-0.03 k2= 0.00031/day

-0.04
-0.05 - .....
0 20 40 60 80 100 120 Time (day)


Figure 3-4. Decomposition stages and reaction rate
coefficients for the four composts studied.






45


four composts studied are quite similar, falling in the range from 3.1x10-4 to 5.7x10-4/day. If the second reaction rate coefficients for the composts studied can be assumed to be representative through the remaining life of the compost, then a rough estimate of the time required for decomposition of 50% of the organic matter (half life) in these composts can be made according to following equation.


t0.5 = 0.693/k (3-5) where:

t0.5 = the half life time of the compost, (day), and
k = the first-order reaction coefficient, (1/day)

The estimated half life time of the composts tested is from 3.3 to 6.1 years. This estimate is comparable to the result reported by Varanka et al. (1976), who showed that it takes approximately 6 years to lose 50% of the sludge organic C when used in the field.

No significant changes in other physical and chemical properties of the composts were observed for the 120 day incubation period used in these studies (Table 3-2).


Effect of H2S on Compost Decomposition


The effect of H2S exposure on compost decomposition is illustrated in Figure 3-5, where CO2 evolved by the composts is expressed as mg-CO2 per g-C of the compost as a function of the H2S concentration (ppmv). The CO2 evolution is significantly increased with the increasing H2S concentration. The rate of this increase is greater at






46










15
1413 12


o 10 010
9
0
9 8
E "
- 7
6
0
5
4
3

0 4 8 12 16 20 24 28 32 (Thousands)
H2S Concentration (ppmv)







Figure 3-5. Effect of H2S exposure on the rate of
compost decomposition as measured by
CO2 respiration.





47


lower H2S concentrations. For example, the CO2 evolved at H2S concentrations near 6,000 ppm is approximately 8.5 mgCO2/g-C added, which is about 3.4 times that evolved when no H2S is present. At higher H2S concentrations the increase of CO2 evolution with H2S concentration is reduced e.g. when the H2S concentration is increased from 12,000 ppm to 32,000 ppm the CO2 evolution increases only by about 17%, or approximately 2 mg-CO2/ g-organic matter.

It can be seen from Figure 3-5 that the CO2 evolution from the compost has a strong dependence on the H2S concentration in the gas to which the compost is exposed. A linear relationship is obtained when plotting CO02 evolved as a function of the square root of H2S concentration in the gas, [H2S]0.5 for the range of H2S concentration less than 17,000 ppmv (Figure 3-6). The regression analysis result for the best fit line is:

CO2 = 2.62 + 0.082[H2S]0.5 (3-6)


where:

CO2 = CO evolved from compost, (mg/g of C added)
[H2S] = 2S concentration in the inlet gas stream,
(ppmv)

The correlation coefficient, R2 for the variables is

0.9234.

Equation 3-6 quantitatively describes the effect of H2S on the decomposition of composts. For example, CO2 evolutions at [H2S] = 0 and [H2S] = 1000 ppmv are calculated





48








15
14
13 v 12CO

10
d 9


7
6
5
o 4
3

0 20 40 60 80 100 120 140 160 180
Sqrt [H2S Concentration (ppmv)]









Figure 3-6. Plot of CO evolution as a function of square root of H2S concentration.





49


to be 2.62 and 5.21 mg-CO2/g-OM, respectively, by equation 3-6. The ratio of these two values is equal to the ratio of the first-order reaction coefficients for the reactions at these two conditions,

C02,1000/C02,0 = k1000/k0 = 1.99.

in other words, the decomposition rate for the compost exposed to 1000 ppmv H2S is 1.99 times of that for the compost not exposed to H2S. The half life times for compost #6 at both conditions are:

t0.5,0 = 0.693/0.00031 = 6.12 (years), and

t0.5,1000 = 6.12/1.99 = 3.08 (years). and for compost #3 are:

t0.5,0 = 0.693/0.00057 = 3.33 (years), and

t0.5,1000 = 3.33/1.99 = 1.67 (years).

No studies of similar effects have been reported in the existing literature. Thus the results obtained in this study can not be compared with the results of other investigations. Taylor et al.(1978) found that the highest S mineralization rates are observed during the period of highest CO2 evolution when 2 to 6% of sewage sludge compost is incubated in soils. Their results and those reported here suggest that the microbial activity of the compost was significantly enhanced by the addition of H2S, especially the activity of the sulfur oxidizing bacteria.

Oxidation of inorganic sulfur compounds is a basic phenomenon in nature. A number of bacteria have been






50
identified in soils and other environments that are capable of oxidizing organic and inorganic sulfur compounds (Roy and Trudinger, 1970). A high population of the oxidizing bacteria can be expected in the composts tested here. With sufficient H2S supply, the bioactivity of the sulfur oxidizing bacteria can be stimulated to result in an increase of microbial population and a corresponding increase in the evolution of CO2. Hydrogen sulfide is finally oxidized to sulfate through various pathways and intermediate stages ( Roy and Trudinger, 1970; Brock and Madigan, 1988; Yang and Allen, 1991; Allen and Yang, 1991).

After the 24 hours reaction period, the color of the compost changed from originally brown to yellowish-white, especially at high H2S concentrations. This feature indicates that a large amount of sulfur has accumulated in the compost.


Conclusions


Among the various biofilter materials, compost is frequently selected as a medium in applying biofiltration to air pollution control due to its unique properties and advantages. Knowledge of the characteristics of compost decomposition are important for both prediction of biofilter operation characteristics and degradation estimates, as well as in deciding on the appropriate disposal treatment and method for used compost. The studies described here indicate that the decomposition rates of the composts tested are much





51

lower than those reported by other researchers, who used fresh composts mixed with soils. Five to ten percent of the total-C in composts were decomposed during the 122 days incubation period. Compost half life times of the order 3 to 6 years are estimated for the composts studied, corresponding to loss 50% of their total-C due to decomposition. A multi-stage first-order reaction sequences is used to describe the decompositions. First-order reaction rate coefficients have been determined.

Decomposition rates are significantly increased when H2S is introduced to the compost. The half life of the compost is significantly reduced as a result of increased biological activity and CO2 respiration. For example, continuous exposure of compost to 1000 ppm H2S can result in reduction of the half life of the compost from about 6 years to 3 years due to enhanced microbiological activity alone. The results suggest that added H2S was oxidized by the sulfur oxidizing bacteria in the compost to sulfate.















CHAPTER 4
DETERMINATION OF THE DESIGN AND
OPERATIONAL PARAMETERS FOR BIOFILTER SYSTEMS Introduction

Extensive experimental work has been carried out in order to determine the design and operational parameters for a biofilter system. This research is essential for best operation as an air pollution control technology and for optimization of the system. This chapter describes the design and construction of lab scale biofilter systems, experimental methodology used and the results obtained.


System Design and Construction


Three biofilter systems were designed and constructed for different investigative purposes. Each system can be operated and controlled separately. Detailed information on each experimental system is presented below. The Dual Tower System


Most of the experimental work was carried out using a dual-tower experimental biofilter system. This configuration, which is shown in Figure 4-1, consists of parallel dual column filters. The two biofilter columns, identified as Tower #1 and Tower #2, can be run


52






53









Orifice Vent


o Orifice Vent House Air Supply



o 0 3
Water Supply House Air
-- Supply = + -- Water Supply
Samplin Ports


0 0




0
H2S H2S



Drain



Figure 4-1. Schematic drawing of the dual tower system.






54

simultaneously and controlled separately. The biofilter bed material is enclosed in transparent rigid Acrylic pipe, with an inner diameter of 0.15 meters (6 inches) and a height of 1.34 meters (4 feet). Each vertically mounted pipe can be packed with the desired compost up to a height of 1.2 meters (3.9 feet). The packed biofilter material is supported by a sieve plate to ensure a homogeneous distribution of the inlet gas stream across the face of the bed. Nonbiodegradable plastic screens are placed between the sieve plate and the biofilter material to avoid separation of smaller compost particles.

Sampling and measurement ports are located along the Acrylic column for compost and gas sampling, and pressure and temperature measurements. The sampling and measurement ports are shown in Figure 4-2a and b for Towers #1 and #2, respectively. An individual sampling/measurement port is identified by a letter-number system, where the letter indicates the function and the number indicates the location of the port. For example, TS11 means this port is used for temperature measurement and solid sampling, and is located on Tower #1 at location 1. All the other filter systems with multi measurement/sampling ports are identified in the same manner.

Room air is forced by a Gast Regenair Model R3105-1 air blower into the humidification chamber. The blower, which is driven by a 1/2 HP motor, generates a maximum flow of 1.5 m3/min (53 cfm) and a maximum pressure/vacuum of






55





015




.14 T = Temperature
-, me P = Pressure Ti G = Gas Sample S = Solid Sample


TS14
0

G13 TSI 3 I 012 TS12J


011
Gil







P10 [ T10










Figure 4-2. Sampling and measurement ports on
towers.
a. Tower #1






56





G25





24 T = Temperature P21 G P = Pressure
T26 G = Gas Sample TS251 s S = Solid Sample



TS24i 023


TS23 ii 22 TS22i



G21
TS21





T20 G20












Figure 4-2. Sampling and measurement ports on
towers.
b. Tower #2






57

1100/1000 mmH20 (43/40 inches) of water column. Humidification of the inlet air is achieved by atomizing water in the spray chamber, through which the room air passes. In addition, Pall rings are stacked in the spray chamber for extending wetted surface area providing better humidification. As a result relative humidities in the range 95 to 100% were routinely and continuously achieved.

Gaseous H2S with a purity of 99+%, which is stored in liquid form under pressure in a cylinder, is continuously leaked and mixed with the prehumidified air in the inlet lines (PVC pipe) to the towers. Plastic screen packing is placed downstream from the H2S introduction point for better mixing. Flow rates of air and H2S are controlled by plastic valves, which are located on the carrier gas inlet lines, and stainless steel needle valves, respectively. The flow rates are measured on calibrated flow meters to obtain the desired H2S concentration and gas flow through the towers. Measurements of temperature, pressure and gas flow rate are discussed in later sections.

A nozzle is installed on the top of Tower #1 in order to introduce water, or other liquid solutions if necessary, to the outlet end of the bed. Gas lines are made from PVC pipes. The towers and pipes are connected by flanges for convenient dismantling of the packed towers and compost changes. Cork-rubber gaskets are used to seal the flanges.






58

Portable Tower #3


The portable tower is made from PVC pipe with an inner diameter of 77 mm (3 in). This tower, which is shown in Figure 4-3, has a total height of 1.2 meters (3.94 ft) with an effective packing height of 1 meter (3.28 ft). The two ends of the pipe are covered by rubber caps and held by pipe clamps. Compost packed in the tower is supported by a packing of non-biodegradable plastic screen. Measurement ports for pressure, temperature and exhaust gas samples are located along the length of the tower. Gas to be tested is introduced through a port at the bottom of the tower. The overall gas flow rate is measured by a pre-calibrated flow meter after the effluent gas passes through a particulate filter. This portable tower was used intensively for pressure drop studies and for investigation of long term operation of compost #16.


Column System #4


A fourth column biofilter system was constructed for investigation of the effects of various operational variables on H2S removal (Figure 4-4). This multicolumn system includes eight compost columns, a manifold for introduction of test gas and several needle valves for flow control. The columns are made from PVC pipes with an inside diameter (ID) of 35 mm (1.25 in). Each column has a length of 300 mm ( 12 in) and an effective packing height of 250 mm





59



Gas Outlet T = Temperature oo a o P = Pressure G = Gas Sample
TG35
P31



TGP34



TGP33



TGP32



TGP31

P30 TG30 8 Gas Inlet









Figure 4-3. Schematic drawing of Tower #3.





60









ThermometerGas Outlet






Compost



Column



as Inlet Manifold Inlet Gas
Sampling Port




Figure 4-4. Schematic drawing of column system #4.






61

(10 in), which provide a 240 mL packing volume. The ends of the packed columns are plugged by rubber stoppers. Thermometers are inserted into the columns to measure temperatures. The gas flow rates are measured by a rotameter at the gas outlets. Effluent gas samples are taken from the outlet of the rotameter.


Measurement Methods


Periodic measurement of temperature, pressure drop and gas flow rate in the biofilter systems are carried out by the following devices.


Temperature


Temperature is measured by mercury in glass thermometers with a range from -20 oC to 110 oC and a minimum scale division of 1 oC.


Pressure Drop


Pressure drop is measured by manometers with a minimum reading of 1 millimeter water column (mmH20). In case the pressure drop is greeter than 1000 mmH20, the pressure drop is measured by mercury manometers with a minimum reading of

1 millimeter of mercury (mmHg). Gas Flow Rate


All the gas flow rates except those of Towers #1 and #2 are measured by pre-calibrated rotameters.






62

The gas flow rates in Towers #1 and #2 are measured by specially designed orifices. Two orifices for each tower were designed and made, one for low flow ranges and the other for high flow ranges. The orifices are made from plastic plate and installed on the outlet gas pipe lines (see Figure 4-1). The pressure drops across the orifices are measured and the flow rates are calculated according to the developed calibration equations.



Sampling Methods

Compost Samples


Compost samples in Towers #1 and #2 are taken from the solid sampling ports shown in Figures 4-2 a and b. The samples are taken at each port in a radial direction to the tower walls so that a representative sample can be obtained for that section. For composts not initially packed in columns, the samples are taken after the compost has been thoroughly mixed and very large particles ( diameter > 10 mm) have been eliminated.


Gas Samples


The inlet and outlet gas samples for each system are obtained directly from the gas sampling ports by extraction using gas-tight syringes. Gas samples extracted from other locations along the towers are obtained by using a gas sampling probe assembly. The gas sampling assembly, as





63

shown in Figure 4-5, consists of a Teflon probe, a Teflon filter and a sampling port connector. The Teflon probe is made from a piece of Teflon tubing (6.35 mm (1/4") in diameter), with 14 holes (1mm diameter) spaced evenly along the probe length. The probe is installed in the towers in such a way that all the holes are perpendicular to the tower's normal axis. Thus, representative gas samples from a cross section of the tower can be obtained. The Teflon filter is used to block out any small particles and water droplets which may be extracted during sampling. Gas samples are obtained through the sampling port located on the end of the assembly (see Figure 4-5) by a gas-tight syringe. The sampling port is sealed by a rubber GC septum.

When taking a sample, at least three full syringes of gas sample are wasted before the actual sample is taken for analysis. This procedure will eliminate residuals of previous gas samples remaining in the Teflon filter holder, in the syringes and in the probe, as well as condition the extraction system to the gas being sampled.

The gas samples are then diluted in 3-L Tedlar sampling bags by pure nitrogen (N2) to an appropriate concentration within the calibration range of the analyzer. The gases in the Tedlar bags are thoroughly mixed by gently kneading the bags and allowing them to sit for at least 10 minutes before analysis. Most of the samples taken are analyzed within 2 hours.





64










Biofilter Tower


Sampling Port Plastic Union











Teflon Filter Teflon Probe








Figure 4-5. Schematic drawing of the gas sampling
assembly.
.. . . . . . . . .
.. . . . . .....
























assembly.






65
In most cases, the H2S concentrations in the outlet gases are so low that the gas samples can be analyzed directly without any dilution. In the latter cases, Tedlar bags are directly connected to the sampling ports. Gas samples are forced into the bags as a result of the positive pressure of gas within the tower system. Also, Teflon filters are replaced by glass wool plugs to reduce the resistance to flow.

Each time after use, the Tedlar bags are purged at least three times with N2 to eliminate residual gas and vapor. The stability of gas samples in the Tedlar bags are discussed in Chapter 6.


Water Samples


Water samples analyzed are mainly biofilter wash waters from the tower drain outlets. When washing a packed tower, the entire wash water is collected in a container. Water samples are obtained from the container after mixing the wash water with a stirrer for a few minutes.


Compost Analysis Methods

Water Content


Two to five grams of wet compost are dried in an aluminum tray at an oven temperature of 70 oC until constant weight is obtained. Compost water content is determined by the difference in weight between the wet and dry composts (Robarge and Fernandez, 1986).






66




A known amount of wet compost is weighed into a 50-mL container. DI water is added to bring the liquid/solid ratio to 10 (Robarge and Fernandez, 1986). The sample is shaken for 30 minutes by a rotary shaker. Measurements of pH are made by a calibrated Corning Model M245 pH meter, which is accurate to � 0.01 pH.


Total Carbon and Total Nitrogen


Finely-ground, oven-dried compost sample (<100 mesh) are analyzed for total carbon and total nitrogen using a Carlo Erba Model NA 1500 CNS Analyzer.


Water Soluble Phosphorus (WSP)


A known amount of wet compost (2.5 g dry weight equivalent) is weighed into 50-mL centrifuge tubes. DI water is added to the tubes to obtain a compost to liquid ratio of 1:10 on a dry weight basis. These samples are allowed to agitate for a period of one hour on a mechanical shaker. The compost suspensions are then centrifuged at 6000 rpm for 15 minutes and filtered through Gelman 0.45 micrometer membrane filters. The filtered solutions are acidified (pH<2.0) with one drop of concentrated H2SO4 and stored at 4 oC until analyzed. The soluble reactive P (SRP) in the filtered extract is determined colorimetrically (APHA, 1989) using a Shimadzu UV-160 spectrophotometer with

1 cm path length at 880 nm wavelength.






67


Acid-extractable Cations


Two and half (2.5) g of finely-ground, oven-dried sample is weighed into 50-mL centrifuge tubes. Twenty five

(25) mL of lM HCI is added and the tubes are shaken for 3 hours on a mechanical shaker. The compost suspensions are centrifuged at 6000 rpm for 15 minutes and filtered through Gelman 0.45 micrometer membrane filters. The solutions are analyzed for Fe, Al, Ca, Mg, Cu and Mn on an Inductively Coupled Argon Plasma Spectrometer (ICAP) (APHA, 1989).


Particle Size Distribution


The compost is dried in oven at 700C for 24 hours. Particle size distribution by weight is measured by passing the dried compost through a series of sieves (U.S.A. Standard Testing Sieve, A.S.T.M. E-11 Specification, Fisher Scientific Company) and weighing the residue. Porosity


Compost porosity is determined according to Danielson and Sutherland (1986).


Organic Matter


After determination of compost water content, the samples are placed in a muffle furnace and baked for 24 hours at 450 oC. Organic matter is determined by the losson-ignition (LOI) (Robarge and Fernandez, 1986).






68

Particle Density


Particle density of the compost is measured according to Blake and Hartge (1986a).


Bulk Density


Bulk density of the compost is measured according to Blake and Hartge (1986b).


Sulfur Analysis Methods


Intensive and detailed laboratory work has been carried out on the analysis of sulfur compounds in order to obtain a better understanding of the biochemical reactions involved in the H2S oxidizing processes occurring in the biofilters. The procedures for determining various sulfur compounds in compost, in water, and in the waste gases are described in this section.


Sulfur in Compost


The analysis of sulfur in compost includes the determination of total sulfur (total-S) and fractionation of the total sulfur into inorganic and organic constituents.

Many wet chemical procedures have been developed to fractionate the total sulfur pool in sediments, soils, and peat into its inorganic and organic constituent compounds. Very little information, however, is available about such analyses for compost. The sulfur analyses conducted in this research include the quantitative determination of acid





69


volatile sulfur ( sulfide-S), water soluble sulfur (soluble sulfate), insoluble sulfate (S042-), elemental sulfur (SO), Pyrite sulfur, ester sulfur (organic-S) and total sulfur.

Each of the wet chemical procedures involved the reduction of S to H2S in a Johnson-Nishita apparatus (Johnson and Nishita, 1952) and trapping the evolved H2S in zinc acetate-sodium acetate (ZnAc-NaAc) solutions. Trapped sulfide is quantified by iodometric titration (APHA, 1989) with a 0.025N iodine solution and 0.025N Na2S203 titrant.

The distillation apparatus incorporated slight modifications of that used by Johnson and Nishita and is similar to that used by Wieder and Lang (1985). Figure 4-6 shows the distillation assembly. The reaction flask is a 250-mL, round-bottom, three-neck flask, with an N2 inlet via a bleed tube inserted in one neck and the central neck is connected to a condenser. Ultra high purity (>99.999%) nitrogen is used to sweep out the H2S and to maintain the reaction flask in a reducing environment. The third neck of the flask is fitted with a stopper to allow introduction of liquid solutions to the flask. The ZnAc-NaAc solution is made by dissolving 50 g of zinc acetate dihydrate [Zn(CH3COO)2.2H20] and 12.5 g of sodium acetate trihydrate (CH3COONa.3H20) in 800 mL of DI water and adjusting the final volume to 1 liter (Tabatabai, 1982). Twenty-five (25) mL of this solution is mixed with 100 mL of DI water and this solution is used to fill two traps used in series. The first trap contains 100 mL and the second one contains 25 mL






70














































Figure 4-6. Photograph of the sulfur distillation
assembly.






71


of the H2S absorption solution. Even though the contents of both traps are analyzed, more than 95% of the H2S is consistently recovered in the first of the two series traps. The material to be analyzed is added to the reaction flask through the side neck. The system is purged with N2 at bubbling rate of 1-2 bubbles per second in the ZnAc-NaAc traps for 10 minutes before the introduction of additional reagents. The materials are boiled for 1 hour, the traps are removed and sulfide titrated.

The procedures and methods used for analysis of the total sulfur and various organic and inorganic sulfur compounds in compost are similar to those used for sulfur analyses in peat, soil and sediments described by Zhabina and Volkov (1978), Tabatabai (1982), and Wieder and Lang (1985). Minor modifications were made for the compost sulfur analyses in this research. The procedures are illustrated in Figure 4-7.

All results are expressed as mg sulfur per gram of compost on a dry basis (mg-S/g). Compost moisture content is determined from a sub-sample by drying the compost at 70 OC to constant weight.

Compost samples are stored in plastic bottles and refrigerated at 2-4 oC before analysis. In most cases, however, compost samples are analyzed immediately after sampling.








_FRESH WET COMPOST

Water Content Sto tor Determination Later Use


Acid Water HCI Preci. Prec- Acetone Prc- CrC12 cHydriodic Digestion Extraction Heating pi Filtration tate Extraction pitat Reduction pitate Reduction i Solution

Reducing Reducing Gas Reducing CrC12 Gas Gas Mixture Mixture Mixture Reduction Reduction Reduction Reduction Gas 4

H2 H2 H2S H2S H2 H2S H2S Absorption Absorption Absorption Absorption Absorption Absorption Absorption



lodometr lodomet ric lodometric lodometric lodometric lodometric lodometric Tition Tration Titttron Titration Titration Titration Ttration



Total-S Water-Soluble-S Sulfide-S Sulfate-S Elemental-S Pyrite-S Ester-S (Sulfate-S)


Figure 4-7. Flow chart of the sulfur analysis procedures for compost.






73

The fresh wet compost sample is divided into five subsamples. The latter are used for the analyses of 1) compost moisture content, 2) total-S, 3) water-soluble-S, 4) inorganic and organic sulfur fractions, and 5) storage for later use.


Total sulfur


Total-S is determined by oxidation (acid digestion) of the various reduced sulfur constituents to sulfate and followed by reduction of the sulfate to H2S. The H2S is then trapped and titrated as described above. In most of the analyses, 0.5-5.0 g of fresh compost is used, depending on the sulfur content. In some analyses, oven dried compost is used. In the latter case, the compost samples are finely grounded (<40 mesh) and a correspondingly smaller size of compost sample is used. The compost is subjected to acid digestion as described by Tabatabai (1982). The digest is quantitatively transferred into a 100-mL volumetric flask and the volume is adjusted with IN HC1. Reduction of the sulfate is carried out by a reducing mixture. The reducing mixture contains 50% hypophosphorous acid, 90% formic acid, and hydriodic acid in a 4:2:1 proportion and is prepared as described by Tabatabai (1982). Depending on the sulfur content, 1 to 5 mL aliquot of the digest is transferred into the distillation flask. With aliquots >2 mL, the volume is reduced to about 2 mL by heating the flask on an electric heating mantle. Five (5) mL of the reducing mixture is






74

added to the flask and the material is subjected to reduction and hydrogen sulfide is liberated, collected and titrated as described above.


Water soluble sulfur


Water-soluble-S is determined by shaking 2-5 g of fresh compost in DI water with a liquid to solid ratio of 10:1 for 30 minutes on a rotary shaker at a speed of 140 /min. An aliquot of the compost extract is then subjected to reduction, H2S absorption, and titration successively, as described previously.

The following analyses are conducted in succession:


Sulfide sulfur


Sulfide-S or acid-volatile sulfur (AVS) is determined by introducing 8 mL of 12N HCI to the compost sample in the reaction flask. Heat is applied after 10 min, the materials are brought to boiling, and after 45 min the traps are removed and the sulfide titrated.


Sulfate sulfur


The content of the reaction flask is filtered by a #42 Whatman filter. The filtrate is then subjected to reduction. H2S is then trapped and titrated. This is an alternative way of carrying out sulfate analysis. The results are comparable to the summation of water soluble-S and P-extractable-S.






75

Elemental sulfur


The precipitate obtained above is dried with filter papers as described by Zhabina and Volkov (1978) and is extracted with analytical grade acetone. The volume of acetone used (mL) is 40 times that of the equivalent dry weight of the compost sample (g). The extraction flask is covered with Parafilm and placed on a rotary shaker at a speed of 140/min for 16 hours. The mixture is filtered and rinsed with additional acetone. Either a fraction of or the entire filtrate are subjected to Cr2+ reduction as described by Zhabina and Volkov (1978).


Pyrite sulfur


The residue left after So extraction is subjected to chromium reduction. The Cr2+ was produced by passing a 1 M solution of CrCl3.6H20 in 0.5 M HC1 through a Jones reductor column containing Zn amalgamated with Hg (Zhabina and Volkov, 1978). Preparation of the Jones reductor is described by Swift (1950) and Patterson and Thomas (1952). Ten (10) mL of ethanol is added to the flask followed by 20 mL of 12N HC1 and 16 mL of lM CrC12 solution. Heat is applied after 30 min, the H2S evolved is absorbed in the traps and titrated as above.






76


Organic sulfur


Following the previous procedure, the insoluble residue in the reaction flask is filtered through a #42 Whatman filter and is washed repeatedly with acidified DI water to remove chromium ions. The residue is then subjected to reduction with the reducing mixture as described above. The organic sulfur determined this way is mainly ester-S.

In addition to the literature mentioned above, similar and dissimilar procedures for determining sulfur constituents in sediments, peat and soil are also described by Smittenberg et al. (1951), Freney (1958) , Johnson and Henderson (1979), David et al. (1982), and Hsieh and Yang (1989). An excellent comparison of some of these methods is reported by Wieder and Lang (1985). Sulfur in the Aqueous Phase


Sulfur in water samples, such as in tower wash water and drainage, are analyzed for sulfide, sulfate, and/or total sulfur.


Sulfate sulfur


Sulfate in most of the water samples is determined by a turbidimetric method (APHA, 1989). A Milton Roy Model Spectronic 21 Spectrometer was used at 420 nm to measure the turbidity. Color or suspended matter in large amounts will interfere with this method. In the case of dark colored






77


water samples, the reduction, H2S absorption and iodometric titration procedures described previously are used. Sulfide sulfur


Sulfide was measured according to APHA (1989). The samples are pretreated to remove interfering substances and to separate insoluble sulfide.


Total-S


Total-S was determined by using 2-5 mL of the aqueous sample depends on its sulfur content. The sample is then analyzed as described for the total sulfur in compost. Sulfur in Waste Gas


Gas samples are analyzed by a commercial gas chromatogragh equipped with a flame photometric detector (GC/FPD), a Tracor Model 250H Analyzer. The detection limits and operational conditions for the analyzer are summarized in Table 4-1. Detailed information concerning the principles and conditions of operation are reported elsewhere (Yang, 1988). In the early stages of this study, the component peaks are recorded by a chart recorder (Texas Instruments, Inc., Model Recti/Riter II) and concentrations of sulfur compounds are determined by measuring the peak heights. Later in the study (from August 1990 to September 1991) the chart recorder was replaced by a Spectra-Physics model SP4290 integrator and the measured concentrations are read






78






Table 4-1. Retention times, limits of detection and
operating conditions for the Tracor 250H
Analyzer.


Compoundsa


Item H2S MM DMS


Retention Time (min) 1.15 2.16 4.02 Detection Limit (ppmv) 0.01 0.01 0.02


Operating Conditions:

Temperatures: Valve: 50 OC Column: 70 OC Detector: 110 OC

Flow Rates: Nitrogen: 80 mL/min (carrier gas) Oxygen: 21 mL/min (flame gas) Hydrogen: 80 mL/min (flame gas) Sample: 40 mL/min

Cylinder
Pressures: Air 40 psig ( for sampling valve activation) Hydrogen: 53 psig Oxygen: 40 psig Nitrogen: 80 psig
a MM: methyl mercaptan.
DMS: dimethyl sulfide.






79

directly from the printout of the integrator. The Tracor analyzer is periodically calibrated for H2S, methyl mercaptan (MM) and dimethyl sulfide (DMS) with standards purchased from National Speciality Gases, Inc.


Results and Discussion


Biofiltration is a process that involves physical, chemical, and biological processes. Many variables, such as temperature, compost water content, specific acidity of compost, sulfate content in the compost etc., can affect the function of the system. It is impossible to obtain optimum, longer term performance from a biofiltration system without an in depth understanding of the system properties and proper control of important variables. Extensive evaluations of the system properties have been conducted during the course of this study. The results presented in this section are divided into several subsections according to specific investigations undertaken. Composts from different sources (Table 3-1) have been used in the laboratory studies. The physical and chemical properties of these composts are summarized in Table 4-2. Applications of each of these composts are mentioned in the corresponding study subsections.


Pressure Drop


The energy consumption obtained in operating a biofiltration system is primarily that required by the







80





Table 4-2. Summary of initial compost properties.


Compost ID #


Property 12 13 14 16 17


pH 2.61 1.60 6.44 6.66 8.10 Bulk Density (g/cc) 0.30 0.27 0.18 0.22 0.20 Particle Den. (g/cc) -- 1.73 1.75 1.90 1.78 Porosity (%) -- 84.3 89.7 88.4 88.7 Water Content (wt%) 45.6 54.9 62.4 56.5 62.7 Organic Matter (wt%) 66.5 66.8 59.3 64.3 62.6 Total-S (mg-S/g) 44.8 70.4 -- -- 0.74 Water-P (mg/kg) 140 223 152 114 167 Total-C (wt%) 34.3 31.0 31.3 30.5 40.9 Total-N (wt%) 1.89 3.24 1.75 4.27 1.30 C/N 18.1 9.57 17.9 7.14 31.5

Metals (mg/kg)
Ca 47400 28900 145000 18000 26700
Mg 280 255 4880 1450 120
Zn 38.0 48.0 201 66.5 18.0 Cu 121 91.0 60.0 11.0 93.5 Mn 5.50 55.0 96.5 89.5 9.50 Fe 1900 805 6160 529 1510






81

blower (fans) to move contaminated air at the specified flow rate through the filter bed. However, the pressure drop across the filter bed increases markedly as the flow rate is increased. Since the pressure drop will be determined by the depth of the filter bed it is necessary that the gas velocity should be kept as low as possible.

The pressure drop across a compost bed filter can be lowered by physical treatment of the compost particle content. Such a procedure is an important requirement in optimizing the operation of a filter bed, because operation at a reduced resistance to flow allows the gas velocity as well as the volumetric flow rate to be significantly increased with little or no change in energy consumption. This will in turn increase the biofilter capacity and reduce the required filter size.

The following physical factors determine pressure drop across the filter bed:



1) Particle size distribution in the compost
2) Condition of the filter packing
3) Height of the filter bed
4) Water content of the compost
5) Gas velocity, and
6) Porosity of the compost.


A representative sample of compost #12 was air dried and the particle size distribution determined by weighing the fractions penetrating a series of standard sieves. The particles are classified into 5 size groups in the range >12 mm to <1.2 mm (see Table 4-3). Compost samples in each of






82








Table 4-3. Particle size range distributions for
selected composts.


Particle Size Range Distribution (wt%)

Compost
ID # A B C D E


12 20.0 22.5 10.0 13.1 34.4 14 27.7 26.9 8.10 11.6 25.7 13 21.4 24.5 6.70 15.8 31.6 17 0.00 33.4 14.4 22.5 29.7


A: diam. > 12 mm
B: 3.35 < diam. <12 mm
C: 2.36 < diam. <3.35 mm D: 1.18 < diam. <2.36 mm
E: diam. <1.18 mm





83


the particle size groups were analyzed for water content then wetted to obtain a water content of about 50%, by spraying and mixing water with the compost samples. These treated samples were then used to determine the pressure drop as a function of the 5 particle size range classes by adjusting air velocities in the range 0.02 m/s to 0.28 m/s. All tests were carried out with the same compost bed height (1 m) and water content (50%).

The results of these studies are presented in Figure 48, where it is seen that the pressure drop increases significantly with increasing gas velocity for a bed of small particle size (<1.2mm). For particles greater than 1.2 mm the pressure drop increases to a much lesser extent with increasing velocity as shown by a comparison of data for particle classes D and E. Clearly significant pressure drops observed in operating filter beds are a result of the presence of small particles with sizes less than 1 mm. For example, at a representative gas velocity of 0.03 m/s, which is equivalent to a loading rate of 110 m3/m2-hr, the pressure drop realized by a 1-m bed of particles size classified as <1.2 mm is 390 mm H20, whereas the pressure drop obtained for the same packing height and the same gas velocity, for particles classified as >12 mm is only 2 mm H20. Thus, under these conditions the pressure drop created by 1 mm or less particles is about 200 times that caused by an equivalent bed composed of 12 mm or greater particles.





84







3.5
Compost Packing Height = 1 meter
3 _ Compost Water Content = 50% (Wt/Wt)
Velocity (m/s):
o 0.02 m/s + 0.06 m/s
2.5 o 0.11 m/s a 0.16 m/s
x 0.22 m/s v 0.28 m/s I2


1.5


0- 1


0.5


A B C D E Particle Size Range


A: d > 12 mm
B: 3.35 < d < 12 mm
C: 2.36 < d < 3.35 mm D: 1.18 < d < 2.36 mm
E: d < 1.18 mm




Figure 4-8. Pressure drop as a function of particle size range for different gas velocities.






85
The relationships between pressure drop and packing height at constant gas velocity for particle size ranges B (3.33 to 12 mm) and E (<1.2 mm), and the parent compost #12 are shown in Figure 4-9. It is seen that the pressure drop increases approximately linearly with packing height for all three samples. It is important to note that the pressure drop values obtained for the parent compost are intermediate between the strong dependence of pressure drop on packing height for small particles (class E) and the weak dependence on packing height for larger particles (class B).

The dependence of pressure drop on compost water content is not as consistent as that for particle size. Qualitatively, sewage sludge treated compost contains more viscous and adhesive small particles than are found in untreated compost. Thus, when the water content of the sewage sludge treated compost is increased, coagulation of small particles is enhanced and the pressure drop increases sharply. However, the build-up in pressure may be suddenly released by channeling, i.e., a breakdown of the overall flow restriction by the formation of a channel of much less resistance caused by a separation of packed materials.

The effects of gas velocity on pressure drop across four typical biofilter materials are shown in Figure 4-10. As expected, the pressure drop increases rapidly with increasing gas velocity. The pressure drop depends on the way the filter is packed. Composts #12 and #13 are similar in nature but filter material #13 was more compacted than






86












450
Gas Velocity = 0.03 m/s
400 - Compost Water Content = 50% (Wt/Wt)
o diam. < 1.18 mm
350 + 3.36 < diam. < 12 mm
0 Compost #12 0300

250
2
O 200 a

150

100

50

0 0.2 0.4 0.6 0.8 1 Compost Packing Height (m)










Figure 4-9. Pressure drop as a function of packing
height for different compost particle
size ranges.




Full Text
153
Effect of Water-Compost Contact Time
on Leaching Efficiency
Water is used as a solvent to determine the effect of
contact time on S042- leaching efficiency. Three (3.0) g of
wet compost #13-1 was weighed into each of nine 50-mL
beakers. Thirty (30) mL of DI water is then added to each
beaker. The beakers with compost and water are allowed to
contact without disturbance. After the desired period of
time (from 5 to 120 minutes), the content of the beakers are
filtered and the filtrate subjected to pH and sulfate
determination. Another container with 3 g compost and 30 mL
of DI water are shaken for 60 minutes as a control.
The results are shown in Figure 5-6. At constant water
to compost weight ratios of 10:1 and without shaking, the
S042- leaching efficiency is between 51% and 68%. The S042-
leaching efficiency increases with water-compost contact
time. The maximum leaching efficiency is achieved in about
one hour.
Effect of Water to Compost Ratio
on S042~ Leaching Efficiency
The effect of water to compost weight ratio on S042-
leaching efficiency was determined by shaking 1, 2 and 3 g
of wet compost in 30 mL of DI water for 30 minutes. The
supernatant was filtered and sulfate was determined in the
filtrates. The results are shown in Figure 5-7. It is
obvious that S042- leaching efficiency is increased
significantly with increasing water to compost weight


Hydrogen Sulfide Removal Efficiency (%)
107
Figure 4-17. Effect of sulfate on H2S removal
efficiency.


95
before changing the operational parameters. No change in
H2S removal capacity was found due to the system transfer.
Data were not taken until system #4 reached a stable
condition after any changes in the operational parameters or
conditions were made. Each column in System #4 was packed
with equal weights of compost. The range of compost water
contents evaluated was from 0%, oven dried compost, to about
62%, the maximum water holding capacity of the compost.
Water content of the compost in each column was adjusted to
a desired range by either adding DI water to the compost or
by gently drying the compost in room air. One of the
composts tested was thoroughly dried in an oven at 110 C
for 24 hours to obtain water free compost. The system was
operated at room temperature with a gas loading rate of
about 15 m3/m2-hr. Inlet H2S concentrations were controlled
in a range between 80 and 110 ppmv. The results are
illustrated in Figure 4-12.
The H2S removal efficiency was maintained at a high
value, 99.9+%, with little variation being observed when
the CWC was varied from 30% to 62%. When the CWC was
reduced below 30%, the H2S removal efficiency decreased
linearly with the CWC. Very little removal of H2S was
observed for the oven dried compost. The residual
effeciency of the latter is probably due only to chemical
oxidation and adsorption of H2S on the compost.
Water is essential for all living organisms. All
biological metabolic processes require water as a medium or


Hydrogen Sulfide Removal Efficiency (%)
100
Compost pH
Figure 4-14. Effect of compost pH on H2S removal
efficiency.
Condition a:
H2S loading rate: 10.5 g/m -hr
Gas loading rate: 15 m3/m2-hr
Condition b:
H2S loading rate: 35.4 g/m3-hr
Gas loading rate: 26.1 nr/m -hr


Hydrogen Sulfide Removal Efficiency (%)
98
Figure 4-13. Time required for dried compost to
recover optimum efficiency.
Original compost water content:
WCE3 = 21.4%,
WCF4 = 14.3%.


161
Waste Gas inlet
SIEBO-stone Gas
Distribution System
Central Shaft
t
Humidifier
Plan View
Section A-A
Figure 6-1. Schematic diagram of the Kanapaha
biofilter bed system.


43
The integrated form of equation 3-3 becomes:
cti = ciexP(kit) (3-4)
where:
= organic-C present at the beginning of a
decomposition stage.
= organic-C present at the end of a decomposition
stage at time = t, and
k^ = the first-order reaction rate coefficient.
Decomposition stages and the corresponding reaction
rate coefficients for the four composts are presented
graphically in Figure 3-4. The decomposition of compost #3
is described in one stage and the decompositions of compost
#2 and #6 are best described in two stages. It can be seen
that the reaction rate coefficient values of k-^ and k2 for
these two composts are very similar. This similarity
indicates that these two composts have similar organic C
composition.
The behavior of Compost #1 is markedly different from
those of the other composts. During the first 3 days of
incubation, C02 evolution is rapid followed by a lag period,
lasting for the following 40 days. A second period of high
decomposition rate was observed between 42 and 70 days.
During the remaining period of incubation (after 70 days)',
the C02 evolution rate for this compost is similar to those
for the other composts. The decomposition rate for compost
#1 may be described as a 3 stage series of first-order
reactions.
Within overal experimental error, reaction rate
coefficients for the final stage of decomposition for the


TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES ix
ABSTRACT xiii
CHAPTERS
1 INTRODUCTION 1
2 BACKGROUND 6
Properties of Hydrogen Sulfide and Regulations .. 6
Physical and Chemical Properties 6
Toxicity of H2S 9
Sources of H2S Emissions and Regulations 11
Biofiltration as an Air Pollution
Control Technology 12
History and Development 13
Applications 15
Theoretical Basis 16
Biological Oxidation of Hydrogen Sulfide 21
3 PROPERTIES OF COMPOSTS AND THEIR DECOMPOSITION .. 25
Introduction 25
Selection of Filter Materials 26
Decomposition of Composts under Aerobic
Conditions 28
Materials and Methods 32
Results and Discussion 39
Decomposition of composts 39
Effect of H2S on compost decomposition 45
Conclusions 50
4 DETERMINATION OF THE DESIGN AND OPERATIONAL
PARAMETERS FOR BIOFILTER SYSTEMS 52
Introduction 52
System Design and Construction 52
IV


The Dual Tower System 52
Portable Tower #3 58
Column System #4 58
Measurement Methods 61
Temperature 61
Pressure Drop 61
Gas Flow Rate 61
Sampling Methods 62
Compost Samples 62
Gas Samples 62
Water Samples 65
Compost Analysis Methods 65
Water Content 65
pH 66
Total Carbon and Total Nitrogen 66
Water Soluble Phosphorus (WSP) 66
Acid-extractable Cations 67
Particle Size Distribution 67
Porosity 67
Organic Matter 67
Particle Density 68
Bulk density 68
Sulfur Analysis Methods 68
Sulfur in Compost 68
Total Sulfur 73
Water soluble sulfur 74
Sulfide sulfur 74
Sulfate sulfur 74
Elemental sulfur 75
Pyrite sulfur 75
Organic sulfur 76
Sulfur in the Aqueous Phase 76
Sulfate sulfur 76
Sulfide sulfur 77
Total-S 77
Sulfur in Waste Gas 77
Results and Discussion 79
Pressure Drop 79
Effect of Gas Retention Time on H2S Removal ... 88
Effect of Concentration of H2S on Its Removal 90
Effect of H2S Loading Rate on Its Removal 92
Effect of Compost Water Content
on H2S Removal 93
Effect of Compost Acidity on H2S Removal 97
Effect of Temperature on H2S Removal 102
Effect of Sulfate on H2S Removal 106
Effect of Nutrient Addition on H2S Removal .... 108
Kinetics of H2S Oxidation in the Biofilter 109
Theoretical considerations 109
Determination of the kinetics of H2S
Oxidation in a biofilter 114
v


114
where C is the effluent gas concentration, CQ is the
influent gas concentration, H is the height of the filter
bed and Ug is the gas velocity. This equation is a
different way of expressing equation 4-4, specifically for a
biofilter.
Under diffusion limiting conditions, the kinetics
expression is:
C/CQ = [1 (H/Ug)(K0Dea/2mC06)^]2 (4-8)
where a is the interfacial area per unit volume, De is the
effective diffusion coefficient, m is the distribution
coefficient of the component, and <5 is the biolayer
thickness.
Determination of the Kinetics of H2S Oxidation in a Biofilter
Hydrogen sulfide elimination rates are measured when
the biofilter has reached a steady state conditions. Gas
samples are taken at different locations along the tower and
the H2S concentrations in the gas samples are analyzed. The
initial H2S concentrations, CQ, in the inlet gas stream are
varied from low to high values in order to determine the
kinetics under different situations as described in the
previous section. The results for compost #17 packed in
Tower #2 are presented in Figures 4-19, 4-20, 4-21, 4-22,
and 4-23.
Under high H2S concentrations in the inlet gas stream
(H2S > 400 ppmv), the reaction appears to follow zero-order


Table 5-1. Sulfur fractionation of original compost #17A and compost at different heights
in the filter.
Original
TS11 (0.125m)
TS13 (0.625m)
TS14 (0.875m)
(mg-S/g) (%)
(mg-S/g) (%)
(mg-S/g) (%)
(mg-S/g) (%)
Total-S
0.740
100
129
100
13.0
100
5.18
100
Organic-S
0.470
63.5
6.05
4.69
3.07
23.6
0.190
3.67
C-bonded-S
0.420
56.8
4.69
3.63
2.87
22.0
0.00
0.00
Ester-S
0.050
6.76
1.36
1.05
0.200
1.54
0.190
3.67
Inorganic-S
0.270
36.5
123
95.3
9.95
76.4
4.99
96.3
FeS2-S
0.050
6.76
22.3
17.3
0.310
2.38
0.800
15.4
FeS-S
0.030
4.05
0.240
0.190
0.090
0.690
0.100
1.93
S-S
0.00
0.00
8.44
6.54
0.24
1.84
0.12
2.32
so4 Z-S
0.19
25.7
92.1
71.3
9.31
71.5
3.97
76.6
Water-sol.-S
0.090
12.2
66.6
51.6
6.71
51.5
3.09
59.6
P-extract.-S
0.00
0.00
18.1
14.0
1.19
9.14
0.440
8.49
Insoluble-S
0.100
13.5
7.36
5.70
1.41
10.8
0.440
8.49
141


29
Table 3-1. Description of composts used for this study.
Compost
ID#
Source
Description
1
Pompano Beacha
Fort Lauderdale sewage sludge compost.
Not completely composted. Seven months
old when first used (used for
decomposition study).
2
Pompano Beach
Two parts yard waste and one part stable
cleaning sewage sludge mixed and
composted. Seven months old when first
used (used for decomposition study).
3
Pompano Beach
Yard trash compost. 13 months old when
first used (used for decomposition
study).
6
Pompano Beach
25% by volume of sewage sludge compost
and 75% of yard trash mixed and composted
about 19 months old when first used (used
for decomposition study).
12
Kanapahab
Pompano Beach compost similar to Compost
#6 mixed with tree bark, yard waste and
sewage sludge; lime was used to adjust pH
before use. Used in Kanapaha filter bed
from 11/20/88. Compost obtained from the
filter bed in 5/16/90.
13
Kanapaha
Same as #12, compost obtained and used in
Tower #1 from 12/20/90.
13-1
Kanapaha
Same as #12, compost obtained in 2/5/91.
13-2
Kanapaha
Same as #12, compost obtained in 3/20/91.
14
Kanapaha
Yard trash, grass and sewage sludge were
mixed and composted; lime was used to
adjust pH; about 2.5 years old when
obtained and used in Tower #2, 12/20/90.


42
organic-C, (i i) proteins, (iii) hemicellulose, (iv)
cellulose, and (v) lignin. In the examples mentioned above,
fresh plant biomass releases much more C02 (especially in
the early stages of the incubation) than the digested ones
because it contains much more easily decomposable organic-C.
In this study, all the composts used were well aged or
completely composted. Most of the easily decomposable
organic-C such as soluble organic-C, starch and proteins
have been decomposed during the composting process. The
main organic-C species remaining in the composts studied
are the more oxidation resistant residues of the original
organic matter (Biddlestone et al., 1987). Also, the four
composts studied are either yard waste compost or mixtures
of sewage sludge with yard wastes such as wood chips, leaves
and tree trimmings, etc.. A high content of cellulose and
lignin can be expected in these materials. This feature may
explain why the decomposition rates for the composts studied
here are relatively low.
Decomposition of a complex substrate C is usually
described by a multistage first-order decomposition sequence
(Reddy et al., 1980; Gilmour et al., 1985). The mathematical
rate equations can be written as follows:
-dCj/dt = kjCi (3-3)
where i refers to a particular stage of decomposition.


(!o/ bJui
44
0
-0.02
_
- Compost #1
-
K, = 0.00042/day \
-
k2= 0.00156/day
0.00057/day
-
0
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-0.07
Figure 3-4. Decomposition stages and reaction rate
coefficients for the four composts studied.


142
The original compost has a total-S content of 0.74 mg-
S/g, 64% of which is organic-S. The total-S and sulfur
constitution of the original compost are close to similar
data reported by David et al.(1982) for surface soils in a
forest. High organic-S content is a good indication of high
biomass and microbial sulfohydrolase activity of the
compost. This correlation was quantitatively determined by
David and coworkers (1982).
Inorganic-S is dominant in the 'used' filter compost
samples (>95%). A lower value for TS13 is probably due to
analytical error. Fifty to sixty percent of the total-S in
the biofilter compost is water soluble-S, which indicates
that the final product of H2S oxidation is H2S04. A large
amount of FeS-S and S is measured in the lower region of
the filter bed (TS11, 0-0.125m of the filter).
The total-S distribution profile is graphically shown
in Figure 5-4. As the test gas flows through the filter bed
in an upward direction, the lower region of the compost bed
is always exposed to higher H2S concentrations than the
upper region. This results in sulfate formation and
accumulation at a higher rate in the lower region of the
bed. The higher fraction of intermediately oxidized sulfur
compounds, FeS and S, in the lower region is a result of
incomplete oxidation of H2S due to high H2S concentrations
and reduced biological activity of the biomass as a result
of lowered pH and high sulfate content in this region.


91
Table 4-5. Effect of H2S concentration on removal
efficiency.
Gas Flow
Rate
(Lpm)
h2s
Reten. Loading H2S
Time Rate Inlet
(s) (g-S/m-hr) (ppmv)
H,S Removal
Outlet Eff.
(ppmv) (%)
30.0
35.4
0.72
5.51
BDL
99.9+
30.4
35.0
1.69
12.8
BDL
99.9+
30.8
34.5
16.9
126
0.01
99.9+
30.8
34.5
26.3
196
1.31
99.3
30.0
35.4
29.6
227
0.01
99.9+
30.0
35.4
38.1
292
0.01
99.9+
31.2
34.0
70.4
518
1.04
99.8


Hydrogen Sulfide Removal Efficiency (%)
96
Compost Water Content (Wt%)
Figure 4-12. Effect of compost water content on H2S
removal efficiency.


196
Smittenberg, J.; Harmsen, G.W.; Quispel, A. and Otzen, D.
"Rapid Methods for Determining Different Types of Sulfur
Compounds in Soil," Plant and Soil 3: 353 (1951).
Sommers, L.E.; Nelson, D.W. Terry, R.E., and Silvieria, D.J.
"Nitrogen and Metal Contamination of Natural Waters from
Sewage Disposal on Land," Tech. Rep. No 89, Purdue Univ.
Water Resour. Res. Center, W. Lafayette, IN, 1976.
Starkey, R.L. "Oxidation and Reduction of Sulfur Compounds
in Soils," Soil Sci. 101: 297 (1966).
Stotzky, G."Microbial Respiration," in Methods of Soil
Analysis, Part 2? Black C.A. Ed, Agronomy 9: 1550 (1965).
Sublette, K.L. and Sylvester, N.D. "Oxidation of Hydrogen
Sulfide by Thiobacillus Denitrificans: Desulfurization of
Natural Gas," Biotechnology and Bioengineering, 29: 249
(1987) .
Sweeney, D.W. and Graetz G.A. "Chemical and Decomposition
Characteristics of Anaerobic Digester Effluents Applied to
soil," J. Environ. Qual. 17: 309 (1988).
Swift, E.H. Introductory Quantitative Analysis, Principles
and Selected Procedures. New York Prentice-Hall, Inc., 1950.
Tabatabai, M.A. "Sulfur," in Methods of Soil Analysis. Part
2. Chemical and Microbiological Properties. Page, A.L. Ed.,
2nd ed, American Society of Agronomy, Madison, Wisconsin,
1982.
Taylor, J.M.; Sikora, L.J.; Tester, C.F. and Parr., J.F.
"Decomposition of Sewage Sludge Compost in Soil: II.
Phosphorus and Sulfur Transformations," J. Environ. Quanl.
7: 119 (1978).
Terasawa, M.; Hira, M. and Kubota, H. "Soil Deodorization
Systems," BioCycle 27: 28 (1986).
Terry, R.E.; Nelson, D.W. and Sommers L.E. "Carbon Cycling
During Sewage Sludge Decomposition in Soils," Soil Sci. Soc.
Am. J. 43: 494 (1979a).
Terry, R.E.; Nelson, D.W. and Sommers, L.E. "Decomposition
of Anaerobically Digested Sewage Sludge as Affected by Soil
Environmental Conditions," J. Environ. Qual. 8: 342 (1979b).
Tester, C.F.; Sikora, L.J.; Taylor, J.M. and Parr, J.F.
"Decomposition of Sewage Sludge in Soil: I. Carbon and
Nitrogen Transformations," J. Environ. Qual. 6: 459 (1977).


105
range. For instance, when the temperature was reduced from
25C to 7.5C, the removal efficiency decreased by
approximately 80%. At 7.5C, only 20% removal efficiency
was observed. On the other hand, the decrease in H2S
removal in the higher temperature range was less marked than
observed at lower temperatures. For example, when the
temperature was increased from 50C to 100C, the H2S
removal efficiency decreased from 97.4% to 40%. The optimum
temperature range determined from these studies is between
30C and 40C, which is the optimum temperature range for
mesophilic bacteria. The removal rates of H2S at high
temperature is probably due to chemical oxidation reactions
in addition to biological oxidation.
Poor performance of a biofilter at low temperatures may
limit their application in cold climates, especially during
the winter. Proper means should be taken to avoid operating
biofilter systems below 10C. For larger biofilters, the
bed temperature can be a few degrees higher than the ambient
air temperature due to biological respiration of the
microbes and the exothermal oxidation reactions in the
filter. Kampbell et al. (1987) reported that soil bed
biofilter functioned well at temperatures in the range 12C
to 24C in Wisconsin. In another study carried out by Rands
et al. (1981) the filter bed temperatures were found to be
10 to 20C higher than ambient air temperatures during
winter times. This type of thermal enhancement, however,
was not observed during this study, probably because the


35
the air stream. Two water scrubbers are used in series to
ensure that the air stream is completely free of alkali and
to resaturate the air with water vapor in order to keep the
water content of the composts constant.
The resaturated, C02-free air stream is then forced
to the manifold where it is split into 15 streams. Each
stream goes into one incubation-absorption unit. Fifty
grams of compost sample is put in each incubation bottle.
The CO2 evolved from each of the compost samples is
collected in two 25-mL, 0.5N NaOH collectors in series. The
total air flow rate is controlled by a needle valve located
in front of the manifold. Syringe needles are used as flow
regulators to equalize the air flow through the 15
incubation units. The air flow rate through each unit is
adjusted to about 152 mL/min. The incubation system is
continuously operated at constant temperature (232C).
After flushing the residual air from the incubation
bottles, the outlet tube of each bottle is attached to the
CO2 collectors. C02 collectors are replaced with fresh
solutions periodically during the incubation period. The
system is leak checked before the incubation. Evolved C02 is
efficiently trapped by two absorption collectors in series.
Tests have shown that the first tube absorbed more than 95%
of the total C02 evolved.
C02 evolution is measured as described by Stotzky
(1965) with minor modifications. After C02 absorption, the
solutions in the two collectors of each unit are mixed and


C0 Evolved (mg-CC^/g of C added)
46
Figure 3-5. Effect of H2S exposure on the rate of
compost decomposition as measured by
C02 respiration.


183
University of Florida to address these problems and the
results are discussed in detail in Chapter 4 of this
dissertation.


160
was installed in the summer and made operational in the fall
of 1988.
System Design and Construction
The biofilter control system consists of an air
collecting system, an anti-corrosive blower and ductware, a
humidifier, a central shaft for flow adjustment and a two-
bed SIEBO-stone air distribution system, as shown in Figure
6-1.
The plant's malodor source, the grit chamber, was
covered and sealed as tight as possible (Figure 6-2). A
negative static pressure inside the grit chamber is
maintained by a blower, which prevents possible leakage of
malodorous air. The air collected by the blower through
0.30 meters diameter ducts is forced to a humidifier, where
water is dispersed by several spray nozzles to saturate the
waste gas stream. All pipes in the waste gas collection and
transmission system are made from PVC to prevent corrosion.
The biofilter bed gas distribution system is
constructed from the German patented SIEBO-stones. This
interlocking sinter block base unit of the biofilter system
is rigid and allows for heavy vehicles to be driven on it
without damaging the system. The SIEBO-stone base system
provides for an even distribution of the inlet air to the
filter bed as well as functioning as a drainage system. The
filter area is 100 square meters (m2) and is divided into
two equal sections, each of which can be operated and


156
ratio.
In practice, the water-compost contact time can be
easily controlled by the water spray rate. The selection of
ideal water to compost ratio, however, depends on the pH,
sulfur content of the compost and frequency of the
treatment. High water to compost ratios favor the leaching
of sulfate but more water is required and the maintenance
cost is consequently increased. The wash water is highly
acidic and has high concentrations of sulfate. Proper
treatment of this water is required.
The water to compost weight ratio and contact time used
for this study are 1:1 and 10 minutes, respectively.
Efficiency of elimination of sulfate in filter compost for a
single wash under this condition is shown in Table 5-4. The
average sulfate elimination efficiency is 36.5% with some
variations at each location in the bed. Sulfate
concentrations in the wash water increased from 0 to 5.25
mg-S/g, which indicates that the sulfate sulfur has been
efficiently transferred from compost to wash water.


101
decreased with decreasing acidity up to pH 5 (82.3%), and
then increased with further decrease in acidity of the
compost. As discussed previously in Chapter 2, sulfur
oxidizing bacteria can live in environments with a wide pH
range (1-8) depending on the species present. Probably the
dominant species present in the biofilter systems studied
are acidophiles which prefer an optimum pH value around 3.
For the higher pH value range, chemical reaction between H2S
and the compost material can significantly enhance its
removal in addition to biological oxidation. As a result of
this dual action, higher removal efficiencies of H2S can be
expected.
The acidity of the compost is traditionally expressed
as 'compost pH' for convenience. A more meaningful and
accurate way to express the compost acidity is the 'specific
acidity of compost (SAC)', which gives the quantity of H+
per unit weight of dry compost, /g-H+/g. Specific acidity of
compost can be calculated from compost pH by the following
equation:
SAC = W/[Cw(100-CWC)]X105-PH (4-1)
where:
SAC = specific acidity of compost, /xg-H+/g-dry compost
W = DI water used for compost pH analysis, mL
Cw = Weight of wet compost used for pH analysis, g
CWC = Compost water content, Wt%
pH = compost pH.
For example, if 2g of wet compost is taken and 20 mL of
DI water is used in H+ extraction for a compost pH analysis,


176
6-2). Influent air measurements indicated that the relative
humidity of the waste gas provided to the biofilter bed
after humidification is at least 95%. No other water was
introduced to the filter bed except that due to rain.
No significant changes in the bulk density and particle
size distribution of the compost were observed during the 27
months of operation (from 11/88 to 2/91). The organic
matter content decreased by about 4.6%, presumably due to
mineralization of the compost. Total nitrogen and total
carbon contents of the compost decreased at different rates
resulting in an increase of the C/N ratio from 11.4 to 23.2.
The most significant changes observed in the compost
during extended operation were the total sulfur content and
pH. The total sulfur, expressed as mg-S/g-compost on a dry
basis, increased considerably from 7.3 to 109. After
prolonged operation, parts of the biofilter bed showed a
pronounced color change in the compost from dark brown to
yellowish white, which was accompanied by observation of a
rotten vegetable odor.
The pH of the compost bed material decreased
significantly from 8.6 to 1.8 as a result of the continuous
removal of H2S and corresponding formation of sulfuric acid
(H2S04) by the biological oxidation of H2S. Although this
acidification of the compost did not appear to have a direct
and noticeable effect on the overall H2S removal efficiency,
serious corrosion of the cement blocks in the retaining wall
and the SIEBO-stone base was observed. Corrosion of the


192
Forster, C.F. and Wase, D.A.J. "Biopossibilities: The Next
Few Years," in Environmental Biology. Forster, C.F. and
Wase, D.A.J. eds, Ellis Horwood Limited Publishers,
Chichester, England, pp.439-448., 1987.
Frechen, F.B.; Kettern, J.Y. "Reduction of Odorous Emissions
from a Hazardous Waste Landfill Site Using Biofiltration and
Other Techniques," Paper # 87-95A.3, presented at the 80th
Annual Meeting of APCA, New York, New York, June, 1987.
Freney, J.R. "Determination of Water-Soluble Sulfate in
Soils," Soil Science 86s 241 (1958).
Gale, P.M. "Decomposition of Organic Waste Products Under
Aerobic and Anaerobic Conditions," PhD Dissertation,
University of Arkansas, 1988.
Grant, W.D. and Long, P.E. Environmental Microbiology.
Halsted Press, New York, 1981.
Gilmour, J.T. ; Clark, M.D. and Sigua G.C. "Estimating Net
Nitrogen Mineralization from Carbon Dioxide Evolution," Soil
Sci. Soc. Am. J. 49: 1398 (1985).
Greyson, J. Carbon. Nitrogen, and Sulfur Pollutants and
their Determination in Air and Water. Marcel Dekker, Inc.,
New York and Basel, 1990.
Hack, P.J.F.M. and Habets, L.H.A. "Experience with Full-
Scale Biopaq U.A.S.B.-Plants Treating Various Types of
Effluent," in Environmental Technology. Proceedings of the
second European Conference on Environmental Technology.
Amsterdam. The Netherlands. June 22-26, 1987. De Waal,
K.J.A. and Van Den Brink, W.J., Eds, Martinus Nijhoff
Publishers, Boston, 1987.
Hartenstein, H.U. "Assessment and Redesign of an Existing
Biofiltration System," Master's Thesis, University of
Florida, Gainesville, Florida, 1987.
Hartenstein, H.U. and Allen, E.R. "Biofiltration, An Odor
Control Technology for a Wastewater Treatment Plant," Report
to Department of Public Works, City of Jacksonville,
Florida, 1986.
Hsieh, Y.P. and Yang, C.H. "Diffusion Methods for the
Determination of Reduced Inorganic Sulfur Species in
Sediments," Limnol. Oceanogr., 34: 1126 (1989).
International Process Systems, Inc. (IPS) "Odor Control,
Completing the Composting Process," 655 Winding Brook Drive,
Glastonbury, CT 06033, 1990.


152
Table 5-3. Performance of defective compost before and
after treatment.
At low H2
S Loading9
At High
H2S Loading13
Compost
Treatment
h2s
Removal
Eff. (%)
h2s
Elimination
(g-S/m3-hr)
h2s
Removal
Eff. (%)
h2s
Elimination
(g-S/m3-hr)
Untreated
(control)
20.6
7.51
17.0
12.0
Water
Washed
99.9+
36.4
96.6
68.6
NaOH
Washed
99.9
36.4
96.7
68.6
NaHCCU
Washed
99.9
36.4
94.9
67.4
a H2S inlet concentration: 437 ppm;
Gas loading rate: 14.4 m3/m-hr.
b H2S inlet concentration: 489 ppm;
Gas loading rate: 30.0 m3/m2-hr.


59
8
Gas Outlet
TG35
-E
oooooo/ooI
&
TGP34
g
TGP33
-E!
TGP32
-E
TGP31

..x xxxx
m
*: :
lili
111
$$PI
pi
: ;
;:i
I
P31
P30
T = Temperature
P = Pressure
G = Gas Sample
Figure 4-3
Schematic drawing of Tower #3.


71
of the H2S absorption solution. Even though the contents
of both traps are analyzed, more than 95% of the H2S is
consistently recovered in the first of the two series traps.
The material to be analyzed is added to the reaction flask
through the side neck. The system is purged with N2 at
bubbling rate of 1-2 bubbles per second in the ZnAc-NaAc
traps for 10 minutes before the introduction of additional
reagents. The materials are boiled for 1 hour, the traps
are removed and sulfide titrated.
The procedures and methods used for analysis of the
total sulfur and various organic and inorganic sulfur
compounds in compost are similar to those used for sulfur
analyses in peat, soil and sediments described by Zhabina
and Volkov (1978), Tabatabai (1982), and Wieder and Lang
(1985) Minor modifications were made for the compost
sulfur analyses in this research. The procedures are
illustrated in Figure 4-7.
All results are expressed as mg sulfur per gram of
compost on a dry basis (mg-S/g). Compost moisture content is
determined from a sub-sample by drying the compost at 70 C
to constant weight.
Compost samples are stored in plastic bottles and
refrigerated at 2-4 C before analysis. In most cases,
however, compost samples are analyzed immediately after
sampling.


12
established independent standards for H2S emissions. These
states, which include California and New Mexico (10 ppm),
and Ohio and Michigan (1670 ppm). California, Kentucky,
Minnesota, Montana, New Mexico, New York, North Dakota and
Pennsylvania also have air quality standards for H2S. The
standards vary from 0.003 ppm for New Mexico to 0.1 ppm for
Pennsylvania, whereas and most of the other states specify a
standard of 0.03 ppm (Urone, 1986).
Since H2S is a highly toxic air pollutant, H2S has been
identified by the USEPA as one of 190 air toxic compounds in
Title III of the 1990 Amendments to the Clean Air Act. In
view of the wide spread exposure to this pollutant, emission
and air quality standards for H2S are going to be set in
the near future by EPA.
Biofiltration as a Air Pollution Control Technology
s
Biological degradation is widely used for treatment of
liquid and, to a lesser extent, solid wastes, but has
received little attention as a means of controlling
emissions of industrial gaseous wastes. Biofiltration is a
relatively new technology for control of air pollutants, in
which the air contaminants from off-gas streams are
biologically removed in a solid biological reactor. While
it is a well established air pollution control technology
in European countries, biofiltration as an air pollution
control technology has received little attention and
application in the United States. Few environmental


92
Effect of H2S Loading Rate on Its Removal
One of the most important observations made in this
study is the relationship between H2S reduction and its
loading rate to the biofilter. It is very important and
necessary to introduce here the concept of H2S loading rate
and the maximum elimination capacity of the filter
material. The H2S loading rate is the amount of H2S that is
being introduced to the system per unit volume of the
packing material per unit time (g-S/m3-hr). The maximum
elimination capacity of a compost is the maximum H2S loading
rate that the compost can bear without inhibiting its
microbial activity, and is expressed in the same units as
those used for H2S loading rate. These two parameters
probably play central roles in biofilter design and system
operation.
The maximum H2S elimination capacity of a compost
depends on the microbial population and activity of sulfur
oxidizing bacteria existing in the compost. The latter, in
turn, are related to the operating conditions of the system,
such as temperature, water content, acidity (pH) of compost,
and the concentrations of nutrients and inhibitory
substances.
Overloading of the biofilter system with H2S is
indicated by the appearance of a finely divided, yellowish-
white colored substance on the compost, a sudden decrease in
the H2S removal efficiency or the occurrence of higher
concentrations of elemental sulfur in the compost.


Regulator
Manifold
Row Meter
Row
Regulator
I
]
Dead 4N DI H2S Gas
Volume NaOH Water Mixer
100g
Compost
200+25ml
1N ZnAc
*
2x25ml Dead
0.5N NaOH Volume
Figure 3-2.
Schematic drawing of the experimental arrangement for the
investigation of the effect of H2S exposure on compost
decomposition.


H2S Concentration (ppmv)
172
Days After Start of Biofilter Bed Operation
Figure 6-5. Off-gas sampling locations on the
biofilter beds and concentrations of
hydrogen sulfide observed as a function
of biofilter operating time.


174
for off-gas samples (low concentrations) stored over a
similar observation period. Since the gas samples are
analyzed within 8 hours after sampling, the concentration
changes in the Tediar bags for both influent and off-gas
samples are considered to be negligible, when estimating
control efficiencies.
The effect of sample purging time for the collection
chamber on concentration measurements of effluent gas from
the biofilter was determined as follows: Off-gas samples
are collected at 0, 5, 15 and 20 minutes after placing the
sampling cover on the compost bed and purging. Each sample
is collected for 5 minutes at a flow rate of 1 Lpm. The
results are illustrated in Figure 6-7. These results
indicate that there is no significant difference in H2S
concentrations obtained for the different purging times
selected. The gas loading rate through the compost bed is
about 780 L/m2-min, and the sampling cover has a volume of
12 liters with a cross-sectional area of 0.06 m2. The dead
volume of the cover, therefore, can be replaced by the off
gas every 0.3 minutes. Because the collections of off-gas
samples from the biofilter are usually initiated 5 minutes
after the sampling chamber is placed at the sampling
locations, it is to be expected and has been demonstrated
that representative off-gas samples are collected from the
biofilter bed under these conditions.
The waste gas humidifier works effectively to keep the
compost bed within a moisture content range of 45-60% (Table


19
contain 1 billion bacteria, 10 million actinomycetes and
10,000 fungi per gram of soil (Bohn and Bohn, 1987). The
role of these microorganisms is to oxidize combined carbon,
nitrogen and sulfur to carbon dioxide, nitrogen and sulfate,
respectively, before the compounds leave the bed. The air
contaminants are, therefore, effectively removed from the
waste gas streams.
For good engineering design and environmental decision
making, it is essential to understand the mechanisms
involved and to reliably predict the kinetics of the
biological reactions taking place in these biofilter
systems. Many general kinetic models have been developed to
predict the behavior of bioreactions in a biological film,
none of these models, however, is specific enough to explain
the biodegradation of hydrogen sulfide in a biofilter
system.
Jennings et al. (1976) developed a mathematical model
to predict the percentage removal of a pure, non-adsorbable,
biodegradable substrate in a submerged biological filter
using the non-linear Monod expression for the substrate
utilization rate. In their model, the authors start from a
biological slime layer coating a spherical particle. The
slime layer is in turn surrounded by a liquid boundary
layer. They concluded that even at relatively high values
of influent substrate concentrations, the biological removal
of a single substrate follows first order kinetics.


62
The gas flow rates in Towers #1 and #2 are measured by
specially designed orifices. Two orifices for each tower
were designed and made, one for low flow ranges and the
other for high flow ranges. The orifices are made from
plastic plate and installed on the outlet gas pipe lines
(see Figure 4-1). The pressure drops across the orifices are
measured and the flow rates are calculated according to the
developed calibration equations.
Sampling Methods
Compost Samples
Compost samples in Towers #1 and #2 are taken from the
solid sampling ports shown in Figures 4-2 a and b. The
samples are taken at each port in a radial direction to the
tower walls so that a representative sample can be obtained
for that section. For composts not initially packed in
columns, the samples are taken after the compost has been
thoroughly mixed and very large particles ( diameter > 10
mm) have been eliminated.
Gas Samples
The inlet and outlet gas samples for each system are
obtained directly from the gas sampling ports by extraction
using gas-tight syringes. Gas samples extracted from
other locations along the towers are obtained by using a gas
sampling probe assembly. The gas sampling assembly, as


CHAPTER 2
BACKGROUND
Properties of Hydrogen Sulfide and Regulations
Physical and Chemical Properties
Hydrogen sulfide is a colorless gas that has a foul
rotten egg odor and is slightly heavier than air. Hydrogen
sulfide is moderately soluble in water. The solubility of
H2S decreases with increasing temperatures. Figure 2-1
shows the solubility of H2S as a function of temperature.
Dissolved H2S dissociates in accordance with the
following reversible ionization reactions:
H2S HS + H+ (2-1)
HS S2_ + H+ (2-2)
The distribution of the above species as a function of
pH is shown in Figure 2-2. It is apparent from Figure 2-2
that the concentration of HS- species is insignificant when
pH values are less than 6. The latter condition is normal
in a biofilter system for control of H2S. S2-, on the other
hand, may not occur at all.
Hydrogen sulfide can serve as a reducing agent,
reacting with sulfuric acid (H2S04) to form sulfur dioxide
(S02) and elemental sulfur (S) (Greyson, 1990) :
6


28
materials and has been involved in most applications (Don,
1985; Eitner, 1989), since it provides favorable conditions
for supporting microbial populations as well as having
superior physical and chemical properties.
The properties of individual composts depend on the
materials from which they are derived and the composition of
the final product. The filter materials used in this
research were mainly yard waste compost and sewage sludge
compost or a combination of both. These composts were
obtained from different sources and used for different
purposes. A general description of the types of composts
and their sources are summarized in Table 3-1. The physical
and chemical analyses data for the composts listed in Table
3-1 are presented in the corresponding chapters where the
use of specific composts is discussed.
Decomposition of Composts under Aerobic Conditions
A number of investigations have been carried out to
study the decomposition of anaerobically digested sewage
sludges in soils (Miller, 1974; Tester et al., 1977; Terry
et al., 1979b; Sweeney and Graetz, 1988; Gale, 1988).
Decomposition of fresh and anaerobically digested plant
biomass in soil is also reported by Moorhead et al. (1987).
Only limited information, however, is available concerning
the decomposition of compost. Tester et al. (1977, 1979)
stated that the decomposition of compost in soil is not only
related to the physical and chemical properties of the


53
Orifice Vent
Figure 4-1. Schematic drawing
system.
of the dual tower


187
properties provide good buffering capacity to various
operational impacts. H2S removal efficiencies of
99.9+% were routinely achieved in both the laboratory
and full scale operations.
10. The decomposition rates of composts are described by
multi-stage first-order kinetics. The decomposition
rate as well as biological activity of the compost are
significantly enhanced by the presence of H2S. First-
order reaction coefficients were determined which can
be used to quantitatively predict the useful life of
the compost.
11. Acidification and accumulation of sulfur, especially
sulfate, is a natural feature of the H2S oxidation
process. Continuous formation of H2S04 results in
significant decline in pH and serious corrosion of the
construction materials. A protective washing
procedure was developed to mitigate this feature and
keep the system operational at its optimized
conditions. Water is determined to be the best choice
for elimination of sulfate. NaHC03 solution is
recommended for pH corrections.
12. System upset is identified by compost dry-out, high
sulfur content in compost accompanied by a yellowish-
white deposit, and extremely low pH (<2) values of
compost. Specific procedures have been developed to
recover the activity of the defective filter material.


88
#12, and the more densely packed material shows a much
higher pressure drop for the same gas velocity. Composts
#14 and #17A have similar pressure drop-gas velocity curves
and although material #17A contains a larger fraction of
small particles than does material #14 (see Table 4-3), the
small particles in compost #17A are mainly sand and grass
fractions, whereas those in compost #14 are sewage sludge
particles, which tend to adhere to each other.
Effect of Gas Retention Time on PUS Removal
The effect of gas retention time on H2S reduction is
studied by varying the gas flow rate through the tower. The
results are presented in Table 4-4. In the first data set
of 6 tests, the H2S loading rate is kept approximately
constant at a low flow range to ensure that the maximum H2S
elimination capacity of the system is not exceeded during
the test. It is clear that there is no apparent effect on
H2S removal as long as retention times are longer than about
23 sec. When the retention time is reduced to 7 seconds,
the H2S removal efficiency decreases by about 6%. This
decrease in H2S is controlled by the macrokinetics of
biofiltration process. Sublette and Sylvester (1987)
reported that H2S can be metabolized by a pure culture of T.
denitrificans within 1-2 seconds. This suggests that the
reduction of H2S removal efficiencies under shorter
residence times is not necessarily due to insufficient
reaction time between the H2S molecules and the biomass, but


97
solvent. Insufficient water supply can limit the activity
of the microorganisms, which in turn reduces the H2S
oxidation rate. It is also possible that when the CWC
reduced below 30%, there is no free water existing in the
pours of the compost particles. This may decrease the rate
of transfer of H2S from the waste gas to the biofilm where
the biological oxidation of H2S takes place.
Biological activity of the "dry" compost can be
recovered if water is supplied to the compost to bring the
CWC to a proper range. Two composts, WCF4 and WCE3 with
original CWCs of 14.3% and 21.4%, respectively, were used to
study this phenomenon. DI water was added to bring the
water contents of composts WCF4 and WCE3 to 56.5% and 50%,
respectively. The biological activity of both composts as
indicated by H2S removal was recovered eventually up to
99.9+%. The time required for recovery of the activity,
however, is inversely proportional to the dryness of the
"dry" compost. As shown in Figure 4-13, it takes 63 hours
for compost WCF4 to recover its H2S removal efficiency to
99.9+% while compost WCE3 took 41 hours to reach the same
level even though both systems followed a similar recovery
pattern.
Effect of Compost Acidity on PUS Removal
The effect of compost acidity on H2S removal was
investigated using the Column System #4. The source of the
compost is the same as described in the previous section.


49
to be 2.62 and 5.21 mg-C02/g-OM, respectively, by equation
3-6. The ratio of these two values is equal to the ratio of
the first-order reaction coefficients for the reactions at
these two conditions,
C02,1000/C02,0 = k1000/k0 = 1*99*
in other words, the decomposition rate for the compost
exposed to 1000 ppmv H2S is 1.99 times of that for the
compost not exposed to H2S. The half life times for compost
#6 at both conditions are:
to.5,0 = *693/0.00031 = 6.12 (years), and
tQ ^ 1000 = 6.12/1.99 = 3.08 (years)
and for compost #3 are:
*"0.5,0 = 0*693/0.00057 = 3.33 (years), and
^0.5,1000 = 3.33/1.99 = 1.67 (years).
No studies of similar effects have been reported in the
existing literature. Thus the results obtained in this study
can not be compared with the results of other
investigations. Taylor et al.(1978) found that the highest S
mineralization rates are observed during the period of
highest C02 evolution when 2 to 6% of sewage sludge compost
is incubated in soils. Their results and those reported here
suggest that the microbial activity of the compost was
significantly enhanced by the addition of H2S, especially
the activity of the sulfur oxidizing bacteria.
Oxidation of inorganic sulfur compounds is a basic
phenomenon in nature. A number of bacteria have been


94
Figure 4-11. Determination of maximum H2S
elimination capacity of compost.


Sulfate Leaching Efficiency (%)
154
Figure 5-6
Effect of water-compost contact time on
sulfate leaching efficiency.


195
Pomeroy, R.D."Biological Treatment of Odorous Air", J. of
the Water Pollution Control Federation, 54(12): 1541 (1982).
Prokop, W.H. and Bohn H.L. "Soil Bed System for Control of
Rendering Plant Odors", Journal of Air Pollution Control
Association, 35: 1332 (1985).
Rands, M.B.; Cooper, D.E.; Woo, C.P.; Flether, G.C. and
Rolfe K.A. "Compost Filters for H^S Removal from Anaerobic
Digestion and Rendering Exhausts," Journal of Water
Pollution Control Federation, 53: 185 (1981).
Reddy, K.R.; Khaleel, R. and Overcash M.R. "Carbon
Transformations in the Land Areas Receiving Organic Wastes
in Relation to Nonpoint Source Pollution: A Conceptual
Model," J. Environ. Qual. 9: 434 (1980).
Rittmann, B.E. and McCarty, P.L. "Variable-Order Model of
Bacterial-Film Kinetics," Journal of the Environmental
Engineering Division, Proceedings of the American Society of
Civil Engineers, 104: EE5, 889 (1978).
Robarge, W.P. and Fernandez I. "Quality Assurance Methods
Manual for Laboratory Analytical Techniques," Prepared for
the USEPA and USDA Forest Service Forest Response Program.
Department of Soil Science, North Carolina State University.
Raleigh NC 27695, and Department of Plant and Soil Science,
University of Maine-Orono, Orono, ME 04469, July, 1986.
Rotman, A. "Use of Biofilter in Odor Control," Unpublished
Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 Rene-
Levesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991
a.
Rotman, A. "Biofilter: Traditional Design, Summary of
Technical Data as Based on German Projects," Unpublished
Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 Rene-
Levesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991
b.
Roy, A.B. and Trudinger, P.A. The Biochemistry of Inorganic
Compounds of Sulfur. University Press, Cambridge, 1970.
Sawyer, C.N. and McCarty, P.L. Chemistry for Sanitary
Engineers. McGraw-Hill, New York, NY, 1967.
Schmidt, S.K.; Simkins, S; Alexander, M. "Models for the
Kinetics of Biodegradation of Organic Compounds not
Supporting Growth," Appl. Environ. Microbiol.50: 323 (1985).
Sikora, L. and Sowers M.A. "Effect of Temperature Control on
the Composting Process," J. Environ. Qual. 14: 434 (1985).


50
identified in soils and other environments that are capable
of oxidizing organic and inorganic sulfur compounds (Roy and
Trudinger, 1970). A high population of the oxidizing
bacteria can be expected in the composts tested here. With
sufficient H2S supply, the bioactivity of the sulfur
oxidizing bacteria can be stimulated to result in an
increase of microbial population and a corresponding
increase in the evolution of C02. Hydrogen sulfide is
finally oxidized to sulfate through various pathways and
intermediate stages ( Roy and Trudinger, 1970; Brock and
Madigan, 1988; Yang and Allen, 1991; Allen and Yang, 1991).
After the 24 hours reaction period, the color of the
compost changed from originally brown to yellowish-white,
especially at high H2S concentrations. This feature
indicates that a large amount of sulfur has accumulated in
the compost.
Conclusions
Among the various biofilter materials, compost is
frequently selected as a medium in applying biofiltration to
air pollution control due to its unique properties and
advantages. Knowledge of the characteristics of compost
decomposition are important for both prediction of biofilter
operation characteristics and degradation estimates, as well
as in deciding on the appropriate disposal treatment and
method for used compost. The studies described here indicate
that the decomposition rates of the composts tested are much


181
Table 6-5. Sulfur fractionation of a typical compost
sample in Kanapaha biofilter bed.
mg-S/g
(Wt%)
Total-S
110
100
Organic-S
22.3
20.3
C-bonded-S
21.1
19.3
Ester-S
1.20
1.10
Inorganic-S
87.2
79.7
FeS2~S
30.0
27.6
FeS-S
0.40
0.40
S-S
3.9
3.6
so42"-s
52.7
48.1
Water Soluble-S
20.9
19.1
P-extractable-S
9.9
9.0
Insoluble-S
21.9
20


Ill
equation 4-2 above, to express the rates of biological
reaction for each particular situation:
1) If the substrate concentration is very high, i.e.,
when Km << C, the rate expression approaches zero-order
kinetics in the substrate concentration:
-dC/dt = VmaxB = kQ (4-3)
where kQ is a zero-order rate coefficient. The integral form
of equation 4-3 becomes:
c = co V (4"4>
where CQ and C are the initial substrate concentration and
the substrate concentration at time t, respectively.
If CQ C is plotted against t, a straight line should
be obtained and the slope kQ is the maximum elimination
capacity of the microbes for the substrate.
2) If the substrate concentration is very low, i.e.,
when C << 1^, a first-order kinetics dependence should be
obtained:
-dC/dt = k-j^C (4-5)
The integral form for equation 4-5 becomes:
C = CQ exp(-k1t) (4-6)
where k-j^ is the first-order reaction rate coefficient.
If ln(0/Co) is plotted against t, a straight line
should be obtained and the first-order reaction coefficient,


57
1100/1000 mmH20 (43/40 inches) of water column.
Humidification of the inlet air is achieved by atomizing
water in the spray chamber, through which the room air
passes. In addition, Pall rings are stacked in the spray
chamber for extending wetted surface area providing better
humidification. As a result relative humidities in the
range 95 to 100% were routinely and continuously achieved.
Gaseous H2S with a purity of 99+%, which is stored in
liquid form under pressure in a cylinder, is continuously
leaked and mixed with the prehumidified air in the inlet
lines (PVC pipe) to the towers. Plastic screen packing is
placed downstream from the H2S introduction point for better
mixing. Flow rates of air and H2S are controlled by plastic
valves, which are located on the carrier gas inlet lines,
and stainless steel needle valves, respectively. The flow
rates are measured on calibrated flow meters to obtain the
desired H2S concentration and gas flow through the towers.
Measurements of temperature, pressure and gas flow rate are
discussed in later sections.
A nozzle is installed on the top of Tower #1 in order
to introduce water, or other liquid solutions if necessary,
to the outlet end of the bed. Gas lines are made from PVC
pipes. The towers and pipes are connected by flanges for
convenient dismantling of the packed towers and compost
changes. Cork-rubber gaskets are used to seal the flanges.


122
Thus, it appears that equation 4-12 can be used to
predict the kinetics of H2S oxidation in the intermediate
H2S concentration ranges.
A set of experimental data (H2S =309 ppmv) are plotted
in Figure 4-22, where the symbols used are experimental data
points, and the solid curve is drawn according to equation
4-12. It can be seen that the experimental data are
adequately described by the diffusion limitation model.
The data shown in Figures 4-19, 4-20, and 4-22 are
plotted in Figure 4-23 in the "Ottengraf-form", i.e., C/CQ
on the y-axis and h/H on the x-axis, where h is the height
of sampling location on the filter and H is the total height
of the filter. It is informative to use this plot in order
to observe the H2S concentration profile in the biofilter.
According to Ottengraf, under reaction limiting
conditions, a straight line should be obtained for this type
of plot. However, the profile is no longer linear if the
substrate elimination rate is controlled partially or
completely by the diffusion rate in the biofilter. The
latter feature is clearly observed in this study of H2S
elimination. The data that are plotted in Figure 4-20 shows
a straight line in Figure 4-23. This linear relation is an
indication of zero-order kinetics and reaction limiting
control by the microbial population. When inlet H2S
concentrations are less than 434 ppmv (the middle curve) ,
the profile is curved (concave) below the straight line,
indicating that the reaction is partially diffusion


167
capacity for a few weeks. Because it was not possible to
acquire the same compost for the other half of the biofilter
system, the Kanapaha wastewater treatment plant staff
decided to mix the existing Pompano Beach compost with
compost obtained from the Buckman wastewater treatment plant
in Jacksonville, Florida and in-house compost. This
procedure was necessary to obtain a sufficient quantity of
compost to completely fill the entire biofilter bed system
to a depth of 1.3 meters. The final filter material used is
a mixture of yard waste compost, pine bark and sewage
sludge. Lime was applied to the compost material prior to
installation to buffer the bed acidity (pH). This mixed
compost (Compost #12) was then used to completely fill both
biofilter beds on November 20, 1988. Analytical results for
the composition of the original Pompano Beach compost
(5/10/88) and for the final compost mixture (11/21/88) used
are presented in Table 6-2.
The biofilter system was brought to full operation on
November 21, 1988. The fully operational system is shown
photographically in Figure 6-4.
Influent and off-gas air samples were collected and
analyzed during the first 16 days that the biofilter was
made fully operational (see Table 6-3). During this start
up period, H2S concentrations in the influent gas stream
varied between 156 ppmv and 229 ppmv, and the average off
gas H2S concentrations were observed to be in the range of
0.05 to 0.4 ppmv. There were, however, some minor


164
Table 6-1. Summary of Kanapaha biofilter bed
design and operation parameters.
Total Flow:
Filter Area:
Filter Height:
Gas Loading Rate:
Retention Time:
Temperature:
Pressure Drop:
79 96 acmm
100 m2
1.3 meter
47 57 m3/m2-hr
82 100 sec.
15 30 C
150 200 mmH20


4-4 Schematic drawing of column system #4 60
4-5 Schematic drawing of the gas sampling assembly ... 64
4-6 Photograph of the sulfur distillation assembly ... 70
4-7 Flow chart of the sulfur analysis procedures
for compost 72
4-8 Pressure drop as a function of particle size
range for different gas velocities 84
4-9 Pressure drop as a function of packing height
for different compost particle size range 86
4-10 Pressure drop as a function of gas velocity for
different types of compost 87
4-11 Determination of maximum H2S elimination
capacity of compost 94
4-12 Effect of compost water content on H2S removal
efficiency 96
4-13 Time required for dried compost to recover
optimum efficiency 98
4-14 Effect of compost pH on H2S removal efficiency.
Condition a: H2S loading rate: 10.5 g/m3-hr
Gas loading rate: 15 m3/m2-hr
Condition b: H2S loading rate: 35.4 g/m3-hr
Gas loading rate: 26.1 nr/m-hr... 100
4-15 Schematic drawing of the experimental arrangement
for investigation of the effect of temperature
on H2S removal efficiency 103
4-16 Effect of temperature on H2S removal efficiency .. 104
4-17 Effect of sulfate on H2S removal efficiency 107
4-18 Effect of nutrient addition on H2S removal.
Total-S content in compost (mg-S/g) A: 17.5;
B: 33.7; C: 20.2; D: 119.7 110
4-19 Linear least squares regression analysis for zero-
order kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 m3/m-hr, compost #17 ... 115
4-20 Linear least squares regression analysis for first-
order kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 m3/m -hr, compost #17 ... 116
x


106
system is not large enough to maintain an adiabatic
condition.
Effect of Sulfate on HnS Removal
The effect of sulfate presence on H2S removal was
evaluated using Column System #4. A preconditioned stable
compost from Tower #2 was divided into 8 equal-weight
portions. Two of the sub-composts with an original sulfate
content of 4.6 mg-S/g were packed directly into two columns
as controls. The other sub-compost portions were mixed with
sodium sulfate (Na2S04) to bring the compost sulfate content
to 24.6, 44.6, 64.6, 84.6, 104.6 and 204.6 mg-S/g,
respectively, and packed in the other 6 columns. After a
few days for acclimation, H2S removal efficiencies for each
column were determined. The system was operated at a gas
loading rate of 15 m3/m2-hr and an H2S loading rate between
6.6 and 8.4 g-S/m3-hr. The results are illustrated in
Figure 4-17.
No effect is observed when the compost sulfate content
is less than 25 mg-S/g. However, a significant effect is
observed at higher sulfate levels. The H2S removal
efficiencies were reduced from 99.94% to about 35% and
remained in the lower removal efficiency range when the
sulfate content was increased from 45 to 200 mg-S/g. These
results suggest that a sulfate content of 25 mg-S/g is a
critical level for the microbial environment. Above this
level sulfate probably reaches a toxic level and the


178
Table 6-4. Gas sampling
biofilter bed
and analysis
, 2/5/91.
for Kanapaha
Sampling
h2s
h2s
Location
Concen.
Removal
(ppmv)
(%)
Inlet
136

1
133
2.57
2
81.3
40.3
3
134
1.40
4
0.04
99.9+
5
0.04
99.9+


171
differences in concentrations at the four sampling locations
(see Figure 6-5). The similarity in the low H2S
concentrations simultaneously measured at all four off-gas
sampling locations indicate that the incoming gas is evenly
distributed across the filter bed. The final biofilter
control system, which used mixed compost, functioned
effectively immediately upon operation with a very high H2S
removal efficiency (99.8%). No initial period of reduced
efficiency or acclimation was observed. This unique feature
is probably due to the fact that half of the compost had
been previously exposed to H2S laden air for a few weeks
prior to use in the full scale operation. The average
efficiency of the biofilter in removing hydrogen sulfide
during the initial study period was determined to be greater
than 99.8%.
Several additional tests were made in order to confirm
the validity of the H2S removal efficiencies achieved by the
biofilter system. These tests included, determination of
the decomposition of hydrogen sulfide in the Tediar sampling
bags and observations of the effect of varying sample
equilibration time in the gas sampling cover prior to
sampling. Results of these tests are presented in Figure 6-
6.
To a certain degree, the concentrations of H2S for the
stored influent gas samples (high concentration) are
observed to decrease gradually over a few days. However,
there was no significant change in the H2S concentrations


65
In most cases, the H2S concentrations in the outlet
gases are so low that the gas samples can be analyzed
directly without any dilution. In the latter cases, Tediar
bags are directly connected to the sampling ports. Gas
samples are forced into the bags as a result of the positive
pressure of gas within the tower system. Also, Teflon
filters are replaced by glass wool plugs to reduce the
resistance to flow.
Each time after use, the Tediar bags are purged at
least three times with N2 to eliminate residual gas and
vapor. The stability of gas samples in the Tediar bags are
discussed in Chapter 6.
Water Samples
Water samples analyzed are mainly biofilter wash waters
from the tower drain outlets. When washing a packed tower,
the entire wash water is collected in a container. Water
samples are obtained from the container after mixing the
wash water with a stirrer for a few minutes.
Compost Analysis Methods
Water Content
Two to five grams of wet compost are dried in an
aluminum tray at an oven temperature of 70 C until constant
weight is obtained. Compost water content is determined by
the difference in weight between the wet and dry composts
(Robarge and Fernandez, 1986).


69
volatile sulfur ( sulfide-S), water soluble sulfur (soluble
sulfate), insoluble sulfate (S042-), elemental sulfur (S),
Pyrite sulfur, ester sulfur (organic-S) and total sulfur.
Each of the wet chemical procedures involved the
reduction of S to H2S in a Johnson-Nishita apparatus
(Johnson and Nishita, 1952) and trapping the evolved H2S in
zinc acetate-sodium acetate (ZnAc-NaAc) solutions. Trapped
sulfide is quantified by iodometric titration (APHA, 1989)
with a 0.025N iodine solution and 0.025N Na2S203 titrant.
The distillation apparatus incorporated slight
modifications of that used by Johnson and Nishita and is
similar to that used by Wieder and Lang (1985). Figure 4-6
shows the distillation assembly. The reaction flask is a
250-mL, round-bottom, three-neck flask, with an N2 inlet via
a bleed tube inserted in one neck and the central neck is
connected to a condenser. Ultra high purity (>99.999%)
nitrogen is used to sweep out the H2S and to maintain the
reaction flask in a reducing environment. The third neck of
the flask is fitted with a stopper to allow introduction of
liquid solutions to the flask. The ZnAc-NaAc solution is
made by dissolving 50 g of zinc acetate dihydrate
[Zn(CH3COO)2.2H20] and 12.5 g of sodium acetate trihydrate
(CH3C00Na. 3H20) in 800 mL of DI water and adjusting the
final volume to 1 liter (Tabatabai, 1982). Twenty-five (25)
mL of this solution is mixed with 100 mL of DI water and
this solution is used to fill two traps used in series. The
first trap contains 100 mL and the second one contains 25 mL


163
controlled individually. During normal operation, each
section treats one-half of the total exhaust gas flow
volume. The total flow can be diverted to one section if
necessary, for example when repairs are needed, without
reducing overall pollutant removal efficiency.
The actual waste air flow rate through the biofilter is
in the range from 79 to 96 actual cubic meter per minute,
which gives an average surface loading rate of 52 m3/m2-h
and a biofilter bed treated gas retention time of 88
seconds. Biofilter design and operation parameters are
summarized in Table 6-1.
Sampling and Analysis Methods
To monitor biofilter bed performance, inlet and off-gas
samples are collected and analyzed for H2S concentrations.
The influent gas samples are taken from the inlet pipes in
the central shaft and collected in Tediar bags. These
samples are later quantitatively diluted with pure nitrogen
(N2) and analyzed. Off-gas samples from the biofilter are
collected by a specially designed gas sampling system
illustrated in Figure 6-3. The plastic collector cover, 275
mm in diameter and 220 mm in height, is open at the bottom
and has five 10 mm holes drilled in the top to allow the
flowing off-gas to purge the collector and escape. A Teflon
sampling probe is inserted from the top center of the cover
and extends to within 100 mm from the open base. This
design ensures representative sampling and avoids any


116
d>
O
5
Figure 4-20. Linear least squares regression
analysis for first-order kinetics of
H2S oxidation in biofilter. Gas loading
rate: 224 m3/m-hr, compost #17.


6-2 Summary of periodic Kanapaha biofilter bed compost
analyses during operational period from
5/10/88 to 2/5/91 168
6-3 Summary of Kanapaha biofilter influent and
effluent gas sample analyses during
three week start-up period 170
6-4 Gas sampling and analysis for Kanapaha
biofilter bed, 2/5/91 178
6-5 Sulfur fractionation of a typical compost sample
in Kanapaha biofilter bed 181
viii


20
Another model developed by Rittmann and McCarty (1978)
is a variable-order model of bacterial-film kinetics which
incorporates liquid-layer mass transport, biofilm molecular
diffusion and Monod kinetics. These investigators concluded
that at low substrate concentrations, the reaction follows
first order kinetics, whereas at high concentrations the
reaction follows one-half order kinetics.
Based on their biophysical model (Figure 2-3),
Ottengraf and Van Den Oever (1983) have developed a
mathematical model to describe the kinetics of organic
compound removal from waste gases for a biofilter system.
The model was developed and tested using a soil bed for the
removal of toluene, butylacetate, ethylacetate and butanol.
From their experimental results, they concluded that all the
carbon sources investigated were eliminated according to a
zero order reaction, even at very low concentrations of the
substrates.
Kampbell et al. (1987) investigated the biodegradation
of propane, isobutane and n-butane by soil biofilter beds.
They suggested that at low concentrations the rate of
biodegradation was proportional to the concentration of the
organic compounds (first order reaction), and at higher
concentrations the rate becomes independent of the organic
compound concentration (zero order reaction). The
degradation kinetics appeared to follow a hyperbolic
function:


85
The relationships between pressure drop and packing
height at constant gas velocity for particle size ranges B
(3.33 to 12 mm) and E (<1.2 mm), and the parent compost #12
are shown in Figure 4-9. It is seen that the pressure drop
increases approximately linearly with packing height for all
three samples. It is important to note that the pressure
drop values obtained for the parent compost are intermediate
between the strong dependence of pressure drop on packing
height for small particles (class E) and the weak dependence
on packing height for larger particles (class B).
The dependence of pressure drop on compost water
content is not as consistent as that for particle size.
Qualitatively, sewage sludge treated compost contains more
viscous and adhesive small particles than are found in
untreated compost. Thus, when the water content of the
sewage sludge treated compost is increased, coagulation of
small particles is enhanced and the pressure drop increases
sharply. However, the build-up in pressure may be suddenly
released by channeling, i.e., a breakdown of the overall
flow restriction by the formation of a channel of much less
resistance caused by a separation of packed materials.
The effects of gas velocity on pressure drop across
four typical biofilter materials are shown in Figure 4-10.
As expected, the pressure drop increases rapidly with
increasing gas velocity. The pressure drop depends on the
way the filter is packed. Composts #12 and #13 are similar
in nature but filter material #13 was more compacted than


159
conditions, inorganic sulfates and sulfites can be easily
reduced to hydrogen sulfide by various types of anaerobic
and facultative bacteria, such as sulfur-reducing bacteria
(SRB).
S042" + 2C + 2H20 SRB^ 2HC03 + H2S (5-1)
Because hydrogen sulfide is volatile and only partially
soluble in water, whenever domestic wastewater from sewer
lines is agitated and exposed to the atmosphere, H2S is
released to the air.
The Kanapaha wastewater treatment plant is the major
municipal sewage treatment facility ( 9 million gallons per
day) for the city of Gainesville, Florida. As is the case
for most city public utilities, this domestic wastewater
treatment facility has been identified as a source of
odorous emissions and has been the recipient of numerous
odor complaints for many years. The most significant source
of the malodor was identified as the plant's grit chamber,
where agitation of the incoming wastewater causes
considerable outgassing of hydrogen sulfide. Attempts to
control the emission of odors at the plant using chemical
treatment of the wastewater were only partially successful
and very expensive. The high costs of the chemicals used,
as well as additional problems encountered in the treatment
process prompted the management and staff of Gainesville
Regional Utilities (GRU) to look for alternative effective
odor control technologies. As a result, a biofilter system


82
Table 4-3. Particle size range distributions for
selected composts.
Particle Size
Range Distribution
(wt%)
Compost
ID #
A
B
C
D
E
12
20.0
22.5
10.0
13.1
34.4
14
27.7
26.9
8.10
11.6
25.7
13
21.4
24.5
6.70
15.8
31.6
17
0.00
33.4
14.4
22.5
29.7
A: diam. > 12 mm
B: 3.35 < diam. <12 mm
C: 2.36 < diam. <3.35 mm
D: 1.18 < diam. <2.36 mm
E: diam. <1.18 mm


109
4-18. For each compost set, H2S removal efficiencies are
plotted for the control compost and the test compost before
and after the addition of nutrient solution. It can be seen
that the H2S removal efficiencies are significantly
decreased for the three composts tested with low sulfur
content when the nutrients are added. The reason for this
decline in efficiency is not clear. No improvement of H2S
removal for the high sulfur containing compost was observed,
either.
Kinetics of H2S Oxidation in a Biofilter
Theoretical Considerations
In general, the substrate utilization rate of a
component by microbial flora as well as the enzymatic
reaction rate are expressed by the Michaelis-Menten
relationship (White et al., 1978? McGilvery and Goldstein,
1983; Schmidt, et al., 1985):
-dC/dt = VmaxCB/ (Kjjj + C) (4-2)
where C is the substrate concentration, B is the population
density, Vmax is the theoretical maximum specific reaction
rate, Km is the half-saturation constant (Michaelis
constant), and t is the reaction time.
Under steady state conditions, i.e., when the microbial
population does not change with time, there are three
situations which may be encountered in a biological reaction
system, and corresponding equations can be derived, from


108
activity of these microorganisms is markedly inhibited.
This observation is very important for biofiltration control
of H2S. Since sulfate is the final product of the
biofiltration process, it may accumulate in the biofilter
bed if no other action is taken. Accumulation of sulfate
can easily reach a level that can significantly reduce the
function of the biofilter. Measures to avoid sulfate
accumulation in filter and to enable recovery of the
deteriorated compost are discussed in Chapter 5.
Effect of Nutrient Addition on H2S Removal
Low and high sulfur-containing composts were used for
investigation of the effects of nutrient addition on H2S
removal. The total-S contents are 17.5, 33.7, 20.2 and
119.7 mg-S/g for compost A, B, C, and D, respectively. Each
compost was packed in two columns, one is used as a control,
and the other was treated with nutrients. Fifty (50) mL of
nutrient solution was mixed with 140 g of the compost tested
in each column. Excess water was removed by exposing the
compost to room air for 24 hours. The nutrient added is
similar to the enrichment medium for sulfur-oxidizing
bacteria suggested by Aaronson (1970). The composition of
the solution is: K2HP04, 1.0 g; MgS04.7H20, 0.5 g; NH4N03,
1.0 g; CaC03, 10 g, and DI water was added to bring the
final volume to 1000 mL.
The results on the effect of nutrient addition on
biofilter performance are presented as a bar graph in Figure


C/C
118
Figure 4-22. Plot showing the fractional-order
kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 nr/m -hr, compost
#17.


75
Elemental sulfur
The precipitate obtained above is dried with filter
papers as described by Zhabina and Volkov (1978) and is
extracted with analytical grade acetone. The volume of
acetone used (mL) is 40 times that of the equivalent dry
weight of the compost sample (g). The extraction flask is
covered with Parafilm and placed on a rotary shaker at a
speed of 140/min for 16 hours. The mixture is filtered and
rinsed with additional acetone. Either a fraction of or the
entire filtrate are subjected to Cr2+ reduction as described
by Zhabina and Volkov (1978).
Pvrite sulfur
The residue left after S extraction is subjected to
chromium reduction. The Cr2+ was produced by passing a 1 M
solution of CrCl3.6H20 in 0.5 M HC1 through a Jones reductor
column containing Zn amalgamated with Hg (Zhabina and
Volkov, 1978). Preparation of the Jones reductor is
described by Swift (1950) and Patterson and Thomas (1952).
Ten (10) mL of ethanol is added to the flask followed by 20
mL of 12N HC1 and 16 mL of 1M CrCl2 solution. Heat is
applied after 30 min, the H2S evolved is absorbed in the
traps and titrated as above.


Total-S (mg-S/g)
143
0.125
0.375 0.625 0.875
Height Above Bed Inlet (m)
Original
Figure 5-4. Total-S distribution profile in
biofilter, Tower #1, after exposure to
H2S for 100 days.


170
Table 6-3. Summary of Kanapaha biofilter influent and
effluent gas sample analyses during three
week start-up period.
Date
Avg. H2S
Influent
(ppmv)
Avg. H2S
Off-gas
(ppmv)
Removal
Efficiency
(%)
11/21/88
195
0.39
99.8
11/22/88
229
0.16
99.9
11/23/88
156
0.32
99.8
11/25/88
208
0.12
99.9
11/29/88
140
0.05
99.9+
12/06/88
175
0.11
99.9+
12/15/88
167
0.07
99.9+


22
colorless sulfur bacteria are believed to play the major
role and their ability to oxidize reduced inorganic sulfur
compounds has been clearly established (Roy and Trudinger,
1970; Kuenen, 1975; Brock and Madigan, 1988).
The oxidation of inorganic sulfur compounds is carried
out by a spectrum of sulfur-oxidizng organisms which include
1) obligately chemolithotrophic organisms, 2) mixotrophs,
3) chemolithotrophic heterotrophs, 4) heterotrophs which do
not gain energy from the oxidation of sulfur compounds but
benefit in other ways from this reaction, and 5)
heterotrophs which do not benefit from the oxidation of
sulfur compounds. Physiological characteristics of some
sulfur-oxidizing bacteria are summarized in Table 2-2.
Options for microbial metabolism of hydrogen sulfide
must employ one or more of the following metabolic pathways:
1) aerobic oxidation, 2) anaerobic oxidation, and 3)
photosynthetic dissimilation. Biofiltration of waste gases
is a process utilizing aerobic conditions in most cases. In
aerobic oxidation, sulfur-oxidizing bacteria oxidize H2S to
elemental sulfur or higher oxidation states using oxygen
(02) as an electron acceptor. The biological steps in the
oxidation of various sulfur compounds are summarized in
Figure 2-4.


78
Table 4-1. Retention times, limits of detection and
operating conditions for the Tracor 250H
Analyzer.
Item
h2s
Compounds3
MM
DMS
Retention
Time
(min)
1.15
2.16
4.02
Detection
Limit
(ppmv)
0.01
0.01
0.02
Operating Conditions:
Temperatures: Valve: 50 C
Column: 70 C
Detector: 110 C
Flow Rates:
Cylinder
Pressures:
Nitrogen:
Oxygen:
Hydrogen:
Sample:
80 mL/min
21 mL/min
80 mL/min
40 mL/min
(carrier gas)
(flame gas)
(flame gas)
Air 40 psig
Hydrogen: 53 psig
Oxygen: 40 psig
Nitrogen: 80 psig
for sampling valve
activation)
a
MM: methyl mercaptan.
DMS: dimethyl sulfide.


77
water samples, the reduction, H2S absorption and iodometric
titration procedures described previously are used.
Sulfide sulfur
Sulfide was measured according to APHA (1989). The
samples are pretreated to remove interfering substances and
to separate insoluble sulfide.
Total-S
Total-S was determined by using 2-5 mL of the aqueous
sample depends on its sulfur content. The sample is then
analyzed as described for the total sulfur in compost.
Sulfur in Waste Gas
Gas samples are analyzed by a commercial gas
chromatogragh equipped with a flame photometric detector
(GC/FPD), a Tracor Model 250H Analyzer. The detection limits
and operational conditions for the analyzer are summarized
in Table 4-1. Detailed information concerning the principles
and conditions of operation are reported elsewhere (Yang,
1988). In the early stages of this study, the component
peaks are recorded by a chart recorder (Texas Instruments,
Inc., Model Recti/Riter II) and concentrations of sulfur
compounds are determined by measuring the peak heights.
Later in the study (from August 1990 to September 1991) the
chart recorder was replaced by a Spectra-Physics model
SP4290 integrator and the measured concentrations are read


39
system was previously tested to obtain absorption
efficiencies of H2S and C02 in the ZnAc traps. The results
showed that the H2S absorption efficiency was greater than
99% and the C02 absorption was less than 2% for the ZnAc
solutions used.
Results and Discussion
Decomposition of Composts
The decomposed C evolved as C02 from the four composts
studied during the 122 day incubation period is shown in
Figure 3-3. The decomposition patterns of composts #2, #3
and #6 are somewhat similar. Decomposition of compost is
initially rapid, from 40 to 52% of the total C02 produced in
the 122 days is evolved in the first 42 days of incubation.
A total of 9.2, 5.7, 6.1, and 4.4% of the original C was
decomposed and released as C02 for composts #1, #2, #3, and
#6, respectively during the total 122 days of incubation.
It appears that decomposition rates of the composts
are inversely proportional to their age, in other words,
the older the compost, the slower the decomposition. All
except compost #1, showed decomposition rates which were
similar. Compost #1, however, was not completely composted
when used. Also the organic matter content of this compost
is higher than that for the others tested. Initial and
delayed higher decomposition rates for compost #1 suggest a
two stage incubation involving an initial 'conditioning'
step followed by a 'conditioned' decomposition. The compost


54
simultaneously and controlled separately. The biofilter bed
material is enclosed in transparent rigid Acrylic pipe, with
an inner diameter of 0.15 meters (6 inches) and a height of
1.34 meters (4 feet). Each vertically mounted pipe can be
packed with the desired compost up to a height of 1.2 meters
(3.9 feet). The packed biofilter material is supported by a
sieve plate to ensure a homogeneous distribution of the
inlet gas stream across the face of the bed. Non-
biodegradable plastic screens are placed between the sieve
plate and the biofilter material to avoid separation of
smaller compost particles.
Sampling and measurement ports are located along the
Acrylic column for compost and gas sampling, and pressure
and temperature measurements. The sampling and measurement
ports are shown in Figure 4-2a and b for Towers #1 and #2,
respectively. An individual sampling/measurement port is
identified by a letter-number system, where the letter
indicates the function and the number indicates the location
of the port. For example, TS11 means this port is used for
temperature measurement and solid sampling, and is located
on Tower #1 at location 1. All the other filter systems
with multi measurement/sampling ports are identified in the
same manner.
Room air is forced by a Gast Regenair Model R3105-1
air blower into the humidification chamber. The blower,
which is driven by a 1/2 HP motor, generates a maximum flow
of 1.5 m /min (53 cfm) and a maximum pressure/vacuum of


CQ, Evolved (mg-CC^ /g of C added)
48
Figure 3-6. Plot of C02 evolution as a function of
square root of H2S concentration.


CHAPTER 6
FULL SCALE APPLICATION OF BIOFILTRATION TO CONTROL
HYDROGEN SULFIDE EMISSIONS AT A WASTEWATER TREATMENT PLANT
A full scale biofilter bed system has been installed
and operated since the fall of 1988 to control H2S emissions
from the grit chamber at the Kanapaha Wastewater Treatment
Plant, Gainesville, Florida. In this chapter is describes
the design, construction and operation of this system.
Experience gained, as well as advantages and disadvantages
observed during operation of the system are discussed.
Introduction
Emissions of objectionable odors are a common problem
encountered at most wastewater treatment plants. The odorous
compounds frequently observed as volatile emissions from
these sources include hydrogen sulfide, ammonia, organo-
sulfur compounds and some volatile organic compounds (VOCs)
(WPCF, 1979; Yang, 1988). The origin of these odorous
chemicals is in the sewer lines where, due to existing
anaerobic conditions and excessively long residence times
for the incoming wastewater, the odorous compounds are
formed and confined.
The predominant odorous compound emitted from municipal
waste water treatment plants is H2S. Under anaerobic
158


H2S Concentration (ppmv)
175
Figure 6-7. Effect of varying purging time for
sample collection chamber prior to
sampling on measured hydrogen sulfide
concentrations.


2
(Bethea et al., 1973; Ferguson, 1975; USEPA, 1985; Lalazary
et al., 1986; Walker et al., 1986; Lindstrom, 1990).
Processes that have been used to remove H2S from waste gas
streams involve either physical treatment or chemical
oxidation. Some methods require addition of chemicals, and
energy expenditure is usually necessary for physical
treatment. Additional environmental problems are
encountered with chemical additions, where resulting
products and by-products require further treatment and
disposal.
Biofiltration can provide for a universal, simple,
economicly feasible, and efficient pollutant-destructive
control technology for a variety of toxic and hazardous
substances in waste gas streams. In recent years biological
filters have been developed and described which have the
potential to simply and effectively control odors, including
H2S emissions (Prokop and Bohn, 1985; Allen et al., 1987a;
Eitner and Gethke, 1987; Hartenstein, 1987 ). Deodorization
methods based upon the activity of microorganisms are
beginning to attract increasing attention in the U.S.
Although the biofiltration technique has been shown to
be an efficient, practical and simple gas cleaning
technology, which is increasingly being used around the
world, the design and operation parameters as well as the
microbial processes involved have not yet been very well
defined. In particular, little research has been directed at
the details of the biofiltration control of H2S. A


90
is possibly due to the slower step of H2S diffusion from the
gas phase into the liquid phase.
In the second data set of 5 tests presented in Table 4-
4, the biofilter inlet gas contains very high H2S
concentrations and the system is operated at lower flow
rates and, therefore, longer retention times. These tests
show that, even when the inlet gas contains an H2S
concentration of 2650 ppmv, the biofilter can successfully
reduce the concentration to 4.6 ppmv with a 99.8% removal
efficiency when the retention time of the flowing gas is
increased to 197 seconds. Thus, it can be seen that, as
long as the H2S loading rate does not exceed the maximum H2S
elimination capacity of the system (discussed in a later
section), then the design engineer or operator can always
deal with high H2S concentrations in the waste gas by
decreasing the gas flow rate to obtain the desired high
level of control efficiency.
Effect of Concentration of FUS on Its Removal
The effect of H2S concentration in the inlet gas on the
H2S removal efficiency has been investigated under constant
gas flow rate conditions. The results of these studies are
presented in Table 4-5. No significant difference in
control efficiencies is observed when H2S concentrations in
the influent gas stream are varied from 5.5 ppm to 518 ppm
as long as the H2S loading rate is less than the maximum
acceptable value for the compost studied.


61
(10 in), which provide a 240 mL packing volume. The ends of
the packed columns are plugged by rubber stoppers.
Thermometers are inserted into the columns to measure
temperatures. The gas flow rates are measured by a
rotameter at the gas outlets. Effluent gas samples are
taken from the outlet of the rotameter.
Measurement Methods
Periodic measurement of temperature, pressure drop and
gas flow rate in the biofilter systems are carried out by
the following devices.
Temperature
Temperature is measured by mercury in glass thermometers
with a range from -20 C to 110 C and a minimum scale
division of 1 C.
Pressure Drop
Pressure drop is measured by manometers with a minimum
reading of 1 millimeter water column (mmH20). In case the
pressure drop is greeter than 1000 mmH20, the pressure drop
is measured by mercury manometers with a minimum reading of
1 millimeter of mercury (mmHg).
Gas Flow Rate
All the gas flow rates except those of Towers #1 and #2
are measured by pre-calibrated rotameters.


139
major role in the oxidation of reduced sulfur. The
colorless sulfur bacteria are divided into three families,
the Thiobacteriaceae, the Beggiatoaceae, and the
Achromatiaceae. The genera, which belong to these families,
have been studied and include Thiobacterium. Macromonas.
Thiovulum. Thiospira. Thiobacillus. Thiomircospira.
Sulfolobus. Beqqiatoa. Thiospirillopisis. Thioploca.
Thiothrix, Thiodendron. and Achromatium (Kuenen, 1975).
The colorless sulfur bacteria are naturally occurring
almost everywhere on the earth. These bacteria live in a
wide pH range from 1 to 8, and temperatures up to 85C
(Kuenen, 1975; Brock and Madigan, 1988).
To distinguish between the active genera and species of
sulfur bacteria is not the goal of this study. The macro
processes of H2S oxidation, which involve physical, chemical
and biological processes are of greatest interest and
application in this study.
A number of oxidation reactions of inorganic sulfur
compounds which are effected by the colorless bacteria,
especially thiobacilli, have been reported (Starkey, 1966).
Some of the important reactions that possibly occur in a
biofilter system are as follows:
H2S + 202 H2S04 (5-1)
2H2S + 02 -* 2S + 2H20 (5-2)
2S + 302 + 2H20 2H2S04 (5-3)
Na2S203 + 202 + H20 - Na2S04 + H2S04 (5-4)
4Na2S203 + 02 + 2H20 2Na2S406 + 4NaOH (5-5)


67
Acid-extractable Cations
Two and half (2.5) g of finely-ground, oven-dried
sample is weighed into 50-mL centrifuge tubes. Twenty five
(25) mL of 1M HC1 is added and the tubes are shaken for 3
hours on a mechanical shaker. The compost suspensions are
centrifuged at 6000 rpm for 15 minutes and filtered through
Gelman 0.45 micrometer membrane filters. The solutions are
analyzed for Fe, Al, Ca, Mg, Cu and Mn on an Inductively
Coupled Argon Plasma Spectrometer (ICAP) (APHA, 1989).
Particle Size Distribution
The compost is dried in oven at 70C for 24 hours.
Particle size distribution by weight is measured by passing
the dried compost through a series of sieves (U.S.A.
Standard Testing Sieve, A.S.T.M. E-ll Specification, Fisher
Scientific Company) and weighing the residue.
Porosity
Compost porosity is determined according to Danielson
and Sutherland (1986).
Organic Matter
After determination of compost water content, the
samples are placed in a muffle furnace and baked for 24
hours at 450 C. Organic matter is determined by the loss-
on-ignition (LOI) (Robarge and Fernandez,
1986).


180
Figure 6-8. Compost samples taken from Kanapaha
Wastewater Treatment Plant biofilter
beds (2/5/91) .
Left: Sample taken from west bed. White
color indicates high sulfur
accumulation.
Right: Sample from east bed. Low
sulfur content compost, color is close
to the original (dark brown).


Pressure Drop (mmh^O)
86
Figure 4-9
Pressure drop as a function of packing
height for different compost particle
size ranges.


Compost pH
134
Cumulative Operation Time (Day)
a TS11 + TS12 O TS13 A TS14 x Wash Water
Figure 5-3. pH changes of compost in different
sections of the biofilter with
operation time,
a) Tower #1, compost #17A.


41
decomposition rates measured here are much slower than those
found for soils. Tester et al. (1977) reported that
approximately 16% of the compost C was evolved as C02 during
54 days of incubation, when 2 to 6% fresh sewage sludge
compost was incubated with soils. Miller (1974) reported
that 20% of added organic carbon is evolved as C02 for a 6
month incubation period under similar conditions. In another
investigation carried out by Moorhead et al. (1987) it was
observed that about 39 and 19% of the total-C for fresh and
digested low-N plant biomass, and 50 and 23% of fresh and
digested high-N plant biomass are released as C02 during 90
days decomposition when these biomasses are added to soils.
Fresh plant biomass evolves as much as twice the organic-C
as C02 when compared to corresponding digested biomass
sludges. These results of other researchers suggest that
the decomposition rate of organic matter strongly depends on
the source and the properties of the available organic
matter. Miller (1974), Sommers et al. (1976) and Terry et
al. (1979a, b) have concluded that sludge composition and
incubation conditions, rather than soil properties control
sludge decomposition.
Reddy et al. (1980) have shown that decomposition of
organic carbon depends on the nature and constitution of the
wastes. Low molecular weight (simple) compounds can be more
easily degraded by microorganisms than more complex organic
compounds. Organic-C components in decreasing order of
biodegradability are: (i) readily oxidizable soluble


186
5. The microorganisms in the biofilter system are
mesophiles. The optimum temperature range for these
organisms to be efficient is from 30 to 40 C. A H2S
removal efficiency of 50% can be achieved when the
temperature range is extended from 10 to 80 C.
6. High concentrations of sulfate are toxic to the
microbial flora. The critical level is between 30 and
40 mg-S/g dry compost. Above this range, the
biological activity of the microorganisms may be
significantly inhibited and result in reduction of the
H2S elimination capacity of the compost.
7. Variation in Compost pH showed no effect on H2S
removal efficiency for values greater than 2.0 pH
units. However, lower pH values for the filter bed
will cause serious corrosion problems.
8. Biofiltration is a comprehensive process which
involves physical, chemical and biological processes.
Ottengraf's model was adopted and has proved to be
successful in describing the macro kinetics of H2S
removal. Kinetic models and equations have been
developed and determined to be appropriate and
accurate in quantitatively describing removal of H2S
by the biofiltration process.
9. Composts from different sources have been
demonstrated to be excellent media for biofilters used
in H2S removal from waste gas streams. Their unique


151
NaC03
The systems were tested under two different operating
conditions, a) at a low H2S loading rate, 36 g-S/m3-hr and
b) at a higher H2S loading rate, 71 g-S/m3-hr. At the lower
loading rate, the H2S removal efficiency increased from
20.6% to 99.9+% for all three treated composts. The H2S
elimination capacities of these composts were determined at
high H2S loading rates. From Table 5-3 it can be seen that
the H2S elimination capacities for all the three treated
composts were increased by a factor of 9, from 7.5 to 68 g-
S/m3-hr.
The NaOH-treated compost looks darker and feels sticky,
probably the structure of the compost was altered by this
aggressive alkali treatment. From an economic point of
view, water is the best choice for the treatment. If
acidity of the compost needs to be corrected, then aqueous
NaHC03 solution is recommended as a treatment chemical.
The shaking-washing method used here is not feasible in
practice for full scale biofilters. The most feasible way
to treat a full scale bed is to spray water or the desired
solution onto the top of the filter. In this case, the
water to compost ratio and contact time become critical for
the effectiveness of the treatment.


76
Organic sulfur
Following the previous procedure, the insoluble residue
in the reaction flask is filtered through a #42 Whatman
filter and is washed repeatedly with acidified DI water to
remove chromium ions. The residue is then subjected to
reduction with the reducing mixture as described above. The
organic sulfur determined this way is mainly ester-S.
In addition to the literature mentioned above, similar
and dissimilar procedures for determining sulfur
constituents in sediments, peat and soil are also described
by Smittenberg et al. (1951), Freney (1958) Johnson and
Henderson (1979), David et al. (1982), and Hsieh and Yang
(1989). An excellent comparison of some of these methods is
reported by Wieder and Lang (1985).
Sulfur in the Aqueous Phase
Sulfur in water samples, such as in tower wash water
and drainage, are analyzed for sulfide, sulfate, and/or
total sulfur.
Sulfate sulfur
Sulfate in most of the water samples is determined by a
turbidimetric method (APHA, 1989). A Milton Roy Model
Spectronic 21 Spectrometer was used at 420 nm to measure the
turbidity. Color or suspended matter in large amounts will
interfere with this method. In the case of dark colored


32
Materials and Methods
Four types of compost samples were investigated for
their decomposition characteristics during the course of
this study. All of the compost samples were obtained from
Broward County Streets and Highways Nursery Division,
Pompano Beach, Florida. The composts were stored in sealed
plastic bags at room temperature (232 C) before use. A
brief description of the composts used in this study is
presented in Table 3-1 ( Composts #1, #2, #3 and #6). The
compost samples were analyzed for their physical and
chemical properties at the beginning and the end of the
investigation. The results of these studies are presented in
Table 3-2. Each cured compost was passed through a 10 mm
screen to remove larger materials. The compost samples are
then placed in 225-mL wide mouth bottles directly for
incubation. The experimental arrangement for the
decomposition study is shown in Figure 3-1. Four types of
compost and one blank, each with three duplicates, were
investigated. Compressed air from the laboratory house air
supply is controlled to about 4 psig by a regulator. The
air stream is passed through a scrubber system consisting of
4N NaOH to remove C02 and distilled water to saturate the
air stream. A dead volume is placed before the 4N NaOH
scrubber as a safety precaution in the event that the air
system causes a backpressure forcing scrubbing solution
against the air system. An empty impinger is placed after
the water scrubber to separate larger water droplets from


24
Cell-Bound
Sulfur Complex
R
Sulfite
S
Sulfide
Elemental Sulfur
S 20g' Thiosulfate
e Electro T ransport
System
SO 4' Sulfate
SO
2-
4
Figure 2-4. Steps in the oxidation of different
compounds by thiobacilli. The sulfite
oxidase pathway is thought to account
for the majority of sulfide oxidized.
(Source: Brock and Madigan, 1989, p.
704) .


193
Jennings, P.A.; Snoeyink, V.L. and Chian, E.S.K.
"Theoretical Model for a Submerged Biological Filter,"
Biotechnology and Bioengineering, 18: 1249 (1976).
Johnson, D.W. and Henderson, G.S. "Sulfate Adsorption and
Sulfur Fractions in a Highly Weathered Soil Under a Mixed
Deciduous Forest," Soil Science, 128: 34 (1979).
Johnson, C.M. and Nishita, H. "Microestimation of Sulfur in
Plant materials, Soils, and Irrigation Waters," Anal. Chem.
24: 736 (1952).
Kampbell, D.H.; Wilson, J.T.; Read, H.W.; Thomas, T.
Stocksdale, T.T. "Removal of Volatile Aliphatic Hydrocarbons
in a Soil Bioreactor," Journal of Air Pollution Control
Association 37: 1236 (1987).
Kuenen, J.G. Colorless Sulfur Bacteria and Their Role in
the Sulfur Cycle," Plant and Soil 43: 49 (1975).
Lalazary, S.; Pirbazari, M. and McGuire, M.J. "Oxidation of
Five Earthy-Musty Taste and Odor Compounds," J. Amer. Water
Works Assoc. 78: 62 (1986).
Leson, G. and Winer, A.M. "Biofiltration: An Innovative Air
Pollution Control Technology for VOC Emissions," J. AW&MA,
41: 1045 (1991).
Lindstrom, K.P. Air Toxic Emissions and POTWs," Workshop
Report and Proceedings, Co-sponsored by WPCF and USEPA,
Alexandria, VA, July 10-11, 1989.
Miller, R.H. "Factors Affecting the Decomposition of an
Aerobically Digested Sewage Sludge in Soil," J. Environ.
Qual. 3: 376 (1974).
Miller, R.D. and Johnson, D.D. "The Effect of Soil Moisture
Tension on Carbon Dioxide Evolution, Nitrification, and
Nitrogen Mineralization," Soil Sci. Soc. Pro. 644-647 ,
1964.
McGilvery, R.W. and Goldstein, G.W. Biochemistry, A
Functional Approach. Third ed.; W.B. Saunders Company,
Philadelphia, PA, 1983.
Moorhead, K.K.; Graets, D.A. and Reddy, K.R. "Decomposition
of Fresh and Anaerobically Digested Plant Biomass in Soil,"
J. Environ. Qual. 16: 25 (1987).
National Research Council (NRC), "Hydrogen Sulfide,"
Subcommittee on Hydrogen Sulfide, Committee on Medical and
Biologic Effects of Environmental Pollutants. University
Parck Press, Baltimore, Maryland, 1979.


H2S Loading Rate (g-S/m3-hr) H2S Concentration (ppmv) H2S Removal
Efficiency (%)
126
Figure 5-1. Biofilter control of H2S during long
term operation,
a) Tower #1, compost #17A.


5 BIOFILTER PERFORMANCE AND CHANGES OF COMPOST
PROPERTIES ASSOCIATED WITH LONG TERM OPERATION 125
Overall Performance of the biofilters 125
Accumulation of Sulfur in Compost and Its
Effect on System Performance 137
System Upset and Recovery 146
Selection of Chemical Solutions 148
Effect of Water-Compost Contact Time on S042-
Leaching Efficiency 153
Effect of Water to Compost Ratio on S042
Leaching Efficiency 153
6 FULL SCALE APPLICATION OF BIOFILTRATION TO
CONTROL H2S EMISSIONS AT A WASTEWATER TREATMENT
PLANT 158
Introduction 158
System Design and Construction 160
Sampling and Analysis Methods 163
Results and Discussion 166
Conclusions 182
7 SUMMARY AND CONCLUSIONS 184
REFERENCES 189
BIOGRAPHICAL SKETCH 199
vi


Pressure Drop (mHgO)
84
Particle Size Range
A: d > 12 mm
B: 3.35 < d < 12 mm
C: 2.36 < d < 3.35 mm
D: 1.18 < d < 2.36 mm
E: d < 1.18 mm
Figure 4-8
Pressure drop as a function of particle
size range for different gas
velocities.


137
1) Premixed lime is effective only temporarily, lime can
be consumed by the accumulated acid very quickly;
2) Since high concentrations of H2S04 are continually
formed, the quantity of CaC03 required to neutralize
the acid is very large. Addition of large amounts of
lime increases the inorganic fraction of the filter
medium and significantly changes the compost
construction and composition.
3) The addition of lime increases the smaller particle
fraction in the filter, which results in a significant
increase in pressure drop across the filter bed, and
4) Most importantly, addition of lime does not solve
the problem of sulfur accumulation in the compost,
which appears to be the main reason for the decline in
H2S removal efficiency.
Accumulation of Sulfur in Compost
and Its Effect on System Performance
Another serious problem which is frequently encountered
in a H2S-biofilter system is the accumulation of sulfur in
the filter material (Carlson and Leiser, 1966; Rands et al.,
1981; Yang and Allen, 1991). This feature has been
routinely observed during the course of this study in both
Towers #1 and #2 since these towers are transparent. During
long term operation of the biofilter the color of the
compost eventually changes from dark brown to a yellowish
white. The color change progresses from the lower region of


89
Table 4-4. Effect of gas retention time on H2S removal
efficiency.
Gas Flow
Rate
(Lpm)
H2S
Reten. Loading H^S
Time Rate Inlet
(s) (g-S/m3-hr) (ppmv)
H2S Removal
Outlet Eff.
(ppmv) (%)
151 j,
7.06 rcmw-
16.3
106
9.98
17.1
75.5
14.1
17.7
46.0
23.1
21.0
30.8
34.5
20.7
15.3
70.0
19.8
15.0
71.0
39.8
15.0
71.0
50.7
9.00
118
56.3
9.00
118
73.2
5.40
197
62.3
24.9
1.62
93.5
37.0
1.30
96.5
53.7
0.77
98.6
105
0.30
99.7
155
0.02
99.9+
297
0.01
99.9+
610
BDL
99.9+
776
2.77
99.6
1440
0.01
99.9+
1870
4.24
99.8
2650
4.58
99.8


162
Figure 6-2. Photograph of the grit chamber at
Kanapaha Wastewater Treatment Plant
(top view). The chamber is covered to
collect the malodorous gas.


81
blower (fans) to move contaminated air at the specified flow
rate through the filter bed. However, the pressure drop
across the filter bed increases markedly as the flow rate is
increased. Since the pressure drop will be determined by
the depth of the filter bed it is necessary that the gas
velocity should be kept as low as possible.
The pressure drop across a compost bed filter can be
lowered by physical treatment of the compost particle
content. Such a procedure is an important requirement in
optimizing the operation of a filter bed, because operation
at a reduced resistance to flow allows the gas velocity as
well as the volumetric flow rate to be significantly
increased with little or no change in energy consumption.
This will in turn increase the biofilter capacity and reduce
the required filter size.
The following physical factors determine pressure drop
across the filter bed:
1) Particle size distribution in the compost
2) Condition of the filter packing
3) Height of the filter bed
4) Water content of the compost
5) Gas velocity, and
6) Porosity of the compost.
A representative sample of compost #12 was air dried
and the particle size distribution determined by weighing
the fractions penetrating a series of standard sieves. The
particles are classified into 5 size groups in the range >12
mm to <1.2 mm (see Table 4-3). Compost samples in each of


165
Teflon
Tubing
Quick Connecto
Sampling
Probe
Plastic
Chamber
3-Way Valve
Vacuum Gauge
P)
Tediar Bag
Biofllter Bed
Figure 6-3. Biofilter off-gas sampling system.
A
Pump


Sulfate Leached Out (mg-S/g)
155
Figure 5-7. Effect of water/compost ratio on
sulfate leaching.



16
observed.
Allen et al. (1987a, b, c) investigated a compost based
tower biofilter system used for odor control in a wastewater
treatment plant. The odor-causing compounds identified were
reduced sulfur compounds such as H2S, methyl mercaptan,
dimethyl sulfide and dimethyl disulfide as well as terpene
hydrocarbons. Removal efficiency for total reduced sulfur
compounds (TRS) was 65 to 72%. The poor performance of this
system was determined to be the short residence time in the
system, poor gas distribution, and improper maintenance.
Biofiltration control of volatile organic compounds
(VOCs) has been reported by Ottengraf (1986), Kampbell et
al. (1987), Bohn (1989), Paul and Castelijn (1987), and Hack
and Habets (1987) .
In recent years, increasing numbers of biological
filters are being used around the world for odor control. It
has been estimated that more than 500 biofilters are
currently operating in Europe (Leson and Winer, 1991).
Excellent summaries of recent applications have been
provided by Bohn and Bohn (1987) and Rotman (1991b).
Theoretical Basis
The concept of a biological-film or 'biofilm' is
freguently used to describe degradation processes in aqueous
systems ( Williamson, 1973; Williamson and McCarty, 1976a,
b; Jennings et al., 1976; Rittmann and McCarty, 1978).
This concept has been adopted and improved to describe the


179
however, the removal efficiency was much lower than
previously recorded values obtained during normal
operations. The large variation in H2S concentrations in
the off gas at different locations indicates uneven
distribution of the waste gas stream throughout the system.
Since not much air is vented through the east bed, the H2S
concentrations measured in the off gas for this bed (sites 4
and 5) are very low and therefore, the calculated removal
efficiencies for this bed (Table 6-4) are questionable. A
strong characteristic rotten egg odor (H2S) was smelled
around both filter beds during sampling.
The color of the compost packed on the west bed had
changed to yellowish-white in comparison to the color of the
I
compost packed on the east bed, which had still retained
its original dark brow color (Figure 6-8). Analysis of the
compost samples from these discolored sites is shown in
Table 6-5. Approximately 20% of the total S is organic-S
and 80% is inorganic-S. Very high fractions of FeS2-S
(27.6%) and insoluble sulfate (19.98%) were determined in
these samples.
As a result of this long term operational corrosion
problem, GRU engineers have decided to replace the existing
SIEBO-stones with similar blocks made from anti-corrosive
materials.


197
Tester, C.F.; Sikora, L.J; Taylor, J.M. and Parr, J.F.
"Decomposition of Sewage Sludge in Soil: III. Carbon,
Nitrogen, and Phosphorus Transformations in Different Sized
Fractions," J.Environ. Qual. 8: 79 (1979).
Urone, P."The pollutants," in Air Pollution. Stern, A.C. ed.
Vol.6, Academic Press, New York, 1986.
USEPA, "Odor and Corrosion Control in Sanitary Sewage
Systems and Treatment Plants," EPA-Design Manual, EPA/625/1-
85/018, 1985.
Van Lith, C. "Design Criteria for Biofilters," Paper # 89-
165.5, Presented at 82nd Annual Meeting & Exhibition, AW&MA,
Anaheim, CA June 25-30, 1989.
Varanka, M.W.; Zablocki, Z.M. and Hinesly, T.D., "The Effect
of Digestion Sludge on Soil Biological Activity," J. Water
Poll. Cont. Fed. 48: 1728 (1976).
Walker, G.S.; Lee, F.P. and Aifa, E.M. "Chlorine Dioxide for
Taste and Odor Control," J. Amer. Water Works Assoc. 78: 84
(1986).
Water Pollution Control Federation (WPCF), "Odor Control for
Wastewater Facilities," Manual of Practice No. 22, Water
Pollution Control Federation, Washington, DC., 1979.
White, A., P. Handler, E.I. Smith, R.L. Hill, I.R. Lehman,
Principles of Biochemistry. Sixth Edition, McGraw-Hill Book
Company, New York, 1978.
Wieder, R.K. and Lang, G.E. "An Evaluation of Wet Chemical
Methods for Quantifying Sulfur Fractions in Freshwater
Wetland Peat," Limnol. Oceanogr., 30: 1109 (1985).
Williamson, K.J. "The Kinetics of Substrate Utilization by
Bacterial Films," PhD Dissertation, Stanford University,
June 1973.
Williamson, K. and McCarty, P.L. "A Model of Substrate
Utilization by Bacterial Films," J. WPCF 48: 9 (1976 a).
Williamson, K. and McCarty, P.L. "Verification Studies of
the Biofilm Model for Bacterial Substrate Utilization," J.
WPCF 48: 281 (1976 b).
Yang, Y. "Odor Emissions and Its Control in a Wastewater
Treatment Plant with Industrial Sources," Master's Thesis,
University of Florida, Gainesville, FL, December 1988.


CHAPTER 5
BIOFILTER PERFORMANCE AND CHANGES IN COMPOST PROPERTIES
ASSOCIATED WITH LONG TERM OPERATION
Towers #1 and #2 were packed with composts #17A and
#17, respectively, and operated continuously for more than
200 days. Tower #3 was packed with compost #16 and operated
for 130 days. Long term performance of these biofilter
tower systems for H2S removal and changes in compost
properties are reported here. Biofilter maintenance
conditions and recommended procedures for developing optimum
performance have been determined through these long term
observations.
Overall Performance of the Biofilters
The overall performance of Towers #1, #2 and #3 are
presented graphically in Figures 5-1, a, b and c,
respectively. In each figure, H2S loading rates,
inlet/outlet H2S concentrations and H2S removal efficiencies
are plotted against the cumulative operation time. Except
for some specific tests, where extreme operating conditions
were used (such as high H2S loading rate, high gas flow
rate, etc) and data are presented and discussed elsewhere,
the data showed in these figures are daily averages for an
individual day when the measurements were made. Usually,
Towers #1 and 2 were operated at a gas loading rate of 100
125


144
Biological activities of the biomass in each region of
the filter bed material were indirectly determined by
measuring sectional H2S removal efficiency. Identical gas
flow rates and inlet H2S concentrations were used for these
measurements. The results are shown in Figure 5-5. It can
be seen that the most effective region in the biofilter is
between 0.2 and 0.4m. The biological activity of the lower
region (0-0.2m) is restricted by the factors mentioned
above. The maximum population of the sulfur oxidizing
bacteria as well as the optimum biological activity occur in
the second region (0.2-0.4m) of the filter and decreases
progressively up the bed. This observation is reasonable
because less and less H2S is available in the gas stream as
it passes upward through the filter and more of the H2S in
the gas stream is eliminated by reaction with the biofilter
in the lower region. The population of the sulfur oxidizing
bacteria and the biological activities appear to show a
modal (Gaussian type) distribution along the filter bed.
With prolonged operation, the mode will move upward from the
lower region of the bed due to the increasing toxicity
caused by accumulation of sulfate and acidification of the
compost in the lower region. The latter effect is referred
to as system upset, which must be avoid to maintain
effective control efficiencies.


117
4-21 Determination of the fractional-order reaction rate
coefficient, kf by linear least squares
regression. Gas loading rate: 224 m3/m2-hr,
compost #17
4-22 Plot showing the fractional-order kinetics of H2S
oxidation in biofilter. Gas loading rate:
224 nr/nr-hr, compost #17 118
4-23 Concentration profiles for H2S as a function of
packing height within the biofilter. Gas loading
rate: 224 m3/m2-hr, compost #17 119
5-1 Biofilter control of H2S during long term
operation, a) Tower #1, compost #17A 126
b) Tower #2, compost #17 127
c) Tower #3, compost #16 128
5-2 Compost water content profile 132
5-3 pH changes of compost in different sections of the
biofilter with operation time.
a) Tower #1, compost #17A 134
b) Tower #2, compost #17 135
5-4 Total-S distribution profile in biofilter,
Tower #1, after exposure to H2S for 100
days 14 3
5-5 H2S removal efficiencies in different regions of
the biofilter, Tower #2. A: 0-0.2 m,
B: 0.2-0.4 m, C: 0.4-0.6 m, D: 0.6-0.8 m,
E: 0.8 1.0 145
5-6 Effect of water-compost contact time on sulfate
leaching efficiency 154
5-7 Effect of water/compost ratio on sulfate leaching 155
6-1 Schematic diagram of the Kanapaha biofilter bed
system 161
6-2 Photograph of the grit chamber at Kanapaha
Wastewater Treatment Plant (top view). The
chamber is covered to collect the
malodorous gas 162
6-3 Biofilter off-gas sampling system 165
6-4 Photograph of the biofilter system at Kanapaha
Wastewater Treatment Plant 169
xi


5
hydrogen sulfide. This research provides a detailed data
base on the effects of system variables on H2S control
efficiency, which provide for optimization of design and
operating conditions.


Compost Water Content (%)
i
132
Height Above Bed Inlet (m)
After Washing PKxa Before Washing
Figure 5-2
Compost water content profile.


185
compost material. Small particles have a low water
holding capacity and are less valuable to the overall
biofiltration process. It is recommended that small
particles be separated from the compost by sieving
before use. To minimize pressure drop effects the
filter should not be compacted unduly. Aged compost,
which contains a larger fraction of fine particles,
due to mineralization and fracture, should not be
reused after change-out, unless it is specifically
treated to remove the fine particle fraction by
sieving or washing.
2. The time reguired for the oxidation of H2S to sulfate
by microorganisms is a few seconds, however, the use
of high gas velocities is not recommended since they
will cause uneven gas distributions and high pressure
drops.
3. The concept of maximum H2S elimination capacity of
compost and H2S loading rate is very important in
terms of system design and operation. When working
within the maximum elimination capacity of the system,
the waste gas flow rate can be adjusted to obtain the
best reduction of H2S for various inlet
concentrations.
4. Low compost water content is fatal to the biological
process. A minimum value of 30% water content by
weight is reguired for proper operation, but 40 to 60%
is recommended.


REFERENCES
Aaronson, S. Experimental Microbiology. Academic Press, New
York, 1970.
Allen, E.R.; Hartenstein, H.U. and Yang, Y. "Review and
Assessment of the Design and Operation of a Compost
Biofilter System for Odor Control," Final Project Report,
Environmental Engineering Sciences Department, University of
Florida, Gainesville, Florida, 1987a.
Allen, E.R.; Hartenstein, H.U. and Yang, Y. "Identification
and Control of Industrial Odorous Emissions at a Municipal
Wastewater Treatment Facility," Paper # 87-95A.4, Presented
at 80th Annual Meeting of APCA, New York, June 21-26, 1987b.
Allen, E.R.; Yang, Y.; Hartenstein, H.U. "Odor Producing
Agents, Their Sources and Control", Final Project Report,
Environmental Engineering Sciences Department, University of
Florida, Gainesville, Florida, 1987c.
Allen, E.R.; Yang, Y. and Hartenstein, H.U. "To Provide
Technical Assistance in the Design and Installation of An
Odor Control Biofilter System at the Kanapaha Wastewater
Treatment Plant," Final Project Report, Environmental
Engineering Sciences Department, University of Florida,
Gainesville, Florida, 1989.
Allen, E.R. and Yang, Y. "Biofiltration Control of Hydrogen
Sulfide Emissions," Paper # 91-103.10, presented at the 84th
Annual Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
American Public Health Association (APHA), Standard Methods
for the Examination of Water and Wastewater. 17th ed.
American Public Health Association, Washington, DC, 1989.
Atlas, J.E. and Bartha, R. Microbial Ecology: Fundamentals
and Applications. Addison-Wesly Publ. Company, Reading, MA,
1981.
Bethea, R.M.; Murthy, B.N. and Carey D.R. "Odor Controls
for Rendering Plants," Environ. Sci. Tech. 7s 504 (1973).
189


15
(Rotman, 1991a).
At present, biofiltration is considered to be a state-
of-the art technology for odor removal in West Germany, and
it has been estimated that 40% of deodorization facilities
at wastewater treatment plants are biofilters (Frechen and
Kettern, 1987) .
Applications
The first systematic study of odor control using
biofiltration in this country was conducted by Carlson and
Leiser (1966). They studied the removal efficiencies of
sewage odors using a laboratory scale soil bed. Using
hydrogen sulfide as the test gas, a 99% removal efficiency
was achieved, and biodegradation was reported to be the
primary removal mechanism.
Prokop and Bohn (1985) reported that a soil bed system
for control of rendering plant odors had been in operation
since September, 1983. The soil bed treats 1100 m3/h of
cooker non-condensable waste gases using a bed surface area
of 420 m2. In this work an odor removal efficiency of
99.9% was obtained.
Rands et al. (1981) reported that a full-scale compost
filter system was constructed in 1978 at Moerewa, New
Zealand, to treat odors from a rendering plant. The system
was designed to treat 900 m3/h of air containing hydrogen
sulfide concentrations up to 1000 parts per million (ppm) by
volume. An average H2S removal efficiency of 99.9% was


(C/C0)
117
Figure 4-21. Determination of the fractional-order
reaction rate coefficient, kf by linear
least squares regression. Gas loading
rate: 224 m3/m-hr, compost #17.


45
four composts studied are quite similar, falling in the
range from 3.1xl0-4 to 5.7xl0-4/day. If the second reaction
rate coefficients for the composts studied can be assumed to
be representative through the remaining life of the compost,
then a rough estimate of the time required for decomposition
of 50% of the organic matter (half life) in these composts
can be made according to following equation.
tQ#5 = 0.693/k (3-5)
where:
tQ5 = the half life time of the compost, (day), and
k = the first-order reaction coefficient, (1/day)
The estimated half life time of the composts tested is
from 3.3 to 6.1 years. This estimate is comparable to the
result reported by Varanka et al. (1976), who showed that it
takes approximately 6 years to lose 50% of the sludge
organic C when used in the field.
No significant changes in other physical and chemical
properties of the composts were observed for the 120 day
incubation period used in these studies (Table 3-2).
Effect of FUS on Compost Decomposition
The effect of H2S exposure on compost decomposition is
illustrated in Figure 3-5, where C02 evolved by the composts
is expressed as mg-C02 per g-C of the compost as a function
of the H2S concentration (ppmv). The C02 evolution is
significantly increased with the increasing H2S
concentration. The rate of this increase is greater at


26
biofilter. The C or N deficient filter material must be
replaced by freshly prepared material and the discarded
filter material has to be properly disposed of with due
caution for environmental impact. One of the most common
options is land application. Determination of the
decomposition characteristics of the filter material is,
therefore, necessary for usage of the biofilter and eventual
land disposal applications.
Considerations necessary for selection of appropriate
filter materials and the decomposition of such materials
under aerobic conditions are discussed in this section.
Selection of Filter Materials
Effective removal of air contaminants using a biofilter
relies on the properties of the filter material, especially
the nature and activity of the biomass. The filter
material provides the necessary environment for
microorganisms to survive, generate, function and allows
the entire sequence of biofiltration processes to be carried
out. The filter material serves as 1) support material for
the microbes, 2) supplemental or alternative nutrient
source, 3) moisture storage reservoir, 4) surface area for
sorption of air pollutants and interaction between the
pollutants and the microorganisms, and 5) a buffer volume
for variations in water content and gas conditions during
operation (Eitner, 1989). In general, the following factors
need to be considered when choosing a suitable filter


147
a filter material is determined at optimized operating
conditions. Changes in operating conditions, such as a
lowered pH, dry-out of the compost, accumulation of sulfur
in the bed material, etc. can significantly decrease the H2S
elimination capacity of the filter medium. Therefore, the
actual H2S elimination capacity of a filter material at a
specific condition is always equal to or less than its
maximum capacity. When the H2S loading rate exceeds the
elimination capacity of the compost, then the system is
overloaded.
An overloaded system is indicated by a high H2S
concentration in the effluent gas stream, a low removal
efficiency, noticeable odor and the occurrence of a compost
color change (white deposit on compost particles) Local
overloading is often observed when channeling occurs in the
system (Rands et al., 1981) or the influent gas stream is
not evenly distributed. In either case, the region of the
filter where the high flow rate occurs is overloaded by H2S.
Local overloading can be cured by correcting the channeling
and the gas distribution system.
If the whole system is overloaded by high H2S input,
the solution to the problem is different. If the system is
temporarily overloaded for a few hours, the performance of
the filter can be recovered by decreasing the H2S loading
rate. A certain fraction of the white deposits on the
compost are intermediate oxidation products, such as S,
FeS2, S 2 0 3 ^-, etc. When H2S loading is decreased, the


H2S Loading Rate (g-S/m3-hr) H2S Concentration (ppmv) H2S Removal
Efficiency (%)
127
Figure 5-1. Biofilter control of H2S during long
term operation,
b) Tower #2, compost #17.


LIST OF FIGURES
Figure Page
2-1 Solubility of H2S in water at 1 atm 7
2-2 Effect of pH on H2S Equilibrium 8
2-3 Biophysical model for the biological filter bed.
The concentration profiles shown in the
biofilm refer to: 1) Reaction limitation,
2) Diffusion limitation 18
2-4 Steps in the oxidation of different compounds by
thiobacilli. The sulfite oxidase pathway is
thought to account for the majority of sulfide
oxidized 24
3-1 Schematic drawing of the experimental arrangement
for the study of compost decomposition 34
3-2 Schematic drawing of the experimental arrangement
for the investigation of the effect of H2S
exposure on compost decomposition 37
3-3 Plot of CO2 evolution from composts during the
122 day incubation 40
3-4 Decomposition stages and reaction rate
coefficients for the four composts studied .... 44
3-5 Effect of H2S exposure on the rate of compost
decomposition as measured by C02 respiration .. 46
3-6 Plot of CO2 evolution as a function of square
root of H2S concentration 48
4-1 Schematic drawing of the dual tower system 53
4-2 Sampling and measurement ports on towers.
a. Tower #1 55
b. Tower #2 56
4-3 Schematic drawing of Tower #3 59
IX


30
Table 3-1 Continued.
Compost
Source
Description
ID#
16
WRRC
Yard trash compost, 3.5 months old when
first used in Tower #3 from 1/27/91.
17
WRR
1:1 by volume of yard trash and grass
composted; about 3.5 months old when
first used in Tower #2 from 1/27/91.
17A
WRR
Compost #17 mixed with 2% lime (CaC03),
by dry weight of compost. Used in Tower
#1 from 1/27/91.
a Broward County Streets and Highways Division, 1600 NW 30th
Avenue, Pompano Beach, FL, 33069.
b Kanapaha Wastewater Treatment Plant, Gainesville, FL 32602.
c Wood Resource Recovery, Inc., Gainesville, FL.


190
Biddlestone, A.J.; Gray, K.R. and Day, C.A. "Composting and
Straw Decomposition," in Environmental Biotechnology,
Forster, C.F. and Wase D.A.J. Eds., John Wiley & Sons, New
York, 1987.
Blake, G.R. and Hartge, K.H. "Particle Density," in Methods
of Soil Analysis. Part I. Physical and Mineralogical
Methods. Klute, A. Ed. 2nd ed, American Society of Agronomy,
Madison, Wisconsin, 1986a.
Blake, G.R. and Hartge, K.H. "Bulk Density," in Methods of
Soil Analysis. Part I. Physical and Mineraloqical Methods.
Klute, A. Ed. 2nd ed, American Society of Agronomy, Madison,
Wisconsin, 1986b.
Bohn, H.L. and Miyamoto, S. "Soil as a Sorbent and Filter of
Waste Gases," in Symposium on land for Waste Management.
Tomlinson, J. ed. Natl. Research Council Canada, Ottawa,
1973.
Bohn, H.L."Soil and Compost Filters of Gases," J. Air
Pollution Control Assoc., 25: 953 (1975).
Bohn, H.L. "Compost Scrubbers of Malodorous Air Streams,"
Compost Sci., 17: 5 (1976).
Bohn, H.L. "Soil Treatment of Organic Waste Gases, Chapter
24, in Soils for Management of Organic Waste and Waste
Waters. ASA-CSSA-SSSA, Madison, WI, 1977.
Bohn, H.L.; Bohn, R.K. "Soil Bed Scrubbing of Fugitive Gas
Releases," J. Environ. Sci. Health, A21: 561 (1986).
Bohn, H.L. and Bohn, R.K. "Biofiltration of Odors from Food
and Waste Processing," Proceedings of Food Processing Waste
Conference, Georgia Tech Research Institute, Sept. 1-2,
1987.
Bohn, H.L. and Bohn, R.K. "Soil Beds Weed Out Air
Pollutants," Chemical Engineering, pp. 73-76, April 1988.
Bohn, H.L. "VOC Removal by Soil Biofilter Beds," Proceedings
of Hazmacon'89, Hazardous Materials Management Conference
and Exhibition, Vol.2, Association of Bay Area Govermments,
April 18-20, 1989.
Brock, T.D. and Madigan, M.T. Biology of Microorganisms.
Fifth Edition, Prentice Hall Inc., Englewood Cliffs, New
Jersey, 1988.
Carlson, D.A. and Leiser, C.P. "Soil Beds for the Control of
Sewage Odors," Journal of the Water Pollution Control
Federation, 38: 829 (1966).


177
latter resulted in blocking of the narrow waste gas
distribution vents between the SIEBO-stones, which caused
channeling and uneven gas distribution throughout the
biofilter system.
Corrosion of the SIEBO-stones and the resultant
blockage of the gas distribution system were first noticed
by Kanapaha Plant workers in late 1990. To solve this
problem, the west bed of the biofilter system was unpacked
and the gas distribution vents between the SIEBO-stones were
cleaned manually. The compost used in this bed was
thoroughly mixed by turning and repacked in the bed after
the gas distribution vents had been cleaned.
On February 5, 1991, both gas and compost samples were
taken at the biofilter beds. Results of this recent gas
sampling and analysis exercise are presented in Table 6-4.
Sampling sites 1, 2 and 3 are located on the west bed; 4 and
5 are located on the east bed. Almost no air was vented
through the east bed, even though the gas flow control
valves to both beds were fully open, because of the
corrosion and blockage of the east bed gas vents. During
cleaning of the east bed air distribution system, the entire
waste gas stream was forced through the west bed. The waste
gas stream, however, was no longer evenly distributed
through this 'cured' system. Channeling caused the waste gas
to blow through the filter without control and resulted in
very high off gas H2S concentrations (sites 1 and 3, Table
6-4). Some removal of the H2S was measured at Site 2,


99
The compost used in each column was treated with either
dilute HC1 or dilute NaOH solutions to bring the pH of the
compost to a desired range. The final pH values obtained
for the composts studied in each column are 1.57, 3.20,
4.42, 5.02, 6.39, 6.75, and 8.76, respectively.
Measurements of inlet and outlet gas sample concentrations
were made when the bed operation became stable. The results
of these studies are shown in Figure 4-14. Repeated
measurements at each pH point are used to estimate the mean
value, and mean 2 standard deviations values (error bars)
indicated by the upper and lower bounds of the vertical line
through the mean. Two operational conditions are used in
this test. Under condition A, lower H2S and gas loading
rates are used. No significant effect of pH of compost on
H2S removal is observed for pH values in the range between
3.2 and 8.76. Removal efficiencies of 99.5+% are
consistently determined with little or no variation in this
range. When the pH was reduced to 1.57, the H2S removal
efficiency fell sharply to about 9%. It should be noted
that the measured H2S removal efficiencies showed larger
variations at the lower pH range. Following the previous
studies, the operational conditions were changed to
condition B. Under the latter conditions, higher H2S and
gas loading rates were used and the effects of varing pH the
compost becomes evident. It can be seen in Figure 4-14,
condition B, that the maximum H2S removal occurred at a
compost pH value of 3.2 (99.2%). The removal efficiency


To my father for making this possible
and
In memory of my mother


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID EL8920Z1W_SP4L7T INGEST_TIME 2015-02-17T19:43:16Z PACKAGE AA00028767_00001
AGREEMENT_INFO ACCOUNT UF PROJECT UFDC
FILES


112
klf can be determined from the slope of the line.
3) In the third situation, when the half-saturation
constant, Km, and the substrate concentration, C, are
comparable, the biological reaction should follow fractional
order kinetics, because equation 4-2 can not be simplified.
Relatively complex equations have to be derived to express
the fractional order kinetics.
A number of theoretical and empirical models have been
reported which describe the kinetics of biodegradation of
organic compounds and reduced sulfur species (Chen and
Morris, 1972; Williamson, 1973; Cooper, 1974; Jennings et
al., 1976; Williamson and McCarty, 1976a,b; O'Brien and
Birkner, 1977; Rittmann et al., 1978; Schmidt et al., 1985;
Kampbel et al., 1987; Caunt and Hester, 1989). In
particular, Ottengraf and coworkers have published a series
of papers that describe the processes involved in the
operation of biofilters (Ottengraf and Van Den Oever, 1983;
Ottengraf et al., 1984; Ottengraf, 1986; Ottengraf et al.,
1986; Ottengraf, 1987). Ottengraf developed a biophysical
model as well as derived mathematical solutions to describe
the kinetics of biodegradation of various organic compounds
in biofilter systems.
As described previously in chapter 2, Ottengraf's model
uses the concept of a bed consisting of solid filter
particles, where each particle is surrounded by a wet,
biologically active layer. When waste air flows around the
particle there is continuous mass transfer of pollutant from


191
Caunt, P. and Hester, K.W. "A Kinetic Model for Volatile
Fatty Acid Biodegradation during Aerobic Treatment of
Piggery Wastes," Biotechnology and Bioengineering, 34: 126
(1989).
Chen, K.Y. and Morris, J.C. "Kinetics of Oxidation of
Aqueous Sulfide by 02," Environmental Science & Technology,
6: 529 (1972) .
Cooper Jr, H.B.H. "Kinetics of Inorganic Sulfur Oxidation
during Black Liquor Oxidation with Oxygen," Tappi 57: 130
(1974) .
Danielson R.E. and Sutherland, P.L. "Porosity," in Methods
of Soil Analysis. Part I. Physical and Mineralooical
Methods. Klute, A. Ed. 2nd ed, American Society of Agronomy,
Madison, Wisconsin, 1986.
David, M.B.; Mitchell, M.J. and Nakas, J.P. "Organic and
Inorganic Sulfur Constituents of a Forest Soil and Their
Relationship to Microbial Activity," Soil Sci. Soc. Am. J.
46: 847 (1982).
Delaune, R.D.; Reddy, C.N. and Patrick Jr., W.H. "Organic
Matter Decomposition in Soil as Influenced by pH and Redox
Conditions," Soil Biol. Biochem. 13: 533 (1981).
Dharmavaram, S. "Biofiltration A Lean Emission Abatement
Technology," Paper # 91-103.2, presented at the 84th Annual
Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
Don, J.A. "The Rapid Development of Biofiltration for the
Purification of diversified Waste Gas Streams," in VDI
Berichte 561: VDI Verlag, Dsseldorf, 1985.
Eitner, D. and Gethke, H.G. "Design, Construction and
Operation of Bio-filters for Odor Control in Sewage
Treatment Plants," Paper # 87-95A.6, Presented at the 80th
Annual Meeting of Air Pollution Control Association, New
York, New York, June 21-26, 1987.
Eitner, D. "Biofilter in Flue Gas Cleaning: Biomasses,
Design, Costs, and Applications," Brennst.-Waerme-Kraft
(German). 41/3, 124., 1989.
Ergas, S.J.; Schroeder, E.D.; Chang, D.P. "VOC Emission
Control from Wastewater Treatment Facilities Using
Biofiltration," Paper # 91-105.4, presented at the 84th
Annual Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
Ferguson, P.A. Hydrogen Sulfide Removal from Gases. Air and
Liquids. Noyes Data Corporation, Park Ridge, NJ., 1975.


73
The fresh wet compost sample is divided into five sub
samples. The latter are used for the analyses of 1) compost
moisture content, 2) total-S, 3) water-soluble-S, 4)
inorganic and organic sulfur fractions, and 5) storage for
later use.
Total sulfur
Total-S is determined by oxidation (acid digestion) of
the various reduced sulfur constituents to sulfate and
followed by reduction of the sulfate to H2S. The H2S is
then trapped and titrated as described above. In most of
the analyses, 0.5-5.0 g of fresh compost is used, depending
on the sulfur content. In some analyses, oven dried compost
is used. In the latter case, the compost samples are finely
grounded (<40 mesh) and a correspondingly smaller size of
compost sample is used. The compost is subjected to acid
digestion as described by Tabatabai (1982). The digest is
quantitatively transferred into a 100-mL volumetric flask
and the volume is adjusted with IN HCl. Reduction of the
sulfate is carried out by a reducing mixture. The reducing
mixture contains 50% hypophosphorous acid, 90% formic acid,
and hydriodic acid in a 4:2:1 proportion and is prepared as
described by Tabatabai (1982). Depending on the sulfur
content, 1 to 5 mL aliquot of the digest is transferred into
the distillation flask. With aliquots >2 mL, the volume is
reduced to about 2 mL by heating the flask on an electric
heating mantle. Five (5) mL of the reducing mixture is


140
2Na2S406 + 702 + 6H20 2Na2S04 + 6H2S04
(5-6)
5H2S + 8KNO3 4K2S04 + H2S04 + 4N2 + 4H20
(5-7)
5S + 6KNO3 + 2H20 3K2S04 + 2H2S04 + 3N2
(5-8)
5Na2S203 + 8NaN03 + H20 -
9Na2S04 + H2S04 + 4N2 (5-9)
From the reactions listed above and Table 2-1 in
Chapter 2, it can be seen that the colorless sulfur bacteria
can oxidize both hydrogen sulfide (H2S) and the intermediate
reduced sulfur compounds to sulfate. Different sulfur
compounds, therefore, in various stages of oxidation can be
expected to be present in the biofiltration system.
The original compost, #17A, and compost samples in the
biofilter, after 3 months continuous operation, were
collected and analyzed for total-S and for fractionation
into various sulfur components. Compost samples in the
biofilter system were taken from the lower (TS11, 0.125m),
middle (TS13, 0.625m), and upper (TS14, 0.875m) regions of
the tower. Total-S and the following sulfur components:
ester-S, FeS2-S, FeS-S. S-S, water-soluble-S,
P-extractable-S and insoluble-S, were analyzed using the
methods described in the previous section. S042-S and
inorganic-S are estimated according to individual analyses.
Organic-S is calculated as the difference between total-S
and inorganic-S, and C-bonded-S is the difference between
organic-S and ester-S. The results are summarized in Table
5-1.


130
water by evaporation from the compost. Drying-out of the
compost, therefore, is a natural feature of any biofilter
system. The drying-out process may be slower for those
biofilters which are used to remove hydrocarbons because
water is one of the products of the biodegradation
reactions. There is no water formed during the biological
oxidation of H2S. As a result, water has to be added to the
system at the effluent end of the bed to keep the compost
water content constant.
Towers #1 and #2 are packed with the same compost (#17)
except that 2% by weight of CaC03 was added to the compost
packed in Tower #1 (Compost #17A) as a pH buffer. It should
be noted that the addition of CaC03 did not affect the
performance of the biofilter.
These two towers have been packed and operated since
January 29, 1991. Both of these towers have been subjected
to large variations in waste gas surface loading and H2S
loading rates during the operation period because the
effects of these variables on H2S removal efficiencies were
evaluated in this dual-tower system. High H2S removal
efficiencies and stable performance, were consistently
observed.
The composts (#17 and #17A) showed good moisture
retention and buffering capacity. In addition to the water
added by pre-humidification of the inlet gas stream, these
two towers were washed biweekly using DI water. The latter
procedure was performed to keep the compost water content in


CHAPTER 4
DETERMINATION OF THE DESIGN AND
OPERATIONAL PARAMETERS FOR BIOFILTER SYSTEMS
Introduction
Extensive experimental work has been carried out in
order to determine the design and operational parameters for
a biofilter system. This research is essential for best
operation as an air pollution control technology and for
optimization of the system. This chapter describes the
design and construction of lab scale biofilter systems,
experimental methodology used and the results obtained.
System Design and Construction
Three biofilter systems were designed and constructed
for different investigative purposes. Each system can be
operated and controlled separately. Detailed information on
each experimental system is presented below.
The Dual Tower System
Most of the experimental work was carried out using a
dual-tower experimental biofilter system. This
configuration, which is shown in Figure 4-1, consists of
parallel dual column filters. The two biofilter columns,
identified as Tower #1 and Tower #2, can be run
52


58
Portable Tower #3
The portable tower is made from PVC pipe with an inner
diameter of 77 mm (3 in) This tower, which is shown in
Figure 4-3, has a total height of 1.2 meters (3.94 ft) with
an effective packing height of 1 meter (3.28 ft). The two
ends of the pipe are covered by rubber caps and held by pipe
clamps. Compost packed in the tower is supported by a
packing of non-biodegradable plastic screen. Measurement
ports for pressure, temperature and exhaust gas samples are
located along the length of the tower. Gas to be tested is
introduced through a port at the bottom of the tower. The
overall gas flow rate is measured by a pre-calibrated flow
meter after the effluent gas passes through a particulate
filter. This portable tower was used intensively for
pressure drop studies and for investigation of long term
operation of compost #16.
Column System #4
A fourth column biofilter system was constructed for
investigation of the effects of various operational
variables on H2S removal (Figure 4-4). This multicolumn
system includes eight compost columns, a manifold for
introduction of test gas and several needle valves for flow
control. The columns are made from PVC pipes with an inside
diameter (ID) of 35 mm (1.25 in). Each column has a length
of 300 mm ( 12 in) and an effective packing height of 250 mm


CHAPTER 7
SUMMARY AND CONCLUSIONS
The intention of this research was to develop a
quantitative knowledge of the operation of a microbial
biofilter system for removal of hydrogen sulfide from waste
gas streams and to optimize and maintain the performance of
such a biofilter system using such knowledge. Optimization
of the system involved quantitatively determining the design
parameters, the operating parameters and predictive
relationships for the control efficiency. Maintaining the
system involves recognizing system deterioration and upset
and providing solutions for prevention of long term
irreversible deterioration.
A lab scale biofilter tower system was constructed and
extensive experimental work was conducted to achieve the
stated goals. In addition, a full-scale compost biofilter
bed system for control of H2S emissions in a wastewater
treatment plant was evaluated during long term operation.
Based on the results of this study, it is concluded that:
1. Significant pressure drops in biofilter materials are
mainly caused by the presence of small particles,
particularly those with diameters less than 1 mm.
These small particles are generally composed of sand
and minerals, as well as decomposed and mineralized
184


3
systematic compilation of data from an operational point of
view is also lacking. Most designs are conservatively based
on blanket rule of thumb' criteria (Forster and Wase,
1987). The performance of biofilter systems, therefore, is
not readily predictable and sometimes these systems are not
operated under suitable conditions. As a result, the
desired odor control efficiency is sometimes not achieved
(Allen et al., 1987b). It is essential that more work be
done to demonstrate the effectiveness of these systems in
order to support further progress in the use of
biofiltration as well as to develop better biofilters, based
on an understanding of the fundamental physical, chemical
and biological processes involved.
A major disadvantage of biofiltration technology is the
limited degradation capacity represented by the volume of
waste gas treated per unit area of filter material per unit
time (m3/m2-hr). This limitation restricts the applicability
of biofiltration systems to handling dilute waste gas
streams and requires the filter bed to be large in order to
handle high volumetric gas flows.
In order to overcome the uncertainties and
disadvantages encountered in the full scale application of
biofiltration technology, an exhaustive study is necessary
for the application of biofiltration technology to control
the emissions of air pollutants.
The objectives of the proposed research were to develop
a quantitative knowledge of the principle and operation of


83
the particle size groups were analyzed for water content
then wetted to obtain a water content of about 50%, by
spraying and mixing water with the compost samples. These
treated samples were then used to determine the pressure
drop as a function of the 5 particle size range classes by
adjusting air velocities in the range 0.02 m/s to 0.28 m/s.
All tests were carried out with the same compost bed height
(1 m) and water content (50%).
The results of these studies are presented in Figure 4-
8, where it is seen that the pressure drop increases
significantly with increasing gas velocity for a bed of
small particle size (<1.2mm). For particles greater than
1.2 mm the pressure drop increases to a much lesser extent
with increasing velocity as shown by a comparison of data
for particle classes D and E. Clearly significant pressure
drops observed in operating filter beds are a result of the
presence of small particles with sizes less than 1 mm. For
example, at a representative gas velocity of 0.03 m/s, which
is equivalent to a loading rate of 110 m3/m2-hr, the
pressure drop realized by a 1-m bed of particles size
classified as <1.2 mm is 390 mm H20, whereas the pressure
drop obtained for the same packing height and the same gas
velocity, for particles classified as >12 mm is only 2 mm
H20. Thus, under these conditions the pressure drop created
by 1 mm or less particles is about 200 times that caused by
an equivalent bed composed of 12 mm or greater particles.


Regulator
Needle
Valve
Manifold
Dead 4N Dl Dl
Volume NaOH Water Water
Impinger 50g 2X25ml
Compost 0.5N NaOH
Figure 3-1. Schematic drawing of the experimental arrangement for the
study of compost decomposition.


138
the bed (inlet) to the upper layer. White deposits on the
surface of compost particles are easily observed.
The rate of sulfur deposition is proportional to the
rate of H2S loading. A sudden increase in H2S loading in a
large concentration range and prolonged operation at high
H2S loading rates can cause the white colored material to
accumulate rapidly and spread from the lower region to the
upper region of the bed. This discoloration of the bed is
accompanied by a rapid drop in pH of the compost. Also, the
temperature of the biofilter system can rise 2-3 C for high
H2S loading rates indicating enhanced biological activity of
the microbes. If no appropriate action is taken to
counteract sulfur accumulations, then the system
performance and H2S removal efficiency will decline rapidly.
In biofiltration processes, H2S is oxidized both
chemically and biologically to sulfate under aerobic
conditions. In nature a variety of reduced inorganic sulfur
compounds (e.g. elemental sulfur, thiosulfate) occur as
intermediates between sulfide and sulfate, the reduced and
oxidized forms of sulfur, respectively. As these compounds
are oxidized only slowly by direct chemical reaction with
oxygen (Kuenen, 1975), it is clear that biological oxidation
must play an important role in the recycling of reduced
sulfur compounds under aerobic conditions. This mechanism
appears to be true, also, in biofiltration systems.
Also, many microorganisms can oxidize reduced sulfur
compounds, the colorless sulfur bacteria are known to play a


11
at 20 ppm (15-minute exposure) for an acceptable ceiling
concentration and 50 ppm for a maximum exposure during an 8-
hour work shift if no other measurable exposure occurs. The
National Institutes of Occupational Safety and Health
(NIOSH) established an H2S exposure level at 10 ppm (10
minutes) as a maximum permissible limit (once per 8-hours
shift), with continuous monitoring reguired where H2S
concentrations could egual or exceed 50 ppm or greater
(NIOSH, 1979).
Hydrogen sulfide is an explosive gas. The lower and
upper explosive limit are 4.5 and 45 percent in air by
volume, respectively.
Hydrogen sulfide can attack materials and cause
discoloration and tarnishing. Materials commonly affected
are paint, copper, zinc and silver (Painter, 1974).
Sources of H2S Emissions and Regulations
Natural emissions are mainly caused by biological decay
of protein materials. The natural global rate of emission
is estimated to be about 84 Tg/year (Urone, 1986).
Anthropogenic emission sources include petroleum
refining, natural gas plants, sewage treatment facilities,
coke ovens, Kraft paper pulp plants, and waste disposal
sites. There are no federal U.S. emission standards for H2S
at present, nor are there federal ambient air guality
standards for this gas, but a number of states have


93
The maximum H2S elimination capacity for a compost is
determined at the optimum operating conditions of the
system. H2S concentrations in the inlet gas stream are
varied increasingly at a constant gas flow rate. The H2S
removal rates (g-S/m3-hr) are plotted v.s. the H2S loading
O
rates (g-S/m -hr). The maximum H2S elimination capacity of
the compost is determined when the curve flattens out
(Ottengraf, 1986). The maximum H2S elimination capacity of
compost #13 is determined to be 11.5 g-S/ m3-hr (Figure 4-
11). The maximum H2S elimination capacity of this compost is
very low because its extremely low pH (1.60) and high
sulfate content (70.4 mg-S/g, see Table 4-2).
The maximum H2S elimination capacity for a compost can
also be determined through kinetics studies. Detailed
information is presented in the section of "Kinetics of H2S
Oxidation in a Biofilter". The maximum H2S elimination
capacity for compost #17 is determined to be 129 g-S/m3-hr.
Effect of Compost Water Content on H2S Removal
The effect of compost water content (CWC) on H2S
removal was evaluated using the Column System #4. The
compost (Compost #17) which was used for this study and
other studies carried out using Column System #4 had been
previously packed in Tower #2 and was considered to be pre
conditioned by H2S to a stable operating condition. After
transfer to System #4, the compost was operated at the same
conditions as used in its parent environment, Tower #2,


80
Table 4-2. Summary of initial compost properties.
Compost ID #
Property
12
13
14
16
17
PH
2.61
1.60
6.44
6.66
8.10
Bulk Density (g/cc)
0.30
0.27
0.18
0.22
0.20
Particle Den. (g/cc)

1.73
1.75
1.90
1.78
Porosity (%)

84.3
89.7
88.4
88.7
Water Content (wt%)
45.6
54.9
62.4
56.5
62.7
Organic Matter (wt%)
66.5
66.8
59.3
64.3
62.6
Total-S (mg-S/g)
44.8
70.4


0.74
Water-P (mg/kg)
140
223
152
114
167
Total-C (wt%)
34.3
31.0
31.3
30.5
40.9
Total-N (wt%)
1.89
3.24
1.75
4.27
1.30
C/N
18.1
9.57
17.9
7.14
31.5
Metals (mg/kg)
Ca
47400
28900
145000
18000
26700
Mg
280
255
4880
1450
120
Zn
38.0
48.0
201
66.5
18.0
Cu
121
91.0
60.0
11.0
93.5
Mn
5.50
55.0
96.5
89.5
9.50
Fe
1900
805
6160
529
1510


119
o
O
O
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
C0= 434 ppmv
Zero-order Kinetics
C0= 206 ppmv
First-order Kinetics
C0 = 309 ppmv
Diffusion limitation
h/H
Figure 4-23. Concentration profiles for t^S as a
function of packing height within the
biofilter. Gas loading rate: 224
itr/m-hr, compost #17.


136
Washing the tower with water effectively mitigates the
pH decline. Since H2S04 is water soluble, a major fraction
of the accumulated acid can be washed out at each washing.
The pH of the wash water for each treatment is lower than
the compost pH (by approximately 1 to 1.5 pH units). A few
measurements of compost pH before and after washing were
conducted and the results of these measurements are
indicated in Figures 5-3 a and b. An increase in compost pH
from 0.2 to 0.5 pH units was achieved during each washing,
thus, if the tower is washed routinely, compost pH can be
kept constant.
The effectiveness of the washing process on pH
stabilization depends on the quality of the water used and
the contact time between the compost and water. Both towers
#1 and #2 are washed by 10 L of DI water each time with a
flow rate of 1 L/min from the top of the tower. The water
is allowed to freely flow downward through the tower under
gravity, as a result, the water-compost contact time is
approximately 10 minutes.
Because the rate of pH decrease is proportional to the
H2S loading rate to the system, with a high H2S loading rate
the tower needs to be washed more frequently.
Addition of lime or CaC03 to eliminate acidification of
the biofilter is not effective and is not recommended.
There are some disadvantages in adding lime to the biofilter
system:


Pressure Drop (mmH0)
87
Gas Loading Rate (m3/m2-hr)
Figure 4-10. Pressure drop as a function of gas
velocity for different types of
compost.


102
and the compost water content is 60% by weight, providing a
compost pH of 1.57, then the dry specific acidity of this
compost is:
SAC = 20/[2(100-60)]xl051,57 = 673 (^q-H+/q).
Effect of Temperature on H2S Removal
The effect of temperature on H2S removal efficiency was
studied using Column System #4 with some minor
modifications. The modified system is shown in Figure 4-15.
Three columns (triplicate) packed with the same compost were
placed in a heating box. A rheostat was used to control the
temperature of the columns. The influent gas stream was
blown through a bubbler, which is placed in a water bath and
heated up to the same or slightly higher (5C) temperature
as the biofilter columns to saturate the gas stream at the
desired temperature. For tests carried out below room
temperature (22C), the reaction columns were placed in a
refrigerator and the temperature adjusted through the
refrigerator thermostat. The temperature range investigated
is between -1.5 to 103C. The results of the triplicate
measurements for each temperature point are plotted means
and 2 standard deviations as error bars in Figure 4-16.
In the range from 25C to 45C, high H2S removal
efficiencies are consistently observed with little
variation. The H2S removal efficiency, however, dropped
rapidly with decreasing temperature in the lower temperature


74
added to the flask and the material is subjected to
reduction and hydrogen sulfide is liberated, collected and
titrated as described above.
Water soluble sulfur
Water-soluble-S is determined by shaking 2-5 g of fresh
compost in DI water with a liguid to solid ratio of 10:1 for
30 minutes on a rotary shaker at a speed of 140 /min. An
aliquot of the compost extract is then subjected to
reduction, H2S absorption, and titration successively, as
described previously.
The following analyses are conducted in succession:
Sulfide sulfur
Sulfide-S or acid-volatile sulfur (AVS) is determined
by introducing 8 mL of 12N HC1 to the compost sample in the
reaction flask. Heat is applied after 10 min, the materials
are brought to boiling, and after 45 min the traps are
removed and the sulfide titrated.
Sulfate sulfur
The content of the reaction flask is filtered by a #42
Whatman filter. The filtrate is then subjected to
reduction. H2S is then trapped and titrated. This is an
alternative way of carrying out sulfate analysis. The
results are comparable to the summation of water soluble-S
and P-extractable-S.


6-5
Off-gas sampling locations on the biofilter beds
and concentrations of hydrogen sulfide observed
as a function of biofilter operating time 172
6-6 Concentration changes for hydrogen sulfide in gas
samples contained in Tediar bags as a function
of container holding time 173
6-7 Effect of varying purging time for sample
collection chamber prior to sampling on
measured hydrogen sulfide concentrations 175
6-8 Compost samples taken from Kanapaha Wastewater
Treatment Plant biofilter beds (2/5/91).
Left: sample taken from west bed. White color
indicates high sulfur accumulation.
Right: sample from east bed. Low sulfur content
compost, color is close to the original
(dark brown) 180
xix


169
Figure 6-4. Photograph of the biofilter system at
Kanapaha Wastewater Treatment Plant.


70
Figure 4-6.
Photograph of the sulfur distillation
assembly.


38
A portion of the hydrogen sulfide is adsorbed and/or
oxidized by the compost and the remaining H2S in the
effluent gas is absorbed by two H2S scrubbers in series
which contain 200 mL and 25 mL IN zinc acetate (ZnAc)
solution, respectively. The absorbing reaction used by the
H2S collectors is:
ZnAc + H2S - ZnSi + HAc
ti
H+ + Ac" (3-2)
Total flow in the system is measured by pre-calibrated
rotameters. H2S concentrations in the inlet gas to the
compost column are controlled by adjusting the flow rates of
mixing for the C02-free air and the pure H2S gas. Gas
samples from the influent gas stream are taken periodically
by gas-tight syringes, diluted with prepurified nitrogen
(N2) and analyzed for H2S content by a Tracor 250H analyzer
(See Chapter 4 for details) The effluent gas from the
filter columns is first passed through two scrubbers in
series containing IN ZnAc to absorb any H2S remaining in the
gas stream. Residual C02 in the effluent gas stream is
subsequently absorbed by 0.5N NaOH solution and titrated as
described previously.
Each compost sample is incubated at a desired H2S
concentration level for 24 hours. After the incubation
period the compost as well as the absorption solutions are
replaced by fresh compost samples and absorbing solutions
for operation at the next H2S concentration level. The


23
Table 2-2. Physiological characteristics of sulfur-
oxidizing bacteria.
Lithotrophic Electron
Donor
pH Range
for Growth
Thiobacillus Species
Growing Poorly in
Organic Media:
1. T. thioparus
H2S, sulfide, S, S2032-
6-8
2. T. denitrificans
h2s, s, s2o32~
6-8
3. T. neapolitanus
s, s2o32"
5-8
4. T. thiooxidans
s
2-5
5. T. ferrooxidans
S, sulfides, Fe2+
1.5-4
Thiobacillus Species
Growing Well in
Organic Media:
1. T. novellus
s2o32"
6-8
2. T. intermedius
s2o32"
3-7
Filamentous Sulfur
lithotrophs
Beaaiatoa
h2s, s2032
6-8
Thiothrix
h2s
6-8
Other Genera
Thiomicrosoira
6-8
Thermothrix
H2S, S2O3 f SO3
6.5-7.5
Sulfolobusa
h2s, s
1-4
a Archaebacterium.
Source: Brock and Madigan, 1988.


Table 3-2. Properties of selected composts before and after incubation.
Compost #1 Compost #2 Compost #3 Compost #6
Property Before After Before After Before After Before After
use
use
use
use
use
use
use
use
PH
8.67
6.79
8.13
8.44
9.22
8.72
7.26
7.55
Water (Wt%)
60.3
59.5
59.7
57.4
61.0
63.0
60.5
62.0
LOI (Wt%)
79.8
73.7
69.1
64.4
73.8
74.4
65.2
64.6
Water-P (mg/kg)
144
96.9
95.2
97.9
53.3
49.3
231
218
Total-C (Wt%)
36.9
39.6
36.8
35.3
39.5
40.9
39.5
36.2
Total-N (Wt%)
3.00
3.57
2.84
2.78
2.26
2.30
3.45
3.73
C/N
12.3
11.1
12.9
12.7
17.5
17.8
11.5
9.71
Metals (mg/kg)
Ca
37400
40300
57500
60000
66500
65000
47700
49800
Mg
4120
4470
5250
5300
3840
4250
5400
5950
Zn
492
557
651
638
81.0
94.5
897
964
Cu
201
194
125
131
8.50
9.50
114
113
Mn
30.0
36.0
66.5
70.0
23.5
25.0
48.5
52.0
Fe
8820
9760
5220
5130
512
541
6740
7260


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Professor of Agricultural
Engineering
This dissertation was submitted to the Graduate Faculty
of the College of Engineering and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May 1992
iLuj &
Winfred M. Phillips
Dean, College of Engineering
Madelyn M. Lockhart
Dean, Graduate School


194
(NIOSH) "Criteria for a Recommended Standard for the
Occupational Exposure to Hydrogen Sulfide," U.S. Department
of Health, Education and Welfare, NIOSH, May 1979.
O'Brien, D.J. and Birkner F.B. "Kinetics of Oxygenation of
Reduced Sulfur Species in Aqueous Solution," Environmental
Science & Technology, 11: 1114 (1987).
Ottengraf, S.P.P. Theoretical Model for a Submerged
Biological Filter," Biotechnology and Bioengineering, 19:
1411 (1977).
Ottengraf S.P.P. and Van Den Oever, H.A.C. "Kinetics of
Organic Compound Removal from Waste Gases with a Biological
Filter," Biotechnology and Bioengineering, 25: 3089 (1983).
Ottengraf, S.P.P.; Van Den Oever, A.H.C. and Kempenaars,
F.J.C.M. "Waste Gas Purification in a Biofilter Bed," in
Innovations in Biotechnology. Houwink, E.H. and Van Dan
Meer, R.R. Eds, Elsevier Science Publishers B.V., Amsterdam
(1984) .
Ottengraf, S.P.P.; Meesters, J.J.P.; Van Den Oever, A.H.C.
and Rozema, H.R. "Biological Elimination of Volatile
Xenobiotic Compounds in Biofilters," Bioprocess Engineering
l: 61 (1986).
Ottengraf, S.P.P. "Exhaust Gas Purification," in
Biotechnology. Rehm, H.J. and Reed, G. Eds, Vol.8; VCH
Verlagsgesellschaft., Weinheim, 1986.
Ottengraf, S.P.P. "Biological Systems for Waste Gas
Elimination," TIBTECH 5: 132 (1987).
Painter, D.E. Air Pollution Technology. Reston Publishing
Company, Inc., Reston, VA, 1974.
Parker, H.W. Air Pollution. Prentice-Hall, Inc., NJ, 1977.
Patterson, Jr., A. and Thomas, H.C. A Textbook of
Quantitative Analysis. Henry Holt and Company, New York,
1952.
Paul, P.G; Castelijn, F.J. "BiofiltrationA Relatively
Cheap and Effective Method of Waste Gas Treatment," in
Environmental Technology. Proceedings of the second European
Conference on Environmental Technology, Amsterdam. The
Netherlands. June 22-26. 1987. De Waal, K.J.A. and Van Den
Brink, W.J., Eds, Martinus Nijhoff Publishers, Boston, 1987.
Piscarcyzyk, K. "Odor Control with Potassium Permanganate,"
Presented at Ohio Water Pollution Control Conference,
Dayton, OH, June 16-18, 1982.


Table 6-2. Summary of periodic Kanapaha biofilter bed compost analyses during
operational period from 5/10/88 to 2/5/91.
Date
Bulk
Density
(g/cc)
Water
Content
(%)
Organic
Matter pH
(%)
Total
N
(%)
Total
C
(%)
Total
S
(mg/g)
Water
Ext. P
(Mg/g)
Particle
Size Distri.(Wt%)
>3.35mm
Interm.
<2.36mm
05/10/88
0.23
52.1
68.1 8.63
3.45
39.5
7.3
231
41.0
6.6
52.4
11/21/88
0.27
54.6
69.3 4.40




43.8
12.9
43.3
05/16/90
0.30
45.6
66.5 2.62
1.89
34.3
44.5
140
47.0
6.1
46.8
12/20/90
0.27
54.6
66.5 1.60
1.58
36.6
71.0
294
45.9
6.7
47.4
02/05/91
57.2
64.7 1.80
109

168


123
controlled at intermediate H2S concentrations. With
further lower H2S concentrations, the profile becomes curved
well below the straight line, and suggests that the dominant
mechanism is diffusion limited in this first-order kinetics
region.
Table 4-6 summarizes the equations describing the
kinetics of H2S oxidation in the biofilter system. These
equations allow for a quantitative description of the basic
processes involved in this biofiltration elimination of H2S
and they allow for an accurate sizing of biofilters for H2S
removal. It should be noted that the macrokinetics of H2S
oxidation in biofilters as well as the reaction coefficients
reported here are related to operational conditions. The
kinetics are valid for a similar compost biofilter system
operating under similar conditions. If operational
conditions such as pH, temperature, sulfur content, etc. are
changed, then the kinetic behavior may be altered. In
practice, the kinetic behavior for a particular compound
should be determined by laboratory or pilot scale studies
(Van Lith, 1989; Leson and Winer, 1991).


125
55
Q15
8
8
8
pii
po c
T15
e-
TS14:
-
TS13;
-
TS12!
--
TS11:
-
Q14
G13
3
8
Q12
i
8
Q11
H
8
Qio ^
no O 3
8
8
T = Temperature
P = Pressure
G = Gas Sample
S = Solid Sample
. Sampling and measurement ports on
towers.
a. Tower #1
Figure 4-2


Hydrogen Sulfide Removal Efficiency (%)
104
Figure 4-16. Effect of temperature on H2S removal
efficiency.


131
the desired range. The main purpose of this procedure,
however, is to reduce the acidity and prevent accumulation
of sulfate in the compost. A typical compost water content
distribution profile in Tower #2 is presented in Figure 5-2.
Samples were taken and analyzed before washing (14 days
after the last washing) and 1 hour after washing the tower.
It can be seen that 1) the water content of the compost
is evenly distributed along the length of the bed, and 2)
the compost has very good water retention and buffering
capacity. Only 3-5% of the compost water content was lost
during the 14 day interval between washings. When the
biofilter tower was operated for 130 days and samples were
taken and analyzed, the water holding capacity of the
compost in the inlet region had decreased slightly compared
to that in the outlet region.
The system showed a good buffering capacity to gas
surface loading changes. No significant reduction in H2S
removal efficiency was observed when the gas surface loading
rate was varied in the range 20 to 500 m3/m2-hr for the
same H2S loading rate.
Also, the buffering capacity for H2S loading rate or
H2S concentration changes was very good. Under the same
conditions of gas surface loading rate ( 100 m3/m2-hr), when
H2S concentrations were changed from 15 ppmv to 775 ppmv
(corresponding to the H2S loading rate being changed from 2
to 50.5 g-S/m3-hr) the H2S removal efficiency did not vary
from 99.5%. Sudden changes in H2S loading rates over a


9
H2S + H2S04 = S02 + S + 2H20 (2-3)
Hydrogen sulfide also burns in air to form sulfur
dioxide and water:
2H2S + 302 = 2S02 + 2H20 (2-4)
Table 2-1 summarizes the physical and chemical
properties and the odor threshould of H2S.
Toxicity of HoS
Hydrogen sulfide is almost as toxic as hydrogen cyanide
(HCN), which is used in prison gas chambers (Parker, 1977).
Human exposure to small amounts of H2S in air can cause
headaches, nausea, and eye irritation, and higher
concentrations can cause paralysis of the respiratory
system, which results in fainting and possible death.
Concentrations of the gas approaching 0.2 percent (2000
ppmv) are fatal to humans after exposure for a few minutes
(NRC, 1979).
Hydrogen sulfide has a characteristic rotten egg smell
at low concentrations. But as levels of H2S increase, a
person's ability to sense dangerous concentrations by smell
is quickly lost. If the concentration is high enough,
unconsciousness will come suddenly, followed by death if
there is not a prompt rescue.
The Occupational Safety and Health Administration
(OSHA) has established limits for work place exposure to H2S


H2S Concentration (ppmv)
173
Days After Sampling
Figure 6-6. Concentration changes for hydrogen
sulfide in gas samples contained in
Tediar bags as a function of container
holding time.


Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE
By
Yonghua Yang
May 1992
Chairman: Dr. Eric R. Allen, Professor
Major Department: Environmental Engineering Sciences
A laboratory scale biological filter system for control
of hydrogen sulfide (H2S) emissions has been developed and
optimum design and operating parameters are evaluated. This
biofiltration system uses yard waste compost as filter
material, and the performance of the system for controlling
waste gas containing H2S is evaluated through long term
operation.
Extensive tests have been conducted to determine the
effect of various filter bed operating parameters such as
pH, temperature, pollutant retention time, pressure drop,
water content, etc. on H2S removal efficiencies.
A biofilm model is used to characterize the macro
kinetics of the biofiltration process. Models for the
kinetics of H2S oxidation are developed that allow one to
quantitatively predict the performance of the biofilter.
X1XX


CHAPTER 3
FILTER MATERIALS AND THEIR DECOMPOSITION
UNDER AEROBIC CONDITIONS
Introduction
Biofiltration systems or biofilters employ physical,
chemical and biological processes such as adsorption,
absorption and microbial digestion and oxidative degradation
to remove air pollutants from waste gas streams. Microbial
degradation and oxidation of the pollutants, however, appear
to be the primary removal mechanisms within a biofilter. In
the biodegradation process, pollutants are consumed by the
microorganisms, providing an energy source or essential
nutrients and are converted usually to, less harmful
compounds. The filter materials used, on the other hand,
must provide the proper environment for microbial growth and
contain materials on which the microbes can feed to ensure
that the microbial population can develop and survive.
The effectiveness of a biofilter material depends on
its physical, chemical and biological characteristics. The
lifetime of a biofilter material mainly depends on its rate
of carbon (C) and nitrogen (N) mineralization. When
available C and N in the filter material are no longer
sufficient to support the microbial population in the
system, then the material is no longer suitable as a
25


Solubility (g-S/L)
7
Figure 2-1. Solubility of H2S in water at 1 atm.
Data adopted from Piscarcyzyk, 1982.


149
introduced to the system at a flow rate of 230 mL/min
(Equivalent to 15 m3/m2-hr surface loading rate). Influent
and effluent gas samples were analyzed to determine the
original performance of the compost before treatment. The
compost is then unpacked and divided into 4 groups, each
group was packed into two columns as duplicates. The first
group was treated with DI water, the second group with 0.05M
NaOH, third group with 0.05M NaHC03 and the fourth group was
left untreated as a control. For each treated group, the
compost was shaken with 10 times the liquid (by weight) for
30 minutes with a rotary shaker. The composts are treated
this way twice. After each shaking treatment, the compost
pH was measured. The total-S of the original defective
compost and the composts after treatment were measured. The
results of these compost analyses are summarized in Table 5-
2. It can be seen that after being washed twice, the pH of
water treated compost was raised 0.3 pH unit, but both the
NaOH and NaHC03 treated composts are neutralized to near pH
7. The total-S content of the compost was successfully
reduced by 72 to 85% in all cases. According to this study,
water appears to be the best washing agent.
After treatment, the composts are repacked into the
columns, and the system subjected to an H2S removal test.
The system was allowed to operate continuously for one week
before gas sampling and analysis was conducted. The delay in
testing was included to eliminate the effects of residual
alkali on H2S removal for those composts treated by NaOH and


27
medium:
1). Density: Too dense material may contain a large
fraction of inorganic materials such as stone and
sand which are unsuitable as carbon and energy
sources for microbial growth.
2). Structure: Structure of the medium will affect
the uniformity of the filter load. Too large
sized materials should be avoided because the
surface-to-volume ratio will be reduced.
3). Particle Size Distribution: Too small particles
affect the pressure drop by compacting and
restricting the gas flow.
4). Pore Volume (void fraction): This property
determines the total surface area available for
reaction, also it will affect pressure drop.
5). Organic Matter Content: The organic matter
controls the microbial population and the useful
service life of the filter media.
6). pH Value: pH will affect the nature and level of
the microbial population and activity.
7). Water Retention Capacity: This property will
determine the consistency in liquid water content
of the filter material, and
8). Economics: Reasonable Capital and operating
expenditures.
All of these requirements can be met by selecting
suitable filter materials. Many kinds of filter materials
have been used in biofiltration applications. Examples
include field soils, compost, peat, bush, clay, volcanic
ash, sand, bark and a combination of such materials (Rands
et al., 1981; Prokop and Bohn, 1985; Terasawa et al., 1986
Frechen and Kettern, 1987). The performance of these
materials, however, can be very different due to the
diversity of their physical and chemical properties. Compost
has been considered to be the best choice for filter


13
professionals in this country appear to be aware of the
'biofiltration1 process and its applications.
Although there are some applications of biofiltration
in the U.S. and some technical papers have been published in
the English language, most of the research and development
work on biofiltration has been conducted in Europe and the
majority of the recent research data have been published in
the German language. Excellent reviews of previous
biofiltration work have been published by Hartenstein
(1987), Leson and Winer (1991), Ergas et al. (1991), and
Dharmavaram (1991).
History and Development
The first deodorization method based upon the use of a
soil bed in the U.S. was developed and patented by Pomeroy
in 1957. Later, Pomeroy (1982) described the deodorization
of waste gases emitted from sewer lines by a soil bed system
used in Los Angeles in 1957. The microbiological degradation
of sulfur-containing gases in the filter bed was observed to
be effective in these studies.
Other early applications of biological treatment of
odorous gases include a soil bed system built in Nrnberg,
West Germany, in 1959 and biofilters built in Geneva,
Switzerland, and Mercer Island, Washington, to remove odors
from wastewater treatment and compost manufacturing,
respectively, in the mid-1960s (Bohn and Bohn, 1987).


I
Reducing
Mixture
Reduction
H2S
Absorption
I
lodometric
Titration
I
Reducing
Mixture
Reduction
I
H2S
Absorption
I
lodometric
Titration
I
Total-S
Water-Soluble-S
(Sulfate-S)
Gas
H2S
Absorption
I
lodometric
Titration
I
Sulfide-S
Figure 4-7
Flow chart of
FRESH WET COMPOST
Store for
Later Use
Hydriodic
Acid
Reduction
Filtration
Precl-
Acetone
Extraction
Preci-
CrCI2
Reduction
Preci-
ptate
pitate
pitate
| Solution
1
Gas
Reducing
Mixture
Reduction
Gas
CrCI2
Reduction
| Gas
H2S
H2S
H2S
H2S
Absorption
Absorption
Absorption
Absorption
1
1
1
lodometric
lodometric
lodometric
lodometric
Titration
Titration
Titration
Titration
1
Sulfate-S
1
Elemental-S
Pyrite-S
1
Ester-S
the sulfur analysis procedures for compost


BIOGRAPHICAL SKETCH
Yonghua Yang was born on March 24, 1949, in Inner
Mongolia, China, and attended local schools until completion
of high school in 1968 He received his Bachelor of
Engineering (equivalent) degree in chemical engineering from
Dalian Institute of Technology, China, in 1977. After
graduation, he worked for 8 years for the Environmental
Protection Institute, Baotou Iron and Steel Corporation in
China as an environmental engineer.
He was accepted as a graduate student in fall 1986 and
received his Master of Engineering degree in air pollution
from the University of Florida, Gainesville, FL, in 1988. He
continued graduate study to pursue the Doctor of Philosophy
degree in the Environmental Engineering Sciences Department,
University of Florida from 1989 through 1991.
He was the recipient of the Axel Hendrickson
Scholarship award from the Air & Waste Management
Association (AW&MA), Florida Section in 1990 and a graduate
scholarship award from AW&MA, in 1991.
199


Percent
8
pH
Figure 2-2. Effect of pH on H2S equilibrium.
Source: Sawyer, 1967.


133
very large range, for example, from a few tens ppmv to a few
hundreds ppmv may cause a temporary reduction in H2S removal
efficiency. However, after only a few hours, the optimum
control efficiency was recovered (Figure 5-1 a and b). This
perturbation is probably due to the uneven and inadequate
initial distribution of the sulfur oxidizing bacteria
population. In other words, the length of the 'active
portion' of the filter is related to the H2S concentration
in the gas stream. This feature will be discussed in a
later section.
One of the most significant changes observed is the
compost pH. Compost pH changes in different sections of
Towers #1 and #2 are shown in Figures 5-3 a and b,
respectively. The product of H2S oxidation is sulfuric acid
(H2S04). This strong acid is soluble in water and
accumulates in the compost, resulting in rapid acidification
of the system. Compost acidity increases very rapidly with
time, e.g. after 32 days of operation, the pH of the bottom
section (inlet) of the compost dropped from 8.0 to 1.5
(TS11, Figure 5-3a). The rate of compost pH change is
proportional to the H2S input. The pH drop in the lower
portion of the compost is much larger than that in the upper
portion of the bed since most of the H2S oxidation reaction
takes place in the former region. When the gas stream flows
through the bed less H2S is left in the gas stream and less
H2S04 is formed as the bed is traversed, therefore, the
acidification is slower in the upper portion of the tower.


188
Although hydrogen sulfide has been selected as a test
gas for this research, the results obtained through this
study should provide for an improved quantitative
understanding of the principles of biofiltration.
Controlling variables in the operation and effectiveness of
biofiltration waste gas control technology may be applicable
not only to hydrogen sulfide but also to other sulfur-
containing compounds, and more generally, to air toxics and
VOCs. Dissemination of this basic information on the
multiple advantages of biofiltration, including low cost,
high destruction-efficiency, energy conservation, ease of
operation and maintenance and universal application will
provide environmetal engineers and federal, state and local
government air pollution control officials with a viable
alternative in controlling emissions of air toxic compounds
from commercial and industrial sources. Biofiltration
control of waste gas streams is a relatively unknown and
little explored control technology in the U.S.. It has the
potential, however, for wide-spread application and
acceptance because of its relative simplicity and low
capital and operating cost, in addition to its great
potential for indiscriminate effectiveness in controlling
multiple pollutants.


182
Conclusions
In spite of the few limitations described, for a full
scale system, biofiltration has been demonstrated to be a
simple, effective and inexpensive method for odor control
at wastewater treatment plants based on the Kanapaha plant
experience. The biofilter system described was successfully
operated at high H2S removal efficiencies with little or no
maintenance for a period of 2.5 years. The system was
effective until the corrosion problem occurred and the gas
distribution system had to be reconstructed. During the
2.5-year operation period, no odor was noticeable even close
to the filter beds. Also, the City of Gainesville did not
receive a single odor complaint during this period. More
than $200,000 per year has been saved in chemicals that were
originally used to provide alternative odor control systems
for the plant. These cost savings have resulted in a one-
year payback on the biofilter system capital and operating
costs (IPS, 1990).
The Kanapaha biofilter experience suggests that
routine monitoring and maintenance are necessary to ensure
proper operation conditions for effective, long term control
of H2S emissions. It was recognized that further research
was needed to solve existing problems, such as progressive
system acidification, accumulation of sulfur in the filter
medium and the eventual decline of H2S removal efficiency.
Appropriate laboratory studies have been conducted at the


Hydrogen Sulfide Removal Efficeincy (%)
110
Figure 4-18. Effect of nutrient addition on H2S
removal.
Total-S content in compost (mg-S/g)
A: 17.5
B: 33.7
C: 20.2
D: 120


198
Yang, Y. and Allen, E.R. "Biofiltration Control of Odor
Emissions in Wastewater Treatment Plants," Paper presented
at the 201st National Meeting of the American Chemical
Society, Atlanta, GA, April 14-19, 1991.
Zhabina, N.N. and Volkov, I.I. "A Method for Determination
of Various Sulfur Compounds in Sea Sediments and Rocks," in
Environmental Biogeochemistrv and Geomicrobioloav. Krumbein
Ed., Ann Arbor Sci., Ann Arbor, MI, 1978.


166
disturbances from spurious ambient air currents near the bed
surface. The probe is connected to a Nutech Model 218
integrated gas sampler by a 7.5 meter length of 6.4 mm
Teflon tubing. Sampling is accomplished by evacuating the
dead space between the inner wall of the canister and the
outer walls of the Tediar bag at a constant rate. The purge
line pump is allowed to run for 5 minutes to flush the
sample line and allow for off-gas equilibrium to be
established inside the sampling cover. The sampling flow
rate is set at about 1 liter per minute (Lpm).
Off-gas samples are collected on the top surface of the
twin biofilter beds at four locations. Each rectangular bed
was divided into two equal-area triangles by drawing
diagonals. The sampling locations were selected at the
centroids of each of the four triangular areas. Analytical
results for these four off-gas samples were averaged later
to obtain a typical off-gas concentration.
The gas samples were transported in the Tediar sampling
bags to the University of Florida (UF) laboratories and
analyzed within eight hours of collection using a Tracor
250H Analyzer.
Results and Discussion
Early in August 1988, the west half of the biofilter
bed was filled up to a grade of 1.3 meters with compost
obtained from Pompano Beach, Florida. The biofilter system
was tested by operating the system at half-filter bed


150
Table 5-2. Effect of washing on compost pH and sulfate
content by DI water, NaOH and NaHC03
solutions.
Reduction
Treatment
Time of
Treatment
pH
Total-S
(mg-S/g)
of Total
(%)
Untreated
(control)
1.59
120

Water
First
1.61


Washed
Second
2.20
17.5
85.4
NaOH
First
1.73


Washed
Second
7.51
33.7
71.9
NaHCO-,
Washed
First
1.86


Second
6.66
20.2
83.1


4
a microbial biofilter system for removal of H2S from waste
gas streams and determine the operating parameters necessary
to optimize the performance of such a biofilter system.
The objectives were achieved through the following
studies:
1. Evaluation of the properties of filter materials and
their decomposition characteristics under aerobic
conditions.
2. Evaluation of the effects of design and operational
parameters on H2S removal efficiencies on laboratory
scale biofilter systems. Variables evaluated
included temperature, pH, compost water and sulfate
content, H2S elimination capacity, pollutant
retention time, etc..
3. Determination of the predictive relationships for
H2S control efficiencies through chemical kinetic
studies.
4. Evaluation of system performance and determination
of optimum maintenance procedures for biofiltration
control of H2S during long term operation.
5. Evaluation of the field performance of a full scale
biofiltration system for control of H2S emissions at
a local waste water treatment plant.
The research reported here focuses mainly on the
utilization, improvement and optimization of a compost
biofiltration tower system. Optimization of this system
has been directed toward the best achievable control of


C0 Evolved (mg-CC>2/g of C added)
40
Time (day)
B Compost #1+ Compost #2 o Compost #3 a Compost #6
Figure 3-3. Plot of C02 evolution from composts
during the 122 day incubation.


56
8
P 21
P20
-e-
T26 !
TS25;
-e-
TS24;
TS23
TS22 ;
TS21 I
<-
T20
"O'
G25
3-
G23
B
G24
G22
B
G21
B
G20
8
T = Temperature
P = Pressure
G = Gas Sample
S = Solid Sample
8
8
Figure 4-2. Sampling and measurement ports on
towers,
b. Tower #2


BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE
BY
YONGHUA YANG
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1992
UNIVERSITY OF FLORIDA LIBRARIES

To my father for making this possible
and
In memory of my mother

ACKNOWLEDGEMENTS
I would like to express my sincerest gratitude and
appreciation to the following people who made this research
possible:
To Dr. E. R. Allen, the doctoral committee chairman,
for his foresight in support of this research, his
encouragement, guidance and invaluable input during the
course of this study and my graduate work.
To Drs. D. A. Lundgren, B. Koopman, K. R. Reddy and D.
P. Chynoweth for their interests in this research, helpful
suggestions and participation on my graduate committee.
To Dr. P. Urone for his friendship, kindness and
valuable suggestions.
To Mr. A. White and the Kanapaha Wastewater Treatment
Plant engineers for their assistance in the research on the
full scale biofilter system.
To Ms. Yu Wang for her help on compost analysis.
To Ms. S. Jordan for her help in construction and set
up of the Lab scale biofilter units.
To Mr. R. Vanderpool for his invaluable friendship, his
ideas and his help in all aspects of my work that have made
my years at the university much easier and so enjoyable.
To my wife, Li, for her continuing support,
encouragement, patience and understanding.
iii

TABLE OF CONTENTS
Page
ACKNOWLEDGEMENTS iii
LIST OF TABLES vii
LIST OF FIGURES ix
ABSTRACT xiii
CHAPTERS
1 INTRODUCTION 1
2 BACKGROUND 6
Properties of Hydrogen Sulfide and Regulations .. 6
Physical and Chemical Properties 6
Toxicity of H2S 9
Sources of H2S Emissions and Regulations 11
Biofiltration as an Air Pollution
Control Technology 12
History and Development 13
Applications 15
Theoretical Basis 16
Biological Oxidation of Hydrogen Sulfide 21
3 PROPERTIES OF COMPOSTS AND THEIR DECOMPOSITION .. 25
Introduction 25
Selection of Filter Materials 26
Decomposition of Composts under Aerobic
Conditions 28
Materials and Methods 32
Results and Discussion 39
Decomposition of composts 39
Effect of H2S on compost decomposition 45
Conclusions 50
4 DETERMINATION OF THE DESIGN AND OPERATIONAL
PARAMETERS FOR BIOFILTER SYSTEMS 52
Introduction 52
System Design and Construction 52
IV

The Dual Tower System 52
Portable Tower #3 58
Column System #4 58
Measurement Methods 61
Temperature 61
Pressure Drop 61
Gas Flow Rate 61
Sampling Methods 62
Compost Samples 62
Gas Samples 62
Water Samples 65
Compost Analysis Methods 65
Water Content 65
pH 66
Total Carbon and Total Nitrogen 66
Water Soluble Phosphorus (WSP) 66
Acid-extractable Cations 67
Particle Size Distribution 67
Porosity 67
Organic Matter 67
Particle Density 68
Bulk density 68
Sulfur Analysis Methods 68
Sulfur in Compost 68
Total Sulfur 73
Water soluble sulfur 74
Sulfide sulfur 74
Sulfate sulfur 74
Elemental sulfur 75
Pyrite sulfur 75
Organic sulfur 76
Sulfur in the Aqueous Phase 76
Sulfate sulfur 76
Sulfide sulfur 77
Total-S 77
Sulfur in Waste Gas 77
Results and Discussion 79
Pressure Drop 79
Effect of Gas Retention Time on H2S Removal ... 88
Effect of Concentration of H2S on Its Removal 90
Effect of H2S Loading Rate on Its Removal 92
Effect of Compost Water Content
on H2S Removal 93
Effect of Compost Acidity on H2S Removal 97
Effect of Temperature on H2S Removal 102
Effect of Sulfate on H2S Removal 106
Effect of Nutrient Addition on H2S Removal .... 108
Kinetics of H2S Oxidation in the Biofilter 109
Theoretical considerations 109
Determination of the kinetics of H2S
Oxidation in a biofilter 114
v

5 BIOFILTER PERFORMANCE AND CHANGES OF COMPOST
PROPERTIES ASSOCIATED WITH LONG TERM OPERATION 125
Overall Performance of the biofilters 125
Accumulation of Sulfur in Compost and Its
Effect on System Performance 137
System Upset and Recovery 146
Selection of Chemical Solutions 148
Effect of Water-Compost Contact Time on S042-
Leaching Efficiency 153
Effect of Water to Compost Ratio on S042
Leaching Efficiency 153
6 FULL SCALE APPLICATION OF BIOFILTRATION TO
CONTROL H2S EMISSIONS AT A WASTEWATER TREATMENT
PLANT 158
Introduction 158
System Design and Construction 160
Sampling and Analysis Methods 163
Results and Discussion 166
Conclusions 182
7 SUMMARY AND CONCLUSIONS 184
REFERENCES 189
BIOGRAPHICAL SKETCH 199
vi

LIST OF TABLES
Table Page
2-1 Physical and chemical properties of H2S 10
2-2 Physiological characteristics of sulfur-oxidizing
bacteria 23
3-1 Description of compost used for this study 29
3-2 Properties of selected composts before and after
incubation 33
4-1 Retention times, limits of detection and operating
conditions for the Tracor 250H analyzer 78
4-2 Summary of initial compost properties 80
4-3 Particle size range distribution for selected
composts 82
4-4 Effect of gas retention time on H2S removal
efficiency 89
4-5 Effect of H2S concentration on removal efficiency 91
4-6 Models for the kinetics of H2S oxidation in
biofilter 124
5-1 Sulfur fractionation of original compost #17A and
compost at different heights in the filter .... 141
5-2 Effect of washing on compost pH and sulfate
content by DI water, NaOH and NaHC03
solutions 150
5-3 Performance of defective compost before and
after treatment 152
5-4 Effect of water washing on elimination of sulfate
in filter compost 157
6-1 Summary of Kanapaha biofilter bed design and
operation parameters 164
vii

6-2 Summary of periodic Kanapaha biofilter bed compost
analyses during operational period from
5/10/88 to 2/5/91 168
6-3 Summary of Kanapaha biofilter influent and
effluent gas sample analyses during
three week start-up period 170
6-4 Gas sampling and analysis for Kanapaha
biofilter bed, 2/5/91 178
6-5 Sulfur fractionation of a typical compost sample
in Kanapaha biofilter bed 181
viii

LIST OF FIGURES
Figure Page
2-1 Solubility of H2S in water at 1 atm 7
2-2 Effect of pH on H2S Equilibrium 8
2-3 Biophysical model for the biological filter bed.
The concentration profiles shown in the
biofilm refer to: 1) Reaction limitation,
2) Diffusion limitation 18
2-4 Steps in the oxidation of different compounds by
thiobacilli. The sulfite oxidase pathway is
thought to account for the majority of sulfide
oxidized 24
3-1 Schematic drawing of the experimental arrangement
for the study of compost decomposition 34
3-2 Schematic drawing of the experimental arrangement
for the investigation of the effect of H2S
exposure on compost decomposition 37
3-3 Plot of CO2 evolution from composts during the
122 day incubation 40
3-4 Decomposition stages and reaction rate
coefficients for the four composts studied .... 44
3-5 Effect of H2S exposure on the rate of compost
decomposition as measured by C02 respiration .. 46
3-6 Plot of CO2 evolution as a function of square
root of H2S concentration 48
4-1 Schematic drawing of the dual tower system 53
4-2 Sampling and measurement ports on towers.
a. Tower #1 55
b. Tower #2 56
4-3 Schematic drawing of Tower #3 59
IX

4-4 Schematic drawing of column system #4 60
4-5 Schematic drawing of the gas sampling assembly ... 64
4-6 Photograph of the sulfur distillation assembly ... 70
4-7 Flow chart of the sulfur analysis procedures
for compost 72
4-8 Pressure drop as a function of particle size
range for different gas velocities 84
4-9 Pressure drop as a function of packing height
for different compost particle size range 86
4-10 Pressure drop as a function of gas velocity for
different types of compost 87
4-11 Determination of maximum H2S elimination
capacity of compost 94
4-12 Effect of compost water content on H2S removal
efficiency 96
4-13 Time required for dried compost to recover
optimum efficiency 98
4-14 Effect of compost pH on H2S removal efficiency.
Condition a: H2S loading rate: 10.5 g/m3-hr
Gas loading rate: 15 m3/m2-hr
Condition b: H2S loading rate: 35.4 g/m3-hr
Gas loading rate: 26.1 nr/m-hr... 100
4-15 Schematic drawing of the experimental arrangement
for investigation of the effect of temperature
on H2S removal efficiency 103
4-16 Effect of temperature on H2S removal efficiency .. 104
4-17 Effect of sulfate on H2S removal efficiency 107
4-18 Effect of nutrient addition on H2S removal.
Total-S content in compost (mg-S/g) A: 17.5;
B: 33.7; C: 20.2; D: 119.7 110
4-19 Linear least squares regression analysis for zero-
order kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 m3/m-hr, compost #17 ... 115
4-20 Linear least squares regression analysis for first-
order kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 m3/m -hr, compost #17 ... 116
x

117
4-21 Determination of the fractional-order reaction rate
coefficient, kf by linear least squares
regression. Gas loading rate: 224 m3/m2-hr,
compost #17
4-22 Plot showing the fractional-order kinetics of H2S
oxidation in biofilter. Gas loading rate:
224 nr/nr-hr, compost #17 118
4-23 Concentration profiles for H2S as a function of
packing height within the biofilter. Gas loading
rate: 224 m3/m2-hr, compost #17 119
5-1 Biofilter control of H2S during long term
operation, a) Tower #1, compost #17A 126
b) Tower #2, compost #17 127
c) Tower #3, compost #16 128
5-2 Compost water content profile 132
5-3 pH changes of compost in different sections of the
biofilter with operation time.
a) Tower #1, compost #17A 134
b) Tower #2, compost #17 135
5-4 Total-S distribution profile in biofilter,
Tower #1, after exposure to H2S for 100
days 14 3
5-5 H2S removal efficiencies in different regions of
the biofilter, Tower #2. A: 0-0.2 m,
B: 0.2-0.4 m, C: 0.4-0.6 m, D: 0.6-0.8 m,
E: 0.8 1.0 145
5-6 Effect of water-compost contact time on sulfate
leaching efficiency 154
5-7 Effect of water/compost ratio on sulfate leaching 155
6-1 Schematic diagram of the Kanapaha biofilter bed
system 161
6-2 Photograph of the grit chamber at Kanapaha
Wastewater Treatment Plant (top view). The
chamber is covered to collect the
malodorous gas 162
6-3 Biofilter off-gas sampling system 165
6-4 Photograph of the biofilter system at Kanapaha
Wastewater Treatment Plant 169
xi

6-5
Off-gas sampling locations on the biofilter beds
and concentrations of hydrogen sulfide observed
as a function of biofilter operating time 172
6-6 Concentration changes for hydrogen sulfide in gas
samples contained in Tediar bags as a function
of container holding time 173
6-7 Effect of varying purging time for sample
collection chamber prior to sampling on
measured hydrogen sulfide concentrations 175
6-8 Compost samples taken from Kanapaha Wastewater
Treatment Plant biofilter beds (2/5/91).
Left: sample taken from west bed. White color
indicates high sulfur accumulation.
Right: sample from east bed. Low sulfur content
compost, color is close to the original
(dark brown) 180
xix

Abstract of Dissertation Presented to the Graduate School
of the University of Florida in Partial Fulfillment of the
Requirements for the Degree of Doctor of Philosophy
BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE
By
Yonghua Yang
May 1992
Chairman: Dr. Eric R. Allen, Professor
Major Department: Environmental Engineering Sciences
A laboratory scale biological filter system for control
of hydrogen sulfide (H2S) emissions has been developed and
optimum design and operating parameters are evaluated. This
biofiltration system uses yard waste compost as filter
material, and the performance of the system for controlling
waste gas containing H2S is evaluated through long term
operation.
Extensive tests have been conducted to determine the
effect of various filter bed operating parameters such as
pH, temperature, pollutant retention time, pressure drop,
water content, etc. on H2S removal efficiencies.
A biofilm model is used to characterize the macro
kinetics of the biofiltration process. Models for the
kinetics of H2S oxidation are developed that allow one to
quantitatively predict the performance of the biofilter.
X1XX

Decomposition of composts under aerobic conditions and
the effect of H2S concentrations on the decomposition rate
are quantitatively determined. The half life time for the
composts tested is estimated to be between 3.3 and 6.1 years.
Hydrogen sulfide loading rate and maximum H2S
elimination capacity of the filter material are emphasized as
important design parameters. The maximum H2S elimination
capacity of a typical yard waste compost is determined to be
130 g-S/m3-hr under optimized conditions.
Hydrogen sulfide is oxidized to sulfuric acid in the
biofilter system, where biological oxidation plays a major
role. Acidification of the biofilter system and accumulation
of sulfate in the filter material are determined to be
natural features of the oxidation process, where the latter
is toxic for biological activities of the microorganisms.
Appropriate methods have been developed to effectively
mitigate this affect.
System 'upset1 is identified as being due to compost
dry-out and system overloading. Methods have been identified
to provide for recovery of the defective filter material.
Operation of a full scale biofilter system at a
wastewater treatment plant has been investigated. Both the
laboratory and full scale systems have demonstrated excellent
performance over substantial operational periods. Hydrogen
sulfide removal efficiencies of 99.9+% have been constantly
achieved when the H2S inlet concentrations are varied from 5
to 2650 ppmv.
xiv

CHAPTER 1
INTRODUCTION
Hydrogen sulfide (H2S) is a highly toxic air pollutant
which has been identified in the list of 190 air toxic
substances in Title III of the 1990 Amendments to the Clean
Air Act.
Considerable amounts of H2S are produced in association
with industrial processes, such as petroleum refining,
rendering, waste water treatment, paper and pulp
manufacturing, food processing, and in the treatment of
"sour" natural gas and other fuels. Hydrogen sulfide is
frequently the main component of most observable odorous
emissions.
Hydrogen sulfide is an odorous gas, and its presence at
low concentrations is easily perceived and recognized due to
its characteristic odor of rotten eggs. Hydrogen sulfide is
perceptible to most people at concentrations in excess of
0.5 parts per billion (ppb) in air. Control of H2S
emissions is essential to protect public health and welfare
as well as to mitigate vegetation and material damage
problems.
Numerous processes involving physico-chemical
principles have been developed in order to effectively
remove hydrogen sulfide from air, waste gases and liquids
1

2
(Bethea et al., 1973; Ferguson, 1975; USEPA, 1985; Lalazary
et al., 1986; Walker et al., 1986; Lindstrom, 1990).
Processes that have been used to remove H2S from waste gas
streams involve either physical treatment or chemical
oxidation. Some methods require addition of chemicals, and
energy expenditure is usually necessary for physical
treatment. Additional environmental problems are
encountered with chemical additions, where resulting
products and by-products require further treatment and
disposal.
Biofiltration can provide for a universal, simple,
economicly feasible, and efficient pollutant-destructive
control technology for a variety of toxic and hazardous
substances in waste gas streams. In recent years biological
filters have been developed and described which have the
potential to simply and effectively control odors, including
H2S emissions (Prokop and Bohn, 1985; Allen et al., 1987a;
Eitner and Gethke, 1987; Hartenstein, 1987 ). Deodorization
methods based upon the activity of microorganisms are
beginning to attract increasing attention in the U.S.
Although the biofiltration technique has been shown to
be an efficient, practical and simple gas cleaning
technology, which is increasingly being used around the
world, the design and operation parameters as well as the
microbial processes involved have not yet been very well
defined. In particular, little research has been directed at
the details of the biofiltration control of H2S. A

3
systematic compilation of data from an operational point of
view is also lacking. Most designs are conservatively based
on blanket rule of thumb' criteria (Forster and Wase,
1987). The performance of biofilter systems, therefore, is
not readily predictable and sometimes these systems are not
operated under suitable conditions. As a result, the
desired odor control efficiency is sometimes not achieved
(Allen et al., 1987b). It is essential that more work be
done to demonstrate the effectiveness of these systems in
order to support further progress in the use of
biofiltration as well as to develop better biofilters, based
on an understanding of the fundamental physical, chemical
and biological processes involved.
A major disadvantage of biofiltration technology is the
limited degradation capacity represented by the volume of
waste gas treated per unit area of filter material per unit
time (m3/m2-hr). This limitation restricts the applicability
of biofiltration systems to handling dilute waste gas
streams and requires the filter bed to be large in order to
handle high volumetric gas flows.
In order to overcome the uncertainties and
disadvantages encountered in the full scale application of
biofiltration technology, an exhaustive study is necessary
for the application of biofiltration technology to control
the emissions of air pollutants.
The objectives of the proposed research were to develop
a quantitative knowledge of the principle and operation of

4
a microbial biofilter system for removal of H2S from waste
gas streams and determine the operating parameters necessary
to optimize the performance of such a biofilter system.
The objectives were achieved through the following
studies:
1. Evaluation of the properties of filter materials and
their decomposition characteristics under aerobic
conditions.
2. Evaluation of the effects of design and operational
parameters on H2S removal efficiencies on laboratory
scale biofilter systems. Variables evaluated
included temperature, pH, compost water and sulfate
content, H2S elimination capacity, pollutant
retention time, etc..
3. Determination of the predictive relationships for
H2S control efficiencies through chemical kinetic
studies.
4. Evaluation of system performance and determination
of optimum maintenance procedures for biofiltration
control of H2S during long term operation.
5. Evaluation of the field performance of a full scale
biofiltration system for control of H2S emissions at
a local waste water treatment plant.
The research reported here focuses mainly on the
utilization, improvement and optimization of a compost
biofiltration tower system. Optimization of this system
has been directed toward the best achievable control of

5
hydrogen sulfide. This research provides a detailed data
base on the effects of system variables on H2S control
efficiency, which provide for optimization of design and
operating conditions.

CHAPTER 2
BACKGROUND
Properties of Hydrogen Sulfide and Regulations
Physical and Chemical Properties
Hydrogen sulfide is a colorless gas that has a foul
rotten egg odor and is slightly heavier than air. Hydrogen
sulfide is moderately soluble in water. The solubility of
H2S decreases with increasing temperatures. Figure 2-1
shows the solubility of H2S as a function of temperature.
Dissolved H2S dissociates in accordance with the
following reversible ionization reactions:
H2S HS + H+ (2-1)
HS S2_ + H+ (2-2)
The distribution of the above species as a function of
pH is shown in Figure 2-2. It is apparent from Figure 2-2
that the concentration of HS- species is insignificant when
pH values are less than 6. The latter condition is normal
in a biofilter system for control of H2S. S2-, on the other
hand, may not occur at all.
Hydrogen sulfide can serve as a reducing agent,
reacting with sulfuric acid (H2S04) to form sulfur dioxide
(S02) and elemental sulfur (S) (Greyson, 1990) :
6

Solubility (g-S/L)
7
Figure 2-1. Solubility of H2S in water at 1 atm.
Data adopted from Piscarcyzyk, 1982.

Percent
8
pH
Figure 2-2. Effect of pH on H2S equilibrium.
Source: Sawyer, 1967.

9
H2S + H2S04 = S02 + S + 2H20 (2-3)
Hydrogen sulfide also burns in air to form sulfur
dioxide and water:
2H2S + 302 = 2S02 + 2H20 (2-4)
Table 2-1 summarizes the physical and chemical
properties and the odor threshould of H2S.
Toxicity of HoS
Hydrogen sulfide is almost as toxic as hydrogen cyanide
(HCN), which is used in prison gas chambers (Parker, 1977).
Human exposure to small amounts of H2S in air can cause
headaches, nausea, and eye irritation, and higher
concentrations can cause paralysis of the respiratory
system, which results in fainting and possible death.
Concentrations of the gas approaching 0.2 percent (2000
ppmv) are fatal to humans after exposure for a few minutes
(NRC, 1979).
Hydrogen sulfide has a characteristic rotten egg smell
at low concentrations. But as levels of H2S increase, a
person's ability to sense dangerous concentrations by smell
is quickly lost. If the concentration is high enough,
unconsciousness will come suddenly, followed by death if
there is not a prompt rescue.
The Occupational Safety and Health Administration
(OSHA) has established limits for work place exposure to H2S

10
Table 2-1. Physical and Chemical Properties of H2Sa.
Molecular Weight
34.08
Boiling Point, C
-60.2
Melting Point, C
-83.8 to -85.5
Vapor Pressure, -0.4C
10 atm
25C
20 atm
Specific Gravity (Relative to Air)
1.192
Auto Ignition Temperature, C
250
Explosive Range in Air, %
4.5 to 45.5
Odor Threshold, ppbv
0.47
a Source: USEPA, 1985.

11
at 20 ppm (15-minute exposure) for an acceptable ceiling
concentration and 50 ppm for a maximum exposure during an 8-
hour work shift if no other measurable exposure occurs. The
National Institutes of Occupational Safety and Health
(NIOSH) established an H2S exposure level at 10 ppm (10
minutes) as a maximum permissible limit (once per 8-hours
shift), with continuous monitoring reguired where H2S
concentrations could egual or exceed 50 ppm or greater
(NIOSH, 1979).
Hydrogen sulfide is an explosive gas. The lower and
upper explosive limit are 4.5 and 45 percent in air by
volume, respectively.
Hydrogen sulfide can attack materials and cause
discoloration and tarnishing. Materials commonly affected
are paint, copper, zinc and silver (Painter, 1974).
Sources of H2S Emissions and Regulations
Natural emissions are mainly caused by biological decay
of protein materials. The natural global rate of emission
is estimated to be about 84 Tg/year (Urone, 1986).
Anthropogenic emission sources include petroleum
refining, natural gas plants, sewage treatment facilities,
coke ovens, Kraft paper pulp plants, and waste disposal
sites. There are no federal U.S. emission standards for H2S
at present, nor are there federal ambient air guality
standards for this gas, but a number of states have

12
established independent standards for H2S emissions. These
states, which include California and New Mexico (10 ppm),
and Ohio and Michigan (1670 ppm). California, Kentucky,
Minnesota, Montana, New Mexico, New York, North Dakota and
Pennsylvania also have air quality standards for H2S. The
standards vary from 0.003 ppm for New Mexico to 0.1 ppm for
Pennsylvania, whereas and most of the other states specify a
standard of 0.03 ppm (Urone, 1986).
Since H2S is a highly toxic air pollutant, H2S has been
identified by the USEPA as one of 190 air toxic compounds in
Title III of the 1990 Amendments to the Clean Air Act. In
view of the wide spread exposure to this pollutant, emission
and air quality standards for H2S are going to be set in
the near future by EPA.
Biofiltration as a Air Pollution Control Technology
s
Biological degradation is widely used for treatment of
liquid and, to a lesser extent, solid wastes, but has
received little attention as a means of controlling
emissions of industrial gaseous wastes. Biofiltration is a
relatively new technology for control of air pollutants, in
which the air contaminants from off-gas streams are
biologically removed in a solid biological reactor. While
it is a well established air pollution control technology
in European countries, biofiltration as an air pollution
control technology has received little attention and
application in the United States. Few environmental

13
professionals in this country appear to be aware of the
'biofiltration1 process and its applications.
Although there are some applications of biofiltration
in the U.S. and some technical papers have been published in
the English language, most of the research and development
work on biofiltration has been conducted in Europe and the
majority of the recent research data have been published in
the German language. Excellent reviews of previous
biofiltration work have been published by Hartenstein
(1987), Leson and Winer (1991), Ergas et al. (1991), and
Dharmavaram (1991).
History and Development
The first deodorization method based upon the use of a
soil bed in the U.S. was developed and patented by Pomeroy
in 1957. Later, Pomeroy (1982) described the deodorization
of waste gases emitted from sewer lines by a soil bed system
used in Los Angeles in 1957. The microbiological degradation
of sulfur-containing gases in the filter bed was observed to
be effective in these studies.
Other early applications of biological treatment of
odorous gases include a soil bed system built in Nrnberg,
West Germany, in 1959 and biofilters built in Geneva,
Switzerland, and Mercer Island, Washington, to remove odors
from wastewater treatment and compost manufacturing,
respectively, in the mid-1960s (Bohn and Bohn, 1987).

14
Additional studies in the US have been carried out by
Carlson and Leiser (1966), Bohn and Miyamoto (1973), Bohn
(1975, 1976, 1977, 1989), Pomeroy (1982), Prokop and Bohn
(1985), Hartenstein and Allen (1986), Bohn and Bohn (1986,
1988), Hartenstein (1987), and Allen et al. (1987a, b, c;
1989).
In spite of the work mentioned above, most of the
research and development in biofiltration technology has
been carried out in Europe, especially in West Germany and
Holland. In the latter countries the principle of
biofiltration has been applied to a wide variety of
environmental problems. Among the many researchers in the
field, Ottengraf and coworkers in Chemical Engineering
Department, The Eindhoven University of Technology, Holland,
have contributed most of the theoretical research in
biofiltration in a series of papers which have been
published in the English language (Ottengraf, 1977;
Ottengraf and Van Den Oever, 1983; Ottengraf et al., 1984;
Ottengraf, 1986; Ottengraf et al. 1986; Ottengraf, 1987).
Also, Eitner in West Germany, has made significant
contributions to the research and development of
biofiltration ( Eitner and Gethke, 1987), although most of
his publications are in the German language (see
Hartenstein, 1987; Leson and Winer, 1991).
The practice and application of biofiltration has also
been reported in other countries such as Japan (Terasawa et
al., 1986), New Zealand (Rands et al.,
1981) and Canada

15
(Rotman, 1991a).
At present, biofiltration is considered to be a state-
of-the art technology for odor removal in West Germany, and
it has been estimated that 40% of deodorization facilities
at wastewater treatment plants are biofilters (Frechen and
Kettern, 1987) .
Applications
The first systematic study of odor control using
biofiltration in this country was conducted by Carlson and
Leiser (1966). They studied the removal efficiencies of
sewage odors using a laboratory scale soil bed. Using
hydrogen sulfide as the test gas, a 99% removal efficiency
was achieved, and biodegradation was reported to be the
primary removal mechanism.
Prokop and Bohn (1985) reported that a soil bed system
for control of rendering plant odors had been in operation
since September, 1983. The soil bed treats 1100 m3/h of
cooker non-condensable waste gases using a bed surface area
of 420 m2. In this work an odor removal efficiency of
99.9% was obtained.
Rands et al. (1981) reported that a full-scale compost
filter system was constructed in 1978 at Moerewa, New
Zealand, to treat odors from a rendering plant. The system
was designed to treat 900 m3/h of air containing hydrogen
sulfide concentrations up to 1000 parts per million (ppm) by
volume. An average H2S removal efficiency of 99.9% was

16
observed.
Allen et al. (1987a, b, c) investigated a compost based
tower biofilter system used for odor control in a wastewater
treatment plant. The odor-causing compounds identified were
reduced sulfur compounds such as H2S, methyl mercaptan,
dimethyl sulfide and dimethyl disulfide as well as terpene
hydrocarbons. Removal efficiency for total reduced sulfur
compounds (TRS) was 65 to 72%. The poor performance of this
system was determined to be the short residence time in the
system, poor gas distribution, and improper maintenance.
Biofiltration control of volatile organic compounds
(VOCs) has been reported by Ottengraf (1986), Kampbell et
al. (1987), Bohn (1989), Paul and Castelijn (1987), and Hack
and Habets (1987) .
In recent years, increasing numbers of biological
filters are being used around the world for odor control. It
has been estimated that more than 500 biofilters are
currently operating in Europe (Leson and Winer, 1991).
Excellent summaries of recent applications have been
provided by Bohn and Bohn (1987) and Rotman (1991b).
Theoretical Basis
The concept of a biological-film or 'biofilm' is
freguently used to describe degradation processes in aqueous
systems ( Williamson, 1973; Williamson and McCarty, 1976a,
b; Jennings et al., 1976; Rittmann and McCarty, 1978).
This concept has been adopted and improved to describe the

17
biofiltration processes (Ottengraf, 1986; Hartenstein, 1987;
Paul and Castelijn, 1987; Van Lith, 1989). In particular,
Ottengraf and coworkers have carried out systematic studies
delineating the overall process and have presented
sufficient experimental data to support the proposed model.
In biofiltration, evenly distributed waste gases are
forced through a biologically active material, such as soil,
peat or compost. Many of the pores of the filter material
particles are filled with water. Microorganisms are attached
to the particle surfaces to form a layer of film. This wet,
biologically active layer surrounding the particles is
called a biofilm. The biophysical model proposed by
Ottengraf for the biofilm is shown in Figure 2-3. The
mechanism of the biological process is derived from a
combination of physical, chemical and biological processes
that occur in the filter material and is related to two
processes in particular; sorption and regeneration. As
waste gases pass through the countless narrow pores of the
filter material, air contaminants as well as oxygen will
adsorb on the surfaces of the pores and dissolve in the
liquid phase of the wet biofilm. The absorbed and adsorbed
gases are quickly degraded by the biofilter's enormous
microbial population. In this way a concentration gradient
is created in the biofilter, which maintains a continuous
mass flow of the component from the gas to the wet biofilm.
Activity of the biofilter depends mainly on the
population of the microorganisms. Soil biofilters can

18
zone
Cal = High concentrations of air pollutants.
Cg2 = Low concentrations of air pollutants.
Figure 2-3. Biophysical model for the biological
filter bed. The concentration profiles
shown in the biofilm refer to: 1)
Reaction limitation, 2) Diffusion
limitation. (Source: Ottengraf, 1986,
p. 436).

19
contain 1 billion bacteria, 10 million actinomycetes and
10,000 fungi per gram of soil (Bohn and Bohn, 1987). The
role of these microorganisms is to oxidize combined carbon,
nitrogen and sulfur to carbon dioxide, nitrogen and sulfate,
respectively, before the compounds leave the bed. The air
contaminants are, therefore, effectively removed from the
waste gas streams.
For good engineering design and environmental decision
making, it is essential to understand the mechanisms
involved and to reliably predict the kinetics of the
biological reactions taking place in these biofilter
systems. Many general kinetic models have been developed to
predict the behavior of bioreactions in a biological film,
none of these models, however, is specific enough to explain
the biodegradation of hydrogen sulfide in a biofilter
system.
Jennings et al. (1976) developed a mathematical model
to predict the percentage removal of a pure, non-adsorbable,
biodegradable substrate in a submerged biological filter
using the non-linear Monod expression for the substrate
utilization rate. In their model, the authors start from a
biological slime layer coating a spherical particle. The
slime layer is in turn surrounded by a liquid boundary
layer. They concluded that even at relatively high values
of influent substrate concentrations, the biological removal
of a single substrate follows first order kinetics.

20
Another model developed by Rittmann and McCarty (1978)
is a variable-order model of bacterial-film kinetics which
incorporates liquid-layer mass transport, biofilm molecular
diffusion and Monod kinetics. These investigators concluded
that at low substrate concentrations, the reaction follows
first order kinetics, whereas at high concentrations the
reaction follows one-half order kinetics.
Based on their biophysical model (Figure 2-3),
Ottengraf and Van Den Oever (1983) have developed a
mathematical model to describe the kinetics of organic
compound removal from waste gases for a biofilter system.
The model was developed and tested using a soil bed for the
removal of toluene, butylacetate, ethylacetate and butanol.
From their experimental results, they concluded that all the
carbon sources investigated were eliminated according to a
zero order reaction, even at very low concentrations of the
substrates.
Kampbell et al. (1987) investigated the biodegradation
of propane, isobutane and n-butane by soil biofilter beds.
They suggested that at low concentrations the rate of
biodegradation was proportional to the concentration of the
organic compounds (first order reaction), and at higher
concentrations the rate becomes independent of the organic
compound concentration (zero order reaction). The
degradation kinetics appeared to follow a hyperbolic
function:

21
1/v = (i/vmax) + ( yvnax)(i/s) (2-5)
wh e r e :
V = the biodegradation rate, mg hydrocarbon/kg soil-h
Vmax = maximum possible biodegradation rate,
mg hydrocarbon/kg soil-h
Kjjj = an empirical constant, half saturation value, ppm
S = the concentration of organic compound in air, ppm
The values of Km and Vmax can be obtained from a
Lineweaver-Burk plot of 1/V against 1/S.
Biological Oxidation of Hydrogen Sulfide
Hydrogen sulfide may be utilized by microorganisms in
three different ways: assimilation, mineralization and
sulfur oxidation (Atlas and Bartha, 1981; Grant and Long,
1981). However, the rates of uptake of hydrogen sulfide
based on the assimilation processes are far too low to
achieve reasonably high removal efficiencies from a highly
loaded waste gas stream. The most important and efficient
way for microorganisms to utilize hydrogen sulfide is by the
oxidation of sulfur to gain energy. In this process,
relatively large quantities of sulfur are oxidized in order
for the microbes to receive sufficient energy. The
microorganisms living in the biofilter materials are usually
mixed cultures. Various groups of microorganisms,
therefore, are involved in the energy conversion process
under aerobic or anaerobic conditions. However,
the

22
colorless sulfur bacteria are believed to play the major
role and their ability to oxidize reduced inorganic sulfur
compounds has been clearly established (Roy and Trudinger,
1970; Kuenen, 1975; Brock and Madigan, 1988).
The oxidation of inorganic sulfur compounds is carried
out by a spectrum of sulfur-oxidizng organisms which include
1) obligately chemolithotrophic organisms, 2) mixotrophs,
3) chemolithotrophic heterotrophs, 4) heterotrophs which do
not gain energy from the oxidation of sulfur compounds but
benefit in other ways from this reaction, and 5)
heterotrophs which do not benefit from the oxidation of
sulfur compounds. Physiological characteristics of some
sulfur-oxidizing bacteria are summarized in Table 2-2.
Options for microbial metabolism of hydrogen sulfide
must employ one or more of the following metabolic pathways:
1) aerobic oxidation, 2) anaerobic oxidation, and 3)
photosynthetic dissimilation. Biofiltration of waste gases
is a process utilizing aerobic conditions in most cases. In
aerobic oxidation, sulfur-oxidizing bacteria oxidize H2S to
elemental sulfur or higher oxidation states using oxygen
(02) as an electron acceptor. The biological steps in the
oxidation of various sulfur compounds are summarized in
Figure 2-4.

23
Table 2-2. Physiological characteristics of sulfur-
oxidizing bacteria.
Lithotrophic Electron
Donor
pH Range
for Growth
Thiobacillus Species
Growing Poorly in
Organic Media:
1. T. thioparus
H2S, sulfide, S, S2032-
6-8
2. T. denitrificans
h2s, s, s2o32~
6-8
3. T. neapolitanus
s, s2o32"
5-8
4. T. thiooxidans
s
2-5
5. T. ferrooxidans
S, sulfides, Fe2+
1.5-4
Thiobacillus Species
Growing Well in
Organic Media:
1. T. novellus
s2o32"
6-8
2. T. intermedius
s2o32"
3-7
Filamentous Sulfur
lithotrophs
Beaaiatoa
h2s, s2032
6-8
Thiothrix
h2s
6-8
Other Genera
Thiomicrosoira
6-8
Thermothrix
H2S, S2O3 f SO3
6.5-7.5
Sulfolobusa
h2s, s
1-4
a Archaebacterium.
Source: Brock and Madigan, 1988.

24
Cell-Bound
Sulfur Complex
R
Sulfite
S
Sulfide
Elemental Sulfur
S 20g' Thiosulfate
e Electro T ransport
System
SO 4' Sulfate
SO
2-
4
Figure 2-4. Steps in the oxidation of different
compounds by thiobacilli. The sulfite
oxidase pathway is thought to account
for the majority of sulfide oxidized.
(Source: Brock and Madigan, 1989, p.
704) .

CHAPTER 3
FILTER MATERIALS AND THEIR DECOMPOSITION
UNDER AEROBIC CONDITIONS
Introduction
Biofiltration systems or biofilters employ physical,
chemical and biological processes such as adsorption,
absorption and microbial digestion and oxidative degradation
to remove air pollutants from waste gas streams. Microbial
degradation and oxidation of the pollutants, however, appear
to be the primary removal mechanisms within a biofilter. In
the biodegradation process, pollutants are consumed by the
microorganisms, providing an energy source or essential
nutrients and are converted usually to, less harmful
compounds. The filter materials used, on the other hand,
must provide the proper environment for microbial growth and
contain materials on which the microbes can feed to ensure
that the microbial population can develop and survive.
The effectiveness of a biofilter material depends on
its physical, chemical and biological characteristics. The
lifetime of a biofilter material mainly depends on its rate
of carbon (C) and nitrogen (N) mineralization. When
available C and N in the filter material are no longer
sufficient to support the microbial population in the
system, then the material is no longer suitable as a
25

26
biofilter. The C or N deficient filter material must be
replaced by freshly prepared material and the discarded
filter material has to be properly disposed of with due
caution for environmental impact. One of the most common
options is land application. Determination of the
decomposition characteristics of the filter material is,
therefore, necessary for usage of the biofilter and eventual
land disposal applications.
Considerations necessary for selection of appropriate
filter materials and the decomposition of such materials
under aerobic conditions are discussed in this section.
Selection of Filter Materials
Effective removal of air contaminants using a biofilter
relies on the properties of the filter material, especially
the nature and activity of the biomass. The filter
material provides the necessary environment for
microorganisms to survive, generate, function and allows
the entire sequence of biofiltration processes to be carried
out. The filter material serves as 1) support material for
the microbes, 2) supplemental or alternative nutrient
source, 3) moisture storage reservoir, 4) surface area for
sorption of air pollutants and interaction between the
pollutants and the microorganisms, and 5) a buffer volume
for variations in water content and gas conditions during
operation (Eitner, 1989). In general, the following factors
need to be considered when choosing a suitable filter

27
medium:
1). Density: Too dense material may contain a large
fraction of inorganic materials such as stone and
sand which are unsuitable as carbon and energy
sources for microbial growth.
2). Structure: Structure of the medium will affect
the uniformity of the filter load. Too large
sized materials should be avoided because the
surface-to-volume ratio will be reduced.
3). Particle Size Distribution: Too small particles
affect the pressure drop by compacting and
restricting the gas flow.
4). Pore Volume (void fraction): This property
determines the total surface area available for
reaction, also it will affect pressure drop.
5). Organic Matter Content: The organic matter
controls the microbial population and the useful
service life of the filter media.
6). pH Value: pH will affect the nature and level of
the microbial population and activity.
7). Water Retention Capacity: This property will
determine the consistency in liquid water content
of the filter material, and
8). Economics: Reasonable Capital and operating
expenditures.
All of these requirements can be met by selecting
suitable filter materials. Many kinds of filter materials
have been used in biofiltration applications. Examples
include field soils, compost, peat, bush, clay, volcanic
ash, sand, bark and a combination of such materials (Rands
et al., 1981; Prokop and Bohn, 1985; Terasawa et al., 1986
Frechen and Kettern, 1987). The performance of these
materials, however, can be very different due to the
diversity of their physical and chemical properties. Compost
has been considered to be the best choice for filter

28
materials and has been involved in most applications (Don,
1985; Eitner, 1989), since it provides favorable conditions
for supporting microbial populations as well as having
superior physical and chemical properties.
The properties of individual composts depend on the
materials from which they are derived and the composition of
the final product. The filter materials used in this
research were mainly yard waste compost and sewage sludge
compost or a combination of both. These composts were
obtained from different sources and used for different
purposes. A general description of the types of composts
and their sources are summarized in Table 3-1. The physical
and chemical analyses data for the composts listed in Table
3-1 are presented in the corresponding chapters where the
use of specific composts is discussed.
Decomposition of Composts under Aerobic Conditions
A number of investigations have been carried out to
study the decomposition of anaerobically digested sewage
sludges in soils (Miller, 1974; Tester et al., 1977; Terry
et al., 1979b; Sweeney and Graetz, 1988; Gale, 1988).
Decomposition of fresh and anaerobically digested plant
biomass in soil is also reported by Moorhead et al. (1987).
Only limited information, however, is available concerning
the decomposition of compost. Tester et al. (1977, 1979)
stated that the decomposition of compost in soil is not only
related to the physical and chemical properties of the

29
Table 3-1. Description of composts used for this study.
Compost
ID#
Source
Description
1
Pompano Beacha
Fort Lauderdale sewage sludge compost.
Not completely composted. Seven months
old when first used (used for
decomposition study).
2
Pompano Beach
Two parts yard waste and one part stable
cleaning sewage sludge mixed and
composted. Seven months old when first
used (used for decomposition study).
3
Pompano Beach
Yard trash compost. 13 months old when
first used (used for decomposition
study).
6
Pompano Beach
25% by volume of sewage sludge compost
and 75% of yard trash mixed and composted
about 19 months old when first used (used
for decomposition study).
12
Kanapahab
Pompano Beach compost similar to Compost
#6 mixed with tree bark, yard waste and
sewage sludge; lime was used to adjust pH
before use. Used in Kanapaha filter bed
from 11/20/88. Compost obtained from the
filter bed in 5/16/90.
13
Kanapaha
Same as #12, compost obtained and used in
Tower #1 from 12/20/90.
13-1
Kanapaha
Same as #12, compost obtained in 2/5/91.
13-2
Kanapaha
Same as #12, compost obtained in 3/20/91.
14
Kanapaha
Yard trash, grass and sewage sludge were
mixed and composted; lime was used to
adjust pH; about 2.5 years old when
obtained and used in Tower #2, 12/20/90.

30
Table 3-1 Continued.
Compost
Source
Description
ID#
16
WRRC
Yard trash compost, 3.5 months old when
first used in Tower #3 from 1/27/91.
17
WRR
1:1 by volume of yard trash and grass
composted; about 3.5 months old when
first used in Tower #2 from 1/27/91.
17A
WRR
Compost #17 mixed with 2% lime (CaC03),
by dry weight of compost. Used in Tower
#1 from 1/27/91.
a Broward County Streets and Highways Division, 1600 NW 30th
Avenue, Pompano Beach, FL, 33069.
b Kanapaha Wastewater Treatment Plant, Gainesville, FL 32602.
c Wood Resource Recovery, Inc., Gainesville, FL.

31
compost but also is a function of the particle size of the
compost. The decomposition was observed to be directly
related to the carbon content in the compost.
Decomposition is affected by a number of environmental
conditions, for instance, pH, moisture content, and the
presence or absence of foreign chemicals (Miller and
Johnson, 1964; Terry et al., 1979a, b; Delaune et al.,
1981). In the application of biofiltration to control H2S
emissions, the compost filter material is subjected to
conditions that are quite different to that for land
applications of compost. In the former case, the compost is
exposed to a gas stream which may contain a variety of
chemicals, especially H2S, at various concentrations. The
presence of xenobiotics in the gas streams and filter
materials could change the population and composition of the
microorganisms in the compost or significantly affect their
metabolic processes. As a result, the decomposition rate of
the compost can be altered. Unfortunately, little
information can be found in the literature related to this
topic.
The objectives of this study were (i) to evaluate the
decomposition of four types of compost by determining the
CO2 evolution, and (ii) to investigate the effect of H2S at
various concentrations on compost decomposition. Such
information is valuable for biofilter design and for
justifying land disposal applications of the compost after
use as a biofilter medium.

32
Materials and Methods
Four types of compost samples were investigated for
their decomposition characteristics during the course of
this study. All of the compost samples were obtained from
Broward County Streets and Highways Nursery Division,
Pompano Beach, Florida. The composts were stored in sealed
plastic bags at room temperature (232 C) before use. A
brief description of the composts used in this study is
presented in Table 3-1 ( Composts #1, #2, #3 and #6). The
compost samples were analyzed for their physical and
chemical properties at the beginning and the end of the
investigation. The results of these studies are presented in
Table 3-2. Each cured compost was passed through a 10 mm
screen to remove larger materials. The compost samples are
then placed in 225-mL wide mouth bottles directly for
incubation. The experimental arrangement for the
decomposition study is shown in Figure 3-1. Four types of
compost and one blank, each with three duplicates, were
investigated. Compressed air from the laboratory house air
supply is controlled to about 4 psig by a regulator. The
air stream is passed through a scrubber system consisting of
4N NaOH to remove C02 and distilled water to saturate the
air stream. A dead volume is placed before the 4N NaOH
scrubber as a safety precaution in the event that the air
system causes a backpressure forcing scrubbing solution
against the air system. An empty impinger is placed after
the water scrubber to separate larger water droplets from

Table 3-2. Properties of selected composts before and after incubation.
Compost #1 Compost #2 Compost #3 Compost #6
Property Before After Before After Before After Before After
use
use
use
use
use
use
use
use
PH
8.67
6.79
8.13
8.44
9.22
8.72
7.26
7.55
Water (Wt%)
60.3
59.5
59.7
57.4
61.0
63.0
60.5
62.0
LOI (Wt%)
79.8
73.7
69.1
64.4
73.8
74.4
65.2
64.6
Water-P (mg/kg)
144
96.9
95.2
97.9
53.3
49.3
231
218
Total-C (Wt%)
36.9
39.6
36.8
35.3
39.5
40.9
39.5
36.2
Total-N (Wt%)
3.00
3.57
2.84
2.78
2.26
2.30
3.45
3.73
C/N
12.3
11.1
12.9
12.7
17.5
17.8
11.5
9.71
Metals (mg/kg)
Ca
37400
40300
57500
60000
66500
65000
47700
49800
Mg
4120
4470
5250
5300
3840
4250
5400
5950
Zn
492
557
651
638
81.0
94.5
897
964
Cu
201
194
125
131
8.50
9.50
114
113
Mn
30.0
36.0
66.5
70.0
23.5
25.0
48.5
52.0
Fe
8820
9760
5220
5130
512
541
6740
7260

Regulator
Needle
Valve
Manifold
Dead 4N Dl Dl
Volume NaOH Water Water
Impinger 50g 2X25ml
Compost 0.5N NaOH
Figure 3-1. Schematic drawing of the experimental arrangement for the
study of compost decomposition.

35
the air stream. Two water scrubbers are used in series to
ensure that the air stream is completely free of alkali and
to resaturate the air with water vapor in order to keep the
water content of the composts constant.
The resaturated, C02-free air stream is then forced
to the manifold where it is split into 15 streams. Each
stream goes into one incubation-absorption unit. Fifty
grams of compost sample is put in each incubation bottle.
The CO2 evolved from each of the compost samples is
collected in two 25-mL, 0.5N NaOH collectors in series. The
total air flow rate is controlled by a needle valve located
in front of the manifold. Syringe needles are used as flow
regulators to equalize the air flow through the 15
incubation units. The air flow rate through each unit is
adjusted to about 152 mL/min. The incubation system is
continuously operated at constant temperature (232C).
After flushing the residual air from the incubation
bottles, the outlet tube of each bottle is attached to the
CO2 collectors. C02 collectors are replaced with fresh
solutions periodically during the incubation period. The
system is leak checked before the incubation. Evolved C02 is
efficiently trapped by two absorption collectors in series.
Tests have shown that the first tube absorbed more than 95%
of the total C02 evolved.
C02 evolution is measured as described by Stotzky
(1965) with minor modifications. After C02 absorption, the
solutions in the two collectors of each unit are mixed and

36
titrated with standard IN HC1. The C02 samples collected
from each of the control bottles are concomitantly titrated.
The CC>2 evolved for individual samples is calculated as
follows (Stotzky, 1965):
C02 = (B V)NE (mg) (3-1)
where:
B = volume of HC1 used to titrate the NaOH in the
controls to the end point, (mL);
V = volume of HC1 used to titrate the NaOH
remaining in the C02 collectors after
treatment to the end point, (mL);
N = normality of the HC1, (meq/mL);
E = equivalent weight, (mg/meq), for C02, E = 22
(mg/meq).
To investigate the effect of H2S concentration on
compost decomposition, one hundred grams of Compost #6 was
used as the test material. The experimental arrangement for
this test is similar to that for the compost decomposition
test with some minor modifications (Figure 3-2) Room air
is forced through a scrubber containing 4N NaOH to absorb
C02 from the air stream. The C02 free air is then saturated
by bubbling through DI water. Pure H2S is then mixed with
the pretreated air stream to obtain the test gas mixture
with the desired H2S concentration. The treated gas stream
is vented through the manifold, where it is split into four
sub-streams: one of these sub-streams is vented to a control
column (empty), and the other three streams to duplicates of
three compost columns. The gas is forced vertically through
the compost from bottom to top at a flow rate of 30 mL/min.

Regulator
Manifold
Row Meter
Row
Regulator
I
]
Dead 4N DI H2S Gas
Volume NaOH Water Mixer
100g
Compost
200+25ml
1N ZnAc
*
2x25ml Dead
0.5N NaOH Volume
Figure 3-2.
Schematic drawing of the experimental arrangement for the
investigation of the effect of H2S exposure on compost
decomposition.

38
A portion of the hydrogen sulfide is adsorbed and/or
oxidized by the compost and the remaining H2S in the
effluent gas is absorbed by two H2S scrubbers in series
which contain 200 mL and 25 mL IN zinc acetate (ZnAc)
solution, respectively. The absorbing reaction used by the
H2S collectors is:
ZnAc + H2S - ZnSi + HAc
ti
H+ + Ac" (3-2)
Total flow in the system is measured by pre-calibrated
rotameters. H2S concentrations in the inlet gas to the
compost column are controlled by adjusting the flow rates of
mixing for the C02-free air and the pure H2S gas. Gas
samples from the influent gas stream are taken periodically
by gas-tight syringes, diluted with prepurified nitrogen
(N2) and analyzed for H2S content by a Tracor 250H analyzer
(See Chapter 4 for details) The effluent gas from the
filter columns is first passed through two scrubbers in
series containing IN ZnAc to absorb any H2S remaining in the
gas stream. Residual C02 in the effluent gas stream is
subsequently absorbed by 0.5N NaOH solution and titrated as
described previously.
Each compost sample is incubated at a desired H2S
concentration level for 24 hours. After the incubation
period the compost as well as the absorption solutions are
replaced by fresh compost samples and absorbing solutions
for operation at the next H2S concentration level. The

39
system was previously tested to obtain absorption
efficiencies of H2S and C02 in the ZnAc traps. The results
showed that the H2S absorption efficiency was greater than
99% and the C02 absorption was less than 2% for the ZnAc
solutions used.
Results and Discussion
Decomposition of Composts
The decomposed C evolved as C02 from the four composts
studied during the 122 day incubation period is shown in
Figure 3-3. The decomposition patterns of composts #2, #3
and #6 are somewhat similar. Decomposition of compost is
initially rapid, from 40 to 52% of the total C02 produced in
the 122 days is evolved in the first 42 days of incubation.
A total of 9.2, 5.7, 6.1, and 4.4% of the original C was
decomposed and released as C02 for composts #1, #2, #3, and
#6, respectively during the total 122 days of incubation.
It appears that decomposition rates of the composts
are inversely proportional to their age, in other words,
the older the compost, the slower the decomposition. All
except compost #1, showed decomposition rates which were
similar. Compost #1, however, was not completely composted
when used. Also the organic matter content of this compost
is higher than that for the others tested. Initial and
delayed higher decomposition rates for compost #1 suggest a
two stage incubation involving an initial 'conditioning'
step followed by a 'conditioned' decomposition. The compost

C0 Evolved (mg-CC>2/g of C added)
40
Time (day)
B Compost #1+ Compost #2 o Compost #3 a Compost #6
Figure 3-3. Plot of C02 evolution from composts
during the 122 day incubation.

41
decomposition rates measured here are much slower than those
found for soils. Tester et al. (1977) reported that
approximately 16% of the compost C was evolved as C02 during
54 days of incubation, when 2 to 6% fresh sewage sludge
compost was incubated with soils. Miller (1974) reported
that 20% of added organic carbon is evolved as C02 for a 6
month incubation period under similar conditions. In another
investigation carried out by Moorhead et al. (1987) it was
observed that about 39 and 19% of the total-C for fresh and
digested low-N plant biomass, and 50 and 23% of fresh and
digested high-N plant biomass are released as C02 during 90
days decomposition when these biomasses are added to soils.
Fresh plant biomass evolves as much as twice the organic-C
as C02 when compared to corresponding digested biomass
sludges. These results of other researchers suggest that
the decomposition rate of organic matter strongly depends on
the source and the properties of the available organic
matter. Miller (1974), Sommers et al. (1976) and Terry et
al. (1979a, b) have concluded that sludge composition and
incubation conditions, rather than soil properties control
sludge decomposition.
Reddy et al. (1980) have shown that decomposition of
organic carbon depends on the nature and constitution of the
wastes. Low molecular weight (simple) compounds can be more
easily degraded by microorganisms than more complex organic
compounds. Organic-C components in decreasing order of
biodegradability are: (i) readily oxidizable soluble

42
organic-C, (i i) proteins, (iii) hemicellulose, (iv)
cellulose, and (v) lignin. In the examples mentioned above,
fresh plant biomass releases much more C02 (especially in
the early stages of the incubation) than the digested ones
because it contains much more easily decomposable organic-C.
In this study, all the composts used were well aged or
completely composted. Most of the easily decomposable
organic-C such as soluble organic-C, starch and proteins
have been decomposed during the composting process. The
main organic-C species remaining in the composts studied
are the more oxidation resistant residues of the original
organic matter (Biddlestone et al., 1987). Also, the four
composts studied are either yard waste compost or mixtures
of sewage sludge with yard wastes such as wood chips, leaves
and tree trimmings, etc.. A high content of cellulose and
lignin can be expected in these materials. This feature may
explain why the decomposition rates for the composts studied
here are relatively low.
Decomposition of a complex substrate C is usually
described by a multistage first-order decomposition sequence
(Reddy et al., 1980; Gilmour et al., 1985). The mathematical
rate equations can be written as follows:
-dCj/dt = kjCi (3-3)
where i refers to a particular stage of decomposition.

43
The integrated form of equation 3-3 becomes:
cti = ciexP(kit) (3-4)
where:
= organic-C present at the beginning of a
decomposition stage.
= organic-C present at the end of a decomposition
stage at time = t, and
k^ = the first-order reaction rate coefficient.
Decomposition stages and the corresponding reaction
rate coefficients for the four composts are presented
graphically in Figure 3-4. The decomposition of compost #3
is described in one stage and the decompositions of compost
#2 and #6 are best described in two stages. It can be seen
that the reaction rate coefficient values of k-^ and k2 for
these two composts are very similar. This similarity
indicates that these two composts have similar organic C
composition.
The behavior of Compost #1 is markedly different from
those of the other composts. During the first 3 days of
incubation, C02 evolution is rapid followed by a lag period,
lasting for the following 40 days. A second period of high
decomposition rate was observed between 42 and 70 days.
During the remaining period of incubation (after 70 days)',
the C02 evolution rate for this compost is similar to those
for the other composts. The decomposition rate for compost
#1 may be described as a 3 stage series of first-order
reactions.
Within overal experimental error, reaction rate
coefficients for the final stage of decomposition for the

(!o/ bJui
44
0
-0.02
_
- Compost #1
-
K, = 0.00042/day \
-
k2= 0.00156/day
0.00057/day
-
0
-0.01
-0.02
-0.03
-0.04
-0.05
-0.06
-0.07
Figure 3-4. Decomposition stages and reaction rate
coefficients for the four composts studied.

45
four composts studied are quite similar, falling in the
range from 3.1xl0-4 to 5.7xl0-4/day. If the second reaction
rate coefficients for the composts studied can be assumed to
be representative through the remaining life of the compost,
then a rough estimate of the time required for decomposition
of 50% of the organic matter (half life) in these composts
can be made according to following equation.
tQ#5 = 0.693/k (3-5)
where:
tQ5 = the half life time of the compost, (day), and
k = the first-order reaction coefficient, (1/day)
The estimated half life time of the composts tested is
from 3.3 to 6.1 years. This estimate is comparable to the
result reported by Varanka et al. (1976), who showed that it
takes approximately 6 years to lose 50% of the sludge
organic C when used in the field.
No significant changes in other physical and chemical
properties of the composts were observed for the 120 day
incubation period used in these studies (Table 3-2).
Effect of FUS on Compost Decomposition
The effect of H2S exposure on compost decomposition is
illustrated in Figure 3-5, where C02 evolved by the composts
is expressed as mg-C02 per g-C of the compost as a function
of the H2S concentration (ppmv). The C02 evolution is
significantly increased with the increasing H2S
concentration. The rate of this increase is greater at

C0 Evolved (mg-CC^/g of C added)
46
Figure 3-5. Effect of H2S exposure on the rate of
compost decomposition as measured by
C02 respiration.

47
lower H2S concentrations. For example, the C02 evolved at
H2S concentrations near 6,000 ppm is approximately 8.5 mg-
C02/g-C added, which is about 3.4 times that evolved when no
H2S is present. At higher H2S concentrations the increase
of C02 evolution with H2S concentration is reduced e.g.
when the H2S concentration is increased from 12,000 ppm to
32,000 ppm the C02 evolution increases only by about 17%, or
approximately 2 mg-C02/ g-organic matter.
It can be seen from Figure 3-5 that the C02 evolution
from the compost has a strong dependence on the H2S
concentration in the gas to which the compost is exposed. A
linear relationship is obtained when plotting C02 evolved as
a function of the square root of H2S concentration in the
gas, [H2S]0,5 for the range of H2S concentration less than
17,000 ppmv (Figure 3-6). The regression analysis result
for the best fit line is:
C02 = 2.62 + 0.082[H2S]0*5 (3-6)
where:
C02 = C02 evolved from compost, (mg/g of C added)
[H^S] = H,S concentration in the inlet gas stream,
(ppmv)
The correlation coefficient, for the variables is
0.9234.
Equation 3-6 quantitatively describes the effect of H2S
on the decomposition of composts. For example, C02
evolutions at [H2S] = 0 and [H2S] = 1000 ppmv are calculated

CQ, Evolved (mg-CC^ /g of C added)
48
Figure 3-6. Plot of C02 evolution as a function of
square root of H2S concentration.

49
to be 2.62 and 5.21 mg-C02/g-OM, respectively, by equation
3-6. The ratio of these two values is equal to the ratio of
the first-order reaction coefficients for the reactions at
these two conditions,
C02,1000/C02,0 = k1000/k0 = 1*99*
in other words, the decomposition rate for the compost
exposed to 1000 ppmv H2S is 1.99 times of that for the
compost not exposed to H2S. The half life times for compost
#6 at both conditions are:
to.5,0 = *693/0.00031 = 6.12 (years), and
tQ ^ 1000 = 6.12/1.99 = 3.08 (years)
and for compost #3 are:
*"0.5,0 = 0*693/0.00057 = 3.33 (years), and
^0.5,1000 = 3.33/1.99 = 1.67 (years).
No studies of similar effects have been reported in the
existing literature. Thus the results obtained in this study
can not be compared with the results of other
investigations. Taylor et al.(1978) found that the highest S
mineralization rates are observed during the period of
highest C02 evolution when 2 to 6% of sewage sludge compost
is incubated in soils. Their results and those reported here
suggest that the microbial activity of the compost was
significantly enhanced by the addition of H2S, especially
the activity of the sulfur oxidizing bacteria.
Oxidation of inorganic sulfur compounds is a basic
phenomenon in nature. A number of bacteria have been

50
identified in soils and other environments that are capable
of oxidizing organic and inorganic sulfur compounds (Roy and
Trudinger, 1970). A high population of the oxidizing
bacteria can be expected in the composts tested here. With
sufficient H2S supply, the bioactivity of the sulfur
oxidizing bacteria can be stimulated to result in an
increase of microbial population and a corresponding
increase in the evolution of C02. Hydrogen sulfide is
finally oxidized to sulfate through various pathways and
intermediate stages ( Roy and Trudinger, 1970; Brock and
Madigan, 1988; Yang and Allen, 1991; Allen and Yang, 1991).
After the 24 hours reaction period, the color of the
compost changed from originally brown to yellowish-white,
especially at high H2S concentrations. This feature
indicates that a large amount of sulfur has accumulated in
the compost.
Conclusions
Among the various biofilter materials, compost is
frequently selected as a medium in applying biofiltration to
air pollution control due to its unique properties and
advantages. Knowledge of the characteristics of compost
decomposition are important for both prediction of biofilter
operation characteristics and degradation estimates, as well
as in deciding on the appropriate disposal treatment and
method for used compost. The studies described here indicate
that the decomposition rates of the composts tested are much

51
lower than those reported by other researchers, who used
fresh composts mixed with soils. Five to ten percent of the
total-C in composts were decomposed during the 122 days
incubation period. Compost half life times of the order 3 to
6 years are estimated for the composts studied,
corresponding to loss 50% of their total-C due to
decomposition. A multi-stage first-order reaction sequences
is used to describe the decompositions. First-order reaction
rate coefficients have been determined.
Decomposition rates are significantly increased when
H2S is introduced to the compost. The half life of the
compost is significantly reduced as a result of increased
biological activity and C02 respiration. For example,
continuous exposure of compost to 1000 ppm H2S can result in
reduction of the half life of the compost from about 6 years
to 3 years due to enhanced microbiological activity alone.
The results suggest that added H2S was oxidized by the
sulfur oxidizing bacteria in the compost to sulfate.

CHAPTER 4
DETERMINATION OF THE DESIGN AND
OPERATIONAL PARAMETERS FOR BIOFILTER SYSTEMS
Introduction
Extensive experimental work has been carried out in
order to determine the design and operational parameters for
a biofilter system. This research is essential for best
operation as an air pollution control technology and for
optimization of the system. This chapter describes the
design and construction of lab scale biofilter systems,
experimental methodology used and the results obtained.
System Design and Construction
Three biofilter systems were designed and constructed
for different investigative purposes. Each system can be
operated and controlled separately. Detailed information on
each experimental system is presented below.
The Dual Tower System
Most of the experimental work was carried out using a
dual-tower experimental biofilter system. This
configuration, which is shown in Figure 4-1, consists of
parallel dual column filters. The two biofilter columns,
identified as Tower #1 and Tower #2, can be run
52

53
Orifice Vent
Figure 4-1. Schematic drawing
system.
of the dual tower

54
simultaneously and controlled separately. The biofilter bed
material is enclosed in transparent rigid Acrylic pipe, with
an inner diameter of 0.15 meters (6 inches) and a height of
1.34 meters (4 feet). Each vertically mounted pipe can be
packed with the desired compost up to a height of 1.2 meters
(3.9 feet). The packed biofilter material is supported by a
sieve plate to ensure a homogeneous distribution of the
inlet gas stream across the face of the bed. Non-
biodegradable plastic screens are placed between the sieve
plate and the biofilter material to avoid separation of
smaller compost particles.
Sampling and measurement ports are located along the
Acrylic column for compost and gas sampling, and pressure
and temperature measurements. The sampling and measurement
ports are shown in Figure 4-2a and b for Towers #1 and #2,
respectively. An individual sampling/measurement port is
identified by a letter-number system, where the letter
indicates the function and the number indicates the location
of the port. For example, TS11 means this port is used for
temperature measurement and solid sampling, and is located
on Tower #1 at location 1. All the other filter systems
with multi measurement/sampling ports are identified in the
same manner.
Room air is forced by a Gast Regenair Model R3105-1
air blower into the humidification chamber. The blower,
which is driven by a 1/2 HP motor, generates a maximum flow
of 1.5 m /min (53 cfm) and a maximum pressure/vacuum of

125
55
Q15
8
8
8
pii
po c
T15
e-
TS14:
-
TS13;
-
TS12!
--
TS11:
-
Q14
G13
3
8
Q12
i
8
Q11
H
8
Qio ^
no O 3
8
8
T = Temperature
P = Pressure
G = Gas Sample
S = Solid Sample
. Sampling and measurement ports on
towers.
a. Tower #1
Figure 4-2

56
8
P 21
P20
-e-
T26 !
TS25;
-e-
TS24;
TS23
TS22 ;
TS21 I
<-
T20
"O'
G25
3-
G23
B
G24
G22
B
G21
B
G20
8
T = Temperature
P = Pressure
G = Gas Sample
S = Solid Sample
8
8
Figure 4-2. Sampling and measurement ports on
towers,
b. Tower #2

57
1100/1000 mmH20 (43/40 inches) of water column.
Humidification of the inlet air is achieved by atomizing
water in the spray chamber, through which the room air
passes. In addition, Pall rings are stacked in the spray
chamber for extending wetted surface area providing better
humidification. As a result relative humidities in the
range 95 to 100% were routinely and continuously achieved.
Gaseous H2S with a purity of 99+%, which is stored in
liquid form under pressure in a cylinder, is continuously
leaked and mixed with the prehumidified air in the inlet
lines (PVC pipe) to the towers. Plastic screen packing is
placed downstream from the H2S introduction point for better
mixing. Flow rates of air and H2S are controlled by plastic
valves, which are located on the carrier gas inlet lines,
and stainless steel needle valves, respectively. The flow
rates are measured on calibrated flow meters to obtain the
desired H2S concentration and gas flow through the towers.
Measurements of temperature, pressure and gas flow rate are
discussed in later sections.
A nozzle is installed on the top of Tower #1 in order
to introduce water, or other liquid solutions if necessary,
to the outlet end of the bed. Gas lines are made from PVC
pipes. The towers and pipes are connected by flanges for
convenient dismantling of the packed towers and compost
changes. Cork-rubber gaskets are used to seal the flanges.

58
Portable Tower #3
The portable tower is made from PVC pipe with an inner
diameter of 77 mm (3 in) This tower, which is shown in
Figure 4-3, has a total height of 1.2 meters (3.94 ft) with
an effective packing height of 1 meter (3.28 ft). The two
ends of the pipe are covered by rubber caps and held by pipe
clamps. Compost packed in the tower is supported by a
packing of non-biodegradable plastic screen. Measurement
ports for pressure, temperature and exhaust gas samples are
located along the length of the tower. Gas to be tested is
introduced through a port at the bottom of the tower. The
overall gas flow rate is measured by a pre-calibrated flow
meter after the effluent gas passes through a particulate
filter. This portable tower was used intensively for
pressure drop studies and for investigation of long term
operation of compost #16.
Column System #4
A fourth column biofilter system was constructed for
investigation of the effects of various operational
variables on H2S removal (Figure 4-4). This multicolumn
system includes eight compost columns, a manifold for
introduction of test gas and several needle valves for flow
control. The columns are made from PVC pipes with an inside
diameter (ID) of 35 mm (1.25 in). Each column has a length
of 300 mm ( 12 in) and an effective packing height of 250 mm

59
8
Gas Outlet
TG35
-E
oooooo/ooI
&
TGP34
g
TGP33
-E!
TGP32
-E
TGP31

..x xxxx
m
*: :
lili
111
$$PI
pi
: ;
;:i
I
P31
P30
T = Temperature
P = Pressure
G = Gas Sample
Figure 4-3
Schematic drawing of Tower #3.

60
Inlet Gas
Sampling Port
Figure 4-4.
Schematic drawing of column system #4.

61
(10 in), which provide a 240 mL packing volume. The ends of
the packed columns are plugged by rubber stoppers.
Thermometers are inserted into the columns to measure
temperatures. The gas flow rates are measured by a
rotameter at the gas outlets. Effluent gas samples are
taken from the outlet of the rotameter.
Measurement Methods
Periodic measurement of temperature, pressure drop and
gas flow rate in the biofilter systems are carried out by
the following devices.
Temperature
Temperature is measured by mercury in glass thermometers
with a range from -20 C to 110 C and a minimum scale
division of 1 C.
Pressure Drop
Pressure drop is measured by manometers with a minimum
reading of 1 millimeter water column (mmH20). In case the
pressure drop is greeter than 1000 mmH20, the pressure drop
is measured by mercury manometers with a minimum reading of
1 millimeter of mercury (mmHg).
Gas Flow Rate
All the gas flow rates except those of Towers #1 and #2
are measured by pre-calibrated rotameters.

62
The gas flow rates in Towers #1 and #2 are measured by
specially designed orifices. Two orifices for each tower
were designed and made, one for low flow ranges and the
other for high flow ranges. The orifices are made from
plastic plate and installed on the outlet gas pipe lines
(see Figure 4-1). The pressure drops across the orifices are
measured and the flow rates are calculated according to the
developed calibration equations.
Sampling Methods
Compost Samples
Compost samples in Towers #1 and #2 are taken from the
solid sampling ports shown in Figures 4-2 a and b. The
samples are taken at each port in a radial direction to the
tower walls so that a representative sample can be obtained
for that section. For composts not initially packed in
columns, the samples are taken after the compost has been
thoroughly mixed and very large particles ( diameter > 10
mm) have been eliminated.
Gas Samples
The inlet and outlet gas samples for each system are
obtained directly from the gas sampling ports by extraction
using gas-tight syringes. Gas samples extracted from
other locations along the towers are obtained by using a gas
sampling probe assembly. The gas sampling assembly, as

63
shown in Figure 4-5, consists of a Teflon probe, a Teflon
filter and a sampling port connector. The Teflon probe is
made from a piece of Teflon tubing (6.35 mm (1/4") in
diameter), with 14 holes (1mm diameter) spaced evenly along
the probe length. The probe is installed in the towers in
such a way that all the holes are perpendicular to the
tower's normal axis. Thus, representative gas samples from
a cross section of the tower can be obtained. The Teflon
filter is used to block out any small particles and water
droplets which may be extracted during sampling. Gas
samples are obtained through the sampling port located on
the end of the assembly (see Figure 4-5) by a gas-tight
syringe. The sampling port is sealed by a rubber GC septum.
When taking a sample, at least three full syringes of
gas sample are wasted before the actual sample is taken for
analysis. This procedure will eliminate residuals of
previous gas samples remaining in the Teflon filter holder,
in the syringes and in the probe, as well as condition the
extraction system to the gas being sampled.
The gas samples are then diluted in 3-L Tediar sampling
bags by pure nitrogen (N2) to an appropriate concentration
within the calibration range of the analyzer. The gases in
the Tediar bags are thoroughly mixed by gently kneading the
bags and allowing them to sit for at least 10 minutes before
analysis. Most of the samples taken are analyzed within 2
hours.

64
Biofilter Tower
Teflon Filter
Teflon Probe
Sampling Port
Plastic Union
Figure 4-5. Schematic drawing of the gas sampling
assembly.

65
In most cases, the H2S concentrations in the outlet
gases are so low that the gas samples can be analyzed
directly without any dilution. In the latter cases, Tediar
bags are directly connected to the sampling ports. Gas
samples are forced into the bags as a result of the positive
pressure of gas within the tower system. Also, Teflon
filters are replaced by glass wool plugs to reduce the
resistance to flow.
Each time after use, the Tediar bags are purged at
least three times with N2 to eliminate residual gas and
vapor. The stability of gas samples in the Tediar bags are
discussed in Chapter 6.
Water Samples
Water samples analyzed are mainly biofilter wash waters
from the tower drain outlets. When washing a packed tower,
the entire wash water is collected in a container. Water
samples are obtained from the container after mixing the
wash water with a stirrer for a few minutes.
Compost Analysis Methods
Water Content
Two to five grams of wet compost are dried in an
aluminum tray at an oven temperature of 70 C until constant
weight is obtained. Compost water content is determined by
the difference in weight between the wet and dry composts
(Robarge and Fernandez, 1986).

66
E
A known amount of wet compost is weighed into a 50-mL
container. DI water is added to bring the liquid/solid
ratio to 10 (Robarge and Fernandez, 1986). The sample is
shaken for 30 minutes by a rotary shaker. Measurements of
pH are made by a calibrated Corning Model M245 pH meter,
which is accurate to 0.01 pH.
Total Carbon and Total Nitrogen
Finely-ground, oven-dried compost sample (<100 mesh)
are analyzed for total carbon and total nitrogen using a
Carlo Erba Model NA 1500 CNS Analyzer.
Water Soluble Phosphorus (WSP)
A known amount of wet compost (2.5 g dry weight
equivalent) is weighed into 50-mL centrifuge tubes. DI
water is added to the tubes to obtain a compost to liquid
ratio of 1:10 on a dry weight basis. These samples are
allowed to agitate for a period of one hour on a mechanical
shaker. The compost suspensions are then centrifuged at
6000 rpm for 15 minutes and filtered through Gelman 0.45
micrometer membrane filters. The filtered solutions are
acidified (pH<2.0) with one drop of concentrated H2S04 and
stored at 4 C until analyzed. The soluble reactive P (SRP)
in the filtered extract is determined colorimetrically
(APHA, 1989) using a Shimadzu UV-160 spectrophotometer with
1 cm path length at 880 nm wavelength.

67
Acid-extractable Cations
Two and half (2.5) g of finely-ground, oven-dried
sample is weighed into 50-mL centrifuge tubes. Twenty five
(25) mL of 1M HC1 is added and the tubes are shaken for 3
hours on a mechanical shaker. The compost suspensions are
centrifuged at 6000 rpm for 15 minutes and filtered through
Gelman 0.45 micrometer membrane filters. The solutions are
analyzed for Fe, Al, Ca, Mg, Cu and Mn on an Inductively
Coupled Argon Plasma Spectrometer (ICAP) (APHA, 1989).
Particle Size Distribution
The compost is dried in oven at 70C for 24 hours.
Particle size distribution by weight is measured by passing
the dried compost through a series of sieves (U.S.A.
Standard Testing Sieve, A.S.T.M. E-ll Specification, Fisher
Scientific Company) and weighing the residue.
Porosity
Compost porosity is determined according to Danielson
and Sutherland (1986).
Organic Matter
After determination of compost water content, the
samples are placed in a muffle furnace and baked for 24
hours at 450 C. Organic matter is determined by the loss-
on-ignition (LOI) (Robarge and Fernandez,
1986).

68
Particle Density
Particle density of the compost is measured according
to Blake and Hartge (1986a).
Bulk Density
Bulk density of the compost is measured according to
Blake and Hartge (1986b).
Sulfur Analysis Methods
Intensive and detailed laboratory work has been carried
out on the analysis of sulfur compounds in order to obtain a
better understanding of the biochemical reactions involved
in the H2S oxidizing processes occurring in the biofilters.
The procedures for determining various sulfur compounds in
compost, in water, and in the waste gases are described in
this section.
Sulfur in Compost
The analysis of sulfur in compost includes the
determination of total sulfur (total-S) and fractionation of
the total sulfur into inorganic and organic constituents.
Many wet chemical procedures have been developed to
fractionate the total sulfur pool in sediments, soils, and
peat into its inorganic and organic constituent compounds.
Very little information, however, is available about such
analyses for compost. The sulfur analyses conducted in this
research include the quantitative determination of acid

69
volatile sulfur ( sulfide-S), water soluble sulfur (soluble
sulfate), insoluble sulfate (S042-), elemental sulfur (S),
Pyrite sulfur, ester sulfur (organic-S) and total sulfur.
Each of the wet chemical procedures involved the
reduction of S to H2S in a Johnson-Nishita apparatus
(Johnson and Nishita, 1952) and trapping the evolved H2S in
zinc acetate-sodium acetate (ZnAc-NaAc) solutions. Trapped
sulfide is quantified by iodometric titration (APHA, 1989)
with a 0.025N iodine solution and 0.025N Na2S203 titrant.
The distillation apparatus incorporated slight
modifications of that used by Johnson and Nishita and is
similar to that used by Wieder and Lang (1985). Figure 4-6
shows the distillation assembly. The reaction flask is a
250-mL, round-bottom, three-neck flask, with an N2 inlet via
a bleed tube inserted in one neck and the central neck is
connected to a condenser. Ultra high purity (>99.999%)
nitrogen is used to sweep out the H2S and to maintain the
reaction flask in a reducing environment. The third neck of
the flask is fitted with a stopper to allow introduction of
liquid solutions to the flask. The ZnAc-NaAc solution is
made by dissolving 50 g of zinc acetate dihydrate
[Zn(CH3COO)2.2H20] and 12.5 g of sodium acetate trihydrate
(CH3C00Na. 3H20) in 800 mL of DI water and adjusting the
final volume to 1 liter (Tabatabai, 1982). Twenty-five (25)
mL of this solution is mixed with 100 mL of DI water and
this solution is used to fill two traps used in series. The
first trap contains 100 mL and the second one contains 25 mL

70
Figure 4-6.
Photograph of the sulfur distillation
assembly.

71
of the H2S absorption solution. Even though the contents
of both traps are analyzed, more than 95% of the H2S is
consistently recovered in the first of the two series traps.
The material to be analyzed is added to the reaction flask
through the side neck. The system is purged with N2 at
bubbling rate of 1-2 bubbles per second in the ZnAc-NaAc
traps for 10 minutes before the introduction of additional
reagents. The materials are boiled for 1 hour, the traps
are removed and sulfide titrated.
The procedures and methods used for analysis of the
total sulfur and various organic and inorganic sulfur
compounds in compost are similar to those used for sulfur
analyses in peat, soil and sediments described by Zhabina
and Volkov (1978), Tabatabai (1982), and Wieder and Lang
(1985) Minor modifications were made for the compost
sulfur analyses in this research. The procedures are
illustrated in Figure 4-7.
All results are expressed as mg sulfur per gram of
compost on a dry basis (mg-S/g). Compost moisture content is
determined from a sub-sample by drying the compost at 70 C
to constant weight.
Compost samples are stored in plastic bottles and
refrigerated at 2-4 C before analysis. In most cases,
however, compost samples are analyzed immediately after
sampling.

I
Reducing
Mixture
Reduction
H2S
Absorption
I
lodometric
Titration
I
Reducing
Mixture
Reduction
I
H2S
Absorption
I
lodometric
Titration
I
Total-S
Water-Soluble-S
(Sulfate-S)
Gas
H2S
Absorption
I
lodometric
Titration
I
Sulfide-S
Figure 4-7
Flow chart of
FRESH WET COMPOST
Store for
Later Use
Hydriodic
Acid
Reduction
Filtration
Precl-
Acetone
Extraction
Preci-
CrCI2
Reduction
Preci-
ptate
pitate
pitate
| Solution
1
Gas
Reducing
Mixture
Reduction
Gas
CrCI2
Reduction
| Gas
H2S
H2S
H2S
H2S
Absorption
Absorption
Absorption
Absorption
1
1
1
lodometric
lodometric
lodometric
lodometric
Titration
Titration
Titration
Titration
1
Sulfate-S
1
Elemental-S
Pyrite-S
1
Ester-S
the sulfur analysis procedures for compost

73
The fresh wet compost sample is divided into five sub
samples. The latter are used for the analyses of 1) compost
moisture content, 2) total-S, 3) water-soluble-S, 4)
inorganic and organic sulfur fractions, and 5) storage for
later use.
Total sulfur
Total-S is determined by oxidation (acid digestion) of
the various reduced sulfur constituents to sulfate and
followed by reduction of the sulfate to H2S. The H2S is
then trapped and titrated as described above. In most of
the analyses, 0.5-5.0 g of fresh compost is used, depending
on the sulfur content. In some analyses, oven dried compost
is used. In the latter case, the compost samples are finely
grounded (<40 mesh) and a correspondingly smaller size of
compost sample is used. The compost is subjected to acid
digestion as described by Tabatabai (1982). The digest is
quantitatively transferred into a 100-mL volumetric flask
and the volume is adjusted with IN HCl. Reduction of the
sulfate is carried out by a reducing mixture. The reducing
mixture contains 50% hypophosphorous acid, 90% formic acid,
and hydriodic acid in a 4:2:1 proportion and is prepared as
described by Tabatabai (1982). Depending on the sulfur
content, 1 to 5 mL aliquot of the digest is transferred into
the distillation flask. With aliquots >2 mL, the volume is
reduced to about 2 mL by heating the flask on an electric
heating mantle. Five (5) mL of the reducing mixture is

74
added to the flask and the material is subjected to
reduction and hydrogen sulfide is liberated, collected and
titrated as described above.
Water soluble sulfur
Water-soluble-S is determined by shaking 2-5 g of fresh
compost in DI water with a liguid to solid ratio of 10:1 for
30 minutes on a rotary shaker at a speed of 140 /min. An
aliquot of the compost extract is then subjected to
reduction, H2S absorption, and titration successively, as
described previously.
The following analyses are conducted in succession:
Sulfide sulfur
Sulfide-S or acid-volatile sulfur (AVS) is determined
by introducing 8 mL of 12N HC1 to the compost sample in the
reaction flask. Heat is applied after 10 min, the materials
are brought to boiling, and after 45 min the traps are
removed and the sulfide titrated.
Sulfate sulfur
The content of the reaction flask is filtered by a #42
Whatman filter. The filtrate is then subjected to
reduction. H2S is then trapped and titrated. This is an
alternative way of carrying out sulfate analysis. The
results are comparable to the summation of water soluble-S
and P-extractable-S.

75
Elemental sulfur
The precipitate obtained above is dried with filter
papers as described by Zhabina and Volkov (1978) and is
extracted with analytical grade acetone. The volume of
acetone used (mL) is 40 times that of the equivalent dry
weight of the compost sample (g). The extraction flask is
covered with Parafilm and placed on a rotary shaker at a
speed of 140/min for 16 hours. The mixture is filtered and
rinsed with additional acetone. Either a fraction of or the
entire filtrate are subjected to Cr2+ reduction as described
by Zhabina and Volkov (1978).
Pvrite sulfur
The residue left after S extraction is subjected to
chromium reduction. The Cr2+ was produced by passing a 1 M
solution of CrCl3.6H20 in 0.5 M HC1 through a Jones reductor
column containing Zn amalgamated with Hg (Zhabina and
Volkov, 1978). Preparation of the Jones reductor is
described by Swift (1950) and Patterson and Thomas (1952).
Ten (10) mL of ethanol is added to the flask followed by 20
mL of 12N HC1 and 16 mL of 1M CrCl2 solution. Heat is
applied after 30 min, the H2S evolved is absorbed in the
traps and titrated as above.

76
Organic sulfur
Following the previous procedure, the insoluble residue
in the reaction flask is filtered through a #42 Whatman
filter and is washed repeatedly with acidified DI water to
remove chromium ions. The residue is then subjected to
reduction with the reducing mixture as described above. The
organic sulfur determined this way is mainly ester-S.
In addition to the literature mentioned above, similar
and dissimilar procedures for determining sulfur
constituents in sediments, peat and soil are also described
by Smittenberg et al. (1951), Freney (1958) Johnson and
Henderson (1979), David et al. (1982), and Hsieh and Yang
(1989). An excellent comparison of some of these methods is
reported by Wieder and Lang (1985).
Sulfur in the Aqueous Phase
Sulfur in water samples, such as in tower wash water
and drainage, are analyzed for sulfide, sulfate, and/or
total sulfur.
Sulfate sulfur
Sulfate in most of the water samples is determined by a
turbidimetric method (APHA, 1989). A Milton Roy Model
Spectronic 21 Spectrometer was used at 420 nm to measure the
turbidity. Color or suspended matter in large amounts will
interfere with this method. In the case of dark colored

77
water samples, the reduction, H2S absorption and iodometric
titration procedures described previously are used.
Sulfide sulfur
Sulfide was measured according to APHA (1989). The
samples are pretreated to remove interfering substances and
to separate insoluble sulfide.
Total-S
Total-S was determined by using 2-5 mL of the aqueous
sample depends on its sulfur content. The sample is then
analyzed as described for the total sulfur in compost.
Sulfur in Waste Gas
Gas samples are analyzed by a commercial gas
chromatogragh equipped with a flame photometric detector
(GC/FPD), a Tracor Model 250H Analyzer. The detection limits
and operational conditions for the analyzer are summarized
in Table 4-1. Detailed information concerning the principles
and conditions of operation are reported elsewhere (Yang,
1988). In the early stages of this study, the component
peaks are recorded by a chart recorder (Texas Instruments,
Inc., Model Recti/Riter II) and concentrations of sulfur
compounds are determined by measuring the peak heights.
Later in the study (from August 1990 to September 1991) the
chart recorder was replaced by a Spectra-Physics model
SP4290 integrator and the measured concentrations are read

78
Table 4-1. Retention times, limits of detection and
operating conditions for the Tracor 250H
Analyzer.
Item
h2s
Compounds3
MM
DMS
Retention
Time
(min)
1.15
2.16
4.02
Detection
Limit
(ppmv)
0.01
0.01
0.02
Operating Conditions:
Temperatures: Valve: 50 C
Column: 70 C
Detector: 110 C
Flow Rates:
Cylinder
Pressures:
Nitrogen:
Oxygen:
Hydrogen:
Sample:
80 mL/min
21 mL/min
80 mL/min
40 mL/min
(carrier gas)
(flame gas)
(flame gas)
Air 40 psig
Hydrogen: 53 psig
Oxygen: 40 psig
Nitrogen: 80 psig
for sampling valve
activation)
a
MM: methyl mercaptan.
DMS: dimethyl sulfide.

79
directly from the printout of the integrator. The Tracor
analyzer is periodically calibrated for H2S, methyl
mercaptan (MM) and dimethyl sulfide (DMS) with standards
purchased from National Speciality Gases, Inc.
Results and Discussion
Biofiltration is a process that involves physical,
chemical, and biological processes. Many variables, such as
temperature, compost water content, specific acidity of
compost, sulfate content in the compost etc., can affect the
function of the system. It is impossible to obtain optimum,
longer term performance from a biofiltration system without
an in depth understanding of the system properties and
proper control of important variables. Extensive
evaluations of the system properties have been conducted
during the course of this study. The results presented in
this section are divided into several subsections according
to specific investigations undertaken. Composts from
different sources (Table 3-1) have been used in the
laboratory studies. The physical and chemical properties of
these composts are summarized in Table 4-2. Applications of
each of these composts are mentioned in the corresponding
study subsections.
Pressure Drop
The energy consumption obtained in operating a
biofiltration system is primarily that required by the

80
Table 4-2. Summary of initial compost properties.
Compost ID #
Property
12
13
14
16
17
PH
2.61
1.60
6.44
6.66
8.10
Bulk Density (g/cc)
0.30
0.27
0.18
0.22
0.20
Particle Den. (g/cc)

1.73
1.75
1.90
1.78
Porosity (%)

84.3
89.7
88.4
88.7
Water Content (wt%)
45.6
54.9
62.4
56.5
62.7
Organic Matter (wt%)
66.5
66.8
59.3
64.3
62.6
Total-S (mg-S/g)
44.8
70.4


0.74
Water-P (mg/kg)
140
223
152
114
167
Total-C (wt%)
34.3
31.0
31.3
30.5
40.9
Total-N (wt%)
1.89
3.24
1.75
4.27
1.30
C/N
18.1
9.57
17.9
7.14
31.5
Metals (mg/kg)
Ca
47400
28900
145000
18000
26700
Mg
280
255
4880
1450
120
Zn
38.0
48.0
201
66.5
18.0
Cu
121
91.0
60.0
11.0
93.5
Mn
5.50
55.0
96.5
89.5
9.50
Fe
1900
805
6160
529
1510

81
blower (fans) to move contaminated air at the specified flow
rate through the filter bed. However, the pressure drop
across the filter bed increases markedly as the flow rate is
increased. Since the pressure drop will be determined by
the depth of the filter bed it is necessary that the gas
velocity should be kept as low as possible.
The pressure drop across a compost bed filter can be
lowered by physical treatment of the compost particle
content. Such a procedure is an important requirement in
optimizing the operation of a filter bed, because operation
at a reduced resistance to flow allows the gas velocity as
well as the volumetric flow rate to be significantly
increased with little or no change in energy consumption.
This will in turn increase the biofilter capacity and reduce
the required filter size.
The following physical factors determine pressure drop
across the filter bed:
1) Particle size distribution in the compost
2) Condition of the filter packing
3) Height of the filter bed
4) Water content of the compost
5) Gas velocity, and
6) Porosity of the compost.
A representative sample of compost #12 was air dried
and the particle size distribution determined by weighing
the fractions penetrating a series of standard sieves. The
particles are classified into 5 size groups in the range >12
mm to <1.2 mm (see Table 4-3). Compost samples in each of

82
Table 4-3. Particle size range distributions for
selected composts.
Particle Size
Range Distribution
(wt%)
Compost
ID #
A
B
C
D
E
12
20.0
22.5
10.0
13.1
34.4
14
27.7
26.9
8.10
11.6
25.7
13
21.4
24.5
6.70
15.8
31.6
17
0.00
33.4
14.4
22.5
29.7
A: diam. > 12 mm
B: 3.35 < diam. <12 mm
C: 2.36 < diam. <3.35 mm
D: 1.18 < diam. <2.36 mm
E: diam. <1.18 mm

83
the particle size groups were analyzed for water content
then wetted to obtain a water content of about 50%, by
spraying and mixing water with the compost samples. These
treated samples were then used to determine the pressure
drop as a function of the 5 particle size range classes by
adjusting air velocities in the range 0.02 m/s to 0.28 m/s.
All tests were carried out with the same compost bed height
(1 m) and water content (50%).
The results of these studies are presented in Figure 4-
8, where it is seen that the pressure drop increases
significantly with increasing gas velocity for a bed of
small particle size (<1.2mm). For particles greater than
1.2 mm the pressure drop increases to a much lesser extent
with increasing velocity as shown by a comparison of data
for particle classes D and E. Clearly significant pressure
drops observed in operating filter beds are a result of the
presence of small particles with sizes less than 1 mm. For
example, at a representative gas velocity of 0.03 m/s, which
is equivalent to a loading rate of 110 m3/m2-hr, the
pressure drop realized by a 1-m bed of particles size
classified as <1.2 mm is 390 mm H20, whereas the pressure
drop obtained for the same packing height and the same gas
velocity, for particles classified as >12 mm is only 2 mm
H20. Thus, under these conditions the pressure drop created
by 1 mm or less particles is about 200 times that caused by
an equivalent bed composed of 12 mm or greater particles.

Pressure Drop (mHgO)
84
Particle Size Range
A: d > 12 mm
B: 3.35 < d < 12 mm
C: 2.36 < d < 3.35 mm
D: 1.18 < d < 2.36 mm
E: d < 1.18 mm
Figure 4-8
Pressure drop as a function of particle
size range for different gas
velocities.

85
The relationships between pressure drop and packing
height at constant gas velocity for particle size ranges B
(3.33 to 12 mm) and E (<1.2 mm), and the parent compost #12
are shown in Figure 4-9. It is seen that the pressure drop
increases approximately linearly with packing height for all
three samples. It is important to note that the pressure
drop values obtained for the parent compost are intermediate
between the strong dependence of pressure drop on packing
height for small particles (class E) and the weak dependence
on packing height for larger particles (class B).
The dependence of pressure drop on compost water
content is not as consistent as that for particle size.
Qualitatively, sewage sludge treated compost contains more
viscous and adhesive small particles than are found in
untreated compost. Thus, when the water content of the
sewage sludge treated compost is increased, coagulation of
small particles is enhanced and the pressure drop increases
sharply. However, the build-up in pressure may be suddenly
released by channeling, i.e., a breakdown of the overall
flow restriction by the formation of a channel of much less
resistance caused by a separation of packed materials.
The effects of gas velocity on pressure drop across
four typical biofilter materials are shown in Figure 4-10.
As expected, the pressure drop increases rapidly with
increasing gas velocity. The pressure drop depends on the
way the filter is packed. Composts #12 and #13 are similar
in nature but filter material #13 was more compacted than

Pressure Drop (mmh^O)
86
Figure 4-9
Pressure drop as a function of packing
height for different compost particle
size ranges.

Pressure Drop (mmH0)
87
Gas Loading Rate (m3/m2-hr)
Figure 4-10. Pressure drop as a function of gas
velocity for different types of
compost.

88
#12, and the more densely packed material shows a much
higher pressure drop for the same gas velocity. Composts
#14 and #17A have similar pressure drop-gas velocity curves
and although material #17A contains a larger fraction of
small particles than does material #14 (see Table 4-3), the
small particles in compost #17A are mainly sand and grass
fractions, whereas those in compost #14 are sewage sludge
particles, which tend to adhere to each other.
Effect of Gas Retention Time on PUS Removal
The effect of gas retention time on H2S reduction is
studied by varying the gas flow rate through the tower. The
results are presented in Table 4-4. In the first data set
of 6 tests, the H2S loading rate is kept approximately
constant at a low flow range to ensure that the maximum H2S
elimination capacity of the system is not exceeded during
the test. It is clear that there is no apparent effect on
H2S removal as long as retention times are longer than about
23 sec. When the retention time is reduced to 7 seconds,
the H2S removal efficiency decreases by about 6%. This
decrease in H2S is controlled by the macrokinetics of
biofiltration process. Sublette and Sylvester (1987)
reported that H2S can be metabolized by a pure culture of T.
denitrificans within 1-2 seconds. This suggests that the
reduction of H2S removal efficiencies under shorter
residence times is not necessarily due to insufficient
reaction time between the H2S molecules and the biomass, but

89
Table 4-4. Effect of gas retention time on H2S removal
efficiency.
Gas Flow
Rate
(Lpm)
H2S
Reten. Loading H^S
Time Rate Inlet
(s) (g-S/m3-hr) (ppmv)
H2S Removal
Outlet Eff.
(ppmv) (%)
151 j,
7.06 rcmw-
16.3
106
9.98
17.1
75.5
14.1
17.7
46.0
23.1
21.0
30.8
34.5
20.7
15.3
70.0
19.8
15.0
71.0
39.8
15.0
71.0
50.7
9.00
118
56.3
9.00
118
73.2
5.40
197
62.3
24.9
1.62
93.5
37.0
1.30
96.5
53.7
0.77
98.6
105
0.30
99.7
155
0.02
99.9+
297
0.01
99.9+
610
BDL
99.9+
776
2.77
99.6
1440
0.01
99.9+
1870
4.24
99.8
2650
4.58
99.8

90
is possibly due to the slower step of H2S diffusion from the
gas phase into the liquid phase.
In the second data set of 5 tests presented in Table 4-
4, the biofilter inlet gas contains very high H2S
concentrations and the system is operated at lower flow
rates and, therefore, longer retention times. These tests
show that, even when the inlet gas contains an H2S
concentration of 2650 ppmv, the biofilter can successfully
reduce the concentration to 4.6 ppmv with a 99.8% removal
efficiency when the retention time of the flowing gas is
increased to 197 seconds. Thus, it can be seen that, as
long as the H2S loading rate does not exceed the maximum H2S
elimination capacity of the system (discussed in a later
section), then the design engineer or operator can always
deal with high H2S concentrations in the waste gas by
decreasing the gas flow rate to obtain the desired high
level of control efficiency.
Effect of Concentration of FUS on Its Removal
The effect of H2S concentration in the inlet gas on the
H2S removal efficiency has been investigated under constant
gas flow rate conditions. The results of these studies are
presented in Table 4-5. No significant difference in
control efficiencies is observed when H2S concentrations in
the influent gas stream are varied from 5.5 ppm to 518 ppm
as long as the H2S loading rate is less than the maximum
acceptable value for the compost studied.

91
Table 4-5. Effect of H2S concentration on removal
efficiency.
Gas Flow
Rate
(Lpm)
h2s
Reten. Loading H2S
Time Rate Inlet
(s) (g-S/m-hr) (ppmv)
H,S Removal
Outlet Eff.
(ppmv) (%)
30.0
35.4
0.72
5.51
BDL
99.9+
30.4
35.0
1.69
12.8
BDL
99.9+
30.8
34.5
16.9
126
0.01
99.9+
30.8
34.5
26.3
196
1.31
99.3
30.0
35.4
29.6
227
0.01
99.9+
30.0
35.4
38.1
292
0.01
99.9+
31.2
34.0
70.4
518
1.04
99.8

92
Effect of H2S Loading Rate on Its Removal
One of the most important observations made in this
study is the relationship between H2S reduction and its
loading rate to the biofilter. It is very important and
necessary to introduce here the concept of H2S loading rate
and the maximum elimination capacity of the filter
material. The H2S loading rate is the amount of H2S that is
being introduced to the system per unit volume of the
packing material per unit time (g-S/m3-hr). The maximum
elimination capacity of a compost is the maximum H2S loading
rate that the compost can bear without inhibiting its
microbial activity, and is expressed in the same units as
those used for H2S loading rate. These two parameters
probably play central roles in biofilter design and system
operation.
The maximum H2S elimination capacity of a compost
depends on the microbial population and activity of sulfur
oxidizing bacteria existing in the compost. The latter, in
turn, are related to the operating conditions of the system,
such as temperature, water content, acidity (pH) of compost,
and the concentrations of nutrients and inhibitory
substances.
Overloading of the biofilter system with H2S is
indicated by the appearance of a finely divided, yellowish-
white colored substance on the compost, a sudden decrease in
the H2S removal efficiency or the occurrence of higher
concentrations of elemental sulfur in the compost.

93
The maximum H2S elimination capacity for a compost is
determined at the optimum operating conditions of the
system. H2S concentrations in the inlet gas stream are
varied increasingly at a constant gas flow rate. The H2S
removal rates (g-S/m3-hr) are plotted v.s. the H2S loading
O
rates (g-S/m -hr). The maximum H2S elimination capacity of
the compost is determined when the curve flattens out
(Ottengraf, 1986). The maximum H2S elimination capacity of
compost #13 is determined to be 11.5 g-S/ m3-hr (Figure 4-
11). The maximum H2S elimination capacity of this compost is
very low because its extremely low pH (1.60) and high
sulfate content (70.4 mg-S/g, see Table 4-2).
The maximum H2S elimination capacity for a compost can
also be determined through kinetics studies. Detailed
information is presented in the section of "Kinetics of H2S
Oxidation in a Biofilter". The maximum H2S elimination
capacity for compost #17 is determined to be 129 g-S/m3-hr.
Effect of Compost Water Content on H2S Removal
The effect of compost water content (CWC) on H2S
removal was evaluated using the Column System #4. The
compost (Compost #17) which was used for this study and
other studies carried out using Column System #4 had been
previously packed in Tower #2 and was considered to be pre
conditioned by H2S to a stable operating condition. After
transfer to System #4, the compost was operated at the same
conditions as used in its parent environment, Tower #2,

94
Figure 4-11. Determination of maximum H2S
elimination capacity of compost.

95
before changing the operational parameters. No change in
H2S removal capacity was found due to the system transfer.
Data were not taken until system #4 reached a stable
condition after any changes in the operational parameters or
conditions were made. Each column in System #4 was packed
with equal weights of compost. The range of compost water
contents evaluated was from 0%, oven dried compost, to about
62%, the maximum water holding capacity of the compost.
Water content of the compost in each column was adjusted to
a desired range by either adding DI water to the compost or
by gently drying the compost in room air. One of the
composts tested was thoroughly dried in an oven at 110 C
for 24 hours to obtain water free compost. The system was
operated at room temperature with a gas loading rate of
about 15 m3/m2-hr. Inlet H2S concentrations were controlled
in a range between 80 and 110 ppmv. The results are
illustrated in Figure 4-12.
The H2S removal efficiency was maintained at a high
value, 99.9+%, with little variation being observed when
the CWC was varied from 30% to 62%. When the CWC was
reduced below 30%, the H2S removal efficiency decreased
linearly with the CWC. Very little removal of H2S was
observed for the oven dried compost. The residual
effeciency of the latter is probably due only to chemical
oxidation and adsorption of H2S on the compost.
Water is essential for all living organisms. All
biological metabolic processes require water as a medium or

Hydrogen Sulfide Removal Efficiency (%)
96
Compost Water Content (Wt%)
Figure 4-12. Effect of compost water content on H2S
removal efficiency.

97
solvent. Insufficient water supply can limit the activity
of the microorganisms, which in turn reduces the H2S
oxidation rate. It is also possible that when the CWC
reduced below 30%, there is no free water existing in the
pours of the compost particles. This may decrease the rate
of transfer of H2S from the waste gas to the biofilm where
the biological oxidation of H2S takes place.
Biological activity of the "dry" compost can be
recovered if water is supplied to the compost to bring the
CWC to a proper range. Two composts, WCF4 and WCE3 with
original CWCs of 14.3% and 21.4%, respectively, were used to
study this phenomenon. DI water was added to bring the
water contents of composts WCF4 and WCE3 to 56.5% and 50%,
respectively. The biological activity of both composts as
indicated by H2S removal was recovered eventually up to
99.9+%. The time required for recovery of the activity,
however, is inversely proportional to the dryness of the
"dry" compost. As shown in Figure 4-13, it takes 63 hours
for compost WCF4 to recover its H2S removal efficiency to
99.9+% while compost WCE3 took 41 hours to reach the same
level even though both systems followed a similar recovery
pattern.
Effect of Compost Acidity on PUS Removal
The effect of compost acidity on H2S removal was
investigated using the Column System #4. The source of the
compost is the same as described in the previous section.

Hydrogen Sulfide Removal Efficiency (%)
98
Figure 4-13. Time required for dried compost to
recover optimum efficiency.
Original compost water content:
WCE3 = 21.4%,
WCF4 = 14.3%.

99
The compost used in each column was treated with either
dilute HC1 or dilute NaOH solutions to bring the pH of the
compost to a desired range. The final pH values obtained
for the composts studied in each column are 1.57, 3.20,
4.42, 5.02, 6.39, 6.75, and 8.76, respectively.
Measurements of inlet and outlet gas sample concentrations
were made when the bed operation became stable. The results
of these studies are shown in Figure 4-14. Repeated
measurements at each pH point are used to estimate the mean
value, and mean 2 standard deviations values (error bars)
indicated by the upper and lower bounds of the vertical line
through the mean. Two operational conditions are used in
this test. Under condition A, lower H2S and gas loading
rates are used. No significant effect of pH of compost on
H2S removal is observed for pH values in the range between
3.2 and 8.76. Removal efficiencies of 99.5+% are
consistently determined with little or no variation in this
range. When the pH was reduced to 1.57, the H2S removal
efficiency fell sharply to about 9%. It should be noted
that the measured H2S removal efficiencies showed larger
variations at the lower pH range. Following the previous
studies, the operational conditions were changed to
condition B. Under the latter conditions, higher H2S and
gas loading rates were used and the effects of varing pH the
compost becomes evident. It can be seen in Figure 4-14,
condition B, that the maximum H2S removal occurred at a
compost pH value of 3.2 (99.2%). The removal efficiency

Hydrogen Sulfide Removal Efficiency (%)
100
Compost pH
Figure 4-14. Effect of compost pH on H2S removal
efficiency.
Condition a:
H2S loading rate: 10.5 g/m -hr
Gas loading rate: 15 m3/m2-hr
Condition b:
H2S loading rate: 35.4 g/m3-hr
Gas loading rate: 26.1 nr/m -hr

101
decreased with decreasing acidity up to pH 5 (82.3%), and
then increased with further decrease in acidity of the
compost. As discussed previously in Chapter 2, sulfur
oxidizing bacteria can live in environments with a wide pH
range (1-8) depending on the species present. Probably the
dominant species present in the biofilter systems studied
are acidophiles which prefer an optimum pH value around 3.
For the higher pH value range, chemical reaction between H2S
and the compost material can significantly enhance its
removal in addition to biological oxidation. As a result of
this dual action, higher removal efficiencies of H2S can be
expected.
The acidity of the compost is traditionally expressed
as 'compost pH' for convenience. A more meaningful and
accurate way to express the compost acidity is the 'specific
acidity of compost (SAC)', which gives the quantity of H+
per unit weight of dry compost, /g-H+/g. Specific acidity of
compost can be calculated from compost pH by the following
equation:
SAC = W/[Cw(100-CWC)]X105-PH (4-1)
where:
SAC = specific acidity of compost, /xg-H+/g-dry compost
W = DI water used for compost pH analysis, mL
Cw = Weight of wet compost used for pH analysis, g
CWC = Compost water content, Wt%
pH = compost pH.
For example, if 2g of wet compost is taken and 20 mL of
DI water is used in H+ extraction for a compost pH analysis,

102
and the compost water content is 60% by weight, providing a
compost pH of 1.57, then the dry specific acidity of this
compost is:
SAC = 20/[2(100-60)]xl051,57 = 673 (^q-H+/q).
Effect of Temperature on H2S Removal
The effect of temperature on H2S removal efficiency was
studied using Column System #4 with some minor
modifications. The modified system is shown in Figure 4-15.
Three columns (triplicate) packed with the same compost were
placed in a heating box. A rheostat was used to control the
temperature of the columns. The influent gas stream was
blown through a bubbler, which is placed in a water bath and
heated up to the same or slightly higher (5C) temperature
as the biofilter columns to saturate the gas stream at the
desired temperature. For tests carried out below room
temperature (22C), the reaction columns were placed in a
refrigerator and the temperature adjusted through the
refrigerator thermostat. The temperature range investigated
is between -1.5 to 103C. The results of the triplicate
measurements for each temperature point are plotted means
and 2 standard deviations as error bars in Figure 4-16.
In the range from 25C to 45C, high H2S removal
efficiencies are consistently observed with little
variation. The H2S removal efficiency, however, dropped
rapidly with decreasing temperature in the lower temperature

103
Inlet Gas
Sampling Port
Figure 4-15. Schematic drawing of the experimental
arrangement for investigation of the
effect of temperature on H2S removal
efficiency.

Hydrogen Sulfide Removal Efficiency (%)
104
Figure 4-16. Effect of temperature on H2S removal
efficiency.

105
range. For instance, when the temperature was reduced from
25C to 7.5C, the removal efficiency decreased by
approximately 80%. At 7.5C, only 20% removal efficiency
was observed. On the other hand, the decrease in H2S
removal in the higher temperature range was less marked than
observed at lower temperatures. For example, when the
temperature was increased from 50C to 100C, the H2S
removal efficiency decreased from 97.4% to 40%. The optimum
temperature range determined from these studies is between
30C and 40C, which is the optimum temperature range for
mesophilic bacteria. The removal rates of H2S at high
temperature is probably due to chemical oxidation reactions
in addition to biological oxidation.
Poor performance of a biofilter at low temperatures may
limit their application in cold climates, especially during
the winter. Proper means should be taken to avoid operating
biofilter systems below 10C. For larger biofilters, the
bed temperature can be a few degrees higher than the ambient
air temperature due to biological respiration of the
microbes and the exothermal oxidation reactions in the
filter. Kampbell et al. (1987) reported that soil bed
biofilter functioned well at temperatures in the range 12C
to 24C in Wisconsin. In another study carried out by Rands
et al. (1981) the filter bed temperatures were found to be
10 to 20C higher than ambient air temperatures during
winter times. This type of thermal enhancement, however,
was not observed during this study, probably because the

106
system is not large enough to maintain an adiabatic
condition.
Effect of Sulfate on HnS Removal
The effect of sulfate presence on H2S removal was
evaluated using Column System #4. A preconditioned stable
compost from Tower #2 was divided into 8 equal-weight
portions. Two of the sub-composts with an original sulfate
content of 4.6 mg-S/g were packed directly into two columns
as controls. The other sub-compost portions were mixed with
sodium sulfate (Na2S04) to bring the compost sulfate content
to 24.6, 44.6, 64.6, 84.6, 104.6 and 204.6 mg-S/g,
respectively, and packed in the other 6 columns. After a
few days for acclimation, H2S removal efficiencies for each
column were determined. The system was operated at a gas
loading rate of 15 m3/m2-hr and an H2S loading rate between
6.6 and 8.4 g-S/m3-hr. The results are illustrated in
Figure 4-17.
No effect is observed when the compost sulfate content
is less than 25 mg-S/g. However, a significant effect is
observed at higher sulfate levels. The H2S removal
efficiencies were reduced from 99.94% to about 35% and
remained in the lower removal efficiency range when the
sulfate content was increased from 45 to 200 mg-S/g. These
results suggest that a sulfate content of 25 mg-S/g is a
critical level for the microbial environment. Above this
level sulfate probably reaches a toxic level and the

Hydrogen Sulfide Removal Efficiency (%)
107
Figure 4-17. Effect of sulfate on H2S removal
efficiency.

108
activity of these microorganisms is markedly inhibited.
This observation is very important for biofiltration control
of H2S. Since sulfate is the final product of the
biofiltration process, it may accumulate in the biofilter
bed if no other action is taken. Accumulation of sulfate
can easily reach a level that can significantly reduce the
function of the biofilter. Measures to avoid sulfate
accumulation in filter and to enable recovery of the
deteriorated compost are discussed in Chapter 5.
Effect of Nutrient Addition on H2S Removal
Low and high sulfur-containing composts were used for
investigation of the effects of nutrient addition on H2S
removal. The total-S contents are 17.5, 33.7, 20.2 and
119.7 mg-S/g for compost A, B, C, and D, respectively. Each
compost was packed in two columns, one is used as a control,
and the other was treated with nutrients. Fifty (50) mL of
nutrient solution was mixed with 140 g of the compost tested
in each column. Excess water was removed by exposing the
compost to room air for 24 hours. The nutrient added is
similar to the enrichment medium for sulfur-oxidizing
bacteria suggested by Aaronson (1970). The composition of
the solution is: K2HP04, 1.0 g; MgS04.7H20, 0.5 g; NH4N03,
1.0 g; CaC03, 10 g, and DI water was added to bring the
final volume to 1000 mL.
The results on the effect of nutrient addition on
biofilter performance are presented as a bar graph in Figure

109
4-18. For each compost set, H2S removal efficiencies are
plotted for the control compost and the test compost before
and after the addition of nutrient solution. It can be seen
that the H2S removal efficiencies are significantly
decreased for the three composts tested with low sulfur
content when the nutrients are added. The reason for this
decline in efficiency is not clear. No improvement of H2S
removal for the high sulfur containing compost was observed,
either.
Kinetics of H2S Oxidation in a Biofilter
Theoretical Considerations
In general, the substrate utilization rate of a
component by microbial flora as well as the enzymatic
reaction rate are expressed by the Michaelis-Menten
relationship (White et al., 1978? McGilvery and Goldstein,
1983; Schmidt, et al., 1985):
-dC/dt = VmaxCB/ (Kjjj + C) (4-2)
where C is the substrate concentration, B is the population
density, Vmax is the theoretical maximum specific reaction
rate, Km is the half-saturation constant (Michaelis
constant), and t is the reaction time.
Under steady state conditions, i.e., when the microbial
population does not change with time, there are three
situations which may be encountered in a biological reaction
system, and corresponding equations can be derived, from

Hydrogen Sulfide Removal Efficeincy (%)
110
Figure 4-18. Effect of nutrient addition on H2S
removal.
Total-S content in compost (mg-S/g)
A: 17.5
B: 33.7
C: 20.2
D: 120

Ill
equation 4-2 above, to express the rates of biological
reaction for each particular situation:
1) If the substrate concentration is very high, i.e.,
when Km << C, the rate expression approaches zero-order
kinetics in the substrate concentration:
-dC/dt = VmaxB = kQ (4-3)
where kQ is a zero-order rate coefficient. The integral form
of equation 4-3 becomes:
c = co V (4"4>
where CQ and C are the initial substrate concentration and
the substrate concentration at time t, respectively.
If CQ C is plotted against t, a straight line should
be obtained and the slope kQ is the maximum elimination
capacity of the microbes for the substrate.
2) If the substrate concentration is very low, i.e.,
when C << 1^, a first-order kinetics dependence should be
obtained:
-dC/dt = k-j^C (4-5)
The integral form for equation 4-5 becomes:
C = CQ exp(-k1t) (4-6)
where k-j^ is the first-order reaction rate coefficient.
If ln(0/Co) is plotted against t, a straight line
should be obtained and the first-order reaction coefficient,

112
klf can be determined from the slope of the line.
3) In the third situation, when the half-saturation
constant, Km, and the substrate concentration, C, are
comparable, the biological reaction should follow fractional
order kinetics, because equation 4-2 can not be simplified.
Relatively complex equations have to be derived to express
the fractional order kinetics.
A number of theoretical and empirical models have been
reported which describe the kinetics of biodegradation of
organic compounds and reduced sulfur species (Chen and
Morris, 1972; Williamson, 1973; Cooper, 1974; Jennings et
al., 1976; Williamson and McCarty, 1976a,b; O'Brien and
Birkner, 1977; Rittmann et al., 1978; Schmidt et al., 1985;
Kampbel et al., 1987; Caunt and Hester, 1989). In
particular, Ottengraf and coworkers have published a series
of papers that describe the processes involved in the
operation of biofilters (Ottengraf and Van Den Oever, 1983;
Ottengraf et al., 1984; Ottengraf, 1986; Ottengraf et al.,
1986; Ottengraf, 1987). Ottengraf developed a biophysical
model as well as derived mathematical solutions to describe
the kinetics of biodegradation of various organic compounds
in biofilter systems.
As described previously in chapter 2, Ottengraf's model
uses the concept of a bed consisting of solid filter
particles, where each particle is surrounded by a wet,
biologically active layer. When waste air flows around the
particle there is continuous mass transfer of pollutant from

113
the gas phase to the biolayer (See Figure 2-3).
The overall kinetic behavior observed in a biofilter is
a result of the interaction between mass transfer phenomena,
the microkinetics of the biological elimination reactions,
the residence time distribution of the gas flow, etc.. This
overall kinetic behavior is termed 1 macrokinetics by
Ottengraf and can be determined experimentally.
Ottengraf divided the macrokinetics in a biofilter into
two classes: first-order reaction and zero-order reaction.
He expressed the first-order reaction kinetics in a form
similar to equation 4-4 described previously. For zero-
order reaction kinetics, however, Ottengraf has
distinguished the following two situations:
1. At gas phase concentrations, C, above a compound
specific, critical concentration (Ccr^t), the film will
be fully saturated (Figure 2-3, Case I) and pollutant
elimination is limited by the biological activity in
the biofilm. This process is defined as reaction
limitation.
2. At concentrations less than Ccr^t, diffusion in the
biofilm will limit compound removal. The biofilm is no
longer fully penetrated (Figure 2-3, Case 2) and the
removal rate decreases with decreasing pollutant
concentration in the waste gas. This process is
referred to as diffusion limitation.
Ottengraf has also derived two equations which describe
the kinetics for either situation.
For the first situation, i.e. under reaction limiting
conditions, the kinetics expression is:
C/C0 = 1 K0H/C0Ug
(4-7)

114
where C is the effluent gas concentration, CQ is the
influent gas concentration, H is the height of the filter
bed and Ug is the gas velocity. This equation is a
different way of expressing equation 4-4, specifically for a
biofilter.
Under diffusion limiting conditions, the kinetics
expression is:
C/CQ = [1 (H/Ug)(K0Dea/2mC06)^]2 (4-8)
where a is the interfacial area per unit volume, De is the
effective diffusion coefficient, m is the distribution
coefficient of the component, and <5 is the biolayer
thickness.
Determination of the Kinetics of H2S Oxidation in a Biofilter
Hydrogen sulfide elimination rates are measured when
the biofilter has reached a steady state conditions. Gas
samples are taken at different locations along the tower and
the H2S concentrations in the gas samples are analyzed. The
initial H2S concentrations, CQ, in the inlet gas stream are
varied from low to high values in order to determine the
kinetics under different situations as described in the
previous section. The results for compost #17 packed in
Tower #2 are presented in Figures 4-19, 4-20, 4-21, 4-22,
and 4-23.
Under high H2S concentrations in the inlet gas stream
(H2S > 400 ppmv), the reaction appears to follow zero-order

r C (ppmv)
115
Reaction Time (sec)
Figure 4-19. Linear least squares regression
analysis for zero-order kinetics of H2S
oxidation in biofilter. Gas loading
rate: 224 m3/m2-hr, compost #17.

116
d>
O
5
Figure 4-20. Linear least squares regression
analysis for first-order kinetics of
H2S oxidation in biofilter. Gas loading
rate: 224 m3/m-hr, compost #17.

(C/C0)
117
Figure 4-21. Determination of the fractional-order
reaction rate coefficient, kf by linear
least squares regression. Gas loading
rate: 224 m3/m-hr, compost #17.

C/C
118
Figure 4-22. Plot showing the fractional-order
kinetics of H2S oxidation in biofilter.
Gas loading rate: 224 nr/m -hr, compost
#17.

119
o
O
O
1.2
1.1
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
C0= 434 ppmv
Zero-order Kinetics
C0= 206 ppmv
First-order Kinetics
C0 = 309 ppmv
Diffusion limitation
h/H
Figure 4-23. Concentration profiles for t^S as a
function of packing height within the
biofilter. Gas loading rate: 224
itr/m-hr, compost #17.

120
kinetics according to equation 4-4 (Figure 4-19). The
regression formula is expressed as:
C = CQ 27.6t + 21.8 (4-9)
with a correlation coefficient R2 of 0.9907. The good
agreement obtained supports the conclusion that the reaction
can be described by zero-order kinetics. The value
obtained for KQ is 27.6 ppmv/s. This zero order coefficient
conresponds to a maximum H2S elimination capacity of 130 g-
S/m3-hr for the compost at the operating conditions
selected. It should be noted that the second and third data
points in Figure 4-19 are below the regression curve. This
deviation is mainly due to the H2S removal efficiency in the
lower portion of the filter has been effected by high
acidity and sulfate content in the compost in this portion
of the bed as a result of prolonged operation.
For lower inlet H2S concentrations (H2S < 200 ppmv),
the reaction appears to follow first-order kinetics as
expressed by equation 4-6 (Figure 4-20). The linear
regression formula is expressed as:
ln(C/CQ) = 0.31 0.57t (4-10)
In this case the value for k-^ is 0.57/s, which is
applicable for inlet H2S concentration of less than 200
ppmv.
When H2S concentrations fall in the intermediate range,
i.e., between 200 ppmv and 400 ppmv, the data cannot be

121
represented by either Equation 4-4 (zero-order kinetics) or
equation 4-6 (first-order kinetics). Several empirical
equations have been tested, and Ottengraf's diffusion
limiting model was found to be the most accurate expression
describing H2S oxidation kinetics in this range for the
operating conditions employed.
In order to conveniently use Ottengraf's model, and
equation 4-8, however, it is necessary to define a new
parameter, the fractional-order reaction coefficient kf:
kf = (k0Dea/2miC05)^ (4-11)
It can be seen that kf is a function of the operating
conditions of the biofilter system, and under steady state
conditions, kf is constant.
Thus, equation 4-8 can be rewritten as:
C/CQ = (l-kft)2 (4-12)
or
(C/Cq)55 -1 = -kft (4-13)
where t = H/Ug is the reaction time.
When (C/C0)^ -1 is plotted against t, a straight line
should be obtained, and the slope of the line should equal
kf. A sample plot of the relation in equation 4-13 for the
concentration range 200 to 400 ppmv H2S is shown in Figure
4-21. The reasonableness of this plot supports the concept
of fractional-order dependence in the stated range of
concentrations.

122
Thus, it appears that equation 4-12 can be used to
predict the kinetics of H2S oxidation in the intermediate
H2S concentration ranges.
A set of experimental data (H2S =309 ppmv) are plotted
in Figure 4-22, where the symbols used are experimental data
points, and the solid curve is drawn according to equation
4-12. It can be seen that the experimental data are
adequately described by the diffusion limitation model.
The data shown in Figures 4-19, 4-20, and 4-22 are
plotted in Figure 4-23 in the "Ottengraf-form", i.e., C/CQ
on the y-axis and h/H on the x-axis, where h is the height
of sampling location on the filter and H is the total height
of the filter. It is informative to use this plot in order
to observe the H2S concentration profile in the biofilter.
According to Ottengraf, under reaction limiting
conditions, a straight line should be obtained for this type
of plot. However, the profile is no longer linear if the
substrate elimination rate is controlled partially or
completely by the diffusion rate in the biofilter. The
latter feature is clearly observed in this study of H2S
elimination. The data that are plotted in Figure 4-20 shows
a straight line in Figure 4-23. This linear relation is an
indication of zero-order kinetics and reaction limiting
control by the microbial population. When inlet H2S
concentrations are less than 434 ppmv (the middle curve) ,
the profile is curved (concave) below the straight line,
indicating that the reaction is partially diffusion

123
controlled at intermediate H2S concentrations. With
further lower H2S concentrations, the profile becomes curved
well below the straight line, and suggests that the dominant
mechanism is diffusion limited in this first-order kinetics
region.
Table 4-6 summarizes the equations describing the
kinetics of H2S oxidation in the biofilter system. These
equations allow for a quantitative description of the basic
processes involved in this biofiltration elimination of H2S
and they allow for an accurate sizing of biofilters for H2S
removal. It should be noted that the macrokinetics of H2S
oxidation in biofilters as well as the reaction coefficients
reported here are related to operational conditions. The
kinetics are valid for a similar compost biofilter system
operating under similar conditions. If operational
conditions such as pH, temperature, sulfur content, etc. are
changed, then the kinetic behavior may be altered. In
practice, the kinetic behavior for a particular compound
should be determined by laboratory or pilot scale studies
(Van Lith, 1989; Leson and Winer, 1991).

124
Table 4-6. Models for the kinetics of H2S oxidation in
biofilter.
H2S
Concen.
Range
(ppmv)
Kinetic
Order
Reaction
Coefficient
Equation
<200
First
k-L = 27.6/s
-dC/dt
c = cG
= ki C
exp(-k^t)
200-400
Fractional
kf = 0.067/s
c = co
(l-kft)2
>400
Zero
kQ = 0.567ppmv/s
-dC/dt
C = CQ
i ii
** **
o o
rt

CHAPTER 5
BIOFILTER PERFORMANCE AND CHANGES IN COMPOST PROPERTIES
ASSOCIATED WITH LONG TERM OPERATION
Towers #1 and #2 were packed with composts #17A and
#17, respectively, and operated continuously for more than
200 days. Tower #3 was packed with compost #16 and operated
for 130 days. Long term performance of these biofilter
tower systems for H2S removal and changes in compost
properties are reported here. Biofilter maintenance
conditions and recommended procedures for developing optimum
performance have been determined through these long term
observations.
Overall Performance of the Biofilters
The overall performance of Towers #1, #2 and #3 are
presented graphically in Figures 5-1, a, b and c,
respectively. In each figure, H2S loading rates,
inlet/outlet H2S concentrations and H2S removal efficiencies
are plotted against the cumulative operation time. Except
for some specific tests, where extreme operating conditions
were used (such as high H2S loading rate, high gas flow
rate, etc) and data are presented and discussed elsewhere,
the data showed in these figures are daily averages for an
individual day when the measurements were made. Usually,
Towers #1 and 2 were operated at a gas loading rate of 100
125

H2S Loading Rate (g-S/m3-hr) H2S Concentration (ppmv) H2S Removal
Efficiency (%)
126
Figure 5-1. Biofilter control of H2S during long
term operation,
a) Tower #1, compost #17A.

H2S Loading Rate (g-S/m3-hr) H2S Concentration (ppmv) H2S Removal
Efficiency (%)
127
Figure 5-1. Biofilter control of H2S during long
term operation,
b) Tower #2, compost #17.

H2S Loading Rate (g-S/m3-hr) HgS Concentration (ppmv) HgS Removal
Efficiency (%)
128
Cumulative Operation Time (Day)
Figure 51. Biofilter control of H2S during long
term operation,
c) Tower #3, compost #16.

129
m3/m2-hr.
All three composts studied showed very good overall
performance characteristics during continuous long term
operation. The acclimation periods for all three composts
are similar, approximately 10 days. After this period, the
filters achieved stable operating conditions.
Tower #3 was operated under a lower gas loading rate,
50 m3/m2-hr (Figure 5-lc). This tower was not washed during
130 days of operation. When unpacked, the color of the
compost at the bottom of the tower (inlet) had changed to
yellowish-white, indicating accumulation of sulfur. The
compost at the bottom of the tower was wet, with a water
content of 55.9%, but the compost became drier as samples
were analyzed at various distances up the tower. The water
content of the compost in the top portion of the biofilter
(exit) was only 25.8%. This characteristic is probably
common for closed biofilter systems with no additional water
supplied to the system except water contained in the
influent gas stream. Although the gas is prehumidified and
almost saturated with water (RH > 95%) at ambient
temperature, the temperature in the compost can be a few
degrees (C) higher than the incoming gas stream due to the
exothermic oxidation reactions occurring in the system and
the biological respiration of the biomass. The rise of
compost temperature is more significant with high H2S
loading rates. The elevated bed temperature causes the gas
stream to become more unsaturated, and results in loss of

130
water by evaporation from the compost. Drying-out of the
compost, therefore, is a natural feature of any biofilter
system. The drying-out process may be slower for those
biofilters which are used to remove hydrocarbons because
water is one of the products of the biodegradation
reactions. There is no water formed during the biological
oxidation of H2S. As a result, water has to be added to the
system at the effluent end of the bed to keep the compost
water content constant.
Towers #1 and #2 are packed with the same compost (#17)
except that 2% by weight of CaC03 was added to the compost
packed in Tower #1 (Compost #17A) as a pH buffer. It should
be noted that the addition of CaC03 did not affect the
performance of the biofilter.
These two towers have been packed and operated since
January 29, 1991. Both of these towers have been subjected
to large variations in waste gas surface loading and H2S
loading rates during the operation period because the
effects of these variables on H2S removal efficiencies were
evaluated in this dual-tower system. High H2S removal
efficiencies and stable performance, were consistently
observed.
The composts (#17 and #17A) showed good moisture
retention and buffering capacity. In addition to the water
added by pre-humidification of the inlet gas stream, these
two towers were washed biweekly using DI water. The latter
procedure was performed to keep the compost water content in

131
the desired range. The main purpose of this procedure,
however, is to reduce the acidity and prevent accumulation
of sulfate in the compost. A typical compost water content
distribution profile in Tower #2 is presented in Figure 5-2.
Samples were taken and analyzed before washing (14 days
after the last washing) and 1 hour after washing the tower.
It can be seen that 1) the water content of the compost
is evenly distributed along the length of the bed, and 2)
the compost has very good water retention and buffering
capacity. Only 3-5% of the compost water content was lost
during the 14 day interval between washings. When the
biofilter tower was operated for 130 days and samples were
taken and analyzed, the water holding capacity of the
compost in the inlet region had decreased slightly compared
to that in the outlet region.
The system showed a good buffering capacity to gas
surface loading changes. No significant reduction in H2S
removal efficiency was observed when the gas surface loading
rate was varied in the range 20 to 500 m3/m2-hr for the
same H2S loading rate.
Also, the buffering capacity for H2S loading rate or
H2S concentration changes was very good. Under the same
conditions of gas surface loading rate ( 100 m3/m2-hr), when
H2S concentrations were changed from 15 ppmv to 775 ppmv
(corresponding to the H2S loading rate being changed from 2
to 50.5 g-S/m3-hr) the H2S removal efficiency did not vary
from 99.5%. Sudden changes in H2S loading rates over a

Compost Water Content (%)
i
132
Height Above Bed Inlet (m)
After Washing PKxa Before Washing
Figure 5-2
Compost water content profile.

133
very large range, for example, from a few tens ppmv to a few
hundreds ppmv may cause a temporary reduction in H2S removal
efficiency. However, after only a few hours, the optimum
control efficiency was recovered (Figure 5-1 a and b). This
perturbation is probably due to the uneven and inadequate
initial distribution of the sulfur oxidizing bacteria
population. In other words, the length of the 'active
portion' of the filter is related to the H2S concentration
in the gas stream. This feature will be discussed in a
later section.
One of the most significant changes observed is the
compost pH. Compost pH changes in different sections of
Towers #1 and #2 are shown in Figures 5-3 a and b,
respectively. The product of H2S oxidation is sulfuric acid
(H2S04). This strong acid is soluble in water and
accumulates in the compost, resulting in rapid acidification
of the system. Compost acidity increases very rapidly with
time, e.g. after 32 days of operation, the pH of the bottom
section (inlet) of the compost dropped from 8.0 to 1.5
(TS11, Figure 5-3a). The rate of compost pH change is
proportional to the H2S input. The pH drop in the lower
portion of the compost is much larger than that in the upper
portion of the bed since most of the H2S oxidation reaction
takes place in the former region. When the gas stream flows
through the bed less H2S is left in the gas stream and less
H2S04 is formed as the bed is traversed, therefore, the
acidification is slower in the upper portion of the tower.

Compost pH
134
Cumulative Operation Time (Day)
a TS11 + TS12 O TS13 A TS14 x Wash Water
Figure 5-3. pH changes of compost in different
sections of the biofilter with
operation time,
a) Tower #1, compost #17A.

Compost pH
135
Cumulative Operation Time (Day)
3 TS21 + TS22 O TS23 a TS24 x Wash Water
Figure 5-3. pH changes of compost in different
sections of the biofilter with
operation time,
b) Tower #2, compost #17.

136
Washing the tower with water effectively mitigates the
pH decline. Since H2S04 is water soluble, a major fraction
of the accumulated acid can be washed out at each washing.
The pH of the wash water for each treatment is lower than
the compost pH (by approximately 1 to 1.5 pH units). A few
measurements of compost pH before and after washing were
conducted and the results of these measurements are
indicated in Figures 5-3 a and b. An increase in compost pH
from 0.2 to 0.5 pH units was achieved during each washing,
thus, if the tower is washed routinely, compost pH can be
kept constant.
The effectiveness of the washing process on pH
stabilization depends on the quality of the water used and
the contact time between the compost and water. Both towers
#1 and #2 are washed by 10 L of DI water each time with a
flow rate of 1 L/min from the top of the tower. The water
is allowed to freely flow downward through the tower under
gravity, as a result, the water-compost contact time is
approximately 10 minutes.
Because the rate of pH decrease is proportional to the
H2S loading rate to the system, with a high H2S loading rate
the tower needs to be washed more frequently.
Addition of lime or CaC03 to eliminate acidification of
the biofilter is not effective and is not recommended.
There are some disadvantages in adding lime to the biofilter
system:

137
1) Premixed lime is effective only temporarily, lime can
be consumed by the accumulated acid very quickly;
2) Since high concentrations of H2S04 are continually
formed, the quantity of CaC03 required to neutralize
the acid is very large. Addition of large amounts of
lime increases the inorganic fraction of the filter
medium and significantly changes the compost
construction and composition.
3) The addition of lime increases the smaller particle
fraction in the filter, which results in a significant
increase in pressure drop across the filter bed, and
4) Most importantly, addition of lime does not solve
the problem of sulfur accumulation in the compost,
which appears to be the main reason for the decline in
H2S removal efficiency.
Accumulation of Sulfur in Compost
and Its Effect on System Performance
Another serious problem which is frequently encountered
in a H2S-biofilter system is the accumulation of sulfur in
the filter material (Carlson and Leiser, 1966; Rands et al.,
1981; Yang and Allen, 1991). This feature has been
routinely observed during the course of this study in both
Towers #1 and #2 since these towers are transparent. During
long term operation of the biofilter the color of the
compost eventually changes from dark brown to a yellowish
white. The color change progresses from the lower region of

138
the bed (inlet) to the upper layer. White deposits on the
surface of compost particles are easily observed.
The rate of sulfur deposition is proportional to the
rate of H2S loading. A sudden increase in H2S loading in a
large concentration range and prolonged operation at high
H2S loading rates can cause the white colored material to
accumulate rapidly and spread from the lower region to the
upper region of the bed. This discoloration of the bed is
accompanied by a rapid drop in pH of the compost. Also, the
temperature of the biofilter system can rise 2-3 C for high
H2S loading rates indicating enhanced biological activity of
the microbes. If no appropriate action is taken to
counteract sulfur accumulations, then the system
performance and H2S removal efficiency will decline rapidly.
In biofiltration processes, H2S is oxidized both
chemically and biologically to sulfate under aerobic
conditions. In nature a variety of reduced inorganic sulfur
compounds (e.g. elemental sulfur, thiosulfate) occur as
intermediates between sulfide and sulfate, the reduced and
oxidized forms of sulfur, respectively. As these compounds
are oxidized only slowly by direct chemical reaction with
oxygen (Kuenen, 1975), it is clear that biological oxidation
must play an important role in the recycling of reduced
sulfur compounds under aerobic conditions. This mechanism
appears to be true, also, in biofiltration systems.
Also, many microorganisms can oxidize reduced sulfur
compounds, the colorless sulfur bacteria are known to play a

139
major role in the oxidation of reduced sulfur. The
colorless sulfur bacteria are divided into three families,
the Thiobacteriaceae, the Beggiatoaceae, and the
Achromatiaceae. The genera, which belong to these families,
have been studied and include Thiobacterium. Macromonas.
Thiovulum. Thiospira. Thiobacillus. Thiomircospira.
Sulfolobus. Beqqiatoa. Thiospirillopisis. Thioploca.
Thiothrix, Thiodendron. and Achromatium (Kuenen, 1975).
The colorless sulfur bacteria are naturally occurring
almost everywhere on the earth. These bacteria live in a
wide pH range from 1 to 8, and temperatures up to 85C
(Kuenen, 1975; Brock and Madigan, 1988).
To distinguish between the active genera and species of
sulfur bacteria is not the goal of this study. The macro
processes of H2S oxidation, which involve physical, chemical
and biological processes are of greatest interest and
application in this study.
A number of oxidation reactions of inorganic sulfur
compounds which are effected by the colorless bacteria,
especially thiobacilli, have been reported (Starkey, 1966).
Some of the important reactions that possibly occur in a
biofilter system are as follows:
H2S + 202 H2S04 (5-1)
2H2S + 02 -* 2S + 2H20 (5-2)
2S + 302 + 2H20 2H2S04 (5-3)
Na2S203 + 202 + H20 - Na2S04 + H2S04 (5-4)
4Na2S203 + 02 + 2H20 2Na2S406 + 4NaOH (5-5)

140
2Na2S406 + 702 + 6H20 2Na2S04 + 6H2S04
(5-6)
5H2S + 8KNO3 4K2S04 + H2S04 + 4N2 + 4H20
(5-7)
5S + 6KNO3 + 2H20 3K2S04 + 2H2S04 + 3N2
(5-8)
5Na2S203 + 8NaN03 + H20 -
9Na2S04 + H2S04 + 4N2 (5-9)
From the reactions listed above and Table 2-1 in
Chapter 2, it can be seen that the colorless sulfur bacteria
can oxidize both hydrogen sulfide (H2S) and the intermediate
reduced sulfur compounds to sulfate. Different sulfur
compounds, therefore, in various stages of oxidation can be
expected to be present in the biofiltration system.
The original compost, #17A, and compost samples in the
biofilter, after 3 months continuous operation, were
collected and analyzed for total-S and for fractionation
into various sulfur components. Compost samples in the
biofilter system were taken from the lower (TS11, 0.125m),
middle (TS13, 0.625m), and upper (TS14, 0.875m) regions of
the tower. Total-S and the following sulfur components:
ester-S, FeS2-S, FeS-S. S-S, water-soluble-S,
P-extractable-S and insoluble-S, were analyzed using the
methods described in the previous section. S042-S and
inorganic-S are estimated according to individual analyses.
Organic-S is calculated as the difference between total-S
and inorganic-S, and C-bonded-S is the difference between
organic-S and ester-S. The results are summarized in Table
5-1.

Table 5-1. Sulfur fractionation of original compost #17A and compost at different heights
in the filter.
Original
TS11 (0.125m)
TS13 (0.625m)
TS14 (0.875m)
(mg-S/g) (%)
(mg-S/g) (%)
(mg-S/g) (%)
(mg-S/g) (%)
Total-S
0.740
100
129
100
13.0
100
5.18
100
Organic-S
0.470
63.5
6.05
4.69
3.07
23.6
0.190
3.67
C-bonded-S
0.420
56.8
4.69
3.63
2.87
22.0
0.00
0.00
Ester-S
0.050
6.76
1.36
1.05
0.200
1.54
0.190
3.67
Inorganic-S
0.270
36.5
123
95.3
9.95
76.4
4.99
96.3
FeS2-S
0.050
6.76
22.3
17.3
0.310
2.38
0.800
15.4
FeS-S
0.030
4.05
0.240
0.190
0.090
0.690
0.100
1.93
S-S
0.00
0.00
8.44
6.54
0.24
1.84
0.12
2.32
so4 Z-S
0.19
25.7
92.1
71.3
9.31
71.5
3.97
76.6
Water-sol.-S
0.090
12.2
66.6
51.6
6.71
51.5
3.09
59.6
P-extract.-S
0.00
0.00
18.1
14.0
1.19
9.14
0.440
8.49
Insoluble-S
0.100
13.5
7.36
5.70
1.41
10.8
0.440
8.49
141

142
The original compost has a total-S content of 0.74 mg-
S/g, 64% of which is organic-S. The total-S and sulfur
constitution of the original compost are close to similar
data reported by David et al.(1982) for surface soils in a
forest. High organic-S content is a good indication of high
biomass and microbial sulfohydrolase activity of the
compost. This correlation was quantitatively determined by
David and coworkers (1982).
Inorganic-S is dominant in the 'used' filter compost
samples (>95%). A lower value for TS13 is probably due to
analytical error. Fifty to sixty percent of the total-S in
the biofilter compost is water soluble-S, which indicates
that the final product of H2S oxidation is H2S04. A large
amount of FeS-S and S is measured in the lower region of
the filter bed (TS11, 0-0.125m of the filter).
The total-S distribution profile is graphically shown
in Figure 5-4. As the test gas flows through the filter bed
in an upward direction, the lower region of the compost bed
is always exposed to higher H2S concentrations than the
upper region. This results in sulfate formation and
accumulation at a higher rate in the lower region of the
bed. The higher fraction of intermediately oxidized sulfur
compounds, FeS and S, in the lower region is a result of
incomplete oxidation of H2S due to high H2S concentrations
and reduced biological activity of the biomass as a result
of lowered pH and high sulfate content in this region.

Total-S (mg-S/g)
143
0.125
0.375 0.625 0.875
Height Above Bed Inlet (m)
Original
Figure 5-4. Total-S distribution profile in
biofilter, Tower #1, after exposure to
H2S for 100 days.

144
Biological activities of the biomass in each region of
the filter bed material were indirectly determined by
measuring sectional H2S removal efficiency. Identical gas
flow rates and inlet H2S concentrations were used for these
measurements. The results are shown in Figure 5-5. It can
be seen that the most effective region in the biofilter is
between 0.2 and 0.4m. The biological activity of the lower
region (0-0.2m) is restricted by the factors mentioned
above. The maximum population of the sulfur oxidizing
bacteria as well as the optimum biological activity occur in
the second region (0.2-0.4m) of the filter and decreases
progressively up the bed. This observation is reasonable
because less and less H2S is available in the gas stream as
it passes upward through the filter and more of the H2S in
the gas stream is eliminated by reaction with the biofilter
in the lower region. The population of the sulfur oxidizing
bacteria and the biological activities appear to show a
modal (Gaussian type) distribution along the filter bed.
With prolonged operation, the mode will move upward from the
lower region of the bed due to the increasing toxicity
caused by accumulation of sulfate and acidification of the
compost in the lower region. The latter effect is referred
to as system upset, which must be avoid to maintain
effective control efficiencies.

H2S Removal Efficiency (%)
145
110
Section of Filter
Figure 5-5. H2S removal efficiencies in different
regions of the biofilter, Tower #2.
A: 0-0.2 m
B: 0.2-0.4 m
C: 0.4-0.6 m
D: 0.6-0.8 m
E: 0.8-1.0 m

146
System Upset and Recovery
System upset is indicated by a sudden decrease in H2S
removal efficiency, increased H2S concentration in the
effluent gas stream and noticeable objectionable odor. The
most common reasons for system upset are compost dry-out and
H2S overloading.
A dry compost system can be easily determined by
measurement of compost water content. As discussed
previously, if the compost water content dropped below 30%,
reduced H2S removal can be expected. For a closed system
without additional water supply a dry region is generally
observed in the upper portion (exit) of the biofilter.
Drying of compost also causes shrinking of the compost and
results in channeling. This particular feature is indicated
by a decreased pressure drop across the filter bed.
The dry-out problem can be solved by spraying water at
the exit (top) of the compost filter. Channels generally
disappear as a result of the compost volume expanding after
watering. The system may need a few days to rebuild the
microbial population and recover its optimum performance
depending on its original dryness, as described in previous
sections.
Another cause of system upset is overloading the system
with high H2S concentrations. The maximum H2S elimination
capacity for a filter medium depends on the nature of the
material and the operating conditions of the system.
Generally speaking, the maximum H2S elimination capacity of

147
a filter material is determined at optimized operating
conditions. Changes in operating conditions, such as a
lowered pH, dry-out of the compost, accumulation of sulfur
in the bed material, etc. can significantly decrease the H2S
elimination capacity of the filter medium. Therefore, the
actual H2S elimination capacity of a filter material at a
specific condition is always equal to or less than its
maximum capacity. When the H2S loading rate exceeds the
elimination capacity of the compost, then the system is
overloaded.
An overloaded system is indicated by a high H2S
concentration in the effluent gas stream, a low removal
efficiency, noticeable odor and the occurrence of a compost
color change (white deposit on compost particles) Local
overloading is often observed when channeling occurs in the
system (Rands et al., 1981) or the influent gas stream is
not evenly distributed. In either case, the region of the
filter where the high flow rate occurs is overloaded by H2S.
Local overloading can be cured by correcting the channeling
and the gas distribution system.
If the whole system is overloaded by high H2S input,
the solution to the problem is different. If the system is
temporarily overloaded for a few hours, the performance of
the filter can be recovered by decreasing the H2S loading
rate. A certain fraction of the white deposits on the
compost are intermediate oxidation products, such as S,
FeS2, S 2 0 3 ^-, etc. When H2S loading is decreased, the

148
microbial population uses these intermediate oxidized
compounds as its energy source and oxidizes the
intermediates to sulfate. As a result, the white color and
other deposits disappear within a few days. However, the
compost pH is significantly decreased and the sulfur content
of the compost is increased due to the formation of large
amounts of sulfuric acid. The latter results in the H2S
elimination capacity of the compost being reduced if no
counter measures are taken.
If the system is continuously overloaded by high H2S
input, the deterioration of biofilter performance can not be
reversed by simply decreasing the H2S loading rate. The
biological activity of the microorganisms in the filter is
strongly inhibited by high sulfate content and very low pH.
In this case, specific treatment is required to recover the
defective compost and the deteriorated system.
In order to determine a proper method to recover the
defective compost, a set of experiments were designed and
conducted. The resulting information is provided in the
following sections.
Selection of Chemical Solutions
The Column System #4 was used for this test. The
defective, white-colored compost was obtained from the
bottom region of Towers #1 and #2 and mixed thoroughly. One
hundred and fifty grams (150 g) of this compost was packed
in each of the 8 columns. Waste gas containing H2S was

149
introduced to the system at a flow rate of 230 mL/min
(Equivalent to 15 m3/m2-hr surface loading rate). Influent
and effluent gas samples were analyzed to determine the
original performance of the compost before treatment. The
compost is then unpacked and divided into 4 groups, each
group was packed into two columns as duplicates. The first
group was treated with DI water, the second group with 0.05M
NaOH, third group with 0.05M NaHC03 and the fourth group was
left untreated as a control. For each treated group, the
compost was shaken with 10 times the liquid (by weight) for
30 minutes with a rotary shaker. The composts are treated
this way twice. After each shaking treatment, the compost
pH was measured. The total-S of the original defective
compost and the composts after treatment were measured. The
results of these compost analyses are summarized in Table 5-
2. It can be seen that after being washed twice, the pH of
water treated compost was raised 0.3 pH unit, but both the
NaOH and NaHC03 treated composts are neutralized to near pH
7. The total-S content of the compost was successfully
reduced by 72 to 85% in all cases. According to this study,
water appears to be the best washing agent.
After treatment, the composts are repacked into the
columns, and the system subjected to an H2S removal test.
The system was allowed to operate continuously for one week
before gas sampling and analysis was conducted. The delay in
testing was included to eliminate the effects of residual
alkali on H2S removal for those composts treated by NaOH and

150
Table 5-2. Effect of washing on compost pH and sulfate
content by DI water, NaOH and NaHC03
solutions.
Reduction
Treatment
Time of
Treatment
pH
Total-S
(mg-S/g)
of Total
(%)
Untreated
(control)
1.59
120

Water
First
1.61


Washed
Second
2.20
17.5
85.4
NaOH
First
1.73


Washed
Second
7.51
33.7
71.9
NaHCO-,
Washed
First
1.86


Second
6.66
20.2
83.1

151
NaC03
The systems were tested under two different operating
conditions, a) at a low H2S loading rate, 36 g-S/m3-hr and
b) at a higher H2S loading rate, 71 g-S/m3-hr. At the lower
loading rate, the H2S removal efficiency increased from
20.6% to 99.9+% for all three treated composts. The H2S
elimination capacities of these composts were determined at
high H2S loading rates. From Table 5-3 it can be seen that
the H2S elimination capacities for all the three treated
composts were increased by a factor of 9, from 7.5 to 68 g-
S/m3-hr.
The NaOH-treated compost looks darker and feels sticky,
probably the structure of the compost was altered by this
aggressive alkali treatment. From an economic point of
view, water is the best choice for the treatment. If
acidity of the compost needs to be corrected, then aqueous
NaHC03 solution is recommended as a treatment chemical.
The shaking-washing method used here is not feasible in
practice for full scale biofilters. The most feasible way
to treat a full scale bed is to spray water or the desired
solution onto the top of the filter. In this case, the
water to compost ratio and contact time become critical for
the effectiveness of the treatment.

152
Table 5-3. Performance of defective compost before and
after treatment.
At low H2
S Loading9
At High
H2S Loading13
Compost
Treatment
h2s
Removal
Eff. (%)
h2s
Elimination
(g-S/m3-hr)
h2s
Removal
Eff. (%)
h2s
Elimination
(g-S/m3-hr)
Untreated
(control)
20.6
7.51
17.0
12.0
Water
Washed
99.9+
36.4
96.6
68.6
NaOH
Washed
99.9
36.4
96.7
68.6
NaHCCU
Washed
99.9
36.4
94.9
67.4
a H2S inlet concentration: 437 ppm;
Gas loading rate: 14.4 m3/m-hr.
b H2S inlet concentration: 489 ppm;
Gas loading rate: 30.0 m3/m2-hr.

153
Effect of Water-Compost Contact Time
on Leaching Efficiency
Water is used as a solvent to determine the effect of
contact time on S042- leaching efficiency. Three (3.0) g of
wet compost #13-1 was weighed into each of nine 50-mL
beakers. Thirty (30) mL of DI water is then added to each
beaker. The beakers with compost and water are allowed to
contact without disturbance. After the desired period of
time (from 5 to 120 minutes), the content of the beakers are
filtered and the filtrate subjected to pH and sulfate
determination. Another container with 3 g compost and 30 mL
of DI water are shaken for 60 minutes as a control.
The results are shown in Figure 5-6. At constant water
to compost weight ratios of 10:1 and without shaking, the
S042- leaching efficiency is between 51% and 68%. The S042-
leaching efficiency increases with water-compost contact
time. The maximum leaching efficiency is achieved in about
one hour.
Effect of Water to Compost Ratio
on S042~ Leaching Efficiency
The effect of water to compost weight ratio on S042-
leaching efficiency was determined by shaking 1, 2 and 3 g
of wet compost in 30 mL of DI water for 30 minutes. The
supernatant was filtered and sulfate was determined in the
filtrates. The results are shown in Figure 5-7. It is
obvious that S042- leaching efficiency is increased
significantly with increasing water to compost weight

Sulfate Leaching Efficiency (%)
154
Figure 5-6
Effect of water-compost contact time on
sulfate leaching efficiency.

Sulfate Leached Out (mg-S/g)
155
Figure 5-7. Effect of water/compost ratio on
sulfate leaching.


156
ratio.
In practice, the water-compost contact time can be
easily controlled by the water spray rate. The selection of
ideal water to compost ratio, however, depends on the pH,
sulfur content of the compost and frequency of the
treatment. High water to compost ratios favor the leaching
of sulfate but more water is required and the maintenance
cost is consequently increased. The wash water is highly
acidic and has high concentrations of sulfate. Proper
treatment of this water is required.
The water to compost weight ratio and contact time used
for this study are 1:1 and 10 minutes, respectively.
Efficiency of elimination of sulfate in filter compost for a
single wash under this condition is shown in Table 5-4. The
average sulfate elimination efficiency is 36.5% with some
variations at each location in the bed. Sulfate
concentrations in the wash water increased from 0 to 5.25
mg-S/g, which indicates that the sulfate sulfur has been
efficiently transferred from compost to wash water.

157
Table 5-4. Effect of water washing on elimination of sulfate
in filter compost.
Samplimg
Location
Sulfate in Compost (mg-S/g)
Sulfate
Elimination
(%)
Before Washing
After Washing
T14
1.67
0.34
79.6
T13
6.71
6.16
8.20
T12
13.4
11.0
17.8
Til
57.1
33.9
40.6
Wash Water
0.00
5.25

CHAPTER 6
FULL SCALE APPLICATION OF BIOFILTRATION TO CONTROL
HYDROGEN SULFIDE EMISSIONS AT A WASTEWATER TREATMENT PLANT
A full scale biofilter bed system has been installed
and operated since the fall of 1988 to control H2S emissions
from the grit chamber at the Kanapaha Wastewater Treatment
Plant, Gainesville, Florida. In this chapter is describes
the design, construction and operation of this system.
Experience gained, as well as advantages and disadvantages
observed during operation of the system are discussed.
Introduction
Emissions of objectionable odors are a common problem
encountered at most wastewater treatment plants. The odorous
compounds frequently observed as volatile emissions from
these sources include hydrogen sulfide, ammonia, organo-
sulfur compounds and some volatile organic compounds (VOCs)
(WPCF, 1979; Yang, 1988). The origin of these odorous
chemicals is in the sewer lines where, due to existing
anaerobic conditions and excessively long residence times
for the incoming wastewater, the odorous compounds are
formed and confined.
The predominant odorous compound emitted from municipal
waste water treatment plants is H2S. Under anaerobic
158

159
conditions, inorganic sulfates and sulfites can be easily
reduced to hydrogen sulfide by various types of anaerobic
and facultative bacteria, such as sulfur-reducing bacteria
(SRB).
S042" + 2C + 2H20 SRB^ 2HC03 + H2S (5-1)
Because hydrogen sulfide is volatile and only partially
soluble in water, whenever domestic wastewater from sewer
lines is agitated and exposed to the atmosphere, H2S is
released to the air.
The Kanapaha wastewater treatment plant is the major
municipal sewage treatment facility ( 9 million gallons per
day) for the city of Gainesville, Florida. As is the case
for most city public utilities, this domestic wastewater
treatment facility has been identified as a source of
odorous emissions and has been the recipient of numerous
odor complaints for many years. The most significant source
of the malodor was identified as the plant's grit chamber,
where agitation of the incoming wastewater causes
considerable outgassing of hydrogen sulfide. Attempts to
control the emission of odors at the plant using chemical
treatment of the wastewater were only partially successful
and very expensive. The high costs of the chemicals used,
as well as additional problems encountered in the treatment
process prompted the management and staff of Gainesville
Regional Utilities (GRU) to look for alternative effective
odor control technologies. As a result, a biofilter system

160
was installed in the summer and made operational in the fall
of 1988.
System Design and Construction
The biofilter control system consists of an air
collecting system, an anti-corrosive blower and ductware, a
humidifier, a central shaft for flow adjustment and a two-
bed SIEBO-stone air distribution system, as shown in Figure
6-1.
The plant's malodor source, the grit chamber, was
covered and sealed as tight as possible (Figure 6-2). A
negative static pressure inside the grit chamber is
maintained by a blower, which prevents possible leakage of
malodorous air. The air collected by the blower through
0.30 meters diameter ducts is forced to a humidifier, where
water is dispersed by several spray nozzles to saturate the
waste gas stream. All pipes in the waste gas collection and
transmission system are made from PVC to prevent corrosion.
The biofilter bed gas distribution system is
constructed from the German patented SIEBO-stones. This
interlocking sinter block base unit of the biofilter system
is rigid and allows for heavy vehicles to be driven on it
without damaging the system. The SIEBO-stone base system
provides for an even distribution of the inlet air to the
filter bed as well as functioning as a drainage system. The
filter area is 100 square meters (m2) and is divided into
two equal sections, each of which can be operated and

161
Waste Gas inlet
SIEBO-stone Gas
Distribution System
Central Shaft
t
Humidifier
Plan View
Section A-A
Figure 6-1. Schematic diagram of the Kanapaha
biofilter bed system.

162
Figure 6-2. Photograph of the grit chamber at
Kanapaha Wastewater Treatment Plant
(top view). The chamber is covered to
collect the malodorous gas.

163
controlled individually. During normal operation, each
section treats one-half of the total exhaust gas flow
volume. The total flow can be diverted to one section if
necessary, for example when repairs are needed, without
reducing overall pollutant removal efficiency.
The actual waste air flow rate through the biofilter is
in the range from 79 to 96 actual cubic meter per minute,
which gives an average surface loading rate of 52 m3/m2-h
and a biofilter bed treated gas retention time of 88
seconds. Biofilter design and operation parameters are
summarized in Table 6-1.
Sampling and Analysis Methods
To monitor biofilter bed performance, inlet and off-gas
samples are collected and analyzed for H2S concentrations.
The influent gas samples are taken from the inlet pipes in
the central shaft and collected in Tediar bags. These
samples are later quantitatively diluted with pure nitrogen
(N2) and analyzed. Off-gas samples from the biofilter are
collected by a specially designed gas sampling system
illustrated in Figure 6-3. The plastic collector cover, 275
mm in diameter and 220 mm in height, is open at the bottom
and has five 10 mm holes drilled in the top to allow the
flowing off-gas to purge the collector and escape. A Teflon
sampling probe is inserted from the top center of the cover
and extends to within 100 mm from the open base. This
design ensures representative sampling and avoids any

164
Table 6-1. Summary of Kanapaha biofilter bed
design and operation parameters.
Total Flow:
Filter Area:
Filter Height:
Gas Loading Rate:
Retention Time:
Temperature:
Pressure Drop:
79 96 acmm
100 m2
1.3 meter
47 57 m3/m2-hr
82 100 sec.
15 30 C
150 200 mmH20

165
Teflon
Tubing
Quick Connecto
Sampling
Probe
Plastic
Chamber
3-Way Valve
Vacuum Gauge
P)
Tediar Bag
Biofllter Bed
Figure 6-3. Biofilter off-gas sampling system.
A
Pump

166
disturbances from spurious ambient air currents near the bed
surface. The probe is connected to a Nutech Model 218
integrated gas sampler by a 7.5 meter length of 6.4 mm
Teflon tubing. Sampling is accomplished by evacuating the
dead space between the inner wall of the canister and the
outer walls of the Tediar bag at a constant rate. The purge
line pump is allowed to run for 5 minutes to flush the
sample line and allow for off-gas equilibrium to be
established inside the sampling cover. The sampling flow
rate is set at about 1 liter per minute (Lpm).
Off-gas samples are collected on the top surface of the
twin biofilter beds at four locations. Each rectangular bed
was divided into two equal-area triangles by drawing
diagonals. The sampling locations were selected at the
centroids of each of the four triangular areas. Analytical
results for these four off-gas samples were averaged later
to obtain a typical off-gas concentration.
The gas samples were transported in the Tediar sampling
bags to the University of Florida (UF) laboratories and
analyzed within eight hours of collection using a Tracor
250H Analyzer.
Results and Discussion
Early in August 1988, the west half of the biofilter
bed was filled up to a grade of 1.3 meters with compost
obtained from Pompano Beach, Florida. The biofilter system
was tested by operating the system at half-filter bed

167
capacity for a few weeks. Because it was not possible to
acquire the same compost for the other half of the biofilter
system, the Kanapaha wastewater treatment plant staff
decided to mix the existing Pompano Beach compost with
compost obtained from the Buckman wastewater treatment plant
in Jacksonville, Florida and in-house compost. This
procedure was necessary to obtain a sufficient quantity of
compost to completely fill the entire biofilter bed system
to a depth of 1.3 meters. The final filter material used is
a mixture of yard waste compost, pine bark and sewage
sludge. Lime was applied to the compost material prior to
installation to buffer the bed acidity (pH). This mixed
compost (Compost #12) was then used to completely fill both
biofilter beds on November 20, 1988. Analytical results for
the composition of the original Pompano Beach compost
(5/10/88) and for the final compost mixture (11/21/88) used
are presented in Table 6-2.
The biofilter system was brought to full operation on
November 21, 1988. The fully operational system is shown
photographically in Figure 6-4.
Influent and off-gas air samples were collected and
analyzed during the first 16 days that the biofilter was
made fully operational (see Table 6-3). During this start
up period, H2S concentrations in the influent gas stream
varied between 156 ppmv and 229 ppmv, and the average off
gas H2S concentrations were observed to be in the range of
0.05 to 0.4 ppmv. There were, however, some minor

Table 6-2. Summary of periodic Kanapaha biofilter bed compost analyses during
operational period from 5/10/88 to 2/5/91.
Date
Bulk
Density
(g/cc)
Water
Content
(%)
Organic
Matter pH
(%)
Total
N
(%)
Total
C
(%)
Total
S
(mg/g)
Water
Ext. P
(Mg/g)
Particle
Size Distri.(Wt%)
>3.35mm
Interm.
<2.36mm
05/10/88
0.23
52.1
68.1 8.63
3.45
39.5
7.3
231
41.0
6.6
52.4
11/21/88
0.27
54.6
69.3 4.40




43.8
12.9
43.3
05/16/90
0.30
45.6
66.5 2.62
1.89
34.3
44.5
140
47.0
6.1
46.8
12/20/90
0.27
54.6
66.5 1.60
1.58
36.6
71.0
294
45.9
6.7
47.4
02/05/91
57.2
64.7 1.80
109

168

169
Figure 6-4. Photograph of the biofilter system at
Kanapaha Wastewater Treatment Plant.

170
Table 6-3. Summary of Kanapaha biofilter influent and
effluent gas sample analyses during three
week start-up period.
Date
Avg. H2S
Influent
(ppmv)
Avg. H2S
Off-gas
(ppmv)
Removal
Efficiency
(%)
11/21/88
195
0.39
99.8
11/22/88
229
0.16
99.9
11/23/88
156
0.32
99.8
11/25/88
208
0.12
99.9
11/29/88
140
0.05
99.9+
12/06/88
175
0.11
99.9+
12/15/88
167
0.07
99.9+

171
differences in concentrations at the four sampling locations
(see Figure 6-5). The similarity in the low H2S
concentrations simultaneously measured at all four off-gas
sampling locations indicate that the incoming gas is evenly
distributed across the filter bed. The final biofilter
control system, which used mixed compost, functioned
effectively immediately upon operation with a very high H2S
removal efficiency (99.8%). No initial period of reduced
efficiency or acclimation was observed. This unique feature
is probably due to the fact that half of the compost had
been previously exposed to H2S laden air for a few weeks
prior to use in the full scale operation. The average
efficiency of the biofilter in removing hydrogen sulfide
during the initial study period was determined to be greater
than 99.8%.
Several additional tests were made in order to confirm
the validity of the H2S removal efficiencies achieved by the
biofilter system. These tests included, determination of
the decomposition of hydrogen sulfide in the Tediar sampling
bags and observations of the effect of varying sample
equilibration time in the gas sampling cover prior to
sampling. Results of these tests are presented in Figure 6-
6.
To a certain degree, the concentrations of H2S for the
stored influent gas samples (high concentration) are
observed to decrease gradually over a few days. However,
there was no significant change in the H2S concentrations

H2S Concentration (ppmv)
172
Days After Start of Biofilter Bed Operation
Figure 6-5. Off-gas sampling locations on the
biofilter beds and concentrations of
hydrogen sulfide observed as a function
of biofilter operating time.

H2S Concentration (ppmv)
173
Days After Sampling
Figure 6-6. Concentration changes for hydrogen
sulfide in gas samples contained in
Tediar bags as a function of container
holding time.

174
for off-gas samples (low concentrations) stored over a
similar observation period. Since the gas samples are
analyzed within 8 hours after sampling, the concentration
changes in the Tediar bags for both influent and off-gas
samples are considered to be negligible, when estimating
control efficiencies.
The effect of sample purging time for the collection
chamber on concentration measurements of effluent gas from
the biofilter was determined as follows: Off-gas samples
are collected at 0, 5, 15 and 20 minutes after placing the
sampling cover on the compost bed and purging. Each sample
is collected for 5 minutes at a flow rate of 1 Lpm. The
results are illustrated in Figure 6-7. These results
indicate that there is no significant difference in H2S
concentrations obtained for the different purging times
selected. The gas loading rate through the compost bed is
about 780 L/m2-min, and the sampling cover has a volume of
12 liters with a cross-sectional area of 0.06 m2. The dead
volume of the cover, therefore, can be replaced by the off
gas every 0.3 minutes. Because the collections of off-gas
samples from the biofilter are usually initiated 5 minutes
after the sampling chamber is placed at the sampling
locations, it is to be expected and has been demonstrated
that representative off-gas samples are collected from the
biofilter bed under these conditions.
The waste gas humidifier works effectively to keep the
compost bed within a moisture content range of 45-60% (Table

H2S Concentration (ppmv)
175
Figure 6-7. Effect of varying purging time for
sample collection chamber prior to
sampling on measured hydrogen sulfide
concentrations.

176
6-2). Influent air measurements indicated that the relative
humidity of the waste gas provided to the biofilter bed
after humidification is at least 95%. No other water was
introduced to the filter bed except that due to rain.
No significant changes in the bulk density and particle
size distribution of the compost were observed during the 27
months of operation (from 11/88 to 2/91). The organic
matter content decreased by about 4.6%, presumably due to
mineralization of the compost. Total nitrogen and total
carbon contents of the compost decreased at different rates
resulting in an increase of the C/N ratio from 11.4 to 23.2.
The most significant changes observed in the compost
during extended operation were the total sulfur content and
pH. The total sulfur, expressed as mg-S/g-compost on a dry
basis, increased considerably from 7.3 to 109. After
prolonged operation, parts of the biofilter bed showed a
pronounced color change in the compost from dark brown to
yellowish white, which was accompanied by observation of a
rotten vegetable odor.
The pH of the compost bed material decreased
significantly from 8.6 to 1.8 as a result of the continuous
removal of H2S and corresponding formation of sulfuric acid
(H2S04) by the biological oxidation of H2S. Although this
acidification of the compost did not appear to have a direct
and noticeable effect on the overall H2S removal efficiency,
serious corrosion of the cement blocks in the retaining wall
and the SIEBO-stone base was observed. Corrosion of the

177
latter resulted in blocking of the narrow waste gas
distribution vents between the SIEBO-stones, which caused
channeling and uneven gas distribution throughout the
biofilter system.
Corrosion of the SIEBO-stones and the resultant
blockage of the gas distribution system were first noticed
by Kanapaha Plant workers in late 1990. To solve this
problem, the west bed of the biofilter system was unpacked
and the gas distribution vents between the SIEBO-stones were
cleaned manually. The compost used in this bed was
thoroughly mixed by turning and repacked in the bed after
the gas distribution vents had been cleaned.
On February 5, 1991, both gas and compost samples were
taken at the biofilter beds. Results of this recent gas
sampling and analysis exercise are presented in Table 6-4.
Sampling sites 1, 2 and 3 are located on the west bed; 4 and
5 are located on the east bed. Almost no air was vented
through the east bed, even though the gas flow control
valves to both beds were fully open, because of the
corrosion and blockage of the east bed gas vents. During
cleaning of the east bed air distribution system, the entire
waste gas stream was forced through the west bed. The waste
gas stream, however, was no longer evenly distributed
through this 'cured' system. Channeling caused the waste gas
to blow through the filter without control and resulted in
very high off gas H2S concentrations (sites 1 and 3, Table
6-4). Some removal of the H2S was measured at Site 2,

178
Table 6-4. Gas sampling
biofilter bed
and analysis
, 2/5/91.
for Kanapaha
Sampling
h2s
h2s
Location
Concen.
Removal
(ppmv)
(%)
Inlet
136

1
133
2.57
2
81.3
40.3
3
134
1.40
4
0.04
99.9+
5
0.04
99.9+

179
however, the removal efficiency was much lower than
previously recorded values obtained during normal
operations. The large variation in H2S concentrations in
the off gas at different locations indicates uneven
distribution of the waste gas stream throughout the system.
Since not much air is vented through the east bed, the H2S
concentrations measured in the off gas for this bed (sites 4
and 5) are very low and therefore, the calculated removal
efficiencies for this bed (Table 6-4) are questionable. A
strong characteristic rotten egg odor (H2S) was smelled
around both filter beds during sampling.
The color of the compost packed on the west bed had
changed to yellowish-white in comparison to the color of the
I
compost packed on the east bed, which had still retained
its original dark brow color (Figure 6-8). Analysis of the
compost samples from these discolored sites is shown in
Table 6-5. Approximately 20% of the total S is organic-S
and 80% is inorganic-S. Very high fractions of FeS2-S
(27.6%) and insoluble sulfate (19.98%) were determined in
these samples.
As a result of this long term operational corrosion
problem, GRU engineers have decided to replace the existing
SIEBO-stones with similar blocks made from anti-corrosive
materials.

180
Figure 6-8. Compost samples taken from Kanapaha
Wastewater Treatment Plant biofilter
beds (2/5/91) .
Left: Sample taken from west bed. White
color indicates high sulfur
accumulation.
Right: Sample from east bed. Low
sulfur content compost, color is close
to the original (dark brown).

181
Table 6-5. Sulfur fractionation of a typical compost
sample in Kanapaha biofilter bed.
mg-S/g
(Wt%)
Total-S
110
100
Organic-S
22.3
20.3
C-bonded-S
21.1
19.3
Ester-S
1.20
1.10
Inorganic-S
87.2
79.7
FeS2~S
30.0
27.6
FeS-S
0.40
0.40
S-S
3.9
3.6
so42"-s
52.7
48.1
Water Soluble-S
20.9
19.1
P-extractable-S
9.9
9.0
Insoluble-S
21.9
20

182
Conclusions
In spite of the few limitations described, for a full
scale system, biofiltration has been demonstrated to be a
simple, effective and inexpensive method for odor control
at wastewater treatment plants based on the Kanapaha plant
experience. The biofilter system described was successfully
operated at high H2S removal efficiencies with little or no
maintenance for a period of 2.5 years. The system was
effective until the corrosion problem occurred and the gas
distribution system had to be reconstructed. During the
2.5-year operation period, no odor was noticeable even close
to the filter beds. Also, the City of Gainesville did not
receive a single odor complaint during this period. More
than $200,000 per year has been saved in chemicals that were
originally used to provide alternative odor control systems
for the plant. These cost savings have resulted in a one-
year payback on the biofilter system capital and operating
costs (IPS, 1990).
The Kanapaha biofilter experience suggests that
routine monitoring and maintenance are necessary to ensure
proper operation conditions for effective, long term control
of H2S emissions. It was recognized that further research
was needed to solve existing problems, such as progressive
system acidification, accumulation of sulfur in the filter
medium and the eventual decline of H2S removal efficiency.
Appropriate laboratory studies have been conducted at the

183
University of Florida to address these problems and the
results are discussed in detail in Chapter 4 of this
dissertation.

CHAPTER 7
SUMMARY AND CONCLUSIONS
The intention of this research was to develop a
quantitative knowledge of the operation of a microbial
biofilter system for removal of hydrogen sulfide from waste
gas streams and to optimize and maintain the performance of
such a biofilter system using such knowledge. Optimization
of the system involved quantitatively determining the design
parameters, the operating parameters and predictive
relationships for the control efficiency. Maintaining the
system involves recognizing system deterioration and upset
and providing solutions for prevention of long term
irreversible deterioration.
A lab scale biofilter tower system was constructed and
extensive experimental work was conducted to achieve the
stated goals. In addition, a full-scale compost biofilter
bed system for control of H2S emissions in a wastewater
treatment plant was evaluated during long term operation.
Based on the results of this study, it is concluded that:
1. Significant pressure drops in biofilter materials are
mainly caused by the presence of small particles,
particularly those with diameters less than 1 mm.
These small particles are generally composed of sand
and minerals, as well as decomposed and mineralized
184

185
compost material. Small particles have a low water
holding capacity and are less valuable to the overall
biofiltration process. It is recommended that small
particles be separated from the compost by sieving
before use. To minimize pressure drop effects the
filter should not be compacted unduly. Aged compost,
which contains a larger fraction of fine particles,
due to mineralization and fracture, should not be
reused after change-out, unless it is specifically
treated to remove the fine particle fraction by
sieving or washing.
2. The time reguired for the oxidation of H2S to sulfate
by microorganisms is a few seconds, however, the use
of high gas velocities is not recommended since they
will cause uneven gas distributions and high pressure
drops.
3. The concept of maximum H2S elimination capacity of
compost and H2S loading rate is very important in
terms of system design and operation. When working
within the maximum elimination capacity of the system,
the waste gas flow rate can be adjusted to obtain the
best reduction of H2S for various inlet
concentrations.
4. Low compost water content is fatal to the biological
process. A minimum value of 30% water content by
weight is reguired for proper operation, but 40 to 60%
is recommended.

186
5. The microorganisms in the biofilter system are
mesophiles. The optimum temperature range for these
organisms to be efficient is from 30 to 40 C. A H2S
removal efficiency of 50% can be achieved when the
temperature range is extended from 10 to 80 C.
6. High concentrations of sulfate are toxic to the
microbial flora. The critical level is between 30 and
40 mg-S/g dry compost. Above this range, the
biological activity of the microorganisms may be
significantly inhibited and result in reduction of the
H2S elimination capacity of the compost.
7. Variation in Compost pH showed no effect on H2S
removal efficiency for values greater than 2.0 pH
units. However, lower pH values for the filter bed
will cause serious corrosion problems.
8. Biofiltration is a comprehensive process which
involves physical, chemical and biological processes.
Ottengraf's model was adopted and has proved to be
successful in describing the macro kinetics of H2S
removal. Kinetic models and equations have been
developed and determined to be appropriate and
accurate in quantitatively describing removal of H2S
by the biofiltration process.
9. Composts from different sources have been
demonstrated to be excellent media for biofilters used
in H2S removal from waste gas streams. Their unique

187
properties provide good buffering capacity to various
operational impacts. H2S removal efficiencies of
99.9+% were routinely achieved in both the laboratory
and full scale operations.
10. The decomposition rates of composts are described by
multi-stage first-order kinetics. The decomposition
rate as well as biological activity of the compost are
significantly enhanced by the presence of H2S. First-
order reaction coefficients were determined which can
be used to quantitatively predict the useful life of
the compost.
11. Acidification and accumulation of sulfur, especially
sulfate, is a natural feature of the H2S oxidation
process. Continuous formation of H2S04 results in
significant decline in pH and serious corrosion of the
construction materials. A protective washing
procedure was developed to mitigate this feature and
keep the system operational at its optimized
conditions. Water is determined to be the best choice
for elimination of sulfate. NaHC03 solution is
recommended for pH corrections.
12. System upset is identified by compost dry-out, high
sulfur content in compost accompanied by a yellowish-
white deposit, and extremely low pH (<2) values of
compost. Specific procedures have been developed to
recover the activity of the defective filter material.

188
Although hydrogen sulfide has been selected as a test
gas for this research, the results obtained through this
study should provide for an improved quantitative
understanding of the principles of biofiltration.
Controlling variables in the operation and effectiveness of
biofiltration waste gas control technology may be applicable
not only to hydrogen sulfide but also to other sulfur-
containing compounds, and more generally, to air toxics and
VOCs. Dissemination of this basic information on the
multiple advantages of biofiltration, including low cost,
high destruction-efficiency, energy conservation, ease of
operation and maintenance and universal application will
provide environmetal engineers and federal, state and local
government air pollution control officials with a viable
alternative in controlling emissions of air toxic compounds
from commercial and industrial sources. Biofiltration
control of waste gas streams is a relatively unknown and
little explored control technology in the U.S.. It has the
potential, however, for wide-spread application and
acceptance because of its relative simplicity and low
capital and operating cost, in addition to its great
potential for indiscriminate effectiveness in controlling
multiple pollutants.

REFERENCES
Aaronson, S. Experimental Microbiology. Academic Press, New
York, 1970.
Allen, E.R.; Hartenstein, H.U. and Yang, Y. "Review and
Assessment of the Design and Operation of a Compost
Biofilter System for Odor Control," Final Project Report,
Environmental Engineering Sciences Department, University of
Florida, Gainesville, Florida, 1987a.
Allen, E.R.; Hartenstein, H.U. and Yang, Y. "Identification
and Control of Industrial Odorous Emissions at a Municipal
Wastewater Treatment Facility," Paper # 87-95A.4, Presented
at 80th Annual Meeting of APCA, New York, June 21-26, 1987b.
Allen, E.R.; Yang, Y.; Hartenstein, H.U. "Odor Producing
Agents, Their Sources and Control", Final Project Report,
Environmental Engineering Sciences Department, University of
Florida, Gainesville, Florida, 1987c.
Allen, E.R.; Yang, Y. and Hartenstein, H.U. "To Provide
Technical Assistance in the Design and Installation of An
Odor Control Biofilter System at the Kanapaha Wastewater
Treatment Plant," Final Project Report, Environmental
Engineering Sciences Department, University of Florida,
Gainesville, Florida, 1989.
Allen, E.R. and Yang, Y. "Biofiltration Control of Hydrogen
Sulfide Emissions," Paper # 91-103.10, presented at the 84th
Annual Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
American Public Health Association (APHA), Standard Methods
for the Examination of Water and Wastewater. 17th ed.
American Public Health Association, Washington, DC, 1989.
Atlas, J.E. and Bartha, R. Microbial Ecology: Fundamentals
and Applications. Addison-Wesly Publ. Company, Reading, MA,
1981.
Bethea, R.M.; Murthy, B.N. and Carey D.R. "Odor Controls
for Rendering Plants," Environ. Sci. Tech. 7s 504 (1973).
189

190
Biddlestone, A.J.; Gray, K.R. and Day, C.A. "Composting and
Straw Decomposition," in Environmental Biotechnology,
Forster, C.F. and Wase D.A.J. Eds., John Wiley & Sons, New
York, 1987.
Blake, G.R. and Hartge, K.H. "Particle Density," in Methods
of Soil Analysis. Part I. Physical and Mineralogical
Methods. Klute, A. Ed. 2nd ed, American Society of Agronomy,
Madison, Wisconsin, 1986a.
Blake, G.R. and Hartge, K.H. "Bulk Density," in Methods of
Soil Analysis. Part I. Physical and Mineraloqical Methods.
Klute, A. Ed. 2nd ed, American Society of Agronomy, Madison,
Wisconsin, 1986b.
Bohn, H.L. and Miyamoto, S. "Soil as a Sorbent and Filter of
Waste Gases," in Symposium on land for Waste Management.
Tomlinson, J. ed. Natl. Research Council Canada, Ottawa,
1973.
Bohn, H.L."Soil and Compost Filters of Gases," J. Air
Pollution Control Assoc., 25: 953 (1975).
Bohn, H.L. "Compost Scrubbers of Malodorous Air Streams,"
Compost Sci., 17: 5 (1976).
Bohn, H.L. "Soil Treatment of Organic Waste Gases, Chapter
24, in Soils for Management of Organic Waste and Waste
Waters. ASA-CSSA-SSSA, Madison, WI, 1977.
Bohn, H.L.; Bohn, R.K. "Soil Bed Scrubbing of Fugitive Gas
Releases," J. Environ. Sci. Health, A21: 561 (1986).
Bohn, H.L. and Bohn, R.K. "Biofiltration of Odors from Food
and Waste Processing," Proceedings of Food Processing Waste
Conference, Georgia Tech Research Institute, Sept. 1-2,
1987.
Bohn, H.L. and Bohn, R.K. "Soil Beds Weed Out Air
Pollutants," Chemical Engineering, pp. 73-76, April 1988.
Bohn, H.L. "VOC Removal by Soil Biofilter Beds," Proceedings
of Hazmacon'89, Hazardous Materials Management Conference
and Exhibition, Vol.2, Association of Bay Area Govermments,
April 18-20, 1989.
Brock, T.D. and Madigan, M.T. Biology of Microorganisms.
Fifth Edition, Prentice Hall Inc., Englewood Cliffs, New
Jersey, 1988.
Carlson, D.A. and Leiser, C.P. "Soil Beds for the Control of
Sewage Odors," Journal of the Water Pollution Control
Federation, 38: 829 (1966).

191
Caunt, P. and Hester, K.W. "A Kinetic Model for Volatile
Fatty Acid Biodegradation during Aerobic Treatment of
Piggery Wastes," Biotechnology and Bioengineering, 34: 126
(1989).
Chen, K.Y. and Morris, J.C. "Kinetics of Oxidation of
Aqueous Sulfide by 02," Environmental Science & Technology,
6: 529 (1972) .
Cooper Jr, H.B.H. "Kinetics of Inorganic Sulfur Oxidation
during Black Liquor Oxidation with Oxygen," Tappi 57: 130
(1974) .
Danielson R.E. and Sutherland, P.L. "Porosity," in Methods
of Soil Analysis. Part I. Physical and Mineralooical
Methods. Klute, A. Ed. 2nd ed, American Society of Agronomy,
Madison, Wisconsin, 1986.
David, M.B.; Mitchell, M.J. and Nakas, J.P. "Organic and
Inorganic Sulfur Constituents of a Forest Soil and Their
Relationship to Microbial Activity," Soil Sci. Soc. Am. J.
46: 847 (1982).
Delaune, R.D.; Reddy, C.N. and Patrick Jr., W.H. "Organic
Matter Decomposition in Soil as Influenced by pH and Redox
Conditions," Soil Biol. Biochem. 13: 533 (1981).
Dharmavaram, S. "Biofiltration A Lean Emission Abatement
Technology," Paper # 91-103.2, presented at the 84th Annual
Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
Don, J.A. "The Rapid Development of Biofiltration for the
Purification of diversified Waste Gas Streams," in VDI
Berichte 561: VDI Verlag, Dsseldorf, 1985.
Eitner, D. and Gethke, H.G. "Design, Construction and
Operation of Bio-filters for Odor Control in Sewage
Treatment Plants," Paper # 87-95A.6, Presented at the 80th
Annual Meeting of Air Pollution Control Association, New
York, New York, June 21-26, 1987.
Eitner, D. "Biofilter in Flue Gas Cleaning: Biomasses,
Design, Costs, and Applications," Brennst.-Waerme-Kraft
(German). 41/3, 124., 1989.
Ergas, S.J.; Schroeder, E.D.; Chang, D.P. "VOC Emission
Control from Wastewater Treatment Facilities Using
Biofiltration," Paper # 91-105.4, presented at the 84th
Annual Meeting of the Air & Waste Management Association,
Vancouver, BC Canada, June 16-21, 1991.
Ferguson, P.A. Hydrogen Sulfide Removal from Gases. Air and
Liquids. Noyes Data Corporation, Park Ridge, NJ., 1975.

192
Forster, C.F. and Wase, D.A.J. "Biopossibilities: The Next
Few Years," in Environmental Biology. Forster, C.F. and
Wase, D.A.J. eds, Ellis Horwood Limited Publishers,
Chichester, England, pp.439-448., 1987.
Frechen, F.B.; Kettern, J.Y. "Reduction of Odorous Emissions
from a Hazardous Waste Landfill Site Using Biofiltration and
Other Techniques," Paper # 87-95A.3, presented at the 80th
Annual Meeting of APCA, New York, New York, June, 1987.
Freney, J.R. "Determination of Water-Soluble Sulfate in
Soils," Soil Science 86s 241 (1958).
Gale, P.M. "Decomposition of Organic Waste Products Under
Aerobic and Anaerobic Conditions," PhD Dissertation,
University of Arkansas, 1988.
Grant, W.D. and Long, P.E. Environmental Microbiology.
Halsted Press, New York, 1981.
Gilmour, J.T. ; Clark, M.D. and Sigua G.C. "Estimating Net
Nitrogen Mineralization from Carbon Dioxide Evolution," Soil
Sci. Soc. Am. J. 49: 1398 (1985).
Greyson, J. Carbon. Nitrogen, and Sulfur Pollutants and
their Determination in Air and Water. Marcel Dekker, Inc.,
New York and Basel, 1990.
Hack, P.J.F.M. and Habets, L.H.A. "Experience with Full-
Scale Biopaq U.A.S.B.-Plants Treating Various Types of
Effluent," in Environmental Technology. Proceedings of the
second European Conference on Environmental Technology.
Amsterdam. The Netherlands. June 22-26, 1987. De Waal,
K.J.A. and Van Den Brink, W.J., Eds, Martinus Nijhoff
Publishers, Boston, 1987.
Hartenstein, H.U. "Assessment and Redesign of an Existing
Biofiltration System," Master's Thesis, University of
Florida, Gainesville, Florida, 1987.
Hartenstein, H.U. and Allen, E.R. "Biofiltration, An Odor
Control Technology for a Wastewater Treatment Plant," Report
to Department of Public Works, City of Jacksonville,
Florida, 1986.
Hsieh, Y.P. and Yang, C.H. "Diffusion Methods for the
Determination of Reduced Inorganic Sulfur Species in
Sediments," Limnol. Oceanogr., 34: 1126 (1989).
International Process Systems, Inc. (IPS) "Odor Control,
Completing the Composting Process," 655 Winding Brook Drive,
Glastonbury, CT 06033, 1990.

193
Jennings, P.A.; Snoeyink, V.L. and Chian, E.S.K.
"Theoretical Model for a Submerged Biological Filter,"
Biotechnology and Bioengineering, 18: 1249 (1976).
Johnson, D.W. and Henderson, G.S. "Sulfate Adsorption and
Sulfur Fractions in a Highly Weathered Soil Under a Mixed
Deciduous Forest," Soil Science, 128: 34 (1979).
Johnson, C.M. and Nishita, H. "Microestimation of Sulfur in
Plant materials, Soils, and Irrigation Waters," Anal. Chem.
24: 736 (1952).
Kampbell, D.H.; Wilson, J.T.; Read, H.W.; Thomas, T.
Stocksdale, T.T. "Removal of Volatile Aliphatic Hydrocarbons
in a Soil Bioreactor," Journal of Air Pollution Control
Association 37: 1236 (1987).
Kuenen, J.G. Colorless Sulfur Bacteria and Their Role in
the Sulfur Cycle," Plant and Soil 43: 49 (1975).
Lalazary, S.; Pirbazari, M. and McGuire, M.J. "Oxidation of
Five Earthy-Musty Taste and Odor Compounds," J. Amer. Water
Works Assoc. 78: 62 (1986).
Leson, G. and Winer, A.M. "Biofiltration: An Innovative Air
Pollution Control Technology for VOC Emissions," J. AW&MA,
41: 1045 (1991).
Lindstrom, K.P. Air Toxic Emissions and POTWs," Workshop
Report and Proceedings, Co-sponsored by WPCF and USEPA,
Alexandria, VA, July 10-11, 1989.
Miller, R.H. "Factors Affecting the Decomposition of an
Aerobically Digested Sewage Sludge in Soil," J. Environ.
Qual. 3: 376 (1974).
Miller, R.D. and Johnson, D.D. "The Effect of Soil Moisture
Tension on Carbon Dioxide Evolution, Nitrification, and
Nitrogen Mineralization," Soil Sci. Soc. Pro. 644-647 ,
1964.
McGilvery, R.W. and Goldstein, G.W. Biochemistry, A
Functional Approach. Third ed.; W.B. Saunders Company,
Philadelphia, PA, 1983.
Moorhead, K.K.; Graets, D.A. and Reddy, K.R. "Decomposition
of Fresh and Anaerobically Digested Plant Biomass in Soil,"
J. Environ. Qual. 16: 25 (1987).
National Research Council (NRC), "Hydrogen Sulfide,"
Subcommittee on Hydrogen Sulfide, Committee on Medical and
Biologic Effects of Environmental Pollutants. University
Parck Press, Baltimore, Maryland, 1979.

194
(NIOSH) "Criteria for a Recommended Standard for the
Occupational Exposure to Hydrogen Sulfide," U.S. Department
of Health, Education and Welfare, NIOSH, May 1979.
O'Brien, D.J. and Birkner F.B. "Kinetics of Oxygenation of
Reduced Sulfur Species in Aqueous Solution," Environmental
Science & Technology, 11: 1114 (1987).
Ottengraf, S.P.P. Theoretical Model for a Submerged
Biological Filter," Biotechnology and Bioengineering, 19:
1411 (1977).
Ottengraf S.P.P. and Van Den Oever, H.A.C. "Kinetics of
Organic Compound Removal from Waste Gases with a Biological
Filter," Biotechnology and Bioengineering, 25: 3089 (1983).
Ottengraf, S.P.P.; Van Den Oever, A.H.C. and Kempenaars,
F.J.C.M. "Waste Gas Purification in a Biofilter Bed," in
Innovations in Biotechnology. Houwink, E.H. and Van Dan
Meer, R.R. Eds, Elsevier Science Publishers B.V., Amsterdam
(1984) .
Ottengraf, S.P.P.; Meesters, J.J.P.; Van Den Oever, A.H.C.
and Rozema, H.R. "Biological Elimination of Volatile
Xenobiotic Compounds in Biofilters," Bioprocess Engineering
l: 61 (1986).
Ottengraf, S.P.P. "Exhaust Gas Purification," in
Biotechnology. Rehm, H.J. and Reed, G. Eds, Vol.8; VCH
Verlagsgesellschaft., Weinheim, 1986.
Ottengraf, S.P.P. "Biological Systems for Waste Gas
Elimination," TIBTECH 5: 132 (1987).
Painter, D.E. Air Pollution Technology. Reston Publishing
Company, Inc., Reston, VA, 1974.
Parker, H.W. Air Pollution. Prentice-Hall, Inc., NJ, 1977.
Patterson, Jr., A. and Thomas, H.C. A Textbook of
Quantitative Analysis. Henry Holt and Company, New York,
1952.
Paul, P.G; Castelijn, F.J. "BiofiltrationA Relatively
Cheap and Effective Method of Waste Gas Treatment," in
Environmental Technology. Proceedings of the second European
Conference on Environmental Technology, Amsterdam. The
Netherlands. June 22-26. 1987. De Waal, K.J.A. and Van Den
Brink, W.J., Eds, Martinus Nijhoff Publishers, Boston, 1987.
Piscarcyzyk, K. "Odor Control with Potassium Permanganate,"
Presented at Ohio Water Pollution Control Conference,
Dayton, OH, June 16-18, 1982.

195
Pomeroy, R.D."Biological Treatment of Odorous Air", J. of
the Water Pollution Control Federation, 54(12): 1541 (1982).
Prokop, W.H. and Bohn H.L. "Soil Bed System for Control of
Rendering Plant Odors", Journal of Air Pollution Control
Association, 35: 1332 (1985).
Rands, M.B.; Cooper, D.E.; Woo, C.P.; Flether, G.C. and
Rolfe K.A. "Compost Filters for H^S Removal from Anaerobic
Digestion and Rendering Exhausts," Journal of Water
Pollution Control Federation, 53: 185 (1981).
Reddy, K.R.; Khaleel, R. and Overcash M.R. "Carbon
Transformations in the Land Areas Receiving Organic Wastes
in Relation to Nonpoint Source Pollution: A Conceptual
Model," J. Environ. Qual. 9: 434 (1980).
Rittmann, B.E. and McCarty, P.L. "Variable-Order Model of
Bacterial-Film Kinetics," Journal of the Environmental
Engineering Division, Proceedings of the American Society of
Civil Engineers, 104: EE5, 889 (1978).
Robarge, W.P. and Fernandez I. "Quality Assurance Methods
Manual for Laboratory Analytical Techniques," Prepared for
the USEPA and USDA Forest Service Forest Response Program.
Department of Soil Science, North Carolina State University.
Raleigh NC 27695, and Department of Plant and Soil Science,
University of Maine-Orono, Orono, ME 04469, July, 1986.
Rotman, A. "Use of Biofilter in Odor Control," Unpublished
Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 Rene-
Levesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991
a.
Rotman, A. "Biofilter: Traditional Design, Summary of
Technical Data as Based on German Projects," Unpublished
Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 Rene-
Levesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991
b.
Roy, A.B. and Trudinger, P.A. The Biochemistry of Inorganic
Compounds of Sulfur. University Press, Cambridge, 1970.
Sawyer, C.N. and McCarty, P.L. Chemistry for Sanitary
Engineers. McGraw-Hill, New York, NY, 1967.
Schmidt, S.K.; Simkins, S; Alexander, M. "Models for the
Kinetics of Biodegradation of Organic Compounds not
Supporting Growth," Appl. Environ. Microbiol.50: 323 (1985).
Sikora, L. and Sowers M.A. "Effect of Temperature Control on
the Composting Process," J. Environ. Qual. 14: 434 (1985).

196
Smittenberg, J.; Harmsen, G.W.; Quispel, A. and Otzen, D.
"Rapid Methods for Determining Different Types of Sulfur
Compounds in Soil," Plant and Soil 3: 353 (1951).
Sommers, L.E.; Nelson, D.W. Terry, R.E., and Silvieria, D.J.
"Nitrogen and Metal Contamination of Natural Waters from
Sewage Disposal on Land," Tech. Rep. No 89, Purdue Univ.
Water Resour. Res. Center, W. Lafayette, IN, 1976.
Starkey, R.L. "Oxidation and Reduction of Sulfur Compounds
in Soils," Soil Sci. 101: 297 (1966).
Stotzky, G."Microbial Respiration," in Methods of Soil
Analysis, Part 2? Black C.A. Ed, Agronomy 9: 1550 (1965).
Sublette, K.L. and Sylvester, N.D. "Oxidation of Hydrogen
Sulfide by Thiobacillus Denitrificans: Desulfurization of
Natural Gas," Biotechnology and Bioengineering, 29: 249
(1987) .
Sweeney, D.W. and Graetz G.A. "Chemical and Decomposition
Characteristics of Anaerobic Digester Effluents Applied to
soil," J. Environ. Qual. 17: 309 (1988).
Swift, E.H. Introductory Quantitative Analysis, Principles
and Selected Procedures. New York Prentice-Hall, Inc., 1950.
Tabatabai, M.A. "Sulfur," in Methods of Soil Analysis. Part
2. Chemical and Microbiological Properties. Page, A.L. Ed.,
2nd ed, American Society of Agronomy, Madison, Wisconsin,
1982.
Taylor, J.M.; Sikora, L.J.; Tester, C.F. and Parr., J.F.
"Decomposition of Sewage Sludge Compost in Soil: II.
Phosphorus and Sulfur Transformations," J. Environ. Quanl.
7: 119 (1978).
Terasawa, M.; Hira, M. and Kubota, H. "Soil Deodorization
Systems," BioCycle 27: 28 (1986).
Terry, R.E.; Nelson, D.W. and Sommers L.E. "Carbon Cycling
During Sewage Sludge Decomposition in Soils," Soil Sci. Soc.
Am. J. 43: 494 (1979a).
Terry, R.E.; Nelson, D.W. and Sommers, L.E. "Decomposition
of Anaerobically Digested Sewage Sludge as Affected by Soil
Environmental Conditions," J. Environ. Qual. 8: 342 (1979b).
Tester, C.F.; Sikora, L.J.; Taylor, J.M. and Parr, J.F.
"Decomposition of Sewage Sludge in Soil: I. Carbon and
Nitrogen Transformations," J. Environ. Qual. 6: 459 (1977).

197
Tester, C.F.; Sikora, L.J; Taylor, J.M. and Parr, J.F.
"Decomposition of Sewage Sludge in Soil: III. Carbon,
Nitrogen, and Phosphorus Transformations in Different Sized
Fractions," J.Environ. Qual. 8: 79 (1979).
Urone, P."The pollutants," in Air Pollution. Stern, A.C. ed.
Vol.6, Academic Press, New York, 1986.
USEPA, "Odor and Corrosion Control in Sanitary Sewage
Systems and Treatment Plants," EPA-Design Manual, EPA/625/1-
85/018, 1985.
Van Lith, C. "Design Criteria for Biofilters," Paper # 89-
165.5, Presented at 82nd Annual Meeting & Exhibition, AW&MA,
Anaheim, CA June 25-30, 1989.
Varanka, M.W.; Zablocki, Z.M. and Hinesly, T.D., "The Effect
of Digestion Sludge on Soil Biological Activity," J. Water
Poll. Cont. Fed. 48: 1728 (1976).
Walker, G.S.; Lee, F.P. and Aifa, E.M. "Chlorine Dioxide for
Taste and Odor Control," J. Amer. Water Works Assoc. 78: 84
(1986).
Water Pollution Control Federation (WPCF), "Odor Control for
Wastewater Facilities," Manual of Practice No. 22, Water
Pollution Control Federation, Washington, DC., 1979.
White, A., P. Handler, E.I. Smith, R.L. Hill, I.R. Lehman,
Principles of Biochemistry. Sixth Edition, McGraw-Hill Book
Company, New York, 1978.
Wieder, R.K. and Lang, G.E. "An Evaluation of Wet Chemical
Methods for Quantifying Sulfur Fractions in Freshwater
Wetland Peat," Limnol. Oceanogr., 30: 1109 (1985).
Williamson, K.J. "The Kinetics of Substrate Utilization by
Bacterial Films," PhD Dissertation, Stanford University,
June 1973.
Williamson, K. and McCarty, P.L. "A Model of Substrate
Utilization by Bacterial Films," J. WPCF 48: 9 (1976 a).
Williamson, K. and McCarty, P.L. "Verification Studies of
the Biofilm Model for Bacterial Substrate Utilization," J.
WPCF 48: 281 (1976 b).
Yang, Y. "Odor Emissions and Its Control in a Wastewater
Treatment Plant with Industrial Sources," Master's Thesis,
University of Florida, Gainesville, FL, December 1988.

198
Yang, Y. and Allen, E.R. "Biofiltration Control of Odor
Emissions in Wastewater Treatment Plants," Paper presented
at the 201st National Meeting of the American Chemical
Society, Atlanta, GA, April 14-19, 1991.
Zhabina, N.N. and Volkov, I.I. "A Method for Determination
of Various Sulfur Compounds in Sea Sediments and Rocks," in
Environmental Biogeochemistrv and Geomicrobioloav. Krumbein
Ed., Ann Arbor Sci., Ann Arbor, MI, 1978.

BIOGRAPHICAL SKETCH
Yonghua Yang was born on March 24, 1949, in Inner
Mongolia, China, and attended local schools until completion
of high school in 1968 He received his Bachelor of
Engineering (equivalent) degree in chemical engineering from
Dalian Institute of Technology, China, in 1977. After
graduation, he worked for 8 years for the Environmental
Protection Institute, Baotou Iron and Steel Corporation in
China as an environmental engineer.
He was accepted as a graduate student in fall 1986 and
received his Master of Engineering degree in air pollution
from the University of Florida, Gainesville, FL, in 1988. He
continued graduate study to pursue the Doctor of Philosophy
degree in the Environmental Engineering Sciences Department,
University of Florida from 1989 through 1991.
He was the recipient of the Axel Hendrickson
Scholarship award from the Air & Waste Management
Association (AW&MA), Florida Section in 1990 and a graduate
scholarship award from AW&MA, in 1991.
199

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Eric R. Allen, Chairman
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Dale A. Lundgren
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Ben Koopman ^
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Q=>:
Konda R. Reddy
Professor of S<

I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Professor of Agricultural
Engineering
This dissertation was submitted to the Graduate Faculty
of the College of Engineering and to the Graduate School and
was accepted as partial fulfillment of the requirements for
the degree of Doctor of Philosophy.
May 1992
iLuj &
Winfred M. Phillips
Dean, College of Engineering
Madelyn M. Lockhart
Dean, Graduate School



21
1/v = (i/vmax) + ( yvnax)(i/s) (2-5)
wh e r e :
V = the biodegradation rate, mg hydrocarbon/kg soil-h
Vmax = maximum possible biodegradation rate,
mg hydrocarbon/kg soil-h
Kjjj = an empirical constant, half saturation value, ppm
S = the concentration of organic compound in air, ppm
The values of Km and Vmax can be obtained from a
Lineweaver-Burk plot of 1/V against 1/S.
Biological Oxidation of Hydrogen Sulfide
Hydrogen sulfide may be utilized by microorganisms in
three different ways: assimilation, mineralization and
sulfur oxidation (Atlas and Bartha, 1981; Grant and Long,
1981). However, the rates of uptake of hydrogen sulfide
based on the assimilation processes are far too low to
achieve reasonably high removal efficiencies from a highly
loaded waste gas stream. The most important and efficient
way for microorganisms to utilize hydrogen sulfide is by the
oxidation of sulfur to gain energy. In this process,
relatively large quantities of sulfur are oxidized in order
for the microbes to receive sufficient energy. The
microorganisms living in the biofilter materials are usually
mixed cultures. Various groups of microorganisms,
therefore, are involved in the energy conversion process
under aerobic or anaerobic conditions. However,
the


LIST OF TABLES
Table Page
2-1 Physical and chemical properties of H2S 10
2-2 Physiological characteristics of sulfur-oxidizing
bacteria 23
3-1 Description of compost used for this study 29
3-2 Properties of selected composts before and after
incubation 33
4-1 Retention times, limits of detection and operating
conditions for the Tracor 250H analyzer 78
4-2 Summary of initial compost properties 80
4-3 Particle size range distribution for selected
composts 82
4-4 Effect of gas retention time on H2S removal
efficiency 89
4-5 Effect of H2S concentration on removal efficiency 91
4-6 Models for the kinetics of H2S oxidation in
biofilter 124
5-1 Sulfur fractionation of original compost #17A and
compost at different heights in the filter .... 141
5-2 Effect of washing on compost pH and sulfate
content by DI water, NaOH and NaHC03
solutions 150
5-3 Performance of defective compost before and
after treatment 152
5-4 Effect of water washing on elimination of sulfate
in filter compost 157
6-1 Summary of Kanapaha biofilter bed design and
operation parameters 164
vii


Decomposition of composts under aerobic conditions and
the effect of H2S concentrations on the decomposition rate
are quantitatively determined. The half life time for the
composts tested is estimated to be between 3.3 and 6.1 years.
Hydrogen sulfide loading rate and maximum H2S
elimination capacity of the filter material are emphasized as
important design parameters. The maximum H2S elimination
capacity of a typical yard waste compost is determined to be
130 g-S/m3-hr under optimized conditions.
Hydrogen sulfide is oxidized to sulfuric acid in the
biofilter system, where biological oxidation plays a major
role. Acidification of the biofilter system and accumulation
of sulfate in the filter material are determined to be
natural features of the oxidation process, where the latter
is toxic for biological activities of the microorganisms.
Appropriate methods have been developed to effectively
mitigate this affect.
System 'upset1 is identified as being due to compost
dry-out and system overloading. Methods have been identified
to provide for recovery of the defective filter material.
Operation of a full scale biofilter system at a
wastewater treatment plant has been investigated. Both the
laboratory and full scale systems have demonstrated excellent
performance over substantial operational periods. Hydrogen
sulfide removal efficiencies of 99.9+% have been constantly
achieved when the H2S inlet concentrations are varied from 5
to 2650 ppmv.
xiv


14
Additional studies in the US have been carried out by
Carlson and Leiser (1966), Bohn and Miyamoto (1973), Bohn
(1975, 1976, 1977, 1989), Pomeroy (1982), Prokop and Bohn
(1985), Hartenstein and Allen (1986), Bohn and Bohn (1986,
1988), Hartenstein (1987), and Allen et al. (1987a, b, c;
1989).
In spite of the work mentioned above, most of the
research and development in biofiltration technology has
been carried out in Europe, especially in West Germany and
Holland. In the latter countries the principle of
biofiltration has been applied to a wide variety of
environmental problems. Among the many researchers in the
field, Ottengraf and coworkers in Chemical Engineering
Department, The Eindhoven University of Technology, Holland,
have contributed most of the theoretical research in
biofiltration in a series of papers which have been
published in the English language (Ottengraf, 1977;
Ottengraf and Van Den Oever, 1983; Ottengraf et al., 1984;
Ottengraf, 1986; Ottengraf et al. 1986; Ottengraf, 1987).
Also, Eitner in West Germany, has made significant
contributions to the research and development of
biofiltration ( Eitner and Gethke, 1987), although most of
his publications are in the German language (see
Hartenstein, 1987; Leson and Winer, 1991).
The practice and application of biofiltration has also
been reported in other countries such as Japan (Terasawa et
al., 1986), New Zealand (Rands et al.,
1981) and Canada


129
m3/m2-hr.
All three composts studied showed very good overall
performance characteristics during continuous long term
operation. The acclimation periods for all three composts
are similar, approximately 10 days. After this period, the
filters achieved stable operating conditions.
Tower #3 was operated under a lower gas loading rate,
50 m3/m2-hr (Figure 5-lc). This tower was not washed during
130 days of operation. When unpacked, the color of the
compost at the bottom of the tower (inlet) had changed to
yellowish-white, indicating accumulation of sulfur. The
compost at the bottom of the tower was wet, with a water
content of 55.9%, but the compost became drier as samples
were analyzed at various distances up the tower. The water
content of the compost in the top portion of the biofilter
(exit) was only 25.8%. This characteristic is probably
common for closed biofilter systems with no additional water
supplied to the system except water contained in the
influent gas stream. Although the gas is prehumidified and
almost saturated with water (RH > 95%) at ambient
temperature, the temperature in the compost can be a few
degrees (C) higher than the incoming gas stream due to the
exothermic oxidation reactions occurring in the system and
the biological respiration of the biomass. The rise of
compost temperature is more significant with high H2S
loading rates. The elevated bed temperature causes the gas
stream to become more unsaturated, and results in loss of


124
Table 4-6. Models for the kinetics of H2S oxidation in
biofilter.
H2S
Concen.
Range
(ppmv)
Kinetic
Order
Reaction
Coefficient
Equation
<200
First
k-L = 27.6/s
-dC/dt
c = cG
= ki C
exp(-k^t)
200-400
Fractional
kf = 0.067/s
c = co
(l-kft)2
>400
Zero
kQ = 0.567ppmv/s
-dC/dt
C = CQ
i ii
** **
o o
rt


H2S Removal Efficiency (%)
145
110
Section of Filter
Figure 5-5. H2S removal efficiencies in different
regions of the biofilter, Tower #2.
A: 0-0.2 m
B: 0.2-0.4 m
C: 0.4-0.6 m
D: 0.6-0.8 m
E: 0.8-1.0 m


121
represented by either Equation 4-4 (zero-order kinetics) or
equation 4-6 (first-order kinetics). Several empirical
equations have been tested, and Ottengraf's diffusion
limiting model was found to be the most accurate expression
describing H2S oxidation kinetics in this range for the
operating conditions employed.
In order to conveniently use Ottengraf's model, and
equation 4-8, however, it is necessary to define a new
parameter, the fractional-order reaction coefficient kf:
kf = (k0Dea/2miC05)^ (4-11)
It can be seen that kf is a function of the operating
conditions of the biofilter system, and under steady state
conditions, kf is constant.
Thus, equation 4-8 can be rewritten as:
C/CQ = (l-kft)2 (4-12)
or
(C/Cq)55 -1 = -kft (4-13)
where t = H/Ug is the reaction time.
When (C/C0)^ -1 is plotted against t, a straight line
should be obtained, and the slope of the line should equal
kf. A sample plot of the relation in equation 4-13 for the
concentration range 200 to 400 ppmv H2S is shown in Figure
4-21. The reasonableness of this plot supports the concept
of fractional-order dependence in the stated range of
concentrations.


148
microbial population uses these intermediate oxidized
compounds as its energy source and oxidizes the
intermediates to sulfate. As a result, the white color and
other deposits disappear within a few days. However, the
compost pH is significantly decreased and the sulfur content
of the compost is increased due to the formation of large
amounts of sulfuric acid. The latter results in the H2S
elimination capacity of the compost being reduced if no
counter measures are taken.
If the system is continuously overloaded by high H2S
input, the deterioration of biofilter performance can not be
reversed by simply decreasing the H2S loading rate. The
biological activity of the microorganisms in the filter is
strongly inhibited by high sulfate content and very low pH.
In this case, specific treatment is required to recover the
defective compost and the deteriorated system.
In order to determine a proper method to recover the
defective compost, a set of experiments were designed and
conducted. The resulting information is provided in the
following sections.
Selection of Chemical Solutions
The Column System #4 was used for this test. The
defective, white-colored compost was obtained from the
bottom region of Towers #1 and #2 and mixed thoroughly. One
hundred and fifty grams (150 g) of this compost was packed
in each of the 8 columns. Waste gas containing H2S was



PAGE 1

BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE BY YONGHUA YANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1992 UNIVERSITY OF FLORIDA LIBRARIES

PAGE 2

To my father for making this possible and In memory of my mother

PAGE 3

ACKNOWLE DGEMENTS I would like to express my sincerest gratitude and appreciation to the following people who made this research possible: To Dr. E. R . Allen, the doctoral committee chairman, for his foresight in support of this research, his encouragement, guidance and invaluable input during the course of this study and my graduate work. To Drs. D. A. Lundgren, B. Koopman, K. R. Reddy and D. P. Chynoweth for their interests in this research, helpful suggestions and participation on my graduate committee. To Dr. P. Urone for his friendship, kindness and valuable suggestions. To Mr. A. White and the Kanapaha Wastewater Treatment Plant engineers for their assistance in the research on the full scale biofilter system. To Ms. Yu Wang for her help on compost analysis. To Ms. S. Jordan for her help in construction and set up of the Lab scale biofilter units. To Mr. R. Vanderpool for his invaluable friendship, his ideas and his help in all aspects of my work that have made my years at the university much easier and so enjoyable. To my wife, Li, for her continuing support, encouragement, patience and understanding.

PAGE 4

TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii LIST OF TABLES vii LIST OF FIGURES ix ABSTRACT xiii CHAPTERS 1 INTRODUCTION 1 2 BACKGROUND 6 Properties of Hydrogen Sulfide and Regulations . . 6 Physical and Chemical Properties 6 Toxicity of H^S 9 Sources of H 2 § Emissions and Regulations 11 Biof iltration as an Air Pollution Control Technology 12 History and Development 13 Applications 15 Theoretical Basis 16 Biological Oxidation of Hydrogen Sulfide 21 3 PROPERTIES OF COMPOSTS AND THEIR DECOMPOSITION . . 25 Introduction 25 Selection of Filter Materials 26 Decomposition of Composts under Aerobic Conditions 28 Materials and Methods 32 Results and Discussion 39 Decomposition of composts 39 Effect of H 2 S on compost decomposition 45 Conclusions 50 4 DETERMINATION OF THE DESIGN AND OPERATIONAL PARAMETERS FOR BIOFILTER SYSTEMS 52 Introduction 52 System Design and Construction 52 iv

PAGE 5

The Dual Tower System 52 Portable Tower #3 58 Column System #4 58 Measurement Methods 61 Temperature 61 Pressure Drop 61 Gas Flow Rate 61 Sampling Methods 62 Compost Samples 62 Gas Samples 62 Water Samples 65 Compost Analysis Methods 65 Water Content 65 pH 66 Total Carbon and Total Nitrogen 66 Water Soluble Phosphorus (WSP) 66 Acid-extractable Cations 67 Particle Size Distribution 67 Porosity 67 Organic Matter 67 Particle Density 68 Bulk density 68 Sulfur Analysis Methods 68 Sulfur in Compost 68 Total Sulfur 73 Water soluble sulfur 74 Sulfide sulfur 74 Sulfate sulfur 74 Elemental sulfur 75 Pyrite sulfur 75 Organic sulfur 76 Sulfur in the Aqueous Phase 76 Sulfate sulfur 76 Sulfide sulfur 77 Total-S 77 Sulfur in Waste Gas 77 Results and Discussion 79 Pressure Drop 79 Effect of Gas Retention Time on H 2 S Removal ... 88 Effect of Concentration of H 2 S on Its Removal . 90 Effect of H 2 S Loading Rate on Its Removal 92 Effect of Compost Water Content on H 2 S Removal 93 Effect of Compost Acidity on H 2 S Removal 97 Effect of Temperature on H 2 S Removal 102 Effect of Sulfate on H 2 S Removal 106 Effect of Nutrient Addition on H 2 S Removal .... 108 Kinetics of H 2 S Oxidation in the Biofilter 109 Theoretical considerations 109 Determination of the kinetics of H 2 S Oxidation in a biofilter 114 v

PAGE 6

5 BIOFILTER PERFORMANCE AND CHANGES OF COMPOST PROPERTIES ASSOCIATED WITH LONG TERM OPERATION 125 Overall Performance of the biofilters 125 Accumulation of Sulfur in Compost and Its Effect on System Performance 137 System Upset and Recovery 146 Selection of Chemical Solutions 148 Effect of Water-Compost Contact Time on S0 4 Leaching Efficiency , 153 Effect of Water to Compost Ratio on S0 4 2 ~ Leaching Efficiency 153 6 FULL SCALE APPLICATION OF BIOFILTRATION TO CONTROL H 2 S EMISSIONS AT A WASTEWATER TREATMENT PLANT 158 Introduction 158 System Design and Construction 160 Sampling and Analysis Methods 163 Results and Discussion 166 Conclusions 182 7 SUMMARY AND CONCLUSIONS 184 REFERENCES 189 BIOGRAPHICAL SKETCH 199 vi

PAGE 7

LIST OF TABLES Table Page 2-1 Physical and chemical properties of H 2 S 10 22 Physiological characteristics of sulfur-oxidizing bacteria 23 31 Description of compost used for this study 29 32 Properties of selected composts before and after incubation 33 41 Retention times, limits of detection and operating conditions for the Tracor 250H analyzer 78 4-2 Summary of initial compost properties 80 4-3 Particle size range distribution for selected composts 82 4-4 Effect of gas retention time on H 2 S removal efficiency 89 4-5 Effect of H 2 S concentration on removal efficiency 91 46 Models for the kinetics of H 2 S oxidation in biofilter 124 51 Sulfur fractionation of original compost #17A and compost at different heights in the filter .... 141 5-2 Effect of washing on compost pH and sulfate content by DI water, NaOH and NaHC0 3 solutions 150 5-3 Performance of defective compost before and after treatment 152 54 Effect of water washing on elimination of sulfate in filter compost 157 61 Summary of Kanapaha biofilter bed design and operation parameters 164 vii

PAGE 8

6-2 Summary of periodic Kanapaha biofilter bed compost analyses during operational period from 5/10/88 to 2/5/91 168 6-3 Summary of Kanapaha biofilter influent and effluent gas sample analyses during three week start-up period 170 6-4 Gas sampling and analysis for Kanapaha biofilter bed, 2/5/91 178 6-5 Sulfur fractionation of a typical compost sample in Kanapaha biofilter bed 181 viii

PAGE 9

LIST OF FIGURES Figure Page 2-1 Solubility of H 2 S in water at 1 atm 7 2-2 Effect of pH on H 2 S Equilibrium 8 2-3 Biophysical model for the biological filter bed. The concentration profiles shown in the biofilm refer to: 1) Reaction limitation, 2) Diffusion limitation 18 24 Steps in the oxidation of different compounds by thiobacilli . The sulfite oxidase pathway is thought to account for the majority of sulfide oxidized 24 31 Schematic drawing of the experimental arrangement for the study of compost decomposition 34 3-2 Schematic drawing of the experimental arrangement for the investigation of the effect of H 2 S exposure on compost decomposition 3-3 Plot of C0 2 evolution from composts during the 122 day incubation 3-4 Decomposition stages and reaction rate coefficients for the four composts studied .... 44 3-5 Effect of H 2 S exposure on the rate of compost decomposition as measured by C0 2 respiration . . 46 36 Plot of C0 2 evolution as a function of square root of H 2 S concentration 48 41 Schematic drawing of the dual tower system 53 4-2 Sampling and measurement ports on towers. a. Tower #1 55 b. Tower #2 56 4-3 Schematic drawing of Tower #3 59 ix

PAGE 10

4-4 Schematic drawing of column system #4 60 4-5 Schematic drawing of the gas sampling assembly ... 64 4-6 Photograph of the sulfur distillation assembly — 70 4-7 Flow chart of the sulfur analysis procedures for compost 72 4-8 Pressure drop as a function of particle size range for different gas velocities 84 4-9 Pressure drop as a function of packing height for different compost particle size range 86 4-10 Pressure drop as a function of gas velocity for different types of compost 87 4-11 Determination of maximum H 2 S elimination capacity of compost 94 4-12 Effect of compost water content on H 2 S removal efficiency 96 4-13 Time reguired for dried compost to recover optimum efficiency 98 4-14 Effect of compost pH on H 2 S removal efficiency. Condition a: H 2 S loading rate: 10.5 g/m J -hr Gas loading rate: 15 nr/nr-hr Condition b: H 2 S loading rate: 35.4 g/m 3 -hr Gas loading rate: 26.1 nr/nr-hr. . . 100 4-15 Schematic drawing of the experimental arrangement for investigation of the effect of temperature on H 2 S removal efficiency 103 4-16 Effect of temperature on H 2 S removal efficiency . . 104 4-17 Effect of sulfate on H 2 S removal efficiency 107 4-18 Effect of nutrient addition on H 2 S removal. Total-S content in compost (mg-S/g) A: 17.5; B: 33.7; C: 20.2; D: 119.7 110 4-19 Linear least sguares regression analysis for zeroorder kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 m 3 /m 2 -hr, compost #17 ... 115 4-20 Linear least sguares regression analysis for firstorder kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 nr/nr-hr, compost #17 ... 116 x

PAGE 11

4-21 Determination of the fractional-order reaction rate coefficient, k f by linear least squares regression. Gas loading rate: 224 m /m 2 -hr, compost #17 117 4-22 Plot showing the fractional-order kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 m 3 /m 2 -hr, compost #17 118 423 Concentration profiles for H 2 S as a function of packing height within the biofilter. Gas loading rate: 224 m 3 /m 2 -hr, compost #17 119 51 Biofilter control of H 2 S during long term operation, a) Tower #1, compost #17A 126 b) Tower #2, compost #17 127 c) Tower #3, compost #16 128 5-2 Compost water content profile 132 5-3 pH changes of compost in different sections of the biofilter with operation time. a) Tower #1, compost #17A 134 b) Tower #2, compost #17 135 5-4 Total-S distribution profile in biofilter, Tower #1, after exposure to H 2 S for 100 days I 4 3 5-5 H 2 S removal efficiencies in different regions of the biofilter, Tower #2. A: 0-0.2 m, B: 0.2-0.4 m, C: 0.4-0.6 m, D: 0.6-0.8 m, E: 0.8 1.0 m 145 5-6 Effect of water-compost contact time on sulfate leaching efficiency 154 57 Effect of water/compost ratio on sulfate leaching 155 61 Schematic diagram of the Kanapaha biofilter bed system 161 6-2 Photograph of the grit chamber at Kanapaha Wastewater Treatment Plant (top view) . The chamber is covered to collect the malodorous gas 162 6-3 Biofilter off-gas sampling system 165 6-4 Photograph of the biofilter system at Kanapaha Wastewater Treatment Plant 169 xi

PAGE 12

6-5 Off-gas sampling locations on the biofilter beds and concentrations of hydrogen sulfide observed as a function of biofilter operating time 172 6-6 Concentration changes for hydrogen sulfide in gas samples contained in Tedlar bags as a function of container holding time 173 6-7 Effect of varying purging time for sample collection chamber prior to sampling on measured hydrogen sulfide concentrations 175 6-8 Compost samples taken from Kanapaha Wastewater Treatment Plant biofilter beds (2/5/91) . Left: sample taken from west bed. White color indicates high sulfur accumulation. Right: sample from east bed. Low sulfur content compost, color is close to the original (dark brown) 180 xii

PAGE 13

Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE By Yonghua Yang May 1992 Chairman: Dr. Eric R. Allen, Professor Major Department: Environmental Engineering Sciences A laboratory scale biological filter system for control of hydrogen sulfide (H 2 S) emissions has been developed and optimum design and operating parameters are evaluated. This biof iltration system uses yard waste compost as filter material, and the performance of the system for controlling waste gas containing H 2 S is evaluated through long term operation. Extensive tests have been conducted to determine the effect of various filter bed operating parameters such as pH, temperature, pollutant retention time, pressure drop, water content, etc. on H 2 S removal efficiencies. A biofilm model is used to characterize the macrokinetics of the biof iltration process. Models for the kinetics of H 2 S oxidation are developed that allow one to quantitatively predict the performance of the biofilter. xiii

PAGE 14

Decomposition of composts under aerobic conditions and the effect of H 2 S concentrations on the decomposition rate are quantitatively determined. The half life time for the composts tested is estimated to be between 3.3 and 6.1 years. Hydrogen sulfide loading rate and maximum H 2 S elimination capacity of the filter material are emphasized as important design parameters. The maximum H 2 S elimination capacity of a typical yard waste compost is determined to be 130 g-S/m 3 -hr under optimized conditions. v v Hydrogen sulfide is oxidized to sulfuric acid in the biofilter system, where biological oxidation plays a major role. Acidification of the biofilter system and accumulation of sulfate in the filter material are determined to be natural features of the oxidation process, where the latter is toxic for biological activities of the microorganisms. Appropriate methods have been developed to effectively mitigate this affect. System 'upset' is identified as being due to compost dry-out and system overloading. Methods have been identified to provide for recovery of the defective filter material. Operation of a full scale biofilter system at a wastewater treatment plant has been investigated. Both the laboratory and full scale systems have demonstrated excellent performance over substantial operational periods. Hydrogen sulfide removal efficiencies of 99.9+% have been constantly achieved when the H 2 S inlet concentrations are varied from 5 to 2650 ppmv. xiv

PAGE 15

CHAPTER 1 INTRODUCTION Hydrogen sulfide (H 2 S) is a highly toxic air pollutant which has been identified in the list of 190 air toxic substances in Title III of the 1990 Amendments to the Clean Air Act. Considerable amounts of H 2 S are produced in association with industrial processes, such as petroleum refining, rendering, waste water treatment, paper and pulp manufacturing, food processing, and in the treatment of "sour" natural gas and other fuels. Hydrogen sulfide is frequently the main component of most observable odorous emissions . Hydrogen sulfide is an odorous gas, and its presence at low concentrations is easily perceived and recognized due to its characteristic odor of rotten eggs. Hydrogen sulfide is perceptible to most people at concentrations in excess of 0.5 parts per billion (ppb) in air. Control of H 2 S emissions is essential to protect public health and welfare as well as to mitigate vegetation and material damage problems . Numerous processes involving physico-chemical principles have been developed in order to effectively remove hydrogen sulfide from air, waste gases and liquids

PAGE 16

(Bethea et al., 1973; Ferguson, 1975; USEPA, 1985; Lalazary et al., 1986; Walker et al., 1986; Lindstrom, 1990). Processes that have been used to remove H 2 S from waste gas streams involve either physical treatment or chemical oxidation. Some methods require addition of chemicals, and energy expenditure is usually necessary for physical treatment. Additional environmental problems are encountered with chemical additions, where resulting products and by-products require further treatment and disposal . Biof iltration can provide for a universal, simple, economicly feasible, and efficient pollutant-destructive control technology for a variety of toxic and hazardous substances in waste gas streams. In recent years biological filters have been developed and described which have the potential to simply and effectively control odors, including H 2 S emissions (Prokop and Bohn, 1985; Allen et al., 1987a; Eitner and Gethke, 1987; Hartenstein, 1987 ). Deodorization methods based upon the activity of microorganisms are beginning to attract increasing attention in the U.S. Although the biof iltration technique has been shown to be an efficient, practical and simple gas cleaning technology, which is increasingly being used around the world, the design and operation parameters as well as the microbial processes involved have not yet been very well defined. In particular, little research has been directed at the details of the biof iltration control of H^S. A

PAGE 17

systematic compilation of data from an operational point of view is also lacking. Most designs are conservatively based on blanket 'rule of thumb 1 criteria (Forster and Wase, 1987) . The performance of biofilter systems, therefore, is not readily predictable and sometimes these systems are not operated under suitable conditions. As a result, the desired odor control efficiency is sometimes not achieved (Allen et al., 1987b). It is essential that more work be done to demonstrate the effectiveness of these systems in order to support further progress in the use of biof iltration as well as to develop better biofilters, based on an understanding of the fundamental physical, chemical and biological processes involved. A major disadvantage of biof iltration technology is the^ limited degradation capacity represented by the volume of; waste gas treated per unit area of filter material per unit time (m 3 /m 2 -hr) . This limitation restricts the applicability of biof iltration systems to handling dilute waste gas streams and requires the filter bed to be large in order to handle high volumetric gas flows. In order to overcome the uncertainties and disadvantages encountered in the full scale application of biof iltration technology, an exhaustive study is necessary for the application of biof iltration technology to control the emissions of air pollutants. The objectives of the proposed research were to develop a quantitative knowledge of the principle and operation of

PAGE 18

4 a microbial biofilter system for removal of H 2 S from waste gas streams and determine the operating parameters necessary to optimize the performance of such a biofilter system. The objectives were achieved through the following studies: 1. Evaluation of the properties of filter materials and their decomposition characteristics under aerobic conditions. 2. Evaluation of the effects of design and operational parameters on H 2 S removal efficiencies on laboratory scale biofilter systems. Variables evaluated included temperature, pH, compost water and sulfate content, H 2 S elimination capacity, pollutant retention time, etc. . 3. Determination of the predictive relationships for H 2 S control efficiencies through chemical kinetic studies. 4. Evaluation of system performance and determination of optimum maintenance procedures for biof iltration control of H 2 S during long term operation. 5. Evaluation of the field performance of a full scale biof iltration system for control of H 2 S emissions at a local waste water treatment plant. The research reported here focuses mainly on the utilization, improvement and optimization of a compost biof iltration tower system. Optimization of this system has been directed toward the best achievable control of

PAGE 19

5 hydrogen sulfide. This research provides a detailed database on the effects of system variables on H 2 S control efficiency, which provide for optimization of design and operating conditions.

PAGE 20

CHAPTER 2 BACKGROUND Properties of Hydrogen Sulfide and Regulations Physical and Chemical Properties Hydrogen sulfide is a colorless gas that has a foul rotten egg odor and is slightly heavier than air. Hydrogen sulfide is moderately soluble in water. The solubility of H 2 S decreases with increasing temperatures. Figure 2-1 shows the solubility of H 2 S as a function of temperature. Dissolved H 2 S dissociates in accordance with the following reversible ionization reactions: H 2 S S5=3* HS" + H + (2-1) HS" ^S 2 " + H + (2-2) The distribution of the above species as a function of pH is shown in Figure 2-2. It is apparent from Figure 2-2 that the concentration of HS~ species is insignificant when pH values are less than 6. The latter condition is normal in a biofilter system for control of H 2 S. S 2 ~, on the other hand, may not occur at all. Hydrogen sulfide can serve as a reducing agent, reacting with sulfuric acid (H 2 S0 4 ) to form sulfur dioxide (S0 2 ) and elemental sulfur (S°) (Greyson, 1990): 6

PAGE 21

Figure 2-1. Solubility of H 2 S in water at 1 atm. Data adopted from Piscarcyzyk, 1982.

PAGE 22

3 Figure 2-2. Effect of pH on H 2 S equilibrium. Source: Sawyer, 1967.

PAGE 23

9 H 2 S + H 2 S0 4 = S0 2 + S° + 2H 2 0 (2-3) Hydrogen sulfide also burns in air to form sulfur dioxide and water: 2H 2 S + 30 2 = 2S0 2 + 2H 2 0 (2-4) Table 2-1 summarizes the physical and chemical properties and the odor threshould of H 2 S. Toxicity of H 2 S Hydrogen sulfide is almost as toxic as hydrogen cyanide (HCN) , which is used in prison gas chambers (Parker, 1977) . Human exposure to small amounts of H 2 S in air can cause headaches, nausea, and eye irritation, and higher concentrations can cause paralysis of the respiratory system, which results in fainting and possible death. Concentrations of the gas approaching 0.2 percent (2000 ppmv) are fatal to humans after exposure for a few minutes (NRC, 1979) . Hydrogen sulfide has a characteristic rotten egg smell at low concentrations. But as levels of H 2 S increase, a person's ability to sense dangerous concentrations by smell is quickly lost. If the concentration is high enough, unconsciousness will come suddenly, followed by death if there is not a prompt rescue. The Occupational Safety and Health Administration (OSHA) has established limits for work place exposure to H 2 S

PAGE 24

10 Table 2-1. Physical and Chemical Properties of H 2 S a . Molecular Weight 34.08 Boiling Point, °C -60.2 Melting Point, °C -83.8 to -85.5 Vapor Pressure, -0.4°C 10 atm 25°C 20 atm Specific Gravity (Relative to Air) 1.192 Auto Ignition Temperature, °C 250 Explosive Range in Air, % 4.5 to 45.5 Odor Threshold, ppbv 0.47 a Source: USEPA, 1985.

PAGE 25

11 at 20 ppm (15-minute exposure) for an acceptable ceiling concentration and 50 ppm for a maximum exposure during an 8hour work shift if no other measurable exposure occurs. The National Institutes of Occupational Safety and Health (NIOSH) established an H 2 S exposure level at 10 ppm (10 minutes) as a maximum permissible limit (once per 8-hours shift) , with continuous monitoring reguired where H 2 S concentrations could egual or exceed 50 ppm or greater (NIOSH, 1979) . Hydrogen sulfide is an explosive gas. The lower and upper explosive limit are 4.5 and 45 percent in air by volume, respectively. Hydrogen sulfide can attack materials and cause discoloration and tarnishing. Materials commonly affected are paint, copper, zinc and silver (Painter, 1974) . Sources of H ^ S Emissions and Regulations Natural emissions are mainly caused by biological decay of protein materials. The natural global rate of emission is estimated to be about 84 Tg/year (Urone, 1986) . Anthropogenic emission sources include petroleum refining, natural gas plants, sewage treatment facilities, coke ovens, Kraft paper pulp plants, and waste disposal sites. There are no federal U.S. emission standards for H 2 S at present, nor are there federal ambient air guality standards for this gas, but a number of states have

PAGE 26

12 established independent standards for H 2 S emissions. These states, which include California and New Mexico (10 ppm) , and Ohio and Michigan (1670 ppm) . California, Kentucky, Minnesota, Montana, New Mexico, New York, North Dakota and Pennsylvania also have air quality standards for H 2 S. The standards vary from 0.003 ppm for New Mexico to 0.1 ppm for Pennsylvania, whereas and most of the other states specify a standard of 0.03 ppm (Urone, 1986). Since H 2 S is a highly toxic air pollutant, H 2 S has been identified by the USEPA as one of 190 air toxic compounds in Title III of the 1990 Amendments to the Clean Air Act. In view of the wide spread exposure to this pollutant, emission and air quality standards for H 2 S are going to be set in the near future by EPA. Biof iltration as a Air Pollution Control Technology Biological degradation is widely used for treatment of liquid and, to a lesser extent, solid wastes, but has received little attention as a means of controlling emissions of industrial gaseous wastes. Biof iltration is a relatively new technology for control of air pollutants, in which the air contaminants from off-gas streams are biologically removed in a solid biological reactor. While it is a well established air pollution control technology in European countries, biof iltration as an air pollution control technology has received little attention and application in the United States. Few environmental

PAGE 27

13 professionals in this country appear to be aware of the •biof iltration 1 process and its applications. Although there are some applications of biof iltration in the U.S. and some technical papers have been published in the English language, most of the research and development work on biof iltration has been conducted in Europe and the majority of the recent research data have been published in the German language. Excellent reviews of previous biof iltration work have been published by Hartenstein (1987), Leson and Winer (1991), Ergas et al. (1991), and Dharmavaram (1991) . History and Development The first deodorization method based upon the use of a soil bed in the U.S. was developed and patented by Pomeroy in 1957. Later, Pomeroy (1982) described the deodorization of waste gases emitted from sewer lines by a soil bed system used in Los Angeles in 1957. The microbiological degradation of sulfur-containing gases in the filter bed was observed to be effective in these studies. Other early applications of biological treatment of odorous gases include a soil bed system built in Nurnberg, West Germany, in 1959 and biofilters built in Geneva, Switzerland, and Mercer Island, Washington, to remove odors from wastewater treatment and compost manufacturing, respectively, in the mid-1960s (Bohn and Bohn, 1987) .

PAGE 28

14 Additional studies in the US have been carried out by Carlson and Leiser (1966), Bohn and Miyamoto (1973), Bohn (1975, 1976, 1977, 1989), Pomeroy (1982), Prokop and Bohn (1985), Hartenstein and Allen (1986), Bohn and Bohn (1986, 1988) , Hartenstein (1987), and Allen et al. (1987a, b, c; 1989) . In spite of the work mentioned above, most of the research and development in biof iltration technology has been carried out in Europe, especially in West Germany and Ho Hand. In the latter countries the principle of biof iltration has been applied to a wide variety of environmental problems. Among the many researchers in the field, Ottengraf and coworkers in Chemical Engineering Department, The Eindhoven University of Technology, Holland, have contributed most of the theoretical research in biof iltration in a series of papers which have been published in the English language (Ottengraf, 1977; Ottengraf and Van Den Oever, 1983; Ottengraf et al., 1984; Ottengraf, 1986; Ottengraf et al. 1986; Ottengraf, 1987). Also, Eitner in West Germany, has made significant contributions to the research and development of biof iltration ( Eitner and Gethke, 1987) , although most of his publications are in the German language (see Hartenstein, 1987; Leson and Winer, 1991). The practice and application of biof iltration has also been reported in other countries such as Japan (Terasawa et al., 1986), New Zealand (Rands et al., 1981) and Canada

PAGE 29

15 (Rotman, 1991a) . At present, biof iltration is considered to be a stateof-the art technology for odor removal in West Germany, and it has been estimated that 40% of deodorization facilities at wastewater treatment plants are biofilters (Frechen and Kettern, 1987) . Applications The first systematic study of odor control using biof iltration in this country was conducted by Carlson and Leiser (1966). They studied the removal efficiencies of sewage odors using a laboratory scale soil bed. Using hydrogen sulfide as the test gas, a 99% removal efficiency was achieved, and biodegradation was reported to be the primary removal mechanism. Prokop and Bohn (1985) reported that a soil bed system for control of rendering plant odors had been in operation since September, 1983. The soil bed treats 1100 m 3 /h of cooker non-condensable waste gases using a bed surface area of 420 m . In this work an odor removal efficiency of 99.9% was obtained. Rands et al. (1981) reported that a full-scale compost filter system was constructed in 1978 at Moerewa, New Zealand, to treat odors from a rendering plant. The system was designed to treat 900 m 3 /h of air containing hydrogen sulfide concentrations up to 1000 parts per million (ppm) by volume. An average H 2 S removal efficiency of 99.9% was

PAGE 30

16 observed. Allen et al. (1987a, b, c) investigated a compost based tower biofilter system used for odor control in a wastewater treatment plant. The odor-causing compounds identified were reduced sulfur compounds such as H 2 S, methyl mercaptan, dimethyl sulfide and dimethyl disulfide as well as terpene hydrocarbons. Removal efficiency for total reduced sulfur compounds (TRS) was 65 to 72%. The poor performance of this system was determined to be the short residence time in the system, poor gas distribution, and improper maintenance. Biof iltration control of volatile organic compounds (VOCs) has been reported by Ottengraf (1986), Kampbell et al. (1987), Bohn (1989), Paul and Castelijn (1987), and Hack and Habets (1987) . In recent years, increasing numbers of biological filters are being used around the world for odor control. It has been estimated that more than 500 biofilters are currently operating in Europe (Leson and Winer, 1991) . Excellent summaries of recent applications have been provided by Bohn and Bohn (1987) and Rotman (1991b) . Theoretical Basis The concept of a biological-film or 'biofilm* is freguently used to describe degradation processes in agueous systems ( Williamson, 1973; Williamson and McCarty, 1976a, b; Jennings et al., 1976; Rittmann and McCarty, 1978). This concept has been adopted and improved to describe the

PAGE 31

17 biof iltration processes (Ottengraf, 1986; Hartenstein, 1987; Paul and Castelijn, 1987; Van Lith, 1989). In particular, Ottengraf and coworkers have carried out systematic studies delineating the overall process and have presented sufficient experimental data to support the proposed model. In biof iltration, evenly distributed waste gases are forced through a biologically active material, such as soil, peat or compost. Many of the pores of the filter material particles are filled with water. Microorganisms are attached to the particle surfaces to form a layer of film. This wet, biologically active layer surrounding the particles is called a biofilm. The biophysical model proposed by Ottengraf for the biofilm is shown in Figure 2-3. The mechanism of the biological process is derived from a combination of physical, chemical and biological processes that occur in the filter material and is related to two processes in particular; sorption and regeneration. As waste gases pass through the countless narrow pores of the filter material, air contaminants as well as oxygen will adsorb on the surfaces of the pores and dissolve in the liguid phase of the wet biofilm. The absorbed and adsorbed gases are guickly degraded by the biofilter 's enormous microbial population. In this way a concentration gradient is created in the biofilter, which maintains a continuous mass flow of the component from the gas to the wet biofilm. Activity of the biofilter depends mainly on the population of the microorganisms. Soil biofilters can

PAGE 32

18 Figure 2-3. Biophysical model for the biological filter bed. The concentration profiles shown in the biofilm refer to: 1) Reaction limitation, 2) Diffusion limitation. (Source: Ottengraf, 1986, p. 436) .

PAGE 33

19 contain 1 billion bacteria, 10 million actinomycetes and 10,000 fungi per gram of soil (Bohn and Bohn, 1987). The role of these microorganisms is to oxidize combined carbon, nitrogen and sulfur to carbon dioxide, nitrogen and sulfate, respectively, before the compounds leave the bed. The air contaminants are, therefore, effectively removed from the waste gas streams. For good engineering design and environmental decision making, it is essential to understand the mechanisms involved and to reliably predict the kinetics of the biological reactions taking place in these biofilter systems. Many general kinetic models have been developed to predict the behavior of bioreactions in a biological film, none of these models, however, is specific enough to explain the biodegradation of hydrogen sulfide in a biofilter system. Jennings et al. (1976) developed a mathematical model to predict the percentage removal of a pure, non-adsorbable, biodegradable substrate in a submerged biological filter using the non-linear Monod expression for the substrate utilization rate. In their model, the authors start from a biological slime layer coating a spherical particle. The slime layer is in turn surrounded by a liguid boundary layer. They concluded that even at relatively high values of influent substrate concentrations, the biological removal of a single substrate follows first order kinetics.

PAGE 34

20 Another model developed by Rittmann and McCarty (1978) is a variable-order model of bacterial-film kinetics which incorporates liquid-layer mass transport, biofilm molecular diffusion and Monod kinetics. These investigators concluded that at low substrate concentrations, the reaction follows first order kinetics, whereas at high concentrations the reaction follows one-half order kinetics. Based on their biophysical model (Figure 2-3), Ottengraf and Van Den Oever (1983) have developed a mathematical model to describe the kinetics of organic compound removal from waste gases for a biofilter system. The model was developed and tested using a soil bed for the removal of toluene, butylacetate, ethylacetate and butanol. From their experimental results, they concluded that all the carbon sources investigated were eliminated according to a zero order reaction, even at very low concentrations of the substrates . Kampbell et al. (1987) investigated the biodegradation of propane, isobutane and n-butane by soil biofilter beds. They suggested that at low concentrations the rate of biodegradation was proportional to the concentration of the organic compounds (first order reaction) , and at higher concentrations the rate becomes independent of the organic compound concentration (zero order reaction). The degradation kinetics appeared to follow a hyperbolic function:

PAGE 35

21 1/V = (1/V max ) +( yV^Jd/S) (2-5) where : V = the biodegradation rate, mg hydrocarbon/kg soil-h v max = ^ e max i mum possible biodegradation rate, mg hydrocarbon/kg soil-h Kjjj = an empirical constant, half saturation value, ppm S = the concentration of organic compound in air, ppm The values of K m and V max can be obtained from a Lineweaver-Burk plot of 1/V against 1/S. Biological Oxidation of Hydrogen Sulfide Hydrogen sulfide may be utilized by microorganisms in three different ways: assimilation, mineralization and sulfur oxidation (Atlas and Bartha, 1981; Grant and Long, 1981) . However, the rates of uptake of hydrogen sulfide based on the assimilation processes are far too low to achieve reasonably high removal efficiencies from a highly loaded waste gas stream. The most important and efficient way for microorganisms to utilize hydrogen sulfide is by the oxidation of sulfur to gain energy. In this process, relatively large quantities of sulfur are oxidized in order for the microbes to receive sufficient energy. The microorganisms living in the biofilter materials are usually mixed cultures. Various groups of microorganisms, therefore, are involved in the energy conversion process under aerobic or anaerobic conditions. However, the

PAGE 36

colorless sulfur bacteria are believed to play the major role and their ability to oxidize reduced inorganic sulfur compounds has been clearly established (Roy and Trudinger, 1970; Kuenen, 1975; Brock and Madigan, 1988). The oxidation of inorganic sulfur compounds is carried out by a spectrum of sulfur-oxidizng organisms which include 1) obligately chemolithotrophic organisms, 2) mixotrophs, 3) chemolithotrophic heterotrophs , 4) heterotrophs which do not gain energy from the oxidation of sulfur compounds but benefit in other ways from this reaction, and 5) heterotrophs which do not benefit from the oxidation of sulfur compounds. Physiological characteristics of some sulfur-oxidizing bacteria are summarized in Table 2-2. Options for microbial metabolism of hydrogen sulfide must employ one or more of the following metabolic pathways: 1) aerobic oxidation, 2) anaerobic oxidation, and 3) photosynthetic dissimilation. Biof iltration of waste gases is a process utilizing aerobic conditions in most cases. In aerobic oxidation, sulfur-oxidizing bacteria oxidize H 2 S to elemental sulfur or higher oxidation states using oxygen (0 2 ) as an electron acceptor. The biological steps in the oxidation of various sulfur compounds are summarized in Figure 2-4.

PAGE 37

23 Table 2-2. Physiological characteristics of sulfuroxidizing bacteria. Lithotrophic Electron pH Range Donor for Growth Thiobacillus Species Growing Poorly in Organic Media: 1. T. thioparus H 2 S, sulfide, S°, S 2 °3 6-8 7. T . denitrif icans H 2 S, S^ / S 2 0 3 26-8 3. T. neapolitanus s°, s 2 o 3 2 5-8 4. T. thiooxidans s° 2-5 5. T. ferrooxidans S°, sulfides, Fe 2+ 1.5-4 Thiobacillus Species Growing Well in Organic Media: l. T. novellus s 2°3 2 " 6-8 2. T. intermedius S 2 °3 3-7 Filamentous Sulfur lithotrophs Beaaiatoa H 2 S, S 2 0 3 2 6-8 Thiothrix H 2 S 6-8 Other Genera Thiomicrospira S 2 0 3 2 , H 2 S 6-8 Thermothrix H2 S / ^2^3 ' SO36.5-7 Sulfolobus a 1-4 .5 a Archaebacterium. Source: Brock and Madigan, 1988.

PAGE 38

24 Sulfide S2Cell-Bound Sulfur Complex 1/ Elemental Sulfur S 2 0|' Thiosulfate Sulfite Electro Transport System 2e ADP Electron Transport Phosphorylation Sulfite Oxidase ATP Adenosine Phosphosulfate (APS) Substrate Level Phosphorylation i ADP SO 2Sulfate so: Figure 2-4. Steps in the oxidation of different compounds by thiobacilli . The sulfite oxidase pathway is thought to account for the majority of sulfide oxidized. (Source: Brock and Madigan, 1989, p. 704) .

PAGE 39

CHAPTER 3 FILTER MATERIALS AND THEIR DECOMPOSITION UNDER AEROBIC CONDITIONS Introduction Biof iltration systems or biofilters employ physical, chemical and biological processes such as adsorption, absorption and microbial digestion and oxidative degradation to remove air pollutants from waste gas streams. Microbial degradation and oxidation of the pollutants, however, appear to be the primary removal mechanisms within a biofilter. In the biodegradation process, pollutants are consumed by the microorganisms, providing an energy source or essential nutrients and are converted usually to, less harmful compounds. The filter materials used, on the other hand, must provide the proper environment for microbial growth and contain materials on which the microbes can feed to ensure that the microbial population can develop and survive. The effectiveness of a biofilter material depends on its physical, chemical and biological characteristics. The lifetime of a biofilter material mainly depends on its rate of carbon (C) and nitrogen (N) mineralization. When available C and N in the filter material are no longer sufficient to support the microbial population in the system, then the material is no longer suitable as a 25

PAGE 40

biofilter. The C or N deficient filter material must be replaced by freshly prepared material and the discarded filter material has to be properly disposed of with due caution for environmental impact. One of the most common options is land application. Determination of the decomposition characteristics of the filter material is, therefore, necessary for usage of the biofilter and eventual land disposal applications. Considerations necessary for selection of appropriate filter materials and the decomposition of such materials under aerobic conditions are discussed in this section. Selection of Filter Materials Effective removal of air contaminants using a biofilter relies on the properties of the filter material, especially the nature and activity of the biomass. The filter material provides the necessary environment for microorganisms to survive, generate, function and allows the entire sequence of biof iltration processes to be carried out. The filter material serves as 1) support material for the microbes, 2) supplemental or alternative nutrient source, 3) moisture storage reservoir, 4) surface area for sorption of air pollutants and interaction between the pollutants and the microorganisms, and 5) a buffer volume for variations in water content and gas conditions during operation (Eitner, 1989). In general, the following factors need to be considered when choosing a suitable filter

PAGE 41

27 medium: 1) . Density: Too dense material may contain a large fraction of inorganic materials such as stone and sand which are unsuitable as carbon and energy sources for microbial growth. 2) . Structure : Structure of the medium will affect the uniformity of the filter load. Too large sized materials should be avoided because the surface-to-volume ratio will be reduced. 3) . Particle Size Distribution: Too small particles affect the pressure drop by compacting and restricting the gas flow. 4) . Pore Volume (void fraction) : This property determines the total surface area available for reaction, also it will affect pressure drop. 5) . Organic Matter Content: The organic matter controls the microbial population and the useful service life of the filter media. 6) . pH Value: pH will affect the nature and level of the microbial population and activity. 7) . Water Retention Capacity: This property will determine the consistency in liguid water content of the filter material, and 8) . Economics: Reasonable Capital and operating expenditures. All of these reguirements can be met by selecting suitable filter materials. Many kinds of filter materials have been used in biof iltration applications. Examples include field soils, compost, peat, bush, clay, volcanic ash, sand, bark and a combination of such materials (Rands et al., 1981; Prokop and Bohn, 1985; Terasawa et al., 1986 Frechen and Kettern, 1987) . The performance of these materials, however, can be very different due to the diversity of their physical and chemical properties. Compost has been considered to be the best choice for filter

PAGE 42

28 materials and has been involved in most applications (Don, 1985; Eitner, 1989), since it provides favorable conditions for supporting microbial populations as well as having superior physical and chemical properties. The properties of individual composts depend on the materials from which they are derived and the composition of the final product. The filter materials used in this research were mainly yard waste compost and sewage sludge compost or a combination of both. These composts were obtained from different sources and used for different purposes. A general description of the types of composts and their sources are summarized in Table 3-1. The physical and chemical analyses data for the composts listed in Table 3-1 are presented in the corresponding chapters where the use of specific composts is discussed. Decomposition of Composts under Aerobic Conditions A number of investigations have been carried out to study the decomposition of anaerobically digested sewage sludges in soils (Miller, 1974; Tester et al., 1977; Terry et al., 1979b; Sweeney and Graetz, 1988; Gale, 1988). Decomposition of fresh and anaerobically digested plant biomass in soil is also reported by Moorhead et al. (1987). Only limited information, however, is available concerning the decomposition of compost. Tester et al. (1977, 1979) stated that the decomposition of compost in soil is not only related to the physical and chemical properties of the

PAGE 43

29 Table 3-1. Description of composts used for this study. Compost ID# Source Description 1 Pompano Beach 4 2 Pompano Beach 3 Pompano Beach 6 Pompano Beach 12 Kanapaha* 13 Kanapaha 13-1 Kanapaha 13-2 Kanapaha 14 Kanapaha Fort Lauderdale sewage sludge compost. Not completely composted. Seven months old when first used (used for decomposition study) . Two parts yard waste and one part stable cleaning sewage sludge mixed and composted. Seven months old when first used (used for decomposition study) . Yard trash compost. 13 months old when first used (used for decomposition study) . 25% by volume of sewage sludge compost and 75% of yard trash mixed and composted about 19 months old when first used (used for decomposition study) . Pompano Beach compost similar to Compost #6 mixed with tree bark, yard waste and sewage sludge; lime was used to adjust pH before use. Used in Kanapaha filter bed from 11/20/88. Compost obtained from the filter bed in 5/16/90. Same as #12, compost obtained and used in Tower #1 from 12/20/90. Same as #12, compost obtained in 2/5/91. Same as #12, compost obtained in 3/20/91. Yard trash, grass and sewage sludge were mixed and composted; lime was used to adjust pH; about 2.5 years old when obtained and used in Tower #2, 12/20/90.

PAGE 44

30 Table 3-1 — Continued, Compost Source Description ID# 16 WRR C Yard trash compost, 3.5 months old when first used in Tower #3 from 1/27/91. 17 WRR 1:1 by volume of yard trash and grass composted; about 3.5 months old when first used in Tower #2 from 1/27/91. 17A WRR Compost #17 mixed with 2% lime (CaC0 3 ) , by dry weight of compost. Used in Tower #1 from 1/27/91. a Broward County Streets and Highways Division, 1600 NW 30th Avenue, Pompano Beach, FL, 33069. b Kanapaha Wastewater Treatment Plant, Gainesville, FL 32602. c Wood Resource Recovery, Inc., Gainesville, FL.

PAGE 45

31 compost but also is a function of the particle size of the compost. The decomposition was observed to be directly related to the carbon content in the compost. Decomposition is affected by a number of environmental conditions, for instance, pH, moisture content, and the presence or absence of foreign chemicals (Miller and Johnson, 1964; Terry et al., 1979a, b; Delaune et al., 1981) . In the application of biof iltration to control H 2 S emissions, the compost filter material is subjected to conditions that are quite different to that for land applications of compost. In the former case, the compost is exposed to a gas stream which may contain a variety of chemicals, especially H 2 S, at various concentrations. The presence of xenobiotics in the gas streams and filter materials could change the population and composition of the microorganisms in the compost or significantly affect their metabolic processes. As a result, the decomposition rate of the compost can be altered. Unfortunately, little information can be found in the literature related to this topic. The objectives of this study were (i) to evaluate the decomposition of four types of compost by determining the C0 2 evolution, and (ii) to investigate the effect of H 2 S at various concentrations on compost decomposition. Such information is valuable for biofilter design and for justifying land disposal applications of the compost after use as a biofilter medium.

PAGE 46

32 Materials and Methods Four types of compost samples were investigated for their decomposition characteristics during the course of this study. All of the compost samples were obtained from Broward County Streets and Highways Nursery Division, Pompano Beach, Florida. The composts were stored in sealed plastic bags at room temperature (23±2 °C) before use. A brief description of the composts used in this study is presented in Table 3-1 ( Composts #1, #2, #3 and #6). The compost samples were analyzed for their physical and chemical properties at the beginning and the end of the investigation. The results of these studies are presented in Table 3-2. Each cured compost was passed through a 10 mm screen to remove larger materials. The compost samples are then placed in 225-mL wide mouth bottles directly for incubation. The experimental arrangement for the decomposition study is shown in Figure 3-1. Four types of compost and one blank, each with three duplicates, were investigated. Compressed air from the laboratory house air supply is controlled to about 4 psig by a regulator. The air stream is passed through a scrubber system consisting of 4N NaOH to remove C0 2 and distilled water to saturate the air stream. A dead volume is placed before the 4N NaOH scrubber as a safety precaution in the event that the air system causes a backpressure forcing scrubbing solution against the air system. An empty impinger is placed after the water scrubber to separate larger water droplets from

PAGE 47

33 • c o •H M O +J a) 0J ID O VO cn n H o o o o cfl VO +J Ul in • • oo • r~ oo in CO • VO ja • H VO • • cn Cn VO rH CM CM < VO VO cn co co CTi in CTi rH in o 4J c in a) •H o ^ o a, o 0) in cn in in in o o in o g a> CN o co cn o 00 o o in O o c co -P • • • • CO • o in • in • rH 3 • cn - in CM • in < oo vo t«* «t cn co co o o o o cu cm 4-> W • • • • r» • o o CO rH • CO e =«= <«-i 3 • I** in • CN o CO CO CO o rH 0 < CO in vo cn co cn rH VO in VO H rin 0 4J w a) T3 O *H o a> a o a> m cm co *r cn o o m o (0 H • • • • co • in in rH in • CM 0 o a) 3 • cn cn in vo • CM r> CM in CM VO CM w t• • • • in • (O t> • VO 0 =**= vo t"» 4| VO in r~ cn co co H «f in rH CO CTi W P a) en a> o •H O M p» ro oo cn o CO o o o o +j a o a) vo • • *t • o • CM CM rH • CM • O Cn vo • rH o o CO a> o a> 3 00 vo r» rH CO CO rH CO CM CO CO U CQ 0 H t^4J -P /kg • sss Cn CM 1 p p £ « rt CU U 55 1 1 1 (fl O a) u U rH rH rH H Q) CO CO (0 CO cn c 3 c (1) 0 4J H 4J 4-> +J 2 -P U S U M S re o CO o o \ oj cu as M S H H U SE

PAGE 48

34 2 o c '5. E 0| — to s I Ij -P O 41 -P c 0) B o cn c <3 5i (T3 P C 1 X! 3 O -P CO M H I n

PAGE 49

35 the air stream. Two water scrubbers are used in series to ensure that the air stream is completely free of alkali and to resaturate the air with water vapor in order to keep the water content of the composts constant. The resaturated, C0 2 -free air stream is then forced to the manifold where it is split into 15 streams. Each stream goes into one incubation-absorption unit. Fifty grams of compost sample is put in each incubation bottle. The C0 2 evolved from each of the compost samples is collected in two 25-mL, 0.5N NaOH collectors in series. The total air flow rate is controlled by a needle valve located in front of the manifold. Syringe needles are used as flow regulators to equalize the air flow through the 15 incubation units. The air flow rate through each unit is adjusted to about 15±2 mL/min. The incubation system is continuously operated at constant temperature (23±2°C) . After flushing the residual air from the incubation bottles, the outlet tube of each bottle is attached to the C0 2 collectors. C0 2 collectors are replaced with fresh solutions periodically during the incubation period. The system is leak checked before the incubation. Evolved C0 2 is efficiently trapped by two absorption collectors in series. Tests have shown that the first tube absorbed more than 95% of the total C0 2 evolved. C0 2 evolution is measured as described by Stotzky (1965) with minor modifications. After C0 2 absorption, the solutions in the two collectors of each unit are mixed and

PAGE 50

36 titrated with standard IN HCl. The C0 2 samples collected from each of the control bottles are concomitantly titrated. The C0 2 evolved for individual samples is calculated as follows (Stotzky, 1965) : C0 2 = (B V)NE (mg) (3-1) where : B = volume of HCl used to titrate the NaOH in the controls to the end point, (mL) ; V = volume of HCl used to titrate the NaOH remaining in the C0 2 collectors after treatment to the end point, (mL) ; N = normality of the HCl, (meq/mL) ; E = eguivalent weight, (mg/meg), for C0 2 , E = 22 (mg/meg) . To investigate the effect of H 2 S concentration on compost decomposition, one hundred grams of Compost #6 was used as the test material. The experimental arrangement for this test is similar to that for the compost decomposition test with some minor modifications (Figure 3-2) . Room air is forced through a scrubber containing 4N NaOH to absorb C0 2 from the air stream. The C0 2 free air is then saturated by bubbling through DI water. Pure H 2 S is then mixed with the pretreated air stream to obtain the test gas mixture with the desired H 2 S concentration. The treated gas stream is vented through the manifold, where it is split into four sub-streams: one of these sub-streams is vented to a control column (empty) , and the other three streams to duplicates of three compost columns. The gas is forced vertically through the compost from bottom to top at a flow rate of 30 mL/min.

PAGE 51

37 IH>5 — i <1 • $ a z Si 0) -P O P u o g P c 0) s o CP c 4 XIX 55 U n o a n) a) P c 0 g •H U o a) a) o JS p tr> c JC > o c CM I ©

PAGE 52

38 A portion of the hydrogen sulfide is adsorbed and/or oxidized by the compost and the remaining H 2 S in the effluent gas is absorbed by two H 2 S scrubbers in series which contain 200 mL and 25 mL IN zinc acetate (ZnAc) solution, respectively. The absorbing reaction used by the H 2 S collectors is: ZnAc + H 2 S -> ZnSi + HAc U H + + Ac" (3-2) Total flow in the system is measured by pre-calibrated rotameters. H 2 S concentrations in the inlet gas to the compost column are controlled by adjusting the flow rates of mixing for the C0 2 -free air and the pure H 2 S gas. Gas samples from the influent gas stream are taken periodically by gas-tight syringes, diluted with prepurified nitrogen (N 2 ) and analyzed for H 2 S content by a Tracor 250H analyzer (See Chapter 4 for details) . The effluent gas from the filter columns is first passed through two scrubbers in series containing IN ZnAc to absorb any H 2 S remaining in the gas stream. Residual C0 2 in the effluent gas stream is subsequently absorbed by 0.5N NaOH solution and titrated as described previously. Each compost sample is incubated at a desired H 2 S concentration level for 24 hours. After the incubation period the compost as well as the absorption solutions are replaced by fresh compost samples and absorbing solutions for operation at the next H 7 S concentration level. The

PAGE 53

39 system was previously tested to obtain absorption efficiencies of H 2 S and C0 2 in the ZnAc traps. The results showed that the H 2 S absorption efficiency was greater than 99% and the C0 2 absorption was less than 2% for the ZnAc solutions used. Results and Discussion Decomposition of Composts The decomposed C evolved as C0 2 from the four composts studied during the 122 day incubation period is shown in Figure 3-3. The decomposition patterns of composts #2, #3 and #6 are somewhat similar. Decomposition of compost is initially rapid, from 40 to 52% of the total C0 2 produced in the 122 days is evolved in the first 42 days of incubation. A total of 9.2, 5.7, 6.1, and 4.4% of the original C was decomposed and released as C0 2 for composts #1, #2, #3, and #6, respectively during the total 122 days of incubation. It appears that decomposition rates of the composts are inversely proportional to their age, in other words, the older the compost, the slower the decomposition. All except compost #1, showed decomposition rates which were similar. Compost #1, however, was not completely composted when used. Also the organic matter content of this compost is higher than that for the others tested. Initial and delayed higher decomposition rates for compost #1 suggest a two stage incubation involving an initial 'conditioning' step followed by a 'conditioned' decomposition. The compost

PAGE 54

40 Time (day) Compost #1+ Compost #2 o Compost #3 a Compost #6 Figure 3-3. Plot of C0 2 evolution from composts during the 122 day incubation.

PAGE 55

41 decomposition rates measured here are much slower than those found for soils. Tester et al. (1977) reported that approximately 16% of the compost C was evolved as C0 2 during 54 days of incubation, when 2 to 6% fresh sewage sludge compost was incubated with soils. Miller (1974) reported that 20% of added organic carbon is evolved as C0 2 for a 6 month incubation period under similar conditions. In another investigation carried out by Moorhead et al. (1987) it was observed that about 39 and 19% of the total-C for fresh and digested low-N plant biomass, and 50 and 23% of fresh and digested high-N plant biomass are released as C0 2 during 90 days decomposition when these biomasses are added to soils. Fresh plant biomass evolves as much as twice the organic-C as C0 2 when compared to corresponding digested biomass sludges. These results of other researchers suggest that the decomposition rate of organic matter strongly depends on the source and the properties of the available organic matter. Miller (1974) , Sommers et al. (1976) and Terry et al. (1979a, b) have concluded that sludge composition and incubation conditions, rather than soil properties control sludge decomposition. Reddy et al. (1980) have shown that decomposition of organic carbon depends on the nature and constitution of the wastes. Low molecular weight (simple) compounds can be more easily degraded by microorganisms than more complex organic compounds. Organic-C components in decreasing order of biodegradabil ity are: (i) readily oxidizable soluble

PAGE 56

42 organic-C, (ii) proteins, (iii) hemicellulose, (iv) cellulose, and (v) lignin. In the examples mentioned above, fresh plant biomass releases much more C0 2 (especially in the early stages of the incubation) than the digested ones because it contains much more easily decomposable organic-C. In this study, all the composts used were well aged or completely composted. Most of the easily decomposable organic-C such as soluble organic-C, starch and proteins have been decomposed during the composting process. The main organic-C species remaining in the composts studied are the more oxidation resistant residues of the original organic matter (Biddlestone et al., 1987). Also, the four composts studied are either yard waste compost or mixtures of sewage sludge with yard wastes such as wood chips, leaves and tree trimmings, etc.. A high content of cellulose and lignin can be expected in these materials. This feature may explain why the decomposition rates for the composts studied here are relatively low. Decomposition of a complex substrate C is usually described by a multistage first-order decomposition sequence (Reddy et al., 1980; Gilmour et al., 1985). The mathematical rate equations can be written as follows: -dCi/dt = kiCi (3-3) where i refers to a particular stage of decomposition.

PAGE 57

43 The integrated form of equation 3-3 becomes: C ti = CjexpC-kjt) (3-4) where : C^ = organic-C present at the beginning of a decomposition stage. = organic-C present at the end of a decomposition stage at time = t, and k^ = the first-order reaction rate coefficient. Decomposition stages and the corresponding reaction rate coefficients for the four composts are presented graphically in Figure 3-4. The decomposition of compost #3 is described in one stage and the decompositions of compost #2 and #6 are best described in two stages. It can be seen that the reaction rate coefficient values of k]_ and k 2 for these two composts are very similar. This similarity indicates that these two composts have similar organic C composition. The behavior of Compost #1 is markedly different from those of the other composts. During the first 3 days of incubation, C0 2 evolution is rapid followed by a lag period, lasting for the following 40 days. A second period of high decomposition rate was observed between 42 and 70 days. During the remaining period of incubation (after 70 days)', the C0 2 evolution rate for this compost is similar to those for the other composts. The decomposition rate for compost #1 may be described as a 3 stage series of first-order reactions. Within overal experimental error, reaction rate coefficients for the final stage of decomposition for the

PAGE 58

44 U cf 5 -0.04 -0.1 0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 0 -0.01 -0.02 -0.03 -0.04 -0.05 -0.06 -0.07 0 -0.01 -0.02 -0.03 -0.04 -0.05 Compost #1 1^ = 0.00042/day\. k 2 = 0.001 56/day ^^l<3= 0.00057/day "^^^ K, = 0.00069/day + Compost #2 k 2 = 0.00037/day Compost #3 k = 0.00057/day^^^ 1^ = 0.00048/day Compost #6 k 2 = 0.00031 /day 20 40 60 80 100 120 Time (day) Figure 3-4. Decomposition stages and reaction rate coefficients for the four composts studied.

PAGE 59

45 four composts studied are quite similar, falling in the range from 3.1xl0~ 4 to 5 . 7xl0 _4 /day . If the second reaction rate coefficients for the composts studied can be assumed to be representative through the remaining life of the compost, then a rough estimate of the time required for decomposition of 50% of the organic matter (half life) in these composts can be made according to following equation. t 0>5 = 0.693/k (3-5) where : t Q>5 = the half life time of the compost, (day), and k = the first-order reaction coefficient, (1/day) The estimated half life time of the composts tested is from 3.3 to 6.1 years. This estimate is comparable to the result reported by Varanka et al. (1976), who showed that it takes approximately 6 years to lose 50% of the sludge organic C when used in the field. No significant changes in other physical and chemical properties of the composts were observed for the 120 day incubation period used in these studies (Table 3-2) . Effect of H n S on Compost Decomposition The effect of H 2 S exposure on compost decomposition is illustrated in Figure 3-5, where C0 2 evolved by the composts is expressed as mg-C0 2 per g-C of the compost as a function of the H 2 S concentration (ppmv) . The C0 2 evolution is significantly increased with the increasing H 2 S concentration. The rate of this increase is greater at

PAGE 60

46 J 1 1 1 1 i i i i i i i i i i i i 0 4 8 12 16 20 24 28 32 (Thousands) H 2 S Concentration (ppmv) Figure 3-5. Effect of H 2 S exposure on the rate of compost decomposition as measured by C0 2 respiration.

PAGE 61

47 lower H 2 S concentrations. For example, the C0 2 evolved at H 2 S concentrations near 6,000 ppm is approximately 8.5 mgC0 2 /g-C added, which is about 3.4 times that evolved when no H 2 S is present. At higher H 2 S concentrations the increase of C0 2 evolution with H 2 S concentration is reduced e.g. when the H 2 S concentration is increased from 12,000 ppm to 32,000 ppm the C0 2 evolution increases only by about 17%, or approximately 2 mg-C0 2 / g-organic matter. It can be seen from Figure 3-5 that the C0 2 evolution from the compost has a strong dependence on the H 2 S concentration in the gas to which the compost is exposed. A linear relationship is obtained when plotting C0 2 evolved as a function of the sguare root of H 2 S concentration in the gas, [H 2 S] 0,5 for the range of H 2 S concentration less than 17,000 ppmv (Figure 3-6). The regression analysis result for the best fit line is: C0 2 = 2.62 + 0.082[H 2 S] 0,5 (3-6) where : C0 2 = C0 2 evolved from compost, (mg/g of C added) [H 2 S] = H 2 S concentration in the inlet gas stream, (ppmv) The correlation coefficient, R 2 for the variables is 0.9234. Equation 3-6 quantitatively describes the effect of H 2 S on the decomposition of composts. For example, C0 2 evolutions at [H 2 S] = 0 and [H 2 S] = 1000 ppmv are calculated

PAGE 62

48 1 1 1 1 1 1 1 1 1 i i i i i 0 20 40 60 80 100 120 140 160 180 Sqrt [HgS Concentration (ppmv)] Figure 3-6. Plot of C0 2 evolution as a function of square root of H 2 S concentration.

PAGE 63

49 to be 2.62 and 5.21 mg-C0 2 /g-OM, respectively, by equation 3-6. The ratio of these two values is equal to the ratio of the first-order reaction coefficients for the reactions at these two conditions, C0 2 , 1000/ C0 2 , 0 = k 100C)/ k 0 = !' 99 in other words, the decomposition rate for the compost exposed to 1000 ppmv H 2 S is 1.99 times of that for the compost not exposed to H 2 S. The half life times for compost #6 at both conditions are: t 0.5,0 = °693/0. 00031 =6.12 (years), and t 0.5,1000 = 6.12/1.99 = 3.08 (years). and for compost #3 are: t 0.5,0 = 0.693/0.00057 = 3.33 (years), and t 0.5,1000 = 3.33/1.99 = 1.67 (years). No studies of similar effects have been reported in the existing literature. Thus the results obtained in this study can not be compared with the results of other investigations. Taylor et al.(1978) found that the highest S mineralization rates are observed during the period of highest C0 2 evolution when 2 to 6% of sewage sludge compost is incubated in soils. Their results and those reported here suggest that the microbial activity of the compost was significantly enhanced by the addition of H 2 S, especially the activity of the sulfur oxidizing bacteria. Oxidation of inorganic sulfur compounds is a basic phenomenon in nature. A number of bacteria have been

PAGE 64

50 identified in soils and other environments that are capable of oxidizing organic and inorganic sulfur compounds (Roy and Trudinger, 1970) . A high population of the oxidizing bacteria can be expected in the composts tested here. With sufficient H 2 S supply, the bioactivity of the sulfur oxidizing bacteria can be stimulated to result in an increase of microbial population and a corresponding increase in the evolution of C0 2 . Hydrogen sulfide is finally oxidized to sulfate through various pathways and intermediate stages ( Roy and Trudinger, 1970; Brock and Madigan, 1988; Yang and Allen, 1991; Allen and Yang, 1991). After the 24 hours reaction period, the color of the compost changed from originally brown to yellowish-white, especially at high H 2 S concentrations. This feature indicates that a large amount of sulfur has accumulated in the compost. Conclusions Among the various biofilter materials, compost is frequently selected as a medium in applying biof iltration to air pollution control due to its unique properties and advantages. Knowledge of the characteristics of compost decomposition are important for both prediction of biofilter operation characteristics and degradation estimates, as well as in deciding on the appropriate disposal treatment and method for used compost. The studies described here indicate that the decomposition rates of the composts tested are much

PAGE 65

51 lower than those reported by other researchers, who used fresh composts mixed with soils. Five to ten percent of the total-C in composts were decomposed during the 122 days incubation period. Compost half life times of the order 3 to 6 years are estimated for the composts studied, corresponding to loss 50% of their total-C due to decomposition. A multi-stage first-order reaction sequences is used to describe the decompositions. First-order reaction rate coefficients have been determined. Decomposition rates are significantly increased when H 2 S is introduced to the compost. The half life of the compost is significantly reduced as a result of increased biological activity and C0 2 respiration. For example, continuous exposure of compost to 1000 ppm H 2 S can result in reduction of the half life of the compost from about 6 years to 3 years due to enhanced microbiological activity alone. The results suggest that added H 2 S was oxidized by the sulfur oxidizing bacteria in the compost to sulfate.

PAGE 66

CHAPTER 4 DETERMINATION OF THE DESIGN AND OPERATIONAL PARAMETERS FOR BIOFILTER SYSTEMS Introduction Extensive experimental work has been carried out in order to determine the design and operational parameters for a biofilter system. This research is essential for best operation as an air pollution control technology and for optimization of the system. This chapter describes the design and construction of lab scale biofilter systems, experimental methodology used and the results obtained. System Design and Construction Three biofilter systems were designed and constructed for different investigative purposes. Each system can be operated and controlled separately. Detailed information on each experimental system is presented below. The Dual Tower System Most of the experimental work was carried out using a dual-tower experimental biofilter system. This configuration, which is shown in Figure 4-1, consists of parallel dual column filters. The two biofilter columns, identified as Tower #1 and Tower #2, can be run 52

PAGE 67

53 Orifice vent Figure 4-1. Schematic drawing of the dual tower system.

PAGE 68

54 simultaneously and controlled separately. The biofilter bed material is enclosed in transparent rigid Acrylic pipe, with an inner diameter of 0.15 meters (6 inches) and a height of 1.34 meters (4 feet) . Each vertically mounted pipe can be packed with the desired compost up to a height of 1.2 meters (3.9 feet). The packed biofilter material is supported by a sieve plate to ensure a homogeneous distribution of the inlet gas stream across the face of the bed. Nonbiodegradable plastic screens are placed between the sieve plate and the biofilter material to avoid separation of smaller compost particles. Sampling and measurement ports are located along the Acrylic column for compost and gas sampling, and pressure and temperature measurements. The sampling and measurement ports are shown in Figure 4-2a and b for Towers #1 and #2, respectively. An individual sampling/measurement port is identified by a letter-number system, where the letter indicates the function and the number indicates the location of the port. For example, TS11 means this port is used for temperature measurement and solid sampling, and is located on Tower #1 at location 1. All the other filter systems with multi measurement/sampling ports are identified in the same manner. Room air is forced by a Gast Regenair Model R3105-1 air blower into the humidif ication chamber. The blower, which is driven by a 1/2 HP motor, generates a maximum flow of 1.5 m /min (53 cfm) and a maximum pressure/vacuum of

PAGE 69

55 Q15 8 3 P11 T15 TS14 -eTS13 TS12! — 4" TS11 01 3 pio c TioO 3 Q14 012 B — 011 010 8 8 T = Temperature P = Pressure G = Gas Sample S = Solid Sample 8 8 Figure 4-2. Sampling and measurement ports on towers . a. Tower #1

PAGE 70

56 G25 8 P21 -eT26 ! TS25; TS24; TS23 TS22 -eTS21 0T20 3G24 G23 B— G22 P20 C""Q 3 G21 B— G20 T = Temperature P = Pressure G = Gas Sample S = Solid Sample s s Figure 4-2. Sampling and measurement ports on towers . b. Tower #2

PAGE 71

57 1100/1000 mmH 2 0 (43/40 inches) of water column. Humidif ication of the inlet air is achieved by atomizing water in the spray chamber, through which the room air passes. In addition, Pall rings are stacked in the spray chamber for extending wetted surface area providing better humidif ication. As a result relative humidities in the range 95 to 100% were routinely and continuously achieved. Gaseous H 2 S with a purity of 99+%, which is stored in liquid form under pressure in a cylinder, is continuously leaked and mixed with the prehumidif ied air in the inlet lines (PVC pipe) to the towers. Plastic screen packing is placed downstream from the H 2 S introduction point for better mixing. Flow rates of air and H 2 S are controlled by plastic valves, which are located on the carrier gas inlet lines, and stainless steel needle valves, respectively. The flow rates are measured on calibrated flow meters to obtain the desired H 2 S concentration and gas flow through the towers. Measurements of temperature, pressure and gas flow rate are discussed in later sections. A nozzle is installed on the top of Tower #1 in order to introduce water, or other liquid solutions if necessary, to the outlet end of the bed. Gas lines are made from PVC pipes. The towers and pipes are connected by flanges for convenient dismantling of the packed towers and compost changes. Cork-rubber gaskets are used to seal the flanges.

PAGE 72

58 Portable Tower #3 The portable tower is made from PVC pipe with an inner diameter of 77 mm (3 in) . This tower, which is shown in Figure 4-3, has a total height of 1.2 meters (3.94 ft) with an effective packing height of 1 meter (3.28 ft). The two ends of the pipe are covered by rubber caps and held by pipe clamps. Compost packed in the tower is supported by a packing of non-biodegradable plastic screen. Measurement ports for pressure, temperature and exhaust gas samples are located along the length of the tower. Gas to be tested is introduced through a port at the bottom of the tower. The overall gas flow rate is measured by a pre-calibrated flow meter after the effluent gas passes through a particulate filter. This portable tower was used intensively for pressure drop studies and for investigation of long term operation of compost #16. Column System #4 A fourth column biofilter system was constructed for investigation of the effects of various operational variables on H 2 S removal (Figure 4-4) . This multicolumn system includes eight compost columns, a manifold for introduction of test gas and several needle valves for flow control. The columns are made from PVC pipes with an inside diameter (ID) of 35 mm (1.25 in). Each column has a length of 300 mm ( 12 in) and an effective packing height of 250 mm

PAGE 73

59 Gas Outlet s D D D D D 1 0/0 0D TG35 -a -e TGP34 -E TGP33 TGP32 -E TGP31 TG30 1 T Temperature P = Pressure G = Gas Sample P31 P30 30 0 0 0 0 1 0 MOO 8 8 Gas Inlet Figure 4-3. Schematic drawing of Tower #3

PAGE 74

60 Thermometer Gas Outlet Compost Column Gas Inlet 4 m 6 1 1 1 Manifold Inlet Gas Sampling Port Figure 4-4. Schematic drawing of column system #4.

PAGE 75

61 (10 in) , which provide a 240 mL packing volume. The ends of the packed columns are plugged by rubber stoppers. Thermometers are inserted into the columns to measure temperatures. The gas flow rates are measured by a rotameter at the gas outlets. Effluent gas samples are taken from the outlet of the rotameter. Measurement Methods Periodic measurement of temperature, pressure drop and gas flow rate in the biofilter systems are carried out by the following devices. Temperature Temperature is measured by mercury in glass thermometers with a range from -20 °C to 110 °C and a minimum scale division of 1 °C. Pressure Drop Pressure drop is measured by manometers with a minimum reading of 1 millimeter water column (mmH 2 0) . In case the pressure drop is greeter than 1000 mmH 2 0, the pressure drop is measured by mercury manometers with a minimum reading of 1 millimeter of mercury (mmHg) . Gas Flow Rate All the gas flow rates except those of Towers #1 and #2 are measured by pre-cal ibrated rotameters.

PAGE 76

62 The gas flow rates in Towers #1 and #2 are measured by specially designed orifices. Two orifices for each tower were designed and made, one for low flow ranges and the other for high flow ranges. The orifices are made from plastic plate and installed on the outlet gas pipe lines (see Figure 4-1) . The pressure drops across the orifices are measured and the flow rates are calculated according to the developed calibration eguations. Sampling Methods Compost Samples Compost samples in Towers #1 and #2 are taken from the solid sampling ports shown in Figures 4-2 a and b. The samples are taken at each port in a radial direction to the tower walls so that a representative sample can be obtained for that section. For composts not initially packed in columns, the samples are taken after the compost has been thoroughly mixed and very large particles ( diameter > 10 mm) have been eliminated. Gas Samples The inlet and outlet gas samples for each system are obtained directly from the gas sampling ports by extraction using gas-tight syringes. Gas samples extracted from other locations along the towers are obtained by using a gas sampling probe assembly. The gas sampling assembly, as

PAGE 77

63 shown in Figure 4-5, consists of a Teflon probe, a Teflon filter and a sampling port connector. The Teflon probe is made from a piece of Teflon tubing (6.35 mm (1/4") in diameter) , with 14 holes (lmm diameter) spaced evenly along the probe length. The probe is installed in the towers in such a way that all the holes are perpendicular to the tower's normal axis. Thus, representative gas samples from a cross section of the tower can be obtained. The Teflon filter is used to block out any small particles and water droplets which may be extracted during sampling. Gas samples are obtained through the sampling port located on the end of the assembly (see Figure 4-5) by a gas-tight syringe. The sampling port is sealed by a rubber GC septum. When taking a sample, at least three full syringes of gas sample are wasted before the actual sample is taken for analysis. This procedure will eliminate residuals of previous gas samples remaining in the Teflon filter holder, in the syringes and in the probe, as well as condition the extraction system to the gas being sampled. The gas samples are then diluted in 3-L Tedlar sampling bags by pure nitrogen (N 2 ) to an appropriate concentration within the calibration range of the analyzer. The gases in the Tedlar bags are thoroughly mixed by gently kneading the bags and allowing them to sit for at least 10 minutes before analysis. Most of the samples taken are analyzed within 2 hours .

PAGE 78

Figure 4-5. Schematic drawing of the gas sampling assembly.

PAGE 79

65 In most cases, the H 2 S concentrations in the outlet gases are so low that the gas samples can be analyzed directly without any dilution. In the latter cases, Tedlar bags are directly connected to the sampling ports. Gas samples are forced into the bags as a result of the positive pressure of gas within the tower system. Also, Teflon filters are replaced by glass wool plugs to reduce the resistance to flow. Each time after use, the Tedlar bags are purged at least three times with N 2 to eliminate residual gas and vapor. The stability of gas samples in the Tedlar bags are discussed in Chapter 6. Water Samples Water samples analyzed are mainly biofilter wash waters from the tower drain outlets. When washing a packed tower, the entire wash water is collected in a container. Water samples are obtained from the container after mixing the wash water with a stirrer for a few minutes. Compost Analysis Methods Water Content Two to five grams of wet compost are dried in an aluminum tray at an oven temperature of 70 °C until constant weight is obtained. Compost water content is determined by the difference in weight between the wet and dry composts (Robarge and Fernandez, 1986).

PAGE 80

66 EH A known amount of wet compost is weighed into a 50-mL container. DI water is added to bring the liquid/solid ratio to 10 (Robarge and Fernandez, 1986) . The sample is shaken for 3 0 minutes by a rotary shaker. Measurements of pH are made by a calibrated Corning Model M245 pH meter, which is accurate to ± 0.01 pH. Total Carbon and Total Nitrogen Finely-ground, oven-dried compost sample (<100 mesh) are analyzed for total carbon and total nitrogen using a Carlo Erba Model NA 1500 CNS Analyzer. Water Soluble Phosphorus (WSP) A known amount of wet compost (2.5 g dry weight equivalent) is weighed into 50-mL centrifuge tubes. DI water is added to the tubes to obtain a compost to liquid ratio of 1:10 on a dry weight basis. These samples are allowed to agitate for a period of one hour on a mechanical shaker. The compost suspensions are then centrifuged at 6000 rpm for 15 minutes and filtered through Gelman 0.45 micrometer membrane filters. The filtered solutions are acidified (pH<2.0) with one drop of concentrated H 2 S0 4 and stored at 4 °C until analyzed. The soluble reactive P (SRP) in the filtered extract is determined colorimetrically (APHA, 1989) using a Shimadzu UV-160 spectrophotometer with 1 cm path length at 880 nm wavelength.

PAGE 81

67 Acid-extractable Cations Two and half (2.5) g of finely-ground, oven-dried sample is weighed into 50-mL centrifuge tubes. Twenty five (25) mL of 1M HC1 is added and the tubes are shaken for 3 hours on a mechanical shaker. The compost suspensions are centrifuged at 6000 rpm for 15 minutes and filtered through Gelman 0.45 micrometer membrane filters. The solutions are analyzed for Fe, Al , Ca, Mg, Cu and Mn on an Inductively Coupled Argon Plasma Spectrometer (ICAP) (APHA, 1989) . Particle Size Distribution The compost is dried in oven at 70°C for 24 hours. Particle size distribution by weight is measured by passing the dried compost through a series of sieves (U.S.A. Standard Testing Sieve, A.S.T.M. E-ll Specification, Fisher Scientific Company) and weighing the residue. Porosity Compost porosity is determined according to Danielson and Sutherland (1986). Organic Matter After determination of compost water content, the samples are placed in a muffle furnace and baked for 24 hours at 450 °C. Organic matter is determined by the losson-ignition (LOI) (Robarge and Fernandez, 1986).

PAGE 82

68 Particle Density Particle density of the compost is measured according to Blake and Hartge (1986a) . Bulk Density Bulk density of the compost is measured according to Blake and Hartge (1986b) . Sulfur Analysis Methods Intensive and detailed laboratory work has been carried out on the analysis of sulfur compounds in order to obtain a better understanding of the biochemical reactions involved in the H 2 S oxidizing processes occurring in the biofilters. The procedures for determining various sulfur compounds in compost, in water, and in the waste gases are described in this section. Sulfur in Compost The analysis of sulfur in compost includes the determination of total sulfur (total-S) and fractionation of the total sulfur into inorganic and organic constituents. Many wet chemical procedures have been developed to fractionate the total sulfur pool in sediments, soils, and peat into its inorganic and organic constituent compounds. Very little information, however, is available about such analyses for compost. The sulfur analyses conducted in this research include the quantitative determination of acid

PAGE 83

69 volatile sulfur ( sulfide-S) , water soluble sulfur (soluble sulfate) , insoluble sulfate (S0 4 2 ~) , elemental sulfur (S°) , Pyrite sulfur, ester sulfur (organic-S) and total sulfur. Each of the wet chemical procedures involved the reduction of S to H 2 S in a Johnson-Nishita apparatus (Johnson and Nishita, 1952) and trapping the evolved H 2 S in zinc acetate-sodium acetate (ZnAc-NaAc) solutions. Trapped sulfide is quantified by iodometric titration (APHA, 1989) with a 0.025N iodine solution and 0.025N Na 2 S 2 0 3 titrant. The distillation apparatus incorporated slight modifications of that used by Johnson and Nishita and is similar to that used by Wieder and Lang (1985) . Figure 4-6 shows the distillation assembly. The reaction flask is a 250-mL, round-bottom, three-neck flask, with an N 2 inlet via a bleed tube inserted in one neck and the central neck is connected to a condenser. Ultra high purity (>99.999%) nitrogen is used to sweep out the H 2 S and to maintain the reaction flask in a reducing environment. The third neck of the flask is fitted with a stopper to allow introduction of liquid solutions to the flask. The ZnAc-NaAc solution is made by dissolving 50 g of zinc acetate dihydrate [Zn(CH 3 COO) 2 . 2H 2 0] and 12.5 g of sodium acetate trihydrate (CH 3 COONa. 3H 2 0) in 800 mL of DI water and adjusting the final volume to 1 liter (Tabatabai, 1982) . Twenty-five (25) mL of this solution is mixed with 100 mL of DI water and this solution is used to fill two traps used in series. The first trap contains 100 mL and the second one contains 25 mL

PAGE 84

Figure 4-6. Photograph of the sulfur distillation assembly.

PAGE 85

71 of the H 2 S absorption solution. Even though the contents of both traps are analyzed, more than 95% of the H 2 S is consistently recovered in the first of the two series traps. The material to be analyzed is added to the reaction flask through the side neck. The system is purged with N 2 at bubbling rate of 1-2 bubbles per second in the ZnAc-NaAc traps for 10 minutes before the introduction of additional reagents. The materials are boiled for 1 hour, the traps are removed and sulfide titrated. The procedures and methods used for analysis of the total sulfur and various organic and inorganic sulfur compounds in compost are similar to those used for sulfur analyses in peat, soil and sediments described by Zhabina and Volkov (1978), Tabatabai (1982), and Wieder and Lang (1985) . Minor modifications were made for the compost sulfur analyses in this research. The procedures are illustrated in Figure 4-7. All results are expressed as mg sulfur per gram of compost on a dry basis (mg-S/g) . Compost moisture content is determined from a sub-sample by drying the compost at 70 °C to constant weight. Compost samples are stored in plastic bottles and refrigerated at 2-4 °C before analysis. In most cases, however, compost samples are analyzed immediately after sampling.

PAGE 86

72 die tion o o dri np X Ac Re jcing jre Jction rption netric on Redi Mixti Redi H2S lodor Titrat I CO P m o o. e o o u o u 3 T3 CD O o u w •H W >1 0 c (0 u 3 <4-l rH 3 w
PAGE 87

73 The fresh wet compost sample is divided into five subsamples. The latter are used for the analyses of 1) compost moisture content, 2) total-S, 3) water-soluble-S , 4) inorganic and organic sulfur fractions, and 5) storage for later use. Total sulfur Total-S is determined by oxidation (acid digestion) of the various reduced sulfur constituents to sulfate and followed by reduction of the sulfate to H 2 S. The H 2 S is then trapped and titrated as described above. In most of the analyses, 0.5-5.0 g of fresh compost is used, depending on the sulfur content. In some analyses, oven dried compost is used. In the latter case, the compost samples are finely grounded (<40 mesh) and a correspondingly smaller size of compost sample is used. The compost is subjected to acid digestion as described by Tabatabai (1982) . The digest is quantitatively transferred into a 100-mL volumetric flask and the volume is adjusted with IN HC1. Reduction of the sulfate is carried out by a reducing mixture. The reducing mixture contains 50% hypophosphorous acid, 90% formic acid, and hydriodic acid in a 4:2:1 proportion and is prepared as described by Tabatabai (1982). Depending on the sulfur content, 1 to 5 mL aliquot of the digest is transferred into the distillation flask. With aliquots >2 mL, the volume is reduced to about 2 mL by heating the flask on an electric heating mantle. Five (5) mL of the reducing mixture is

PAGE 88

74 added to the flask and the material is subjected to reduction and hydrogen sulfide is liberated, collected and titrated as described above. Water soluble sulfur Water-soluble-S is determined by shaking 2-5 g of fresh compost in DI water with a liguid to solid ratio of 10:1 for 30 minutes on a rotary shaker at a speed of 140 /min. An aliguot of the compost extract is then subjected to reduction, H 2 S absorption, and titration successively, as described previously. The following analyses are conducted in succession: Sulfide sulfur Sulfide-S or acid-volatile sulfur (AVS) is determined by introducing 8 mL of 12N HC1 to the compost sample in the reaction flask. Heat is applied after 10 min, the materials are brought to boiling, and after 45 min the traps are removed and the sulfide titrated. Sulfate sulfur The content of the reaction flask is filtered by a #42 Whatman filter. The filtrate is then subjected to reduction. H 2 S is then trapped and titrated. This is an alternative way of carrying out sulfate analysis. The results are comparable to the summation of water soluble-S and P-extractable-S.

PAGE 89

75 Elemental sulfur The precipitate obtained above is dried with filter papers as described by Zhabina and Volkov (1978) and is extracted with analytical grade acetone. The volume of acetone used (mL) is 40 times that of the equivalent dry weight of the compost sample (g) . The extraction flask is covered with Parafilm and placed on a rotary shaker at a speed of 140/min for 16 hours. The mixture is filtered and rinsed with additional acetone. Either a fraction of or the entire filtrate are subjected to Cr 2+ reduction as described by Zhabina and Volkov (1978). Pvrite sulfur The residue left after S° extraction is subjected to chromium reduction. The Cr 2+ was produced by passing a 1 M solution of CrCl 3 .6H 2 0 in 0.5 M HC1 through a Jones reductor column containing Zn amalgamated with Hg (Zhabina and Volkov, 1978) . Preparation of the Jones reductor is described by Swift (1950) and Patterson and Thomas (1952) . Ten (10) mL of ethanol is added to the flask followed by 20 mL of 12N HC1 and 16 mL of 1M CrCl 2 solution. Heat is applied after 30 min, the H 2 S evolved is absorbed in the traps and titrated as above.

PAGE 90

76 Organic sulfur Following the previous procedure, the insoluble residue in the reaction flask is filtered through a #42 Whatman filter and is washed repeatedly with acidified DI water to remove chromium ions. The residue is then subjected to reduction with the reducing mixture as described above. The organic sulfur determined this way is mainly ester-S. In addition to the literature mentioned above, similar and dissimilar procedures for determining sulfur constituents in sediments, peat and soil are also described by Smittenberg et al. (1951), Freney (1958) , Johnson and Henderson (1979), David et al. (1982), and Hsieh and Yang (1989) . An excellent comparison of some of these methods is reported by Wieder and Lang (1985) . Sulfur in the Aqueous Phase Sulfur in water samples, such as in tower wash water and drainage, are analyzed for sulfide, sulfate, and/or total sulfur. Sulfate sulfur Sulfate in most of the water samples is determined by a turbidimetric method (APHA, 1989). A Milton Roy Model Spectronic 21 Spectrometer was used at 420 nm to measure the turbidity. Color or suspended matter in large amounts will interfere with this method. In the case of dark colored

PAGE 91

77 water samples, the reduction, H 2 S absorption and iodometric titration procedures described previously are used. Sulfide sulfur Sulfide was measured according to APHA (1989). The samples are pretreated to remove interfering substances and to separate insoluble sulfide. Total-S Total-S was determined by using 2-5 mL of the agueous sample depends on its sulfur content. The sample is then analyzed as described for the total sulfur in compost. Sulfur in Waste Gas Gas samples are analyzed by a commercial gas chromatogragh eguipped with a flame photometric detector (GC/FPD) , a Tracor Model 250H Analyzer. The detection limits and operational conditions for the analyzer are summarized in Table 4-1. Detailed information concerning the principles and conditions of operation are reported elsewhere (Yang, 1988) . In the early stages of this study, the component peaks are recorded by a chart recorder (Texas Instruments, Inc., Model Recti/Riter II) and concentrations of sulfur compounds are determined by measuring the peak heights. Later in the study (from August 1990 to September 1991) the chart recorder was replaced by a Spectra-Physics model SP4290 integrator and the measured concentrations are read

PAGE 92

78 Table 4-1. Retention times, limits of detection and operating conditions for the Tracor 250H Analyzer. Compounds 3 Item H 2 S MM DMS Retention Detection Time (min) Limit (ppmv) 1.15 0.01 2.16 0.01 4.02 0.02 Operating Conditions: Temperatures : 50 °C Valve: Column: Detector: 110 70 °C Flow Rates: Cylinder Pressures: Nitrogen: 80 mL/min (carrier gas) Oxygen: 21 mL/min (flame gas) Hydrogen: 80 mL/min (flame gas) Sample: 40 mL/min Air 40 psig ( for sampling valve activation) Hydrogen: 53 psig Oxygen: 40 psig Nitrogen: 80 psig a MM: methyl mercaptan. DMS: dimethyl sulfide.

PAGE 93

directly from the printout of the integrator. The Tracor analyzer is periodically calibrated for H 2 S, methyl mercaptan (MM) and dimethyl sulfide (DMS) with standards purchased from National Speciality Gases, Inc. Results and Discussion Biof iltration is a process that involves physical, chemical, and biological processes. Many variables, such as temperature, compost water content, specific acidity of compost, sulfate content in the compost etc. , can affect the function of the system. It is impossible to obtain optimum, longer term performance from a biof iltration system without an in depth understanding of the system properties and proper control of important variables. Extensive evaluations of the system properties have been conducted during the course of this study. The results presented in this section are divided into several subsections according to specific investigations undertaken. Composts from different sources (Table 3-1) have been used in the laboratory studies. The physical and chemical properties of these composts are summarized in Table 4-2. Applications of each of these composts are mentioned in the corresponding study subsections. Pressure Drop The energy consumption obtained in operating a biof iltration system is primarily that required by the

PAGE 94

80 Table 4-2. Summary of initial compost properties. Compost ID # Property 12 13 14 16 17 pH 2 . 61 1. 60 6.44 6. 66 8.10 Bulk Density (g/cc) 0.30 0.27 0.18 0.22 0.20 Particle Den. (g/cc) 1.73 1.75 1.90 1.78 Porosity (%) 84.3 89.7 88.4 88.7 Water Content (wt%) 45.6 54.9 62.4 56.5 62.7 Organic Matter (wt%) 66.5 66.8 59.3 64.3 62.6 Total-S (mg-S/g) 44.8 70.4 0.74 Water-P (mg/kg) 140 223 152 114 167 Total-C (wt%) 34.3 31.0 31.3 30.5 40.9 Total-N (wt%) 1.89 3.24 1.75 4.27 1.30 C/N 18. 1 9.57 17.9 7.14 31.5 Metals (mg/kg) Ca 47400 28900 145000 18000 2670 Mg 280 255 4880 1450 120 Zn 38.0 48.0 201 66.5 18.0 Cu 121 91.0 60.0 11.0 93.5 Mn 5.50 55.0 96.5 89.5 9.50 Fe 1900 805 6160 529 1510

PAGE 95

81 blower (fans) to move contaminated air at the specified flow rate through the filter bed. However, the pressure drop across the filter bed increases markedly as the flow rate is increased. Since the pressure drop will be determined by the depth of the filter bed it is necessary that the gas velocity should be kept as low as possible. The pressure drop across a compost bed filter can be lowered by physical treatment of the compost particle content. Such a procedure is an important reguirement in optimizing the operation of a filter bed, because operation at a reduced resistance to flow allows the gas velocity as well as the volumetric flow rate to be significantly increased with little or no change in energy consumption. This will in turn increase the biofilter capacity and reduce the reguired filter size. The following physical factors determine pressure drop across the filter bed: 1) Particle size distribution in the compost 2) Condition of the filter packing 3) Height of the filter bed 4) Water content of the compost 5) Gas velocity, and 6) Porosity of the compost. A representative sample of compost #12 was air dried and the particle size distribution determined by weighing the fractions penetrating a series of standard sieves. The particles are classified into 5 size groups in the range >12 mm to <1.2 mm (see Table 4-3). Compost samples in each of

PAGE 96

82 Table 4-3. Particle size range distributions for selected composts. Particle Size Range Distribution (wt%) Compost ID # A B C D E 12 20.0 22.5 14 27.7 26.9 13 21.4 24.5 17 0.00 33.4 10.0 13.1 34.4 8.10 11.6 25.7 6.70 15.8 31.6 14.4 22.5 29.7 A: diam. > 12 mm B: 3.35 < diam. <12 mm C: 2.36 < diam. <3.35 mm D: 1.18 < diam. <2.36 mm E: diam. <1.18 mm

PAGE 97

83 the particle size groups were analyzed for water content then wetted to obtain a water content of about 50%, by spraying and mixing water with the compost samples. These treated samples were then used to determine the pressure drop as a function of the 5 particle size range classes by adjusting air velocities in the range 0.02 m/s to 0.28 m/s. All tests were carried out with the same compost bed height (1 m) and water content (50%) . The results of these studies are presented in Figure 48, where it is seen that the pressure drop increases significantly with increasing gas velocity for a bed of small particle size (<1.2mm). For particles greater than 1.2 mm the pressure drop increases to a much lesser extent with increasing velocity as shown by a comparison of data for particle classes D and E. Clearly significant pressure drops observed in operating filter beds are a result of the presence of small particles with sizes less than 1 mm. For example, at a representative gas velocity of 0.03 m/s, which is equivalent to a loading rate of 110 m 3 /m 2 -hr, the pressure drop realized by a 1-m bed of particles size classified as <1.2 mm is 390 mm H 2 0, whereas the pressure drop obtained for the same packing height and the same gas velocity, for particles classified as >12 mm is only 2 mm H 2 0. Thus, under these conditions the pressure drop created by 1 mm or less particles is about 200 times that caused by an equivalent bed composed of 12 mm or greater particles.

PAGE 98

84 3.5 Particle Size Range A: d > 1 2 mm B: 3.35 < d < 12 mm C: 2.36 < d < 3.35 mm D: 1.18 < d < 2.36 mm E: d < 1.18 mm Figure 4-8. Pressure drop as a function of particle size range for different gas velocities .

PAGE 99

85 The relationships between pressure drop and packing height at constant gas velocity for particle size ranges B (3.33 to 12 mm) and E (<1.2 mm) , and the parent compost #12 are shown in Figure 4-9. It is seen that the pressure drop increases approximately linearly with packing height for all three samples. It is important to note that the pressure drop values obtained for the parent compost are intermediate between the strong dependence of pressure drop on packing height for small particles (class E) and the weak dependence on packing height for larger particles (class B) . The dependence of pressure drop on compost water content is not as consistent as that for particle size. Qualitatively, sewage sludge treated compost contains more viscous and adhesive small particles than are found in untreated compost. Thus, when the water content of the sewage sludge treated compost is increased, coagulation of small particles is enhanced and the pressure drop increases sharply. However, the build-up in pressure may be suddenly released by channeling, i.e., a breakdown of the overall flow restriction by the formation of a channel of much less resistance caused by a separation of packed materials. The effects of gas velocity on pressure drop across four typical biofilter materials are shown in Figure 4-10. As expected, the pressure drop increases rapidly with increasing gas velocity. The pressure drop depends on the way the filter is packed. Composts #12 and #13 are similar in nature but filter material #13 was more compacted than

PAGE 100

86 0 0.2 0.4 0.6 0.8 1 Compost Packing Height (m) Figure 4-9. Pressure drop as a function of packing height for different compost particle size ranges.

PAGE 101

87 900 800 700 600 500 400 300 200 100 0 100 Gas Loading Rate (m 3 /m 2 -hr) 200 300 400 500 600 700 [ I I I I I I I x Compost #13 II II i I Compost #12 o Compost #17A o Compost #14 / /x O C^IT 1 I I 1 1 ! I 1 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 Gas Velocity (m/s) Figure 4-10. Pressure drop as a function of gas velocity for different types of compost .

PAGE 102

88 #12, and the more densely packed material shows a much higher pressure drop for the same gas velocity. Composts #14 and #17A have similar pressure drop-gas velocity curves and although material #17A contains a larger fraction of small particles than does material #14 (see Table 4-3) , the small particles in compost #17A are mainly sand and grass fractions, whereas those in compost #14 are sewage sludge particles, which tend to adhere to each other. Effect of Gas Retention Time on FUS Removal The effect of gas retention time on H 2 S reduction is studied by varying the gas flow rate through the tower. The results are presented in Table 4-4. In the first data set of 6 tests, the H 2 S loading rate is kept approximately constant at a low flow range to ensure that the maximum H 2 S elimination capacity of the system is not exceeded during the test. It is clear that there is no apparent effect on H 2 S removal as long as retention times are longer than about 23 sec. When the retention time is reduced to 7 seconds, the H 2 S removal efficiency decreases by about 6%. This decrease in H 2 S is controlled by the macrokinetics of biof iltration process. Sublette and Sylvester (1987) reported that H 2 S can be metabolized by a pure culture of T. denitrif icans within 1-2 seconds. This suggests that the reduction of H 2 S removal efficiencies under shorter residence times is not necessarily due to insufficient reaction time between the H 2 S molecules and the biomass, but

PAGE 103

89 Table 4-4. Effect of gas retention time on H 2 S removal efficiency. H 2 S Gas Flow Reten. Loading HjS H 2 S Removal Rate Time Rate Inlet Outlet Eff. (Lpm) (s) (g-S/m 3 -hr) (ppmv) (ppmv) (%) 151 106 75.5 46.0 30.8 15.3 15.0 15.0 9.00 9.00 5.40 9.98 14.1 23.1 34.5 70.0 71.0 71.0 118 118 197 16.3 17.1 17.7 21.0 20.7 19.8 39.8 50.7 56.3 73.2 62.3 24.9 37.0 53 .7 105 155 297 610 776 1440 1870 2650 1.62 1.30 0.77 0.30 0.02 0.01 BDL 2.77 0.01 4.24 4.58 93.5 96.5 98.6 99.7 99.9+ 99.9+ 99.9+ 99.6 99.9+ 99.8 99.8

PAGE 104

90 is possibly due to the slower step of H 2 S diffusion from the gas phase into the liquid phase. In the second data set of 5 tests presented in Table 44, the biofilter inlet gas contains very high H 2 S concentrations and the system is operated at lower flow rates and, therefore, longer retention times. These tests show that, even when the inlet gas contains an H 2 S concentration of 2650 ppmv, the biofilter can successfully reduce the concentration to 4.6 ppmv with a 99.8% removal efficiency when the retention time of the flowing gas is increased to 197 seconds. Thus, it can be seen that, as long as the H 2 S loading rate does not exceed the maximum H 2 S elimination capacity of the system (discussed in a later section) , then the design engineer or operator can always deal with high H 2 S concentrations in the waste gas by decreasing the gas flow rate to obtain the desired high level of control efficiency. Effect of Concentration of H 2 S on Its Removal The effect of H 2 S concentration in the inlet gas on the H 2 S removal efficiency has been investigated under constant gas flow rate conditions. The results of these studies are presented in Table 4-5. No significant difference in control efficiencies is observed when H 2 S concentrations in the influent gas stream are varied from 5.5 ppm to 518 ppm as long as the H 2 S loading rate is less than the maximum acceptable value for the compost studied.

PAGE 105

91 Table 4-5. Effect of H 2 S concentration on removal efficiency. H 2 S Gas Flow Reten. Loading H 2 S H 2 S Outlet Removal Rate Time Rate Inlet Eff . (Lpm) (s) (g-S/m 3 -hr) (ppmv) (ppmv) (%) 30.0 35.4 0.72 5.51 BDL 99.9+ 30.4 35.0 1.69 12.8 BDL 99.9+ 30.8 34.5 16.9 126 0.01 99.9+ 30.8 34.5 26.3 196 1.31 99.3 30.0 35.4 29.6 227 0.01 99.9+ 30.0 35.4 38.1 292 0.01 99.9+ 31.2 34.0 70.4 518 1.04 99.8

PAGE 106

92 Effect of HoS Loading Rate on Its Removal One of the most important observations made in this study is the relationship between H 2 S reduction and its loading rate to the biofilter. It is very important and necessary to introduce here the concept of H 2 S loading rate and the maximum elimination capacity of the filter material. The H 2 S loading rate is the amount of H 2 S that is being introduced to the system per unit volume of the packing material per unit time (g-S/m 3 -hr) . The maximum elimination capacity of a compost is the maximum H 2 S loading rate that the compost can bear without inhibiting its microbial activity, and is expressed in the same units as those used for H 2 S loading rate. These two parameters probably play central roles in biofilter design and system operation. The maximum H 2 S elimination capacity of a compost depends on the microbial population and activity of sulfur oxidizing bacteria existing in the compost. The latter, in turn, are related to the operating conditions of the system, such as temperature, water content, acidity (pH) of compost, and the concentrations of nutrients and inhibitory substances. Overloading of the biofilter system with H 2 S is indicated by the appearance of a finely divided, yellowishwhite colored substance on the compost, a sudden decrease in the H 2 S removal efficiency or the occurrence of higher concentrations of elemental sulfur in the compost.

PAGE 107

93 The maximum H 2 S elimination capacity for a compost is determined at the optimum operating conditions of the system. H 2 S concentrations in the inlet gas stream are varied increasingly at a constant gas flow rate. The H 2 S removal rates (g-S/m 3 -hr) are plotted v.s. the H 2 S loading rates (g-S/m 3 -hr) . The maximum H 2 S elimination capacity of the compost is determined when the curve flattens out (Ottengraf , 1986) . The maximum H 2 S elimination capacity of compost #13 is determined to be 11.5 g-S/ m 3 -hr (Figure 411) . The maximum H 2 S elimination capacity of this compost is very low because its extremely low pH (1.60) and high sulfate content (70.4 mg-S/g, see Table 4-2). The maximum H 2 S elimination capacity for a compost can also be determined through kinetics studies. Detailed information is presented in the section of "Kinetics of H 2 S Oxidation in a Biofilter". The maximum H 2 S elimination capacity for compost #17 is determined to be 129 g-S/m 3 -hr. Effect of Compost Water Content on H 2 S Removal The effect of compost water content (CWC) on H 2 S removal was evaluated using the Column System #4. The compost (Compost #17) which was used for this study and other studies carried out using Column System #4 had been previously packed in Tower #2 and was considered to be preconditioned by H 2 S to a stable operating condition. After transfer to System #4, the compost was operated at the same conditions as used in its parent environment, Tower #2,

PAGE 108

94 Figure 4-11. Determination of maximum H 2 S elimination capacity of compost.

PAGE 109

95 before changing the operational parameters. No change in H 2 S removal capacity was found due to the system transfer. Data were not taken until system #4 reached a stable condition after any changes in the operational parameters or conditions were made. Each column in System #4 was packed with equal weights of compost. The range of compost water contents evaluated was from 0%, oven dried compost, to about 62%, the maximum water holding capacity of the compost. Water content of the compost in each column was adjusted to a desired range by either adding DI water to the compost or by gently drying the compost in room air. One of the composts tested was thoroughly dried in an oven at 110 °C for 24 hours to obtain water free compost. The system was operated at room temperature with a gas loading rate of 3 o about 15 m-yni ~hr. Inlet H 2 S concentrations were controlled in a range between 80 and 110 ppmv. The results are illustrated in Figure 4-12. The H 2 S removal efficiency was maintained at a high value, 99.9+%, with little variation being observed when the CWC was varied from 30% to 62%. When the CWC was reduced below 30%, the H 2 S removal efficiency decreased linearly with the CWC. Very little removal of H 2 S was observed for the oven dried compost. The residual effeciency of the latter is probably due only to chemical oxidation and adsorption of H 2 S on the compost. Water is essential for all living organisms. All biological metabolic processes require water as a medium or

PAGE 110

96 Figure 4-12. Effect of compost water content on H 2 S removal efficiency.

PAGE 111

97 solvent. Insufficient water supply can limit the activity of the microorganisms, which in turn reduces the H 2 S oxidation rate. It is also possible that when the CWC reduced below 3 0%, there is no free water existing in the pours of the compost particles. This may decrease the rate of transfer of H 2 S from the waste gas to the biofilm where the biological oxidation of H 2 S takes place. Biological activity of the "dry" compost can be recovered if water is supplied to the compost to bring the CWC to a proper range. Two composts, WCF4 and WCE3 with original CWCs of 14.3% and 21.4%, respectively, were used to study this phenomenon. DI water was added to bring the water contents of composts WCF4 and WCE3 to 56.5% and 50%, respectively. The biological activity of both composts as indicated by H 2 S removal was recovered eventually up to 99.9+%. The time required for recovery of the activity, however, is inversely proportional to the dryness of the "dry" compost. As shown in Figure 4-13, it takes 63 hours for compost WCF4 to recover its H 2 S removal efficiency to 99.9+% while compost WCE3 took 41 hours to reach the same level even though both systems followed a similar recovery pattern. Effect of Compost Acidity on H 2 S Removal The effect of compost acidity on H 2 S removal was investigated using the Column System #4. The source of the compost is the same as described in the previous section.

PAGE 112

98 Figure 4-13. Time required for dried compost to recover optimum efficiency. Original compost water content: WCE3 = 21.4%, WCF4 = 14.3%.

PAGE 113

99 The compost used in each column was treated with either dilute HC1 or dilute NaOH solutions to bring the pH of the compost to a desired range. The final pH values obtained for the composts studied in each column are 1.57, 3.20, 4.42, 5.02, 6.39, 6.75, and 8.76, respectively. Measurements of inlet and outlet gas sample concentrations were made when the bed operation became stable. The results of these studies are shown in Figure 4-14. Repeated measurements at each pH point are used to estimate the mean value, and mean ± 2 standard deviations values (error bars) indicated by the upper and lower bounds of the vertical line through the mean. Two operational conditions are used in this test. Under condition A, lower H 2 S and gas loading rates are used. No significant effect of pH of compost on H 2 S removal is observed for pH values in the range between 3.2 and 8.76. Removal efficiencies of 99.5+% are consistently determined with little or no variation in this range. When the pH was reduced to 1.57, the H 2 S removal efficiency fell sharply to about 9%. It should be noted that the measured H 2 S removal efficiencies showed larger variations at the lower pH range. Following the previous studies, the operational conditions were changed to condition B. Under the latter conditions, higher H 2 S and gas loading rates were used and the effects of varing pH the compost becomes evident. It can be seen in Figure 4-14, condition B, that the maximum H 2 S removal occurred at a compost pH value of 3.2 (99.2%). The removal efficiency

PAGE 114

100 100 90 80 70 60 30 0 -10 -20 a—* 1— — — a ~~ — a Condition A I 1 I 1 [ I 1 I 1 100 90 80 70 60 50 40 30 20 10 0 o Condition B J u J I I L_ Compost pH Figure 4-14. Effect of compost pH on H 2 S removal efficiency. Condition a: 10.5a/m 3 -hr H 2 S loading rate: Gas loading rate: Condition b: H 2 S loading rate: 15 m 3 /m 2 -hr 35.4 g/m 3 -hr Gas loading rate: 26.1 m J /nr-hr

PAGE 115

101 decreased with decreasing acidity up to pH 5 (82.3%), and then increased with further decrease in acidity of the compost. As discussed previously in Chapter 2, sulfur oxidizing bacteria can live in environments with a wide pH range (1-8) depending on the species present. Probably the dominant species present in the biofilter systems studied are acidophiles which prefer an optimum pH value around 3. For the higher pH value range, chemical reaction between H 2 S and the compost material can significantly enhance its removal in addition to biological oxidation. As a result of this dual action, higher removal efficiencies of H 2 S can be expected. The acidity of the compost is traditionally expressed as 'compost pH 1 for convenience. A more meaningful and accurate way to express the compost acidity is the 'specific acidity of compost (SAC)', which gives the quantity of H + per unit weight of dry compost, nq-E + /q. Specific acidity of compost can be calculated from compost pH by the following equation: SAC = W/[C W (100-CWC) ]X10 5 "P H (4-1) wh e r e : SAC = specific acidity of compost, /xg-H + /9~dry compost W = DI water used for compost pH analysis, mL = Weight of wet compost used for pH analysis, g CWC = Compost water content, Wt% pH = compost pH. For example, if 2g of wet compost is taken and 20 mL of DI water is used in H + extraction for a compost pH analysis,

PAGE 116

102 and the compost water content is 60% by weight, providing a compost pH of 1.57, then the dry specific acidity of this compost is: SAC = 20/[2 (100-60) jXlO 5 " 1 * 57 = 673 (nq-H + /q) . Effect of Temperature on H 2 S Removal The effect of temperature on H 2 S removal efficiency was studied using Column System #4 with some minor modifications. The modified system is shown in Figure 4-15. Three columns (triplicate) packed with the same compost were placed in a heating box. A rheostat was used to control the temperature of the columns. The influent gas stream was blown through a bubbler, which is placed in a water bath and heated up to the same or slightly higher (5°C) temperature as the biofilter columns to saturate the gas stream at the desired temperature. For tests carried out below room temperature (2 2°C) , the reaction columns were placed in a refrigerator and the temperature adjusted through the refrigerator thermostat. The temperature range investigated is between -1.5 to 103°C. The results of the triplicate measurements for each temperature point are plotted means and ± 2 standard deviations as error bars in Figure 4-16. In the range from 25°C to 45°C, high H 2 S removal efficiencies are consistently observed with little variation. The H 2 S removal efficiency, however, dropped rapidly with decreasing temperature in the lower temperature

PAGE 117

103 Thermometer Gas OutletColumn — Compost Rheostat 1 ! c Inlet Gas Sampling Port 1 %\ t\ M Gas Inlet Bubbler o o Water Bath Figure 4-15. Schematic drawing of the experimental arrangement for investigation of the effect of temperature on H 2 S removal efficiency.

PAGE 118

104 Figure 4-16. Effect of temperature on H 2 S removal efficiency.

PAGE 119

105 range. For instance, when the temperature was reduced from 25°C to 7.5°C, the removal efficiency decreased by approximately 80%. At 7.5°C, only 20% removal efficiency was observed. On the other hand, the decrease in H 2 S removal in the higher temperature range was less marked than observed at lower temperatures. For example, when the temperature was increased from 50°C to 100°C, the H 2 S removal efficiency decreased from 97.4% to 40%. The optimum temperature range determined from these studies is between 30°C and 40°C, which is the optimum temperature range for mesophilic bacteria. The removal rates of H 2 S at high temperature is probably due to chemical oxidation reactions in addition to biological oxidation. Poor performance of a biofilter at low temperatures may limit their application in cold climates, especially during the winter. Proper means should be taken to avoid operating biofilter systems below 10°C. For larger biofilters, the bed temperature can be a few degrees higher than the ambient air temperature due to biological respiration of the microbes and the exothermal oxidation reactions in the filter. Kampbell et al. (1987) reported that soil bed biofilter functioned well at temperatures in the range 12°C to 24°C in Wisconsin. In another study carried out by Rands et al. (1981) the filter bed temperatures were found to be 10 to 20°C higher than ambient air temperatures during winter times. This type of thermal enhancement, however, was not observed during this study, probably because the

PAGE 120

106 system is not large enough to maintain an adiabatic condition. Effect of Sulfate on H 2 S Removal The effect of sulfate presence on H 2 S removal was evaluated using Column System #4. A preconditioned stable compost from Tower #2 was divided into 8 equal-weight portions. Two of the sub-composts with an original sulfate content of 4.6 mg-S/g were packed directly into two columns as controls. The other sub-compost portions were mixed with sodium sulfate (Na 2 S0 4 ) to bring the compost sulfate content to 24.6, 44.6, 64.6, 84.6, 104.6 and 204.6 mg-S/g, respectively, and packed in the other 6 columns. After a few days for acclimation, H 2 S removal efficiencies for each column were determined. The system was operated at a gas loading rate of 15 m 3 /m 2 -hr and an H 2 S loading rate between 6.6 and 8.4 g-S/m 3 -hr. The results are illustrated in Figure 4-17. No effect is observed when the compost sulfate content is less than 25 mg-S/g. However, a significant effect is observed at higher sulfate levels. The H 2 S removal efficiencies were reduced from 99.94% to about 35% and remained in the lower removal efficiency range when the sulfate content was increased from 45 to 200 mg-S/g. These results suggest that a sulfate content of 25 mg-S/g is a critical level for the microbial environment. Above this level sulfate probably reaches a toxic level and the

PAGE 121

107 20 40 60 80 100 120 140 160 180 200 Sulfate in Compost (mg-S/g) Figure 4-17. Effect of sulfate on H 2 S removal efficiency.

PAGE 122

108 activity of these microorganisms is markedly inhibited. This observation is very important for biof iltration control of H 2 S. Since sulfate is the final product of the biof iltration process, it may accumulate in the biofilter bed if no other action is taken. Accumulation of sulfate can easily reach a level that can significantly reduce the function of the biofilter. Measures to avoid sulfate accumulation in filter and to enable recovery of the deteriorated compost are discussed in Chapter 5. Effect of Nutrient Addition on H 2 S Removal Low and high sulfur-containing composts were used for investigation of the effects of nutrient addition on H 2 S removal. The total-S contents are 17.5, 33.7, 20.2 and 119.7 mg-S/g for compost A, B, C, and D, respectively. Each compost was packed in two columns, one is used as a control, and the other was treated with nutrients. Fifty (50) mL of nutrient solution was mixed with 140 g of the compost tested in each column. Excess water was removed by exposing the compost to room air for 24 hours. The nutrient added is similar to the enrichment medium for sulfur-oxidizing bacteria suggested by Aaronson (1970) . The composition of the solution is: K 2 HP0 4 , 1.0 g; MgS0 4 .7H 2 0, 0.5 g; NH 4 N0 3 , 1.0 g; CaC0 3 , 10 g, and DI water was added to bring the final volume to 1000 mL. The results on the effect of nutrient addition on biofilter performance are presented as a bar graph in Figure

PAGE 123

109 4-18. For each compost set, H 2 S removal efficiencies are plotted for the control compost and the test compost before and after the addition of nutrient solution. It can be seen that the H 2 S removal efficiencies are significantly decreased for the three composts tested with low sulfur content when the nutrients are added. The reason for this decline in efficiency is not clear. No improvement of H 2 S removal for the high sulfur containing compost was observed, either. Kinetics of H 2 S Oxidation in a Biofilter Theoretical Considerations In general, the substrate utilization rate of a component by microbial flora as well as the enzymatic reaction rate are expressed by the Michael is-Menten relationship (White et al., 1978; McGilvery and Goldstein, 1983; Schmidt, et al., 1985): -dC/dt = V^CB/d^ + C) (4-2) where C is the substrate concentration, B is the population density, V max is the theoretical maximum specific reaction rate, K m is the hal f -saturation constant (Michaelis constant) , and t is the reaction time. Under steady state conditions, i.e., when the microbial population does not change with time, there are three situations which may be encountered in a biological reaction system, and corresponding eguations can be derived, from

PAGE 124

110 Figure 4-18. Effect of nutrient addition on H 2 S removal . Total-S content in compost (mg-S/g) A: 17.5 B: 33.7 C: 20.2 D: 120

PAGE 125

Ill equation 4-2 above, to express the rates of biological reaction for each particular situation: 1) If the substrate concentration is very high, i.e., when K m << C, the rate expression approaches zero-order kinetics in the substrate concentration: -dC/dt = V max B = k Q (4-3) where k Q is a zero-order rate coefficient. The integral form of equation 4-3 becomes: c " c o " V (4-4) where C Q and C are the initial substrate concentration and the substrate concentration at time t, respectively. If C Q C is plotted against t, a straight line should be obtained and the slope k Q is the maximum elimination capacity of the microbes for the substrate. 2) If the substrate concentration is very low, i.e., when C << Kjjj, a first-order kinetics dependence should be obtained: -dC/dt = kjC (4-5) The integral form for equation 4-5 becomes: C = C Q exp(-k x t) (4-6) where kj^ is the first-order reaction rate coefficient. If ln(C/C Q ) is plotted against t, a straight line should be obtained and the first-order reaction coefficient,

PAGE 126

112 ki, can be determined from the slope of the line. 3) In the third situation, when the half-saturation constant, K m , and the substrate concentration, C, are comparable, the biological reaction should follow fractional order kinetics, because equation 4-2 can not be simplified. Relatively complex equations have to be derived to express the fractional order kinetics. A number of theoretical and empirical models have been reported which describe the kinetics of biodegradation of organic compounds and reduced sulfur species (Chen and Morris, 1972; Williamson, 1973; Cooper, 1974; Jennings et al., 1976; Williamson and McCarty, 1976a, b; O'Brien and Birkner, 1977; Rittmann et al., 1978; Schmidt et al., 1985; Kampbel et al., 1987; Caunt and Hester, 1989). In particular, Ottengraf and coworkers have published a series of papers that describe the processes involved in the operation of biofilters (Ottengraf and Van Den Oever, 1983; Ottengraf et al., 1984; Ottengraf, 1986; Ottengraf et al., 1986; Ottengraf, 1987). Ottengraf developed a biophysical model as well as derived mathematical solutions to describe the kinetics of biodegradation of various organic compounds in biofilter systems. As described previously in chapter 2, Ottengraf s model uses the concept of a bed consisting of solid filter particles, where each particle is surrounded by a wet, biologically active layer. When waste air flows around the particle there is continuous mass transfer of pollutant from

PAGE 127

113 the gas phase to the biolayer (See Figure 2-3) . The overall kinetic behavior observed in a biofilter is a result of the interaction between mass transfer phenomena, the microkinetics of the biological elimination reactions, the residence time distribution of the gas flow, etc.. This overall kinetic behavior is termed 1 macrokinetics 1 by Ottengraf and can be determined experimentally. Ottengraf divided the macrokinetics in a biofilter into two classes: first-order reaction and zero-order reaction. He expressed the first-order reaction kinetics in a form similar to eguation 4-4 described previously. For zeroorder reaction kinetics, however, Ottengraf has distinguished the following two situations: 1. At gas phase concentrations, C, above a compound specific, critical concentration (C cr ^*.) ' the fil m will be fully saturated (Figure 2-3, Case 1) and pollutant elimination is limited by the biological activity in the biofilm. This process is defined as reaction limitation. 2. At concentrations less than C cr ^ t , diffusion in the biofilm will limit compound removal. The biofilm is no longer fully penetrated (Figure 2-3, Case 2) and the removal rate decreases with decreasing pollutant concentration in the waste gas. This process is referred to as diffusion limitation. Ottengraf has also derived two eguations which describe the kinetics for either situation. For the first situation, i.e. under reaction limiting conditions, the kinetics expression is: C/C Q = 1 K Q H/C 0 U g (4-7)

PAGE 128

114 where C is the effluent gas concentration, C Q is the influent gas concentration, H is the height of the filter bed and Ug is the gas velocity. This eguation is a different way of expressing equation 4-4, specifically for a biof ilter. Under diffusion limiting conditions, the kinetics expression is: C/C Q = [1 (H/U g ) (K 0 D e a/2mC 0 6)^] 2 (4-8) where a is the interfacial area per unit volume, D e is the effective diffusion coefficient, m is the distribution coefficient of the component, and S is the biolayer thickness. Determination of the Kinetics of H 2 S Oxidation in a Biofilter Hydrogen sulfide elimination rates are measured when the biofilter has reached a steady state conditions. Gas samples are taken at different locations along the tower and the H 2 S concentrations in the gas samples are analyzed. The initial H 2 S concentrations, C Q , in the inlet gas stream are varied from low to high values in order to determine the kinetics under different situations as described in the previous section. The results for compost #17 packed in Tower #2 are presented in Figures 4-19, 4-20, 4-21, 4-22, and 4-23. Under high H 2 S concentrations in the inlet gas stream (H 2 S > 400 ppmv) , the reaction appears to follow zero-order

PAGE 129

115 Figure 4-19. Linear least squares regression analysis for zero-order kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 m 3 /m -hr, compost #17.

PAGE 130

116 Figure 4-20. Linear least squares regression analysis for first-order kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 m 3 /m 2 -hr, compost #17.

PAGE 131

117 Figure 4-21. Determination of the fractional-order reaction rate coefficient, k f by linear least squares regression. Gas loading rate: 224 m 3 /m -hr, compost #17.

PAGE 132

118 Reaction Time (sec) Figure 4-22. Plot showing the fractional-order kinetics of H 2 S oxidation in biofilter. Gas loading rate: 224 m 3 /m 2 -hr, compost #17.

PAGE 133

119 1.2 0 0.2 0.4 0.6 0.8 1 h/H Figure 4-23. Concentration profiles for H 2 S as a function of packing height within the biofilter. Gas loading rate: 224 nr/m 2 -hr, compost #17.

PAGE 134

120 kinetics according to equation 4-4 (Figure 4-19) . The regression formula is expressed as: C = C Q 27. 6t + 21.8 (4-9) with a correlation coefficient R 2 of 0.9907. The good agreement obtained supports the conclusion that the reaction can be described by zero-order kinetics. The value obtained for K Q is 27.6 ppmv/s. This zero order coefficient conresponds to a maximum H 2 S elimination capacity of 130 gS/m 3 -hr for the compost at the operating conditions selected. It should be noted that the second and third data points in Figure 4-19 are below the regression curve. This deviation is mainly due to the H 2 S removal efficiency in the lower portion of the filter has been effected by high acidity and sulfate content in the compost in this portion of the bed as a result of prolonged operation. For lower inlet H 2 S concentrations (H 2 S < 200 ppmv) , the reaction appears to follow first-order kinetics as expressed by equation 4-6 (Figure 4-20). The linear regression formula is expressed as: ln(C/C 0 ) = 0.31 0.57t (4-10) In this case the value for is 0.57/s, which is applicable for inlet H 2 S concentration of less than 200 ppmv. When H 2 S concentrations fall in the intermediate range, i.e., between 200 ppmv and 400 ppmv, the data cannot be

PAGE 135

121 represented by either Equation 4-4 (zero-order kinetics) or equation 4-6 (first-order kinetics) . Several empirical equations have been tested, and Ottengraf 's diffusion limitinq model was found to be the most accurate expression describinq H 2 S oxidation kinetics in this ranqe for the operatinq conditions employed. In order to conveniently use Ottenqraf's model, and equation 4-8, however, it is necessary to define a new parameter, the fractional-order reaction coefficient k f : k f = (k 0 D e a/2m i c 0 fi)^ (4-11) It can be seen that kf is a function of the operatinq conditions of the biofilter system, and under steady state conditions, kf is constant. Thus, equation 4-8 can be rewritten as: C/C Q = (l-k f t) 2 (4-12) or (C/C Q ) h -1 = -k f t (4-13) where t = H/Ug is the reaction time. When (C/Cq) 33 -1 is plotted aqainst t, a straiqht line should be obtained, and the slope of the line should equal k f . A sample plot of the relation in equation 4-13 for the concentration ranqe 2 00 to 4 00 ppmv H 2 S is shown in Fiqure 4-21. The reasonableness of this plot supports the concept of fractional-order dependence in the stated ranqe of concentrations.

PAGE 136

122 Thus, it appears that equation 4-12 can be used to predict the kinetics of H 2 S oxidation in the intermediate H 2 S concentration ranges. A set of experimental data (H 2 S =309 ppmv) are plotted in Figure 4-22, where the symbols used are experimental data points, and the solid curve is drawn according to equation 4-12. It can be seen that the experimental data are adequately described by the diffusion limitation model. The data shown in Figures 4-19, 4-20, and 4-22 are plotted in Figure 4-23 in the "Ottengraf-f orm" , i.e., C/C Q on the y-axis and h/H on the x-axis, where h is the height of sampling location on the filter and H is the total height of the filter. It is informative to use this plot in order to observe the H 2 S concentration profile in the biofilter. According to Ottengraf, under reaction limiting conditions, a straight line should be obtained for this type of plot. However, the profile is no longer linear if the substrate elimination rate is controlled partially or completely by the diffusion rate in the biofilter. The latter feature is clearly observed in this study of H 2 S elimination. The data that are plotted in Figure 4-20 shows a straight line in Figure 4-23. This linear relation is an indication of zero-order kinetics and reaction limiting control by the microbial population. When inlet H 2 S concentrations are less than 434 ppmv (the middle curve), the profile is curved (concave) below the straight line, indicating that the reaction is partially diffusion

PAGE 137

123 controlled at intermediate H 2 S concentrations. With further lower H 2 S concentrations, the profile becomes curved well below the straight line, and suggests that the dominant mechanism is diffusion limited in this first-order kinetics region. Table 4-6 summarizes the eguations describing the kinetics of H 2 S oxidation in the biofilter system. These equations allow for a quantitative description of the basic processes involved in this biof iltration elimination of H 2 S and they allow for an accurate sizing of biofilters for H 2 S removal. It should be noted that the macrokinetics of H 2 S oxidation in biofilters as well as the reaction coefficients reported here are related to operational conditions. The kinetics are valid for a similar compost biofilter system operating under similar conditions. If operational conditions such as pH, temperature, sulfur content, etc. are changed, then the kinetic behavior may be altered. In practice, the kinetic behavior for a particular compound should be determined by laboratory or pilot scale studies (Van Lith, 1989; Leson and Winer, 1991).

PAGE 138

124 Table 4-6. Models for the kinetics of H 2 S oxidation in biof ilter . H 2 S Concen. Range (ppmv) Kinetic Order Reaction Coefficient Equation <200 First ^ = 27.6/s -dC/dt = k,C C = C Q expt-^t) 200-400 Fractional k f = 0.067/s C = C Q (l-k f t) 2 >400 Zero k Q = 0.567ppmv/s -dC/dt = k Q c " c o " M

PAGE 139

CHAPTER 5 BIOFILTER PERFORMANCE AND CHANGES IN COMPOST PROPERTIES ASSOCIATED WITH LONG TERM OPERATION Towers #1 and #2 were packed with composts #17A and #17, respectively, and operated continuously for more than 200 days. Tower #3 was packed with compost #16 and operated for 130 days. Long term performance of these biofilter tower systems for H 2 S removal and changes in compost properties are reported here. Biofilter maintenance conditions and recommended procedures for developing optimum performance have been determined through these long term observations. Overall Performance of the Biofilters The overall performance of Towers #1, #2 and #3 are presented graphically in Figures 5-1, a, b and c, respectively. In each figure, H 2 S loading rates, inlet/outlet H 2 S concentrations and H 2 S removal efficiencies are plotted against the cumulative operation time. Except for some specific tests, where extreme operating conditions were used (such as high H 2 S loading rate, high gas flow rate, etc) and data are presented and discussed elsewhere, the data showed in these figures are daily averages for an individual day when the measurements were made. Usually, Towers #1 and 2 were operated at a gas loading rate of 100 125

PAGE 140

126 Figure 5-1. Biofilter control of H 2 S during long term operation, a) Tower #1, compost #17A.

PAGE 141

127 Figure 5-1. Biofilter control of H 2 S during long term operation, b) Tower #2, compost #17.

PAGE 142

128 Figure 5-1. Biofilter control of H 2 S during long term operation. c) Tower #3, compost #16.

PAGE 143

129 m 3 /m 2 -hr. All three composts studied showed very good overall performance characteristics during continuous long term operation. The acclimation periods for all three composts are similar, approximately 10 days. After this period, the filters achieved stable operating conditions. Tower #3 was operated under a lower gas loading rate, 50 m 3 /m 2 -hr (Figure 5-lc) . This tower was not washed during 130 days of operation. When unpacked, the color of the compost at the bottom of the tower (inlet) had changed to yellowish-white, indicating accumulation of sulfur. The compost at the bottom of the tower was wet, with a water content of 55.9%, but the compost became drier as samples were analyzed at various distances up the tower. The water content of the compost in the top portion of the biofilter (exit) was only 25.8%. This characteristic is probably common for closed biofilter systems with no additional water supplied to the system except water contained in the influent gas stream. Although the gas is prehumidif ied and almost saturated with water (RH > 95%) at ambient temperature, the temperature in the compost can be a few degrees (°C) higher than the incoming gas stream due to the exothermic oxidation reactions occurring in the system and the biological respiration of the biomass. The rise of compost temperature is more significant with high H 2 S loading rates. The elevated bed temperature causes the gas stream to become more unsaturated, and results in loss of

PAGE 144

130 water by evaporation from the compost. Drying-out of the compost, therefore, is a natural feature of any biofilter system. The drying-out process may be slower for those biofilters which are used to remove hydrocarbons because water is one of the products of the biodegradation reactions. There is no water formed during the biological oxidation of H 2 S. As a result, water has to be added to the system at the effluent end of the bed to keep the compost water content constant. Towers #1 and #2 are packed with the same compost (#17) except that 2% by weight of CaC0 3 was added to the compost packed in Tower #1 (Compost #17A) as a pH buffer. It should be noted that the addition of CaC0 3 did not affect the performance of the biofilter. These two towers have been packed and operated since January 29, 1991. Both of these towers have been subjected to large variations in waste gas surface loading and H 2 S loading rates during the operation period because the effects of these variables on H 2 S removal efficiencies were evaluated in this dual-tower system. High H 2 S removal efficiencies and stable performance, were consistently observed. The composts (#17 and #17A) showed good moisture retention and buffering capacity. In addition to the water added by pre-humidif ication of the inlet gas stream, these two towers were washed biweekly using DI water. The latter procedure was performed to keep the compost water content in

PAGE 145

131 the desired range. The main purpose of this procedure, however, is to reduce the acidity and prevent accumulation of sulfate in the compost. A typical compost water content distribution profile in Tower #2 is presented in Figure 5-2. Samples were taken and analyzed before washing (14 days after the last washing) and 1 hour after washing the tower. It can be seen that 1) the water content of the compost is evenly distributed along the length of the bed, and 2) the compost has very good water retention and buffering capacity. Only 3-5% of the compost water content was lost during the 14 day interval between washings. When the biofilter tower was operated for 130 days and samples were taken and analyzed, the water holding capacity of the compost in the inlet region had decreased slightly compared to that in the outlet region. The system showed a good buffering capacity to gas surface loading changes. No significant reduction in H 2 S removal efficiency was observed when the gas surface loading rate was varied in the range 20 to 500 m 3 /m 2 -hr for the same H 2 S loading rate. Also, the buffering capacity for H 2 S loading rate or H 2 S concentration changes was very good. Under the same conditions of gas surface loading rate ( 100 m 3 /m 2 -hr) , when H 2 S concentrations were changed from 15 ppmv to 775 ppmv (corresponding to the H 2 S loading rate being changed from 2 to 50.5 g-S/m -hr) the H 2 S removal efficiency did not vary from 99.5%. Sudden changes in H 2 S loading rates over a

PAGE 146

132 Figure 5-2. Compost water content profile.

PAGE 147

133 very large range, for example, from a few tens ppmv to a few hundreds ppmv may cause a temporary reduction in H 2 S removal efficiency. However, after only a few hours, the optimum control efficiency was recovered (Figure 5-1 a and b) . This perturbation is probably due to the uneven and inadequate initial distribution of the sulfur oxidizing bacteria population. In other words, the length of the 'active portion' of the filter is related to the H 2 S concentration in the gas stream. This feature will be discussed in a later section. One of the most significant changes observed is the compost pH. Compost pH changes in different sections of Towers #1 and #2 are shown in Figures 5-3 a and b, respectively. The product of H 2 S oxidation is sulfuric acid (H 2 S0 4 ). This strong acid is soluble in water and accumulates in the compost, resulting in rapid acidification of the system. Compost acidity increases very rapidly with time, e.g. after 32 days of operation, the pH of the bottom section (inlet) of the compost dropped from 8.0 to 1.5 (TS11, Figure 5-3a) . The rate of compost pH change is proportional to the H 2 S input. The pH drop in the lower portion of the compost is much larger than that in the upper portion of the bed since most of the H 2 S oxidation reaction takes place in the former region. When the gas stream flows through the bed less H 2 S is left in the gas stream and less H 2 S0 4 is formed as the bed is traversed, therefore, the acidification is slower in the upper portion of the tower.

PAGE 148

134 0 20 40 60 80 100 120 Cumulative Operation Time (Day) a TS11 + TS12 o TS13 a TS14 x Wash Water Figure 5-3. pH changes of compost in different sections of the biofilter with operation time, a) Tower #1, compost #17A.

PAGE 149

135 Figure 5-3. pH changes of compost in different sections of the biofilter with operation time, b) Tower #2, compost #17.

PAGE 150

136 Washing the tower with water effectively mitigates the pH decline. Since H 2 S0 4 is water soluble, a major fraction of the accumulated acid can be washed out at each washing. The pH of the wash water for each treatment is lower than the compost pH (by approximately 1 to 1.5 pH units). A few measurements of compost pH before and after washing were conducted and the results of these measurements are indicated in Figures 5-3 a and b. An increase in compost pH from 0.2 to 0.5 pH units was achieved during each washing, thus, if the tower is washed routinely, compost pH can be kept constant. The effectiveness of the washing process on pH stabilization depends on the guality of the water used and the contact time between the compost and water. Both towers #1 and #2 are washed by 10 L of DI water each time with a flow rate of 1 L/min from the top of the tower. The water is allowed to freely flow downward through the tower under gravity, as a result, the water-compost contact time is approximately 10 minutes. Because the rate of pH decrease is proportional to the H 2 S loading rate to the system, with a high H 2 S loading rate the tower needs to be washed more freguently. Addition of lime or CaC0 3 to eliminate acidification of the biofilter is not effective and is not recommended. There are some disadvantages in adding lime to the biofilter system:

PAGE 151

137 1) Premixed lime is effective only temporarily, lime can be consumed by the accumulated acid very quickly; 2) Since high concentrations of H 2 S0 4 are continually formed, the quantity of CaC0 3 required to neutralize the acid is very large. Addition of large amounts of lime increases the inorganic fraction of the filter medium and significantly changes the compost construction and composition. 3) The addition of lime increases the smaller particle fraction in the filter, which results in a significant increase in pressure drop across the filter bed, and 4) Most importantly, addition of lime does not solve the problem of sulfur accumulation in the compost, which appears to be the main reason for the decline in H 2 S removal efficiency. Accumulation of Sulfur in Compost and Its Effect on System Performance Another serious problem which is frequently encountered in a H 2 S-biof ilter system is the accumulation of sulfur in the filter material (Carlson and Leiser, 1966; Rands et al., 1981; Yang and Allen, 1991). This feature has been routinely observed during the course of this study in both Towers #1 and #2 since these towers are transparent. During long term operation of the biofilter the color of the compost eventually changes from dark brown to a yellowish white. The color change progresses from the lower region of

PAGE 152

138 the bed (inlet) to the upper layer. White deposits on the surface of compost particles are easily observed. The rate of sulfur deposition is proportional to the rate of H 2 S loading. A sudden increase in H 2 S loading in a large concentration range and prolonged operation at high H 2 S loading rates can cause the white colored material to accumulate rapidly and spread from the lower region to the upper region of the bed. This discoloration of the bed is accompanied by a rapid drop in pH of the compost. Also, the temperature of the biofilter system can rise 2-3 °C for high H 2 S loading rates indicating enhanced biological activity of the microbes. If no appropriate action is taken to counteract sulfur accumulations, then the system performance and H 2 S removal efficiency will decline rapidly. In biof iltration processes, H 2 S is oxidized both chemically and biologically to sulfate under aerobic conditions. In nature a variety of reduced inorganic sulfur compounds (e.g. elemental sulfur, thiosulfate) occur as intermediates between sulfide and sulfate, the reduced and oxidized forms of sulfur, respectively. As these compounds are oxidized only slowly by direct chemical reaction with oxygen (Kuenen, 1975) , it is clear that biological oxidation must play an important role in the recycling of reduced sulfur compounds under aerobic conditions. This mechanism appears to be true, also, in biof iltration systems. Also, many microorganisms can oxidize reduced sulfur compounds, the colorless sulfur bacteria are known to play a

PAGE 153

139 major role in the oxidation of reduced sulfur. The colorless sulfur bacteria are divided into three families, the Thiobacter iaceae , the Begg iatoaceae , and the Achromatiaceae. The genera, which belong to these families, have been studied and include Thiobacterium . Macromonas . Thiovulum , Thiospira . Thiobacillus . Thiomircosoira . Sulf olobus f Beqgiatoa ff Thiospirillopisis . Thioploca . Thiothrix . Thiodendron . and Achromatium (Kuenen, 1975) . The colorless sulfur bacteria are naturally occurring almost everywhere on the earth. These bacteria live in a wide pH range from 1 to 8 , and temperatures up to 85°C (Kuenen, 1975; Brock and Madigan, 1988). To distinguish between the active genera and species of sulfur bacteria is not the goal of this study. The macro processes of H 2 S oxidation, which involve physical, chemical and biological processes are of greatest interest and application in this study. A number of oxidation reactions of inorganic sulfur compounds which are effected by the colorless bacteria, especially thiobacilli, have been reported (Starkey, 1966) . Some of the important reactions that possibly occur in a biofilter system are as follows: H 2 S + 20 2 H 2 S0 4 (5-1) 2H 2 S + 0 2 2S° + 2H 2 0 (5-2) 2S° + 30 2 + 2H 2 0 2H 2 S0 4 (5-3) Na 2 S 2 0 3 + 20 2 + H 2 0 -» Na 2 S0 4 + H 2 S0 4 (5-4) 4Na 2 S 2 0 3 + 0 2 + 2H 2 0 2Na 2 S 4 0 6 + 4NaOH (5-5)

PAGE 154

140 2Na 2 S 4 0 6 + 70 2 + 6H 2 0 2Na 2 S0 4 + 6H 2 S0 4 (5-6) 5H 2 S + 8KNO3 4K 2 S0 4 + H 2 S0 4 + 4N 2 + 4H 2 0 5S° + 6KNO3 + 2H 2 0 3K 2 S0 4 + 2H 2 S0 4 + 3N 2 (5-7) (5-8) 5Na 2 S 2 0 3 + 8NaN0 3 + H 2 0 -» 9Na 2 S0 4 + H 2 S0 4 + 4N 2 (5-9) From the reactions listed above and Table 2-1 in Chapter 2, it can be seen that the colorless sulfur bacteria can oxidize both hydrogen sulfide (H 2 S) and the intermediate reduced sulfur compounds to sulfate. Different sulfur compounds, therefore, in various stages of oxidation can be expected to be present in the biof iltration system. The original compost, #17A, and compost samples in the biofilter, after 3 months continuous operation, were collected and analyzed for total-S and for fractionation into various sulfur components. Compost samples in the biofilter system were taken from the lower (TS11, 0.125m), middle (TS13, 0.625m), and upper (TS14, 0.875m) regions of the tower. Total-S and the following sulfur components: ester-S, FeS 2 -s, FeS-S. S°-S, water-soluble-S , P-extractable-S and insoluble-S, were analyzed using the methods described in the previous section. S0 4 2 "-S and inorganic-S are estimated according to individual analyses. Organic-S is calculated as the difference between total-S and inorganic-S, and C-bonded-S is the difference between organic-S and ester-S. The results are summarized in Table 5-1.

PAGE 155

141 ro ro^rocNVOvocncn o vo o VO • • 0"> CO • • ^J* ^J 4 n o • • • vo in • • vo o\ • • p if) o\° H ro o ro oHHCNr-incoco Cn 00 H • CD o ^ ( O o o o o o A N Cr> 00 CTi O CT\ CTlOOfNr^CTV^f^' \ H H O o ro r\j • VO • • • H H • O in *• — H cm eg H P (N U 10 U o O o o g s — \ o O IT)HO\fHHCfiH 0 CO • O CO CM 0 ro 1 ro H 0> H ro (N o CnOOOCT>vOHH -a CO c (0 < o ^ — . a\ ro in nnm'Trivooo H <*> O vo vo o • • H in • • • o • • • in • • h rjioo (0 c H O CO 1 «* *r o CMOOOHOOH 0 • O o o o oooooooo H U P CD O P (0 rH CO CO rH -H 1 1 CO • • 1 m
PAGE 156

142 The original compost has a total-S content of 0.74 mgS/g, 64% of which is organic-S. The total-S and sulfur constitution of the original compost are close to similar data reported by David et al. (1982) for surface soils in a forest. High organic-S content is a good indication of high biomass and microbial sul f ohydrolase activity of the compost. This correlation was quantitatively determined by David and coworkers (1982) . Inorganic-S is dominant in the 'used' filter compost samples (>95%) . A lower value for TS13 is probably due to analytical error. Fifty to sixty percent of the total-S in the biofilter compost is water soluble-S, which indicates that the final product of H 2 S oxidation is H 2 S0 4 . A large amount of FeS-S and S° is measured in the lower region of the filter bed (TS11, 0-0. 125m of the filter). The total-S distribution profile is graphically shown in Figure 5-4. As the test gas flows through the filter bed in an upward direction, the lower region of the compost bed is always exposed to higher H 2 S concentrations than the upper region. This results in sulfate formation and accumulation at a higher rate in the lower region of the bed. The higher fraction of intermediately oxidized sulfur compounds, FeS and S°, in the lower region is a result of incomplete oxidation of H 2 S due to high H 2 S concentrations and reduced biological activity of the biomass as a result of lowered pH and high sulfate content in this region.

PAGE 157

143 140 Original Height Above Bed Inlet (m) Figure 5-4. Total-S distribution profile in biofilter, Tower #1, after exposure to H 2 S for 100 days.

PAGE 158

144 Biological activities of the biomass in each region of the filter bed material were indirectly determined by measuring sectional H 2 S removal efficiency. Identical gas flow rates and inlet H 2 S concentrations were used for these measurements. The results are shown in Figure 5-5. It can be seen that the most effective region in the biofilter is between 0.2 and 0.4m. The biological activity of the lower region (0-0. 2m) is restricted by the factors mentioned above. The maximum population of the sulfur oxidizing bacteria as well as the optimum biological activity occur in the second region (0.2-0. 4m) of the filter and decreases progressively up the bed. This observation is reasonable because less and less H 2 S is available in the gas stream as it passes upward through the filter and more of the H 2 S in the gas stream is eliminated by reaction with the biofilter in the lower region. The population of the sulfur oxidizing bacteria and the biological activities appear to show a modal (Gaussian type) distribution along the filter bed. With prolonged operation, the mode will move upward from the lower region of the bed due to the increasing toxicity caused by accumulation of sulfate and acidification of the compost in the lower region. The latter effect is referred to as system upset, which must be avoid to maintain effective control efficiencies.

PAGE 159

145 Figure 5-5. H 2 S removal efficiencies in different regions of the biofilter, Tower #2. A: 0-0.2 m B: 0.2-0.4 m C: 0.4-0.6 m D: 0.6-0.8 m E: 0.8-1.0 m

PAGE 160

146 System Upset and Recovery System upset is indicated by a sudden decrease in H 2 S removal efficiency, increased H 2 S concentration in the effluent gas stream and noticeable objectionable odor. The most common reasons for system upset are compost dry-out and H 2 S overloading. A dry compost system can be easily determined by measurement of compost water content. As discussed previously, if the compost water content dropped below 30%, reduced H 2 S removal can be expected. For a closed system without additional water supply a dry region is generally observed in the upper portion (exit) of the biofilter. Drying of compost also causes shrinking of the compost and results in channeling. This particular feature is indicated by a decreased pressure drop across the filter bed. The dry-out problem can be solved by spraying water at the exit (top) of the compost filter. Channels generally disappear as a result of the compost volume expanding after watering. The system may need a few days to rebuild the microbial population and recover its optimum performance depending on its original dryness, as described in previous sections. Another cause of system upset is overloading the system with high H 2 S concentrations. The maximum H 2 S elimination capacity for a filter medium depends on the nature of the material and the operating conditions of the system. Generally speaking, the maximum H 2 S elimination capacity of

PAGE 161

147 a filter material is determined at optimized operating conditions. Changes in operating conditions, such as a lowered pH, dry-out of the compost, accumulation of sulfur in the bed material, etc. can significantly decrease the H 2 S elimination capacity of the filter medium. Therefore, the actual H 2 S elimination capacity of a filter material at a specific condition is always egual to or less than its maximum capacity. When the H 2 S loading rate exceeds the elimination capacity of the compost, then the system is overloaded. An overloaded system is indicated by a high H 2 S concentration in the effluent gas stream, a low removal efficiency, noticeable odor and the occurrence of a compost color change (white deposit on compost particles) . Local overloading is often observed when channeling occurs in the system (Rands et al., 1981) or the influent gas stream is not evenly distributed. In either case, the region of the filter where the high flow rate occurs is overloaded by H 2 S. Local overloading can be cured by correcting the channeling and the gas distribution system. If the whole system is overloaded by high H 2 S input, the solution to the problem is different. If the system is temporarily overloaded for a few hours, the performance of the filter can be recovered by decreasing the H 2 S loading rate. A certain fraction of the white deposits on the compost are intermediate oxidation products, such as S°, FeS 2 , S 2 0 3 2 ~, etc. When H 2 S loading is decreased, the

PAGE 162

148 microbial population uses these intermediate oxidized compounds as its energy source and oxidizes the intermediates to sulfate. As a result, the white color and other deposits disappear within a few days. However, the compost pH is significantly decreased and the sulfur content of the compost is increased due to the formation of large amounts of sulfuric acid. The latter results in the H 2 S elimination capacity of the compost being reduced if no counter measures are taken. If the system is continuously overloaded by high H 2 S input, the deterioration of biofilter performance can not be reversed by simply decreasing the H 2 S loading rate. The biological activity of the microorganisms in the filter is strongly inhibited by high sulfate content and very low pH. In this case, specific treatment is reguired to recover the defective compost and the deteriorated system. In order to determine a proper method to recover the defective compost, a set of experiments were designed and conducted. The resulting information is provided in the following sections. Selecti on of Chemical Solutions The Column System #4 was used for this test. The defective, white-colored compost was obtained from the bottom region of Towers #1 and #2 and mixed thoroughly. One hundred and fifty grams (150 g) of this compost was packed in each of the 8 columns. Waste gas containing H 2 S was

PAGE 163

149 introduced to the system at a flow rate of 230 raL/min (Equivalent to 15 m 3 /m 2 -hr surface loading rate) . Influent and effluent gas samples were analyzed to determine the original performance of the compost before treatment. The compost is then unpacked and divided into 4 groups, each group was packed into two columns as duplicates. The first group was treated with DI water, the second group with 0.05M NaOH, third group with 0.05M NaHC0 3 and the fourth group was left untreated as a control. For each treated group, the compost was shaken with 10 times the liquid (by weight) for 30 minutes with a rotary shaker. The composts are treated this way twice. After each shaking treatment, the compost pH was measured. The total-S of the original defective compost and the composts after treatment were measured. The results of these compost analyses are summarized in Table 52. It can be seen that after being washed twice, the pH of water treated compost was raised 0.3 pH unit, but both the NaOH and NaHC0 3 treated composts are neutralized to near pH 7. The total-S content of the compost was successfully reduced by 72 to 85% in all cases. According to this study, water appears to be the best washing agent. After treatment, the composts are repacked into the columns, and the system subjected to an H 2 S removal test. The system was allowed to operate continuously for one week before gas sampling and analysis was conducted. The delay in testing was included to eliminate the effects of residual alkali on H 2 S removal for those composts treated by NaOH and

PAGE 164

1 150 Table 5-2. Effect of washing on compost pH and sulfate content by DI water, NaOH and NaHC0 3 solutions. Treatment Time of Treatment pH Total-S (mg-S/g) Reduction of Total-S (%) Untreated (control) 1.59 120 Water First 1.61 Washed Second 2.20 17.5 85.4 NaOH First 1.73 Washed Second 7.51 33.7 71.9 NaHCO-i First 1.86 Washed Second 6.66 20.2 83.1

PAGE 165

151 NaC0 3 • The systems were tested under two different operating conditions, a) at a low H 2 S loading rate, 36 g-S/m 3 -hr and b) at a higher H 2 S loading rate, 71 g-S/m 3 -hr. At the lower loading rate, the H 2 S removal efficiency increased from 20.6% to 99.9+% for all three treated composts. The H 2 S elimination capacities of these composts were determined at high H 2 S loading rates. From Table 5-3 it can be seen that the H 2 S elimination capacities for all the three treated composts were increased by a factor of 9, from 7.5 to 68 gS/m 3 -hr. The NaOH-treated compost looks darker and feels sticky, probably the structure of the compost was altered by this aggressive alkali treatment. From an economic point of view, water is the best choice for the treatment. If acidity of the compost needs to be corrected, then aqueous NaHCO-j solution is recommended as a treatment chemical. The shaking-washing method used here is not feasible in practice for full scale biofilters. The most feasible way to treat a full scale bed is to spray water or the desired solution onto the top of the filter. In this case, the water to compost ratio and contact time become critical for the effectiveness of the treatment.

PAGE 166

152 Table 5-3. Performance of defective compost before and after treatment. At low H 2 S Loading 3 At High H 2 — * a • b S Loading Compost Treatment H 2 S Removal Eff. (%) H 2 S Elimination (g-S/m 3 -hr) H 2 S Removal Eff. (%) H 2 S Elimination (g-S/m 3 -hr) Untreated (control) 20.6 7.51 17.0 12.0 Water Washed 99.9+ 36.4 96.6 68.6 NaOH Washed 99.9 36.4 96.7 68.6 NaHCOo Washed 99.9 36.4 94.9 67.4 H 2 S inlet concentration: 437 ppm; Gas loading rate: 14.4 m /m 2 -hr. H 2 S inlet concentration: 489 ppm; Gas loading rate: 30.0 m 3 /m -hr.

PAGE 167

153 Effect of Water-Compost Contact Time on S0 4 z " Leaching Efficiency Water is used as a solvent to determine the effect of contact time on S0 4 2 ~ leaching efficiency. Three (3.0) g of wet compost #13-1 was weighed into each of nine 50-mL beakers. Thirty (3 0) mL of DI water is then added to each beaker. The beakers with compost and water are allowed to contact without disturbance. After the desired period of time (from 5 to 120 minutes) , the content of the beakers are filtered and the filtrate subjected to pH and sulfate determination. Another container with 3 g compost and 30 mL of DI water are shaken for 60 minutes as a control. The results are shown in Figure 5-6. At constant water to compost weight ratios of 10:1 and without shaking, the S0 4 2 " leaching efficiency is between 51% and 68%. The S0 4 2 ~ leaching efficiency increases with water-compost contact time. The maximum leaching efficiency is achieved in about one hour. Effect of Water to Compost Ratio on S0 4 ^ Leaching Efficiency The effect of water to compost weight ratio on S0 4 2 ~ leaching efficiency was determined by shaking 1, 2 and 3 g of wet compost in 3 0 mL of DI water for 3 0 minutes. The supernatant was filtered and sulfate was determined in the filtrates. The results are shown in Figure 5-7. It is obvious that S0 4 leaching efficiency is increased significantly with increasing water to compost weight

PAGE 168

154 Figure 5-6. Effect of water-compost contact time on sulfate leaching efficiency.

PAGE 169

155 50 40 30 20 10 0 10 15 30 Water/Compost Ratio (Wt/Wt) Figure 5-7. Effect of water/compost ratio on sulfate leaching.

PAGE 170

156 ratio. In practice, the water-compost contact time can be easily controlled by the water spray rate. The selection of ideal water to compost ratio, however, depends on the pH, sulfur content of the compost and frequency of the treatment. High water to compost ratios favor the leaching of sulfate but more water is required and the maintenance cost is consequently increased. The wash water is highly acidic and has high concentrations of sulfate. Proper treatment of this water is required. The water to compost weight ratio and contact time used for this study are 1:1 and 10 minutes, respectively. Efficiency of elimination of sulfate in filter compost for a single wash under this condition is shown in Table 5-4. The average sulfate elimination efficiency is 36.5% with some variations at each location in the bed. Sulfate concentrations in the wash water increased from 0 to 5.25 mg-S/g, which indicates that the sulfate sulfur has been efficiently transferred from compost to wash water.

PAGE 171

157 Table 5-4. Effect of water washing on elimination of sulfate in filter compost. Samplimg Location Sulfate in Compost (mg-S/g) Sulfate Elimination (%) Before Washing After Washing T14 1.67 0.34 79.6 T13 6.71 6.16 8.20 T12 13.4 11.0 17.8 Til 57.1 33.9 40.6 Wash Water 0.00 5.25

PAGE 172

CHAPTER 6 FULL SCALE APPLICATION OF BIOFILTRATION TO CONTROL HYDROGEN SULFIDE EMISSIONS AT A WASTEWATER TREATMENT PLANT A full scale biofilter bed system has been installed and operated since the fall of 1988 to control H 2 S emissions from the grit chamber at the Kanapaha Wastewater Treatment Plant, Gainesville, Florida. In this chapter is describes the design, construction and operation of this system. Experience gained, as well as advantages and disadvantages observed during operation of the system are discussed. Introduction Emissions of objectionable odors are a common problem encountered at most wastewater treatment plants. The odorous compounds freguently observed as volatile emissions from these sources include hydrogen sulfide, ammonia, organosulfur compounds and some volatile organic compounds (VOCs) (WPCF, 1979; Yang, 1988). The origin of these odorous chemicals is in the sewer lines where, due to existing anaerobic conditions and excessively long residence times for the incoming wastewater, the odorous compounds are formed and confined. The predominant odorous compound emitted from municipal waste water treatment plants is H,S. Under anaerobic 158

PAGE 173

159 conditions, inorganic sulfates and sulfites can be easily reduced to hydrogen sulfide by various types of anaerobic and facultative bacteria, such as sulfur-reducing bacteria (SRB) . S0 4 2 " + 2C + 2H 2 0 SRB ^ 2HC0 3 ~ + H 2 S (5-1) Because hydrogen sulfide is volatile and only partially soluble in water, whenever domestic wastewater from sewer lines is agitated and exposed to the atmosphere, H 2 S is released to the air. The Kanapaha wastewater treatment plant is the major municipal sewage treatment facility ( 9 million gallons per day) for the city of Gainesville, Florida. As is the case for most city public utilities, this domestic wastewater treatment facility has been identified as a source of odorous emissions and has been the recipient of numerous odor complaints for many years. The most significant source of the malodor was identified as the plant's grit chamber, where agitation of the incoming wastewater causes considerable outgassing of hydrogen sulfide. Attempts to control the emission of odors at the plant using chemical treatment of the wastewater were only partially successful and very expensive. The high costs of the chemicals used, as well as additional problems encountered in the treatment process prompted the management and staff of Gainesville Regional Utilities (GRU) to look for alternative effective odor control technologies. As a result, a biofilter system

PAGE 174

160 was installed in the summer and made operational in the fall of 1988. System Design and Construction The biofilter control system consists of an air collecting system, an anti-corrosive blower and ductware, a humidifier, a central shaft for flow adjustment and a twobed SIEBO-stone air distribution system, as shown in Figure 6-1. The plant's malodor source, the grit chamber, was covered and sealed as tight as possible (Figure 6-2) . A negative static pressure inside the grit chamber is maintained by a blower, which prevents possible leakage of malodorous air. The air collected by the blower through 0.30 meters diameter ducts is forced to a humidifier, where water is dispersed by several spray nozzles to saturate the waste gas stream. All pipes in the waste gas collection and transmission system are made from PVC to prevent corrosion. The biofilter bed gas distribution system is constructed from the German patented SIEBO-stones . This interlocking sinter block base unit of the biofilter system is rigid and allows for heavy vehicles to be driven on it without damaging the system. The SIEBO-stone base system provides for an even distribution of the inlet air to the filter bed as well as functioning as a drainage system. The filter area is 100 square meters (m 2 ) and is divided into two equal sections, each of which can be operated and

PAGE 175

161 SIEBO-stone Gas Distribution System Central Shaft Waste Gas Inlet t \ rnvn«w n i x ». > i > Compost Filter Humidifier J,V Compost Filter * » « c > I « 1 « I * « I 'I • x I £ 1 X * x < !i * t Drainage 4 Cinder Block Retaining Wall Plan View Section A-A Figure 6-1. Schematic diagram of the Kanapaha biofilter bed system.

PAGE 176

162 Figure 6-2. Photograph of the grit chamber at Kanapaha Wastewater Treatment Plant (top view) . The chamber is covered to collect the malodorous gas.

PAGE 177

163 controlled individually. During normal operation, each section treats one-half of the total exhaust gas flow volume. The total flow can be diverted to one section if necessary, for example when repairs are needed, without reducing overall pollutant removal efficiency. The actual waste air flow rate through the biofilter is in the range from 79 to 96 actual cubic meter per minute, which gives an average surface loading rate of 52 m 3 /m 2 -h and a biofilter bed treated gas retention time of 88 seconds. Biofilter design and operation parameters are summarized in Table 6-1. Sampling and Analysis Methods To monitor biofilter bed performance, inlet and off-gas samples are collected and analyzed for H 2 S concentrations. The influent gas samples are taken from the inlet pipes in the central shaft and collected in Tedlar bags. These samples are later quantitatively diluted with pure nitrogen (N 2 ) and analyzed. Off-gas samples from the biofilter are collected by a specially designed gas sampling system illustrated in Figure 6-3. The plastic collector cover, 275 mm in diameter and 220 mm in height, is open at the bottom and has five 10 mm holes drilled in the top to allow the flowing off -gas to purge the collector and escape. A Teflon sampling probe is inserted from the top center of the cover and extends to within 100 mm from the open base. This design ensures representative sampling and avoids any

PAGE 178

164 Table 6-1. Summary of Kanapaha biofilter bed design and operation parameters. Total Flow: 79-96 acmm Filter Area: 100 m 2 Filter Height: 1.3 meter Gas Loading Rate: 47 57 m 3 /m 2 -hr Retention Time: 82 100 sec. Temperature: 15 30 °C Pressure Drop: 150 200 mmH 2 0

PAGE 179

165 Figure 6-3. Biofilter off-gas sampling system.

PAGE 180

166 disturbances from spurious ambient air currents near the bed surface. The probe is connected to a Nutech Model 218 integrated gas sampler by a 7.5 meter length of 6.4 mm Teflon tubing. Sampling is accomplished by evacuating the dead space between the inner wall of the canister and the outer walls of the Tedlar bag at a constant rate. The purge line pump is allowed to run for 5 minutes to flush the sample line and allow for off-gas equilibrium to be established inside the sampling cover. The sampling flow rate is set at about 1 liter per minute (Lpm) . Off-gas samples are collected on the top surface of the twin biofilter beds at four locations. Each rectangular bed was divided into two equal-area triangles by drawing diagonals. The sampling locations were selected at the centroids of each of the four triangular areas. Analytical results for these four off-gas samples were averaged later to obtain a typical off-gas concentration. The gas samples were transported in the Tedlar sampling bags to the University of Florida (UF) laboratories and analyzed within eight hours of collection using a Tracor 250H Analyzer. Results and Discussion Early in August 1988, the west half of the biofilter bed was filled up to a grade of 1.3 meters with compost obtained from Pompano Beach, Florida. The biofilter system was tested by operating the system at half-filter bed

PAGE 181

167 capacity for a few weeks. Because it was not possible to acquire the same compost for the other half of the biofilter system, the Kanapaha wastewater treatment plant staff decided to mix the existing Pompano Beach compost with compost obtained from the Buckman wastewater treatment plant in Jacksonville, Florida and in-house compost. This procedure was necessary to obtain a sufficient quantity of compost to completely fill the entire biofilter bed system to a depth of 1.3 meters. The final filter material used is a mixture of yard waste compost, pine bark and sewage sludge. Lime was applied to the compost material prior to installation to buffer the bed acidity (pH) . This mixed compost (Compost #12) was then used to completely fill both biofilter beds on November 20, 1988. Analytical results for the composition of the original Pompano Beach compost (5/10/88) and for the final compost mixture (11/21/88) used are presented in Table 6-2. The biofilter system was brought to full operation on November 21, 1988. The fully operational system is shown photographically in Figure 6-4. Influent and off-gas air samples were collected and analyzed during the first 16 days that the biofilter was made fully operational (see Table 6-3). During this startup period, H 2 S concentrations in the influent gas stream varied between 156 ppmv and 229 ppmv, and the average offgas H 2 S concentrations were observed to be in the range of 0.05 to 0.4 ppmv. There were, however, some minor

PAGE 182

16 Cn c -H M TS BO l rH c P u o a E o u T3 0) -q u ai p \ in •H O O •H CO X! CO \ (0 o X! H c e (0 o O •H 73 TJ O O -H •H M m a) a rH o c o >1-H nj m e h o (0 ^ -P 2 <*> o En o ^ •H c «— o CJ p s >1 -p —» o co o 3 C \ CQ a) Cn Q w 0) P a n eo f • • • • i (N ft VO I in vo • h r-» • eg . . | VO i-l vo VO I O CO O CTi • • • • | H n ma i ^ ^ H O ^ in o • I * H O in n vo • I • • I CTl I Tf VO I n mm in ov co i co in i • i • • i m o r\j o o VO VO VO CO ffl^NHri H n in in r> co a> vo vo vo vo vo vo vo H VO VO VO (N n in in in tj> m in n o M N n (M • • • • o o o o CO CO O O H co co cfi a\ o\ w w\ o h vo o in H N H N O in h in eg cm O i— I O H O

PAGE 183

169 Figure 6-4. Photograph of the biofilter system at Kanapaha Wastewater Treatment Plant.

PAGE 184

170 Table 6-3. Summary of Kanapaha biofilter influent and effluent gas sample analyses during three week start-up period. Date Avg. H 2 S Influent (ppmv) H 2 S Avg. Off-gas (ppmv) Removal Efficiency (%) 11/21/88 11/22/88 11/23/88 11/25/88 11/29/88 12/06/88 12/15/88 195 229 156 208 140 175 167 0.39 0.16 0.32 0.12 0.05 0. 11 0.07 99.8 99.9 99.8 99.9 99.9+ 99.9+ 99.9+

PAGE 185

171 differences in concentrations at the four sampling locations (see Figure 6-5) . The similarity in the low H 2 S concentrations simultaneously measured at all four off-gas sampling locations indicate that the incoming gas is evenly distributed across the filter bed. The final biofilter control system, which used mixed compost, functioned effectively immediately upon operation with a very high H 2 S removal efficiency (99.8%). No initial period of reduced efficiency or acclimation was observed. This unigue feature is probably due to the fact that half of the compost had been previously exposed to H 2 S laden air for a few weeks prior to use in the full scale operation. The average efficiency of the biofilter in removing hydrogen sulfide during the initial study period was determined to be greater than 99.8%. Several additional tests were made in order to confirm the validity of the H 2 S removal efficiencies achieved by the biofilter system. These tests included, determination of the decomposition of hydrogen sulfide in the Tedlar sampling bags and observations of the effect of varying sample eguilibration time in the gas sampling cover prior to sampling. Results of these tests are presented in Figure 66. To a certain degree, the concentrations of H 2 S for the stored influent gas samples (high concentration) are observed to decrease gradually over a few days. However, there was no significant change in the H P S concentrations

PAGE 186

172 Figure 6-5. Off-gas sampling locations on the biofilter beds and concentrations of hydrogen sulfide observed as a function of biofilter operating time.

PAGE 187

173 Figure 6-6. Concentration changes for hydrogen sulfide in gas samples contained in Tedlar bags as a function of container holding time.

PAGE 188

174 for off-gas samples (low concentrations) stored over a similar observation period. Since the gas samples are analyzed within 8 hours after sampling, the concentration changes in the Tedlar bags for both influent and off-gas samples are considered to be negligible, when estimating control efficiencies. The effect of sample purging time for the collection chamber on concentration measurements of effluent gas from the biofilter was determined as follows: Off-gas samples are collected at 0, 5, 15 and 20 minutes after placing the sampling cover on the compost bed and purging. Each sample is collected for 5 minutes at a flow rate of 1 Lpm. The results are illustrated in Figure 6-7. These results indicate that there is no significant difference in H 2 S concentrations obtained for the different purging times selected. The gas loading rate through the compost bed is about 780 L/m -min, and the sampling cover has a volume of 12 liters with a cross-sectional area of 0.06 m 2 . The dead volume of the cover, therefore, can be replaced by the offgas every 0.3 minutes. Because the collections of off -gas samples from the biofilter are usually initiated 5 minutes after the sampling chamber is placed at the sampling locations, it is to be expected and has been demonstrated that representative off-gas samples are collected from the biofilter bed under these conditions. The waste gas humidifier works effectively to keep the compost bed within a moisture content range of 45-60% (Table

PAGE 189

175 0.20 0.10 0.00 Purge Time (min) Figure 6-7. Effect of varying purging time for sample collection chamber prior to sampling on measured hydrogen sulfide concentrations .

PAGE 190

176 6-2) . Influent air measurements indicated that the relative humidity of the waste gas provided to the biofilter bed after humidif ication is at least 95%. No other water was introduced to the filter bed except that due to rain. No significant changes in the bulk density and particle size distribution of the compost were observed during the 27 months of operation (from 11/88 to 2/91) . The organic matter content decreased by about 4.6%, presumably due to mineralization of the compost. Total nitrogen and total carbon contents of the compost decreased at different rates resulting in an increase of the C/N ratio from 11.4 to 23.2. The most significant changes observed in the compost during extended operation were the total sulfur content and pH. The total sulfur, expressed as mg-S/g-compost on a dry basis, increased considerably from 7.3 to 109. After prolonged operation, parts of the biofilter bed showed a pronounced color change in the compost from dark brown to yellowish white, which was accompanied by observation of a rotten vegetable odor. The pH of the compost bed material decreased significantly from 8.6 to 1.8 as a result of the continuous removal of H 2 S and corresponding formation of sulfuric acid (H 2 S0 4 ) by the biological oxidation of H 2 S. Although this acidification of the compost did not appear to have a direct and noticeable effect on the overall H 2 S removal efficiency, serious corrosion of the cement blocks in the retaining wall and the SIEBO-stone base was observed. Corrosion of the

PAGE 191

177 latter resulted in blocking of the narrow waste gas distribution vents between the SIEBO-stones , which caused channeling and uneven gas distribution throughout the biofilter system. Corrosion of the SIEBO-stones and the resultant blockage of the gas distribution system were first noticed by Kanapaha Plant workers in late 1990. To solve this problem, the west bed of the biofilter system was unpacked and the gas distribution vents between the SIEBO-stones were cleaned manually. The compost used in this bed was thoroughly mixed by turning and repacked in the bed after the gas distribution vents had been cleaned. On February 5, 1991, both gas and compost samples were taken at the biofilter beds. Results of this recent gas sampling and analysis exercise are presented in Table 6-4. Sampling sites 1, 2 and 3 are located on the west bed; 4 and 5 are located on the east bed. Almost no air was vented through the east bed, even though the gas flow control valves to both beds were fully open, because of the corrosion and blockage of the east bed gas vents. During cleaning of the east bed air distribution system, the entire waste gas stream was forced through the west bed. The waste gas stream, however, was no longer evenly distributed through this 'cured' system. Channeling caused the waste gas to blow through the filter without control and resulted in very high off gas H 2 S concentrations (sites 1 and 3, Table 6-4). Some removal of the H 2 S was measured at Site 2,

PAGE 192

178 Table 6-4. Gas sampling and analysis for Kanapaha biofilter bed, 2/5/91. Sampling H 2 S H 2 S Location Concen. Removal (ppmv) (%) Inlet 136 1 133 2.57 2 81.3 40.3 3 134 1.40 4 0.04 99.9+ 5 0.04 99.9+

PAGE 193

179 however, the removal efficiency was much lower than previously recorded values obtained during normal operations. The large variation in H 2 S concentrations in the off gas at different locations indicates uneven distribution of the waste gas stream throughout the system. Since not much air is vented through the east bed, the H 2 S concentrations measured in the off gas for this bed (sites 4 and 5) are very low and , therefore, the calculated removal efficiencies for this bed (Table 6-4) are questionable. A strong characteristic rotten egg odor (H 2 S) was smelled around both filter beds during sampling. The color of the compost packed on the west bed had changed to yellowish-white in comparison to the color of the compost packed on the east bed, which had still retained its original dark brow color (Figure 6-8) . Analysis of the compost samples from these discolored sites is shown in Table 6-5. Approximately 20% of the total S is organic-S and 80% is inorganic-S. Very high fractions of FeS 2 -S (27.6%) and insoluble sulfate (19.98%) were determined in these samples. As a result of this long term operational corrosion problem, GRU engineers have decided to replace the existing SIEBO-stones with similar blocks made from anti-corrosive materials.

PAGE 194

180 Figure 6-8. Compost samples taken from Kanapaha Wastewater Treatment Plant biofilter beds (2/5/91) . Left: Sample taken from west bed. White color indicates high sulfur accumulation . Right: Sample from east bed. Low sulfur content compost, color is close to the original (dark brown) .

PAGE 195

181 Table 6-5. Sulfur fractionation of a typical compost sample in Kanapaha biofilter bed. mg-S/g (Wt%) Total-S 110 100 Organic-S 22.3 20.3 C-bonded-S 21.1 19.3 Ester-S 1.20 1.10 Inorganic-S 87.2 79.7 FeS 2 -S 30.0 27.6 FeS-S 0.40 0.40 s°-s 3.9 3.6 so 4 2_ -s 52.7 48. 1 Water Soluble-S 20.9 19. 1 P-extractable-S 9.9 9.0 Insoluble-S 21.9 20

PAGE 196

182 Conclusions In spite of the few limitations described, for a full scale system, biof iltration has been demonstrated to be a simple, effective and inexpensive method for odor control at wastewater treatment plants based on the Kanapaha plant experience. The biofilter system described was successfully operated at high H 2 S removal efficiencies with little or no maintenance for a period of 2.5 years. The system was effective until the corrosion problem occurred and the gas distribution system had to be reconstructed. During the 2.5-year operation period, no odor was noticeable even close to the filter beds. Also, the City of Gainesville did not receive a single odor complaint during this period. More than $200,000 per year has been saved in chemicals that were originally used to provide alternative odor control systems for the plant. These cost savings have resulted in a oneyear payback on the biofilter system capital and operating costs (IPS, 1990). The Kanapaha biofilter experience suggests that routine monitoring and maintenance are necessary to ensure proper operation conditions for effective, long term control of H 2 S emissions. It was recognized that further research was needed to solve existing problems, such as progressive system acidification, accumulation of sulfur in the filter medium and the eventual decline of H 2 S removal efficiency. Appropriate laboratory studies have been conducted at the

PAGE 197

183 University of Florida to address these problems and the results are discussed in detail in Chapter 4 of this dissertation.

PAGE 198

CHAPTER 7 SUMMARY AND CONCLUSIONS The intention of this research was to develop a quantitative knowledge of the operation of a microbial biofilter system for removal of hydrogen sulfide from waste gas streams and to optimize and maintain the performance of such a biofilter system using such knowledge. Optimization of the system involved quantitatively determining the design parameters, the operating parameters and predictive relationships for the control efficiency. Maintaining the system involves recognizing system deterioration and upset and providing solutions for prevention of long term irreversible deterioration. A lab scale biofilter tower system was constructed and extensive experimental work was conducted to achieve the stated goals. In addition, a full-scale compost biofilter bed system for control of H 2 S emissions in a wastewater treatment plant was evaluated during long term operation. Based on the results of this study, it is concluded that: 1. Significant pressure drops in biofilter materials are mainly caused by the presence of small particles, particularly those with diameters less than 1 mm. These small particles are generally composed of sand and minerals, as well as decomposed and mineralized 184

PAGE 199

185 compost material. Small particles have a low water holding capacity and are less valuable to the overall biof iltration process. It is recommended that small particles be separated from the compost by sieving before use. To minimize pressure drop effects the filter should not be compacted unduly. Aged compost, which contains a larger fraction of fine particles, due to mineralization and fracture, should not be reused after change-out, unless it is specifically treated to remove the fine particle fraction by sieving or washing. 2. The time reguired for the oxidation of H 2 S to sulfate by microorganisms is a few seconds, however, the use of high gas velocities is not recommended since they will cause uneven gas distributions and high pressure drops . 3. The concept of maximum H 2 S elimination capacity of compost and H 2 S loading rate is very important in terms of system design and operation. When working within the maximum elimination capacity of the system, the waste gas flow rate can be adjusted to obtain the best reduction of H 2 S for various inlet concentrations . 4. Low compost water content is fatal to the biological process. A minimum value of 30% water content by weight is reguired for proper operation, but 40 to 60% is recommended.

PAGE 200

186 The microorganisms in the biofilter system are mesophiles. The optimum temperature range for these organisms to be efficient is from 30 to 40 °C. A H 2 S removal efficiency of 50% can be achieved when the temperature range is extended from 10 to 80 °C. High concentrations of sulfate are toxic to the microbial flora. The critical level is between 30 and 40 mg-S/g dry compost. Above this range, the biological activity of the microorganisms may be significantly inhibited and result in reduction of the H 2 S elimination capacity of the compost. Variation in Compost pH showed no effect on H 2 S removal efficiency for values greater than 2.0 pH units. However, lower pH values for the filter bed will cause serious corrosion problems. Biof iltrat ion is a comprehensive process which involves physical, chemical and biological processes. Ottengraf's model was adopted and has proved to be successful in describing the macro kinetics of H 2 S removal. Kinetic models and equations have been developed and determined to be appropriate and accurate in quantitatively describing removal of H 2 S by the biof iltration process. Composts from different sources have been demonstrated to be excellent media for biofilters used in H 2 S removal from waste gas streams. Their unique

PAGE 201

187 properties provide good buffering capacity to various operational impacts. H 2 S removal efficiencies of 99.9+% were routinely achieved in both the laboratory and full scale operations. The decomposition rates of composts are described by multi-stage first-order kinetics. The decomposition rate as well as biological activity of the compost are significantly enhanced by the presence of H 2 S. Firstorder reaction coefficients were determined which can be used to guantitatively predict the useful life of the compost. Acidification and accumulation of sulfur, especially sulfate, is a natural feature of the H 2 S oxidation process. Continuous formation of H 2 S0 4 results in significant decline in pH and serious corrosion of the construction materials. A protective washing procedure was developed to mitigate this feature and keep the system operational at its optimized conditions. Water is determined to be the best choice for elimination of sulfate. NaHC0 3 solution is recommended for pH corrections. System upset is identified by compost dry-out, high sulfur content in compost accompanied by a yellowishwhite deposit, and extremely low pH (<2) values of compost. Specific procedures have been developed to recover the activity of the defective filter material.

PAGE 202

188 Although hydrogen sulfide has been selected as a test gas for this research, the results obtained through this study should provide for an improved quantitative understanding of the principles of biof iltration . Controlling variables in the operation and effectiveness of biof iltration waste gas control technology may be applicable not only to hydrogen sulfide but also to other sulfurcontaining compounds, and more generally, to air toxics and VOCs. Dissemination of this basic information on the multiple advantages of biof iltration, including low cost, high destruction-efficiency, energy conservation, ease of operation and maintenance and universal application will provide environmetal engineers and federal, state and local government air pollution control officials with a viable alternative in controlling emissions of air toxic compounds from commercial and industrial sources. Biof iltration control of waste gas streams is a relatively unknown and little explored control technology in the U.S.. It has the potential, however, for wide-spread application and acceptance because of its relative simplicity and low capital and operating cost, in addition to its great potential for indiscriminate effectiveness in controlling multiple pollutants.

PAGE 203

REFERENCES Aaronson, S. Experimental Microbiology . Academic Press, New York, 1970. Allen, E.R.; Hartenstein, H.U. and Yang, Y. "Review and Assessment of the Design and Operation of a Compost Biofilter System for Odor Control," Final Project Report, Environmental Engineering Sciences Department, University of Florida, Gainesville, Florida, 1987a. Allen, E.R. ; Hartenstein, H.U. and Yang, Y. "Identification and Control of Industrial Odorous Emissions at a Municipal Wastewater Treatment Facility," Paper # 87-95A.4, Presented at 80th Annual Meeting of APCA, New York, June 21-26, 1987b. Allen, E.R.; Yang, Y.; Hartenstein, H.U. "Odor Producing Agents, Their Sources and Control", Final Project Report, Environmental Engineering Sciences Department, University of Florida, Gainesville, Florida, 1987c. Allen, E.R.; Yang, Y. and Hartenstein, H.U. "To Provide Technical Assistance in the Design and Installation of An Odor Control Biofilter System at the Kanapaha Wastewater Treatment Plant," Final Project Report, Environmental Engineering Sciences Department, University of Florida, Gainesville, Florida, 1989. Allen, E.R. and Yang, Y. "Biof iltration Control of Hydrogen Sulfide Emissions," Paper # 91-103.10, presented at the 84th Annual Meeting of the Air & Waste Management Association, Vancouver, BC Canada, June 16-21, 1991. American Public Health Association (APHA) , Standard Methods for the Exami nation of Water and Wastewater . 17th ed. American Public Health Association, Washington, DC, 1989. Atlas, J.E. and Bartha, R. Microbial Ecology: Fundamentals and Applications, Addison-Wesly Publ. Company, Reading, MA, 1981. Bethea, R.M. ; Murthy, B.N. and Carey D.R. "Odor Controls for Rendering Plants," Environ. Sci. Tech. 7: 504 (1973). 189

PAGE 204

190 Biddlestone, A.J. ; Gray, K.R. and Day, C.A. "Composting and Straw Decomposition," in Environmental Biotechnology . Forster, C.F. and Wase D.A.J. Eds., John Wiley & Sons, New York, 1987. Blake, G.R. and Hartge, K.H. "Particle Density," in Methods of Soil Analysis. Part I. Physical and Mineraloqical Methods, Klute, A. Ed. 2nd ed, American Society of Agronomy, Madison, Wisconsin, 1986a. Blake, G.R. and Hartge, K.H. "Bulk Density," in Methods of Soil A nalysis. Part I. Physical and Mineraloqical Methods . Klute, A. Ed. 2nd ed, American Society of Agronomy, Madison, Wisconsin, 1986b. Bonn, H.L. and Miyamoto, S. "Soil as a Sorbent and Filter of Waste Gases," in Symposium on land for Waste Management f Tomlinson, J. ed. , Natl. Research Council Canada, Ottawa, 1973. Bohn, H.L. "Soil and Compost Filters of Gases," J. Air Pollution Control Assoc., 25: 953 (1975). Bohn, H.L. "Compost Scrubbers of Malodorous Air Streams," Compost Sci., 17: 5 (1976). Bohn, H.L. "Soil Treatment of Organic Waste Gases, " Chapter 24, in Soils for Management of Organic Waste and Waste Waters, ASA-CSSA-SSSA, Madison, WI, 1977. Bohn, H.L.; Bohn, R.K. "Soil Bed Scrubbing of Fugitive Gas Releases," J. Environ. Sci. Health, A21: 561 (1986). Bohn, H.L. and Bohn, R.K. "Biof iltration of Odors from Food and Waste Processing," Proceedings of Food Processing Waste Conference, Georgia Tech Research Institute, Sept. 1-2, 1987. Bohn, H.L. and Bohn, R.K. "Soil Beds Weed Out Air Pollutants," Chemical Engineering, pp. 73-76, April 1988. Bohn, H.L. "VOC Removal by Soil Biofilter Beds," Proceedings of Hazmacon'89, Hazardous Materials Management Conference and Exhibition, Vol.2, Association of Bay Area Governments, April 18-20, 1989. Brock, T.D. and Madigan, M.T. Biology of Microorganisms r Fifth Edition, Prentice Hall Inc., Englewood Cliffs, New Jersey, 1988. Carlson, D.A. and Leiser, CP. "Soil Beds for the Control of Sewage Odors," Journal of the Water Pollution Control Federation, 38: 829 (1966) .

PAGE 205

191 Caunt, P. and Hester, K.W. "A Kinetic Model for Volatile Fatty Acid Biodegradat ion during Aerobic Treatment of Piggery Wastes," Biotechnology and Bioengineering, 34: 126 (1989) . Chen, K.Y. and Morris, J.C. "Kinetics of Oxidation of Aqueous Sulfide by 0 2 ," Environmental Science & Technology, 6: 529 (1972). Cooper Jr, H.B.H. "Kinetics of Inorganic Sulfur Oxidation during Black Liguor Oxidation with Oxygen," Tappi 57: 13 0 (1974) . Danielson R.E. and Sutherland, P.L. "Porosity," in Methods of Soil Analysis, Part I. Physical and Mineraloaical Methods, Klute, A. Ed. 2nd ed, American Society of Agronomy, Madison, Wisconsin, 1986. David, M.B.; Mitchell, M.J. and Nakas, J. P. "Organic and Inorganic Sulfur Constituents of a Forest Soil and Their Relationship to Microbial Activity," Soil Sci. Soc. Am. J. 46: 847 (1982) . Delaune, R.D.; Reddy, C.N. and Patrick Jr., W.H. "Organic Matter Decomposition in Soil as Influenced by pH and Redox Conditions," Soil Biol. Biochem. 13: 533 (1981). Dharmavaram, S. "Biof iltration A Lean Emission Abatement Technology," Paper # 91-103.2, presented at the 84th Annual Meeting of the Air & Waste Management Association, Vancouver, BC Canada, June 16-21, 1991. Don, J. A. "The Rapid Development of Biof iltration for the Purification of diversified Waste Gas Streams," in VDI Berichte 561 ; VDI Verlag, Dusseldorf, 1985. Eitner, D. and Gethke, H.G. "Design, Construction and Operation of Bio-filters for Odor Control in Sewage Treatment Plants," Paper # 87-95A.6, Presented at the 80th Annual Meeting of Air Pollution Control Association, New York, New York, June 21-26, 1987. Eitner, D. "Biofilter in Flue Gas Cleaning: Biomasses, Design, Costs, and Applications," Brennst . -Waerme-Kraf t (German). 41/3, 124., 1989. Ergas, S.J.; Schroeder, E.D.; Chang, D.P. "VOC Emission Control from Wastewater Treatment Facilities Using Biof iltration, " Paper # 91-105.4, presented at the 84th Annual Meeting of the Air & Waste Management Association, Vancouver, BC Canada, June 16-21, 1991. Ferguson, P. A. Hvdroaen Sulfide Removal from Gases. Air and Liquids, Noyes Data Corporation, Park Ridge, NJ. , 1975.

PAGE 206

192 Forster, C.F. and Wase, D.A.J. "Biopossibilities: The Next Few Years," in Environmental Biology , Forster, C.F. and Wase, D.A.J, eds, Ellis Horwood Limited Publishers, Chichester, England, pp. 439-448., 1987. Frechen, F.B.; Kettern, J.Y. "Reduction of Odorous Emissions from a Hazardous Waste Landfill Site Using Biof iltration and Other Techniques," Paper # 87-95A.3, presented at the 80th Annual Meeting of APCA, New York, New York, June, 1987. Freney, J.R. "Determination of Water-Soluble Sulfate in Soils," Soil Science 86: 241 (1958). Gale, P.M. "Decomposition of Organic Waste Products Under Aerobic and Anaerobic Conditions," PhD Dissertation, University of Arkansas, 1988. Grant, W.D. and Long, P.E. Environmental Microbiology . Halsted Press, New York, 1981. Gilmour, J.T.; Clark, M.D. and Sigua G.C. "Estimating Net Nitrogen Mineralization from Carbon Dioxide Evolution," Soil Sci. SOC. Am. J. 49: 1398 (1985). Greyson, J. Carbon. Nitrogen, and Sulfur Pollutants and their D etermination in Air and Water . Marcel Dekker, Inc., New York and Basel, 1990. Hack, P.J.F.M. and Habets, L.H.A. "Experience with FullScale Biopaq U. A. S . B. -Plants Treating Various Types of Effluent," in Environmental Technology. Proceedings of the second Europ ean Conference on Environmental Technology. Amsterd am. The Netherlands. June 22-26. 1987 . De Waal, K.J. A. and Van Den Brink, W.J., Eds, Martinus Nijhoff Publishers, Boston, 1987. Hartenstein, H.U. "Assessment and Redesign of an Existing Biof iltration System," Master's Thesis, University of Florida, Gainesville, Florida, 1987. Hartenstein, H.U. and Allen, E.R. "Biof iltration, An Odor Control Technology for a Wastewater Treatment Plant," Report to Department of Public Works, City of Jacksonville, Florida, 1986. Hsieh, Y.P. and Yang, C.H. "Diffusion Methods for the Determination of Reduced Inorganic Sulfur Species in Sediments," Limnol. Oceanogr. , 34: 1126 (1989). International Process Systems, Inc. (IPS) "Odor Control, Completing the Composting Process," 655 Winding Brook Drive' Glastonbury, CT 06033, 1990.

PAGE 207

193 Jennings, P. A.; Snoeyink, V.L. and Chian, E.S.K. "Theoretical Model for a Submerged Biological Filter," Biotechnology and Bioengineering, 18: 1249 (1976). Johnson, D.W. and Henderson, G.S. "Sulfate Adsorption and Sulfur Fractions in a Highly Weathered Soil Under a Mixed Deciduous Forest," Soil Science, 128: 34 (1979). Johnson, CM. and Nishita, H. "Microestimation of Sulfur in Plant materials, Soils, and Irrigation Waters," Anal. Chem. 24: 736 (1952). Kampbell, D.H.; Wilson, J.T.; Read, H.W.; Thomas, T. Stocksdale, T.T. "Removal of Volatile Aliphatic Hydrocarbons in a Soil Bioreactor," Journal of Air Pollution Control Association 37: 1236 (1987). Kuenen, J.G. 11 Colorless Sulfur Bacteria and Their Role in the Sulfur Cycle," Plant and Soil 43: 49 (1975). Lalazary, S.; Pirbazari, M. and McGuire, M.J. "Oxidation of Five Earthy-Musty Taste and Odor Compounds," J. Amer. Water Works Assoc. 78: 62 (1986) . Leson, G. and Winer, A.M. "Biof iltration: An Innovative Air Pollution Control Technology for VOC Emissions," J. AW&MA, 41: 1045 (1991). Lindstrom, K.P. " Air Toxic Emissions and POTWs," Workshop Report and Proceedings, Co-sponsored by WPCF and USEPA, Alexandria, VA, July 10-11, 1989. Miller, R.H. "Factors Affecting the Decomposition of an Aerobically Digested Sewage Sludge in Soil," J. Environ. Qual. 3: 376 (1974). Miller, R.D. and Johnson, D.D. "The Effect of Soil Moisture Tension on Carbon Dioxide Evolution, Nitrification, and Nitrogen Mineralization," Soil Sci. Soc. Pro. 644-647 . 1964. McGilvery, R.W. and Goldstein, G.W. Biochemistry. A Functional Approach. Third ed. ; W.B. Saunders Company, Philadelphia, PA, 1983. Moorhead, K.K. ; Graets, D.A. and Reddy, K.R. "Decomposition of Fresh and Anaerobically Digested Plant Biomass in Soil," J. Environ. Qual. 16: 25 (1987). National Research Council (NRC) , "Hydrogen Sulfide," Subcommittee on Hydrogen Sulfide, Committee on Medical and Biologic Effects of Environmental Pollutants. University Parck Press, Baltimore, Maryland, 1979.

PAGE 208

194 (NIOSH) "Criteria for a Recommended Standard for the Occupational Exposure to Hydrogen Sulfide," U.S. Department of Health, Education and Welfare, NIOSH, May 1979. O'Brien, D.J. and Birkner F.B. "Kinetics of Oxygenation of Reduced Sulfur Species in Aqueous Solution," Environmental Science & Technology, 11: 1114 (1987) . Ottengraf, S.P.P. " Theoretical Model for a Submerged Biological Filter," Biotechnology and Bioengineering, 19 : 1411 (1977). Ottengraf S.P.P. and Van Den Oever, H.A.C. "Kinetics of Organic Compound Removal from Waste Gases with a Biological Filter," Biotechnology and Bioengineering, 25: 3089 (1983). Ottengraf, S.P.P.; Van Den Oever, A.H.C. and Kempenaars, F.J. CM. "Waste Gas Purification in a Biofilter Bed," in Innovations in Biotechnology . Houwink, E.H. and Van Dan Meer, R.R. Eds, Elsevier Science Publishers B.V., Amsterdam (1984) . Ottengraf, S.P.P.; Meesters, J. J. P.; Van Den Oever, A.H.C. and Rozema, H.R. "Biological Elimination of Volatile Xenobiotic Compounds in Biofilters," Bioprocess Enqineerinq II 61 (1986). y Ottengraf, S.P.P. "Exhaust Gas Purification," in Biotechnology, Rehm, H.J. and Reed, G. Eds, Vol.8; VCH Verlagsgesellschaft. , Weinheim, 1986. Ottengraf, S.P.P. "Biological Systems for Waste Gas Elimination," TIBTECH 5: 132 (1987). Painter, D.E. Air Pollution Technology . Reston Publishing Company, Inc., Reston, VA, 1974. Parker, H.W. Air Pollution, Prentice-Hall, Inc., NJ, 1977. Patterson, Jr., A. and Thomas, H.C. A Textbook of Quantitative Analysis, Henry Holt and Company, New York, Paul, P.G; Castelijn, F.J. "Biof iltration~A Relatively Cheap and Effective Method of Waste Gas Treatment," in Environmental Technology. Proceedings of the second European Conference on Environme n tal Technology. Amsterdam. The Netherlands, June 22-26. 1987 . De Waal, K.J. A. and Van Den Brink, W.J., Eds, Martinus Nijhoff Publishers, Boston, 1987. Piscarcyzyk, K. "Odor Control with Potassium Permanganate," Presented at Ohio Water Pollution Control Conference Dayton, OH, June 16-18, 1982. '

PAGE 209

195 Pomeroy, R. D. "Biological Treatment of Odorous Air", J. of the Water Pollution Control Federation, 54(12): 1541 (1982). Prokop, W.H. and Bohn H.L. "Soil Bed System for Control of Rendering Plant Odors", Journal of Air Pollution Control Association, 35: 1332 (1985) . Rands, M.B.; Cooper, D.E.; Woo, CP. ; Flether, G.C. and Rolfe , K.A. "Compost Filters for H^S Removal from Anaerobic Digestion and Rendering Exhausts," Journal of Water Pollution Control Federation, 53: 185 (1981). Reddy, K.R.; Khaleel, R. and Overcash , M.R. "Carbon Transformations in the Land Areas Receiving Organic Wastes in Relation to Nonpoint Source Pollution: A Conceptual Model," J. Environ. Qual. 9: 434 (1980). Rittmann, B.E. and McCarty, P.L. "Variable-Order Model of Bacterial-Film Kinetics," Journal of the Environmental Engineering Division, Proceedings of the American Society of Civil Engineers, 104: EE5, 889 (1978). Robarge, W.P. and Fernandez I. "Quality Assurance Methods Manual for Laboratory Analytical Techniques," Prepared for the USEPA and USDA Forest Service Forest Response Program. Department of Soil Science, North Carolina State University. Raleigh NC 27695, and Department of Plant and Soil Science, University of Maine-Orono, Orono, ME 04469, July, 1986. Rotman, A. "Use of Biofilter in Odor Control," Unpublished Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 ReneLevesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991 a. Rotman, A. "Biofilter: Traditional Design, Summary of Technical Data as Based on German Projects," Unpublished Paper, Hydrogeo Canada, Inc., Lavalin Group, 1100 ReneLevesque Blvd. West, Montreal, Quebec, Canada, H3B 4P3, 1991 Roy, A.B. and Trudinger, P. A. The Biochemistry of Inorganic Compounds of Sulfur. University Press, Cambridge, 1970. Sawyer, C.N. and McCarty, P.L. Chemistry for Sanitary Engineers . McGraw-Hill, New York, NY, 1967. Schmidt, S.K.; Simkins, S; Alexander, M. "Models for the Kinetics of Biodegradat ion of Organic Compounds not Supporting Growth," Appl. Environ. Microbiol . 50 : 323 (1985). Sikora, L. and Sowers M.A. "Effect of Temperature Control on the Composting Process," J. Environ. Qual. 14: 434 (1985).

PAGE 210

196 Smittenberg, J.; Harmsen, G.W. ; Quispel, A. and Otzen, D. "Rapid Methods for Determining Different Types of Sulfur Compounds in Soil," Plant and Soil 3: 353 (1951). Sommers, L.E. ; Nelson, D.W. Terry, R.E., and Silvieria, D.J. "Nitrogen and Metal Contamination of Natural Waters from Sewage Disposal on Land," Tech. Rep. No 89, Purdue Univ. Water Resour. Res. Center, W. Lafayette, IN, 1976. Starkey, R.L. "Oxidation and Reduction of Sulfur Compounds in Soils," Soil Sci. 101: 297 (1966). Stotzky, G. "Microbial Respiration," in Methods of Soil Analysis . Part 2 ; Black C.A. Ed, Agronomy 9: 1550 (1965). Sublette, K.L. and Sylvester, N.D. "Oxidation of Hydrogen Sulfide by Thiobacillus Denitrif icans : Desulfurization of Natural Gas," Biotechnology and Bioengineering, 29: 249 (1987) . Sweeney, D.W. and Graetz G.A. "Chemical and Decomposition Characteristics of Anaerobic Digester Effluents Applied to soil," J. Environ. Qual. 17: 309 (1988). Swift, E.H. Introductory Quantitative Analysis. Principles and Sele cted Procedures . New York Prentice-Hall, Inc., 1950. Tabatabai, M.A. "Sulfur," in Methods of Soil Analysis. Part 2. Chemical an d Microbiological Properties r Page, A.L. Ed., 2nd ed, American Society of Agronomy, Madison, Wisconsin, 1982. Taylor, J.M. ; Sikora, L.J.; Tester, C.F. and Parr., J.F. "Decomposition of Sewage Sludge Compost in Soil: II. Phosphorus and Sulfur Transformations," J. Environ. Quanl. 7: 119 (1978). Terasawa, M. ; Hira, M. and Kubota, H. "Soil Deodorization Systems," BioCycle 27: 28 (1986). Terry, R.E.; Nelson, D.W. and Sommers , L.E. "Carbon Cycling During Sewage Sludge Decomposition in Soils," Soil Sci. Soc. Am. J. 43: 494 (1979a) . Terry, R.E.; Nelson, D.W. and Sommers, L.E. "Decomposition of Anaerobically Digested Sewage Sludge as Affected by Soil Environmental Conditions," J. Environ. Qual. 8: 342 (1979b). Tester, C.F.; Sikora, L. J. ; Taylor, J.M. and Parr, J.F. "Decomposition of Sewage Sludge in Soil: I. Carbon and Nitrogen Transformations," J. Environ. Qual. 6: 459 (1977).

PAGE 211

197 Tester, C.F.; Sikora, L.J; Taylor, J.M. and Parr, J.F. "Decomposition of Sewage Sludge in Soil: III. Carbon, Nitrogen, and Phosphorus Transformations in Different Sized Fractions," J. Environ. Qual. 8: 79 (1979). Urone, P. "The pollutants," in Air Pollution . Stern, A.C. ed. Vol.6, Academic Press, New York, 1986. USEPA, "Odor and Corrosion Control in Sanitary Sewage Systems and Treatment Plants," EPA-Design Manual, EPA/625/185/018, 1985. Van Lith, C. "Design Criteria for Biof ilters, " Paper # 89165.5, Presented at 82nd Annual Meeting & Exhibition, AW&MA, Anaheim, CA June 25-30, 1989. Varanka, M.W. ; Zablocki, Z.M. and Hinesly, T.D., "The Effect of Digestion Sludge on Soil Biological Activity," J. Water Poll. Cont. Fed. 48: 1728 (1976). Walker, G.S.; Lee, F.P. and Aifa, E.M. "Chlorine Dioxide for Taste and Odor Control," J. Amer. Water Works Assoc. 78: 84 (1986) . Water Pollution Control Federation (WPCF) , "Odor Control for Wastewater Facilities," Manual of Practice No. 22, Water Pollution Control Federation, Washington, DC, 1979. White, A., P. Handler, E.I. Smith, R.L. Hill, I.R. Lehman, Principles of B iochemistry , sixth Edition, McGraw-Hill Book Company, New York, 1978. Wieder, R.K. and Lang, G.E. "An Evaluation of Wet Chemical Methods for Quantifying Sulfur Fractions in Freshwater Wetland Peat," Limnol. Oceanogr. , 30: 1109 (1985). Williamson, K.J. "The Kinetics of Substrate Utilization by Bacterial Films," PhD Dissertation, Stanford University. June 1973. Williamson, K. and Mccarty, P.L. "A Model of Substrate Utilization by Bacterial Films," J. WPCF 48: 9 (1976 a). Williamson, K. and Mccarty, P.L. "Verification Studies of the Biofilm Model for Bacterial Substrate Utilization," J. WPCF 48: 281 (1976 b) . Yang, Y. "Odor Emissions and Its Control in a Wastewater Treatment Plant with Industrial Sources," Master's Thesis, University of Florida, Gainesville, FL, December 1988.

PAGE 212

198 Yang, Y . and Allen, E.R. "Biof iltration Control of Odor Emissions in Wastewater Treatment Plants," Paper presented at the 201st National Meeting of the American Chemical Society, Atlanta, GA, April 14-19, 1991. Zhabina, N.N. and Volkov, I.I. "A Method for Determination of Various Sulfur Compounds in Sea Sediments and Rocks, " in Enviro nmental Bioaeochemistrv and Geomicrobioloqy . Krumbein Ed., Ann Arbor Sci., Ann Arbor, MI, 1978.

PAGE 213

BIOGRAPHICAL SKETCH Yonghua Yang was born on March 24, 1949, in Inner Mongolia, China, and attended local schools until completion of high school in 1968. He received his Bachelor of Engineering (equivalent) degree in chemical engineering from Dalian Institute of Technology, China, in 1977. After graduation, he worked for 8 years for the Environmental Protection Institute, Baotou Iron and Steel Corporation in China as an environmental engineer. He was accepted as a graduate student in fall 1986 and received his Master of Engineering degree in air pollution from the University of Florida, Gainesville, FL, in 1988. He continued graduate study to pursue the Doctor of Philosophy degree in the Environmental Engineering Sciences Department, University of Florida from 1989 through 1991. He was the recipient of the Axel Hendrickson Scholarship award from the Air & Waste Management Association (AW&MA) , Florida Section in 1990 and a graduate scholarship award from AW&MA, in 1991. 199

PAGE 214

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Eric R. Allen, Chairman Professor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Dale A. Lundgren £r Professor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Ben Koopman Professor of Environmental Engineering Sciences I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Konda R. Rede Professor of Soil Science

PAGE 215

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Professor of Agricultural Engineering This dissertation was submitted to the Graduate Faculty of the College of Engineering and to the Graduate School and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. May 1992 iLjUJQj. /L-oLt/ Winfred M. Phillips 7i Dean, College of Engineering Madelyn M. Lockhart Dean, Graduate School


47
lower H2S concentrations. For example, the C02 evolved at
H2S concentrations near 6,000 ppm is approximately 8.5 mg-
C02/g-C added, which is about 3.4 times that evolved when no
H2S is present. At higher H2S concentrations the increase
of C02 evolution with H2S concentration is reduced e.g.
when the H2S concentration is increased from 12,000 ppm to
32,000 ppm the C02 evolution increases only by about 17%, or
approximately 2 mg-C02/ g-organic matter.
It can be seen from Figure 3-5 that the C02 evolution
from the compost has a strong dependence on the H2S
concentration in the gas to which the compost is exposed. A
linear relationship is obtained when plotting C02 evolved as
a function of the square root of H2S concentration in the
gas, [H2S]0,5 for the range of H2S concentration less than
17,000 ppmv (Figure 3-6). The regression analysis result
for the best fit line is:
C02 = 2.62 + 0.082[H2S]0*5 (3-6)
where:
C02 = C02 evolved from compost, (mg/g of C added)
[H^S] = H,S concentration in the inlet gas stream,
(ppmv)
The correlation coefficient, for the variables is
0.9234.
Equation 3-6 quantitatively describes the effect of H2S
on the decomposition of composts. For example, C02
evolutions at [H2S] = 0 and [H2S] = 1000 ppmv are calculated


CHAPTER 1
INTRODUCTION
Hydrogen sulfide (H2S) is a highly toxic air pollutant
which has been identified in the list of 190 air toxic
substances in Title III of the 1990 Amendments to the Clean
Air Act.
Considerable amounts of H2S are produced in association
with industrial processes, such as petroleum refining,
rendering, waste water treatment, paper and pulp
manufacturing, food processing, and in the treatment of
"sour" natural gas and other fuels. Hydrogen sulfide is
frequently the main component of most observable odorous
emissions.
Hydrogen sulfide is an odorous gas, and its presence at
low concentrations is easily perceived and recognized due to
its characteristic odor of rotten eggs. Hydrogen sulfide is
perceptible to most people at concentrations in excess of
0.5 parts per billion (ppb) in air. Control of H2S
emissions is essential to protect public health and welfare
as well as to mitigate vegetation and material damage
problems.
Numerous processes involving physico-chemical
principles have been developed in order to effectively
remove hydrogen sulfide from air, waste gases and liquids
1


BIOFILTRATION FOR CONTROL OF HYDROGEN SULFIDE
BY
YONGHUA YANG
A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL
OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT
OF THE REQUIREMENTS FOR THE DEGREE OF
DOCTOR OF PHILOSOPHY
UNIVERSITY OF FLORIDA
1992
UNIVERSITY OF FLORIDA LIBRARIES


Compost pH
135
Cumulative Operation Time (Day)
3 TS21 + TS22 O TS23 a TS24 x Wash Water
Figure 5-3. pH changes of compost in different
sections of the biofilter with
operation time,
b) Tower #2, compost #17.


79
directly from the printout of the integrator. The Tracor
analyzer is periodically calibrated for H2S, methyl
mercaptan (MM) and dimethyl sulfide (DMS) with standards
purchased from National Speciality Gases, Inc.
Results and Discussion
Biofiltration is a process that involves physical,
chemical, and biological processes. Many variables, such as
temperature, compost water content, specific acidity of
compost, sulfate content in the compost etc., can affect the
function of the system. It is impossible to obtain optimum,
longer term performance from a biofiltration system without
an in depth understanding of the system properties and
proper control of important variables. Extensive
evaluations of the system properties have been conducted
during the course of this study. The results presented in
this section are divided into several subsections according
to specific investigations undertaken. Composts from
different sources (Table 3-1) have been used in the
laboratory studies. The physical and chemical properties of
these composts are summarized in Table 4-2. Applications of
each of these composts are mentioned in the corresponding
study subsections.
Pressure Drop
The energy consumption obtained in operating a
biofiltration system is primarily that required by the


10
Table 2-1. Physical and Chemical Properties of H2Sa.
Molecular Weight
34.08
Boiling Point, C
-60.2
Melting Point, C
-83.8 to -85.5
Vapor Pressure, -0.4C
10 atm
25C
20 atm
Specific Gravity (Relative to Air)
1.192
Auto Ignition Temperature, C
250
Explosive Range in Air, %
4.5 to 45.5
Odor Threshold, ppbv
0.47
a Source: USEPA, 1985.


31
compost but also is a function of the particle size of the
compost. The decomposition was observed to be directly
related to the carbon content in the compost.
Decomposition is affected by a number of environmental
conditions, for instance, pH, moisture content, and the
presence or absence of foreign chemicals (Miller and
Johnson, 1964; Terry et al., 1979a, b; Delaune et al.,
1981). In the application of biofiltration to control H2S
emissions, the compost filter material is subjected to
conditions that are quite different to that for land
applications of compost. In the former case, the compost is
exposed to a gas stream which may contain a variety of
chemicals, especially H2S, at various concentrations. The
presence of xenobiotics in the gas streams and filter
materials could change the population and composition of the
microorganisms in the compost or significantly affect their
metabolic processes. As a result, the decomposition rate of
the compost can be altered. Unfortunately, little
information can be found in the literature related to this
topic.
The objectives of this study were (i) to evaluate the
decomposition of four types of compost by determining the
CO2 evolution, and (ii) to investigate the effect of H2S at
various concentrations on compost decomposition. Such
information is valuable for biofilter design and for
justifying land disposal applications of the compost after
use as a biofilter medium.


68
Particle Density
Particle density of the compost is measured according
to Blake and Hartge (1986a).
Bulk Density
Bulk density of the compost is measured according to
Blake and Hartge (1986b).
Sulfur Analysis Methods
Intensive and detailed laboratory work has been carried
out on the analysis of sulfur compounds in order to obtain a
better understanding of the biochemical reactions involved
in the H2S oxidizing processes occurring in the biofilters.
The procedures for determining various sulfur compounds in
compost, in water, and in the waste gases are described in
this section.
Sulfur in Compost
The analysis of sulfur in compost includes the
determination of total sulfur (total-S) and fractionation of
the total sulfur into inorganic and organic constituents.
Many wet chemical procedures have been developed to
fractionate the total sulfur pool in sediments, soils, and
peat into its inorganic and organic constituent compounds.
Very little information, however, is available about such
analyses for compost. The sulfur analyses conducted in this
research include the quantitative determination of acid


I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Eric R. Allen, Chairman
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Dale A. Lundgren
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Ben Koopman ^
Professor of Environmental
Engineering Sciences
I certify that I have read this study and that in my
opinion it conforms to acceptable standards of scholarly
presentation and is fully adequate, in scope and quality, as
a dissertation for the degree of Doctor of Philosophy.
Q=>:
Konda R. Reddy
Professor of S<


H2S Loading Rate (g-S/m3-hr) HgS Concentration (ppmv) HgS Removal
Efficiency (%)
128
Cumulative Operation Time (Day)
Figure 51. Biofilter control of H2S during long
term operation,
c) Tower #3, compost #16.


63
shown in Figure 4-5, consists of a Teflon probe, a Teflon
filter and a sampling port connector. The Teflon probe is
made from a piece of Teflon tubing (6.35 mm (1/4") in
diameter), with 14 holes (1mm diameter) spaced evenly along
the probe length. The probe is installed in the towers in
such a way that all the holes are perpendicular to the
tower's normal axis. Thus, representative gas samples from
a cross section of the tower can be obtained. The Teflon
filter is used to block out any small particles and water
droplets which may be extracted during sampling. Gas
samples are obtained through the sampling port located on
the end of the assembly (see Figure 4-5) by a gas-tight
syringe. The sampling port is sealed by a rubber GC septum.
When taking a sample, at least three full syringes of
gas sample are wasted before the actual sample is taken for
analysis. This procedure will eliminate residuals of
previous gas samples remaining in the Teflon filter holder,
in the syringes and in the probe, as well as condition the
extraction system to the gas being sampled.
The gas samples are then diluted in 3-L Tediar sampling
bags by pure nitrogen (N2) to an appropriate concentration
within the calibration range of the analyzer. The gases in
the Tediar bags are thoroughly mixed by gently kneading the
bags and allowing them to sit for at least 10 minutes before
analysis. Most of the samples taken are analyzed within 2
hours.


157
Table 5-4. Effect of water washing on elimination of sulfate
in filter compost.
Samplimg
Location
Sulfate in Compost (mg-S/g)
Sulfate
Elimination
(%)
Before Washing
After Washing
T14
1.67
0.34
79.6
T13
6.71
6.16
8.20
T12
13.4
11.0
17.8
Til
57.1
33.9
40.6
Wash Water
0.00
5.25


113
the gas phase to the biolayer (See Figure 2-3).
The overall kinetic behavior observed in a biofilter is
a result of the interaction between mass transfer phenomena,
the microkinetics of the biological elimination reactions,
the residence time distribution of the gas flow, etc.. This
overall kinetic behavior is termed 1 macrokinetics by
Ottengraf and can be determined experimentally.
Ottengraf divided the macrokinetics in a biofilter into
two classes: first-order reaction and zero-order reaction.
He expressed the first-order reaction kinetics in a form
similar to equation 4-4 described previously. For zero-
order reaction kinetics, however, Ottengraf has
distinguished the following two situations:
1. At gas phase concentrations, C, above a compound
specific, critical concentration (Ccr^t), the film will
be fully saturated (Figure 2-3, Case I) and pollutant
elimination is limited by the biological activity in
the biofilm. This process is defined as reaction
limitation.
2. At concentrations less than Ccr^t, diffusion in the
biofilm will limit compound removal. The biofilm is no
longer fully penetrated (Figure 2-3, Case 2) and the
removal rate decreases with decreasing pollutant
concentration in the waste gas. This process is
referred to as diffusion limitation.
Ottengraf has also derived two equations which describe
the kinetics for either situation.
For the first situation, i.e. under reaction limiting
conditions, the kinetics expression is:
C/C0 = 1 K0H/C0Ug
(4-7)


66
E
A known amount of wet compost is weighed into a 50-mL
container. DI water is added to bring the liquid/solid
ratio to 10 (Robarge and Fernandez, 1986). The sample is
shaken for 30 minutes by a rotary shaker. Measurements of
pH are made by a calibrated Corning Model M245 pH meter,
which is accurate to 0.01 pH.
Total Carbon and Total Nitrogen
Finely-ground, oven-dried compost sample (<100 mesh)
are analyzed for total carbon and total nitrogen using a
Carlo Erba Model NA 1500 CNS Analyzer.
Water Soluble Phosphorus (WSP)
A known amount of wet compost (2.5 g dry weight
equivalent) is weighed into 50-mL centrifuge tubes. DI
water is added to the tubes to obtain a compost to liquid
ratio of 1:10 on a dry weight basis. These samples are
allowed to agitate for a period of one hour on a mechanical
shaker. The compost suspensions are then centrifuged at
6000 rpm for 15 minutes and filtered through Gelman 0.45
micrometer membrane filters. The filtered solutions are
acidified (pH<2.0) with one drop of concentrated H2S04 and
stored at 4 C until analyzed. The soluble reactive P (SRP)
in the filtered extract is determined colorimetrically
(APHA, 1989) using a Shimadzu UV-160 spectrophotometer with
1 cm path length at 880 nm wavelength.


146
System Upset and Recovery
System upset is indicated by a sudden decrease in H2S
removal efficiency, increased H2S concentration in the
effluent gas stream and noticeable objectionable odor. The
most common reasons for system upset are compost dry-out and
H2S overloading.
A dry compost system can be easily determined by
measurement of compost water content. As discussed
previously, if the compost water content dropped below 30%,
reduced H2S removal can be expected. For a closed system
without additional water supply a dry region is generally
observed in the upper portion (exit) of the biofilter.
Drying of compost also causes shrinking of the compost and
results in channeling. This particular feature is indicated
by a decreased pressure drop across the filter bed.
The dry-out problem can be solved by spraying water at
the exit (top) of the compost filter. Channels generally
disappear as a result of the compost volume expanding after
watering. The system may need a few days to rebuild the
microbial population and recover its optimum performance
depending on its original dryness, as described in previous
sections.
Another cause of system upset is overloading the system
with high H2S concentrations. The maximum H2S elimination
capacity for a filter medium depends on the nature of the
material and the operating conditions of the system.
Generally speaking, the maximum H2S elimination capacity of


103
Inlet Gas
Sampling Port
Figure 4-15. Schematic drawing of the experimental
arrangement for investigation of the
effect of temperature on H2S removal
efficiency.


120
kinetics according to equation 4-4 (Figure 4-19). The
regression formula is expressed as:
C = CQ 27.6t + 21.8 (4-9)
with a correlation coefficient R2 of 0.9907. The good
agreement obtained supports the conclusion that the reaction
can be described by zero-order kinetics. The value
obtained for KQ is 27.6 ppmv/s. This zero order coefficient
conresponds to a maximum H2S elimination capacity of 130 g-
S/m3-hr for the compost at the operating conditions
selected. It should be noted that the second and third data
points in Figure 4-19 are below the regression curve. This
deviation is mainly due to the H2S removal efficiency in the
lower portion of the filter has been effected by high
acidity and sulfate content in the compost in this portion
of the bed as a result of prolonged operation.
For lower inlet H2S concentrations (H2S < 200 ppmv),
the reaction appears to follow first-order kinetics as
expressed by equation 4-6 (Figure 4-20). The linear
regression formula is expressed as:
ln(C/CQ) = 0.31 0.57t (4-10)
In this case the value for k-^ is 0.57/s, which is
applicable for inlet H2S concentration of less than 200
ppmv.
When H2S concentrations fall in the intermediate range,
i.e., between 200 ppmv and 400 ppmv, the data cannot be


18
zone
Cal = High concentrations of air pollutants.
Cg2 = Low concentrations of air pollutants.
Figure 2-3. Biophysical model for the biological
filter bed. The concentration profiles
shown in the biofilm refer to: 1)
Reaction limitation, 2) Diffusion
limitation. (Source: Ottengraf, 1986,
p. 436).


ACKNOWLEDGEMENTS
I would like to express my sincerest gratitude and
appreciation to the following people who made this research
possible:
To Dr. E. R. Allen, the doctoral committee chairman,
for his foresight in support of this research, his
encouragement, guidance and invaluable input during the
course of this study and my graduate work.
To Drs. D. A. Lundgren, B. Koopman, K. R. Reddy and D.
P. Chynoweth for their interests in this research, helpful
suggestions and participation on my graduate committee.
To Dr. P. Urone for his friendship, kindness and
valuable suggestions.
To Mr. A. White and the Kanapaha Wastewater Treatment
Plant engineers for their assistance in the research on the
full scale biofilter system.
To Ms. Yu Wang for her help on compost analysis.
To Ms. S. Jordan for her help in construction and set
up of the Lab scale biofilter units.
To Mr. R. Vanderpool for his invaluable friendship, his
ideas and his help in all aspects of my work that have made
my years at the university much easier and so enjoyable.
To my wife, Li, for her continuing support,
encouragement, patience and understanding.
iii


60
Inlet Gas
Sampling Port
Figure 4-4.
Schematic drawing of column system #4.


r C (ppmv)
115
Reaction Time (sec)
Figure 4-19. Linear least squares regression
analysis for zero-order kinetics of H2S
oxidation in biofilter. Gas loading
rate: 224 m3/m2-hr, compost #17.


36
titrated with standard IN HC1. The C02 samples collected
from each of the control bottles are concomitantly titrated.
The CC>2 evolved for individual samples is calculated as
follows (Stotzky, 1965):
C02 = (B V)NE (mg) (3-1)
where:
B = volume of HC1 used to titrate the NaOH in the
controls to the end point, (mL);
V = volume of HC1 used to titrate the NaOH
remaining in the C02 collectors after
treatment to the end point, (mL);
N = normality of the HC1, (meq/mL);
E = equivalent weight, (mg/meq), for C02, E = 22
(mg/meq).
To investigate the effect of H2S concentration on
compost decomposition, one hundred grams of Compost #6 was
used as the test material. The experimental arrangement for
this test is similar to that for the compost decomposition
test with some minor modifications (Figure 3-2) Room air
is forced through a scrubber containing 4N NaOH to absorb
C02 from the air stream. The C02 free air is then saturated
by bubbling through DI water. Pure H2S is then mixed with
the pretreated air stream to obtain the test gas mixture
with the desired H2S concentration. The treated gas stream
is vented through the manifold, where it is split into four
sub-streams: one of these sub-streams is vented to a control
column (empty), and the other three streams to duplicates of
three compost columns. The gas is forced vertically through
the compost from bottom to top at a flow rate of 30 mL/min.


17
biofiltration processes (Ottengraf, 1986; Hartenstein, 1987;
Paul and Castelijn, 1987; Van Lith, 1989). In particular,
Ottengraf and coworkers have carried out systematic studies
delineating the overall process and have presented
sufficient experimental data to support the proposed model.
In biofiltration, evenly distributed waste gases are
forced through a biologically active material, such as soil,
peat or compost. Many of the pores of the filter material
particles are filled with water. Microorganisms are attached
to the particle surfaces to form a layer of film. This wet,
biologically active layer surrounding the particles is
called a biofilm. The biophysical model proposed by
Ottengraf for the biofilm is shown in Figure 2-3. The
mechanism of the biological process is derived from a
combination of physical, chemical and biological processes
that occur in the filter material and is related to two
processes in particular; sorption and regeneration. As
waste gases pass through the countless narrow pores of the
filter material, air contaminants as well as oxygen will
adsorb on the surfaces of the pores and dissolve in the
liquid phase of the wet biofilm. The absorbed and adsorbed
gases are quickly degraded by the biofilter's enormous
microbial population. In this way a concentration gradient
is created in the biofilter, which maintains a continuous
mass flow of the component from the gas to the wet biofilm.
Activity of the biofilter depends mainly on the
population of the microorganisms. Soil biofilters can


64
Biofilter Tower
Teflon Filter
Teflon Probe
Sampling Port
Plastic Union
Figure 4-5. Schematic drawing of the gas sampling
assembly.


51
lower than those reported by other researchers, who used
fresh composts mixed with soils. Five to ten percent of the
total-C in composts were decomposed during the 122 days
incubation period. Compost half life times of the order 3 to
6 years are estimated for the composts studied,
corresponding to loss 50% of their total-C due to
decomposition. A multi-stage first-order reaction sequences
is used to describe the decompositions. First-order reaction
rate coefficients have been determined.
Decomposition rates are significantly increased when
H2S is introduced to the compost. The half life of the
compost is significantly reduced as a result of increased
biological activity and C02 respiration. For example,
continuous exposure of compost to 1000 ppm H2S can result in
reduction of the half life of the compost from about 6 years
to 3 years due to enhanced microbiological activity alone.
The results suggest that added H2S was oxidized by the
sulfur oxidizing bacteria in the compost to sulfate.