Oxidation of carbon and nitrogen in the root-zone of emergent macrophytes grown in wetland microcosms

MISSING IMAGE

Material Information

Title:
Oxidation of carbon and nitrogen in the root-zone of emergent macrophytes grown in wetland microcosms
Physical Description:
xviii, 219 leaves : ill. ; 29 cm.
Language:
English
Creator:
Burgoon, Peter Smiley, 1956-
Publication Date:

Subjects

Subjects / Keywords:
Soil and Water Science thesis Ph. D
Dissertations, Academic -- Soil and Water Science -- UF
Genre:
bibliography   ( marcgt )
non-fiction   ( marcgt )

Notes

Thesis:
Thesis (Ph. D.)--University of Florida, 1993.
Bibliography:
Includes bibliographical references (leaves 206-218).
General Note:
Typescript.
General Note:
Vita.
Statement of Responsibility:
by Peter Smiley Burgoon.

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 030180663
oclc - 30335716
System ID:
AA00022703:00001

Table of Contents
    Title Page
        Page i
        Page ii
    Acknowledgement
        Page iii
        Page iv
    Table of Contents
        Page v
        Page vi
        Page vii
        Page viii
    List of Tables
        Page ix
        Page x
        Page xi
        Page xii
        Page xiii
    List of Figures
        Page xiv
        Page xv
        Page xvi
        Page xvii
        Page xviii
    Abstract
        Page xix
        Page xx
    Chapter 1. Introduction
        Page 1
        Page 2
        Page 3
        Page 4
        Page 5
        Page 6
        Page 7
        Page 8
        Page 9
    Chapter 2. Performance of vegetated submerged beds under batch-loading and continuous-flow conditions
        Page 10
        Page 11
        Page 12
        Page 13
        Page 14
        Page 15
        Page 16
        Page 17
        Page 18
        Page 19
        Page 20
        Page 21
        Page 22
        Page 23
        Page 24
        Page 25
        Page 26
        Page 27
        Page 28
        Page 29
        Page 30
        Page 31
        Page 32
        Page 33
        Page 34
        Page 35
        Page 36
        Page 37
        Page 38
        Page 39
        Page 40
        Page 41
        Page 42
        Page 43
        Page 44
        Page 45
        Page 46
        Page 47
        Page 48
        Page 49
        Page 50
        Page 51
        Page 52
        Page 53
        Page 54
    Chapter 3. Design of continuous-flow stirred tank reactors for simulation of anaerobic wetland environments
        Page 55
        Page 56
        Page 57
        Page 58
        Page 59
        Page 60
        Page 61
        Page 62
        Page 63
        Page 64
        Page 65
        Page 66
        Page 67
        Page 68
        Page 69
        Page 70
        Page 71
        Page 72
        Page 73
        Page 74
        Page 75
        Page 76
        Page 77
        Page 78
        Page 79
        Page 80
        Page 81
        Page 82
        Page 83
        Page 84
        Page 85
        Page 86
        Page 87
        Page 88
        Page 89
        Page 90
        Page 91
        Page 92
        Page 93
        Page 94
        Page 95
        Page 96
        Page 97
        Page 98
        Page 99
        Page 100
        Page 101
        Page 102
        Page 103
        Page 104
        Page 105
        Page 106
        Page 107
        Page 108
        Page 109
        Page 110
        Page 111
        Page 112
        Page 113
        Page 114
        Page 115
        Page 116
        Page 117
        Page 118
        Page 119
        Page 120
        Page 121
        Page 122
        Page 123
        Page 124
        Page 125
        Page 126
        Page 127
        Page 128
        Page 129
        Page 130
        Page 131
        Page 132
    Chapter 4. Estimation of O2 transport through scrirpus validus based on oxidation of carbon and nitrogen in anaerobic wetland microcosms
        Page 133
        Page 134
        Page 135
        Page 136
        Page 137
        Page 138
        Page 139
        Page 140
        Page 141
        Page 142
        Page 143
        Page 144
        Page 145
        Page 146
        Page 147
        Page 148
        Page 149
        Page 150
        Page 151
        Page 152
        Page 153
        Page 154
        Page 155
        Page 156
        Page 157
        Page 158
        Page 159
        Page 160
        Page 161
        Page 162
        Page 163
        Page 164
        Page 165
        Page 166
        Page 167
        Page 168
        Page 169
        Page 170
        Page 171
        Page 172
        Page 173
        Page 174
        Page 175
        Page 176
        Page 177
        Page 178
        Page 179
        Page 180
        Page 181
        Page 182
        Page 183
    Chapter 5. Summary and synthesis
        Page 184
        Page 185
        Page 186
        Page 187
        Page 188
        Page 189
        Page 190
        Page 191
        Page 192
        Page 193
        Page 194
        Page 195
        Page 196
        Page 197
        Page 198
        Page 199
    Appendix. Summary data tables
        Page 200
        Page 201
        Page 202
        Page 203
        Page 204
        Page 205
    Reference list
        Page 206
        Page 207
        Page 208
        Page 209
        Page 210
        Page 211
        Page 212
        Page 213
        Page 214
        Page 215
        Page 216
        Page 217
        Page 218
    Biographical sketch
        Page 219
        Page 220
        Page 221
Full Text












OXIDATION OF CARBON AND NITROGEN IN THE ROOT-ZONE OF EMERGENT
NACROPHYTES GROWN IN WETLAND MICROCOSMS





















By

PETER SMILEY BURGOON




















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 1993























At times one must bend until surely one will break

but instead of breaking something profoundly new and

unexpected manifests. Fears, blinding illusions, disappear in the light of new insights. Life becomes not just bending, but surrendering, dying, and being reborn. Hail! the joys and pains of life and death.

P. S. Burgoon





True end is not the reaching of the limit, but in a completion that is limitless.

Rabindranath Tagore














ACKNOWLEDGMENTS


First and foremost I would like to thank Ramesh Reddy for his indefatigable patience and willingness to provide opportunities for, and encouragement in my professional development. Dr. Koopman has my greatest thanks for helping refine the scope and direction of my research goals, and helping to question and understand my fundamental hypotheses. I would also like to thank Dr. Rao for his encouragement, insights, and honesty. I thank Dr. Graetz for his pleasant hospitality as I ransacked his lab and wore out the Dionex columns. Dr. Zoltek has been a reminder to stay pragmatic, which I appreciate.

I would like to thank Orlando and Ms. Yu for always

providing a friendly hand as I lugged N2 cylinders downstairs to the basement. It is difficult to thank each person who has contributed in small but significant ways, so I would like to thank all the people in the lab, for their friendship, help and interest in my work.

The field research in chapter 2 could not have been done without the assistance in experimental design and operations from Tom DeBusk and Michael Langston. The pleasant interactions and exchange of ideas with Tom and


iii








Mike made the weeks of camping in a trailer at the sewage treatment plant bearable and rewarding.

Although at times a graduate dissertation seems

entirely self-consuming, my respites from the front lines of academic rigors were possible due to my friendships outside of the lab. I would like to thank Jennifer and Flick for keeping a regular sitting schedule and sharing meals and Ginnan; thank God for her objective ears and support. Many thanks to John Tilton for his constant encouragement to visit and find rest at the Temple of the Universe. My sincere thanks and respect go to all involved at the Temple for their perseverance in, and practice of the sanctity of life.

I thank my parents for my foundations in perseverance and hard work, and their continued support in my education.

Thank God.





















iv















TABLE OF CONTENTS



ACKNOWLEDGEMENTS.................... . .. .. .. .. ...

LIST OF TABLES........................x

LIST OF FIGURES........................xiv

ABSTRACT.........................xix

CHAPTER 1 INTRODUCTION...................1

Statement of the Problem................2
Need for Research...................3
objectives.............................7
Dissertation Format..................9

CHAPTER 2 PERFORMANCE OF VEGETATED SUBMERGED BEDS
UNDER BATCH-LOADING AND CONTINUOUS-FLOW
CONDITIONS........................10

Introduction.....................10
Materials and Methods................13
Experiment I. Batch-load VSBs ..........13
Operation and experimental setup .... 13
Water sampling and analysis........16 Physico-chemical measurements .......18 Methane generation.................19
Experiment II. Continuous-flow VSBs and FA~s 20
Experimental setup...............20
Water sampling and analysis ........22 Hydraulic characteristics.........22
Statistical methods...............24
Results........................................24
Experiment I. Batch-load VSBs...........24
Temperature................24
oxidation-reduction potential .......27 CBOD removal in batch-load VSBs .... 27
Suspended solids...............33
Nitrogen removal....................36
Experiment II Continuous-Flow VSBs and FA~s 38
Ef fect of rooting matrix on BOD5 removal 38 Nitrogen removal...............41


V








Hydraulic characteristics..........41
Discussion ......................44
CBOD Removal..................44
Oxygen Transport ................47
suspended Solids ................50
Effect of Porosity on BOD Removal .. ......51
Conclusions.....................53

CHAPTER 3 DESIGN OF CONTINUOUS-FLOW STIRRED TANK
REACTORS FOR SIMULATION OF ANAEROBIC WETLAND
ENVIRONMENTS .....................55

Introduction .....................55
Oxidation-Reduction Potential ..........56
Reactor Design for Simulating Wetland
Environments. ...............60
Control of the oxidation-reduction
potential. .......................60
Hydraulic considerations in reactor
design ................62
Materials and Methods ................64
Experiment I. Design and Hydraulic
Characteristics of Continuous- Flow Stirred Tank Reactors with Anaerobic
Biofilms...................64
Reactor construction and operation .. 64
Experiment II. Steady State Culture of
Nitrate- and Sulfate-reducing Biofilms
with Scirpus validus ............72
Experimental setup. ............72
Bacterial inoculation...........73
Nutrient solution .............73
Plant growth. .......... .....80
Water sampling and analysis.... ......82 Total and microbial solids ..........85 Statistical methods............86
Results ..... *.*..*.*.*..*...................87
Experiment I. Design and Hydraulic
Characterization of CSTRs..........87 Hydraulics..................87
Chemical homogeneity ............87
Experiment II. Steady State Nitrate- and
Sulfate-reducing Cultures..........89
Experiment Ila. Steady state nitratereducing cultures ...........89
Experiment IIb. Steady state sulf atereducing cultures ...........99
Discussion. ..........................116
Design and HydraulicCharacterization of
C STR ...................116
Homogeneity................116
Steady state concentrations ........117

vi








Steady state Eh and pH.........118 Bacterial environment ..........120 Control of Eh and pH. ..........120
Plant growth and environmental factors 125
Comparison to natural and constructed
wetland systems. ..........127
Wetland simulation without soil?. .....128
Conclusions.....................129

CHAPTER 4 ESTIMATION OF 02 TRANSPORT THROUGH SCIRPUS
VALIDUS BASED ON OXIDATION OF CARBON AND NITROGEN
IN ANAEROBIC WETLAND MICROCOSMS ..........133

Introduction ....................133
Methods and Materials. ...............138
Theoretical Basis for Mass Balance on
Acetate..................138
Nitrate-reducing biofilms .........139 Sulfate-reducing biofilms .........140 Assumptions about anaerobic biofilms 140
Mass balance for acetate in anaerobic
biofilms ..................141
Estimation of Oxygen Transport Based on
Oxidation of NH4-N .............145
Water Sampling.................146
Water sampling for acetate mass balance 149
Water samples for 15NH4 oxidation and
.DNRA.................149
Sampling for methane..............150
Gas Sampling for 15 N2..............151
Dissimlatory Reduction of Nitrate to Ammonium 154 Total Solids and Bacterial Biomass .......154 Plant Analysis.................155
Statistical Methods ...............155
Results .. .. .. ..................155
Reactor Water Chemistry during*02 Transport
Studies..................155
Electron donors and acceptors .......155 Physico-chemical parameters: Eh and pH .156 Methane production. ...........156
Inorganic nitrogen. ...........160
Estimation of Oxygen Transport Based on
Acetate oxidation .................160
Estimation of 02 Transport Based on Oxidation
of 15 NH4..................166
Accumulation of 15N. ...........166
Estimation of 02 transport based on N
oxidation ................166
Mass Balance for 15 NH4 Addition.........166
Discussion......................169
Acetate Mass Balance to Estimate 02 Transport 169


vii








15 NH4 Oxidation to Estimate 02 Transport . 171
Critical Analysis of Methods for Estimating
02 Transport . . . . . . 172
Reliability of mass balance . . . 172
Dissimilatory nitrate reduction to ammonium
. . . . . . . 177
Plant p eduction . . . . . . 177
Estimated Oxygen transport . . . . 178
conclusions . . . . . . . . . 181

CHAPTER 5 SUMMARY AND SYNTHESIS . . . . . 184

Summary . . . . . . . . . . 184
Field Experiments . . . . . . . 184
Laboratory Experiments . . . . . 188
Synthesis . . . . . . . . . . 193
Field Studies . . . . . . . . 193
Laboratory Studies . . . . . . 195
Reactor design . . . . . . 195
Oxygen transport . . . . . . 196
Application of reactors for environmental
research . . . . . . . 198

APPENDIX SUMMARY DATA TABLES . . . . . . 200

REFERENCE LIST . . . . . . . . . . 206

BIOGRAPHICAL SKETCH . . . . . . . . 219


























viii














LIST OF TABLES


Table Page

1.1. Basic processes involved in constructed
wetlands used for wastewater treatment. . 4

2.1. Mean oxidation-reduction potentials in VSBs
and NVSBs receiving primary and
secondary wastewater (3 and 6 day hydraulic
retention times, respectively) ......... .. 25

2.2. First-order CBOD removal rate coefficients
for batch and continuous flow VSBs
(calculated for T = 200C, k20 =
k0/(l.06^(T-20)) (Reed et al. 1988). Values
in parentheses are sample standard
deviations. ... ............... 29

2.3. Methane production at 10 and 30 cm depths in
vegetated submerged beds receiving primary
and secondary wastewater effluent. . . 32

2.4. Mean effluent BOD and CBOD, and CBOD mass
removal rates from the vegetated submerged
beds (VSB) and floating aquatic macrophyte
ponds (FAM) receiving primary wastewater
effluent (sample standard deviations are in
parentheses) ..... ................. 34

2.5. Total nitrogen effluent concentrations and
mass removal rates for vegetated submerged
beds (with gravel and plastic rooting matrix) and floating aquatic macrophyte ponds (FAMs).
Standard deviations are in parentheses. . 39

2.6. Nominal retention times for Rhodamine WT dye
added to vegetated submerged beds (VSB) 6 and 12 months after application of wastewater had
begun (hydraulic loading rate was 10
cm/day) ...... ................... 42





ix








Table pagfe

3.1. Oxidation-reduction potentials for redeox
couples common in anaerobic wetland soils.
Standard cell potentials (Eh0) are compared
to Eh7 values representative of natural systems (corrected to pH = 7 at 25 CC). Standard cell potentials are adapted from Zehnder and Stum
(1988); values for natural systems are from
(Ponnamperuma 1972) .. ... ... ... ..58

3.2. Physical characteristics of continuous-flow
stirred tank reactors used to simulate
nitrate- and wetland environments. Two
different sized reactors were designed, each had the same basic configuration as in Figure 3.1. In the text, reactors will be referred
to by void volume (i.e. 2.2 and 4.9 L). .. 66

3.3. Treatments and experimental conditions during
an experiment with anaerobic nitrate- and sulfate-reducing (NRB and SRB respectively)
continuous-flow stirred tank reactors during
Experiment II ..................74

3.4. Hydraulic retention times (HRT) and mass
loading rates (mg/L*hr) of carbon (C), sulfur
as S04-S (S) and nitrogen as N03-N to the
continuous-flow stirred tank reactors during
four phases of operation. ............75

3.5. Nutrient solution pH for sulfate- and
nitrate-reducing reactors during four phases
of operation. Volumes of concentrated hydrochloric acid added to each 8 liter
carboy of nutrient solution are shown. 77

3.6. Basic composition of nutrient solution for
nitrate- and sulfate-reducing continuous-flow
stirred tank reactors. Note in Tables 3.8,
and 3.10 that influent concentrations of
acetate, sulfate, and nitrate change during
different phases of operation. .........79

3.7. Depth profile of oxidation-reduction
potential (Eh7) in two 2.2 L continuous-flow stirred tank reactors with sulfate-reducing
biofilms (n=3). No plants were in the
reactors. ....................91




x








Table p~age

3.8. Mean influent and effluent concentrations of
water quality parameters in 2.2 and 4.9 liter
continuous-flow stirred tank reactors with
nitrate-reducing biofilms during four phases
of operation (Table 3.3) .. ...........92

3.9. Total and f errous (Fe+2) iron concentrations
in nitrate- and sulfate-reducing continuousf low stirred tank reactors during four phases of operation. Experimental conditions during
phases are summarized in Table 3.3. Sample
standard deviations are in parentheses. .. 94

3.10. Mean influent and effluent concentrations of
water quality parameters in 2.2 and 4.9 liter
continuous-flow stirred tank reactors with
sulfate-reducing biofilm. during four phases
of operation (Table 3.3)............100

3.11. Summary of plants effects on pH and Eh in
nitrate-and sulfate-reducing continuous-flow
stirred tank reactors during four phases of operation. Analysis of variance (SAS 1987)
compared mean differences between treatments with and without plants, or before and after
plants were placed in reactors. Acetate,
S04-S, and N03-N concentrations were not
effected by plant treatments. .........113

3.12. Accumulation rates of total and bacterial
solids' in continuous-f low stirred tank
reactors (CSTRs) with nitrate- and sulfatereducing biofilms during four phases of
operation ....................115

4.1. Influent and effluent concentrations of major
electron donors and acceptorsin the nitratereducing reactors during two operational
periods .. ....................147

4.2. Influent and effluent concentrations of major
electron donors and acceptors in the sulf atereducing reactors during two operational
periods .. ....................148

4.3. Mean Eh, and pH of in continuous-f low stirred
tank reactors (CSTR) with nitrate- and
sulfate-reducing biofilms receiving low and
high acetate loading rates...........158


xi








Table page

4.4. Production of CH4 and carbon (C) loads to
continuous-flow stirred tank reactors (CSTR)
with nitrate- and sulfate-reducing biofilm
receiving low and high acetate loading
rates ........ .................... ..159

4.5. Accumulation of 15NH4-N due to dissimilatory
nitrate reduction to ammonium in continuousflow stirred tank reactors (CSTR) with
nitrate- and sulfate-reducing biofilms. The
nutrient solution entering the reactors
contained about 10 % Hoaglands solution, 620
75 mg/L acetate, and about 13 mg/L 15N03-N
which was 99 % 15N ..... .............. ..161

4.6. Percent acetate oxidized via anaerobic
respiration (N03-N ,Fe+3 and S04-S reduction, and methanogenesis) and/or incorporated into microbial biomass in continuous-flow stirred
tank reactors with nitrate-reducing biofilms 162

4.7. Percent acetate oxidized via anaerobic
respiration (Fe+3 and S04-S reduction, and methanogenesis) and/or incorporated into
microbial biomass in continuous-flow stirred
tank reactors with sulfate-reducing
biofilms ......... ............... 163

4.8. Calculated 02 transport through Scirpus
validus based on acetate mass balance in
continuous-flow stirred tank reactors (CSTR) with nitrate- and sulfate-reducing biofilms
receiving to high and low acetate loading
rates ........ .................... ..167

4.9. Increase in atom % 15N2 due to nitrification
and denitrification in two continuous-flow
stirred tank reactors with nitrate- and
sulfate- reducing biofilms. Reactors had no
plants (NPL) or were planted with Sciryus
validus (PL) ...... ................. 167

4.10. Mass balance on 15NH4-N added to the nitrateand sulfate-reducing continuous-flow stirred
tank reactors with and without Scirpus
validus ....... .................. .. 168




xii








Table joace

4.11. Summary of 02 transport estimated using C and
N oxidation (and accumulation of 02), from field and laboratory experiments comparing
wastewater treatment with and without plants.
02 transport is presented using two types of
units: i) per unit area and ii) per unit pore
volume ......................179











































xiii














LIST OF FIGURES


Figure page

2.1. Vegetated submerged bed plumbed to receive
batch loads of primary or secondary
wastewater effluent. Placement of Eh
electrodes and tubes for collecting water and
gas samples are shown. Water level was kept
2-3 cm below the top surface of the gravel. 15

2.2. Redox potentials (Eh) and water levels at 10
and 30 cm depths within vegetated submerged
beds filled with secondary wastewater
effluent ....... ................... 26

2.3. Carbonaceous biochemical oxygen demand (CBOD)
measured at 30 cm depth in batch-load vegetated
(Scirpus pungens) and nonvegetated submerged beds
receiving primary wastewater. a) Effluent CBOD
concentration; b) Mass removal of CBOD; c) Removal of soluble CBOD ..... ............... 30

2.4. Suspended solids (SS) and effluent flow rate
from batch-load vegetated and nonvegetated
submerged beds during a 15 minute drain
cycle, a) SS and effleunt flow rate; b) Flow
weighted SS concentrations ............ ...35

2.5. Ammonium nitrogen removal from batch-load
vegetated and nonvegetated submerged beds
receiving secondary wastewater. a) Effluent NH4-N concentrations; b) Total NH4-N removal; c) NH4-N mass removal rate (influent NH4-N =
25 mg/L) ....... ................... 37

2.6. Effluent CBOD concentrations in vegetated
submerged beds and floating aquatic
macrophyte ponds receiving a continuous flow of primary wastewater (hydraulic loading rate
= 20 cm/day, mean influent = 210 13 mg/L). 40





xiv








Figure page

2.7. Hydraulic characteristic curves for vegetated
(Sagittaria latifolia) submergeds beds (VSBs)
receiving a continuous flow of primary or
secondary wastewater (Hydraulic loading rate
= 10 cm/day). Rooting matrix in VSB was a)
gravel; b) Norpak plastic media ......... 43

2.8. Comparison of first-order removal-rate
coefficients from systems reviewed by Conley et al. (1990) to those calculated for batch
and continuous load experiments ......... 46

3.1. Exploded view of physical components of a
continuous-flow stirred tank reactor used to
simulate nitrate- and sulfate-reducing
wetland environments ... ............. ... 65

3.2. Tri-pak plastic media used as a surface for
attachment of nitrate- and sulfate-reducing
bacteria, and rooting matrix for Scirpus
validus grown in continuous-flow stirred tank
reactors ....... ................... 69

3.3. Detail of interior of continuous-flow stirred
tank reactor filled with buoyant Tri-pak
plastic balls. Retaining rod and vexar mesh were used to hold plastic 2-3 cm below water level, a) Cross section of reactor; b) Plan
view of vexar mesh and retaining rod ........ 70

3.4. Sulfide levels in the sulfate-reducing
reactors were controlled by maintaining low partial pressure of H2S, precipitation with
Fe+2, and adjustment of pH ........... .. 81

3.5. Hydraulic characteristic curves for the 4.9 L
continuous-flow stirred tank reactors (CSTR)
operated at three different recycle ratios. A
conservative tracer, Rhodamine WT, was added to the CSTRs. Experimental data are compared
to a theoretical curve for an ideal CSTR. 88

3.6. Depth profile of nonconservative elements in
4.9 and 2.2 L continuous-flow stirred tank reactors with sulfate-reducing biofilms a)
S04-S; b) acetate .... .............. 90





xv








Figure page

3.7. Oxidation-reduction potential and pH during
PHASE II in two 2.2 L continuous-flow stirred tank reactors with nitrate-reducing biofilms.
Scirpus validus was planted 27 days after
start-up. a) pH; b) Eh7. ... .......... 95

3.8. Oxidation-reduction potential during PHASE
III for two 2.2 L continuous-flow stirred
tank reactors with nitrate-reducing biofilms.
Scirpus validus was grown for the entire
period. . .... . . . .. .................. 96

3.9. Oxidation-reduction potential during PHASE IV
for a 4.9 L continuous-flow stirred tank
reactor with nitrate-reducing biofilms.
Scirpus validus was removed 27 days after the
start of PHASE IV. a) pH; b) Eh7. . . 97

3.10. Concentration of electron donor and acceptor
in two 2.2 L continuous-flow stirred tank
reactors with sulfate-reducing biofilms
during PHASE I. No plants were present in
the reactors. a) Acetate; b) S04-S ..... 101

3.11. Influent and effluent S04-S concentrations in two
2.2 L continuous-flow stirred tank reactors with
sulfate-reducing biofilms. Acetate loading rate was 9 mg/hr to reactors with and without plants
(PHASE III) ...... ................. 102

3.12. Influent and effluent concentrations of total
Fe and Fe 2 during PHASE II. Plants were
placed in reactors after 27 days. Data is
representative of Fe and Fe+2 concentrations
during PHASEs I, III, and IV ........... ...103

3.13. Concentration of total and soluble sulfides
in two continuous-flow stirred tank reactors
with sulfate-reducing biofilms. Levels of
H2S were maintained at low concentrations to
reduce toxicity to Scirpus validus. a). Total
soluble sulfides; b). soluble H2S ........ ..105

3.14. Oxidation-reduction potential and pH during
PHASE I in two 2.2 L continuous-flow stirred tank reactors with sulfate-reducing biofilms.
No plants were present in the reactors. a)
pH; b) Eh7 ...... .................. ..106



xvi








Figure oacre

3.15. oxidation-reduction potential and pH during
PHASE TI in two 2.2 L continuous-flow stirred tank reactors with sulfate-reducing biofilms.
Scirpus validus was planted 27 days after
start-up. a) pH; b) Eh7............108

3.16. oxidation-reduction and pH during PHASE III
in two 2.2 L continuous-flow stirred tank
reactors with sulfate-reducing biofilms.
Scirpus validus was planted for the entire
period, a) pH; b) Eh7............109

3.17. Oxidation-reduction potential and pH during PHASE
IV in a 4.9 L continuous-flow stirred tank reactor
with sulfate-reducing biofilms. Scirpus validus
was removed 27 days after start-up. . . 110

3.18. comparison of measured and predicted Eh
during two acetate loading rates in nitratereducing reactors. Predicted values were
calculated using the Nernst equation solved
with measured concentrations of N03-N,
atmospheric partial pressure nitrogen, and
pH. a) PHASE II; b) PHASE III . . . 123

3.19. Comparison of measured and predicted Eh
values for two acetate loading rates in
sulfate-reducing reactors. Predicted values
were calculated using the Nernst equation
solved with measured concentrations of S04-S1
HS-', H2S, and pH. a) PHASE II; b) PHASE 111. 124

4.1. Oxidative pathways for acetate which are
dominant (thick lines), accounted for (solid
lines) or assumed to be insignificant (dotted
lines) in an acetate mass balance in a
continuous-flow stirred tank reactor with
nitrate-reducing biofilms . . .. .. ..142

4.2. Oxidative pathways for acetate which are
dominant (thick lines), accounted for (solid
lines), or assumed to be insignificant
(dotted lines) in an acetate mass balance in
a continuous-flow stirred tank reactor with
sulfate-reducing biofilm .. ..........143






xvii








Figure page

4.3. Plexiglas cover placed over plants for
determining CH4 and 15N2 production in
continuous-flow stirred tank reactors with
N03-N and S04-S reducing biofilms .... ..... 152

4.4. Linear production of CH4 during a 3 hour
sampling period from continuous-flow stirred tank reactors with sulfate-reducing biofilms receiving about 30 mg/hr acetate. No CH4 was
measured during an loading rate of 9 mg/hr. 153

4.5. Influent and effluent NH4-N concentrations
from continuous-flow stirred tank reactors
with nitrate- and sulfate-reducing biofilms
receiving a low acetate loading rate. Dashed
lines indicate addition of 15NH4 (99 %) .... 157

































xviii














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 OXIDATION OF CARBON AND NITROGEN IN THE ROOT-ZONE OF
EMERGENT MACROPHYTES GROWN IN WETLAND MICROCOSMS


By

Peter Smiley Burgoon

May 1993

Chairperson: Dr. K. R. Reddy Major Department: Soil and Water Science

Use of constructed wetlands designed for wastewater treatment is becoming common throughout the world. Oxidation of carbon (C) and nitrogen (N) may be increased by

transport of oxygen (02) into the rhizosphere of aquatic plants and periodic draining of the wetland. Field studies were conducted to determine: i) the influence of plants and draining of batch-load vegetated submerged beds (VSBs) on the oxidation of C and N from wastewater, ii) the effect of hydraulic retention time (HRT) in VSBs on first-order CBOD removal rate coefficients and, iii) the efficiency of C removal in VSBs and floating aquatic macrophyte systems (FA~s). Plants had significant effects on C and N oxidation, however, after an 18 hour HRT there was no difference between VSBS with and without plants. Draining


xix








VSBs had no effect on oxidation of C and N. Greater than 90% CBOD removal occurred within 18 hours. The first-order CBOD removal rate model was not appropriate after 24 hours. At a hydraulic loading rate of 20 cm/day, VSBs removed more CBOD than FAMs. Oxygen transport in VSBs estimated from C oxidation was 28.6 g/m2 day, and for N oxidation 2.4 g/m2 day. These rates agree with published 02 transport rates.

Laboratory studies evaluated C and N oxidation in the

rhizopshere of plants grown in wetland microcosms. Nitrateand sulfate-reducing biofilms were established on plastic matrix in anaerobic continuous-flow stirred tank reactors with and without Scirpus validus. Steady state pH and Eh were maintained for at least 40 days in reactors. Redox potential was not affected by the presence of plants. An acetate mass balance accounted for reduction of N03-, S04-2, and Fe+3, and methanogenesis. Differences in acetate mass balance between reactors with and without plants, were used to estimate oxygen transport through the plants. Oxygen transport rate estimated from acetate mass balance was about 10 g 02/m2 day; and the estimated rate from 15NH4 oxidation was 0.5 g 02/m2 day. This reactor design for simulating anaerobic wetland environments may be a useful tool for studying plant/microbial interactions, and for studying treatment of wastewater in wetlands.





xx














CHAPTER 1
INTRODUCTION


Use of constructed and natural wetlands for wastewater treatment has increased immensely in the last decade. Constructed wetlands are being used for municipal, agricultural, and industrial waste treatment in the United States (Reed and Brown 1992; Conley et al. 1991) and around the world (Reddy and Smith 1987; Hammer 1989; Cooper and Findlater 1990; Cooper 1990; IAWQ 1992). These systems represent an inexpensive yet effective method for treating waste in rural environments or for ecosystem restoration. The types of designs and waste treatment systems are diverse. Successful, and experimental operations are scattered across the country. A few examples include: (1) use in coal mining areas where impacts of acid mine drainage with high concentrations of sulfate have been ameliorated with wetlands (Kleinmann and Girts 1987); (2) municipal waste treatment in the estuaries of Humbolt Bay, California (Gearheart et al. 1989); (3) septage waste treatment in artificial wetland ecosystems in a greenhouse on Cape Cod (Teal and Peterson 1993); and (4) a 1500 acre marsh being built for the restoration of Lake Apopka in Florida (Lowe et al. 1989). Constructed wetlands may be designed to provide


1








2

secondary, advanced or tertiary waste treatment (Reed et al. 1988). Accompanying benefits are enhanced wildlife habitat, flood amelioration and aesthetic beauty.

There are two basic types of constructed wetlands:

free- water surface systems (FWS), with flow of water over the surface of the wetland, and subsurface flow, in which water flows laterally through the root matrix. Vegetated Submerged Bed (VSB) refers to subsurface flow wetlands (Reed et al. 1988), this terminology will be used throughout this dissertation. In Europe and the United States, VSBs have been a popular treatment option for small community waste treatment (Cooper and Findlater 1990). More than 90% of the

VSBS in the United States treat less than 3800 m3/d (1 mgd); the FWS wetlands receive larger f lows, up to 76000 m3/day (20 mgd) (Reed and Brown 1992).


Statement of the Problem


A decade of research and operational experience have

provided a foundation for development of engineering design manuals in the United States (Reed et al. 1988; USEPA 1988; WPCF 1990) and Europe (Cooper 1990). However, in a current review of constructed wetlands in the United States, Reed and Brown (1992) note that for FWS and VSB wetlands, "the design approach . ranges from 'seat of the pants' empirical to more rational models from a limited data base"(p 778). Although design information is available for








3

BOD and nitrogen (N) removal, basic processes for removal of carbon (C) and N are poorly understood.

Due to limited information on removal processes, design engineers choose the most conservative approach to ensure adequate treatment. This eliminates data from systems that have above average performance, and which may actually represent the true capacity of a wetland system to remove pollutants. For example, Reed et al. (1988) suggest use of first-order BOD removal rate coefficients which are an order of magnitude smaller than reported coefficients (Burgoon et al. 1991; Conley et al. 1991). Further work is necessary to understand and optimize rate constants and other relationships that govern system performance (Reed and Brown 1992). Reed and Brown (1992) note that design models for N removal are based on "lumped data from FWS and VSBs and may be useful only as a first approximation"(p 779). Increased understanding of biochemical and physico-chemical processes functioning in the system will enhance design optimization and ensure both economic and environmental success of a constructed wetland.


Need for Research


Processes active in wetlands used for wastewater may be generalized as physical, chemical and biological (Table

1.1). Physical processes have dominant effects in systems receiving wastewater high in solids, such as










-P


-P


4J
4 ro tp 0,
r. ro $4
z 0 -10
0 Cq En r
-P 0 + 0 + g (d
p 0
-P w 4 m
ro r. 0 :3 p w
a) -14 u % f--q I (o o
> P4 en 0 C', u -11 -P
0 0 + m 600 0
4 10 H w M 4 -H V.
0 r. z 44 ro -P 0
44 fo (d >4 m 4 P
a) ol m4 0) 0 a)
o r-i z p T 0 04 H -P
4 04 (1) (1) Z ro EA rl 4-)
(d 6 FT4 4-) 0) D (d
g ro 'd u 'd N (0 4J En -H rd R
0 0) ro a) r-I N (d u 41 0 r.
rd 4 U > (n 0 0 z -A o (a U
rc$ r z b-H r-I w 9 u P Z H 4J lq z H
r. 0 a) (1) r. 0 r. 0 4) p (d 4 0 9
ro +. % 4 (d
U 04 rd fd EO 0 -r4 4J 4-J
V) 0 tp til "1 41 P 4-J r-i 0 Z + 0 tP
41 a) 0 -r-j p -r-j g (d m m 4 p w :e6 0 p
4-) M q 0 04 U P4 0 Z U (d X Z P4 Z 0

r-4 C14 M r-I N (Y) T-4 N m 19t
4-)

0 4
-P -1 W 0
U) 41 z 0) 4J P
rl 0 U) P 4-) J
0 :j Q) (d u
u 10 ro u 4 M a)
a) r P4 r--l
r P M 44 C 4 m U a)
.,4 0 0 O-H
Ea r4 r. ro N 9 En
Ea 0 0 4-4 9 Q) Z 44 -r-I (0 (0
0) r-i r-i 0 (d N 0 04 t3)
> U 4-) 9 4-J rq p p rd
r-j 0 tp M 0 (z C: 9 4J 0 4J 0 Q)
0 P 9 4-J H 4-) 0 0 F (d p p (1) (n
> P4 -H r 4J -r-I -r-I -r-I Z "I t) 0 4 w :1
r. a) 04 04 4-) 4J -r-I P4 4J W
.r-j U H r P r-I (d (d 0 P4 4J EO (d a)
.r-i (0 -H 0 u rc5 rc5 Z -r-i 9 z 0 0 4 0
rj) 44 P rd En a) f-i "A En ro (0 (0 4-J r-A rz
a) -1 -P Q) ro 4 X X 4-4 W 9 H -l 9 0) 0 0
rA 0 Ul M (d 04 00 0 f< ou P4 H --i P4 -P r(j
U) W
QU) r 4 r C4 4 C4 C4
0
p
P4 U)
0)
U U)
.H U)
U) (D
0
4
04 fu 0

4 H
H H 0 134
m 9 r-4
>4 0
4 r-l
P4
E-4








5

settled wastewater. In systems used for advanced and tertiary wastewater treatment soluble forms of C and N and phosphorus (P) are the pollutants of concern. These are most affected by chemical and biological removal processes.

Plants are the most conspicuous component of wetlands. A wide variety of herbaceous and woody plants permeate the natural wetland landscape. Polycultures of species are common in FWS wetlands, whereas monocultures of a few specific species are used in VSBs. In Europe, the plant of choice has been the cold tolerant, common reed, Phragmites australis. Although experiments by Gersberg et al. (1983, 1986), have shown Scirpus validus to be superior to P. australis and Typha latifolia for C and N removal in VSBs, all three plants are commonly used in VSBs around the world.

Plants assimilate nutrients into above- and belowground biomass, with as much as 33% of removed N stored in root biomass (Rogers et al. 1991). Young growing plants may assimilate up to 85% of wastewater N (Rogers et al. 1991). However, full-sized treatment systems must operate with mature plants, consequently nutrient uptake is significantly reduced. Generally, harvesting plant biomass is not a practical method for N removal (Crites and Tchobanoglous 1992).

The primary function of plants in constructed wetlands is to transport 02 into the rhizosphere. Gersberg et al. (1986, 1983) found that VSBs with plants removed








6

significantly more N than without plants. Many species are able to oxygenate their roots (Armstrong 1967; Armstrong and Armstrong 1988; Grosse 1989; Moorhead and Reddy 1988) however the ability is highly variable and well studied in only one species, Phragmites australis (Armstrong and Armstrong 1990a,b; Armstrong et al. 1988; 1990). The

capacity for 02 transport into the rhizosphere is well documented (Armstrong and Armstrong 1990a; Moorhead and Reddy 1988; Grosse 1988; Dacey 1980). However, Brix (1990)

has shown that no 02 leaks into the roots of P. australis grown in a constructed wetland. Specific rates of 02 transport which may be used in a rational design process are not available (Reed and Brown 1992; Cooper 1990).

The major source of 02 in the FWS wetlands is from reaeration at the exposed water surface (Reed and Brown 1992). Oxygen transport is considered less critical in FWS wetlands because water passes over the surface of the

wetland. The major source of 02 for treatment in VSBs is through the plants (Reed and Brown 1992). A fundamental

design consideration is the amount of 02 that leaks out of the roots, and is available to the aerobic hetero- and autotrophic microbes associated with the roots.

Environmental factors influencing 02 transport in P.

australis are sunlight, humidity, and wind. Thermo-osmotic pressure differentials induced by sunlight have been found to be the major driving force of convective flows of air








7

into the roots of many emergent aquatic plants (Grosse 1989; Armstrong and Armstrong 1990b). This mechanism provides 02 for root respiration. However, Mendelssohn et al. (1981) and McKee et al. (1989) found that root respiration is dependent on the oxidation-reduction potential of the soil.

This implies that 02 transport may not satisfy the respiratory demand of roots if conditions are highly reduced. If this is so, then wetlands may need to be designed to encourage specific redox environments for optimal treatment.


Objectives


The overall objective of this dissertation is to

estimate 02 transport through an emergent aquatic plant based on the oxidation of C and N in the root zone. To

evaluate the role of 02 transport in wastewater treatment, field experiments were designed to address the following questions:

M How do plants affect removal of C and N in VSBs

that receive primary and secondary effluent?

(ii) Does draining the VSB introduce enough air to

improve C and N oxidation in a VSB? is it more

effective than plants?

(iii) What are estimated 02 transport rates through plants

based on oxidation of C and N ?








8

objectives in experiments designed to address these questions were

Mi Operate VSBs in batch-load and continuous modes to

determine the effect of draining and filling on

bringing air into the rhizosphere.

(ii) Determine the first-order removal rates of CBOD in

the VSBs.

(iii) Estimate 02 transport based on removal of C and N in

the VSBs.

Laboratory experiments were designed to examine the

effects of anaerobiosis on 02 transport under controlled environmental conditions. Questions posed were Mi IS 02 transport through plants affected by the

anaerobic condition of the rhizosphere?

(ii) Can 02 transport through plants grown in anaerobic

environments be accurately estimated from C and N

oxidation?

The objectives in the experiments designed to address these questions were

Mi Design an anaerobic biofilm reactor with steady state

Eh and pH which simulates anaerobic wetland

environments.

(ii) Establish conditions in the reactors which are

conducive to growth of an emergent aquatic plant. (iii) Estimate 02 transport by Scirpus validus based on

oxidation of C and N in anaerobic environments.








9

Laboratory and field studies were based on the

following hypotheses: (i) Plants transport 02 into the root zone in excess of root respiratory requirements; (ii) excess 02 is available to aerobic hetero- and autotrophic bacteria; (iii) 02 transport is independent of the oxidation-reduction environment of the rhizosphere.


Dissertation Format


The progression of this research moves from field studies to the use of laboratory microcosms for a better understanding the results observed in field-scale systems. Chapter 2 presents field studies using VSBs to remove C and N from primary and secondary wastewater. Chapters 3 and 4 present the laboratory studies. Chapter 3 consists of two sections, the first covers reactor construction, and evaluation of complete-mix and steady state conditions in the reactors. The second section present results of establishing the anaerobic biofilms, and reactor chemistry.

Chapter 4 presents results of estimating 0. transport through the S. validus planted in the reactors. oxygen transport was based on (i) mass balance of C in reactors with and without plants and; (ii) oxidation of 15NH4. Chapter 5 is a summary of the dissertation followed by a discussion of the usefulness of the research for studying contemporary environmental problems.














CHAPTER 2
PERFORMANCE OF VEGETATED SUBMERGED BEDS UNDER BATCH-LOADING
AND CONTINUOUS-FLOW CONDITIONS Introduction


Constructed wetlands have become a popular waste

treatment alternative for small communities in the United States and Europe (Cooper and Findlater 1991; Hammer 1989; Reddy and Smith 1987). Ideally, they offer effective treatment with low initial capital and operational costs. A vegetated submerged bed (VSB) is a type of constructed wetland in which wastewater flows through the plant rooting media (via sub-surface flow) instead of over the surface of the wetland. Numerous studies have verified the effectiveness of VSBs in treating wastewater (Gersberg et al. 1983; Hammer 1989; Cooper and Findlater 1990) and have provided a data base for the development of design manuals (Reed et al. 1988; USEPA 1988; Cooper 1990). Design criteria developed for these systems have been based on changes in concentration of selected parameters such as biological oxygen demand (BOD), nitrogen (N), and phosphorus

(P), with little attention to processes functioning in the wetland. The design information for carbon (C) and N removal is further restricted by conflicting experimental


10








11

data from batch and continuous flow systems (Burgoon et al. 1991). Basic understanding of the processes functioning within VSBs that account for removal of C and N is critical for optimizing VSBs to attain maximum treatment efficiency.

Pilot-scale studies with batch-load VSBs provide a simple, rapid method for obtaining performance data which can be used to design wetlands for wastewater treatment. However, batch-load systems have yielded first-order BOD removal rate coefficients that are significantly larger than those obtained for field-scale continuous-flow VSBs (Wolverton et al. 1983; Reed et al.1988; Burgoon et al. 1991). Although BOD removal in VSBs is assumed to follow first-order kinetics, the validity of this assumption has not been demonstrated.

Alternate draining and flooding of the VSB during batch-loading may result in aeration of porewater and biofilms within the soil and gravel matrix (entrainment) (Seidel 1976; Burgoon 1989; Brix and Schierup 1990). Carbon and N oxidation and removal may be improved if significant amounts of 02 are available to the microbiota after draining and refilling of the VSB with wastewater. Availability of air will increase if enough 02 diffuses into the biofilm to significantly increase oxidation of C and N by heterotrophic and autotrophic bacteria. Diffusion of 02 into biofilms in trickling filters has been quantified with 02 microelectrodes (Bungay et al. 1969; Kuenen et al. 1986).








12

The amount of 02 saturating the biof ilm depends on the specific surface area of the gravel/soil matrix and plant roots, depth of the biofilm, microbial metabolism (Kuenen et al. 1986), and length of time that the VSB is drained.

Active transport of 02 into the root zone through

plants was shown to have significant effects on oxidation of C and N in ponds and wetlands (Reddy et al. 1990; Moorhead and Reddy 1988; Gersberg et al. 1983; Wolverton et al. 1983). Howes et al. (1984) concluded that mass flow of air into the soil matrix during tidal excursions in an estuary resulted in significantly greater oxidation of organic

matter than would occur due to transport of 02 through emergent macrophytes. The relative importance of 02 transport through plants versus entrainment of air in VSBs operated in batch-load has not been evaluated.

The objectives of this study are to determine: (1)

the effects of emergent aquatic plants on the oxidation of C and N in batch-load VSBs and, (2) the influence of hydraulic retention time (HRT) on first-order removal rate coefficients for continuous and batch-load VSBs. The hypothesis being tested was that batch-load operation (vs continuous flow) of a VSB will augment the oxidation of C

and N. If entrainment of 02 (due to drain and f ill operation) is an effective aeration mechanism in VSBs, a batch-load VSB should remove more C and N than a continuous flow VSB.








13


Materials and Methods


The constructed wetland systems were set-up outdoors, at the Reedy Creek Improvement District Wastewater Treatment Plant located in Lake Buena Vista, Florida. Description of experiments are as follows:

Experiment I. Batch-load VSBs

This experiment was designed to test the hypothesis that (i) batch-load operation of a VSB will introduce air into the root zone and consequently increase the rate of oxidation of C and N, and (ii) first-order BOD removal rate coefficients are different for batch and continuous flow VSBs.

Operation and experimental setup

The VSBs (6 x 1 x 0.4 m) contained river gravel

(diameter = 3 cm, porosity = 28%, specific surface area = 100 m2/m3) as the rooting medium. The six VSBs used in this study had previously received a continuous flow of primary settled wastewater (CBOD = 220 mg/L, SS = 70 mg/L) or secondary effluent (CBOD = 15 mg/L, NH4-N = 25 mg/L) from water hyacinth (Eichhornia crassipes [Mort] solm) ponds for two years (November 1987 to December 1989) prior to beginning these experiments. The emergent macrophyte, Scirpus pungens L. (swordstem bulrush) was established during a two year period in the VSBs. Prior to initiation








14

of this experiment, dead plant biomass was removed from the gravel surface. Four of the six VSBs were plumbed for batch-loading of wastewater (Figure 2.1). Two of the batch-load VSBs were left nonvegetated (NVSB) to serve as controls. A standpipe on the outside of the batch-load VSB was used to set the water level at 35 cm in the batchload VSBs. A #14 rubber stopper at the bottom of the pipe was used to drain the VSBs at regular intervals. Batch-load VSBs were operated for six weeks before experimental data was collected. Batch-load VSBs which received primary wastewater were drained and refilled after a three day retention time; VSBs which received secondary effluent were drained and refilled after six days. Two of the six VSBs remained in operation as continuous flow VSBs. One received primary wastewater at hydraulic loading rates of 10 cm/day (January 1, 1990 to June 1 1990) and 20 cm/day (June 1 to July 15, 1990); the other received secondary wastewater at a hydraulic loading of 10 cm/day. Use of two hydraulic loading rates of settled sewage allowed comparison of CBOD removal rate coefficients from the batch and continuous-flow VSBs at 0.5 and 1.0 day hydraulic retention times (20 and 10 cm/day hydraulic loading rates respectively).

At each sampling time, water loss due to

evapotranspiration (ET) was measured by monitoring the water level changes in the stand pipe. Experiments in VSBs with 3 day HRTs were terminated if rainfall occurred within the 3










5

0 u
44
cl
0 W CQ

p lu 124 0 w 14 rd 9

04
44 Q)
0 ro H 0 0 4 >
ro -P a) cn H
cn (D 0 a)
c > CD
w
(D Lcr CD3: 4-)
u 4 (v 4-)
CL
44
0 C 0
CY) 0
t5 > 4-)
:3 t -H 9 0


C U) c: p U W
0 It (d P
+ 0 L) 0 r-I (d 0)
ao 4-; a4 >
cr)
0) w to
c 4
0)
44-) P4
0 z g w
0) 04 co r-i 0 W 4J
04 r-I
0 c)) 44 tO 4-4
M rO 4-4 0 0
cy
-60 00 g p rd u
LO w 9 (a
Ea Q) -P M 44
C:
% -iF tP (d 4
c C: co
-P -P
F 0 4 th (U P4
0 m : 0 w -P
>4 r. w P -ri 4 4J (d -P -P (a ro u 4J 9 0 (D 0 r-i 0 b) U r-4 r-I w 0 0 w > w u q










Pr4








16

day HRT. Experiments on VS2Bs with 6 day HRTs were terminated when rainfall caused the VSB to flood.

When these experiments were initiated, the root mass of the Sciripus 'Pungens extended 10 to 15 cm deep into the gravel bed. Standing crop biomass in each VSB was determined by complete harvest of the VSBs at the end of the experiment.

Water sampling and analysis

All VSBs and NVSBs had sampling tubes (1.25 cm PVC) set at two depths, one in the root zone (10 cm) and the other below the root zone (30 cm) (Figure 2.1). Water samples were collected from March 1 to July 15, 1990, for the batch-load and continuous-flow VSBs. Influent samples were taken as the VSBs filled with wastewater. Batch-load VSBs receiving primary settled wastewater were sampled at t

- 6, 12, 24, 36, 48, and 72 hours; batch-load VSBs receiving secondary effluent were sampled once a day for six days. A hand-held suction pump was used to draw samples from the 10 and 30 cm sampling tubes. After several tube volumes had been displaced (about 200 mL), a water sample was collected

and analyzed for CBOD' (method 507, APHA 1989) within one half hour. Water samples were filtered through glass fiber filters (1 gm), frozen and analyzed within one month for




'CBOD = oxygen demand due to heterotrophic oxidation of organic matter, a nitrification inhibitor is added to Stop 02 demand of nitrifying bacteria.








17

NH4-N and (NO3 + N02)-N using the automated salicylate method (method 361.2 EPA 1983) and the automated cadmium reduction method (method 418 C, APHA 1989) respectively. All EPA check samples used to confirm methods for analyzing CBOD, NH4-N and (NO3 + N02)-N were within the accepted 95% confidence limits.

Suspended solids (SS) and flowrate of the effluent were measured at regular intervals throughout the 15 minute drain time. Flow rates were measured using a graduated cylinder and stop watch. Suspended solids samples were analyzed according to method 2540 D (APHA 1989). Concentrations of SS varied during drainage; therefore, flow weighted concentrations were calculated.

After four sets of CBOD removal data were obtained from VSBs receiving primary effluent, soluble CBOD removal rates were evaluated. Influent wastewater and pore water samples were filtered through 0.45 Am (used as a pre-filter) and 0.2 Am millipore filters for removal of particulate organic matter. Bacteria are removed by filtration through the 0.2 Am filter; therefore, an inoculant was used in the CBOD analysis. Inoculant was prepared using 5 mL of primary sewage effluent (filtered through a glass fiber filter) diluted in 1 L of deionized water. One milliliter of this solution was used to inoculate the 300 mL CBOD bottles. Loss of dissolved 02 in CBOD bottles with 1 mL of inoculant was less than 0.5 mg/L. Analysis of soluble CBOD enabled








18

estimation of the CBOD removed by sedimentation and straining of organic material within the gravel matrix. If sedimentation is a primary mechanism for CBOD removal we expect no change in soluble CBOD over the three day HRT.

First-order removal rate coefficients were calculated using Equation [2-1]. Coefficients were calculated, then converted to a coefficient expected at 200c using Equation [2-2] (Reed et al. 1988)


ln (Cf/Ci)
k= [2-1]
t

Cf = effluent CBOD, mg/L
Ci = influent CBOD, mg/L
t = treatment time within the VSB, days

kT
k20 = [2-2]
(1. 06~2)

T = water temperature within the VSB, 0c kT = first-order removal rate coefficient at T c. Physico-chemical measurements

Platinum electrodes for measurement of oxidationreduction potentials (Eh) were placed at 10 and 30 cm depths in the VSBs. Electrodes consisted of a 1 cm piece of platinum wire soldered to 16 gauge copper wire and a reference calomel electrode (Faulkner et al. 1989). All data are reported relative to a standard hydrogen half cell (Eh..,urm + 242 mV). Influent wastewater flowed through a 2.5 cm orifice as it entered the batch-load VSB and NVSB. The








19

steady stream of water splashed on the gravel as the tanks

were filled. Dissolved 02 of the influent wastewater was measured by placing a YSI 02 probe in a pool of water below the NVSB inlet as primary sewage effluent splashed into the NVSB.

Concentrations of S04-S and Fe in the influent

wastewater were analyzed since they may also be used as electron acceptors in the oxidation of organic matter. Water samples collected during two of the CBOD studies were filtered through 1 Am glass fiber filters, frozen, then

analyzed for S04-S within two weeks on a Dionex Ion Chromatograph (method 4110, APHA 1989). Iron was a concern

because FeCd3 was added to the raw sewage as it entered the waste treatment plant. Therefore, primary wastewater used in the VSBs had a high Fe content (10 20 mg/L total Fe). Oxidized and reduced forms of Fe in the influent were analyzed using the Phenanthroline method (3500-F3 D, APHA 1989). Oxidized and reduced forms of Fe were also measured immediately after the influent wastewater had splashed into the VSB to determine changes in oxidation of Fe+2 due to aeration. Iron measurements were made two times, both after the CBOD studies were finished. Methane generation

Two ports for sampling dissolved gases (Figure 2.1) were placed at 10 and 30 cm depths in the batch-load VSBs (Faulkner et al. 1989). The air space within the gas








20

sampling ports was purged with helium gas immediately after the VSB was filled with wastewater. After 12 or 24 hours about 8 mL of gas sample was pulled from the ports; 5 ml of each sample was stored in a vaccutainer. After each gas sampling, the ports were again purged with helium and sampled after 12 or 24 hours, for up to 72 hours. Gas samples were analyzed for OH4 using a Hewlett-Packard 5840A gas chroinatograph equipped with a 1.8 m x 2 mm I.D. stainless steel column packed with porpak Q (80 -100 mesh) and a thermal conductivity detector. Carrier gas (hydrogen)

flow-rate of 10 mL/min, injector temperature of 100'C, oven temperature of 40*C and the detector temperature of 1000C were used.

Experiment II. Continuous-flow VSBs and FAMs

This experiment tests the hypothesis that (i) C and N

removal in VSBs is not affected by the specific surface area or porosity of the rooting matrix, (ii) longer hydraulic retention times in floating aquatic macrophyte ponds (FAMs) will improve C and N removal relative to a VSB, and (iii) accumulation of solids in a VSB can be related to changes in hydraulic characteristic curves. Experimental setup

Experiment Iha. Four 6 m2 (6 x 1 x 0.4 m) VSBs were

used for comparison of two different rooting matrices. Two tanks were filled with 5 cm river gravel (same as experiment








21
I) and two were filled with 5 cm Nor-pak plastic packing2 (porosity = 90%, specific surface area 145 m2/m3). One set of two VSBs (1 gravel and 1 plastic rooting matrix) received a continuous flow of primary settled wastewater, while the other two VSBs received secondary effluent. Hydraulic loading rate of primary and secondary effluent was 10 cm/day. Hydraulic retention times in gravel and plastic filled VSBs were 24 and 80 hours respectively. These tanks were operated from November 1987 to December 1989.

Experiment IIb. This experiment compared treatment of primary effluent in two 6 M2 tanks planted with the floating aquatic macrophyte, pennywort (Hydrocotyle umbellata) to treatment in three 6 M2 VSBs (5 cm gravel) planted with the common arrowhead, Sagittaria latifolia (two tanks), and Scirpus punens (one tank). Hydraulic loading rate of the wastewater was 20 cm/day for a 1.5 month period (June 1, 1990, and July 15, 1990); hydraulic retention time was 12 hours. All five VSBs and FAMs previously received primary wastewater (10 cm/day) for a two year period. Results for these experimental systems operated with a hydraulic loading rate of 10 cm/day have been reported previously in DeBusk et al. (1990).

Seedlings of Sagittaria latifolia and Scirpus pungens were planted (11 plants/m2) during initial startup of the



2Donated by Jeager Inc., Spring, Texas.








22

VSBs (November 1987). Tanks without rooting matrix (FAMs) were stocked with Hydrocotyle umbellata. Periodic application of insecticide was needed to control aphids on the Sagittaria and Hydrocotyle. Water samplincr and analysis

Influent and effluent water samples for Experiment IIa were collected weekly for two years (November 1987-November 1989) and analyzed for BOD (method 507, APHA 1989), NH4-N, and (NO2 + N03)-N. Analytical methods for NH4-N and (NO2 + N03)-N were the same as used in Experiment I. When the hydraulic load was increased to 20 cm/day in Experiment IIb, influent and effluent samples were analyzed for CBOD (method 507, APHA 1985). Analysis of samples with and without nitrification inhibitor (CBOD and BOD respectively) revealed that BOD results were affected by nitrification in the BOD bottles. A CBOD/BOD conversion ratio was calculated (based on comparison of samples evaluated for BOD and CBOD) so that BOD data from VSBs receiving 10 cm/day could be compared to CBOD data from VSBs receiving 20 cm/day. Hydraulic characteristics

Fluorescent dye (Rhodamine WT) was used to characterize the hydraulic flow patterns within the gravel and plastic filled VSBs. Dye was added to the VSBs six months and one year after the VSBs were planted and began receiving wastewater effluent (10 cm/day). Fifty milliliters of dye (2400 mg/L) were added at the influent of the VSBs.








23

Effluent samples were collected at three hour intervals until the effluent fluorescence was less than 1 gg/L. Fluorescence was analyzed using a Turner Designs Fluorometer. Effluent concentration and sampling times were converted to their dimensionless equivalents using methods (C"CI diagrams) from Middlebrooks et al. (1982). Pore volume of the VSBs were calculated once (one year after receiving wastewater) by measuring the time required to fill a VSB when water flowed in at a constant rate.

Nominal retention time of the dye within the VSB was calculated using the following equation:



on 1(c/c,) ede[23

f (C/ Cj)dO

e, nominal retention time (dimensionless) o dimensionless time (time of sampling divided by
theoretical hydraulic retention time) C =effluent dye concentration, mg/L C, mass of dye added per tank pore volume, mg/L

Hypothesized changes in hydraulic patterns may be

estimated using the nominal retention time as an indicator of clogging in the VSB. Theoretically, as the pores fill with solids the nominal retention time should decrease. changes in nominal retention times should indicate changes in the hydraulic properties of the bed (due to accumulation of solids and/or root growth).








24

Statistical methods

comparisons were made using the Statistical Analysis Systems version 6.04 (SAS 1987). A Generalized Linear Model t-test was used for comparison of treatment means, unless stated otherwise. All differences are reported as significant in a rejection region of p 0.05.


Results


Experiment I. Batch-load VSBs Temperature

Plants shaded the gravel and consequently moderated

temperature variations of the batch-load VSBs. The largest variation in temperature recorded over a 72 hour period was from 22 to 300C in the NVSB, and 23 to 270C in the VSB. The mean temperatures within the VSBs and NVSBs during the batch-load studies were 25.0 1.2 and 25.6 2.10C respectively. VSBs receiving a continuous-flow of wastewater also showed small variation in temperature due to shading by the plants. Mean temperatures within the continuous-flow VSBs and NVSBs were 26.3 0.9 and 25.4

0. 60C respectively.

mean evapotranspiration (ET) was 0.9 cm/day. This is equivalent to a change in water depth of 3 cm/day. Evapotranspiration over a 6 day HRT resulted in as much as 45% water loss from the tank.










25







p 0 t I Io 0c
P4) LO It


5-40 Id~ 0C
rr

*H0 0 )

ZI- I uI I f
m- I H
9) 0 P W>v

Hd 0 +1 +1 fd I I u
N~C rd-H 4
o 0 sI co H

md~ 0 rl I I1+ I d I d

idO) rI .,I


cld0 rd 0 a

I ii 4 -)~

H4-) Cd 0 1 040>'
-P 1 +1 +1 +1 +1 I


0 U) X 0 1 H H- H H s-uH


4-) Iu
0 ) I +
-H p ~ IQ 41, P
0 )a) N
Urdo 4J t I H- N Cl % 4IJ
Q)mI N r- 0) N
*H + +1 +1 +1 0 rU) r. 4-) 4J 0 I 1t3
0 0h0 1 N (
-H~d (001 o H v -HC) UH
4 :PI I CN N CN 02

Id I- H C~
9H s-d 0 0I

(dC () >
ril) 4-)

r.a Wu0 04J
H~ ~ *H0-00)Id (1)4-)U) 4- (
tic a) p4a

-~0 *0 044
P + Im








26








400 14

300 12E
> 30 Cm 12c
S 200 ET..... E

0 100
0~
CC

-100
0
4 a
~' -200 U
o0
-300 2

-400 -0
3 24 48 72 96
Time, hours


















Figure 2.2. Redox potentials (Eh) and water levels at 10
and 30 cm depths within vegetated submerged
beds filled with secondary wastewater
effluent.








27

oxidation-reduction potential (Eh)

The Eh of the VSBs was affected by water loss due to

ET. High ET caused water levels to drop below the platinum electrode placed at the 10 cm depth, resulting in Eh values typical of oxidized conditions (Figure 2.2). These conditions resulted in high variability in Eh at the 10 cm depth in the VSB receiving primary and secondary effluent (Table 2.1). The Eh values were consistently less than -150 mV at the 30 cm depth in VSBs receiving primary or secondary effluent (Table 2.1). The Eh remained constant at the 30 cm depth (during the 3 day HRT) in batch-load VSBs and NVSBs despite greater than 95% removal of CBOD within one day.

The Eh in continuous-flow VSBs was less than -150 mV. Wastewater flow into the continuous-flow VSBs was greater than ET losses; therefore, the water level did not drop below the tip of the electrode placed 10 cm deep into the root zone. There was no difference in Eh at the influent and effluent end of the continuous-flow VSBs receiving primary or secondary effluent (data not shown). There was also no significant variation in Eh at 10 and 30 cm depths in the continuous flow VSB (Table 2.1). CBOD removal in batch-load VSBs

After 6 and 12 hours, CEOD was significantly lower in

the VSB than in the NVSB (Figure 2.3a). Within 6 hours, 78%

of the CBOD (64 4 g CBOD /m' day) in the VSB was removed compared to 55% (48 12 g CBOD /m2 day) in the NSVB. After








28

6 and 12 hours, CBOD was significantly lower in the VSB than in the NVSB (Figure 2.3a). However, within 18 hours the difference between the VSB and NVSB was not significant and CBOD had decreased from 220 mg/L to less than 20 mg/L in both VSBs and NVSBs. After 24 hours, greater than 95% of the CBOD was removed in both VSBs and NVSBs, with a mean removal of 21 0.3 g CBOD /m2 day (Figure 2.3b). These results suggest that plants improved CBOD removal in the VSBs only at HRTs less than or equal to 12 hours. There was no difference in CBOD removal between the samples taken at 10 (in root zone) and 30 cm (below root zone) depths.

Removal of soluble CBOD. Soluble CBOD of influent wastewater averaged 125 mg/L. Removal of soluble CBOD followed the same trends observed for total CBOD. After 6 hours, soluble CBOD concentration in the VSB was 33 mg/L, significantly lower than 66 mg/L, in the NVSB (Figure 2.3c). After 12 hours the soluble CBOD from the VSB and NVSB averaged 12 2 mg/L. The rapid removal of soluble CBOD implies that VSBs do not operate simply as physical filters for straining and sedimentation of influent organic material.

First-order CBOD removal rate coefficients. Firstorder CBOD removal rate coefficients were similar for batch and continuous-flow VSBs with a HRT of 1 day (Table 2.2). The results show that CBOD is removed rapidly within the








29

Table 2.2 First-order CBOD removal rate coefficients for
batch and continuous flow VSBs (calculated for T
200C, k20 = kt/(1.06A(T-20)) (Reed et al. 1988).
Values in parentheses are sample standard
deviations.

Hydraulic retention time, hours
Treatment 6 n 12 n 18 n 24 n

--- ------------------days-'------------------Batch-load
S. pungens 4.4 9 2.9 9 2.1 9 1.7 9
(0.8) (0.4) (0.1) (0.2)
No plants 2.2 9 2.8 9 2.7 9 2.4 9
(1.1) (0.4) (0.1) (0.5)
Continuous
S. pungens na na 5.0 24 na na 2.0 6 (0.9) (0.2)

n = number of data points na = data not available








30




250
a
S200 -- S. pungens
.. No plants
E 1500^
o 100
50

0
0 12 24 36 48 60 72
140
b
120 E 100

80
o
E 60
40
m 20 C 0
0 12 24 36 48 60 72
140
J C
D 120 C
E
E 100
0 80 C 60
2 40
LE 20

0
0 12 24 36 48
Time, hours



Figure 2.3 Carbonaceous biochemical oxygen demand (CBOD)
measured at 30 cm depth in batch-load
vegetated (Scirpus pungens) and nonvegetated submerged beds receiving primary wastewater.
a) Effluent CBOD concentration; b) Mass
removal of CBOD; c) Removal of soluble CBOD.








31
batch-load and continuous-flow VSBs (Figure 2.3); therefore, removal rate coefficients decreased as HRT increased (Table

2.2). Removal rate coefficients for continuous-flow VSBs were significantly larger than batch-load VSBs also operated at an HRT of 12 hours.

Effect of electron acceptors on OBOD removal. Electron acceptors (02, N03-, Fe 3+ 1So42-) entering VSBs and NVSBs were measured to determine their relative contribution to oxidation of organic matter in the VSB. Dissolved 0. content of the influent, after splashing into a NVSB was 2

mg/L. This was assumed as the 02 concentration of influent entering VSBs and NVSBs. At the sewage treatment plant, Fedl3 is added to the raw sewage prior to sedimentation in order to control odor and remove phosphorus. Analysis of primary sewage effluent entering VSBs and NVSBs showed that >98% of Fe was in reduced form (Fe12 ) as it entered VSBs and NVSBs. After splashing onto the gravel surface, approximately 13% of the Fe was oxidized. This is in agreement with an expected oxidation rate of Fe at pH = 6.8 in an aerated environment (Benefield et al. 1982). Nitrates were not detected in the influent; sulfates were less than

1.5 mg/L. Electron acceptors (02, N03-, Fe+3 and S04-2) entering the VSB via the influent were calculated to contribute to less than 2% of the total CBOD removal within VSBs. Therefore >98% of the CBOD was removed by a








32

Table 2.3 Methane production at 10 and 30 cm depths in
vegetated submerged beds receiving primary and
secondary wastewater effluent.

Rate CH4 increase

Primary Secondary

HRT, Depth, mmol CH4 mmol CH4
hours cm CH4 /hr cm2 CH4 /hr cm2

12 10 3.9 4.38 0.0 0.0
30 3.5 3.93 0.01 0.0
24 10 6.4 7.14 1.0 1.12

30 3.9 4.38 1.0 1.12
48 10 9.7 5.36 0.7 0.71
30 12.9 7.14 0.7 0.76
72 10 16.2 8.97 0.1 0.0
30 13.4 7.41 0.2 0.09

+Percent CH4 in sampling ports which were filled with helium and allowed to equilibrate with porewater CH4 for a period of 12 or 24 hours after each HRT.








33

combination of three possible mechanisms, 1) methanogenesis, 2) diffusion of air at the water interface, and 3) 02 transport through the aquatic plants.

Table 2.3 shows the rate of CH4 accumulation within the VSBs receiving primary and secondary effluent. There were

no consistent differences between rates of CH4 production at the 10 and 30 cm depths, within and below the root zone (respectively) of Sciryus puncgens. There was significant

production of CH4 within VSB receiving primary effluent after a 72 hour retention time even though >95% of the influent CEOD had been removed after 24 hours (Figure 2.3a). Methane production was less in the VSBs receiving secondary effluent. Production reached a maximum after 24 hours, then decreased to almost zero after 72 hours. Suspended solids

The effluent flow rate and the concentration of

suspended solids (SS) in effluent decreased during the 15 minute drain period (Figure 2.4a). The flow rate out of the tank varied due to the falling pressure head (initial water depth was 0.35 in). Flow-weighted concentrations of SS within the VSB and NVSB varied from 20 to 280 mg/L (Figure

2.4b). High SS concentrations were observed after about the 10th batch load. The very high peak (280 mg/L) was not aberrant, but was indicative of a cyclic pattern observed in the batch tanks during four months of previous operation. High concentrations of SS were also observed in the pore








34
Table 2.4 Mean effluent BOD and CBOD, and CBOD mass removal
rates from the vegetated submerged beds (VSB) and floating aquatic macrophyte ponds (FAM) receiving
primary wastewater effluent (sample standard
deviations are in parentheses).

Hydraulic loading rate, cm/day Treatment 10 10 20 10 20
BOD CBOD
--------- mg/L ---------- -- g/m2 day -VSB, gravel
S. latifolia 18 (10) 7 (4) 5 (1) 17 34
S. pungens 18 (4) 6 (1) 7 (1) 17 34
No plants 20 (7) 7 (1) na 17 na
VSB, plastic
S. latifolia 21 (12) 7 (4) na 17 na
FAM
H. umbellata 25 (10) 15 (1) 21 (7) 16 31

Influent CBOD and BOD concentration = 176 mg/L. +CBOD for the systems with a HLR of 10 cm/day was calculated using the following CBOD/BOD conversion ratios; S. latifolia = 0.374 0.106 (n = 6), a. pungens = 0.333 0.098 (n = 3), H.umbellata = 0.624 0.136 (n = 6). Data used for conversion ratios is shown in Appendix A-2.1. CBOD for 20 cm/day represent actual measurements. na = not available








35



250 1.2
A Suspended solids a

-- 200 -flow rate 1
E "
@- 0.8 8
2 150

-0.6
100a .A 0.4
")
A, 0
CD 50- u.
u,."- 0 .2
-A
........ ...... ..... .. ...

0 I I I 0
0 2 4 6 8 10 12 14 16 Time, minutes 350
b
300 Scirpus pungens
-j
0 No plants
E 250U)-
"0
S200

e150
CD

100
50
U)


50

0
1 2 3 4 5 6 7 8 9 10 11 12 13 Batch-load cycle (every 3 days)

Figure 2.4 Suspended solids (SS) and effluent flow rate
from batch-load vegetated and nonvegetated
submerged beds during a 15 minute drain
cycle, a) SS and effleunt flow rate; b) Flow
weighted SS concentrations.








36

spaces in the continuous-flow VSBs. However, SS were not washed out of the continuous-flow VSB resulting in effluent concentrations less than 30 mg/L. Nitrogen removal

VSBs receiving primary effluent. Effluent

concentrations of NH4-N in the VSBs (20.7 mg/L) and NVSBs (25.9 mg/L) were higher than the influent NH4-N (20.3 mg/L). This implies that although 95% of the CBOD was removed within 24 hours in the VSB, 02 transport through the plants was inadequate for significant oxidation of N during a three day HRT.

VSBs receiving secondary effluent. Effluent

concentrations of NH4-N were lower in the batch-load VSBs than NVSBs (Figure 2.5a). Concentration is a poor indicator of performance of the systems because ET loss from the VSB caused NH4-N concentrations to be higher than expected. Thus mass removal rates would be a better indicator of system performance. Total N removal increased throughout the 144 hour HRT (Figure 2.5b); however, the mass removal rate of N decreased (Figure 2.5c). If 02 transport through the plants was a constant process, N removal rate should also have remained constant throughout the 144 hour HRT. The removal rate remained relatively constant in the NVSB (Figure 2.5c). It is possible that low water levels restricted NH4-N removal by the plants. Although plant








37



28
28... ... Scirpus pungens
26-- No plants a
,,j 24S22
E
20 T
i18
z16
14
120 24 48 72 96 120 144 168

12
10
E
(D 8L -I.......
C2
C92 6 4



-z 1
0
1 0 -24 48 72 96 120 144 168

E
C
o0.8

>
0 0.6 ... T
............
E ~ .. ..,.,.
~.04 ... ......

E 0.2
z
I
z 0 24 48 72 96 120 144 168
Time, hours


Figure 2.5. Ammonium nitrogen removal from batch-load
vegetated and nonvegetated submerged beds
receiving secondary wastewater. a) Effluent NH4-N concentrations; b) Total NH4-N removal; c) NH4-N mass removal rate (influent NH4-N =
25 mg/L).








38
roots extended the entire depth of the tank, the portion in contact with the water decreased as ET increased. Experiment II Continuous-Flow VSBs and FAMs Effect of rooting matrix on BOD removal

Mean effluent BOD5 concentrations for VSBs with gravel and plastic rooting matrix were equal to or less than 20 mg/L (Table 2.4). Difference in specific surface area between the gravel and plastic rooting media (100 vs 145 m2/m3 respectively) had no significant effect on BOD5 removal at a loading rate of 18 g/m2 day (Table 2.4).

For the initial two years of operation, the BOD mass loading rate for the VSBs and FAMS was 18 g/m2 day. When the hydraulic loading rate to VSBs receiving primary sewage effluent was increased to 20 cm/day, the BOD mass load was 35 g/m2 day. At this higher loading rate, VSBs consistently produced a higher quality effluent than FAMs (Table 2.4). Effluent CBOD concentrations from VSBs were consistently below 15 mg/L, while effluent from FAMs was variable and often above 30 mg/L (Figure 2.6). CBOD mass removal from VSBs were also higher and less variable than CBOD mass removal from FAMs (Table 2.4).

These data indicate that the VSBs consistently produce an effluent low in CBOD at a loading rate of 35 g/m2 day. However, the front end of the VSBs became clogged and wastewater flowed over the first meter of the VSB surface








39
Table 2.5 Total nitrogen effluent concentrations and mass
removal rates for vegetated submerged beds (with gravel and plastic rooting matrix) and floating
aquatic macrophyte ponds (FAMs). Standard
deviations are in parentheses.


Hydraulic loading rate = 10 cm/day Primary Secondary
Treatment n mg/L g/m2 mg/L g/m2
day day
Influent 20 38.1 a --- 32.8 a --(5.8) (4.2)
Gravel
No plants 10 36.5 a 0.16 29.8 a 0.30
(4.3) (5.1)
S. pungens 20 34.3 b 0.37 27.9 ab 0.49
(3.1) (7.6)
S. latifolia 20 31.1 b 0.70 22.3 bc 1.05
(4.2) (6.3)
Plastic
S. latifolia 10 32.8 b 0.53 24.6 bc 0.82
(5.0) (6.8)
FAMs
H. umbellata 20 33.4 a 0.47 20.5 c 1.23
(2.7) (7.0)

Note: A General Linear Model t-test (LSD) was used for comparison of treatment means. Different letters in a column are significantly different (p < 0.05).








40





40
H. umbellata

S. latifolia

30 S. pungens



S20
o 0
o q

10
0
~ ~~~..... .. . ...: .


0 I I I I
0
150 160 170 180 190 200
Julian days



















Figure 2.6. Effluent CBOD concentrations in vegetated
submerged beds and floating aquatic
macrophyte ponds receiving a continuous flow of primary wastewater (hydraulic loading rate
= 20 cm/day, mean influent = 210 13 mg/L).








41

1.5 months after receiving 20 cm/day of wastewater; however, VSBs had been receiving primary sewage effluent for two years prior at a hydraulic loading rate = 10 cm/day. Although clogging did not affect removal of DOD or SS, it does imply that solids management is necessary to maintain flow through the VSB when a high DOD loading rate is applied.

Nitrogen removal

Rooting matrix (gravel or plastic) had no significant effect on total N concentration in VSBs receiving primary or secondary effluent (Table 2.5). Mass removal rate of N was higher for S. latifolia planted in gravel compared to plastic when primary or secondary effluent was applied to VSBs. Total N removal rate was higher when influent wastewater was low in CBOD.

Hydraulic characteristics

Hydraulic characteristic curves for VSBs filled with gravel and plastic rooting media (Figure 2.7a and 2.7b, respectively) indicate that neither operate as an ideal CSTR or a plug flow reactor. This is expected in a packed-bed reactor with hydrodynamic dispersion (mass flow and diffusion). The nominal retention time (Eqn [3]) of dye placed in VSBs are shown in Table 2.6. There are no apparent differences between VS~s receiving primary and secondary effluent after six months of receiving wastewater. After one year there were small differences between








42

Table 2.6 Nominal retention times for Rhodamine WT dye added
to vegetated submerged beds (VSB) 6 and 12 months
after application of wastewater had begun
(hydraulic loading rate was 10 cm/day).

Nominal HRT
Treatment Type of 6 month 12 month
effluent

S. latifolia primary 1.5 1.3
secondary 1.7 1.4
S. pungens primary 1.6 na
secondary 1.7 na
No plants primary 1.4 na
secondary 1.5 na
VSB, plasticV
S. latifolia primary 1.0 na
secondary 1.0 na
+ The nominal retention time of the dye in the VSB is expressed in units of dimensionless time (sampling time/hydraulic retention time). v Nor-pak plastic media na = not available








43







1
Plug flow a
a
0.8 --CSTR

Primary c0.6 E
0 0. Secondary


0
4

uo 0.2-'..
00 @ 00
.. .. ...--


0 1 2 3 4 5 6
1
0".8 CTR Plug flow b
~0.4
S0CSTR
o 0





0 0.2 -0
0
0.








0 ~........................... ............... ,
00







0 0.5 1 1.5 2 2.5
Retention time, (t/to)





Figure 2.7. Hydraulic characteristic curves for vegetated
(Sagittaria latifolia) submergeds beds (VSBs)
receiving a continuous flow of primary or
secondary wastewater (Hydraulic loading rate
= 10 cm/day). Rooting matrix in VSB was a)
gravel; b) Norpak plastic media.








44

hydraulic characteristic curves for VSBs planted with Sagittaria receiving primary or secondary effluent (Table

2.6).

Nominal retention times greater than 1 imply that there was adsorption of the dye onto the gravel and associated solids. Since the nominal retention time of the plastic filled VSB is close to 1 it may be inferred that plastic matrix and roots do not interact with the dye, whereas gravel and solids may have adsorbed dye.


Discussion


CBOD Removal

It is a common assumption that BOD removal in VSBs can be modeled with first-order kinetics (Reed et al. 1988; Wood 1990; USEPA 1990; Cooper 1990; Conley et al 1991). Curve fitting our data to a linearized first-order model gave best fit when HRTs were less than 18 hours. This implies that after 18 hours the assumption of first-order kinetics is not valid, and a different kinetic model is required. Although first-order kinetics were not operating after 18 hours, coefficients were calculated for all data collected for comparison with published data. The rate coefficients observed were higher (some by an order of magnitude) than those previously reported in design manuals for natural systems (Reed et al. 1989; USEPA 1988). The rate coefficients reported by Conley et al. (1991) for VSBs and








45

root zone systems agree closely with those observed in the present study (Figure 2.8). These results suggest that as the HRT increases, the removal rate coefficient decreases. Therefore, VSBs may be operated at shorter HRTs without a loss in CBOD removal efficiency. Shorter HRTs also imply that less land area is required to attain secondary effluent standards for CBOD removal.

Less than 2% of the CBOD was calculated to be oxidized by electron acceptors entering the VSB in the wastewater. Other mechanisms for the removal of CBOD are: (i) methanogenesis, (ii) 02 transport though the plants, and/or (iii) diffusion of 02 across the air water interface. An approximation of CBOD removal due to methanogenesis within VSBs receiving primary effluent is possible if, 1) the values in Table 2.3 are assumed to represent the rate at which CH4 diffuses into the atmosphere at the air water interface, and 2) the pore water is saturated with CH4 (32.8ml/L). Also, assuming that CBOD = COD, and 1 g COD/393 mL CH4, it was estimated that 54% of the CBOD removed within 12 hours was due to methanogenesis. These calculations may underestimate the CBOD removal by methanogenesis since CH4 transport through the plants was not accounted for and may represent 79% of the CH4 produced in a soil (Cicerone 1981). Fermentation should also significantly reduce the CBOD. The stable methanogenic redox conditions and steady production of CH4 implies that methanogenesis and fermentation were








46












Conley et al. 1991
4
o No plants
S. Pungens
3
Cz


2
A 0



---------------------------------------2 3 4
Time, days
















Figure 2.8. Comparison of first-order removal-rate
coefficients from systems reviewed by Conley
et al. (1990) to those calculated for batch
and continuous load experiments.








47

predominant pathways for oxidation of C. High removal rates of BOD have been reported for anaerobic gravel beds without plants. Young and McCarty (1969) reported a BOD removal

rate of 776 g BOD/m3 day in an anaerobic gravel bed with an HRT of 19 hours. The CBOD removal rate in the VSB after 12

hours was about 400 g CBOD/m3 day. Comparison to Young and McCarty's research (1969) implies that the VSBs may be capable of higher removal rates, independent of the presence of the plants.

Oxygren Transport

It has been hypothesized that drain and fill cycles

would introduce air into the pore spaces of a VSB, thereby enhancing oxidation of organic matter and nitrogen (Seidel 1976; Burgoon et al. 1991; Brix and Schierup 1990). Similar processes have been shown to occur in nature. For example, Howes et al. (1984) has shown that "tidal excursions" in an estuary enhance the oxidation of organic material in an esturarine marsh. During ebb tide, mass flow of air into

the pore space provides 02 which diffuses into the pore water and is consumed in microbial respiration and oxidation of organic matter. It is possible that a similar process may enhance waste treatment in a VSB. Brix and Schierup (1990) hypothesized that mass flow of air into the pore spaces of a rooting matrix will occur more rapidly as the pore space of the soil matrix increases. However, comparison of removal rate coefficients from this research








48

shows that the drain and fill process did not improve waste treatment in the VSB. Analysis of possible mechanisms

involved in the storage of 02 within the gravel during a drain and fill cycle shows why improvement in treatment is unexpected. Since the majority of pore spaces in gravel are large, very little water will remain in the VSB when it is drained (i.e. specific yield is high, Hillel 1982).

Therefore, storage of 02 in pore water retained within micropores, as occurs in the fine textured sediments of the tidal marsh, will occur to a minor extent in a gravel bed.

Another possible mechanism for storage of 02 in the gravel bed would be saturation of microbial biofilms with 02 while the VSB is drained. An estimate of the amount of 02 stored in biofilm can be made by assuming a 2 mm thick layer of biofilm covering the gravel. Assuming a water content of 95% in the biofilm, complete saturation (with dissolved 02) of the biofilm at 250c, and a specific surface area of 100 m2/m3 for the gravel, then less than 2% of the CBOD load

would be removed due to entrained 02. Therefore, both analysis of 02 transport mechanisms, and results obtained in this study show that batch-load processes do not improve oxidation of organic matter and N.

An estimate of 02 transport by the plants into the root zone can be calculated if we assume that the difference in C removal between the VSB and NVSB removal was due to 02 transport through Scirpus Puncrens. This difference which








49

will be referred to as the apparent 0. transport rate, is equivalent to 28. 6 g CBOD/ (e day) This value for CBOD removal is similar to the value of 20 g 02/mn2 day, that is suggested for removal of BOD in a VSB (Reed et al. 1989).

An enigmatic aspect of the results was that plants

significantly affected CBOD and N removal only in the first 12 hours of the treatment period; the effect of plants diminished as the HRT increased. After easily degradable OBOD was removed (95% was oxidized within one day), that 02 should be available for nitrification. However, there were

no changes in NH4-N or Eh over the 3 day HRT in the VSB or NVSB. The fact that the CBOD continued to decrease over the entire 72 hour HRT in the NVSB, and not in the VSB, also implies that organic compounds may be released by the plants. other researchers have shown significant differences in C and N removal between VSBs and NVSBs which

indicated 02 transport through the plants (Gersberg et al. 1983; Wolverton et al. 1983). In the present study nitrification was not observed in the VSBs receiving primary effluent with a HRT of 3 days. However, there were

significant differences in NH4-N removal due to nitrification in VSBs receiving secondary effluent.

Nitrogen mass balance in the VSBs receiving secondary

effluent shows that less than 10% of the NH4-N was assimilated into plants during the 144 hour HRT. If the remainder of the N was removed via nitrification and








50

denitrification in the root zone (Reddy et al. 1988), maximum 02 transport rate would be about 2.4 g 02/m2 day. This rate was calculated using the N removal after a 24 hour HRT (Figure 2.5c) and subtracting the N removal for the

control. This calculated 02 transport rate is less than that calculated from the OBOD removal in the VSB. The effect of plants on N removal also decreased as HRTs

increased. This method may underestimate 02 transport because it does not account for oxidation of organic matter. Carbonacaeous BOD was low; however, surrogate methods such as CBOD do not allow accurate accounting because organic material may be released from the roots. Suspended Solids

A common problem caused by accumulation of solids

within VSBs is clogging of pore spaces. Accumulation of solids within VSBs depends on solids loading rate, microbial growth rate, and sedimentation of solids. Operation of VSBs in drain and fill cycles resulted in periodic flushing of solids which accumulated within the VSB. Flushing of solids from the VSB during a drain cycle did not negatively affect subsequent treatment of wastewater. The accumulation of solids in an anaerobic gravel bed has also been reported by Young and McCarty (1969). Young and McCarty (1969) used a synthetic wastewater with low suspended solids; therefore, they suggested solids accumulation was due to biological synthesis. Our results shows that periodic draining of








51

continuous-flow VSBs may remove the accumulated solids and could extend the operating life of the VSB. Further research is needed to evaluate the effects on treatment and disposal of the high solids anaerobic waste effluent.

Accumulation of solids in the continuous flow VSB could not be assessed using hydraulic characteristic curves. A difference was expected between characteristic curves for VSBs receiving primary and secondary effluent due to different loading rates of BOD and SS. The VSBs receiving

primary wastewater had BOD and SS loads of 17.6 g BOD/m2 day and 5 g SS/m2 day compared to BOD and SS loading rates of

3.3 g BOD /M2 day and 0. 5 g SS/m2 day for VSBs receiving secondary effluent. Accumulation of solids should result in short circuiting of water over the surface of the VSB and a hydraulic characteristic curve increasingly similar to a complete-mix reactor. However, within one year it was not possible to use nominal retention times to assess accumulation of solids in the VSB. Further research is needed on the effects of SS and BOD loading and their affects on clogging of VSBs.

Effect of Porosity on BOD Removal

These results are similar to studies by Burgoon et al. (1991) which showed that within a range of 100 250 m2/m3, specific surface area does not have a significant effect on BOD removal within a VSB receiving BOD loading rates less

than about 15 gfm2 day. Due to the large dif ference in HRT








52
(80 versus 24 hours for the plastic and gravel respectively) the plastic filled VSB would be expected to remove more BOD than the gravel filled VSBs. These studies showed that BOD removal was not effected by HRT or specific surface area of the rooting media.

Current design manuals suggest that BOD removal is a function of the porosity of the rooting matrix (Reed et al. 1988; USEPA 1990). The following empirical equation for the first-order BOD removal rate coefficient was derived by Reed et al. (1988):

k20 = k, (37.31 n 4.172 [2-4]

k2o = first-order removal rate coefficient at 201c k. = first-order removal rate coefficient in field. n = porosity

The general applicability of this relationship has not

been tested for substrates with a wide range of porosities (such as the plastic and gravel media). The calculated first-order removal rate coefficient (using equation [2-4]) for the plastic rooting matrix is equal to 44.2 d-1 compared to a measured removal rate coefficient of 0.58 d-I (calculated from the experimental data with plastic media). This shows that in general, Equation [2-4] cannot be used to calculate removal rate coefficients based on porosity alone. Although this experiment compares two media with extremely different porosities (0.28 vs 0.90 for gravel and plastic respectively), the empirical relationship in Equation [2-43 may also fail if comparisons were made between gravel and a








53
fine textured soil (e.g. clay with a relatively high porosity but low hydraulic conductivity). The low hydraulic conductivity of the fine textured soil will restrict the flow of wastewater through the pores which would result in surface flow of wastewater (Bucksteeg 1987). Consequently DOD removal in a VSB with a fine textured soil would be related to surface area of plant stems not the specific surface area or porosity of the rooting matrix. These results imply that in general, porosity should not be used as a design variable for BOD removal in VSBs.


Conclusions


There were no significant differences between firstorder CBOD removal rate coefficients for batch-load and continuous flow VSBs operated with a HRT of one day. The assumption of first-order kinetics was valid only at HRTs less than one day. These data concur with other published literature and show that BOD removal rate coefficients for VSBs may be an order of magnitude larger than recommended in current design manuals. Although BOD removal was significantly higher after 12 hours in the VSB with plants, at longer HRTs plants did not improve the BOD removal.

Periodic draining and filling of the VSB to introduce air into gravel pore spaces did not have any effect on BOD or N removal. Regular draining of the VSB however did cause periodic flushing of solids from the systems.








54
oxygen transport rates through plants agreed with rates

(20 g/m2 day) recommended by Reed et al. (1988) for removal of BOD by VSBs. However the apparent effect of plants on BOD removal decreased as the HRT increased. Oxygen transport rates calculated for N were considerably lower than rates calculated for BOD removal in VSBs, and also decreased as the HRT increased. These results imply that 02 transport is not constant in time, and the rates based on C and N oxidation are significanlty different.

Two rooting matrices with different porosities (0.28 vs

0.90) and specific surface areas (100 vs 140 m2/m3) had similar effects on BOD removal. Porosity needs to be used carefully when used as a variable for VSB design.














CHAPTER 3
DESIGN OF CONTINUOUS-FLOW STIRRED TANK REACTORS FOR
SIMULATION OF ANAEROBIC WETLAND ENVIRONMENTS


Introduction


During the past decade there has been a growing

interest in the use of constructed and natural wetlands for wastewater treatment and water quality improvement (Reddy and Smith 1987; Hammer 1989; Cooper and Findlater 1990). Successful design, construction and use of these systems requires an understanding of the biogeochemical cycles active in the wetlands. oxygen demand within a wetland soil

is generally much higher than diffusion of 02 through the overlying water. Therefore, molecular 02 is depleted within the first few centimeters of soil depth. Consequently wetland soils are characteristically anaerobic.

The rates of cycling of carbon (C), nitrogen (N), and sulfur (S) depend on the type of microbial populations present in the wetland. The metabolic needs of these organisms depend on the flux of electron acceptors and donors into the wetland. Types of microbial populations which function in the wetlands are; aerobes, facultative anaerobes, and obligate anaerobes. The bacteria live in a matrix of soil organic and mineral matter permeated by roots 55








56
of wetland plants. The metabolism and physiology of wetland plants are affected by the extent of anaerobiosis in the soil (Mendelssohn et al. 1981; Mckee et al. 1989).

Wetland plants adapt to soil anaerobiosis by pumping 02 into the root zone. Oxygen in excess of the root respiratory requirements may diffuse into the rhizosphere and be used by autotrophic and heterotrophic bacteria in the oxidation of C and N (Reddy et al. 1990; Moorhead and Reddy 1988; 1989). Oxygen transport through emergent wetland plants has been verified (Armstrong 1967; Dacey 1981; Grosse and Mevi-Schutz 1987; Armstrong and Armstrong 1990; Armstrong et al. 1990). However, anaerobic root respiration has been shown to be dominant in some wetland plants (Mendelssohn et al. 1981; McKee et al. 1989). These studies have shown that amounts of the fermentative enzyme alcohol dehydrogenase (ADH) in roots are related to the oxidationreduction (redox) potential of the soil. Oxidation-Reduction Potential

The most common parameter used to characterize wetland soil anaerobic conditions has been the oxidation-reduction potential (Eh). Redox potential is a measure of the ease with which electrons are transferred from an electron donor to an acceptor. The common donors in soil are organic compounds released during decomposition of plant and animal material. The common electron acceptors in a soil are 02, N03, Fe+3, S042, and CO2. The amount and types of electron








57

acceptors depends on the organic and mineral content of the soil, and the contents of the hydraulic flows entering the wetland. The driving force for the exchange of electrons between donors and acceptors is the tendency of the free energy of the system to decrease until at equilibrium, when the free energies of the products equals that of the reactants (Ponnamperuma 1972). The driving force can be measured in calories or volts. The change in free energy for the reduction reaction, Ox + ne-+ mH-" = Red, is given by AG =AGO +RT ln (red) [3-1]
nF (Ox) (H+)m

where (red) and (Ox) are the activities of the reduced and

oxidized species, and AG. is the free energy when activities are unity (Snoeyink and Jenkins 1980). Converting calories to volts using the relationship AG = -nEF, and measuring Eh with a standard hydrogen electrode, Equation [3-1] becomes

Eh = E0 RT ln (Red) + xnRT ln (HI+) [3-2]
nF (Ox) nF

Equation [3-2] shows that Eh is a function of the activities of the reduced and oxidized species, and the pH of the solution. The standard electrode potentials of the reduction reactions of common electron acceptors in wetlands are given in Table 3.1. The redox couples with the highest free energy are the strongest oxidants. The energy yield available to the bacteria which mediate these








58

Table 3.1 Oxidation-reduction potentials for redeox
couples common in anaerobic wetland soils.
Standard cell potentials (Eho) are compared to
Eh7 values representative of natural systems
(corrected to pH = 7 at 25 OC). Standard cell potentials are adapted from Zehnder and Stumm
(1988); values for natural systems are from
(Ponnamperuma 1972).


Reduction half reaction Eho Eh7
Volts
02 + 4 H+ + 4 e"- t 2 H20 1.23 0.33
N03- + 6/5 H+ + 5e- f 1/2 N2 + 3H20 1.25 0.22
MnO2 + 4H+ + 2e- Mn+2 + 2H20 1.22 0.20
Fe(OH)3 + e # Fe+2 + 30H- 1.06 0.12
SO4= + 9H+ + 8e- HS- + 4H20 0.33 -0.15
CO2 + 8 H+ + 8 e- 7CH4 + 2H20 0.17 na

na = not available








59

reactions are also greatest from reduction of the strongest oxidants (Stumm 1966).

The reduction of the electron acceptors in wetlands

generally occurs in the sequence predicted by thermodynamics and energy yields (Ponnamperuma 1972; Patrick 1981). The observed Eh values listed in Table 3.1 provide a rough guide to the progress of reduction (Ponnamperuma 1972) in a natural soil. Note that potentials measured in natural systems do not agree with expected thermodynamic potentials. Most natural systems consist of a mixture of electron acceptors and donors within a heterogenous mix of microenvironments (Zehnder and Stumm 1988). Therefore, Eh measurements from natural systems represent mixed potentials and may not be interpreted quantitatively (Stumm 1966; Ponnamperuma 1972; Whitfield 1974; Vershinin and Razanov 1983). Despite this fundamental ambiguity in measuring Eh, it is common practice to characterize a wetland soil by measuring the Eh.

Eh has been a useful tool for evaluating spatial patterns and physiological and population responses of wetland plants to anaerobic conditions (Howes et al. 1981; McKee et al. 1989). Simulation of anaerobic wetland conditions in the laboratory have facilitated studies on adaptations of plants to anaerobiosis (Koch et al. 1990; Reddy et al. 1976) and wetland biogeochemistry (Masscheleyn et al. 1991; Patrick et al. 1973; Olila 1992).








60

In natural systems the availability of electron donors and acceptors, and activity of anaerobic bacteria, determine the Eh. The section that follows introduces an approach to simulating anaerobic wetland environments in the laboratory by establishing anaerobic microbial populations (verified by water chemistry via loss of electron acceptors) as an indicator of anaerobiosis instead of Eh. Accurate simulation of wetland environments requires establishment of anaerobic bacteria and emergent aquatic plants. Reactor Desigin for Simulatingr Wetland Environments Control of the oxidation-reduction potential

Batch-load reactors have been designed which allow control of Eh and pH in soil suspensions (Patrick 1966; Patrick et al. 1973). These controlled systems have contributed to a basic understanding of redox processes in wetland soils. In these systems, the Eh and pH may be set and controlled at any desired level. The Eh was monitored with platinum and calomel (Hg/HgCl) standard electrodes. Generally, organic carbon (rice straw) was added to the reactors to ensure an excess of electron donors in solution. The inherent direction of thermodynamic equilibrium was towards a methanogenic redox environment. However, in the controlled redox reactor developed by Patrick et al. (1973), when the Eh falls below a desired setpoint (e.g. +200 mV) a meter relay activates an air pump which pumps air into the soil suspension (Delaune et al. 1984). The Eh then








61

increases to the desired set point. Alkaline by-products (from oxidation of organic matter) which accumulate are neutralized by addition of acid to maintain the desired pH level. Pseudo-steady state conditions are maintained within the reactors for about 15 days. Under these conditions several redox couples in different soils have been studied (Connell and Patrick 1968; Moraghan and Patrick 1974; Cleemput et al. 1975). Current research with these reactors has addressed arsenic and selenium chemistry as affected by sediment Eh and pH (Masscheleyn et al. 1991; 1990). A modified version of the reactor (Reddy et al. 1976) allows for introduction of plants for studying effects of Eh on plant growth, root respiration, and nutrient cycling (Delaune et al. 1984). They have also been used to study phosphorus cycling in a eutrophic lake (Olila 1992).

Although this approach for simulating redox

environments has been useful in studying the basic redox processes in wetland soils (Patrick 1981), the approach used to control Eh is a poor simulation of natural wetlands. The major drawback is that Eh is controlled using 02 in the simulated anaerobic environment. This does not occur in

reduced environments. Pulsing 02 into a soil system affects decomposition rates of organic matter and rapidly oxidizes reduced compounds (e.g. NH4-N, Fe +3 Mn +2 etc.). The bacterial ecology may also be altered since aerobic bacteria may become established, Studying oxidation of organic








62
matter becomes obfuscated since facultative anaerobes prefer

02, and 02 is toxic to obligate anaerobes (Widdel 1988).

This chapter introduces a reactor designed to control Eh and pH in an anaerobic environment. The goal was to create an environment which was representative of an anaerobic wetland. The Eh was poised by the oxidants and reductants in solutions which are representative of redox processes as they occur in wetlands. The Eh of the systems was defined by the chemical species and pH of the solution according to the Nernst equation ([3-3]). Hydraulic considerations in reactor desicfn

Two ideal hydraulic conditions which are used to model reactors are plug flow and complete mix hydraulics (Metcalf and Eddy 1979). The hydraulic conditions affect the metabolic status of the bacteria and chemical equilibrium in the reactors. The bacteria and chemistry of the reactors can be in a steady state (conditions do not change in time) or dynamic state (changing with time). If a solution continuously flows into the reactor, whether it is plug flow or complete-mix, the bacteria and consequently the chemistry will come to an equilibrium which will be referred to as steady state. The reactor designed by Patrick (1966) to control Eh and pH in soil suspensions was a complete mix batch reactor. Bacterial properties and chemistry in batch reactors constantly change as the resources are depleted and by products accumulate, they are therefore, dynamic (change








63

in time). Plug flow may be simulated in columns packed with porous media. Complete-mix hydraulics are simulated in reactors which are continuously stirred. An advantage of complete mix conditions in a continuous-flow stirred tank reactor (CSTR) is that steady state conditions are established. The reactor design being presented consists of a column packed with porous media, an arrangement which will generally result in hydraulic conditions similar to plug flow. However, complete-mix conditions can be created in a packed column by recycling the reactor effluent at a recycle

rate (QR) which is much f aster than the inf luent f low rate (QJ (Characklis and Marshall 1990). This technique has been used in columns used to simulate ground water environments in packed columns (Rittman et al. 1986; Wrenn 1992). A continuous-flow of electron donors and acceptors through the completely-mixed reactor theoretically creates conditions of steady state chemistry and bacterial metabolism.

The objectives of this study were to: (i) develop a continuous-flow stirred tank reactor for growing emergent aquatic plants (Experiment 1); (ii) establish steady state biofilms of nitrate- and sulfate-reducing bacterial communities with controlled Eh and pH within the reactors (Experiment 2a and b); and (iii) establish steady state levels of oxidizable C and N which may be used as indicators

of 02 transport through the plants.








64


Materials and Methods


Experiment I. Design and Hydraulic Characteristics of
Continuous-flow Stirred Tank Reactors with Anaerobic
Biofilms

The objective of this experiment was to (i) develop design of the CSTR; and (ii) determine if the designed reactor functions as a CSTR. The hypothesis was that by varying the recycle ratio, complete-mix conditions could be created in a reactor packed with porous media. Reactor construction and operation

An expanded view of the reactor, and supporting

equipment are shown in Figure 3.1. A total of 6 reactors were built, two had 4.6 L pore volumes, the other four had

2.2 L pore volumes. Physical and operational characteristics of the two reactor types are shown in Table

3.2.

The reactor was built with 15.2 cm (6 in.) I.D.

schedule 40 PVC pipe. A 15.2 cm PVC end cap was glued to one end of the pipe. A 2.54 cm (1 in.) section of 15.2 cm I.D. pipe was glued to the end cap to create a stand for the reactor. The lid for the reactor was made by gluing a 0.64 cm (1/4 in.) thick Plexiglas disc on top of one half of a 15.2 cm I.D. PVC pipe couple. silicone grease was applied to the inside flange of the couple to create an airtight seal when the lid was placed over the pipe. Holes were cut into the Plexiglas disc (2 cm I.D.) for placement of







65








12a
5


000
2 3 0 0
11 0



6 7
8


49


5

10


6. Recycle line 7. Effluent line
8. 15.2 cm (6 in) I.D, PVC pipe 1. Nutrient solution delivery 9. 15.2 (6 in) L.D. PVC end cap 2. N2 gas line 10, 15.2 cm (6 in) I.D. PVC pipe
3. N2 gas to purge headspace 11. 1/2 of a 10.2 cm PVC couple joint
4. N2 gas, 99.9 % pure 12. 3.2 mm (1/8 in) plexiglas lid
5. Cole-Parmer peristaltic pump 12 a. Holes in plexiglas lid






Figure 3.1. Exploded view of physical components of a
continuous-flow stirred tank reactor used to
simulate nitrate- and sulfate-reducing wetland
environments.








66

Table 3.2 Physical characteristics of continuous-flow
stirred tank reactors used to simulate nitrateand sulfate-reducing wetland environments. Two
different sized reactors were designed, each had
the same basic configuration as in Figure 3.1.
In the text, reactors will be referred to by
void volume (i.e. 2.0 and 4.6 L).

Dimension Symbol Design 1 Design 2
(units)

Reactor length L, (cm) 30 12
Diameter d, (cm) 15.2 15.2
Cross-sectional A, (cm2) 182 182
area
Total volume VT, (Cm3) 5550 2240
Void volume, VV % 84 90
measured (cm3)
4600 2000
Void volume, V, (%) 90 90

specific surface A, 2.8 2.8
area of Tri-Pak' (cm2/ cm3) Specific surface cm2/cM3 4.6 4.4
area of reactors2
Flow rate Qi, 1.63 1.15
(L/day)
Retention time HRT, 2.8 1.7
(day)
Recycle flow QR 315 315
rate (L/day)
Recycle ratio QR/Qm 193 274


'Void volume and specific surface area of plastic as reported by manufacturer (Jaeger Tri-Paks, Inc., Fountain Valley, CA.
2Calculated using (V, manufacturer/V, measured)*A,








67
influent nutrient solution lines, gas lines, pH and Eh electrodes, and plants (Figure 3.1). Stoppers (size no. 3) were placed in the holes of the lid for the reactor which did not have plants.

Masterflex variable speed peristaltic pumps (Model

7567-70, Cole-Parmer, Chicago IL.) were used to continuously feed the nutrient solution to the reactors and to recycle the solution in the reactor. The recycle flow was returned to the top of the reactor 1.5 cm below the water level. Water level in the reactor was maintained about 2 cm below the lid of the reactor by adjusting the height of the effluent tube (Figure 3.1). The influent nutrient solution dripped into the reactor at the same point that the recycle flow entered the reactor to facilitate mixing. Norprene flexible tubing (Cole-Parmer size 18, I.D.= 7.9 mm) was used for the recycle lines. The headspace was continuously purged with 99.99% N2 at a flow rate of 80 100 ml/min to minimize flow of 02 into the reactor headspace and remove gaseous H2S.

Surface for attachment of biofilms. Bacterial cultures can be established either as a suspension, or with bacteria attached to a fixed surface (biofilms). This reactor was designed as a CSTR with anaerobic biofilms.

The CSTRs were filled with Tri-pak plastic packing media (Jaeger Inc., Spring, Texas). The Tri-pak plastic balls are 2.5 cm in diameter with a specific surface area of








68
280 m2/m3, and a measured pore volume of 50 % (Figure 3.2). The plastic media was chosen instead of a soil for the following reasons: (i) the low surface area relative to a soil (280 m2/M3 for plastic vs 6000 M2/M3 for a sandy soil with a mean diameter of 0.1 cm) limits the size of the bacterial population in the reactor 3, (ii) high porosity facilitates high recycle rates, and (iii) plastic is an inert media and does not supply C to the bacteria. Therefore, C balances and steady state chemistry are more easily achieved.

The Tri-pak are slightly buoyant, therefore, vexar

plastic mesh was used to submerge the tri-pak 2-3 cm below the water level in the reactor (Figure 3.3).

Reactor hydraulics. The hydraulic flow patterns were

tested by adding a conservative tracer (Rhodamine WT) to the two 4.6 L reactors. Theoretically, a conservative tracer is not decomposed or adsorbed within the reactor. This was a good assumption because sulfate-reducing bacteria isolated from wastewater metabolize only small volatile fatty acids (Widdell 1988), and Rhodamine WT has low adsorptive properties (Smart and Laidlaw 1977). Effluent concentrations of tracer were compared to the curve for an






3Nonod model for microbial growth shows that the rate of substrate utilization is directly proportional to microbial biomass (Metcalf and Eddy 1979).







69












PLASTIC JAEGER TRI-PACKSO

High performance column packing






























Figure 3.2. Tri-pak plastic media used as a surface for
attachment of nitrate- and sulfate-reducing
bacteria, and rooting matrix for Scirpus validus
grown in continuous-flow stirred tank reactors.







70












Retaining r Vexar mesh
Open area for piar ts

b). Plan view

Influent line
.. . ............ ........! ",










Water level- ........
Retaining rodt- ? e e
Vexar meslARecycle line Tri-Pakk plati

. .- . -. -. -.. .. - . . '. '. '. '














a). Cross-section




Figure 3.3. Detail of interior of continuous-flow stirred
tank reactor filled with buoyant Tri-pak plastic
balls. Retaining rod and vexar mesh were used
to hold plastic 2-3 cm below water level, a)
Cross section of reactor; b) Plan view of vexar
mesh and retaining rod.








71

ideal CSTR (equation [3], Metcalf and Eddy 1979) using methods from Middlebrooks et al. (1982): Ce = CO* e-V' [3-3]
where C, = effluent concentration
C, = influent concentration
t time of sampling
to. hydraulic retention time.
The influent flow rate was set at 3.5 mL/min while the recycle rate was adjusted on the variable speed Cole-Parmer peristaltic pump. Recycle ratios tested were 62, 70, 77, and 88. After a recycle ratio was set, 300 ML of 2.4 X 103 mg/L Rhodamine WT fluorescent dye was injected into the reactor at the same point that the influent entered the reactor. Effluent samples were collected (as it dripped out of the effluent line) every 15 minutes for the first 3 hours, then on the hour for the next nine hours. After the initial 12 hours samples were collected intermittently for up to 5 days (5 HRTs). Fluorescence of the effluent samples was measured using a Turner Design Fluorometer.

A homogeneous chemical environment is also an indicator of complete mix conditions in a CSTR. Homogeneity was evaluated by testing for stratification within the reactor. Nonconservative reactants such as electron donors (acetate)

and acceptors (S04-S) in the reactors were sampled at 5 cm depth increments in the 4.6 and 2.0 L reactors. Oxidationreduction potential was also evaluated with depth. There








72

were no plants in the reactors during the studies on hydraulic characterization.

EXPERIMENT II. Steady State Culture of Nitrate- arid Sulfatereducing Biofilms with Scirpus validus

The experimental objective was establishment of steady state nitrate- and sulfate- reducing biofilms within the CSTRs designed in Experiment 1. The hypothesis was that introduction of emergent aquatic plants into the steady state anaerobic microcosms will cause detectable changes in

concentrations of acetate, S04-S, N03-N, Eh, and/or pH. Experimental setup

One 4.6 L, and two 2.0 L reactors were used for growing the sulfate-reducing biofilms, a similar set of reactors was used for culturing bacteria. The hydraulic studies conducted on the 4.6 L reactors showed that a minimal recycle ratio of 88 was necessary for complete mix conditions. The recycle ratios for reactors used in these experiments were greater than 190 (Table 3.2). This was accomplished by using influent flow rates which were lower than those used in experiment 1 (1.1 or 0.8 vs 3.2 mL/min).

Data presented are for periods when reactors were

stable and approached steady state conditions for at least 30 days (Table 3.3). During PHASEs I, II, and III, two 2.0 L reactors were used to culture SRB. During PHASEs II and III, two 2.0 L reactors were used to culture NRB. During PHASE IV one 4.6 L reactor was used to culture SRB and one








73

to culture NRB. Reactors in PHASE I were left without plants to establish baseline operations. Twenty-seven days after initiation of PHASE II, Scirpus validus was planted in one NRB and one SRB reactor. The plants were in the reactors for all of PHASE III. After 37 days plants were removed from the reactors in PHASE IV. Relative loading

rates of acetate, N03-N and S04-S varied from 1 to 1/3 during PHASEs I through IV. The effects of different loading rates of C, N, and S were varied (actual mass loading rates are shown in Table 3.4) and analyzed for affects on steady state Eh, pH, and effluent concentrations of electron donors and acceptors.

Bacterial inoculation

The reactors were initially filled with settled

secondary wastewater (nitrate-reducing biofilm reactor [NRB]), or settled primary wastewater (sulfate-reducing biofilm reactors, [SRB]) obtained from the University of Florida Wastewater Treatment Plant. Over a period of three weeks the feed solution was changed from 100% settled wastewater to 100% nutrient solution. Nutrient solution

Nutrient solutions were prepared in opaque brown 8 L

carboys. A #9 rubber stopper was place in the mouth of the carboy. Three 2 mm holes were bored in the stopper, holes

were f or a N2 gas line, an escape hole for N2 gas, and a nutrient solution feed line (Figure 3-1). Nutrient solution










74



0
4
w r--i
M 44
CN LO N

0 -P
10 0
-P 0 -P
0
4J
4J r.
9 0
CY)
01
4 >4 N
(1) r-4 H I o U
04 a) Z
> Q)
rA ol >
4-) ri H p
C.) 4J I 4J 0
0) .,j (d 44
M Q4 ro 1 --1
Q)
-P
ol P4 cn Cl)
4J
r-I
0) r-I
u
ro w 4-)l 1
I I p
4-) a) E-1
ro 9 rl a)
0 co 0
.r-l p 4J 0 Eq.. to to p
-P pq p (1) 9 0 o rcj p
.H 041 Q) 4J 4J -to 4-1 M -P M 4) 0
r 4-) r; 4-) r-; 4-) 4J 4-3
rO Z 04 E IIJ fu fu r. to
Z X 4J r-I r-i 9 r-4 (d r-i 0 M U
0 N (0 (a W P 0 N
Q) 134 P.. H 04 r-I Q1,-j $4 W 0 0) 0)
4 0 0 04 0 04 0 P4 W 4-) $4 --A -14 Pi 0 P 0 N
(13 0 -H t 04 (a tp
4J 0 P 4J (1) Z
ro 0 4) p -r-4 N Nr
(1) ro + p u 0)
0 p (d tp 0 EA
M In >4 Z rcl N p
4 Q) H W -H Q) r-i 0
a) -P 0 4J U 4 --4-)
P4 M 4J 0 0 rd rd rd 0 1
4-4 U 4-) rd 0)
1: 4J th (1)
rO W 4 m 4J 10 1 44 0
04 W H 4-)
rd 9 z z z 44 4-) ;J U 0
r. r. 0 0 W 0 4
rd (d p (V 4-4
4-) 4J U) 4-) Q) P
r 1 0 H 0 N cq o Cl Q) -H 4 rd
0 4J r. 4J Z 0)
Q) Q) rcl Lo 4J M 4J 0) 0 0) 4J 4) fu r4 >
0 -P a) 0) 4-) --l
4J (d P 4-) m r- LO 0 p 0 0 0
fa p p fo N H r, N 4-) 4-) 4-) 9
W 4.J -r-I Q N 1-1 1-1 m 0 Q)
p -H -P 00 H cq r, LO r. >4 >4 P4 $4
E-4 r. ttl rq r-I r-i
'o rcl IZ 9 (D 0)
0 (0 0 0 4 P
0 a) a)
W
H tp r--l -ri -r-I u) w
H F-%' > (0 0 0 4J 4J
0) F-I H ;J 4-) 4J r. 9
-P p p fu (0
u (1) 0 r-A r-q
4 P4 P4 P., P4
Eri + w It 4 @








75

Table 3.4 Hydraulic retention times (HRT) and mass loading
rates (mg/L*hr) of carbon (C), sulfur as S04-S (S), and nitrogen as N03-N to the continuousflow stirred tank reactors during four phases of
operation.

PHASE I PHASE II PHASE III PHASE IV

Qi.
(mL/hr) 48 48 48 66
volume 2.0 2.0 2.0 4.6
(L)
HRT (hrs) 41.7 41.7 41.7 69.7
S04-S Mass loading,
REDUCING ---------------mg/L hr---------------Acetate 12.2 11.5 3.7 8.5
C 5.0 4.7 1.5 3.5
S 4.1 2.3 0.8 2.3
N03-N
REDUCING
Acetate na 12.2 3.7 8.5
C na 5.0 1.5 3.5
N03-N na 3.1 1.0 1.9
S na 0.2 0.07 0.11

Mass load = Cm*Q./V = Ck/HRT








76

was pumped continuously at a rate of 1.1 and 1.8 mL/min into the 4.6 liter reactors and 0.8 mL/min into the 2.0 liter reactors. Norprene flexible tubing4 (Cole-Parmer size #13, I.D. = 0.8 mm) was used to deliver nutrient solution to each reactor.

Anaerobic conditions were maintained by continuously purging the nutrient solution with 99.99% pure N2 gas (80 90 ml/min). A wooden fine bubble diffuser was used to enhance stripping of 02 from the nutrient solution. Copper tubing was used for all gas lines from the compressed gas cylinder to minimize diffusion of 02 into the N2 gas.

Maximum specific growth rate of the bacteria was not desired, therefore the nutrient solution was formulated to supply the minimum nutritional requirements for microbial growth (Widdel and Pfennig 1989). Minerals were also added to provide the minimal requirements of plants. Nutritional guidelines for wetlands plants are not established, therefore nutritional requirements of vegetable crops presented by Resh (1981) were used as a guide for preparation of the nutrient solution. The pH of the nutrient solution in the 8 L carboys was adjusted with concentrated hydrochloric acid (Table 3.5).





4oxygen permeability of the norprene tubing is 5.49 mL/mm thickness/cm2 day atm (90 ml/mil thickness/100 in2*day*atm) as reported by Cole-Parmer Inc..








77

Table 3.5 Nutrient solution pH for nitrate- and sulfatereducing reactors during four phases of
operation. Volumes of concentrated hydrochloric
acid added to each 8 liter carboy of nutrient
solution are shown.

Sulfate-reducing CSTRs+ Nitrate-reducing CSTRs

PHASE' NPL* PL PL NPL PL PL

PHASE I 2.92 2.78 na 2.96 2.93 na
(.16) (.23) (.72) (.68)

mL HCL 14 14 na 14.5 14.5 na

PHASE II 3.01 3.01 na 2.15 2.36 na
(.01) (.01) (.12) (.71)

mL HCL 8.8 8.8 na 13.0 13.0 na

PHASE III 3.16 3.37 na 2.74 2.71 na
(.11) (.30) (.08) (.02)

mL HCL 3.9 3.5 na 4.6 4.6 na

PHASE IV na na 3.08 na na 2.33
(.29) (.30)

mL HCL na na 5 na na 7


na = not applicable
+Mean and standard deviation (in parentheses) '2.2 L reactors were used during PHASES I, II, and III;
4.9 L reactor was used during PHASE IV
*NPL = no plants, PL = Scirpus validus.








78

Nutrient solution for sulfate-reducing cultures.

Sulfate was added as the only electron acceptor (Table 3.6). Yeast extract was added (1 % of carbonaceous chemical 02 demand) to supply required growth factors (Widdel and Pfennig 1984). Use of acetate as an electron donor favored selection of acetate-oxidizing SRB. Concentrations of macro- and micronutrients used in the basic nutrient solutions are shown in Table 3.6. The influent concentration of acetate, S04-S, N03-N, and Fe +2 was varied during the phases of operation (Tables 3.8, 3.10). Composition of stock solutions and volumes added to the 8 L carboys are shown in Table A-3.1.

Control of sulf ides. Microbial reduction of one mole of SO4 results in production of one mole of sulfide. Establishing plants in the sulfate-reducing reactors requires sulfide concentrations at levels nontoxic to the emergent aquatic plants. Methods to control sulf ides in solution were based on sulfide chemistry. Speciation of sulf ides in solution are dependent on pH as shown in Equation [3-4]:

H2S(.) : H'2S() # HS- + H+ # S= + H+ [3-4]

As pH decreases gaseous hydrogen sulfide is formed. In current reactor conditions (conductivity = 2200 gimhos/cmn and T = 250c) the percent of total sulfur as HS(, at pH = 6,

6.5 and 7.0, are 90, 72 and 50 % respectively (APHA 1989).

Volatilization of the H2SM is a function of the partial








79

Table 3.6 Basic composition of nutrient solution for
nitrate- and sulfate-reducing continuous-flow
stirred tank reactors. Note in Tables 3.8, and
3.10 that influent concentrations of acetate, sulfate, and nitrate change during different
phases of operation.
Salt Nutrient S04-S reducing N03-N reducing
CSTR CSTR
--------- mg/L---------NaC2H402 Na 272 272
C2H402 694 694
K2S04 K 142 0
S-SO4 58 0
NH4SO4 N-NH4 1.5 0
S-SO4 3.5 0
FeSO47H20 Fe+2 107 2.2
S-SO4 62 1.2
K2HPO4 K 71 71
P 28 28
CaCI22H20 Ca 57 57
Cl 100 100
MgC126H20 Mg 19 19
Cl 55 55
MnCI24H20 Mn 1.2 1.2
Cl 1.5 1.5
ZnCl2 Zn 1.2 1.2
Cl 0.1 0.1
KI K 0.5 0.5
I 1.6 1.6
NH4Cl N-NH4 12 0
Cl 4.2 0
KNO3 K 0 485
N-NO3 0 173
NH4NO3 N-NH4 0 9
N-NO3 0 9
Yeast extract -- 3 0
Preparation of stock solutions added to the 8L carboys to attain the above concentrations are presented in Table A-3.1








80

pressure of H2S(I) in the headspace above the reactor. Therefore the concentration of sulf ides in solution can be controlled by adjustment of pH, and continuous purging of the head space in the CSTR which maintains low partial

pressure of H2S(,, which enhances volatilization of H2SW. Ferrous iron was also added to the nutrient solution for precipitation of sulf ides (Snoeyink and Jenkins 1980) (Figure 3.6). The amount of Fe +2 added was in excess of the requirements of plants and bacteria to insure effective precipitation of sulfide.

Nutrient solution for nitrate-reducing~ bacteria.

Nitrate was added as the sole electron acceptor, with acetate as the only available electron donor. Ferrous Fe

and S04-S were added at levels required as micronutrients for plants (Resh 1981) and bacteria (Austin 1988). Other nutrient additions were the same as those added to the sulfate-reducing reactors, except that no yeast extract was added (Table 3.6) because NRB do not need the growth factors (Widdel and Pfennig 1984). Plant growth

The longstem bulrush, Scirpus validus, was propagated from seed, planted in gravel and grown in a shaded greenhouse. Plants were fed a mixture of nutrient solutions used for growing the nitrate- and SRB. Five segments of roots and rhizome (about 3 cm each) with 2 to 5 tillers/rhizome were removed from the gravel. The roots were