Engineered Biofiltration for Enhanced Hydraulic and Water Treatment Performance

MISSING IMAGE

Material Information

Title:
Engineered Biofiltration for Enhanced Hydraulic and Water Treatment Performance
Physical Description:
1 online resource (231 p.)
Language:
english
Creator:
Lauderdale,Chance V
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Doctorate ( Ph.D.)
Degree Grantor:
University of Florida
Degree Disciplines:
Environmental Engineering Sciences
Committee Chair:
Chadik, Paul A
Committee Co-Chair:
Lindner, Angela S
Committee Members:
Mazyck, David W
Svoronos, Spyros
Kirisits, Mary Jo

Subjects

Subjects / Keywords:
biofiltration -- biological -- drinking -- manganese -- methylisoborneol -- nutrient -- organic -- water
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre:
Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
Currently, biofiltration is largely operated as a passive process in the water treatment industry. Particle/turbidity removal and headloss drive the design and operation of conventional filtration as well as biofiltration. Thus, biofilter design parameters are typically limited to media configuration, backwash strategy, and loading rate. The biological removal of dissolved organic and inorganic contaminants is an anticipated benefit of biofiltration. However, common design and operational practice does not seek to enhance the biological activity responsible for those mechanisms. Indeed, in an effort to improve filter productivity and minimize headloss, many utilities employ chlorinated backwashes and other biomass control strategies. However, these are often to the detriment of biological activity and may be ineffective at removing a primary foulant of biofilters ? extracellular polymeric substances (EPS). This study focused on identifying enhancement strategies to improve both water quality and hydraulic performance of drinking water biofilters by increasing microbial activity while decreasing biological fouling. These strategies included biofilter substrate-, nutrient-, and oxidant-enhancement. Of the strategies tested, nutrient-enhancement and oxidant-enhancement showed the most promise for drinking water biofilter applications. The nutrient enhancement strategy is elegant in its simplicity: operate a given biofiltration process so that an approximate bioavailable carbon:Ammonia-nitrogen:Orthophosphate-phosphorus (C:N:P) molar ratio of 100:10:1 is maintained. Achieving a nutrient balance decreased biofilter terminal headloss by ~15 percent relative to the control which is possibly the result of reduced EPS formation in the filter. Nutrient enhancement also sustainably decreased contaminant breakthrough relative to the control biofilter, including 2-methylisoborneol (MIB), manganese (Mn), and dissolved organic carbon (DOC). A preliminary evaluation of the oxidant-enhancement strategy was implemented by providing a low dose (1 mg/L) of hydrogen peroxide to the biofilter feed over a 2-week test period. The objective of this strategy was to enhance the oxidative action of biofilter microorganisms and to promote the oxidation of inactive biomass. The test filter demonstrated ~15 percent removal of filter feed DOC (7 percent less breakthrough than control), and removal of Mn and MIB to non-detect levels. The oxidant enhancement strategy also decreased terminal headloss to a mean 2.2 feet, or ~60 percent of the control. Both strategies showed enhanced biofilter water treatment performance without compromising unit filter run volumes or particulate removal performance.
General Note:
In the series University of Florida Digital Collections.
General Note:
Includes vita.
Bibliography:
Includes bibliographical references.
Source of Description:
Description based on online resource; title from PDF title page.
Source of Description:
This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility:
by Chance V Lauderdale.
Thesis:
Thesis (Ph.D.)--University of Florida, 2011.
Local:
Adviser: Chadik, Paul A.
Local:
Co-adviser: Lindner, Angela S.

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2011
System ID:
UFE0042819:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 ENGINEERED BIOFILTRATION FOR ENHANCED HYDRAULIC AND WATER TREATMENT PERFORMANCE By CHANCE V. LAUDERDALE 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 201 1

PAGE 2

2 201 1 Chance V. Lauderdale

PAGE 3

3 To Courtney and Marley, I love you both. T his is yours more than it is mine. I know I have a lot of making up to do!

PAGE 4

4 ACKNOWLEDGMENTS This work was made possible through the financial support of Water Research Foundation, the City of Arlington Water Utilities Ozonia, and the University of Florida (thanks Angela!) I would also like to express gratitude to Paul Chadik and my research committee for their guidance and assistance through my unconventional path. My parents also deserve recognition for their continued support through all of my education and f or constantly reminding me how neat this opportunity has been. A final thank you goes to Jess Brown, for being the best mentor someone could have, and Bob Cushing for being a champion for this endeavor, despite the fact that it led me to Texas.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS .................................................................................................. 4 LIST OF TABLES .......................................................................................................... 11 LIST OF FIGURES ........................................................................................................ 14 LIST OF ABBREVIATIONS ........................................................................................... 17 ABSTRACT ................................................................................................................... 2 2 CHAPTER 1 INTRODUCTION AND OBJECTIVES ..................................................................... 24 Problem Statement ................................................................................................. 24 Hydraulic Performance ..................................................................................... 26 Water Treatment Performance Concerns ......................................................... 30 Hypothesis and Objectives ..................................................................................... 33 Approach ................................................................................................................ 34 2 LITERATURE REVIEW .......................................................................................... 35 Historical Background ............................................................................................. 35 Configurat ions of Biological Treamtent ................................................................... 37 Contaminant Treatability ......................................................................................... 38 Overview .......................................................................................................... 38 Natural Organic Matter ..................................................................................... 38 MIB and Geosmin ............................................................................................. 39 Manganese Removal ....................................................................................... 41 Pharmaceuticals and Pesticides ....................................................................... 43 Convention Biofiltration Design Considerations ...................................................... 45 Flow Equalization ............................................................................................. 45 Pre treatment ................................................................................................... 45 Media Selection ................................................................................................ 46 Backwash System ............................................................................................ 47 Conventional Bio filtration Operation Considerations ............................................... 48 Loading Rate and Contact Time ....................................................................... 48 Acclimation Procedures .................................................................................... 48 Backwash Protocol ........................................................................................... 49 Engineered Bio filtration Operational Considerations .............................................. 51 Primary Substrate Augmentation ...................................................................... 51 Nutrient Augmentation ...................................................................................... 53 The role of EPS in microbial systems ............................................................... 53 EPSrelated challenges in water treatment ...................................................... 54

PAGE 6

6 Oxidant Augmentation with Peroxide ............................................................... 56 Monitoring Techniques ............................................................................................ 59 Real Time Monitoring Tools ............................................................................. 59 Water quality analyses ............................................................................... 59 Headloss tracking ...................................................................................... 60 ATP analysis .............................................................................................. 60 Non Real Time Monitoring Tools ...................................................................... 61 Wa ter quality analyses ............................................................................... 61 Annular reactor study ................................................................................. 61 Filter coring ................................................................................................ 62 Biological activity analyses ......................................................................... 62 Microbial community analysis .................................................................... 63 Microscopy ................................................................................................. 63 Post treatment analysis ............................................................................. 64 3 MATERIALS AND METHODS ................................................................................ 65 Backgrou nd ............................................................................................................. 65 Pilot Biofiltration System ......................................................................................... 65 Pilot Biofiltration System ................................................................................... 65 General Process and Control ........................................................................... 66 Supplemental Chemical Dosing ....................................................................... 66 Contaminant Spik ing ........................................................................................ 67 Biofilter Augmentation and Contaminant Feed Chemicals ............................... 67 Backwash Protocol ........................................................................................... 68 Experimental D esign ............................................................................................... 69 Research Testing Plan ..................................................................................... 69 Operational Setup ............................................................................................ 70 Sampling Plan .................................................................................................. 70 Baseline Biofiltration Characterization .............................................................. 70 Biofiltration SubstrateEnh ancement Evaluation ............................................. 71 Biofiltration Nutrient Enhancement Evaluation ................................................ 74 Oxidant Enhancement Evaluation ................................................................... 74 Microbial Tracking ............................................................................................ 74 FullScale Process Integration Assessment and Economic Evaluation ............ 75 Data Analyses ......................................................................................................... 75 General ............................................................................................................. 75 Paired T Test .................................................................................................... 75 Analytical Methods .................................................................................................. 76 General Water Quality Parameters .................................................................. 76 Sample collection for laboratory analyses .................................................. 76 Turbidity ..................................................................................................... 76 pH .............................................................................................................. 77 Dissolved oxygen and temperature ............................................................ 77 Free and total chlorine ............................................................................... 77 Hydrogen peroxide ..................................................................................... 77 Ozone residual ........................................................................................... 77

PAGE 7

7 Total and dissolved organic carbon ........................................................... 77 UV254 .......................................................................................................... 78 Iron ............................................................................................................. 78 Manganese ................................................................................................ 78 Nitrates, nitrites .......................................................................................... 78 Ammonia .................................................................................................... 78 Orthophosphate ......................................................................................... 79 Regulated microbial parameters ................................................................ 79 Disinfection by products ............................................................................. 79 Tastes and odors ....................................................................................... 79 Pesticides and Pharmaceuticals ....................................................................... 79 Sample collection ....................................................................................... 80 Chemicals and reagents ............................................................................ 84 Sample preparation (solid phase extraction) .............................................. 84 LC/TOF MS analyses of pesticides and pharmaceuticals .......................... 85 LC/MS MS analyses of pharmaceuticals at low level (ppt concentration) .. 86 Biofilter Media Microbial Characterization and Analyses .................................. 87 Sample collection for ATP analyses ........................................................... 87 ATP analyses ............................................................................................. 87 Sample collection for other microbial tracking analyses ............................. 87 Scanning electron microscopy ................................................................... 88 Biofilter media heterotrophic plate count .................................................... 88 Crystal violet (CV) assay (biofilm formation potential) ................................ 89 Phenol sulfuric acid assay (EPS quantification) ......................................... 89 Terminal restriction fragment length polymorphism ................................... 90 Clone libraries ............................................................................................ 91 Summary of Responsible Parties for Analytical Work Performed ..................... 92 4 BASELINE CHARACTERIZATION AND CONTROL STUDIES ............................. 93 Objectives ............................................................................................................... 93 Biofilter Configuration .............................................................................................. 93 Biofilte r Backwash Strategy Development .............................................................. 94 Wood Based Gac Media Evaluation ....................................................................... 94 Hydraulic Characterization ...................................................................................... 95 Water Quality Characterization ............................................................................... 97 Ge neral ............................................................................................................. 97 Turbidity ............................................................................................................ 98 Dissolved Organic Carbon ................................................................................ 99 Nutrients ......................................................................................................... 102 Metals ............................................................................................................. 103 Tastes and Odors ........................................................................................... 106 Pharmaceuticals and Pesticides ..................................................................... 107 Summary .............................................................................................................. 108

PAGE 8

8 5 SUBSTRATE ENHANCEMENT STUDIES ........................................................... 109 Objectives ............................................................................................................. 109 Chemical Feed Observations ................................................................................ 109 Hydraulic Characterization .................................................................................... 110 Water Quality Characterization ............................................................................. 111 General ........................................................................................................... 111 Turbidity .......................................................................................................... 112 Dissolved Organic Carbon .............................................................................. 113 Nutrie nts ......................................................................................................... 116 Metals ............................................................................................................. 116 Taste and Odor .............................................................................................. 117 Pharmaceuticals and Pesticides ..................................................................... 119 Biological Activity .................................................................................................. 120 Summary .............................................................................................................. 120 6 NUTRIENT ENHANCEMENT STUDIES ............................................................... 122 Objectives ............................................................................................................. 122 Hydraulic Characterization ............................................................................. 123 Nutrient enhancement testing .................................................................. 123 Substrate and nutrient enhancement testing ............................................ 124 Nutrient enhancement validation .............................................................. 127 Water Quality Characterization ............................................................................. 128 General ........................................................................................................... 128 Turbidity .......................................................................................................... 128 DOC ............................................................................................................... 129 Nutrients ......................................................................................................... 134 Nutrient balance evaluation ...................................................................... 134 Nutrient breakthrough evaluation ............................................................. 136 Metals ............................................................................................................. 140 Taste and Odor .............................................................................................. 141 Chloramine Decay and DBPFP Testing ......................................................... 145 Pharmaceuticals and Pesticides ..................................................................... 146 Biological Activity .................................................................................................. 147 Robustness T esting .............................................................................................. 148 Summary .............................................................................................................. 148 7 OXIDANT ENHANCEMENT STUDIES ................................................................. 151 Objectives ............................................................................................................. 151 Testing Conditions ................................................................................................ 151 Hydraulic Characterization .................................................................................... 152 Water Quality Characterization ............................................................................. 154 General ........................................................................................................... 154 Hydrogen Peroxide ......................................................................................... 154 Turbidity .......................................................................................................... 154

PAGE 9

9 DOC ............................................................................................................... 155 Nutrients ......................................................................................................... 156 Metals ............................................................................................................. 157 Taste and Odor .............................................................................................. 158 Biological Activity .................................................................................................. 159 Summary .............................................................................................................. 160 8 MICROBIAL TRACKING ....................................................................................... 161 Objectives ............................................................................................................. 161 Scanning Electron Microscopy .............................................................................. 161 General Observations .................................................................................... 161 Fullscale Biofilter ........................................................................................... 162 Biofilter Control ............................................................................................... 163 Substrate Enhancement Studies .................................................................... 166 Nutrient Enhancement Studies ....................................................................... 168 Ox idation Enhancement Studies .................................................................... 170 Summary of SEM Micrographs ....................................................................... 170 Plate Count ..................................................................................................... 171 Biofilm Formation Characterization ................................................................ 172 EPS Quantification ......................................................................................... 174 ATP Characterization ............................................................................................ 176 Baseline Control ............................................................................................. 176 Substrate Enhancement Studies .................................................................... 176 Nu trient Enhancement Studies ....................................................................... 177 Oxidant Enhancement Studies ....................................................................... 178 T RFLP ................................................................................................................. 179 Clone Libraries ...................................................................................................... 181 9 FULL SCALE PRO CESS INTEGRATION ASSESSMENT AND ECONOMIC EVALUATION ....................................................................................................... 189 Objectives ............................................................................................................. 189 Process Integration ............................................................................................... 189 Conceptual Design and Implementation ......................................................... 189 Process Monitoring ......................................................................................... 191 Additional Considerations ............................................................................... 191 Cost Assessment .................................................................................................. 192 Assumptions ................................................................................................... 192 Capital Cost .................................................................................................... 192 Operation and Chemical Cost ........................................................................ 193 Total Estimated Cost for Implementation ........................................................ 193 Potential Net Costs and Cost Savings .................................................................. 193 10 SUMMARY AND CONCLUSIONS ........................................................................ 196 Problem Statement and Hypothesis ..................................................................... 197

PAGE 10

10 Objectives ............................................................................................................. 198 Results .................................................................................................................. 199 Baseline Biofiltration C haracterization ............................................................ 199 Substrate Enhancement Studies .................................................................... 201 Nutrient Enhancement Studies ....................................................................... 203 Oxidant Enhancement Studies ....................................................................... 205 Microbial Tracking .......................................................................................... 207 Scanning electron microscopy ................................................................. 207 Biofilter media HPC .................................................................................. 208 Biofilm formation potential ........................................................................ 2 08 EPS quantification .................................................................................... 208 Adenosine triphosphate measurements ................................................... 209 16S rRNA and clone libraries ................................................................... 209 Fullscale Process Integration Assessment and Economic Evaluation .......... 210 Summary .............................................................................................................. 211 11 FUTURE WORK ................................................................................................... 213 REFERENCES ............................................................................................................ 217 BIOGRAPHICAL SKETCH .......................................................................................... 231

PAGE 11

11 LIST OF TABLES Table page 1 1 Summary of calculated hydraulic parameters for biofilters ................................. 28 1 2 Summary of pharmaceuticals and pesticides measured in the City of Arlingtons raw and finished waters .................................................................... 32 2 1 Summary of minimum % removal of selected pharmaceuticals and pesticides by ozonation at drinking water dosages ............................................................. 44 2 2 Summary of % removal of selected pharmaceuticals and pesticides by pilot biofilters (with ozone pretreatment) .................................................................... 44 3 1 Backwash protocols ............................................................................................ 69 3 2 Pilot biofilter operational parameters .................................................................. 72 3 3 Approximate sampling schedule for routine analyses ......................................... 73 3 4 Limit of detection for screened pharmaceuticals and pesticides assuming 100% recovery by solid phase extraction. Individual water sample matrices may vary ............................................................................................................. 80 3 5 Responsible Parties fo r Analytical Work Performed ........................................... 92 4 1 Baseline characterization of pilot biofilter headloss ............................................ 96 4 2 City of Arlington biofilter water treatment objectives ........................................... 97 4 3 City of Arlington fullscale biofilter performance.................................................. 98 4 4 Baseline characterization of biofilter turbidity removal ........................................ 99 4 5 Baseline characterization of biofilter DOC removal .......................................... 101 4 6 Baseline characterization of nutrient feed and biofilter utilization ..................... 104 4 7 Baseline characterization of biofilter Mn and Fe removal ................................. 104 4 8 Baseline characterization of biofilter taste and odor removal ........................... 105 4 9 Baseline performance comparison of the pilot and full scale filters .................. 107 4 10 Pilot biofilter treatment performance for spiked atrazine, carbamazepine, and caffeine ............................................................................................................. 108

PAGE 12

12 5 1 Characterization of substrateenhanced pilot biofilter headloss ........................ 111 5 2 Substrate enhanced biofilter DOC removal characterization ............................ 114 5 3 Substrate enhanced biofilter normalized DOC removal characterization .......... 114 5 4 Substrate enhanced bio filter Fe removal characterization ................................ 117 5 5 Substrate enhanced biofilter Mn removal characterization ............................... 117 5 6 Substrate enhanced biofilter MIB removal characterization .............................. 118 5 7 Substrate enhanced biofilter evaluation of media ATP concentrations ............. 120 6 1 Nutrient enhanced biofilter hydraulic performance ........................................... 125 6 2 Substrate and nutrient enhanc ed biofilter hydraulic performance ................... 126 6 3 Nutrient enhancement validation on biofilter hydraulic performance ................ 127 6 4 Baseline characterization of nutrient enhanced biofilter turbidity breakthrough 129 6 5 Substrate and nutrient enhanced biofilter DOC removal performance ............ 130 6 6 Substrate and nutrient enhanced biofilter normalized DOC removal characterization ................................................................................................ 133 6 7 Nutrient enhancement studies: nutrient balances ............................................. 135 6 8 Nutrient enhancement studies: PO4P characterization .................................... 137 6 9 Nutrient enhancement studies: NH4N characteri zation ................................... 138 6 10 Biofilter nitrification characterization after ammonium chloride supplementation ............................................................................................... 139 7 1 Oxidant enhancement validation on biofilter hydraulic performance ................ 153 7 2 Oxidant enhanced biofilter DOC removal performance .................................... 156 8 1 Summary of observations made during SEM analysis of biofilter media .......... 171 8 2 Diversity and similarity indices for various biofilter media samples ................... 181 8 3 Biofilter control clone library ............................................................................. 183 8 4 Substrate and nutrient enhanced biofilter clone library (carbon limited) ........... 184 8 5 Oxidant enhanced biofilter clone library ........................................................... 185

PAGE 13

13 8 6 Ammonialimited, nutrient enhanced biofilter clone library ............................... 186 9 1 Cost assessment assumptions ......................................................................... 192 9 2 Backwash water production estimates ............................................................. 194 9 3 Chemical costs ($/MG) to retreat backwash wastewater .................................. 194 9 4 Pumping costs to recycle backwash wastewater .............................................. 194 9 5 Nutrient enhancement implementation net costs or savings ............................ 195

PAGE 14

14 LIST OF FIGURES Figure page 1 1 Conceptual process schematic for the JKWTP and PBSWTP ............................ 25 1 2 Historical UFRVs for the JKWTP ........................................................................ 27 1 3 Failed IMS cap removed from a JKWTP biofilter with stripped anchoring screw .................................................................................................................. 29 1 4 Photograph of failed IMS cap removed from a JKWTP biofilter with blown mastic seal .......................................................................................................... 29 3 1 Process flow schematic for pilot biofiltration system ........................................... 67 4 1 Control pilot biofilter headloss profi les impacted by sludge pond recycle to JKWTP influent ................................................................................................... 96 4 2 Pilot control biofilter effluent turbidity profiles across two filter runs .................. 100 4 3 Pilot control biofilter and full scale biofilter steady state DOC removal performance ..................................................................................................... 102 4 4 Pilot control biofilter steady state and peak load Mn removal performance ...... 105 5 1 Comparison of substrate enhanced (MicroC) and biofilter control headloss profiles .............................................................................................................. 112 5 2 Comparison of substrate enhanced and biofilter control turbidity profiles for a typical week of filter runs .................................................................................. 113 5 3 Comparison of substrate enhanced and biofilter control normalized DOC concentrations .................................................................................................. 115 5 4 Comparison of substrate enhanced and biofilter control normalized MIB concentrations .................................................................................................. 119 6 1 Comparison of nutrient enhanced (PO4P) and biofilter control headloss profi les .............................................................................................................. 125 6 2 Effect of ammonium chloride supplementation on substrate and nutrient enhanced biofilter operated with NH4N limitation ............................................ 126 6 3 Nutrient enhanced biofilter turbidity profiles for typical filter runs ...................... 129 6 4 Comparison of nutrient enhanced and biofilter control normalized DOC removals .......................................................................................................... 130

PAGE 15

15 6 5 DOC removal performance improvement with nutrient enhancement of (previous) biofilter control ................................................................................. 132 6 6 Characterization of normalized DOC removal for substrateand nutrient enhanced biofilters ........................................................................................... 134 6 7 Biofilter nitrification ch aracterization after ammonium chloride supplementation to the substrate (ethanol) and nutrient enhanced biofilter .... 140 6 8 Mn removal performance for the nutrient enhancement strategies during simulated moderate long term loading event .................................................... 142 6 9 Mn removal per formance for the nutrient enhancement strategies during simulated high short term loading event ........................................................... 143 6 10 MIB removal performance for the nutrient enhancement strategies during simulated moderate long term loading event .................................................... 144 6 11 MIB removal performance for the nut rient enhancement strategies during simulated high short term loading event ........................................................... 144 6 12 Chloramine decay results for the nutrient enhanced biofilter and biofilter control ............................................................................................................... 145 6 13 DBPFP results for the nutrient enhanced biofilter and biofilter control ............. 146 6 14 Normalized pharmaceutical and pesticide removal performance during the nutrient enhancement studies .......................................................................... 147 7 1 Effect of oxidant enhancement on biofilter headloss profiles ............................ 153 7 2 Oxidant enhanced biofilter turbidity profiles for typical filter runs ...................... 155 7 3 The effect of oxidant enhancement on DOC removal performance .................. 157 7 4 Oxidant enhanced biofilter Mn removal performance ....................................... 158 7 5 Oxidant enhanced biofilter MIB removal performance ...................................... 159 8 1 Fullscale biofilter GAC media SEM micrograph............................................... 163 8 2 Biofilter control media SEM micrograph: 2 weeks of pilot operation ................. 164 8 3 Biofilter control media SEM micrograph: 38 weeks of pilot operation ............... 165 8 4 Biofilter control media SEM micrograph: white granules .................................. 166 8 5 Substrate enhanced, nutrient limited biofilter: MicroC substrate ..................... 167

PAGE 16

16 8 6 Substrate enhanced, nutrient limited biofilter media SEM micrograph: ethanol substrate ........................................................................................................... 167 8 7 Nutrient enhanced biofilter media SEM micrograph: phosphoric acid .............. 168 8 8 Substrate and nutrient enhanced biofilter media SEM micrograph: ethanol substrate, phosphoric acid, ammonium chloride ............................................... 169 8 9 Substrate enhanced biofilter media SEM micrograph: ethanol substrate, phosphoric acid ................................................................................................ 169 8 10 Oxidant enhanced biofilter media SEM micrograph: hydrogen peroxide .......... 170 8 11 Biofilter media HPC per mL of phosphate buffered saline media samples: biofilter control and nutrient enhanced bi ofilter ................................................. 172 8 12 Relative biofilm formation potential between biofilter control and nutrient enhanced biofilter ............................................................................................. 173 8 13 Relative biofilm formation potential between biofilter control, substrateenhanced biofilter, and substrateand nutrient e nhanced biofilter ................... 174 8 14 Nutrient enhancement influences on biofilter media EPS relative to the control biofilter .................................................................................................. 175 8 15 Substrate enhancement greatly increased EPS concentrations under nutrient limited conditions .............................................................................................. 176 8 16 Effects of nutrient supplementation on substrateenhanced biofilter media ATP concentrations .......................................................................................... 178 8 17 Nutrient enhancement and nutrient and substrateenhancement ATP characterization ................................................................................................ 178 8 18 Nutrient enhancement and nutrient and substrateenhancement ATP characterization ................................................................................................ 179 8 19 Comparison of biofilter media clone libraries under phosphorus limited and carbonlimited conditions .................................................................................. 188 8 20 Relative abundance (via T RFLP) of Bradyrhizobium before and after phosphoric acid supplementation (0.02 mg/L as P) .......................................... 188 9 1 Conceptual integration schematic for nutrient enhancement ............................ 190

PAGE 17

17 LIST OF ABBREVIATION S AAS atomic adsorption spectrophotometer AOC assimilable organic carbon AP Associated Press ATP adenosine triphos phate BDOC biologically degradable organic carbon BOM biodegradable organic matter C carbon CFU colony forming units City City of Arlington, Texas C:N:P bioavailable carbon: ammonium N :orthophosphateP CT contact time CV crystal violet C degree Celsius DBP disinfection by product DBPFP disinfection by product formation potential DEET N,NDiethyl meta tolumide DI deionized water DNA deoxyribonucleic acid DO dissolved oxygen DOC dissolved organic carbon DPD N,NDiethyl p Phenylenediamine DSA deposit and surface analysis DWU Dallas Water Utilities EBCT empty bed contact time

PAGE 18

18 EDCs endocrine disruptor ing compounds EDS energy dispersive spectroscopy EPS extra cellular polymeric substances ESI electrospray ionization FDA Food and Drug Administration Fe iron FISH fluorescence in situ hybridization FTIR Fourier transform infrared ft feet or foot GAC granular activated carbon GC gas chromatography GC/MS gas chromatography/mass spectrometry GHz gigahertz gpd gallons per day gpm/ft2 gallons per minute per square foot HLR hydraulic loading rate HMI human machine interface HO, hydroxyl radicals HPCs heterotrophic plate counts HPLC high performance liquid chromograph hr hour(s) ID inner diameter ISE ion selective electrode JKWTP John F. Kubala Water Treatment Plant LB Luri a Bertani

PAGE 19

19 L/d (filter) lengthto diameter (particle of media) L/min liters per minute LC/MS liquid chromatography/mass spectrometry LC/MS/MS liquid chromatography/quadr upole mass spectrometry LOD limit of detection MCL maximum contaminant level g/L micrograms per liter g/mL micrograms per milliliter mg milligrams MG million gallons mg/L milligrams per liter mgd million gallons per day MIB 2 methylisoborneol min minute(s) mL milliliter mL/min milliliter per minute mm millimeter mM millimolar Mn manganese MRLs minimum reporting levels mRNA messenger ribonucleic acid MWH Montgomery Watson Harza ng/L nanograms per liter NH4N ammonianitrogen nm nanometer

PAGE 20

20 NOM natural organic matter NSF National Sanitation Foundation NTU Nephelometric Turbidity Units O3 ozone ORD Office of Research and Development PO4P orthophosphatephosphorus PAC powdered activated carbon PBS phosphatebuffered saline PBSWTP Pierce Burch South Water Treatment Plant PCR polymerase chain reaction PLC p rogrammable l ogic controler PVC polyvinyl chloride qPCR quant it ative polymerase chain reaction RLU relative light units RNA ribonucleic acid RPM revolutions per minute rRNA ribosomal ribonucleic acid RSD relative standard deviation RT PCR reverse transcription polymerase chain reaction SCADA supervisory control and data acquisition SDSDBP simulated distribution system disinfection by product SEM scanning electron microscopy SI S rensen index SM standard method SWI ShannonWeaver Index

PAGE 21

21 T&O taste and odor TCCP tris (2 chloroisopropyl) phosphate TCEP tris (2 charboxyethyl) phosphine TCPP tris (2 chlorophropyl) phosphate TEM transmission electron microscopy TOC total organic carbon T RFLP terminal restriction fragment length polymorphism TRWD Tarrant Regional Water District U.S. United States UFRV unit filter run volume UHPLC ultra highpressure liquid chromotograph USEPA United States Environmental Protection Agency V volt VFDs variable frequency drives WTP w ater treatment plant

PAGE 22

22 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 ENGINEERED BIOFILTRATION FOR ENHANCED HYDRAULIC AND WATER TREATMENT PERFORMANCE By Chance V. Lauderdale August 2011 Chair: Paul Chadik Cochair: Angela Lindner Major: Environmental Engineering Sciences Currently, biofiltration is largely operated as a passive process in the wat er treatment industry. Particle removal and headloss drive the design and operation of conventional filtration and biofiltration. Thus, biofilter design parameters are typically limited to media configuration, backwash strategy, and loading rate. The removal of dissolv ed organic and inorganic contaminants is an anticipated benefit of biofiltration. However, common design and operational practice does not seek to enhance the biological activity responsible for those mechanisms. Indeed, in an effort to improve filter productivity and minimize headloss, many utilities employ chlorinated backwashes and other biomass control strategies. However, these are often to the detriment of biological activity and may be ineffective at removing a primary foulant of biofilters extracellular polymeric substances (EPS). This study focused on identi fying enhancement strategies to improve both water quality and hydraulic performance of drinking water biofilters by increasing microbial activity while dec reasing biological fouling These strategies included biofilter substrate, nutrient and oxidant enhancement. Of the strategies tested, nutrient enhancement and

PAGE 23

23 oxidant enhancement showed the most promise for drinking water biofilter applications. The nutrient enhancement strategy is elegant in its simplicity: operate a given biofiltration process so that an approximate bioavailable carbon: a mmonianitrogen: orthophosphatephosphorus molar ratio of 100:10:1 is maintained. Achieving a nutrient balance decreased terminal headloss by ~15 percent relative to the control possibly the result of reduced EPS formation. Nutrient enhancement also sustainably decreased contaminant breakthrough relative to the control biofilter, including 2 methylisoborneol (MIB) manganese (Mn), and dissolved organic carbon (DOC) A preliminary evaluation of the oxidant enhancement strategy was implemented by providing a 1 mg/L dose of hydrogen peroxide to the biofilter feed over a 2week test period. The objective was to enhance the oxidative action and respons e of biofilter microor ganisms and to promote the oxi dation of inactive biomass. The filter demonstrated ~15 percent removal of filter feed DOC (7 percent less breakthrough than control), and removal of Mn and MIB to nondetect levels. The oxidant enhancement strategy also decreased terminal headloss to a mean 2.2 feet, or ~60% of the control. Both strategies showed enhanced water treatment performance without compromising filter productivity or particulate removal performance.

PAGE 24

24 CHAPTER 1 INTRODUCTION AND OBJ ECTIVES Problem Statement The City of Arlington, Texas (City) owns and operates two ozone/biofiltration drinking water treatment facilities, the Pierce Burch South Water Treatment Plant ( PBSWTP) and the John F. Kubala Water ( JKWTP ) Bo th facilities receive water from the Tarrant Regional Water District system (TRWD). The TRWD system pumps water from the Richland Chambers, Benbrook, and Cedar Creek reservoirs. Water quality is similar among the reservoirs, and changes in blending ratios have some impact on facility performance. The JKWTP receives water directly from the TRWD system. However, Lake Arlington is used as terminal storage for TRWD water before it is pumped to the PBSWTP. Lake Arlington contributes additional seasonal loads of taste and odor (T&O) and manganese ( Mn ) to the PBSWTP. The City implemented ozone/biofiltration processes at these facilities in 2001 to remove t astes and o dors ( T&O ) iron (Fe), manganese( Mn ) turbidity, and to minimize disinfection by product formation potential (DBPFP) distribution system chloramine demand, and regrowth potential. Currently, the PBSWTP has a production capacity of 72 million gallons per day (mgd) with a process train that includes coagulation, flocculation, sedimentation, biofiltratio n (40 inches of granular activated carbon [GAC], 6 inches of sand, and Leopold IMS cap) 1 1 The PBSWTP filter design was altered in 2009, during this study, to replace the Leopold IMS caps with gravel over the underdrains. and final disinfection. Disinfection residual consists of a chlorination step to achieve free chlorine contact time followed by downstream ammonia addition to produce chloramines for distribution. Primary

PAGE 25

25 disinfection is obtained through the intermediate ozonation process. The City practices a nonchlorinated/chloraminated backwash (BW) at both the JKWTP and PBSWTP. Backwash wastewater (BWW) is conveyed to onsite lag oons. The JKWTP has a capacity of 97.5 mgd and includes the same treatment process scheme as the PBSWTP plant, with the exception that JKWTP biofilters contain 48 inches of GAC, 8 inches of sand, and 12 inches of gravel over the underdrain. Figure 11 provi des a conceptual process schematic for the JKWTP and PBSWTP. Figure 11 Conceptual process schematic for the JKWTP and PBSWTP While the PBSWTP and JKWTP have performed well for many years, recently observed biofilter hydraulic and water treatment performance disruptions led the City to evaluate their biofiltration systems to identify potential enhancement strategies. Specific performance disruptions included underdrain clogging, increased chloramines residual decay rates (i.e., organic carbon break through from the biofilters, specific to PBSWTP), decreased Mn removal efficiency, and decreased T&O removal efficiency resulting in odor complaints. In addition, a 2007 survey of JKWTP and PBSWTP raw and finished Preozonation Contactor Sedimentation Filtration Chlorine Disinfection Intermediate Ozone Contactor Flocculation BW NaOH, Ammonia Alum Polymer BWW Ozone Ozone To Pilot Biofilters

PAGE 26

26 waters identified low levels (ng/L) of pharmaceuticals and endocrine disrupting compounds. Through one sampling effort, the City analyzed both raw and treated water for both the JKWTP and PBSWTP. The City determined that their treatment processes removed all but one identified pharmaceutical present in the raw water to below parts per trillion. Therefore, the City desired to determine if enhancing biofiltration would further remove this compound and other potential compounds through participation in this study. Detailed discussions on the PBSWTP an d JKWTP hydraulic and water treatment performance upsets are provided in the subsequent sections Hydraulic Performance The PBSWTP and JKWTP biofilters were designed as rising level/constant rate, inter filter backwashing filters, with a Leopold universal air/water underdrain and IMS cap. The inter filter backwash uses the effluent and discharge head from filters in production mode for the backwash process. Although both facilities produce exceptional filtered water quality (<0.08 Nephelometric Turbidity U nits [NTU] effluent), filter productivity has declined through the years. As shown in Figure 12, the average unit filter run volume (UFRV) for JKWTP decreased by approximately 50% from January 2001 to May 2005. The largest observed decrease in productivit y occurred between 2001 and 2002, coinciding with the ozone/biofiltration system going online.

PAGE 27

27 Figure 12 Historical UFRVs for the JKWTP It was determined that filter media mud ball formation and underdrain cap fouling were contributing factors for decreased filter productivity. These factors also limit the efficacy of the backwash system, creating a self exacerbating condition. Cleanbed headloss has increased through the years, resulting in higher water elevations within the filters at the start of filter runs. Table 11 summarizes an analysis of six selected JKWTP biofilter runs for three biofilters operated from May 2006 through September 2006. There is a total of eight feet of head available for filtration. The average biofilter clean bed headlos s ranged from 5.3 to 7.9 feet, leaving 2. 7 to 0.1 feet of head available for filtration. Based on media characteristics, bed depth, and loading rates, the calculated expected clean bed headloss for the media configuration at the JKWTP is 0.9 feet (calculat ion provided in Table 11 notes) The discrepancy 0 2,000 4,000 6,000 8,000 10,000 12,000 14,000 2001 2002 2003 2004 2005 Average UFRV ( gal/ ft 2 /day) Year of Operation Ave UFRV Min UFRV

PAGE 28

28 between actual and expected clean bed headloss reflects considerable fouling of the media and porous plate in the underdrain. Indeed, at both the JKWTP and the PBSWTP, headloss accumulation across the biofi lters has led to underdrain failures, including blown out mastic seals and stripped IMS cap anchoring screws. Figures 13 and 14 provide photographs of failed IMS caps removed from the JKWTP. Compromised IMS caps create short circuiting through the bio filters, which diminishes treatment performance and backwash effectiveness. Consequently, additional particles and biomass accumulate, exacerbating the original short circuiting problem. The City of Arlington had the failed JKWTP IMS Caps autopsied to determine cause of failure. The autopsies were performed by Cyrus Rice Water Consultants (Pittsburg h Pa.) and included microbial analyses and scanning electron microscopy (SEM) coupled with an energy dispersive x ray. The results of the se tests suggested that the primary foulant was microorganisms and associated biological materials. It was believed that the presence of excess biological materials in the caps was accelerating the collection and entrapment of inorganic foulants as well. D uring operation of the pilot study discussed in this dissertation, Arlington elected to remove all biofilter IMS caps and replaced them with gravel underdrains. Table 1 1 Summary of calculated hydraulic parameters for biofilters Filter no. Average clean bed headloss (ft) Resulting available head for production (ft) 1 5.3 2.1 5 7.9 0.1 9 5.7 2.9 Data collected for JKWTP Expansion II Project Calculated clean bed headloss (Darcys Equation) = 0.5 ft (GAC) + 0.3 ft. (sand) + 0.1 underdrain = 0.9 ft. Average of six different filter runs from May 15, June 14, July 15, August 15, and September 15, 2006, using data from instrument readings for water level in filter boxes, backwash weir, and elevations from record drawings.

PAGE 29

29 Figure 13 Failed IMS cap removed from a JKWTP biofilter with stripped anchoring screw Figure 14 Photograph of failed IMS cap removed from a JKWTP biofilter with blown mastic seal Stripped Anchors Blown Mastic Seal

PAGE 30

30 Water Treatment Performance Concerns The PBSWTP and JKWTP provide regulatory compliant, highquality finished waters. However, occasional process upsets and seasonal loading events have negatively affected effluent aesthetics and stability. Contaminant breakthrough has included Mn an d T&O compounds. Unstable chloramine levels also have been observed at various locations in the distribution system. Manganese JKWTP and PBSWTP effluent Mn concentrations rarely exceed the secondary maximum contaminant level (MCL) of 0.05 mg/L. The ozonation/biofiltration process historically has performed well removing large fractions (>85% ) of soluble Mn via oxidation/filtration. However, seasonally high loads at the PBSWTP and extended periods of low level breakthrough have led to an accumulation of M n precipitates in the distribution system. Small decreases in distribution system redox potential have caused dissolution of the particulate Mn resulting in coloredwater episodes. Distribution system redox can be impacted by decreasing chloramines residu al and/or decreasing dissolved oxygen (DO) caused by microbial regrowth. The City received approximately 47 customer complaints related to colored tap water from January 2008 through October 2009. Tastes and o dors The impounded reservoirs that provide raw water to the PBSWTP and the JKWTP experience seasonal bluegreen algae blooms. These blooms produce metabolic by products that impart T&O to the water source. Two of the most prevalent T&O causing bluegreen algae metabolites detected in PBSWTP and JKWTP finished waters are MIB and trans 1,10 dimethyl trans 9 decalol (geosmin). Although the ozone/biofiltration process may remove 80 to 90% of these compounds, finished water concentrations above the commonly accepted odor threshold

PAGE 31

31 concentration (OTC) of 10 ng/L may still lead to customer complaints. The City received approximately 63 customer complaints related to T&O from January 2008 through October 2009. Chloramine i nstability The City has observed chloramine instability in some parts of the distribution system. Chloramine instability in ozone/biofiltration effluent may have multiple causes, including the breakthrough of particulates, organic carbon, or microbial cells. Pharmaceuticals and p esticides. A quick scan of any newspaper across the United States (U.S.) in early March 2008 reveals the potential impact that pharmaceuticals may have on the drinking water industry. On March 9, 2008, the Associated Press (AP) released a report indicating that their investigative team discovered the presence of pharmac euticals in the drinking water supplies of at least 41million Americans in 24 major metropolitan areas across the country. The City had proactively sampled its raw and finished waters in 2006 to characterize the presence of pharmaceuticals and pesticides. The results from this characterization were disclosed to the AP and released as part of the March 9 report. While detected concentrations were very low and health implications are far from well understood, the concern about drugs in drinking water is now ever present in the minds of utilities, consumers, regulators, and drinking water professionals in general. Congress has already called on the EPA to establish a national task force to study this issue further. In other words, while regulations covering many of these compounds are not on the immediate horizon, utilities must begin to better understand the presence and removal of pharmaceuticals from drinking water supplies. Many pesticides are currently regulated under the Safe

PAGE 32

32 Drinking Water Act and have m aximum contaminant levels ( MCLs) that are included in the Primary Drinking Water Standards. However, the mere presence of these compounds at detectable concentrations in drinking water supplies may also promote negative media and public attention. Trace le vels of pharmaceuticals and pesticides were detected in the Citys source and finished water supplies in the 2006 study Table 1 2 provides a summary of the results from that survey. Although none of the current water quality concerns threatens compliance, the City sought to improve effluent water aesthetics and stability. Table 1 2 Summary of pharmaceuticals and pesticides measured in the City of Arlingtons raw and finished waters Contaminant JKWTP PBSWTP Raw water (ng/L) Finished water (ng/L) Raw water (ng/L) Finished water (ng/L) Sulfamethoxazole 1.4 <0.25 0.62 <0.25 Meprobamate 3.5 1.4 2.6 1.0 Dilantin 1.9 <1.0 1.0 <1.0 Carbamazepine 1.6 <0.50 1.8 <0.50 Naproxen 0.85 < 0.50 <0.50 <0.50 Estrone 0.63 <0.20 0.28 <0.20 Estradiol <0.50 <0.50 17 <0.50 Atrazine 780 240 460 220 DEET 35 <25 <25 <25 TCEP 57 <0.50 <0.50 <0.50 TCPP (Fyrol PCF) 71 <0.50 <0.50 <0.50 Nonylphenol 72 <0.50 83 <0.50 Samples collected and sent for analyses on October 30, 2006 and analyzed at the Southern Nevada Water Authority by Shane Snyder. N,N Diethyl metatolumide. tris (2 charboxyethyl) phosphine. tris (2 chloropropyl) phosphate.

PAGE 33

33 Hypothesis and Objectiv es In conventional practice, turbidity removal and headloss drive the design and operation of both conventional filtration and biofiltration for drinking water treatment in surface water applications Thus, biofilter design parameters are often limited to media configuration, loading rate, and backwash strategy. While these parameters can significantly impact biofilter performance, their influence on improved biological activity is largely passive. The biological removal of dissolved organic and inorganic contaminants is an anticipated benefit of biofiltration. However, common design and operational practice does not seek to enhance the biological activity responsible for those mechanisms. Furthermore, current biofilter operational practices primarily focus on biofilm control to maintain hydraulic performance, often to the detriment of biological activity (and optimal water treatment) The hypothesis of this work is both water treatment and hydraulic performance of a biofilter can be improved by modifying influent conditions for enhanced biological activity. The purpose of this research was to identify strategies to enhance the biological activity in a biofilter without compromising productivity or particulate removal performance. Specific objectives included: Evaluate potential biofilter enhancement strategies comprised of dosing low levels of common drinking water treatment chemicals at a feed point just upstream of a biofiltration process. These chemicals were added to provide substrate, nutrient, and/or oxidant optimization of the biofilter process influent. Investigate biological drinking water treatment process fundamentals (e.g., microbial ecology, bacterial metabolism, and contaminant removal mechanisms) to understand how DOC, MIB, geosmin, and Mn c an be removed effectively in a single treatment step Biological clogging (filter headloss) can be minimized

PAGE 34

34 The ultimate goal of this work is to shift an industry accepted paradigm so that the design and operation of biofilters are driven not only by filtration but also by biological treatment objectives. Approach The research study included ten months of biofiltration enhancement pilot scale testing at the JKWTP to evaluate methods for restoring and enhancing the performance of the Citys ozone/biofil tration process. This evaluation entailed both a characterization and evaluation of biological activity in the filters and an examination of potential enhancement strategies. The premise of this work is that small changes in filter feed conditions could gr eatly improve the health and activity of the microbial community in the biofilters, consequently enhancing performance. The strategies tested were selected based on previously published literature and industry experience. The following studies were perform ed to meet research objectives: 1. Characterize the baseline performance of the JKWTP operating under existing conditions. This included an assessment of the systems ability to meet the Citys current treatment objectives. 2. Evaluate primary substrate augmentation enhancement strategies. 3. Evaluate nutrient augmentation enhancement strategies. 4. Evaluate peroxide supplementation for augmenting the oxidative action and response of the biofiltration process 5. Identify and track the microbial communiti es active present in the biofilters. 6 Develop full scale design and operating parameters for implementation of the recommended modified biofiltration process at the Citys WTPs and estimate the associated capital and production costs

PAGE 35

35 CHAPTER 2 LIT ERATURE REVIEW The purpose of this chapter is to provide a summary of both conventional and state of the art understanding with respect to biofiltration fundamentals, contaminant applicability biofilter design, operations and monitoring. The basis for th e tested enhancement strategies is also discussed. Historical Background Perhaps the first recognized commercial use of filtration for water treatment occurred in 1804 at a bleachery in Paisley, Scotland (Huisman and Wood 1974). The treatment system was an experimental slow sand filter, designed and built by John Gibb. Surplus water produced by the filters was sold to the public at a halfpenny per gallon (Baker 1948). These early slow sand filters could effectively remove turbidity and pathogenic organisms. Furthermore, these slow sand filters were not only the first effective municipal water treatment system they were also likely the first example of drinking water biological treatment. The design offered indigenous micr oorganisms ample surface area and detention time to degrade nutrients and organic compound s present in the source water. The success of this system to improve water aesthetics and public health lead to metropolitan use of slow sand filters across Europe and the United States by the late nineteenth century ( Johnson, 1914; Huisman and Wood, 1974). However, the loading rates of slow sand filters were low, typically 0.04 to 0.08 gallons per minute per square foot (gpm/ft2). Thus, large areas of land were required to accommodate the facilities for even modest urban communities. Reliable slow sand filter treatment also required low turbidity influent (<10 NTU) and high operator effort for media cleaning and replacement. As populations and water demands increased, the

PAGE 36

36 space, labor, and source water requirements for slow sand filters became unfeasible in many applications. In 1896, the first rapid rate sand filters were installed in the United States, utilizing an expandedbed backwash with water and air supply (Baker 1948). Rapid rate sand filters provided filtration at rates from 2 to >10 gpm/ft2, requiring significantly less space than earlier filtration technologies. In addition, rapid rate filters provided better performance for treating water with high, or flashy, influent turbidity (EPA, 1990). Coincidentally, implementation of rapid rate filtration coincided with the widespread use of chlorine as a primary disinfect ant for cholera and typhoid fever control (Leal 1909). By 1925, there were approximately 18 million U.S. residents served by utilities employing rapid rate sand filters, and only 5 million served those with slow sand filtration (Gillespie, 1925). Thus, most drinking water filters in the U.S. moved away from bi ological treatment to conventional filtration by the early twentieth century. Today, suppressedbiological (chlorine present) rapid rate granular media filtration remains the most common system to remove particulate matter in water treatment systems. Indee d, biological drinking water treatment was limited in the United States until recently (Urfer et al ., 1997), likely due to public perception and resistance to the intentional use of microorganisms within water treatment processes (Evans 2009). However, new regulatory and technological developments are driving more utilities to consider the use of biological processes to treat their drinking water. These developments include (1) the promulgation of regulations further limiting the formation of DBPs; (2) the emergence of ozone for taste, odor, and color control; (3) the increased awareness that biologically stable water must be produced in the treatment plant to avoid the negative impacts of biological activity in the distribution system (e.g.,

PAGE 37

37 disinfectant demand, aesthetic upsets, and corrosion); and (4) the push for green technologies. Configurations of Biological Treamtent There are numerous forms and configurations of biological treatment processes for drinking water (Rittman n and McCarty 2001). A small number of drinking water biological treatment technologies operate as suspended growth systems, in which freefloating bacteria and other microorganisms are hydraulically maintained within a reactor. However, most are operated as biofilm systems, including fixed bed biofilm systems, fluidizedbed biofilm systems, and hollow fiber membrane biofilm reactors (Brown, 2007a). Among the various biological treatment configurations used for drinking water applications, fixedbed biofilm reactors are the most common. This process includes a biogrowth support medium such as sand, anthracite, and/or GAC on which microbial communities attach and grow. The most common fixedbed biofilm processes are positioned just upstream of final disinfection and serve not only as bioreactors to oxidize organic and inorganic contaminants but also as filters to remove particles, thus the name biofilter. In typical configurations, biofilters are fed post coagulated/flocculated settled water (with or without ozonation). The biofilters s erve as the final step to remove remaining suspended particles and unsettled floc. Therefore, the biological action of these systems must not compromise their ability to meet filtration design criteria (i.e., turbidity limits and unit filter run volumes). The following sections provide additional detail on the contaminant treatability design parameters, conventional and enhanced operational procedures, and monitoring techniques for biofilters in water treatment applications.

PAGE 38

38 Contaminant Treatability Overview Biofiltration can be used to treat a wide range of organic and inorganic contaminants in surface water and groundwater. The following sections provide a general description of the current state of knowledge on the effectiveness of drinking water biofiltration for the contaminants tracked during this study. Natural Organic Matter Nearly all surface water utilities and many groundwater utilities are affected by organic carbon, which can (1) cause organic or biological fouling within any unit process, (2) exert an oxidant demand, (3) serve as a DBP precursor, and (4) cause biological regrowth in distribution systems (LeChevallier et al. 1992; Escobar et al. 2001) recommended that distributed assimilable organic carbon ( AOC ) concentrations be less than 100 g/L to limit regrowth. Numerous studies have investigated the design and operation of biological filters to reduce organic carbon concentrations. Biofilter organic carbon removal efficiencies have been reported to vary bet ween 5 and 75% with typical removal s of 10 to 20% (Bouwer and Crowe, 1990, Raymond et al. 1995 ; Urfer et al. 1997). Organic carbon removal can be correlated to biofilter activity, as measured by dissolved oxygen (DO) uptake, adenosine triphosphate (ATP) and/or phospholipid fatty acids, which may be a ffected by feed water source and characteristics, operating temperature, backwash protocol, disinfectant applicat ion, and ozone dose (Huck et al. 2000; Liu et al. 2001 ; Westerhoff et al. 2005 ; Wert et al. 2008).

PAGE 39

39 MIB and Geosmin An increase in blooms of cyanobacteria (bluegreen algae ) in freshwaters has been a global phenomenon in recent years. Blue green algae blooms occur on every continent, excluding Antarctica, and in over 30 states in the U.S. (Carmi chael 2001). These blooms are responsible for producing algal metabolites that can impart tastes, odors, and toxicity to surface waters. In areas that rely heavily on surface water for drinking water supply, these metabolites may limit a critical resource by degrading water aesthetics and safety. The presence of objectionable T&O compounds in surface water supplies is a growing problem facing drinking water utilities across the U.S. and worldwide. Two of the most common surface water T&O causing compounds, MIB and geosmin, are metabolites of bluegreen algae. MIB and geosmin have an earthy/musty T&O that can be detected by human senses at concentrations as low as 10 ng/L. Though MIB and geosmin are not regulated, the importance of their removal from drinkin g water sources is critical for maintaining consumer confidence. Mallevialle and Suffet (1987) suggested that the majority of consumers judge the safety of their water by appearance, taste, and smell. Therefore, utilities must remove T&O compounds, though they are under no regulatory pressure to do so. Suffet et al (1995) also noted that 128 of 800 utilities surveyed in the U.S. experience serious T&O problems and that, on average, U.S. utilities spend approximately 4.5% of their total treatment budget on T&O control. Conventional approaches for mitigating blue green algae metabolite impacts on drinking water include (1) bluegreen algae population control through reservoir management, and (2) metabolite removal at the water treatment facility. Reservoir management (e.g., copper dosing) has shown some success at reducing algal

PAGE 40

40 populations and the resulting production of bluegreen algae me tabolites. However, the growth of copper resistant bluegreen algae strains and increased nutrient loading to surface waters limit the effectiveness of this approach. Furthermore, conventional treatment methods, such as preand post chlorination, coagulat ion, sedimentation, and filtration, are marginally effective at reducing algal metabolite concentrations. Powdered activated carbon (PAC) is often supplemented to these processes to achieve adequate removal, though its use can be cost prohibitive over leng thy or intense algal events. Recently, many utilities have begun to employ the combination of ozone and biofiltration as a strategy to mitigate MIB, geosmin, and other T&O compounds. In this process, ozonation typically achieves 30 to 75% removal of MIB an d 40 to 60 % removal of geosmin at dosages of 1.5 mg/L (Lundgren et al. ; 1998 ; Nerenbe rg et al. 2000; Westerhoff et al. 2005). Biofilters operated downstream of ozonation typically achieve an additional 50 to 80 % of the remaining MIB and geosmin (Nerenbe r g et al. 2000; Westerhoff et al. 2005 ; Elhadi et al. 2006). MIB is similar in structure to the bicyclic monoterpene, camphor Camphor is metabolized by pathways that involve the sequential cleavage of each ring by formation of unstable lactones (Trudgi ll, 1984). MIB biodegradation probably follows a similar pathway, i.e., ring cleavage following formation of each lactone. However, isoborneol (MIB without the methyl group bonded to the same carbon as the hydroxyl group) was degraded at a faster rate than MIB and was unable to enrich microbial populations for MIB degradation (Izaguirre et al. 1988a,1988b). The tertiary alcohol of MIB may be responsible for the slower kinetics by making it more difficult for enzymes to oxidize the alicyclic ring to the ali cyclic ketone prior to lactone formation. Despite relative

PAGE 41

41 recalcitrance, microbial isolates capable of degrading MIB to levels below the odor threshold concentration ( OTC ) have been identified. Previously isolated MIB degrading bacteria include Pseudomonas sp Flavobacterium sp and Bacillus subtilis (Ishida and Miyaji, 1992; Egashira et al. 1992; Lauderdale, 2004) Manganese Removal The presence of Mn in drinking water creates aesthetic challenges for utilities. Mn may produce brown, black or purplecolored water events, depending on the concentration and oxidative reduction potential (ORP). A 200 1 study summarized the available utility Mn occurrence data, indicating that 20% and 4.4% of drinking water systems treating groundwater and surf ace water, respectively, had raw water Mn levels above 0.05 mg/L (WaterRF, 200 1 ) Effluent Mn concentrations less than the secondary MCL may also present concerns for utilities as low level breakthrough can lead to an accumulation of Mn precipitates in the distribution system. Small decreases in distribution system ORP may cause dissolution (resuspension) of the particulate Mn resulting in colored water episodes (Kohl and Medlar 2006). Distribution system ORP can be impacted by changing disinfectant residual and/or by biofilm induced DO fluctuations. Particulate Mn (Mn+4) is easily managed in most WTPs using sedimentation and filtration processes (Knocke et al. 1990). However, soluble Mn (most commonly Mn+2 in surficial waters) is much more difficult to treat. The conventional approach for soluble Mn treatment is the continuous application of free chlorine to increase the ORP sufficient ly to oxidize Mn+2 to Mn+4. This change in oxidation state precipitates t he soluble Mn which then plates on the filter media. This strategy requires continuous maintenance of the applied chemical oxidant, as the particles formed are only stable at

PAGE 42

42 the ORP at which they were formed. Interruption of oxidant application leads to the reduction of Mn+4 to Mn+2 and the release of soluble Mn in the finished water. Recently, DBP concerns have forced utilities to consider alternative oxidants such as ozone. O zone oxidation of Mn may lead to the formation of colloidal Mn which may be to o small to be effectively removed during filtration (Kohl and Medlar, 2006). By adjusting the ozone dose, it is possible to improve the rate and extent of particulate Mn formation, thereby minimizing colloidal Mn release to the distribution system. However the optimal dose is controlled by multiple factors including ozone demand and water temperature. An alternative Mn treatment technology is biofiltration. Biofilters take advantage of the Fe and Mn oxidizing ability of certain autotrophic bacteria (Kohl et al. 2006). The bacteria (e.g. Metallogenium sp ., Burkholderia sp Pseudomonas sp p Bacillus sp p ) oxidize soluble Mn+2 to Mn+ 4, which readily precipitates as MnO2 under most filter conditions Mn oxidation may be driven through oxidase enzymes (intra and/or extracellular) or be promoted by metabolic by products that sufficiently elevate pH or ORP ( Diem and Stumm 1984; Nealson 1992). The biologically oxidized Mn becomes integrated (adsorbed) within the biofilm filter media matrix. Surficial ch arges within the matrix maintain Mn stability, preventing desorption ( Sahabi et al. 2009). Excess biofilm and sloughed microbial aggregates containing Mn are removed from the filter media during backwash procedures. Drinking water biofiltration for Mn removal has been practiced for decades in the U.S. and Europe (WaterRF, 2001) Despite success, there are process challenges during facility startup and operational upsets. Indeed, nonacclimated or disrupted systems may take months before desired steady stat e Mn removal performance is achieved. The potential

PAGE 43

43 rapid release of Mn from filter media matrices following an ORP shift remains a significant concern for utilities considering conversion of existing chlorinetreated filters to a biological mode. Pharmaceuticals and Pesticides A quick scan of any newspaper across the U.S. in early March 2008 revealed the potential impact that pharmaceuticals and pesticides may have on the drinking water industry. On March 9, 2008, the AP released a report indicating that their investigative team discovered the presence of pharmaceutical drugs in the drinking water supplies of at least 41 million Americans, in 24 major metropolitan areas across the c ountry (AP 2008). While detected concentrations were very low and h ealth implications are far from well understood (Snyder et al. 2003; Snyder et al. 2005; Snyder et al. 2007 the concern about drugs in drinking water persists with utilities, consumers, regulators, and drinking water professionals in general. Congress has already called on the EPA to establish a national taskforce to study this issue further. While regulations covering pharmaceuticals and pesticides are not on the immediate horizon, utilities must begin to better understand the presence and removal of p harmaceuticals and pesticides from drinking water supplies. The literature suggests that ozonebiofiltration may effectively remove some trace organic compounds (Snyder et al. 2003; Snyder et al. 2005; Al Rifai et al. 2007; Guay et al. 2007; Snyder et al. 2007). Generally, ozonation alone may effectively treat compounds containing amine groups, phenolic groups, and unsaturated carbon structures. However, many other trace compounds are not removed effectively by ozonation. Snyder et al. (2007) summarize d the removal of selected trace organics with ozone. Table 21 presents the summarized results.

PAGE 44

44 Table 2 1. Summary of minimum % removal of selected pharmaceuticals and pesticides by ozonation at drinking water dosages (Snyder et al. 2007) >95% removal > 80% removal 50 80% removal 20 50% removal <20% removal Acetaminophen Androstenedione DEET Atrazine TCEP Carbamazepine Caffeine Diazepam Iopromide Diclofenac Pentoxifylline Dilantin Meprobamate Erythromycin H 2 0 Progesterone Ibuprofen Estradiol Testosterone Estriol Estrone Ethynylestradiol Fluoxetine Gemfobrozil Hydrocodone Naproxen Oxybenzone Sulfamethoxazole Triclosan Trimethoprim Biofiltration is an effective post treatment for ozonation, as many pharmaceuticals and pesticides may be removed through biotransformation (Fuerhacker et al. 2001 ; Snyder et al. 2003 ; Snyder et al. 2005; Snyder et al. 2007 ; Al Rifai et al. 2007; Guay et al. 2007). Snyder et al. (2007) summarized the removal of trace organics from ozonebiofiltration pilot testing. In general, the removals of the tested trace organic compounds were low. Table 22 presents a summary of the results. Table 2 2 Summary of % removal of selected pharmaceuticals and pesticides by pilot biofilters (with ozone pretreatment) (Snyder et al. 2007) >95% removal >80% removal 50 80% removal 20 50% removal <20% removal Fluoxetine Benzo[a]pyrene Progesterone Androstenedione Acetaminophen DDT Erythromycin H 2 0 Atrazine Fluorene Caffeine Testosterone Carbamazepine Triclosan DEET Trimethoprim Diazepam Diclofenac Dilantin Estradiol Estriol Contaminants detected (<1 g/L) in City of Arlington finished waters

PAGE 45

45 Convention Biofiltration Design Considerations Flow Equalization In general, water treatment processes operate the most effectively under uniform, steady hydraulic conditions. Research has shown that transient hydraulic and biodegradable organic matter (BOM) loading episodes can negatively impact biofiltration performance ( Manem and Rittmann, 1992; Ahmad et al ; 1998). Therefore, biofilters should be operated in a constant rate mode, and adding flow equalization capabilities upstream of biofilters might be considered if a given system often experiences wide swings in demand. Some degree of flow equalization can also be achieved by designing redundancy into a biofilter process. Filters can be taken in and out of service as necessary to buffer changes in hydraulic conditions caused by backwashing events, changes in demand, or filter repair. However, redundancy must be balanced by the resulting biofilter downtime schedule, as downtime can have a deleterious effect on biological activity ( Niquette et al. 1998; Vokes 2007). Other design features that can help minimize flow changes include: (1) increasing the size of filter to waste piping to allow operation at the capacity of an online filter, (2) installing automated controls and a filter to waste modulating valve to allow for a smooth transition when a filter is taken out of service for backwashing, and (3) adding a flow control valve to the filter waste washwater line. Pre treatment Coagulation, flocculation, and sedimentation processes can greatly decrease the particle loading rate to biofilters, thereby increasing biofilt er run times. Coagulant or filter aid polymer can also be added just upstream of a biofilter to minimize filtration turbidity removal ripening periods. Water utilities commonly couple biological filtration with

PAGE 46

46 ozonation for improved water treatment perfor mance (Krasner et al. 1993; Price et al. 1993; Westerhoff et al. 2005; Emelko et al. 2006; Wert et al. 2008). After ozone is dosed to the process, it immediately creates both molecular ozone (O3) and hydroxyl radicals (*HO), both of which are strong oxidants. The O3 and *HO oxidize complex, recalcitrant organic matter into lower molecular weight organic matter, rendering it more labile. Furthermore, the increased concentration of BOM from ozonation increases biomass in the filter, which can improve the degradation of trace organic contaminants (Dewaters and DiGiano, 1990 ; LeChevallier et al. 1992; Rittmann 1995). The increase in BOM across an ozonation pretreatment step is a function of the ozone to TOC ratio (Juhna and Melin, 2006). Chlorination can also be used to increase BOM upstream of biofiltration, but a dechlorination step must be included to minimize the impact to the microbial community in the biofilters ( LeChevallier et al. 1992; Zappia et al. 2007). Unlike free chlorine, ozone residual is consumed within a few minutes after application and poses little risk of microbial inactivation in the biofilter Media Selection Biofilter granular media serves as both the particle screening mechanism and the biogrowth support structure. Biofilters are typically dual media with sand as the base layer and anthracite or GAC as the top layer. Anthracite/GAC selection drivers include cost, need for adsorptive capacity, and fluidization characteristics (Najm et al. 2005). Another important consideration is the medias capacity to support biological activity. Media size selection plays an important role in the available surface area for biological attachment which has implications for improving BOM degradation (Zhang and Huck 1996). However, improved BOM d egradation must be balanced with filter hydraulic performance when considering media size. Media type will also affect

PAGE 47

47 available surface area. GAC has been shown to house three to eight times more biomass than does anthracite, which is likely due to the increased surface area available for microbial attachment (LeChevallier et al. 1992 ; Wang et al. 1995). It has been demonstrated that temperature, chlorinated backwash water, and transient hydraulic conditions can be detrimental to biofiltration performanc e ( Manem and Rittman, 1992; Miltner et al. 1995; Ahmad et al. 1998; Moll et al. 1999; Andersson et al. 2001; Emelko et al. 2006). GAC may provide more robustness over anthracite under these challenging conditions due to high levels of biological activity (Emelko et al. 2006 ; Liu et al. 2001; Wang et al. 1995). Media hardness must also be considered when selecting media, as attrition is a larger concern for GAC than for anthracite. Backwash System An effective backwash system is critical to the successful hydraulic and water treatment performance of any biofilter. Therefore, careful consideration must be given to its design. Some studies have observed a reduction in BOM removal in biofilters over filter run time, suggesting that filter run time (backwash interval) may play an important role in biological treatment performance (Ahmad et al. 1998). Application of chlorine and chloramines to biofilter backwash has been shown to have detrimental effects to biofilter BOM and soluble Mn removal and should be avoided for optimal biofilter performance (Moll et al. 1999; Liu et al. 2001 ; Vokes, 2007). Chlorine application solubilize s Mn matrices present on the filter, as well as biofilm containing absorbed mang an ese. Other backwash design factors include source water, backwashing duration, backwashing rate, hydraulic loading rate, backwash wastewater handling, air scour, and surface wash.

PAGE 48

48 Conventional Biofiltration Operation Considerations Loading Rate and Contact Time Together with the L/d (length [depth] of media/media particle diameter) criteri on, hydraulic loading rate (HLR flow rate applied per unit filter area) is a key design parameter for sizing conventional granular media filters. Excellent particulate filter effluent water quality can be achieved even at high HLRs (10 gpm/ft2), though headloss may become a constraint (AWWA 1999). The HLR and filter bed depth determine the empty bed contact time (EBCT), which is considered a vital criterion for effective biofiltration (DeWaters and Digiano, 1990; LeChevallier et al. 1992; Zhang and Huck 1996b; Urfer et al. 1997; Carlson and Amy, 1998; Huck et al. 2000; Westerhoff et al. 2005). The selection of a design EBCT is depend e nt on hydraulic constraints and ultimate water quality objectives. Studies have shown that 90% of biofilter influent BOM may be removed with a design EBCT of 1020 min (Provost et al. 1995). EBCT is particularly important for biofilter applications targeting trace organic compounds, such as MIB, geosmin, endocrine disruptors, and pharmaceutically active compounds, which are more recalcitrant to biodegradation than is typical BOM (Weste rhoff et al. 2005 ; Brown 2006 ; Lim et al. 2008). Acclimation Procedures One consideration for any biological process is the amount of time required to reach steady state biological activity. This is a concern both during the initial start up and after any process upsets or downtime, as significant breakthrough of organic compo unds may occur. Liu et al. (2001) found that 20 to 40 days was required to reach steady state biofilter removal of easily degradable BOM at 20C. If GAC is used as biofilter support media, most of the initial DOC removal occurs through physical

PAGE 49

49 adsorption while the bacteria in the associated biofilm are acclimating. This synergism may last approximately 2 to 3 months until the GAC is exhausted and/or the biological activity has reached steady state (Servais et al. 1994). The elimination of chlorine in the backwash water may decrease acclimation time and improve the general robustness of biofilter, particularly at low temperatures. When targeting trace organic compounds such as MIB and geosmin, acclimation periods can be several months ( Westerhoff et al. 2005; Chae et al. 2006). This suggests that acclimation may promote an increase in the microbial populations involved in trace organic degradation. Seeding virgin biofilters with biological media that have already been acclimated to the biodegradation of MI B and geosmin may reduce start up times (Westerhoff et al. 2005) although it remains unknown whether the acclimation is to MIB and geosmin (secondary substrates) or to the background NOM (primary substrates) Nutrient and/or primary substrate addition al so ha ve the potential to decrease biological acclimation time due to increased rates of cell growth (Hozalski and Bouwer 2001; Lauderdale and Brown, 2007) Lastly, minimizing biofilter downtime is important in maintaining sufficient biological activity (N iquette et al. 1998 ; Vokes, 2007). Thus, an appropriate cycling of standby and duty biofilters must be established. Backwash Protocol A common practice to control growth in biofilters is backwashing. Numerous backwashing criteria can be considered, including frequency, air scour, surface wash, backwash rate, duration, % bed expansion, addition of chlorine, and pulsing. Removing bound organic/inorganic substances helps maintain uniform hydraulic conditions and controls the headloss across the biofilter bed. Backwashing removes considerable BOM and inactive microorganisms from the biofilter (Brown, 2006) and decreases the local

PAGE 50

50 nutrient concentration around the microbial biofilm by using post filtered water. Therefore, backwashing exposes the biofilm to a potential adverse environmental condition with limited nutrient availability (Simpson, 2008). Although a fraction of the active bacterial biomass fixed on GAC may also be eliminated during backwashing (Dussert and Van Stone, 1994 ; Kim and Logan, 2000), the biofilm is generally very resistant to the high shear stress rates that are directed onto the GAC surface (Scholz and Martin, 1997). However, over aggressive backwashing (intensity or frequency) can leave insufficient biological activity to meet treatm ent objectives, can lead to BOM breakthrough, and can hamper the stability of final disinfectants. Substantial research has been conducted to evaluate the impacts of various backwashing criteria on biological treatment performance ( Miltner et al. 1995; Wa ng et al. 1995; Urfer et al. 1997; Emelko et al. 2006; Huck et al. 1998; Ahmad et al. 1998; Choi et al. 2007 ; Vokes 2007; Brown 2007b ; Marda et al. 2008). In the water treatment industry, biofilter backwash frequency may range from less than 12 hours to more than 48 hours. Backwash episodes are typically triggered by high headloss across the biofilter, biofilter run times, or turbidity breakthrough. As state d previously, biofilm growth may also be controlled through application of chlorine/chloramines to backwash water. However, this practice may inhibit optimal biological treatment performance through the inactivation of viable organisms. Indeed, the presenc e of chlorine/chloramines in biofilters may result in both a decrease in net biological activity and a shift in microbial communities to organisms that may disrupt distribution stability. Norton and LeChevallier (2000) found an increased occurrence of gram positive organisms in biofilters regularly conditioned with chlorine application.

PAGE 51

51 Gram positive organisms have a thick pept id oglycan layer in the cell wall (relative to gram negative bacteria) that may increase resistance to final disinfection (Norton and LeChevallier 2000; Madigan et al. 2009) In addition, the application of chloramines may select for chlorine resistant nitrifying bacteria (Eichler et al. 2006). Engineered Biofiltration Operational Considerations Primary Substrate Augmentation By medi ating the transfer of electrons between electron donor substrates and electron acceptor substrates, microorganisms gain the energy they need to grow and maintain cell metabolism. Primary substrates provide energy during cell metabolism. As described by Monod in the 1940s, the rate of cell synthesis is proportional to the concentration of the ratelimiting primary substrate (i.e., concentration of limiting electron donor or electron acceptor), the cell yield, and the maximum specific primary substrate utiliz ation rate. Equation 21 below further describes Monods equation. S K Ss (2 1) Where: : specific growth rate : maximum growth rate S : concentration of limiting substrate (electron donor or acceptor) sK : half saturation coefficient for the limiting substrate (electron donor or acceptor) The minimum concentration of primary substrate that can support steady state biomass is known as Smin. When the ratelimiting primary substrate concentration equals Smin, the rate of cell synthesis equals the rate of cell decay. Any substrate that is present below its Smin concentration is known as a secondary substrate. Though

PAGE 52

52 secondary substrates c an be biodegraded, bacteria gain little to no energy in doing so, which means that a primary substrate must be biodegraded simultaneously if any biodegradation of the secondary substrate is to be achieved (Kobayashi and Rittmann, 1982; Stratton et al. 1983; Namkung and Rittmann, 1987a, 1987b). The rate of secondary substrate degradation is proportional to the concentration of active biomass present, which is a function of the concentration of primary substrate, and other variables. The kinetics of secondary substrate degradation is driven by Equation 22 : X S K S q rs s s ^ s ut (2 2) Where: rut : the rate of utilization of a secondary substrate ( mg( secondary substrate) cm3h1) sS : Concentration of the secondary substrate (mgs cm3) ^ sq : maximum specific rate of secondary substrate utilization (mgs mgx 1h1) sK : half maximum rate c onstant for the secondary substrate (mgs cm3) X : concentration of active bacteria (mg/ cm3) Because trace organic compounds like MIB, geosmin, endocrine disruptors (EDCs), and pharmaceutically active compounds are present at ng/L or g/L concentrations in most natural waters, they likely are biodegraded as secondary substrates. Thus, biological t reatment processes designed to biodegrade these compounds require the presence of a primary substrate (Lauderdale et al. 2007). Increasing the concentration of primary substrate may increase the rate of trace organic compound degradation (Lim et al. 2008), decrease bioacclimation time requirements (where bioacclimation is the time required after startup for a biofilter to reach steady state treatment performance) and enhance the robustness of the system during process

PAGE 53

53 upsets (Lauderdale and Brown, 20 07). It can also enhance the removal rate of slowly degradable organic material (Hozalski and Bouwer 2001). Primary substrate augmentation is most commonly conducted in biofilters with ozone pretreatment. Ozonation breaks large natural organic molecules i nto smaller, more readily biodegradable organic molecules (Becker and OMelia, 1996), thereby increasing the concentration of primary substrate. Thus, adjusting the ozone dose may improve the rate of both primary substrate (i.e., macroorganic carbon that contributes to biologically degradable organic carbon [BDOC], DBP formation potential, and chloramine demand) and secondary substrate ( e g ., MIB, geosmin, EDC removal) degradation. Nutrient Augmentation The r ole of EPS in m icrobial s ystems Nutrients play an important role in the microbial production and secretion of EPS (Liu et al. 2006), which form an adhesive gelatinous matrix comprising mostly insoluble exopolysaccharides. EPS may also consist of proteins, nucleic acids, lipids, and humic substances (Fl emming et al. 2000) The presence of EPS can be observed through microscopy as a rough, matted material or as filamentous strands (Flemming and Wingender 2001). EPS is an essential component of biofilms, as it promotes surface attachment and maintains st ructural stability. The EPS matrix creates the biofilm (or floc culant ) that provides a competitive advantage to all organisms present. EPS in biofilm drives the surface charge, hydrophobicity, and noncovalent interactions of microbial cells (Flemming et al 2000). Bacteria held within a biofilm may establish and maintain their position over a long period of time, compared to those that are planktonic. Biofilm EPS also provides bacteria with protection from xenobiotics, high/low pH, oxidants, shear, and other forms of stress for the associated microbial community

PAGE 54

54 (Sutherland, 2001; Mauclaire et al. 2004; Wang et al. 2008). Indeed, the stimulation of expression of EPS in biofilms is affected greatly by these environmental parameters and the availability of nutrients ( Sutherland 2001; Liu et al. 2006; Fang et al. 2009). Studies have shown that the amount of EPS synthesis within a biofilm is dependent on the (1) availability of carbon substrates, and (2) balance between those substrates and other nutrients, s uch as nitrogen (ammonia) and phosphorus (orthophosphate) ( Sutherland, 1977; Sutherland, 2001 ; Priester et al. 2006; Fang et al. 2009). Indeed, the production of EPS serves as a competitive advantage in low nutrient conditions as it may help sequester nutrients and other factors necessary for microbial growth (Decho, 1990). EPSr elated c hallenges in w ater t reatment EPS may be the primary contributor to biofilter clogging. In a study by Mauclaire et al (2004), biological material reduced the pore space of a clogged biofilter by 7 to 13% whereas pore space reduction due to particle deposition was not greater than 7% The volume of microbial cells did not exceed 0.2% of the pore volume, indicating that e xtracellular biological materials play a significantly larger role in filter clogging than do the microbial cells themselves. In the Mauclaire et al. (2004) study, filter clogging was defined as a condition where hydraulic conductivity is reduced by a fact or of 4. Other studies have suggested that the proportion of EPS in biofilms can be as high as 90% of the total organic matter (Christensen and Characklis 1990; Nielsen et al. 1997). EPS also accelerates the sequestration of minerals into the biofilm, which may also adversely affect the biofilter production and further increase clogging (Ali et al. 1985 ; Rinck Pfeiffer et al. 2000). One study has shown that calcium and other mineral precipitation can occur preferentially in EPS matrices, via proton exchange and/or sequestion of suspended precipitates and subsequent crystallization of dissolved

PAGE 55

55 species, leading to deposits with remarkable mechanical stability (Flemming and Wingender 2001). EPS may also cause microbial influenced corrosion (Flemming and W ingender 2001b). The interaction and sequestration of Fe and other metals in EPS present in distribution pipes can increase the kinetics of metal corrosion (Beech et al. 1997). Indeed, high levels of EPS have been found on copper surfaces creating signif icant corrosion; it is possible that excess amounts of EPS provide microorganisms a protective mechanism against the toxicity of copper (Flemming and Wingender 2001b) Nutrient limitations can be determined based on the relative concentrations of carbon and nutrients in the biofilter feed. The empirical formula for a bacterial cell is C55H77O22N11P (Metcalf and Eddy 2002), indicating that a stoichiometric C:N:P molar ratio of approximately 55:11:1 is required to avoid a nutrient limited condition. However, nitrogen and phosphorus are more conservative than carbon substrate in an aerobic biological system, which may oxidize to carbon dioxide. In environmental systems, C:N:P ratios as low as 100:10:1 have been shown to eliminate microbial stress and p romote optimal growth (USEPA 1991; LeChevallier et al. 1991). The 100:10:1 C:N:P stoichiometric ratio is equivalent to a concentration ratio of 1 mg/L bioavailable carbon substrate: 0.117 mg/L NH4N: 0.026 mg/L PO4P. Nitrate (NO3 -N) may also serve as an inorganic source of nitrogen in aerobic environments (Madigan et al. 2009). However, NH4N can be incorporated directly into amino acids by most bacteria, without requiring intracellular reduction (Fenchel et al. 2000). Nutrient limitations are an important consideration for most drinking water biofilters due to the common pretreatment processes of coagulation and sedimentation. Common coagulants such as aluminum

PAGE 56

56 phosphate, polyaluminum phosphate, and ferric chloride are extremely effective at removing phosphorus (Tchobanoglous et al. 1991, WEF 1998). Biofilters downstream of these processes will likely receive feed waters that have phosphorus concentrations below 0.01 mg/L (Nishijima et al. 1997; WEF 1998), thereby creating a phosphorus limited condition if the concentration of biodegradable organic substrates is greater than 0.38 mg/L. Nutrient supplementation may have a twofold benefit to water utilities operating biofilters under nutrient limited conditions: (1) it may enhance biofilter biologic al activity both increased biomass and substrate utilization, which could lead to improved water treatment performance, and (2) it may inhibit EPS production (Nishijima et al. 1997 ; Sang et al. 2003; Ryu et al. 2004; Brown et al. 2009 ; Li et al. 2009), which may improve biofilter hydraulic performance. The transition from a phosphorus/nitrogen limitation to a biodegradable carbon limitation is believed to be the driver behind these benefits. As the carbon substrate becomes limited with nutrient s upplementation, microbial cells may utilize more carbon for cell synthesis and energy production instead of producing the EPS (Mauclaire et al. 2004). Studies have shown that nutrient supplementation can increase microbial concentrations and may enhance t he ability of microorganisms to degrade contaminants (Juhna and Rubulis 2004). Furthermore, the decreased biofilter EPS creates a biofilm structure that may be more susceptible for removal through filter backwashing. Oxidant Augmentation with Peroxide An other approach for enhancing biological activity and performance in a biofilter is through oxidant augmentation with peroxide. Hydrogen peroxide provides microorganisms with an additional source of dissolved oxygen and induces certain

PAGE 57

57 microbial communities in the bioreactor to express peroxidasefamily oxidoreductase enzymes such as peroxidase that catalyze the oxidation of organics present in the biofilter feed (Pardieck et al. 1992). Indeed, a wide variety carbon substrates are oxidized in microbial cell s by peroxidase family oxidoreductase enzymes (Schumb et al. 1955). Peroxidasecatalyzed transformation of aromatic compounds, such as phenols, amines, and polyphenols, from wastewaters was first reported by Klibanov et al. (1980) and has since received m uch attention. When peroxidase and hydrogen peroxide are added to water containing these compounds, the enzyme undergoes a catalytic cycle that leads to the release of free radicals. These radicals undergo various reactions. Free radicals may participate i n the oxidation of recalcitrant organic compounds. Auriol et al. (2007) demonstrated that four endocrinedisrupting compounds can be oxidized using horseradish peroxidase. Based on this evidence, the peroxidase enzymes naturally present within microbial sy stems may catalyze reactions between peroxide and organic substrates forming free radicals that lead to the removal of TOC, MIB, geosmin, and EDCs Furthermore, compounds that are not readily available substrates to the enzyme may be oxidized through radic al transfer. In addition, studies suggest that low dosages of hydrogen peroxide (<1 mg/L) may effectively oxidize and remove inactive biomass and EPS, while preserving biological activity (Christensen et al. 1990 ; Neyens et al. 2002). EPS oxidative reductive depolymerization results in cleavage of linkages in the polymeric backbone (Christensen et al. 1990). This reaction involves a series of free radical reactions, which ultimately lead the cleavage of polymer chains and efficient removal. The mechanis m for this action may be driven through Fentons reaction (ferrous catalysts). However,

PAGE 58

58 Fentons reaction may be limited in most drinking water biofilter applications, as efficiency drops significantly (first order, k = >1 min1) at pH higher than 6 The d rop in efficiency at higher pH is attributed to the transition of Fe from a hydrated ferrous ion to a colloidal ferric species. Colloidal ferric catalytically decomposes the hydrogen peroxide into oxygen and water, without forming hydroxyl radicals. Fentons reaction is provided in Equation 2 3 Fe2+ + H2O2 3+ + *OH + OH (2 3) EPS may also be degraded through direct oxidation by hydrogen peroxide or by free radicals produced through oxdioreductive enzyme activation. Direct oxidation of polysaccharides by peroxide may occur at conditions feasible in drinking water biofilters (neutral pH, 30C, no catalyst, low hydrogen peroxide concentrations) (Miller 1986). One of the major considerations for using hydrogen peroxide as a catalyst for enzyme stimulation is its inherent toxicity toward microorganisms. Hydrogen peroxide concentrations in the 10 to 1000 mg/L range have been found to inhibit bacterial growth at room temperatures, while higher concentrations destroy the organisms (Schumb et al. 1955). Howev er, a study by Urfer and Huck (1999) showed that hydrogen peroxide residuals of less than 1 mg/L did not lead to an appreciable inhibition of substrate utilization in a biological filter. The oxidoreductase enzymes present in some bacteria serve as intracellular defense systems against free radicals. Peroxides may operate within a biofilm, degrading EPS, while sparing active microorganisms that can express these enzymes (Christensen et al. 1990). Therefore, hydrogen peroxide supplementation may be an effective approach to improve water treatment and hydraulic performance in a biofilter.

PAGE 59

59 Monitoring Techniques Tracking the operation and health of a given biofiltration process enables operators to recognize and anticipate performance problems and subsequent ly implement the necessary corrective actions. There are numerous monitoring tools available, and they can be divided into those that provide feedback in real time and those that require more time to provide feedback. Real Time Monitoring Tools Water q uali ty a nalyses In line turbidimeters can be used to (1) demonstrate compliance with turbidity regulations, (2) help identify appropriate run times, backwash strategies, and ripening periods, and (3) reveal hydraulic anomalies such as short circuiting. In line particle counters can be added to characterize particle breakthrough as a function of size (Goldgrabe et al. 1993). DO can be measured inline, and since bacteria couple the oxidation of BOM with the reduction of DO, it can be used as an indirect indicat or of biological activity (Huck et al. 2000). A change in the DO consumption across the biofilter may indicate that (1) the BOM loading has changed, (2) activity in the microbial community has changed, and/or (3) biofilter hydraulics have been altered. UV254 (UV light absorbance at a wavelength of 254 nm) can also be measured inline or using a benchtop spectrophotometer and can serve as a surrogate for TOC concentration (Howe and Clark 2002). Thus, to some extent, organic carbon removal and breakthrough can be monitored in near real time. A correlation must be developed between UV254 and TOC concentration at a given site for this tool to provide meaningful data.

PAGE 60

60 Headloss t racking As particles accumulate and bacteria grow over the course of a run, headloss builds across a biofilter, which can be measured using pressure transmitters or a piezometric method. Changes in headloss trends within a single run or over long operational periods may indicate (1) a change in particle or BOM loading, (2) deficiencies in the backwash protocol, (3) short circuiting or other hydraulic anomalies, (4) changes in the effective size of the granular media (e.g., caused by abrasion), or (5) changes in microbial activity. Therefore, tracking of headloss trends can reveal significant insights into the overall effectiveness and microbial health of a biofiltration process and can be used to anticipate changes in treatment performance. ATP a nalysis Being able to accurately quantify biological activity is critical for characterizing t he health of any biofilter process. One means of achieving a real time assessment of biological activity is by measuring the quantity of ATP present in the biofilm. As the primary energy carrier for all living organisms, ATP provides an indication of biomass growth rate, substrate removal rate, and physiological state of the bacteria (Nouvion et al. 1987). Active biomass concentration generally matches BOM removal (Liu et al. 2001). Most biological activity measurements (see below) require hours to days f or analysis. On the other hand, methods have been developed for ATP measurements that require only a few minutes (LuminUltra 2008). The method is simple, fast (15 minutes), and reproducible. ATP is measured using the firefly luciferase assay, wherein a sa mple containing ATP is introduced to a mixture containing luciferins and the enzyme luciferase. The luciferase catalyzes a reaction between ATP, DO, and luciferins in the sample to produce light that can be detected in a luminometer as Relative Light Units

PAGE 61

61 (RLU). The RLU produced correlates with the quantity of ATP in the sample based on the following relationship: light in oxylucifer PPi AMP luciferin O ATPluciferase Mg 2 (2 2) Where: ATP = Adenosine Triphosphate AMP = Adenosine Monophosphate PPi = Pyrophosphate Mg2+ = Magnesium ion As shown in Equation 22, chemical energy produced from the breakdown of ATP is converted into light. Each molecule of ATP consumed in the reaction produces one photon of light (Greer and Szalay, 2002) Non Real Time Monitoring Tools Water q uality a nalyses A variety of water quality parameters can be routinely monitored as grab samples across a biofiltration process to assess microbial health and performance, including TOC, DOC, assimilable organic carbon (AOC), BDOC, and heterotrophic plate counts (HPCs). S im ulated distribution system disinfection by product (SDSDBP) tests and DBPFP tests can also be used to reveal organic carbon breakthrough from biofilters. All these analyses are typically performed in a laboratory and require hours to days for processing. Annular r eactor s tudy Distribution system biological regrowth potential, which can be assessed indirectly by DOC, BDOC, and AOC data, can be measured directly using benchtop annular

PAGE 62

62 reactors (Volk and LeChevallier, 1999). Annular reactors consist of a rot or inside a stationary outer cylinder that can simulate detention times, shear stresses, and water velocities typical of drinking water distribution systems. The reactors allow the collection of both water samples and coupons (made of distribution system pipe material) from which biofilm growth may be determined and analyzed. Filter c oring Depthwise cores of a given biofilter can be extracted using any number of filter coring devices. Analyses of filter coring data can provide abundant biofilter system heal th and performance data. Visual observation of a core will detect clumping or mudballing, indicating a deficiency in the backwash process. Media samples can then be taken for microbial community analyses, which can be monitored over time and correlated to system performance. Floc retention profiles can be developed for each cored biofilter, before and after backwashing, to reveal the distribution of particles and biomass across a given bed and assess backwash effectiveness. Sieve analysis of a biofilter cor e will reveal media intermixing, attrition/degradation, and will provide accurate L/d values for comparison against the design criteria. Lastly, SEM microscopic examination of biofilter core sections can reveal key information related to elemental (e.g., F e and Mn) and cell morphologies and distribution. Biological a ctivity a nalyses Quantifying and tracking biological activity can be used not only to assess the health of a biofilter process but also to anticipate and correct performance deficiencies before they become substantial. Non real time biological activity measurement methods include phospholipids analysis to quantify and characterize active cell biomass (Liu et al. 2000), tetrazolium reduction assays to evaluate cellular activity, via colorimetric

PAGE 63

63 measurement of reduced tetrazolium(Fonseca et al. 2001), and nucleic acid based activity methods, including the determination of RNA/DNA ratios (Chcharo and Chcharo, 2008), quantification of precursor rRNA levels (Oerther, 2000), and evaluation of trans criptional activity by quantifying messenger RNA (mRNA) levels for housekeeping genes using reverse transcription quantitative polymerase chain reaction (RT qPCR) (Nielsen and Boye, 2005). All of these methods require specialized instrumentation, are typic ally performed in a laboratory, and require hours to days for processing Microbial c ommunity a nalysis To better understand how microorganisms contribute to contaminant removal, it is important to identify the key microbial populations present in biofilters and link microbial population dynamics to operational performance measures. Microbial community composition can be characterized by constructing clone libraries of the small subunit rRNA genes or selected functional genes in biomass samples, DNA sequencing, and phylogenetic analyses (Briones et al. 2007). Microbial population dynamics and quantification of population abundance in different niches in biofilters can be determined using real time qPCR. Microscopy Biofilter cores can be analyzed using SEM an d transmission electron microscopy (TEM). SEM can be performed under low voltage, low vacuum conditions to minimize sample disruption. SEM provides surface images down to the 0.5m scale and can be used to assess biofilm structure, identify the presence o f protozoa, and assess the elemental composition of filter associated particles. TEM, which provides images down to the 20 nm scale, can evaluate sectional layers of biofilm, identify the presence of

PAGE 64

64 viruses, and accurately quantify biofilm thickness a key input parameter to most biofilm models. To determine the spatial distribution of specific bacterial populations, fluorescence in situ hybridization (FISH) targeting the small subunit rRNA can be used in combination with confocal laser scanning microscopy, to obtain information on possible niche differentiation of microbial populations (Amann and Fuchs 2008). Post t reatment a nalysis In addition to the water quality analyses discussed above, it is important to evaluate the potential of microbial contamination of finished drinking water. Therefore, the levels of bacteria in biofilter effluents, including soon after backwashing, need to be determined. HPCs are commonly used for this, given the USEPA Surface Water Treatment Rule for finished drinking water ( i.e., less than 500 colony forming units ( CFU) /mL) Biofilter effluents have may have HPC counts comparable to those in surface water sources (5106 in a Mississippi River sample (Norton and LeChevallier 2000). Therefore, it is important to study inactivation kinetics of mixed communities of microorganisms present in biofilter effluents to better assess disinfection methods. With a few exceptions (Pernitsky et al. 1995), most research efforts dealing with inactivation kinetics have used pure cultures of m icroorganisms and many of these studies are performed at conditions dissimilar to those commonly found in the drinking water field, such as the study performed by Berry et al. ( 2008) It will also be important to expand inactivation kinetics analyses beyond simple HPC based evaluations (Pernitsky et al. 1995) for evaluation of biological safety.

PAGE 65

65 CHAPTER 3 MATERIALS AND METHOD S Background The purpose of this chapter is to provide a summary of the materials and methods that were followed to perform this work. The chapter includes 1) a broad overview of the pilot system and support equipment that were employed to evaluate the baseline biofilter performance and various enhancement strategies, 2) the followed experimental design, and 3) the analytical m ethods that were used for the hydraulic, water quality, and microbial characterizations. Pilot Biofiltration System Pilot Biofiltration System The pilot skid (Intuitech, Salt Lake City, Utah) included four parallel 6 in diameter biofilters. The biofilters w ere operated as a closed (pressurized) system. Each biofilter contained the same sample media configuration as the full scale filters (40 in of GAC on top of 8 in of sand). Three pilot biofilters contained Norit GAC 820 (effective size 1.1 mm) (Marshall, TX ) and sand (effective size 0.55 mm) media from the full scale biofilters at the JKWTP. These media were collected from a full scale biofilter (Filter No. 2) using shovels, with an attempt to obtain a homogenized sample of each medium. The remaining pilot biofilter media configuration included 40 in of virgin MeadWestvaco Bionuchar GAC (Richmond, VA) and 8 in of sand (effective size 0.55 mm). The media in each pilot biofilter was supported by a Leopold IMS cap and S type underdrain system (Zelienople, PA) Each biofilter had an independent influent pump with automatic flow control. A polyethylene 150gallon effluent break tank served as a backwash water supply. The backwash system also included a dedicated pump and air scour system.

PAGE 66

66 Pilot instrumentation in cluded an inline effluent turbidimeter (1720C, HACH, Loveland, CO ), flow transmitter (3 2551PO 12, GF Signet, El Monte, CA ), and piezometric sensor for each biofilters to measure headloss. Pilot equipment and instrumentation were monitored and controlled by an HMI (Human Machine Interface) that communicated with a small programmable logic controller (PLC) in the control panel. Other features included automatic data logging of key parameters (flow, turbidity, and headloss), remote monitoring, and control us ing a standard web browser, and email and text message alarm notifications. General Process and Control A portion of the full scale biofilter feed water (post intermediate ozone) was pumped to a 525gallon break tank located in the chemical room at the JKW TP. Flow from the break tank was then pumped to the pilot skid by the four independent feed pumps. The four biofilters were operated under full scale average loading conditions (4.5 gpm/ft2), except during selected optimization and robustness tests. Biofil ter effluent was gravity fed to the effluent break tank before being discharged to the sanitary sewer. The pilot was operated in an automated mode. Parameters such as loading rate, backwashing protocol, and backwash triggers were controlled and selected through the PLC. Backwashing was initiated by the operator in the manual mode, or on runtime, headloss, or effluent turbidity triggers (Table 3 1) in the automatic mode. Only one filter could be backwashed at a time. Figure 31 provides a conceptual process flow schematic of the pilot biofiltration system. Supplemental Chemical Dosing Three chemical dosing modules were used during enhancement studies to provide supplemental chemicals to pilot biofilters. Each module consisted of a diaphragm

PAGE 67

67 metering feed pump (Grundfos City, PA) and 10L tank. Chemical dosing modules could be flow paced to individual pilot biofilters. Figure 31 Process flow schematic for pilot biofiltration system Contaminant Spiking A peristaltic pump (Masterflex, Vernon Hills, IL ) was u sed to spike contaminants from a 40 L chemical tank to a feed point upstream of the biofiltration skid. A static mixer was installed downstream of the injection point to promote mixing before flow was diverted to individual pilot biofilters. Biofilter Augm entation and Contaminant Feed Chemicals Caffeine and carbamazepine as neat chemicals and MIB and geosmin stock solutions in methanol were purchased from Sigma Chemical Company (St. Louis MO ). Chem. Dose Pilot Biofilter Influent Break Tank Process water from top of full scale biofilters Pilot Biofilter Effluent Break Tank and Backwash Reservoir Chem. Dose Chem. Dose Chem. Dose Contaminant Spike Backwash Waste to Floor Drain Air Compressor

PAGE 68

68 All other chemicals used for contaminant spiking were reagent grade or better. Phosphorus supplementation was performed using National Sanitation Foundation International (NSF) certified 83 % phosphoric acid. Ammonia supplementation was performed using reagent grade ammonium chloride. The supplemental substrates used in the substrate enhanced biofiltration studies included NSF certified glacial acetic acid, beverage grade 95% ethanol, Food and Drug Administration (FDA) certified food grade molasses, and a high grade glycerin product, MicroCglycerin from Environmental Operating Solutions (Bourne, MA ). Food grade, 3% hydrogen peroxide was used for the oxidant enhancement studies. A Water & Power Technologies reverse osmosis system (Arlington, TX ) was used to deionize all water used for stock solutions of contaminants and supplements. Stock solutions were made to provide approximately 1 week of chemical feed before tank switch out (under average flow and dosage conditions). The target stock concentrations for hydrogen peroxide and all C substrates was 0. 13% (w/v). The stoc k concentrations for PO4P and NH4N were 0. 0025% (w/v) and 0.014% (w/v), respectively. Backwash Protocol All biofilters were operated at an 18hour filter run interval, and a uniform backwash strategy was used. Maintaining a consistent run time and backwash strategy allowed meaningful hydraulic performance comparisons between operational conditions. The pilot biofilter run time and backwash strategies were modified from current full scale operation to obtain consistent hydraulic performance. The full scale biofilter backwashing protocols led to excessive media attrition when implemented on the pilot biofilters. The effect was believed to be inherent to hydraulic and mechanical limitations of the scaled down pilot design. A backwash strategy was found to provide

PAGE 69

69 acceptable cleanbed headloss, consistent headloss profiles, and limited media attrition for both control and enhanced test conditions. Attrition was limited through by extending the air scour step to provide additional particle/media col lisions for mudball destruction. Table 31 provides a summary of backwash protocols for the pilot and full scale biofilters. Table 3 1 Backwash p rotocols Pilot s cale b iofilters Full s cale b iofilters Backwash t riggers Filter run time (hr) 18 24 Headloss (ft) 13.5 18 Turbidity (NTU) 1.00 1.00 Backwash p arameters Air Scour Rate 3 scfm 3 scfm Duration (min) 10 5 Air scour/low rate combined backwash Duration (min) 3 ~1 High rate backwash Rate (gpm/ ft 2 ) 30.5 18 Duration (min) 8 20 Low rate backwash Rate (gpm/ ft 2 ) 10 6 Duration (min) 5 5 Pilot backwash protocol was developed iteratively over first two weeks of pilot testing to minimize cleanbed headloss, media attrition, backwash duration, and wastewater production. Plant staff manually initiates backwash every 18 to 24 hours if other conditions are not met first. Pilot headloss is limited by influent feed pump capacity. Experimental Design Research Testing Plan The research testing plan included multiple studies to characterize the hydraulic and water treatment performance at the JKWTP (full scale and pilot scale biofilters under control conditions) and to identify and evaluate potential improvements through modifications of the biofiltration process at th e pilot scale. The research studies included:

PAGE 70

70 Baseline Characterization Substrate Enhancement Evaluation Nutrient Enhancement Evaluation Peroxide Enhancement Evaluation Microbial Tracking FullScale Process Integration Assessment and Economic Evaluation. Operational Setup Multiple research studies were conducted simultaneously as the four pilot biofilters were operated in parallel. A summary of the operational objectives targeted by each pilot biofilter is provided in Table 3 1 Sampling Plan Water qualit y and biofilter media samples were collected from pilot and full scale systems during each study The sampling plan prov id ed in Table 32 provides an approximate summary of the samples taken at each process point for various analyses. Baseline Biofiltration Characterization The objective of this study was to characterize baseline water treatment performance, confirming similitude between pilot and full scale treatments, and to provide control data for parallel tests occurring with other pilot f ilters. Pilot system similitude to fullscale performance was characterized by operating the pilot under full scale conditions and monitoring hydraulic performance and effluent water quality. The baseline performance characterization study was conducted by operating the pilot control biofilter under full scale operational conditions at steady state for five months. The control was operated continuously with no carbon, nutrient, or hydrogen peroxide supplementation.

PAGE 71

71 Biofiltration Substrate Enhancement Eval uation The objective of this study was to evaluate various primary substrate augmentation strategies for enhancing biofiltration performance. This study evaluated four substrates for improved biodegradation of organic carbon (trace organic compounds and background DOC). The substrates tested included acetic acid, molasses, MicroC (a proprietary glycerinbased product), and ethanol. The substrates were tested individually through four successive 1to 2 month phases, each dosed at 1 mg/L as carbon (C). Biof iltration substrateenhancement was evaluated by monitoring changes in biofilm appearance, hydraulic and water treatment performance, and microbial activity.

PAGE 72

72 Table 3 2 Pilot biofilter operational parameters Parameter Biofilter 1 Biofilter 2 Biofi lter 3 Biofilter 4 Pilot biofilter operational objective (duration) Study 2: substrate enhancement (6 months) Study 3: nutrient enhancement (1 month) Study 4: Peroxide enhancement (2 weeks) Studies 2 and 3: substrate and nutrient enhancement (9 months) Study 1: Full scale simulation for baseline characterization with different GAC support media (9 months) Study 1: Full scale simulation for baseline characterization (5 months steady state, 2 months robustness testing) Validation of Study 3: nut rient enhancement (2 weeks) Media configuration 8 inches sand 40inches bituminous based GAC obtained from JKWTP biofilter 8 inches sand 40inches bituminous based GAC obtained from JKWTP biofilter 8 inches sand 40inches virgin woodbased based GAC 8 inches sand 40inches bituminous based GAC obtained from JKWTP biofilter

PAGE 73

73 Table 3 3 Approximate sampling schedule for routine analyses Parameter Sampling location Sampling frequency *, Turbidity Full scale biofilter influent 7/week Full scale biofilter effluent 7/week Pilot biofilter influent 1/ week Pilot BF effluent Continuously (every 5 min/biofilter) pH Full scale biofilter influent 7/week Full scale biofilter effluent 7/week Pilot biofilter influent 1/week/biofilter Pilot biofilter effluent 1/week/biofilter Temperature Full scale raw water 7/week Dissolved oxygen Full scale biofilter influent 1/month Full scale biofilter effluent 1/month Pilot biofilter influent 1/week/biofilter Pilot biofilter effluent 1/week/biofilter Ozone residual Full scale ozone contactor cell 8 7/week Full scale biofilter influent 7/week Total coliforms, fecal coliforms, heterotrophic plate count Full scale biofilter effluent 3/study Pilot biofilter effluent 3/study DOC Raw water 1/week Full scale filter influent 1/week Pilot biofilter influent 2/week Pilot biofilter effluent 2/week/biofilter Orthophosphate and ammonianitrogen Full scale biofilter influent 1/week Full Scale biofilter effluent 1/week Pilot biofilter influent 2/week Pilot biofilter effluent 2/week/biofilter Iron and manganese Raw water 1/week Ozone influent 1/week Pilot biofilter influent 2/week Pilot biofilter effluent 2/week/biofilter MIB and geosmin Raw water 1/week Ozone influent 1/week Pilot biofilter influent 2/week Pilot biofilter effluent 2/week/biofilter Pesticide and pharmaceutical suite Raw water 4/study Ozone influent 4/study Full scale biofilter influent 4/study Full scale biofilter effluent 4/study Pilot biofilter influent 1/month Pilot biofilter effluent 1/month/biofilter Adenosine triphosphate Full scale biofilter GAC 4/study Pilot biofilter GAC 1/month/biofilter SEM and other microbial tracking assays Full scale biofilter GAC 5/study Pilot biofilter GAC 1/month/biofilter The sample frequency represents an average minimum collection frequency. Samples were taken with much higher frequencies for many analytes during high sensitivity testing. Fullscale biofilter influent was equival ent to pilot biofilter influent when operated under control conditions. In these instances, only one sample was collected and analyzed for the full scale sample location. Turbidity, pH, temperature, and ozone residual measurements were performed at least daily by JKWTP plant operators for all full scale treatment sample locations.

PAGE 74

74 Biofiltration Nutrient Enhancement Evaluation The objective of this study was to evaluate various nutrient augmentation strategies for enhancing biofiltration performance. Phosphoric acid and/or ammonia were dosed to pilot biological filters to achieve a target molar ratio of bioavailble C:N:P of 100:10:1, where bioavailble C was determined iteratively by the amount of DOC removed in the biofilter The nutrient enhancement strategy was evaluated by monitoring changes in biofilm appearance, hydraulic and water treatment performance, and microbial activity. Oxidant Enhancement Evaluation The obj ective of this study was to evaluate peroxide supplementation for augmenting the oxidative action and response of the biofiltration process. Many microorganisms express a class of enzymes, known as oxidoreductases, when exposed to hydrogen peroxide. Prelim inary testing of this strategy was conducted by dosing hydrogen peroxide to a pilot biofilter (with no other nutrient of carbon supplementation) at 1 mg/L to evaluate the peroxide enhancement strategy. This condition was operated continuously for two weeks The hydrogen peroxide enhancement strategy was evaluated by monitoring changes in biofilm appearance, hydraulic and water treatment performance, and microbial activity. Microbial Tracking The objective of this study was to correlate the microbial ecology and activity in the biofilters to hydraulic and treatment performance. Media samples were collected from the pilot and full scale biofilters after each process change and analyzed for microbial activity, speciation, and morphology. The microbial tracking study included the following

PAGE 75

75 analyses: SEM, ATP quantification, biofilm morphological characterization, T RFLP, and clone libraries. Full Scale Process Integration Assessment and Economic Evaluation A fullscale process integration assessment was performed for the nutrient enhancement strategy, as it was the most effective and best characterized enhancement strategy Capital and operation cost estimates were also developed for integrated enhancement strategies, including the estimated cost savings that may be realized during operation. Data Analyses General All collected data sets were described by determining mean, standard deviation, maximum, and minimum values. The error bars presented in all figures in this dissertation represent the standard deviation o f the data set. This standard deviation accounts for operational variability (i.e., feed water conditions) and sampling/analytical error for that data s et All nondetects were accounted for as 50% of the detection limit in all statistical analyses. Paired T Test The paired t test was the statistical method used to determine if there were statistical differences between sets of collected data from the full scale biofilter, pilot biofilter control, and various pilot enhanced biofilters. The paired t test is a variation of the standard t test and is used to compare two treatment methods where experiments are performed in pairs and the differences are of interest. Since sample collection was performed in pairs in the pilot biofilter studies and the differences in the collected data sets are of interest, the paired t test was appropriate to use. For the purposes of this

PAGE 76

76 study, two means with a paired t test p significant difference. All paired t tests were calculated with two tails (two sided p value). Analytical Methods General Water Quality Parameters Water quality data were collected and analyzed using inline pilot instrumentation, field equipment, and laboratory equipment. Turbidity, pH, DO, temperature, free and total chlorine, ozone residual, and hydrogen peroxide analyses were performed onsite at the JKWTP. The City performed all laboratory analyses except for N nitrosodimethylamine and haloacetic acids, which were performed by Montgomery Watson Harza (MWH) Laboratories (Monrovia, CA ). Sample c ollection for l aboratory a nalyses Aqueous samples were collected from the pilot and full scale system in bottles provided by the laboratory. Full scale biofilter samples were collec ted from JKWTP Filter No. 2. Samples were either delivered same day to the City laboratory, shipped overnight in a cooler packed with ice to MWH, or delivered twice weekly to the Dallas Water Utility Central Wastewater Plant Water Quality Lab (MIB and g eos min). Replicate measurements were performed for each sample set, and replicate samples were collected monthly. Turbidity In line nephelometers were used to perform continuous turbidity measurement for the pilot (Hach 1720E, Loveland, CO ) and full scale biological filters (Hach 1720D, Loveland, Colorado ). A desktop nephelometer was used to measure raw and settled water turbidities (Hach 2100N, Loveland, CO ). Formazin standards were used for instrument calibration in accordance to manufacturers protocols.

PAGE 77

77 pH pH measurements were performed using an Orion pH Electrode (Thermo Fischer Scientific Inc., Waltham, MA ) per the manufacturers protocol. A 3point calibration of the pH electrode was performed daily. Slope limits were between 92 to 102% Dissolved o xygen and t emperature DO and temperature were measured on site using a YSI 55 dissolved oxygen DO probe (Yellow Springs, OH ) Measurements and calibration were conducted per the manufacturers protocol. Free and t otal c hlorine Free and total chlorine were measured on site using a Hach DR890 colorimeter (Loveland, CO ). Free chlorine was analyzed using the EPA N,N Diethylp Phenylenediamine (DPD) Method 8021. Total chlorine was analyzed using the EPA DPD method 8167. Hydrogen p eroxide Hydrogen peroxide was measured on site using a CHEMets Colorimetric Hydrogen Peroxide Test Kit (Chemtech International, Media, PA) according to the manufacturers instructions The colorimetric test is based upon the ferric thiocyanate method. Ozone r esidual Ozone residual was measure d in pilot influent. All measurements were collected using a Hach DR890 colorimeter (Loveland, CO ), following EPA Indigo Method 8311. Total and d issolved o rganic c arbon The City laboratory performed TOC and DOC measurements in accordance with the Standard Methods for Examination of Water and Wastewater 21st Edition (2005),

PAGE 78

78 Standard Method (SM) 5130B (2005) A Shimadzu TOC analyzer (Kyoto, Japan) was used for the analyses. UV254 The City laboratory performed UV254 measurements in accordance with SM 5910B (2 005) UV254 measurements were performed using a Bausch & Lomb spectrophotometer (Rochester, NY ). Iron The City laboratory performed total Fe measurements in accordance with SM 3111B (2005). A Varian atomic adsorption spectrophotometer (AAS) (Palo Alto, CA ) was used to perform these analyses. All measurements were of total Fe. Manganese The City laboratory performed total Mn measurements in accordance with SM 3111B (2005). A Varian AAS (Palo Alto, CA ) was used to perform these analyses. All measurements were of total Mn. Nitrates, n itrites The City laboratory performed nitrate and nitrate measurements in accordance with EPA Method 300.0. A Dionex ion chromatograph ( Sunnyvale, CA ) was used to analyze nitrates and nitrites. Ammonia The City laboratory performed ammonia measurements with a Thermo Electron Corporation ammonium ionselective electrode (ISE) All measurements were performed following the manufacturers protocol.

PAGE 79

79 Orthophosphate The City laboratory performed orthophosphate measurements in accordance with EPA Method 300.0 A Dionex ion chromatograph ( Sunnyvale, CA ) was used tor all orthophosphate measurements. Regulated m icrobial p arameters The City laboratory performed all analyses for aqueous microbial parameters. Mea surements for heterotrophic plate counts, total coliforms, and fecal coliforms were performed in accordance with SM 9215 (1998), 9222 (1998), and 9221 (1998), respectively. Disinfection b y p roducts The City laboratory performed total trihalomethane measurements in accordance with EPA Meth od 501.1, using a gas chromatography (GC). MWH laboratories performed haloacetic acid measurements following SM 6251B (1998) Tastes and o dors MIB and geosmin analyses were performed by the Dallas Water Utilities (DWU) analytical laboratory. The protocol f or analyses followed SM 6040 D, Odor Causing Compounds MIB, g eosmin by gas chromatography/mass spectrometry (GC/MS) (2005). The limit of detection (LOD) for both MIB and g eosmin was 0.4 ng/L. Pesticides and P harmaceuticals All pesticide and pharmaceutical analyses were performed at the University of ColoradoBoulder. The compounds included in the initial screening and continued monitoring are provided in Table 34 with their respective LODs. The methods described herein were derived from Thurman et al. (2006).

PAGE 80

80 Sample c ollection Aqueous samples were collected in 1L amber glass bottles. Full scale biofilter effluent samples were collected from JKWTP Filter No. 2. Samples were stored o n ice and shipped overnight to the University of ColoradoBoulder for pestic ide and pharmaceutical analyses. Table 3 4 Limit of detection for screened pharmaceuticals and pesticides assuming 100% recovery by solid phase extraction. Individual water sample matrices may vary Contaminant LOD (ng/L) Pharmaceuticals 1,7 Dimethylxanthine 100 Acetaminophen 50 Albuterol (Salbutamol) 10 Ampicillin 10 Azithromycin 10 Caffeine 10 Carbamazepine 5 Cefotaxime 10 Cimetidine 10 Ciprofloxacin 5 Clarithromycin 10 Cloxacillin 50 Cotinine 10 Dehydronifedipine 2 Diclofenac 20 Digoxigenin 20 Digoxin 10 Diltiazem 5 Diphenhydramine 2 Enrofloxacin 5 Erythromycin 5 Flumequine 10 Fluoxetine 10 Gemfibrozil 10 Ibuprofen 50 Lincomycin 5 Lomefloxacin 5

PAGE 81

81 Table 3 4. Continued Contaminant LOD (ng/L) Metformin 50 Miconazole 5 Naproxen 50 Norfloxacin 5 Ofloxacin 5 Oxacillin 20 Oxolinic Acid 20 Ranitidine 10 Roxithromycin 5 Sarafloxacin 5 Sulfachloropyridazine 20 Sulfadiazine 50 Sulfadimethoxine 5 Sulfamerazine 20 Sulfamethazine 5 Sulfamethizole 20 Sulfamethoxazole 10 Sulfanilamide 50 Thiabendazole 5 Triclocarban 20 Trimethoprim 5 Tylosin 10 Viginiamycin 20 Warfarin 10 Sulfadimethoxine 5 Hormones Estrone 1 Estradiol 1 Ethinylestradiol 1 Cholesterol 1 Coprastanol 1 Pesticides Acetamiprid 3 Acetochlor 2 Alachlor 3 Aldicarb 5 Aldicarb sulfone 3 Aldicarb sulfoxide 4 Azoxystrobin 0.1

PAGE 82

82 Table 3 4. Continued Contaminant LOD (ng/L) Atrazine 1 Benalaxyl 0.04 Bendiocarb 5 Bensultap 281 Bromoxynil 20 Bromuconazole 0.3 Buprofezin 0.6 Butylate 3 Captan 15 Carbaryl 3 Carbendazim 0.8 Carbofuran 4 Cartap 15 Chlorfenvinphos 0.2 Chlorpyrifos methyl 30 Cyanazine 2 Cyproconazole 1 Cyromazine 9 Deethylatrazine 2 Deethylterbuthylazine 1.5 Deisopropylatrazine 2 Diazinon 0.05 Dichlorvos 0.5 Difeconazole 0.5 Difenoxuron 0.4 Diflubenzuron 12 Dimethenamide 1 Table 34. Continued Dimethoate 1.5 Dimethomorph 4 Diuron 0.6 Ethiofencarb 4 Fenamiphos 0.1 Fenuron 10 Flufenacet 3 Flufenoxuron 6 Fluoroacetamide 80 Fluoroxypyr 45 Hexaflumuron 8

PAGE 83

83 Table 3 4. Continued Contaminant LOD (ng/L) Hydroxyatrazine 0.4 Imazalil 0.3 Imazapyr 5 Imazaquin 0.7 Imidacloprid 2 Ioxynil 15 Iprodione 4 Irgarol 1051 0.1 Irgarol metabolite 0.5 Isoproturon 0.7 Lenacil 9 Lufenuron 9 Malathion 1.5 Mebendazole 0.8 Metalaxyl 0.2 Metamitron 3 Methidathion 15 Methiocarb 0.7 Methiocarb sulfone 9 Methomyl 2 Metolachlor 0.8 Metolcarb 12 Metribuzin 0.6 Molinate 1.5 Monuron 0.7 Nicosulfuron 0.8 Nitenpyram 0.2 Oxadixyl 14 Parathion ethyl 17 Pendimethalin 11 Phosmet 0.9 Prochloraz 0.7 Profenofos 1 Promecarb 3 Prometon 1 Prometryn 0.3 Propachlor 0.5 Propanil 0.7 Propiconazole 0.3

PAGE 84

84 Table 3 4. Continued Contaminant LOD (ng/L) Prosulfocarb 2 Simazine 0.4 Spinosad A 0.9 Spinosad D 6 Spiromesifen 120 Spiroxamine 8 Chemicals and r eagents Analytical standards were purchased from SigmaAldrich ( St. Louis, MO ) and Supelco ( Billerica, PA) Individual stock solutions (1000 g/mL) were prepared in pure methanol and stored at 18C. HPLC grade acetonitrile and methanol were obtained from Burdick & Jakson ( Muskegon, MI ). Formic acid was obtained from SigmaAldrich. A MilliQ Plus ultra pure water system from Millipore ( Billerica, MA ) was used throughout the study to obtai n the HPLC grade water used during the analyses. Sample p reparation ( s olid p hase e xtraction) An off line solid phase extraction ( SPE) was used for the preconcentration of the water samples. All the extraction experiments were performed using an automated sample preparation with extraction column system (GX 271 ASPEC, Gilson, Inc., Middleton, WI) fitted with a 25mL syringe pump for dis pensing the water samples through the SPE cartridges. Disposable cartridge columns packed with 500 mg of Oasis HLB sorbent were used. The cartridges were conditioned with 4 mL of methanol followed by 6 mL of HPLC water at a flow rate of 1 mL/min. The water samples (200 mL) were loaded at a flow rate of 10 mL/min. Elution of the analytes from the cartridge was carried out with 5 mL of methanol. The solvent was evaporated to 0.5mL with a stream of nitrogen at a te mperature of 45C in a water bath using a Tur bovap

PAGE 85

85 concentration workstation (Caliper Life Sciences, Palo Alto, CA) The extracts were analyzed by liquid chromatography/mass spectrometry (LC/MS). LC/TOFMS a nalyses of p esticides and p harmaceuticals The separation of pesticides and pharmaceuticals was carried out using a high performance liquid chromatography (HPLC) system (consisting of vacuum degasser, autosampler and a binary pump) (Agilent Series 1200, Agilent Technologies, Santa Clara, CA) equipped with a reversed phase C8 analytical column of 150 mm x 4.6 mm and 5 m particle size (Zorbax Eclipse XDB C8). Column t emperature was maintained at 25 C. The injected sample volume was 50 L. Mobile phases A and B were acetonitrile and water with 0.1% formic acid, respectively. The optimized chromatographic method held the initial mobile phase composition (10% A) constant for 5 min, followed by a linear gradient to 100 % A after 30 min. The flow rate used was 0.6 mL/min. A 10 min post run time was used after each analysis. This HPLC system was connected to a time of flight mass spectrometer Agilent 6220 MSD TOF equipped with a dual electrospray interface operating in positive ion mode, using the following operation parameters: capillary voltage: 4000 V; nebulizer pressure: 45 psig; drying gas : 9 L/min; gas t emperature: 300C; fragmentor voltage: 190 V; skimmer voltage: 60 V; octopole RF: 250 V. LC/MS accurate mass spectra were recorded across the range 501000 m/z at 4GHz. The data recorded were processed with MassHunter software. Accurate mass measurements o f each peak from the total ion chromatograms were obtained by means of an automated calibrant delivery system using a dual nebulizer electrospray ionization ESI source that introduces the flow from the outlet of the chromatograph together with a low flow of a calibrating solution (calibrant solution A, Agilent Technologies Santa Clara, CA ), which contains the internal reference masses

PAGE 86

86 (purine (C5H4N4 at m/z 121.0509 and HP 921 [hexakis (1H,1H,3H tetrafluoro pentoxy)phosphazene] (C18H18O6N3P3F24) at m/z 922.0098. The instrument provided a typical mass resolving power of 15000500 (m/z 922). Quanti fic ation was carried out using external standard curves. LC/MS MS a nalyses of p harmaceuticals at l ow l evel (ppt concentration) A special method for the detection of two pharmaceuticals (meprobamate and carbamazepine) at low levels of concentration was developed. For this purpose, an LC/MS/MS triple quadrupole Agilent Model 6460 with Jet Stream technology was used. Three different transitions (one for quantitation and two for confirmation) for each compound were used for quantification and qualification of each compound. Fifteen microliters of sample extract were analyzed for each sample. Chromatography consisted of UHPLC, ultra hig h pressure liquid chromatography, using the Agilent Infinity system 1290 Model (Agilent Technologies, Inc., Santa Clara, CA ). The mobile phases were 0.1% formic acid and acetonitrile. The column was the Zorbax Eclipse C18 2.1 mm x 50 mm with 1.8 micron pac king. The limits of detection with this method were 0.1 ng/L for both compounds. The relative standard deviation for this method is 6% Microbial Tracking ATP analyses were performed on site. The SEM work was performed at the United States Environmental P rotection Agency Office of Research and Development (USEPA ORD, Cincinnati, OH ). All other microbial tracking analyses were performed at the University of Texas Austin including filter media HPC, EPS quantification, biofilm formation potential, terminal r estriction length fragment polymorphism (T RFLP), and clone libraries

PAGE 87

87 Biofilter Media Microbial Characterization and Analyses Sample c ollection for ATP a nalyses GAC media samples were collected from pilot scale biofilters at the beginning and end of a fi lter run for ATP analyses. Prior to sample collection the pilot filters were taken offline and partially drained. The top 6 inches of media were then homogenized via stirring. Approximately 50 grams of GAC were then removed from the top of a given biofilte r. GAC obtained from the full scale biofilters was used to replace the removed sample. ATP a nalyses The ATP analyses were conducted using the LuminUltra (NB, CA) Deposit and Surface Analysis (DSA) test kit. The DSA test kit is based on the premise that ATP is a direct and interference free indicator of total active biomass (Luminultra, 2008) Collected biofilter media was decanted and measured into 1.1 mL test samples. The samples were then weighed for reference and analyzed following the DSA manufacturers protocols. RLU were measured using a Kikkoman C 100 LumiTester (Tokyo, Japan). Triplicate samples were run for each sample set. Sample c ollection for other m icrobial t racking a nalyses GAC media samples were collected from pilot scale and full scale bio filters at the beginning of a filter run for analyses. Full scale biofilter GAC media samples were collected from JKWTP Filter No. 2. Prior to sample collection, the pilot biofilters were taken offline and partially drained. The top 6 inches of media were then homogenized via stirring. Approximately 100 grams of GAC were then removed from the top of the biofilter. Samples of GAC were then placed in 60mL amber glass vials, which w ere then filled with pilot filter feed water. The vials were stored o n ice and shipped overnight

PAGE 88

88 to the USEPA ORD for SEM analyses and to the University of Texas Austin for all other microbial tracking analyses. Scanning e lectron m icroscopy Upon receipt, samples were logged and stored at 4C. Prior to analyses, samples were portioned into processing containers and rinsed in deionized water (DI) for 30 seconds to remove extra debris. Samples were then fixed with 2.5 % gluteraldehyde, 4% paraformaldehyde solution in a 0.1M cacodylate buffer that was adjusted to pH 7.3. Samples were allowed to fix for 6 hours and then washed twice for 15 minutes with the cacodylate buffer. Samples were then washed again in DI water for an additional 15 minutes. Next, the samples were postfixed for an additional hour in a 1% osmium tetroxide solution. The postfixed samples were then washed three additional times for 15 minutes each. After osmium fixation the samples were dried using a dilution series of ethanol (25, 50, 75, 95, 100, 100) for 30 minutes each. Samples were then chemically dried using two exchanges of propylene oxide (30 minutes each). Excess propylene oxide was removed and granules were air dried in a chemical hood for an hour on blotter paper then transferred to a desiccation jar. Prior to imaging, several granules were mounted on aluminum SEM stubs and sputter coated with gold palladium for 90 seconds. Biofilms were imaged at 15kV under high vacuum using the JEOL JEM6490LV scanning electron microscope (EPA, Cincinnati, OH). Multiple sets were examined to ensure continuity of imaging in the samples. In all, SEM was performed on 32 samples, with typically 25 to 30 media granules per sample. Biofilter m edia h eterotrophic p late count Upon receipt, samples were logged and stored at 4C. Two grams of sample were susp ended in 10 mL of phosphatebuffered saline (PBS), submerged in a sonicator bath

PAGE 89

89 for one minute, put on ice for one minute, and vortexed vigorously for five seconds. This procedure was repeated five times to dislodge the biofilm from the activated carbon. Ten fold serial dilutions were prepared through the 106 dilution. The dilutions were plated in triplicate on R2A agar and incubated at 30C for 24 hours. Crystal v iolet (CV) a ssay ( b iofilm for mation potential) The biofilm formation capacity of filter medi a biofilms was evaluated using the crystal violet (CV) assay as described by OToole and Kolter (1998). T he CV assay results are heavily dependent on the initial cell concentration of the inoculums. Therefore, the biofilm formation capacity of the filter samples was compared among inocula with similar cell concentrations (as determined by CFU counts). One L from each dilution prepared for CFU counts was taken to seed 100 L of R2A broth on a 96well microtitre plate. This was done in triplicate for each dilution. These cultures (three per dilution of biologically active carbon sample) were allowed to grow statically at 30C for 24 hours. After incubation, the medium was poured off and the remaining biofilm was stained with CV The plates were rinsed with water and dried, and then 200 l of 96 % ethanol was used to solubilize the crystal violet in each well. Absorbance was measured at 600nm using a spectrophotometer. H igher absorbance measurements indicate d greater biofilm formation Phenol s ulfuric a cid a ssay (EPS q uantification) This assay was performed following the method of Dubois (1956). Two grams of sample were suspended in 10 mL of PBS, submerged in a sonicator bath for one minute, put on ice for one minute, and vortexed vigorously for five seconds. This procedure was repeated five times to dislodge the biofilm from the activated carbon. Eight milliliters of the suspension were transferred to a clean tube and centrifuged at

PAGE 90

90 10,000 RPM at 4C. The resulting supernatant was then transferred to another test tube, and the pellet was resuspended in a buffer (10 mM Tris/HCl, pH 8, 10 mM EDTA, 2.5 % NaCl) and incubated for 8 hours at room temperature. To measure free EPS, 2 mL of the supernatant was transferred to a tube where it was mixed with 50% phenol sol ution and 5 mL of concentrated sulfuric acid. A yellow color was developed, and absorbance was read at 480 nm. To measure bound EPS, the resuspended pellet was centrifuged and processed as described above for the supernatant. A glucose calibration curve was constructed for data analyses. Terminal r estriction f ragment l ength p olymorphism DNA was extracted in triplicate from each sample using the MoBio UltraClean Microbial DNA Isolation kit (MoBio Laboratories, Inc., Carlsbad, CA ) according to the manufactur ers instructions. Approximately two hundred milligrams of biologically active carbon was used per extraction. DNA from the triplicate extractions was combined and used for p olymerase chain r eaction (PCR). DNA was amplified in duplicate reactions using 8F (FAMlabeled at the 5 end) and 1492R targeting the 16S rRNA gene. Each 50L PCR reaction contained 1.25 U Taq DNA polymerase, 10 mM Tris HCl, 50 mM KCl, 1.5 mM MgCl2, 0.2 mM of each deoxynucleoside triphosphate, 0.4 M of each primer, and 100 ng of tem plate DNA. The reactions were run under the following amplication conditions: denaturation at 94oC for 3 min, followed by 20 cycles of denaturation at 94oC for 30 s, annealing at 52oC f or 30 s, and extension at 72oC for 1 min, and a final extension at 72 C for 7 min. Amplicon from the duplicate reactions was combined, and 1 L of amplicon was added to duplicate nested PCR reactions. These nested PCR reactions, with primers 8F (FAM labeled at the 5 end) and 926R, were performed to increase the concentratio n and

PAGE 91

91 specificity of the amplicon; with the exception that 0.5 M of each primer was added to the reactions the reaction conditions and thermal cycling parameters were the same as aforementioned. Amplicon from the duplicate nested PCR reactions was combined and visualized on a 1% agarose gel stained with ethidium bromide. The amplicon (1 g for each sample) was treated with the Klenow enzyme as described in Egert and Friedrich (2005) and purified with the MoBio Ultraclean PCR cleanup kit (MoBio Laboratori es, Inc., Carlsbad, CA ). One hundred nanograms of purified amplicon was digested with 40 U of HhaI in a 20L reaction at 37 C for 3 h and purified with a centrifugal filter (YM 30, Millipore Corp., Billerica, MA .). The digested sample was sent to the University of Texas at Austin Institute for Cellular and Molecular Biology core facility for fragment analysis on an ABI 3130 DNA analyzer. The electropherograms were processed using GeneMarker 1.70 (SoftGenetics, LLC, State College, PA); bands greater than 60 bases in length and greater than the intensity threshold of 40 were included in the analysis. The diversity of each sample was determined by the ShannonWeaver index (SWI) (Wani et al. 2006). Community similarity between samples was assessed with th e S renson index (SI) using 0.5bp comparisons (Wawrik et al. 2005). Clone l ibraries DNA was extracted, and the 16S rRNA gene was amplified as described for T RFLP, with the exception that nonlabeled 8F was used for PCR. The cloning reactions were carrie d out using a T OPO TA Cloning Kit with OneShot Top 10 chemically competent cells (Invitrogen, Carlsbad, CA ) following the manufacturers instructions. The transformed E.coli were transferred to LuriaBertani (LB) plates containing a 50 g/mL kanamycin an d then incubated at 37C overnight. Colonies were selected

PAGE 92

92 randomly and used to inoculate a 96well microplate for each sample. The 96m icroplates were incubated at 37 C for 3 days with shaking at 200 rpm. Plasmids were purified with the QIAprep 96 Turbo M iniprep Kit (Qiagen Inc., Valencia, C A ). The purified plasmids were sent to the University of Texas at Austin Institute for Cellular and Molecular Biology core facility and sequenced with the T7 primer. For sequence analyses, the PCR primer site was identi fied using Geneious Pro 4.8.5 (Biomatters Ltd., Auckland, New Zealand), and the vector sequence was removed. The sequences were submitted to BLAST (blastn and megablast queries) at http://blast.ncbi.nlm .nih.gov/Blast.cgi to identify library sequences most closely resembling the query sequence. Summary of Responsible Parties f or Analytical Work Performed Table 3 5. Responsible Parties for Analytical Work Performed Ana l ytical Parameter Responsible Party for Analysis Performance Temperature On site, Chance Lauderdale Dissolved Oxygen On site, Chance Lauderdale pH On site, Chance Lauderdale UV254 On site, Chance Lauderdale Dissolved Organic Carbon Arlington Water Utilities Analytical Laboratory Ortho phosphate Arlington Water Utilities Analytical Laboratory Ammonia Nitrogen Arlington Water Utilities Analytical Laboratory Manganese Arlington Water Utilities Analytical Laboratory Iron Arlington Water Utilities Analytical Laboratory TTHM & HAA Arlington Water Utilities Analytical Laboratory MIB Dallas Water Utilities Analytical Laboratory Geosmin Dallas Water Utilities Analytical Laboratory Color Arlington Water Utilities Analytical Laboratory Total organic carbon Arlington Water Utilities Analytical Laboratory Selected EDCs Univerisity of Colorado Boulder Biofilter Effluent HPC (R2A) Arlington Water Utilities Analytical Laboratory Total and Fecal Coliforms Arlington Water Utilities Analytical Laboratory Scanning Electron Microscopy United States Environmental Protection Agency Office of Research and Development Biofilter Media HPC University of Texas Austin EPS University of Texas Austin Biofilm Formation Potential University of Texas Austin ATP On site, Chance Lauderdale Gen etic analyses University of Texas Austin

PAGE 93

93 CHAPTER 4 BASELINE CHARACTERIZ ATION AND CONTROL ST UDIES Objectives The objective of this study was to characterize the hydraulic and water treatment performance of the pilot biofilter operated under baseline conditions. This allowed for the confirmation of similitude between pilot and full scale biofiltration performance and provided control data for parallel tests occurring with other pilot biofilters operating under various enhancement strategies. Similitude was characterized by operating a control biofilter under full scale operational conditions (i.e., no enhancement strategies were used for the control biofilter) for approximately seven months. The seven months of operation included approximately one month of acclimation, five months of steady state operation, and one month of robustness testing. Contaminant loading (Mn, MIB, atrazine) was also performed to assess biofilter baseline performance, as full scale influent concentrations sufficient for sustained removal characterization were not observed during the study. Monitored water quality parameters included turbidity, TOC, DOC, Fe, Mn, NH4N, PO4P, MIB, geosmin, and a suite of over 150 pharmaceuticals and pesticides. Biofilter Configuration All pilot biofilters were configured and operated under conditions equivalent to the fullscale biofiltration facility; including filter feed loading rat e (4.5 gpm/ft2), media configuration (8 inches of sand, 40 inches of bituminous GAC obtained from full scale filters, 6 years in previous operation). The pilot biofilters were operated without a filter to waste step, as the full scale biofilters do not use filter to waste. Contaminant spiking

PAGE 94

94 was performed on all pilot biofilters after the first month of operation and continued intermittently under a managed experimental plan through the remainder of the study. Biofilter Backwash Strategy Development The in itial pilot biofilter backwash protocol was developed to mirror the full scale biofilter backwash protocol However, mudballs, media attrition, and increasing cleanbed headloss were observed during the first two weeks of operation, and a more aggressive pi lot backwash protocol was implemented. The backwash duration and intensity were increased over current full scale operation, and pilot biofilter run times were limited to 18 hours for the remainder of the study. These modifications provided consistent clea n bed headloss and biofilter hydraulic profiles. In addition, 18hour pilot biofilter run times provided straightforward hydraulic performance comparisons between test conditions, through the observation of terminal headloss. Table 31 provides a summary o f the full scale and modified pilot backwash strategies. Wood Based Gac Media Evaluation A pilot biofilter ( Biofilter 3 ) was operated under baseline conditions for the duration of the study to evaluate a woodbased GAC as biofilter support media. Unlike the bituminous GAC used in the other three biofilters, the woodbased GAC in Biofilter 3 was virgin media at the start of the study. Complete breakthrough of DOC or the tracked contaminants was not observed during the study. The adsorption capacity of the vi rgin GAC was not fully characterized, and therefore Biofilter 3 results were not comparable to the results of the other three biofilters. In addition, the mechanisms for contaminant removal (e.g., biological transformation versus adsorption) were not evaluated. Therefore, the extent of biological activity and its effects on performance remain unknown. The results of the woodbased GAC media evaluation do not support the

PAGE 95

95 project hypothesis nor project objectives, thus they are not discussed further in the di ssertation. Any future comparative evaluations of biofilters composed of virgin GAC support media should include preexhaustion step (via high contaminant and/or NOM loading until complete breakthrough is observed). Hydraulic Characterization Clean bed headloss of the control biofilter was consistent through steady state testing. Table 41 provides the average headloss data for the control biofilter on a month by month basis for five months of steady state testing. Monthly clean bed headloss means remained within 9 % of the calculated theoretical headloss (0.90 ft) based on Darcys equation and information provided by the underdrain manufacturer and pilot fabricator. Generally, the pilot control biofilter showed consistent headloss profiles for 18hour filte r runs during steady state operation. However, terminal headloss values did fluctuate. Increased terminal headloss most often occurred during periods of high biofilter feed turbidity. High turbidity episodes often coincided with operator initiated sludge pond recycle. During these events, decant from the sludge pond (a backwash wastewater and sedimentationcoagulation sludge holding tank) is recirculated to the plant influent, increasing solids loading to the JKWTP and thus, to the pilot biofilters. Settled water turbidity excursions could rise from an average 0.8 NTU to as high as 2.2 NTU (Table 4 4). Events significant enough to increase 18hr terminal headloss by greater than 25% occurred three times during the fivemonth steady state evaluation. These events provided a realistic robustness characterization for turbidity removal and hydraulic performance for all pilot biofilters. Figure 41 illustrates the impact of a high

PAGE 96

96 solids loading event on the control biofilter through consecutive filter run headloss profiles. Table 4 1 Baseline characterization of pilot biofilter headloss Month of steady state operation Clean bed headloss (ft) Terminal headloss (ft) at 18 hr Mean *, Min. Max. Mean *, Min. Max. 1 (Apr) 0.85 0.05 0.76 0.94 5.5 0.9 4.23 7.8 2 (May) 0.92 0.04 0.81 0.97 7.1 1.4 4.5 9.3 3 (Jun) 0.98 0.04 0.85 1.05 4.7 0.5 4.0 6.2 4 (Jul) 0.97 0.05 0.81 1.05 4.9 0.5 4.1 6.0 5 (Aug) 0.91 0.03 0.83 0.98 5.9 0.6 4.7 7.0 Over 5 month study 0.92 0.06 0.76 1.05 5.6 1.2 4.0 9.3 Means are provided with the standard deviation of the data sets as value error. Hydraulic data from approximately four filter runs each month were omitted due to routine (<4 hour) process interruption for pilot maintenance or media collection. Loading rate held to 4.5 gpm/ft 2 during testing. Figure 41 Control pilot biofilter headloss profiles impacted by sludge pond recycle to JKWTP influent 0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 0 1 2 3 4 5 6 7 8 9 10 Turbidity (NTU) Headloss (ft) and Loading Rate ( gpm/ ft 2 ) Date Headloss Biofilter Effluent Turbidity Biofilter Feed Turbidity Operating Conditions Loading Rate: 4.5 gpm/ ft2Run time: 18 hours

PAGE 97

97 Water Quality Characterization General The baseline biofilter water treatment performance characterization was performed over five months of steady state operation. Full scale and pilot biofilter effluent samples were monitored for DOC, Fe, Mn, NH4N, PO4P, MIB, geosmin, and a suite of over 150 pharmaceuticals and pesticides. These constituents were tracked to characterize biofilter feed water quality and to verify that the Citys treatme nt objectives were met in the filter effluent. Table 42 summarizes the Citys water quality objectives and Table 43 provides the mean full scale raw and biofilter effluent values for those criteria during pilot studies. Table 4 2 City of Arlington biofilter water treatment objectives Constituent Finished Water Objective/target concentration Turbidity < 0.3 NTU, 95 % of samples, max sample 1 NTU DOC Remove 10 % of biofilter feed DOC concentration Fe* < 300 g/L Mn* < 50 g/L MIB* Total odor number < 3 (~10 ng/L) Geosmin* Total odor number < 3 (~10 ng/L) Atrazine* < 3 g/L Misc. pharmaceuticals and pesticides* Unspecified concentration reduction Background metal and trace organic loads to the biofilters were low/below detection through most of the study; therefore, pilot biofilter feed contaminant spikes were performed to characterize removal performance.

PAGE 98

98 Table 4 3 City of Arlington full scale biofilter performance Constituent Biofilter feed mean water quality Biofilter effluent mean water quality Turbidity (NTU) 0.8 0.4 NTU 0.06 0.01 NTU DOC (mg/L) 3.50 0.2 3.17 0.12 (~10 % ) Fe (g/L) 34 75 < 8.3 g/L Mn (g/L) 8 5 < 2.4 g/L MIB (ng/L) < 1.4 g/L < 1.4 g/L Geosmin (ng/L) < 1.4 g/L < 1.4 g/L Atrazine (g/L) 0.39 0.07 0.47 0.20 Misc. pharmaceuticals and pesticides Provided in Table 1.2 (< 1 g/L for all measured) Provided in Table 1.2 (< 1 g/L for all measured) Data collected from April 2009 through August 2009. Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Background metal and trace organic loads to the biofilter s were low/below detection through most of the study; therefore, pilot biofilter feed contaminant spikes were performed to characterize removal performance. Turbidity Control biofilter effluent turbidity readings remained stable and mean monthly values w ere similar to full scale values throughout steady state testing. Table 44 summarizes feed turbidity trends along with turbidity breakthrough from the pilot and fullscale biofilters. Turbidities maintained compliance with the USEPA Surface Water Treatment Rule, as greater than 95% of the effluent samples had turbidities less than 0.3 NTU. No turbidity values over 1 NTU were observed in pilot or full scale biofilter effluents. Figure 4 1 illustrates the biofilter ripening period for the pilot control biofilter over approximately eight days of steady state operation. The ripening period was the time required for a filter to meet effluent turbidity objectives after it is put into service. Figure 42 provides biofilter effluent turbidity values across two filter runs.

PAGE 99

99 Table 4 4 Bas eline characterization of biofilter turbidity removal Month of steady state operation Biofilter feed *, (NTU) Pilot control biofilter effluent (NTU) Fullscale biofilter effluent*, (NTU) Mean ** Min. Max. Mean ** Min. Max. Mean ** Min. Max. 1 (Apr) 1.4 0.6 1 2.2 0.1 0.06 0.07 0.62 0.06 0.02 0.04 0.08 2 (May) 0.9 0.4 0.5 1.4 0.08 0.06 0.05 0.52 0.06 0.02 0.04 0.08 3 (Jun) 0.7 0.5 0.4 1.4 0.07 0.02 0.06 0.42 0.06 0.01 0.05 0.08 4 (Jul) 0.5 0.3 0.3 0.8 0.05 0.02 0.04 0.30 0.06 0.01 0.04 0.07 5 (Aug) 0.6 0.3 0.3 1.0 0.05 0.02 0.05 0.30 0.05 0.01 0.04 0.06 Over 5 month study 0.8 0.8 0.3 2.2 0.07 0.06 0.04 0.62 0.06 0.02 0.04 0.08 Biofilter feed water was JKWTP settled/ozonated water for pilot and full scale biofilters. Biofilter feed and composite full scale biofilter effluent turbidities were measured using a desktop turbidimeter. Pilot biofilter effluents were measured continuously (5 minute intervals) using inline instrumentation ( Chapter 3) Pilot control biofilter effluent maximum values were observed during filter ripening. Fullscale biofilter effluent turbidities were measured using a composite sample (blended from all active biofilters). These samples were collected and measured every four hours (over a 24h our period) by plant staff. ** Means are provided with two standard deviations (to capture 95% of the data distribution) to show regulatory compliance. Fullscale biofilter effluent composite samples did not show high turbidity breakthrough during ripening, as only one biofilter is backwashed at a time. Dissolved Organic Carbon After turbidity reduction, biodegradable organic matter (BOM) removal is typically the most important water treatment objective for biofilters. Organic carbon provides an energy and carbon source for the heterotrophic bacteria that populate the biofilter. DOC measurement is a relatively inexpensive analysis that can be performed by most water quality laboratories and provides an indication of BOM reduction. Therefore, observed improvements in DOC removal may suggest an increase in microbial activity in biofilters and a reduction of regrowth potential in the distribution system. Table 45 provides a summary of DOC data for the pilot control and full scale biofilters.

PAGE 100

100 Figure 42 Pilot control biofilter effluent turbidity profiles across two filter runs The data provided in Table 45 show low variability in the mean full scale and pilot control biofilter effluent DOC concentrations over the fivemonth steady state evaluation. Figure 43 presents average monthly DOC removal for the pilot control and full scale biofilters, further illustrating the pilot/full scale similitude. Figure 4 3 also shows the average temperatures for each month of operation. The test period captured a 67% increase in average temperature. However, this appeared to have only provided a small improvement in DOC removal in the full scale biofilter performance. No significant correlation between pilot biofilter DOC removal performance and temperature was 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Pilot Biofilter Control Headloss Pilot Biofilter Control Effluent Turbidity Headloss (ft) and Loading Rate (gpm/ ft 2 ) Turbidity (NTU) Operating Conditions Loading Rate: 4.5 gpm/ ft2Run time: 18 hoursDate

PAGE 101

101 found, which is likely due to the limited range of temperature variation and the variability in the DOC values across the experimental period. The control pilot and full scale biofilters removed an average of approximately 10% of the influent DOC, which amount ed to approximately 0.4 mg/L. If 0.4 mg/L of DOC represented all BOM, 47 g/L NH4N and 10 g/L PO4P would be necessary in the biofilter influent to establish the target 100:10:1 C:N:P molar ratio. However, it is likely that additional nondegraded BOM is present; therefore, th e nutrient requirement s must be determined iteratively if additional BOM removal is observed. Table 4 5 Baseline characterization of biofilter DOC removal Month of steady state operation Settled Water Temp (C) Biofilter feed (mg/L) Pilot control biofilter effluent (mg/L) Full scale biofilter effluent (mg/L) Mean Mean Min. Max. Mean Min. Max. Mean Min. Max. 1 (Apr) 18 1.7 3.43 0.19 3.12 3.66 3.08 0.11 2.89 3.21 3.19 0.14 3.00 3.37 2 (May) 24 1.2 3.45 0.25 2.88 3.66 3.09 0.10 2.98 3.25 3.20 0.05 3.12 3.29 3 (Jun) 28 0.5 3.71 0.18 3.60 4.07 3.12 0.12 3.01 3.30 3.25 0.1 3.08 3.33 4 (Jul) 30 0.3 3.51 0.13 3.33 3.68 3.15 0.25 2.80 3.55 3.13 0.21 2.83 3.40 5 (Aug) 30 0.3 3.52 0.09 3.40 3.69 3.08 0.18 2.87 3.44 3.11 0.09 2.95 3.20 Over 5 month study 26 4.8 3.50 0.19 2.88 4.07 3.10 0.15 2.80 3.55 3.17 0.12 2.84 3.40 Pilot was housed in a heated room; however, pilot effluent temperatures remained within 2 C of settled water temperatures. Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations.

PAGE 102

102 Figure 43 Pilot control biofilter and full scale biofilter steady state DOC removal performance2Nutrients Nutrients (NH4N, PO4P) were monitored throughout the baseline characterization and control studies to identify potential limitations. Table 46 provides a summary of the collected nutrient data. The biofilter feed PO4P concentrations were typically below the method detection limit of 7 g/ L, suggesting that the pilot control and full scale biofilters were operated with a PO4P limitation. Furt hermore, NH4N concentrations varied significantly during testing, indicating that occasional periods of N limitation had occurred. 1 The error bars presented in all figures in this Chapter represent the standard deviation of the data set. This standard deviation accounts for operational variability (i.e. feed water conditions) and sampling/analytical error. 0 5 10 15 20 25 30 35 0.7 0.8 0.9 1 1.1 1.2 April May June July August Temperature (Degree C) Average Monthly Normalized DOC (C/Co) Month of Operation Pilot Biofilter Effluent DOC Full Scale Biofilter Effluent DOC Biofilter Feed Temperature

PAGE 103

103 Metals Metals characterization included monitoring Mn and Fe in pilot control and full scale biofilter feed and effluents. A s discussed in Chapter 1, Mn is a primary water quality concern for the City. During the fivemonth steady state analysis, background Mn concentrations in the biofilter feed were well below the 50 g/L secondary MCL. Therefore, 50 g/L Mn was spiked to the pilot control biofilter over the last month of testing to fully characterize baseline removal capabilities under moderate loading conditions. Subsequently, the biofilter feed Mn spike was increased to approximately 200 g/L for a period of one week of ro bustness testing. Table 47 summarizes the baseline Fe and Mn removal performance. The data presented in Tables 46 and 47 demonstrate similitude between the pilot and full scale filters for Mn and Fe removal performance at background feed concentrations Average Mn and Fe effluent concentrations for the pilot control and full scale biofilters were below detection. However, biofilter feed background loadings remained low throughout this phase of testing. Spiking did provide evidence of Mn removal in the pilot control biofilter, yielding average feed reductions of 76 44 and 89 6% (error as standard deviation) for the moderate and high spike tests, respectively. Despite the high level of treatment, Mn breakthrough near the secondary MCL was observed. It should also be noted that Mn breakthrough below the secondary MCL may also be problematic, as it may contribute to long term accumulation of Mn oxide precipitates in the distribution system, potentially leading to periodic sloughing and colored water event s. Figure 44 illustrates pilot baseline control mean Mn feed and effluent concentrations during the spiking tests.

PAGE 104

104 Table 4 6 Bas eline characterization of nutrient feed and biofilter utilization Nutrient Biofilter feed *, Pilot control biofilter effluent *, Full scale biofilter effluent *, Mean Min. Max. Mean Min. Max. Mean Min. Max. PO 4 P (g/L)
PAGE 105

105 Figure 44 Pilot control biofilter steady state and peak load Mn removal performance Table 4 8 Bas eline characterization of biofilter taste and odor removal Contaminant Background biofilter feed Pilot control biofilter effluent Full scale biofilter effluent Mean Min. Max. Mean Min. Max. Mean Min. Max. ** MIB, (ng/L)
PAGE 106

106 Tastes and Odors The City of Arlingtons water treatment objectives include the removal of MIB and geosmin. Background biofilter feed levels for these contaminants remained low, relative to historical values and their treatment objective (as described in Table 41). The pilot control biofilter was oper ated with unmodified feed for six weeks to evaluate T&O removal at the low level background concentrations. After this initial study, 40 ng/L of MIB was dosed to the pilot feed during five months of steady state operation to better characterize removal per formance. An intermittent high load (>100 ng/L) MIB spiking test was conducted at the end of steady state operation. Table 48 summarizes the pilot control and full scale biofilter T&O removal performance at background concentrations and pilot control perf ormance during the spiking tests. Pilot control and full scale biofilter MIB and geosmin breakthrough remained low under background loading conditions. However, the average influent concentration for both contaminants remained below their MDLs. Spiking did provide evidence of MIB removal in the pilot control biofilter, yielding average feed reductions of 66 32 and 86 4 % (error as standard deviation) for the moderate and high spike tests, respectively. Despite the high level of treatment, the average MIB breakthrough for both high and moderate loads remained above the odor threshold concentration of 10 ng/L. Indeed, approximately 50% of effluent samples (N=38) collected from the baseline control pilot filter during the moderate spiking tests showed MIB concentrations over 10 ng/L. The observed MIB breakthrough in the pilot study illustrates the limitations of the existing biofiltration processes at the JKWTP and PBWTP. Historically, MIB and geosmin analyses on City finished water have been limited; however the high number of

PAGE 107

107 seasonal consumer complaints for earthy musty smelling water suggests insufficient performance (Hunt, 2009). Pharmaceuticals and Pesticides Background biofilter feed concentrations of pharmaceuticals and pesticides remained below 1 g/ L for all tracked parameters. Table 49 provides a summary of the pharmaceuticals and pesticides detected in the biofilter feed and the effluent concentrations observed in the pilot control and full scale biofilters. Pilot biofilter feed spiking of carbamazepine, atrazine, and caffeine was conducted for one week to better characterize their removal acros s the pilot biofilters. Table 410 provides a summary of the results from this spiking test. The data in Table 410 indicate that the pilot control biofilter was capable of removing a portion of the spiked contaminants (22 to 40 % ), though end products were not identified. Table 4 9 Baseline performance comparison of the pilot and full scale filters Contaminant Mean background biofilter feed *, Mean pilot control biofilter effluent *, Mean full scale biofilter effluent *,, Atrazine (g/L) 0.39 0.07 0.42 0.12 0.47 0.20 Deethylatrazine (g/L) 0.56 0.15 0.58 0.10 0.54 0.12 Deisopropylatrazine (g/L) 0.26 0.08 0.26 0.05 0.24 0.05 Hydroxyatrazine (g/L) 0.21 0.22 0.17 0.10 0.12 0.03 Simazine (g/L) 0.24 0.44 0.17 0.23 0.03 0.07 Metolachlor (g/L) 0.03 0.02 0.03 0.02 0.03 0.02 Meprobamate (ng/L) 0.28 0.04 0.23 0.04 0.25 0.0 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Samples were tested against a suite of over 150 pharma ceuticals and pesticides, those with measurable concentrations are included in this table. Samples collected from the effluent of full scale filter No. 2. Data set includes five months of steady state operation (atrazine, N = 6; meprobamate N=2, all ot hers N = 9).

PAGE 108

108 Table 4 10 Pilot biofilter treatment performance for spiked atrazine, carbamazepine, and caffeine Contaminant Background biofilter feed *, Pilot control biofilter effluent *, Atrazine (g/L) 2.8 0.0 2.2 0.2 Carbamazepine ( g/L) 0.5 .3 0.3 0.1 Caffeine (g/L) 2.4 0.2 1.5 0.0 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Data set includes 1 week of steady state operation (all compounds, N=3). Summary The objective of this study was to characterize baseline water treatment performance, confirming similitude between pilot and full scale treatment and to provide control data for parallel tests occurring with other pilot filters. Similitude was characterized by operating a control biofilter under full scale operational conditions (i.e., no enhancement strategies were used for the control bi ofilter) from March 5, 2009 through October 1, 2009. These tests provided approximately seven months of steady state data that confirmed and further elucidated the treatment capabilities of the existing process at the JKWTP. Monitored water quality paramet ers included DOC, total Mn, ammonianitrogen (NH4N), orthophosphatephosphorus (PO4P), MIB, and a suite of over 150 pharmaceuticals and pesticides. These water quality data demonstrated treatment performance similitude between the pilot and full scale fi lters. Measured water quality values between the control biofilter and the full scale biofilter were within the standard deviations of their respective data sets. The baseline characterization also confirmed the process limitations of the existing full sca le system. Both Mn and MIB breakthroughs were observed under moderate and high biofilter influent load conditions.

PAGE 109

109 CHAPTER 5 SUBSTRATE ENHANCEMEN T STUDIES Objectives The objective of this study was to evaluate various primary substrate augmentation strategies for enhancing biofiltration performance. Though secondary substrates (e.g., recalcitrant DOC, MIB, geosmin, pesticides and pharmaceuticals) can be biodegraded, bacteria gain little to no energy in doing so, which means a primary substrate must be biodegraded simultaneously if any biodegradation of the secondary substrate is to be achieved. The rate of secondary substrate degradation is proportional to the concentration of active biomass present, which is, in part, a function of the concentration of primary substrate. Thus, biological treatment processes designed to biodegrade these compounds require the presence of a primary substrate. Increasing the concentration of primary substrate (e.g., by intermediate ozonation, which increases the amount o f biodegradable organic matter, or primary substrate augmentation), can increase the rate of trace organic compound degradation (Lim et al. 2008). It can also enhance the removal rate of slowly degradable natural organic matter (NOM) (Hozalski and Bouwer 2001). This study evaluated the addition of four primary substrates, including acetic acid, molasses, MicroC (a proprietary glycerinbased product produced by Environmental Operating Solutions, Inc, Bourne, MA), and ethanol. The substrates were evaluated individually during four successive twoto six week phases, each dosed at a target of 1 mg/L as carbon (C). Chemical Feed Observations Dosed substrate concentration was validated twice weekly by comparing substrateenhanced DOC concentration in the biofi lter feed against the background

PAGE 110

110 (pilot control) biofilter feed DOC concentration. The measured dose delivered to the pilot column often deviated from the target. Analyses of acetic acid, MicroC, and ethanol influent samples revealed that doses of these s ubstrates were approximately 10 to 40% higher than the target, which may have been due to inadequate mixing in the biofilter feed or siphoning from the chemical feed tanks The mean molasses dose was approximately 130% greater than the target. Stock soluti on checks indicated that molasses was settling in the chemical tank, yielding higher concentrations of DOC at the tank bottom where chemical was drawn. This high dose was likely a factor in the high headloss and DOC breakthrough observations discussed later in this chapter. Hydraulic Characterization Substrate augmentation strategies generally impaired biofilter hydraulic performance. Clean bed headloss was unaffected by substrate addition and generally remained below 1 ft for all substrates tested. However terminal headloss consistently exceeded that of the control biofilter. Indeed, many substrateenhanced biofilter filter runs were terminated prematurely (prior to the typical 18hour run) due to headloss greater than 13.5 ft, the maximum safe total dynam ic head load for the pilot feed pumps. Typically, terminal headloss increased after successive filter runs for each substrate tested. Additional biodegradable carbon increases the relative nitrogen and phosphorus limitations in the biofilter feed. It is believed that the observed increase in headloss in the substrateenhanced columns was due primarily to the increased microbial production of EPS, driven by nutrient stress. This phenomenon was described in Chapter 2 and is investigated further in Chapters 6 and 8. Table 51 provides the average terminal headloss data for the substrateenhanced biofilters compared to that of the biofilter control over the same time Figure 51

PAGE 111

111 provides an illustration of substrateenhanced and biofilter control headloss profil es for a typical week of filter runs for MicroC augmentation. Water Quality Characterization General Water treatment performance characterization included routine sampling and wate r quality analyses (Table 43). pH was also closely monitored during the acetic acid substrateenhanced biofiltration studies; however, no variation greater than 0.1 pH unit was observed relative to the biofilter control. Table 5 1 Characterization of substrate enhanced pilot biofilter headloss Supplemental substrate tested Substrate enhanced biofilter terminal headloss (ft) Biofilter control terminal headloss (ft) Mean Min. Max. Mean Min. Max. Acetic acid 10.3 2.4 5.5 >13.5 5.4 1.2 3.1 8.8 Molasses >13.5 10.8 >13.5 7.2 1.3 4.5 9.0 MicroC ** 8.8 2.8 4.3 >13.5 4.7 0.4 4.1 6.0 Ethanol >13.5 12 >13.5 5.3 0.6 4.7 6.8 Biofilter control operated without supplemental substrate. Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Includes 37 biofilter runs. Six acetic acid supplemented biofilter filter runs were terminated because headlos s exceeded 13.5 feet. Includes 38 biofilter runs Approximately 85% of molasses supplemented biofilter filter runs were terminated prematurely because headloss exceeded 13.5 feet. ** Includes 45 biofilter runs Clogged biofilter effluent lines artificially elevated headloss through six filter runs with MicroC supplementation, these runs were not included in the above analysis. Three additional MicroC supplemented biofilter runs were terminated because headloss exceeded 13.5 feet. Lines were cl ogged with white gelatinous material believed to be related to biological growth. Includes 12 filter runs 90% of ethanol supplemented biofilter runs were terminated prematurely because headloss exceeded 13.5 feet.

PAGE 112

112 Figure 51 Comparison of substra te enhanced (MicroC) and biofilter control headloss profiles Turbidity Substrate enhanced biofilter effluent turbidity readings remained stable and were similar to fullscale and pilot biofilter control values throughout testing. All turbidities maintaine d compliance with the USEPA Surface Water Treatment Rule, as greater than 95% of the effluent turbidity samples were less than 0.3 NTU. Indeed, mean effluent turbidities remained below 0.08 NTU for all conditions tested. No turbidity values over 1 NTU were observed in substrateenhanced biofilter effluent. Temporary headloss drops and excess turbidity was not observed during high headloss events (>13 feet), suggesting that significant short cir cuiting did not occur. Figure 52 illustrates typical substrateenhanced effluent turbidity profiles for a typical week of filter runs. Although 0 2 4 6 8 10 12 14 16 7/7/09 7/8/09 7/9/09 7/10/09 7/11/09 7/12/09 7/13/09 7/14/09 Headloss (ft) Substrate Enhanced Biofilter Control Biofilter Operating Conditions Loading Rate: 4.5 gpm/ ft 2 Run time: 18 hours Substrate Target Dose: MicroC 1 mg/L as C

PAGE 113

113 not shown in Figure 52, occasional turbidity breakthroughs as high as 0.6 NTU were observed in all biofilters immediately following a backwash (less than 15 minutes of produc tion). Figure 52 Comparison of substrate enhanced and biofilter control turbidity profiles for a typical week of filter runs Dissolved Organic Carbon Table 52 provides biofilter feed DOC (includes dosed substrate concentration) and effluent DOC for each substrate test. Table 53 compares substrateenhanced biofilter effluent DOC concentrations with the control biofilter effluent DOC concentration during the same period. Effluent DOC concentrations shown are normalized to the background biofilter feed DOC (NOM) and total DOC concentrations (NOM + substrate, where added). Effluent DOC normalization to the background biofilter feed provided 0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 7/7/09 7/8/09 7/9/09 7/10/09 7/11/09 7/12/09 7/13/09 7/14/09 Turbidity (NTU) Substrate Enhanced Biofilter Control Biofilter Operating Conditions Loading Rate: 4.5 gpm/ ft2Run time: 18 hours Substrate Target Dose: MicroC1 mg/L as C

PAGE 114

114 simple metric to determine if substrateenhanced biofiltration improved (or diminished) NOM removal relative to the control biofilter. The hypothesis for the substrateenhanced biofiltration strategy assumed that the substrates selected would be easily degradable and be removed in addition to the bioavailable carbon present in the biofilter feed. Figure 53 provides a graphic representation of the data presented in Table 53. Table 5 2 Substrate enhanced biofilter DOC removal characterization Substrate tested Background DOC (mg/L) Dose Biofilter feed (background + dosed DO C) (mg/L) Substrate enhanced biofilter efflue nt DOC (mg/L) Mean Mean Mean Min. Max Mean Min. Max Acetic acid 3.5 0.2 1.4 0.5 4.9 0.5 4.3 5.4 3.1 0.1 3.0 3.3 Molasses 3.6 0.1 2.3 0.1 5.9 0.1 5.9 6.0 3.6 0.0 3.6 3.6 MicroC 3.5 0.1 1.4 0.7 4.9 0.7 4.0 6.7 3.1 0.2 2.9 3.4 Ethanol 3.6 0.1 1.1 0.3 4.6 0.3 4.3 4.8 3.0 0.2 2.8 3.2 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Molasses partitioning in the chemical feed tank was observed. Concentrations higher than target were dosed to the substrate enhanced biofilter. Table 5 3 Substrate enhanced biofilter normalized DOC removal characterization Supplement al substrate tested Substrate enhanced biofilter effluent normalized to feed C (C/Co[Background + Dosed] ) Substrate enhanced biofilter effluent normalized to background C (C Effluent /C o,Background ) Biofilter control effluent normalized to background C (C Effluent /C o,Background ) Mean Min. Max. Mean Min. Max. Mean Min. Max. Acetic acid 0.62 0.09 0.49 0.74 0.91 0.04 0.88 1.00 0.89 0.01 0.88 1.00 Molasses 0.61 0.14 0.60 0.62 0.99 0.00 0.99 0.99 0.86 0.01 0.86 0.86 MicroC 0.64 0.07 0.48 0.77 0.88 0.04 0.82 0.93 0.88 0.02 0.81 1.03 Ethanol 0.65 0.05 0.60 0.70 0.85 0.02 0.82 0.87 0.90 0.3 0.84 0.93 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations.

PAGE 115

115 Figure 53 Comparison of substrate enhanced and biofilter control normalized DOC concentrations3As shown in Table 53 and Figure 53 DOC removals (normalized to the background concentration) varied between the substrates tested. Acetic acid and glycerin biofilter substrateenhancement yielded effluent DOC concentrations similar to the biofilter control. Molasses supplem entation resulted in additional biofilter DOC breakthrough. This is likely due to the inconsistent dosing and the heterogeneous nature of molasses product used (i.e., molasses contains humic materials that are likely more recalcitrant than the other substr ate tested).On average, ethanol appeared to perform better than the biofilter control. However, a paired t test analyses of DOC 3 The error bars presented i n all figures in this Chapter represent the standard deviation of the data set. This standard deviation accounts for operational variability (i.e. feed water conditions) and sampling/analytical error. 0.70 0.75 0.80 0.85 0.90 0.95 1.00 1.05 Acetic Acid Molasses Glycerin Ethanol Effluent DOC Normalized to Background DOC (C/Co) Substrate Test Phase Substrate Enhanced Biofilter Control Biofilter

PAGE 116

116 removal data from the ethanol enhanced biofilter and the control indicated that the difference between the means were not statis tically significant [ t ( 3) = 3.18, two tail p = 0.17 (p>0.05)]. Thus, the substrateenhancement studies failed to identify a sole substrate that would improve biofilter DOC removal upon supplementation. Under all substrates tested, the addition of bioavail able C exacerbated the existing nutrient limitation in the biofilter feed. Nutrient limitation may have diminished potential substrate utilitization, thus DOC removal. Nutrients The addition of approximately 1 mg/L biodegradable organic carbon increased th e relative nitrogen and phosphorus limitations in a given biofilter feed (Table 43). The C:N:P ratio for the substrate enhanced biofilter feed was approximately 100:2.6:0, significantly offset from the 100:10:1 target ratio. The effect of this offset will be discussed in Chapter 6. Metals Fe and Mn removal were observed for the substrateenhanced and control biofilters under each test condition. However, metals removal performance was difficult to compare among substrates due to generally low and varied fe ed concentrations. Mean biofilter feed Fe concentrations were less than 100 g/L for all substrates tested with three excursions over 300 g/L (all during acetic acid testing). Table 54 summarizes the substrateenhanced biofilter Fe removal data. Mean bi ofilter feed Mn levels remained below the limit of detection (10 g/L) during the substrate enhancement tests (Table 54). Therefore, moderate Mn (~50g/L) spiking tests were conducted to resolve removal performance with the ethanol substrateenhanced bio filter. Mn spiking was conducted for three weeks (N=6) with a

PAGE 117

117 mean dose of 63 g/L to the ethanol substrateenhanced biofilter and biofilter control. As shown in Table 55, the ethanol enhanced and control biofilters removed Mn to below detection. Table 5 4 Substrate enhanced biofilter Fe removal characterization Substrate tested Biofilter feed Fe (g/L)* Substrateenhanced biofilter effluent Fe (g/L)* Biofilter control effluent Fe (g/L)* Mean Min. Max Mean Min. Max Mean Min. Max Acetic acid 50 93
PAGE 118

118 provides the results for the MIB removal characterization during the substrate enhancement studies. Table 5 6 Substrate enhanced biofilter MIB remo val characterization Substrate tested Biofilter feed (ng/L) *, Substrateenhanced biofilter effluent (ng/L) *, Biofilter control effluent (ng/L) *, Difference between the means for substrateenhanced and control biofilters Mean Min. Max. Mean Min. Max. Mean Min. Max. P Value Acetic acid 37 24 20 101 25 12 4 42 23 12
PAGE 119

119 nutrient limitations. The effects of nutrient limitations and supplementation on MIB removal are discussed further in Chapter 6. Figure 54 Comparison of substrate enhanced and biofilter control normalized MIB concentrations Pharmaceuticals and Pesticides Select biofilter effluent samples were collected during each of the substrate enhancement tests for pharmaceutic al and pesticide monitoring. Atrazine, deethylatrazine deisopropylatrazine, hydroxyatrazine, simazine, metolachlor, and meprobamate were detected in the biofilter feed at ng/L concentrations. Biofiltration removals for these contaminants were not affected (favorably or adversely) by substrateenhancement, regardless of the substrate tested. 0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1.0 1.1 1.2 Acetic Acid Molasses MicroC Ethanol Normalized Effluent MIB Concentration (C/Co) Substrate Test Phase Substrate Enhanced Biofilter Control Biofilter

PAGE 120

120 Biological Activity ATP concentrations in biofilter media were monitored during the substrateenhancement studies. T hese tests indicated that biofilter terminal (end of filter run) biological activity (on a per media volume basis) was modestly improved (7 to 29 % ) relative to the biofilter control with substrate supplementation. The increased ATP concentrations measured in substrateenhanced biofilter media samples corresp onded to higher overall DOC removals (Table 52, Table 53) Table 5.4 summarizes the biofilter media ATP characterization for the substrateenhanced biofiltration strategy. Table 5 7. Substrate enhanced biofilter evaluation of media ATP concentrations Substrate tested Substrateenhanced biofilter measured media ATP Biofilter control measured media ATP Effect of substrate enhancement on ATP concentration relative to biofilter control Start of filter run (pg/L) End of filter run (pg/L) Start of filter run (pg/L) End of filter run (pg/L) Start of filter run (% increase ) End of filter run (% increase) Acetic acid 480,763 768,390 400,159 596,732 20 29 Molasses 532,202 713,713 488,889 667,347 9 7 MicroC 747,156 1 ,209,556 503,305 1 ,068,987 48 13 Ethanol 567,224 880,759 471,214 742,181 21 19 ATP test error was determined to be ~7.5 % through triplicate analyses, performed monthly Single sample sets performed for each substrate tested Additional data from the ATP tests (Figure 8 15) and other microbial tracking data for the substrateenhancement studies are presented in Chapter 8 Microbial Tracking Summary The objective of this study was to evaluate various primary substrate augmentation strategies for enhancing biofiltration performance. Incr easing the concentration of primary substrate (through intermediate ozonation or primary substrate augmentation) in a biological treatment application may increase the rate of recalcitrant

PAGE 121

121 and trace organic compound degradation through secondary substrate metabolism and/or cometabolism. This study evaluated four primary substrates, including acetic acid, molasses, MicroC, and ethanol. The substrates were evaluated individually through four successive oneto two month phases, each dosed at target of 1 mg/ L as C The impact of filter history (i.e., the order in which the substrates were augmented to the filter) was not characterized. However, some substrates, such as molasses, appeared to have an immediate impact on biofilter headloss performance. The subst rate enhanced biofilters yielded 25 to 150% higher terminal headloss than the control biofilter during 18hour filter runs (Table 51). However, sustainable (>1 week continuous operation) 18hour filter runs were not achieved with any substrate tested due to excessive headloss. Substrate augmentation strategies did not provide significant improvement of trace organics or metals removal relative to the control biofilter. Percent DOC removals were also comparable between the control and substrateenhanced bi ofilters when acetic acid and MicroC were used (Table 52 and Table 53 ). The biofilter supplemented with molasses showed twice as much DOC breakthrough as the control biofilter, while the biofilter operated with ethanol supplementation showed an average of 50 % higher background DOC removal relative to the control, though the differences in the mean breakthroughs were not shown to be statistically significant. The increase in net biofilter DOC removal (background + dosed carbon) corresponded to an increase d ATP concentration in the substrateenhanced biofilter media. .

PAGE 122

122 CHAPTER 6 NUTRIENT ENHANCEMENT STUDIES Objectives The objective of this study was to evaluate various nutrient augmentation strategies for enhancing biofiltration performance. Optimal microbial growth is depend e nt on a nutrient balance of carbon, NH4N, and PO4P. This balance is typically targeted at a molar ratio of 100:10:1, bioavailable C:N:P. The molar ratio translates to a concentration ratio of 1 mg/L: 0.117 mg/L: 0.026 mg/L, C:N:P. As discussed in Chapter 4, the biofilter feed at the JKWTP contained no detectable amounts of phosphorus (<0.01 mg/L), which is likely due to general source water limitation and to incidental phosphorus removal through enhanced coagulation. The background NH4N concentrations varied significantly during testing, indicating that occasional periods of N limitation had occurred (Table 46 and Table 6 7). A minimum of 0.010 mg/L PO4P and 0.0 47 mg/L of NH4N are necessary to prevent a biofilter nutrient limitation with background bioavailable carbon levels entering the filter process at ~0.4 mg/L C. Nutrient enhancement was performed by dosing PO4P (as phosphoric acid) and/ or NH4N (as ammonium chloride) to sufficiently eliminate nutrient limitation, thereby creating a carbon, or substrate, limitation. Nutrient enhancement was evaluated through the following tests: Nutrient Enhanced Biofilter Testing : The purpose of this test was to satisfy the baseline PO4P limitation by dosing phosphoric acid at a target of 0.020 mg/L P (200 % of stoichiometric requirement) to a pilot biofilter operated with assumed 0.4 mg/L of background bioavailable C (mean C removed in the pilot control biofilter). This test was conducted in parallel to biofilter control operation. The duration of this test was 6 weeks, with approximately 2 weeks of steady state operation. This test was evaluated across all hydraulic and water tre atment performance criteria, as described in Chapters 3 and 4.

PAGE 123

123 Substrate and Nutrient Enhanced Biofilter Testing: The purpose of this test was to satisfy the phosphorus limitation caused by substrate supplementation by dosing phosphoric acid at a target of 0.070 mg/L PO4P (200% of stoichiometric requirement) to a pilot biofilter operated with ~1.4 mg/L of bioavailable C (0.4 mg/L of background + 1 mg/L of supplemental C). This substrateand nutrient enhanced biofilter was operated in parallel with a sub strate only enhanced biofilter (same substrate) and the biofilter control. Substrate and nutrient enhancement testing was performed for each substrate tested (acetic acid, molasses, MicroC, and ethanol). In addition, one biofilter was operated for a twow eek period with ethanol, phosphoric acid, and ammonium chloride supplementation. This test was evaluated across hydraulic performance criteria and DOC removal. Nutrient EnhancedBiofilter Validation Testing : During the final two weeks of pilot testing, a t arget 0.020 mg/L phosphoric acid as P (200% of stoichiometric requirement) was dosed to the biofilter control (thus sacrificing it as a control). The purpose of this test was to validate previous observations by evaluating nutrient enhancement on a different biofilter. This test was evaluated across hydraulic performance criteria and DOC removal. Hydraulic Characterization Nutrient e nhancement t esting While clean bed headloss was unaffected by nutrient addition, phosphoric acid supplementation decreased bio filter terminal headloss (following 18 hours of operation) by approximately 15% (as average decrease over 17 consecutive runs) relative to the biofilter control ( p = 0.01, p 0.05) In addition, phosphoric acid supplementation provided more consistent biofil ter runs (50 % decrease in terminal headloss standard deviation). The hydraulic improvement was sustained throughout the two weeks of testing. Table 61 provides a summary of the hydraulic characterization during the nutrient enhancement tests. Figure 6 1 illustrates a selection of steady state headloss profiles during the final 9 runs of this study. Terminal headloss in the nutrient enhanced biofilter appeared to be trending down during the final filter runs, suggesting that additional hydraulic improvem ent may be possible with continued operation.

PAGE 124

124 Substrate and n utrient e nhancement t esting The substrateenhanced biofilter operated with phosphoric acid supplementation generally showed lower headloss profiles than the biofilter operated with supplemental substrate alone (for each substrate tested). However, the mean substrateand nutrient enhanced biofilters terminal headloss still exceeded that of the control biofilter for each of the substrates and multiple biofilter runs were prematurely terminated due to excessive headloss (greater than 13.5 ft). Although the phosphorous nutrient requirement was satisfied, the substrate (~1.4 mg/L as C) and nutrient enhanced (~0.070 mg/L PO4P) biofilter was operated with a NH4N feed limitation (C:N:P ~100:3:2). To sat isfy the NH4N limitation, ammonium chloride was fed (0.10 mg/L as NH4N) to the ethanol and nutrient enhanced biofilter during the final two weeks of testing. The addition of ammonium chloride decreased the mean terminal headloss by more than 35% as compared to ethanol and PO4P enhancements alone. Furthermore, the terminal headloss saw an immediate decrease of over 50% relative to the previous four filter runs. This observation confirmed that both PO4P and NH4N limitations may diminish biofilter hydraul ic performance. Table 62 provides a summary of the hydraulic characterization during the substrate and nutrient enhancement tests. Figure 62 illustrates the hydraulic performance improvement realized after ammonium chloride supplementation was implemente d on the substrateand nutrient enhanced biofilter.

PAGE 125

125 Table 6 1 Nutrient enhanced biofilter hydraulic performance Nutrient enhanced* biofilter terminal headloss (ft) Biofilter control terminal headloss (ft) Difference between the means for nutrient enhanced and control biofilters Mean Min. Max. Mean Min. Max. p value 5.1 0. 7 3.0 6.6 5.9 1.4 3.2 7.8 0.0 1 Target PO 4 P feed in the nutrient enhanced biofilter was 0.020 mg/L as P. Biofilter control operated without supplemental phosphoric acid. Includes 17 biofilter runs, means are provided with the standard deviation of the data sets as value error. Figure 61 Comparison of nutrient enhanced (PO4P) and biofilter control headloss profiles 0 1 2 3 4 5 6 7 8 9 10 5/2/2009 5/3/2009 5/4/2009 5/5/2009 5/6/2009 5/7/2009 5/8/2009 5/9/2009 5/10/2009 Headloss (ft) Biofilter Control Nutrient Enhanced Biofilter Operating Conditions Loading Rate: 4.5 gpm/ ft2Run time: 18 hours Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P

PAGE 126

126 Table 6 2. Substrate and nutrient enhanced biofilter hydraulic performance Condition tested Substrateenhanced biofilter terminal headloss (ft) Substrate and PO 4 P enhanced biofilter terminal headloss (ft) Substrate PO 4 P and NH4N enhanced biofilter terminal headloss (ft) Mean Min. Max. Mean Min. Max. Mean Min. Max. Acetic acid 10.3 2.4 5.5 >13.5 10.2 2.4 5.0 >13.5 NA NA NA Molasses >13.5 10.8 >13.5 >13.5 6.6 >13.5 NA NA NA MicroC** 8.8 2.8 4.3 >13.5 8.4 4.3 2.9 >13.5 NA NA NA Ethanol >13.5 12 >13.5 9.9 2.6 6.3 >13.5 6.3 0.6 5.1 7.0 Substrate and nutrient enhanced biofilters operated with a target phosphoric acid dose of 0.070 mg/L as P. Means are provided with the standard deviation of the data sets as value error Includes 24 (consecutive) filter runs. Includes 38 biofilter runs. Approximately 85% of biofilter filter runs terminated prematurely due to headloss exceeding 13.5 feet for substrate enhanced and substrate and nutrient enhanced biofilters ** Includes 45 biofilter runs. Clogged biofilter effluent lines artificially elevated headloss through six filter runs with MicroC and twelve filter runs with MicroC and phosphoric acid supplementation, these runs were not included in the above analysis. T he lines were clogged with a white gelatinous material believed to be a product of biological growth. Includes 12 filter runs. 90 % of ethanol supplemented and 33% of ethanol and phosphoric acid biofilter runs were terminated prematurely due to headloss exceeding 13.5 feet. Includes 16 filter runs NA = Not applicable. No parallel operation of biofilter control. Figure 62 Effect of ammonium chloride supplementation on substrate and nutrient enhanced biofilter operated with NH4N limitation

PAGE 127

127 Nutrient e nhancement v alidation Validation of the nutrient enhancement strategy was performed by dosing the biofilter control with phosphoric acid (200% of the stoichiometric requirement) to satisfy the baseline PO4P limitation. This required the sacrific e of the sole biofilter control. This experiment was performed at the end of the pilot testing and use of the control filter for validation was more appropriate than using other pilot filters that had received substrate and nutrient enhancements in the recent past. Therefore, hydraulic performance was evaluated by comparing the terminal headloss of the biofilter with the previous month of steady state hydraulic data. Table 63 provides a summary of the hydraulic characterization during the nutrient enhance ment validation test. As shown in Table 63, terminal headloss data suggest that phosphoric acid supplementation may have improved the hydraulic performance of the former biofilter control (p value = 0.04, p 5) Table 6 3. Nutrient enhancement valida tion on biofilter hydraulic performance Nutrient enhanced* biofilter terminal headloss (ft) Biofilter control terminal headloss (ft) Difference between the means for nutrient enhanced and control biofilters Mean Min. Max. Mean Min. Max. p value 5.4 0.4 4.1 6.1 5.9 0.6 4.7 7.0 0.04 Target PO 4 P feed in the nutrient enhanced biofilter was 0.020 mg/L as P. Table 6 7 summarizes measured PO4P dosages. Data taken from last month of steady state operation (August 2009). Includes 14 biofilter runs, means are provided with the standard deviation of the data sets as value error. Includes 36 biofilter runs, means are provided with the standard deviation of the data sets as value error.

PAGE 128

128 Water Quality Characterization General Water treatment per formance characterization included routine sampling and water quality analyses ( Chapter 4, Table 43). Selected samples were also collected for chloramine stability and DBPFP tests. Nitrite and nitrate analyses were performed on samples collected from the substrate and nutrient enhanced biofilter to characterize nitrification after ammonium chloride supplementation. Turbidity All turbidities maintained compliance with the USEPA Surface Water Treatment Rule, as greater than 95% of the effluent turbidity sam ples were less than 0.3 NTU. Mean turbidity breakthroughs remained below 0.08 NTU for all conditions tested. No turbidity values over 1 NTU were observed in nutrient enhanced biofilter effluent. However, substrateand nutrient enhanced biofilter turbidity excursions over mean biofilter control turbidities were observed during periods of frequent backwashing due to high headloss. Figure 63 illustrates nutrient enhanced effluent turbidity profiles for two typical filter runs. Table 64 summarizes mean effl uent turbidity breakthroughs for the nutrient enhanced and control biofilters during parallel operation.

PAGE 129

129 Figure 63 Nutrient enhanced biofilter turbidity profiles for typical filter runs Table 6 4 Bas eline characterization of nutrient enhanced biofilter turbidity breakthrough Biofilter Feed*, (NTU) Nutrient enhanced biofilter effluent (NTU) Biofilter control effluent (NTU) Mean Min. Max. Mean Min. Max. Mean Min. Max. 0.6 0.2 0.3 0.9 0.06 0.02 0.05 0.14 0.07 0.02 0.06 0.17 Biofilter feed water was JKWTP settled/ozonated water for pilot and full scale biofilters. Biofilter feed turbidities were measured using a desktop turbidimeter. Pilot biofilter effluents were measured continuously (5 minute intervals) using inline instr umentation ( Chapter 3). Means are provided with two standard deviations (to capture 95% of the data distribution) to show regulatory compliance. Pilot biofilter effluent maximum values were observed during filter ripening. DOC Low level phosphoric acid supplementation improved biofilter DOC removal performance during the nutrient enhanced biofilter tests. The nutrient enhanced biofilter removed a mean of 75 % more DOC than the biofilter control (0.7 mg/L with nutrient enhancement as compared to 0.4 mg/L in the control). A paired t test analyses of DOC removal data collected from the nutrient enhanced biofilter and the control indicated that the difference in t he means was statistically significant [ t ( 6) = 6.14, p = 0.0009]. This 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0 2 4 6 8 10 12 14 16 09/03/09 09/03/09 09/04/09 09/04/09 09/04/09 09/04/09 09/04/09 09/05/09 09/05/09 Turbidity (NTU) Headloss (ft) Date Nutrient Enhanced Biofilter Headloss Nutrient Enhanced Biofilter Effluent Turbidity Operating Conditions Loading Rate: 4.5 gpm/ ft 2 Run time: 18 hours Target Substrate Dose: none Target Nutrient Dose: Phosphoric acid 0.02 mg/L as P

PAGE 130

130 performance was sustained over 2 weeks of routine operation. A summary of the nutrient enhanced biofilter DOC removal performance is provided in Table 65. All DOC removal data was als o normalized (Ceffluent/Co,backgound) to illustrate changes in background BOM removal across the biofilters. Normalized DOC results for the nutrient enhancement study are provided in Figure 64. Table 6 5 Substrate and nutrient enhanced biofilter DOC r emoval performance Biofilter feed (background C, mg/L) Nutrient enhanced biofilter effluent (mg/L) Biofilter control effluent (mg/L) Mean* Min. Max Mean* Min. Max Mean* Min. Max 3.6 0.1 3.5 3.7 2.9 0.1 2.8 3.1 3.2 0.2 3.0 3.4 Target phosphoric acid dose for the nutrient enhanced biofilter was 0.020 mg/L as P. Table 6.7 summarizes measured PO4P dosages. Includes two weeks of steady state data (N=7). Means are provided with the standard deviation of the data sets = Figure 64 Comparison of nutrient enhanced and biofilter control normalized DOC removals 4 4 The error bars presented in all figures in this Chap ter represent the standard deviation of the data set. This standard deviation accounts for operational variability (i.e. feed water conditions) and sampling/analytical error.

PAGE 131

131 The results provided in Table 65 were validated when phosphoric acid was fed to the biofilter control during the last two weeks of pilot testing (Figure 6 5 ). DOC removal (as % of influent) increased by 35% when comparing the two weeks of nutrient enhancement against the preceding two weeks of biofilter control operation (p = 0.0001, p 0.05). Data collected during the substrateand nutrient enhanced biofilter tests suggest that nutrient enhancement also improves DOC removal for the substrateenhanced biofilters. Table 6 6 compares normalized effluent DOC concentrations for the nutrient and substrateenhanced biofilter, the substrateenhanced biofilter, and biofilter control for samples collected during parallel operation. Normalized biofilter effluent DOC data are also provided for the substrateand nutrient enhanced biofilter operated with phosphoric acid and ammonium chloride supplementation. Figure 6 6 illustrates the d ata collected during the parallel studies described in Table 66. The data presented in Table 66 and Figure 6 6 show no DOC removal improvement with nutrient and substrateenhancement relative to nutrient enhancement alone. A hypothesis for this observat ion is that most labile BOM was effectively removed with nutrient supplementation alone (to achieve substrate limitation). In addition, the data suggest that the supplemental substrates tested did not promote significant secondary substrate metabolism or c ometabolism of recalcitrant BOM.

PAGE 132

132 Figure 65 DOC removal performance improvement with nutrient enhancement of (previous) biofilter control Parallel studies between nutrient and substrateenhanced and nutrient enhanced biofiltration were limited; there fore, additional testing must be performed to better characterize relative performance. As shown in Table 65 and Figure 66, relative DOC removals between the substrateenhanced biofilter and substrate and nutrient enhanced biofilter were similar for all conditions tested excluding molasses. Steady state water treatment performance (DOC removal) with supplemental molasses was only achieved with supplemental phosphorus. The biofilter operated with supplemental ethanol and phosphoric acid removed an additional 0.2 mg/L of background DOC relative to the biofilter control (Figure 6 6) A

PAGE 133

133 paired t test analyses of this DOC removal data indicated that the difference between the means were statistically significant [ t ( 7) = 3.3, two tail p = 0.013]. However, paired t test analyses of all other substrates tested showed no statistically significant difference (p>0.05) from the substrateenhanced biofilter, the substrateand nutrient enhanced biofilter, and the biofilter control. The molasses and ethanol substrateand nutrient enhanced biofilter data support the nutrient enhancement tests, suggesting that PO4P limitations during supplemental substrate addition may inhibit optimal DOC removal in biofilters. Furthermore, the PO4P limitations may be resolved with phosphoric acid supplementation for improved DOC removal performance. The data in Table 66 also suggest that the other substrates may be less labile than ethanol or limit background DOC utilization. Table 6 6 Substrate and nutrient enhanced biofilter norm alized DOC removal characterization Condition tested *,, Substrateenhanced biofilter effluent (C effluent /C o,backgound ) Substrate and nutrient enhanced biofilter effluent (C effluent /C o,backgound ) Biofilter control effluent (C effluent /C o,backgound ) Mean ** Min. Max. Mean ** Min. Max. Mean ** Min. Max Acetic acid 0.91 0.04 0.88 1.00 0.89 0.03 0.86 0.95 0.89 0.01 0.88 1.00 Molasses 0.99 0.00 0.99 0.99 0.85 0.03 0.80 0.90 0.86 0.01 0.86 0.86 MicroC 0.88 0.04 0.82 0.93 0.85 0.03 0.80 0.90 0.88 0.02 0.81 1.03 Ethanol 0.85 0.02 0.82 0.87 0.83 0.04 0.74 0.88 0.89 0.03 0.84 0.93 Substrate (Ethanol), NH4N and PO 4 P NA NA NA 0.84 0.04 0.80 0.88 NA NA NA Target substrate dosage was 1 mg/L as C for all substrate enhanced and substrate and nutrient enhanced biofilter conditions tested. Target phosphoric acid dose was 0.070 mg/L as PO4P for all substrateand nutrient enhanced biofilter conditions tested. Mean biofilter feed PO4P concentrations are provided in Table 68. Target ammonium chloride dose was 0.10 mg/L as NH4N, mean dosage provided in Table 69. Effluents normalized to background feed DOC concentrations. ** Means are provided with the standard deviation of the data sets as value error. Molasses partitioning in the chemical feed tank was observed. Concentrations higher than target were dosed to the substrateenhanced biofilter. NA = Not applicable. No parallel operation of substrate enhanced biofilter or biofilter control.

PAGE 134

134 Figur e 66 Characterization of normalized DOC removal for substrateand nutrient enhanced biofilters Nutrients Nutrient b alance e valuation Biofilter effluent PO4P and NH4N concentrations were monitored during the nutrient enhancement studies to validate dosage, characterize utilization, and monitor breakthrough. The primary objective of these analyses was to verify whether sufficient PO4P and NH4N concentrations were established in the biofilter feed to shift the C:N:P ratio from a nutrient limited condit ion to a carbonlimited condition. Table 6 7 presents the nutrient ratio for each nutrient enhancement test by providing the mean bio available C, PO4P, and NH4N for each biofilter feed. The mean DOC removed was used as a 0.7 0.8 0.9 1.0 Acetic Acid Molasses MicroC Ethanol Normalized Effluent DOC (C/Co) Substrate Test Phase Substrate Enhanced Biofilter Effluent Substrate and Nutrient Enhanced Biofilter Effluent Biofilter Control Operating Conditions Substrate and Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.07 mg/L as P Target Substrate Dose: Various, 1 mg/L as C Substrate Enhancement Target Substrate Dose: Various, 1 mg/L as C

PAGE 135

135 conservative estimate of bio available C to identify the minimum nutrient concentrations required to satisfy the 100:10:1 bioavailable C:N:P ratio. The data in Table 67 indicate that under all conditions tested, phosphoric acid supplementation was sufficient to meet minimum PO4P req uirements. The 5month mean background NH4N feed concentration (0.043 mg/L, see Chapter 4, Table 46) met nutrient requirements under control conditions. However, periodic fluctuations created nutrient limitations during the nutrient enhanced biofilter validation tests. Substrate enhancement further exacerbated the observed NH4N limitations by increasing the concentration of bioavailable C. Table 6 7 Nutrient enhancement studies: nutrient balances Condition tested Feed bioavailabl e C* (mg/L) Feed NH4N (g/L) Feed PO4P* (g/L) C:N:P stoichio metric ratio Limitation Hydraulic Performance* Nutrient Enhanced Biofilter Initial phosphoric acid test 0.7 77 19 100:10:1 None Highly Improved Validation w/phosphoric acid (performed at end of study, control biofilter + PO 4 P) 0.5 32 23 100:6:1 NH 4 N Improved Substrate and Nutrient Enhanced Biofilter Acetic acid w/phosphoric acid 1.3 61 170 100:4:4 NH4N Highly Decreased Molasses w/phosphoric acid 1.0 25 211 100:2:8 NH 4 N Highly Decreased MicroC w/phosphoric acid 1.1 31 91 100:2:3 NH 4 N Highly Decreased Ethanol w/phosphoric acid 1.7 59 82 100:3:2 NH 4 N Highly Decreased Ethanol w/phosphoric acid and ammonium chloride 0.8 130 51 100:14:2 Carbon Similar to Control

PAGE 136

136 Table6 7. Continued Means for data collected during test duration. Value error is not provided in table for clarity, it can be found in Tables 64, 6 5, 67, and 68 as standard deviation. Target phosphoric acid dose was 0.020 mg/L as PO4P for all nutrient enhanced biofilter conditions tested. Target phosphoric acid dose was 0.070 mg/L as PO4P for all substrateand nutrient enhanced biofilter conditions tested. Poor chemical flow control observed during acetic acid and molasses testing. However, mean PO4P dosages well exceed minimum requirement to satisfy nutrient limitation. ** Relative to control biofilter Nutrient b reakthrough e valuation Biofilter effluent NH4N and PO4P concentrations were monitored to characterize breakthrough during nutrient enhancement studies. Under some conditions, excessive nutrient breakthrough during biofilter nutrient enhancement may contribute to biological regrowth in the distribution system. The PO4P data collected indicate that nutrient enhanced biofilters show 55 to 65% breakthrough of feed PO4P (mean effluents of ~14 to 16 g/L). The observed breakthrough was likely due to multiple factors including excess feed (200% of nutrient requirement fed under most conditions), fluctuations in chemical delivery, and fluctuations with alum floc culant carryover (thus precipitation of PO4 P in filter media). These PO4P breakthrough levels were considered low as they were below the minimum reporting limits (MRLs) for many utilities, including the City However, the lo ng term impact of low level PO4P breakthrough on the Citys distribution system is unknown and, therefore, must be studied further. Mean PO4P breakthroughs of 55 to 94% (mean effluents of ~38 to 160 g/L) were observed in the substrateand nutrient enha nced biofilter, although biofilter influent overfeeding (>200% ) was observed during some tests. It is important to note that PO4P breakthrough does not necessarily suggest excess PO4P biofilter feed or nonutilization. PO4P does not biologically transfo rm and may appear in effluent samples as

PAGE 137

137 sloughed biomass or extracellular materials Table 68 summarizes the PO4P characterization for the nutrient enhancement studies. Table 6 8 Nutrient enhancement studies: PO 4 P characterization Condition tested Feed PO 4 P (g/L) Biofilter effluent PO4P concentration (g/L) Normalized biofilter effluent PO4P concentration (C/C o ) Mean Min. Max. Mean Min. Max. Mean Min. Max Nutrient enhanced biofilters Initial phosphoric acid test 20 5 12 25 14 4 11 20 0.55 0.36 0.95 0.08 Validation w/phosphoric acid 23 1 14 36 16 5 10 25 0.65 0.19 0.81 0.28 Substrateand nutrient enhanced biofilters Acetic acid w/phosphoric acid 170 80 186 147 160 23 132 197 0.94 0.08 0.84 1.0 Molasses w/phosphoric acid 211 14 86 360 98 73 33 178 0.67 0.50 0.09 0.97 MicroC w/phosphoric acid 91 79 255 0 57 62 0 194 0.55 0.22 0.27 0.95 Ethanol w/phosphoric acid 82 33 104 25 52 22 22 77 0.59 0.19 0.28 0.88 Ethanol w/phosphoric acid and ammonium chloride 51 15 40 72 38 21 16 66 0.72 0.22 0.40 0.91 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Target phosphoric acid dose was 0.020 mg/L as PO4P for all nutrient enhanced biofilter conditions tested. Target phosphoric acid dose was 0.070 mg/L as PO4P for all substrateand nutrient enhanced biofilter conditions t ested. Poor chemical flow control was observed during acetic acid and molasses testing. However, mean PO4P dosages well exceeded minimum requirement to satisfy nutrient limitation. Table 69 provides NH4N characterization through the nutrient enhancement studies. Under NH4N limiting conditions, NH4N mean utilizations (assimilation and oxidation) ranged from 28 to 52% across the conditions tested. The NH4N breakthrough levels were higher than expected; however, the limitation may shift

PAGE 138

138 biofilter microbial communities to those requiring less NH4N than assumed in the nutrient balance. ( Sekar et al. 2002; Davidson et al. 2007). Interestingly, NH4N supplementation increased utilization to a mean of 85% This increase is possibly due to a biofi lter population shift to nitrifying bacteria. Table 6 9 Nutrient enhancement studies: NH 4 N characterization Condition tested Feed NH 4 N (g/L) Biofilter effluent NH4N concentration (g/L) Normalized biofilter effluent NH4N concentration (C/C o ) Mean Min. Max. Mean Min. Max. Mean Min. Max Nutrient enhancement Initial phosphoric acid test 77 41 46 161 39 22 16 65 0.65 0.41 0.1 1.0 Validation w/phosphoric acid 32 10 18 42 15 4 12 29 0.48 0.15 0.36 0.70 Substrate and nutrient enhancement Acetic acid w/phosphoric acid 61 56 16 180 24 11 11 42 0.72 0.15 0.56 0.88 Molasses w/phosphoric acid 25 10 13 39 17 10 10 34 0.38 0.16 0.16 0.59 MicroC w/phosphoric acid 31 13 14 50 16 6 10 27 0.31 0.07 0.22 0.40 Ethanol w/phosphoric acid 59 17 42 88 21 5 17 30 0.49 0.22 0.31 0.66 Ethanol w/phosphoric acid and ammonium chloride 130 72 44 210 16 5 9 21 0.15 0.06 0.08 0.20 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Feed NH4N levels include background only. Feed NH 4 N levels include background and dosed ammonium ch loride. Table 610 summarizes biofilter effluent nitrogen speciation during the ammoni um chloride supplementation. At nonchloraminating utilities, NH4N breakthrough may decrease effluent stability and increase chlorine demand. However, the data show near

PAGE 139

139 complete nitrification of all NH4N to NO3N in the test biofilter. Nitrification in the biofilter is likely driven by the oxidation processes of autotrophic bacteria that utilize ammonia and nitrite as electron donors. It is important to note, NH 4 N is not oxidized during assimilation. However, assimilated NH 4 N (e.g., proteins, amino acids nucleotides ) may be cycled within the biofilter and ultimately nitrified. Effluent nitrate remained well below current drinking water limits (10 mg/L as N). Figure 67 illus trates nitrogen speciation before and after treatment through substrateand nutrient enhanced biofiltration. The, a portion of the influent NO3 -N would be retained in the filter as part of the biomass. Table 6 10 Biofilter nitrification characterizati on after ammonium chloride supplementation Nitrogen species Mean feed NH 4 N (g/L) Mean biofilter effluent NH 4 N (g/L) NH4N 130 70 16 5 111 65 NO2N 4 5 2 1 2 5 NO3N 442 37 585 75 + 143 80 Means are provided with the standard deviation of the data sets as value error. Statistical analyses included assigning onehalf the limit of detection/quantification values to constituents with nondetected concentrations. Feed NH4N levels include background and dosed ammonium chloride. Feed NO 2 N and NO 3 N levels included background only.

PAGE 140

140 Figure 67 Biofilter nitrification characterization after ammonium chloride supplementation to the substrate (ethanol) and nutrient enhanced biofilter Metals Fe and Mn removal were observed for the nutrient enhanced, nutrient and substrateenhanced, and control biofilters under each test condition. However, relative metals removal performance was difficult to establish because of generally low and varied feed concentrat ions. Mean effluent concentrations remained below the Mn and Fe MDLs for all conditions tested including the control perf ormance (Chapter 5, Table 54). Therefore, selected nutrient enhancement test conditions were further characterized for metals removal performance. The nutrient enhanced biofilter and the substrateand nutrient enhanced biofilter (ethanol and phosphoric acid) were evaluated in parallel with the biofilter control over the following conditions:

PAGE 141

141 Simulated long term, moderate Mn load event 1 month of steady state operation with a Mn load of ~50 ng/L Mn. Simulated short term, high Mn load event 1 week of steady state operation with a Mn load of ~220 ng/L. Figure 68 shows the results of long term, moderate load test. The average removals i n the enhanced biofilters were slightly lower than that observed in the control biofilter. The control biofilter also demonstrated greater fluctuations in Mn removal. Figure 69 shows the results of the short term, high load test. The nutrient enhanced and the substrateand nutrient enhanced biofilter showed no Mn breakthrough while the control biofilter showed an average Mn breakthrough of 25 g/L. Maximum breakthrough for the biofilter control was near the secondary MCL for both loading conditions tested (40 to 48 g/L), supporting the observation of occasional colored water complaints during high load events at the JKWTP and PBSWTP. These data suggest that nutrient limitations may create unfavorable conditions for the microorganisms responsible for Mn ox idation, thus inhibiting its overall removal. Figures 6 8 and 69 summarize the results from Mn removal characterization study. Chapter 8 discusses the identified microbial communities in the pilot biofilters and their possible roles in contaminant cycling Taste and Odor MIB and geosmin removal were observed for the nutrient enhanced, nutrient and substrateenhanced, and control biofilters under each test condition. However, relative taste and odor removal performance was difficult to establish due to generally low and varied feed concentrations. Mean background geosmin and MIB feed concentrations remained below their MDLs, as described in Chapter 4, Table 45. Therefore, MIB spiking tests were conducted for additional characterization.

PAGE 142

142 MIB spiking was performed on the substrateand nutrient enhanced biofilters and substrateenhanced biofilters while in parallel operation. These tests were performed to identify whether PO4P limitations may inhibit MIB degradation in the presence of supplemental organic s ubstrates. Phosphoric acid supplementation marginally improved MIB removal performance when acetic acid, MicroC, and ethanol were the dosed substrates. The nutrient enhanced biofilter ( PO4P ) was also evaluated for MIB removal performance across both long and short term MIB loading events. This biofilter was operated in parallel with the substrateand nutrient enhanced biofilter and the biofilter control over the following conditions: Long term, moderate load 1 month of steady state operation with an M IB load of ~30 ng/L. Short term high load 1 week of steady state operation with an MIB load of ~160 ng/L. Figure 68 Mn removal performance for the nutrient enhancement strategies during simulated moderate long term loading event

PAGE 143

143 Figure 69 Mn removal performance for the nutrient enhancement strategies during simulated high short term loading event Figures 610 and 611 show the results of the long term/moderate MIB load and short term/high MIB load tests, respectively. During the simulated long term, moderate MIB loading event, mean biofilter effluent MIB concentrations remained below the OTC for all strategies tested. Paired t tests showed the difference between biofilter mean effluent values were not statistically significant between any test conditions However, the biofilter control showed significant breakthrough (~200% of OTC) during the short term, high MIB loading event, while minimal breakthrough (~ 60% of OTC) was observed in the nutrient enhanced biofilter. These data suggest that nut rient limitations may create unfavorable conditions for the microorganisms responsible MIB degradation, thus inhibiting its overall removal. Figures 610 and 611 summarize the results from MIB removal characterization study. Chapter 8Microbial Tracking d iscusses the

PAGE 144

144 identified microbial communities in the pilot biofilters and their possible roles in contaminant cycling in further detail. Figure 610. MIB removal performance for the nutrient enhancement strategies during simulated moderate long term loa ding event Figure 611. MIB removal performance for the nutrient enhancement strategies during simulated high short term loading event

PAGE 145

145 Chloramine Decay and DBPFP Testing Aqueous samples were collected from batch chloramine study bottles at the end of th e 200 hr test and analyzed for DBPs. Nutrient enhancement provided no stability performance was not improved over the 200 hour maximum detention time. As shown in Figure 613, the results from these analyses indicated that the nutrient enhanced and control biofilter effluent samples had DBPFP values that were not significantly different, and both were well below the HAA5 and TTHM MCLs. NDMA was also non detect (<2 ng/L) in both samples. Figure 612. Chloramine decay results for the nutrient enhanced biofilter and biofilter control 3 3.5 4 4.5 5 5.5 6 0 50 100 150 200 250 Chloramine Concentration (mg/L) Batch Test Elapsed Time (hr) Nutrient Enhanced Biofilter Biofilter Control Operating Conditions pH adjusted to 8.3 Hypochlorite: Ammonia = 4.5:1 Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P Data Points Represent Average of Two Batch Studies

PAGE 146

146 Figure 613. DBPFP results for the nutrient enhanced biofilter and biofilter control Pharmaceuticals and Pesticides Biofilter feed and effluent samples were collected during the nutrient enhancement tests and monitored for pharmaceuticals and pesticides. Atrazine, deethylatrazine deisopropylatrazine, hydroxyatrazine, simazine, metolachlor, and meprobamate were detected in the biofilter feed at concentrations below 1 g/L. Due to the low concentrations; the nutrient enhance ment strategies could not provide a discernable improvement to pharmaceutical and pesticide removal relative to the control under background loading conditions (Chapter 4, Table 45). Therefore, pharmaceutical and pesticide spiking was performed to differentiate the removal performance of the nutrient enhanced biofilter, the substrateand nutrient enhanced biofilter (ethanol and phosphoric acid), and biofilter control. Atrazine, carbamazepine, and caffeine were spiked for one week at low g/L levels to the pilot biofilter feed. These contaminants were monitored along with the degradative atrazine byproducts, deisopropylatrazine 0 20 40 60 80 100 120 HAA5 TTHM DBP Concentration (g/L) Biofilter Control Effluent DBPFP Nutrient Enhanced Biofilter DBPFP Total Trihalomethane MCL Halo Acetic Acid 5 MCL Operating Conditions Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P

PAGE 147

147 and deethylatrazine, and background levels of meprobamate. The results of the spiking studies (Figure 614) showed that nutrient enhancement provided negligible benefit for removal of the contaminants tested. However, each biofilter provided at least 20% removal for all tested contaminants. In addition, concentrations of deisopropylatrazine and deethylatrazine increased through all of the biofilters, possibly reflecting the biotransformation of atrazine. Figure 614. Normalized pharmaceutical and pesticide removal performance during the nutrient enhancement studies Biological Activity ATP concentrations in biofilter media were monit ored during all nutrient enhancement tests. Biofilter ATP concentrations were consistently higher in biofilters with nutrient

PAGE 148

148 supplementation. The nutrient enhanced biofilter showed 30% higher terminal (end of filter run) ATP concentrations relative to the biofilter control. The substrateand nutrient enhanced biofilter showed 70 to 250% (varied by substrate) higher terminal ATP concentrations relative to the substrateenhanced biofilter (Figure 8 15) The increase in ATP correlated with higher DOC removal s, suggesting that PO4P supplementation may enhance cell synthesis or cellular metabolic activity in PO4P limited conditions. Additional data from the ATP tests and other microbial tracking data for the substrateenhancement studies are presented in Chap ter 8 Microbial Tracking. Robustness Testing Limited robustness testing was performed on the nutrient enhanced biofilter with PO4P supplementation. The purpose of these test s w as to determine if nutrient enhancement affected biofiltration process robust ness. Robustness tests included a 48hour complete pilot shutdown and a 48hour phosphoric acid feed shutdown. Neither test created an observable hydraulic nor water treatment upset across any of the monitored parameters. Additional investigation must be c onducted to further stress the nutrient enhanced biofiltration system and characterize improvements and/or degradation of process robustness. Summary The objective of this study was to evaluate various nutrient augmentation strategies for enhancing biofiltration performance. Optimal microbial growth is dependent on a nutrient balance of carbon, ammonia(or nitrate ) nitrogen, and orthophosphatephosphorus. This balance is typically targeted at a molar ratio of 100:10:1, bioavailable C:N:P. This study evaluated nutrient enhancement by dosing

PAGE 149

149 0.020 mg/L phosphoric acid as P to satisfy the biofilter feed nutrient requirement for 0.8 mg/L of bioavailable DO C. Phosphorus supplementation decreased biofilter terminal headloss by approximately 15% relative to the control (Table 63) This improvement could translate to longer filter run times and decreased backwash frequencies. Furthermore, this strategy may ext end underdrain life by decreasing the formation of EPS, which could be responsible for underdrain cap clogging, as experienced by Arlington and other utilities (Zhu, 2010). Phosphorus supplementation also enhanced biofilter water treatment performance. The phosphorus enhanced biofilter showed higher removals of multiple constituents relative to the control biofilter, including spiked MIB, spiked Mn, and DOC (Table 65. Figures 68 through 611) The results of the nutrient enhancement biofilter tests were validated at the end of the pilot study by dosing phosphoric acid to the biofilter control. Improved filter operation through decreased headloss development was observed during the validation study. Another component of this study was to evaluate the significance of NH4 N as a limiting nutrient for biofilter operation. Testing was performed by dosing ethanol and phosphoric acid to a biofilter at concentrations that created an artificial NH4N limitation (C:N:P equaled 100:3:2). This condition was main tained for 7 weeks and resulted in decreased hydraulic performance as measured by gradual increase in 18hr terminal headloss during the pilot filter runs. The average terminal headloss for this biofilter was 9.9 feet (60 % greater than that of the control) with many runs ending before 18hours due to a headloss trigger. Ammonium chloride was then dosed to eliminate the NH4N limitation. Consequently, average terminal headloss in the enhanced biofilter column decreased to

PAGE 150

150 6 feet (56 % improvement), indicating that ammonia limitations may be as deleterious to hydraulic performance as phosphorus limitations during biofilter operation (Figure 6 2) These results suggest that both N and P limitations may increase microbial EPS production in drinking water biofilt ers. ATP measurements were performed on the nutrient enhanced biofilter media to characterize biological activity relative to the control. The data show that nutrient supplementation to achieve a nutrient balance increased biofilter media ATP concentrations by up to 30% These data suggest that the additional DOC, MIB, and Mn removal observed in the nutrient enhanced biofilter is the result of increased biological activity. This leads to the important conclusion that appropriate nutrient supplementation pro duces more biological activity but at the same time improves filter performance because the bacterial production of EPS is reduced with the removal of the nutrient limitation stress.

PAGE 151

151 CHAPTER 7 OXIDANT ENHANCEMENT STUDIES Objectives The objective of this task was to screen hydrogen peroxide supplementation as a potential strategy for improved biofilter hydraulic and water treatment performance. The intention is to augment the oxidative action and response of the biofiltration Many microorganisms express a class of enzymes, known as oxidoreductases, when exposed to hydrogen peroxide. The resulting reaction neutralizes the hydrogen peroxide, and some (e.g., peroxidase) releas e free radicals. These free radicals may then oxidize natural organic matter including recalcitrant forms thereby further reducing the concentration of DOC and trace organics. In addition, inactive microorganisms and/or their extracellular materials may also be oxidized by the hydrogen peroxide and free radicals, thereby improving biofilter hydraulics. Testing Conditions Preliminary testing of this strategy was conducted by dosing 1 mg/L of hydrogen peroxide to a pilot biofilter for approximately ten days. The test biofilter had previously been operated under the nutrient enhancem ent strategy with phosphoric acid supplementation (0.02 mg/L as PO4P). The phosphoric acid feed was suspended one biofilter run prior to oxidant enhancement testing. No other nutrients or substrates were dosed to the oxidant enhanced biofilter during test ing. The oxidant enhanced biofilter was operated in parallel with the nutrient enhanced (validation testing) biofilter and substrateand nutrient enhanced biofilter.

PAGE 152

152 Hydraulic Characterization Clean bed headloss generally remained below 1 ft for all condi tions tested. However, headloss profiles over a given run decreased dramatically immediately after hydrogen peroxide supplementation was initiated. Figure 71 presents serial headloss profiles before and after implementation of the oxidant enhancement strategy. Oxidant enhanced biofilter terminal headloss remained below three ft throughout the 10day testing period, yielding the lowest mean terminal headloss observed for all enhancement strategies tested during the entire pilot study. The oxidant enhanced b iofilter hydraulic profiles appear more consistent than the biofilter control and other strategies tested, with the lowest observed deviations between biofilter terminal headloss. Table 71 summarizes the terminal headloss data collected during the oxidant enhancement strategy testing. The data in Table 71 is a parallel comparison of the oxidant enhancement study to the nutrient enhancement validation study and historical control biofilter operation. As shown in Table 71 and Figure 71, headloss data sugg est that hydrogen peroxide supplementation may significantly improve biofilter hydraulic performance. Indeed, the oxidant enhanced biofilter mean terminal headloss remained 2.4 ft below the lowest monthly mean terminal headloss observed for the biofilter c ontrol (Chapter 4, Table 41).

PAGE 153

153 Table 7 1 Oxidant enhancement validation on biofilter hydraulic performance Oxidant enhanced biofilter terminal headloss (ft) Nutrient enhanced biofilter terminal headloss (ft) Biofilter control terminal headloss (ft) Mean Min. Max. Mean Min. Max. Mean ** Min. Max. 2.3 0.2 2.0 2.5 5.4 0.4 4.1 6.1 5.9 0.6 4.7 7.0 Target hydrogen peroxide feed in the oxidant enhanced biofilter was 1 mg/L. Target PO4P feed in the nutrient enhanced biofilter (validation study) was 0.020 mg/L as P. Biofilter control terminal headloss data from last month of steady state operation (August 2009). Includes 14 biofilter runs, means are provided with the standard deviation of the data sets as value error. ** Includes 36 biofilter runs, means are provided with the standard deviation of the data sets as value error. Figure 71 Effect of oxidant enhancement on biofilter headloss profiles

PAGE 154

154 Water Quality Characterization General The oxidant enhanced biofilter water treatment performance characterization included routine sampling and water quality a nalyses (Chapter 4, Table 43). The study period was ten days. Therefore, the data provided below should be considered preliminary. Further evaluation is required to fully characterize performance of the oxidant enhancement strategy. Hydrogen Peroxide Dosed hydrogen peroxide concentrations were verified by sampling oxidant enhanced biofilter feed daily. Measured concentrations remained on the target dose of 1 mg/L throughout testing (MDL of 0.1 mg/L). Oxidant enhanced biofilter effluent was also tested daily for hydrogen peroxide residual; however, it was never detected. Turbidity All turbidities maintained compliance with the USEPA Surface Water Treatment Rule, as greater than 95% of the effluent turbidity samples were less than 0.3 NTU. No turbidity values over 1 NTU were observed in oxidant enhanced biofilter effluent. Figure 7 2 illustrates oxidant enhanced biofilter effluent turbidity profiles for two typical filter runs.

PAGE 155

155 Fig ure 72 Oxidant enhanced biofilter turbidity profiles for typical filter runs DOC Influent and effluent oxidant enhanced biofilter DOC data were collected during the tenday oxidant enhancement study. These preliminary data showed that DOC removals during hydrogen peroxide supplementation tests were similar to those observed during nutrient enhancement conditions. Furthermore, mean oxidant enhanced biofilter DOC removals also remained below the historical means for the biofilter control. Table 72 summari zes influent and effluent DOC data for the oxidant enhance biofilter and the parallel nutrient enhanced biofilter (validation study, see Chapter 6). DOC removal performance was also characterized by comparing the oxidant enhanced biofilter effluent against that of the preceding 2 weeks of operation under nutrient enhancement conditions. During this study, hydrogen peroxide supplementation appeared to have little observed impact on DOC removal performance as compared to the nutrient enhanced study data (Figure 7 3). 0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 10/1/2009 10/1/2009 10/2/2009 10/2/2009 10/2/2009 10/2/2009 10/2/2009 10/3/2009 10/3/2009 Turbidity (NTU) Oxidant Enhanced Biofilter Operating Conditions Oxidative Enhancement Target Oxidant Dose: Hydrogen Peroxide, 1 mg/L

PAGE 156

156 Nutrients The oxidant enhanced biofilter was operated without nutrient supplementation. The mean biofilter feed nutrient concentrations during the study were 0.0325 NH4N and less than 0.010 PO4P, yielding an approximate C:N:P ratio of 100:3:0. Thus, the oxidant enhanced biofilter was operated with both NH4N and PO4P limitations relative to the available organic carbon. Table 7 2 Oxidant enhanced biofilter DOC removal performance Biofilter feed* (background C) (mg/L) Oxida nt enhanced biofilter effluent (mg/L) Nutrient enhanced biofilter (mg/L) Historic biofilter control (mg/L) Mean Min. Max Mean Min. Max Mean Min. Max Mean Min. Max 3.4 0.1 3.3 3.5 2.9 0.1 2.8 3.0 2.9 0.1 2.8 3.0 3. 2 0.1 2.8 3.4 Samples were collected of biofilter feed before and after hydrogen peroxide addition to evaluate direct mineralization of background DOC. None was observed. Target hydrogen peroxide dose for the oxidant enhanced biofilter was 1 mg/L. Target phosphoric acid dose for the nutrient enhanced biofilter was 0.020 mg/L as P. Includes ten days of steady state data (N=4). Means are provided with the standard deviation of the data sets as value error.

PAGE 157

157 Figure 73 The effect of oxidant enhancement on DOC removal performance Metals Background biofilter feed Fe and Mn levels remained below their respective MDLs (10 g/L) during the oxidant enhancement studies. Therefore, high biofilter feed Mn loading (~180 g/L) was performed to characterize metals removal performance. The oxidant enhanced biofilter successful removed all loaded Mn to nondetect levels (less than 10 g/L). Figure 7 4 illustrates the extent of Mn removal through the oxidant enhanced biofilter. Mn speciation of oxidant enhanced biofilter influ ent was not performed; therefore, it is unknown whether direct oxidation of Mn by peroxide had occurred. The parallel operating nutrient enhanced biofilter performed similarly. The tenday test duration and limited sample set (N=4) provide only a preliminary characterization of Mn removal performance. Therefore, additional testing and

PAGE 158

158 characterization are necessary. Furthermore, the mechanisms for Mn oxidation were uncharacterized. Therefore, it is unknown whether hydrogen peroxide supplementation improved removal via direct oxidation or improved conditions for microbial oxidation. Figure 74 Oxidant enhanced biofilter Mn removal performance5Taste and Odor Background biofilter feed geosmin and MIB levels remained below their respective MDLs (1.4 ng/L) during the oxidant enhancement studies. Therefore, moderate biofilter feed MIB loading (~40 g/L) was performed throughout the oxidationenhancement studies. Samples were collected of biofilter feed before and after hydrogen peroxide addition to evaluate di rect transformation of MIB. No MIB was observed in the effluent. The oxidant enhanced biofilter successfully removed all loaded MIB to nondetect levels 5 The error bars presented in all figures in this Chapter represent the standard deviation of the data set. This standard deviation accounts for operational variability (i.e. feed water conditions) and sampling/analytical error. 0 50 100 150 200 250 Manganese Concentration (ug/L) Spiked Biofilter Feed Oxidant Enhanced Biofilter Effluent <2.4 g/L MDL

PAGE 159

159 (less than 1.4 ng/L). Figure 75 illustrates the extent of MIB removal across the oxidant enhanced biof ilter. The parallel operating nutrient enhanced biofilter performed similarly. Analyses of oxidant enhanced biofilter influent samples suggested that direct peroxide oxidation of MIB did not occur (no degradation observed). The tenday test duration and limited sample set (N=4) provide only a preliminary characterization of MIB removal performance Therefore; further testing and characterization are necessary. Figure 75 Oxidant enhanced biofilter MIB removal performance Biological Activity ATP measurem ents were performed on the oxidant enhanced biofilter media to characterize biological activity relative to the (historical) control. Terminal filter media ATP concentrations for the oxidant enhanced and historical biofilter control were 1,007,000 and 1,009,000 pg/L, respectively. The data suggest that hydrogen peroxide supplementation did not decrease biofilter media ATP concentrations to levels below 0 10 20 30 40 50 60 70 MIB Concentration (ng/L) Spiked Biofilter Feed Oxidant Enhanced Biofilter Effluent Operating Conditions Oxidative Enhancement Target Oxidant Dose: Hydrogen Peroxide, 1 mg/L <1.4 ng/L MDL

PAGE 160

160 those of the biofilter control (Figure 8 16) These data suggest that biological activity is not negativel y impacted by 1 mg/L hydrogen peroxide supplementation. The drastic decrease in headloss might be due to peroxide oxidation of inactive microorganisms and EPS (direct or via activated peroxide/free radicals). Additional microbial tracking data for the oxidant enhancement studies are presented Chapter 8 Microbial Tracking Summary The objective of this task was to evaluate peroxide supplementation for augmenting the oxidative action and response of the biofiltration process. Preliminary testing of this strategy was conducted by dosing 1 mg/L of hydrogen peroxide to a pilot filter for two weeks. The test filter demonstrated 15% removal of filter feed DOC (50% greater than that observed with the control and full scale filters), and complete removal of Mn and MIB (Figure 72). The oxidant enhancement strategy also decreased terminal headloss to 2.2 feet, or 66% less than the control biofilter (Figure 71). ATP measurements were performed on the oxidant enhanced biofilter media to characterize biological act ivity relative to the control. The data showed that hydrogen peroxide supplementation did not decrease ATP concentrations in the biofilter media. These data suggested that biological activity is not negatively affected by 1 mg/L hydrogen peroxide supplementation. Additional microbial tracking data for the oxidant enhancement studies were performed as part of m icrobial tracking study (Chapter 8).

PAGE 161

161 CHAPTER 8 MICROBIAL TRACKING Objectives The objective of this task was identify preliminary correlations between microbial ecology/activity in the biofilters and hydraulic and water treatment performance. The microbial tracking task included the following analyses: SEM, biofilter media ATP quantification, HPCs, biofilm formation potential, EPS quantification, T RFLP, and clone libraries. Scanning Electron Microscopy Multiple SEM micrographs, at varying magnifications, were taken of biofilter media samples collected during each test condition in the pilot scale experiments. SEM micrographs also were taken from fullscale biofilter and biofilter control media samples. Morphological changes were observed on pilot scale biofilter biofilms as they matured through the study and were exposed to various treatments. These observations were used only as a qualitative reference for potential biological responses to the conditions of the biofilter feed (i.e., nutrient limitation/supplementation, hydrogen peroxide supplementation, substrate supplementation, etc.). The sections below provide a summary of the observations made for the full scale and pilot scale biofilters. General Observations Bacteria were observed in all biofilter media samples, colonizing large regions of the surface. In addition, a biofilm matrix , was observed in all samples at varying levels. While the biofilm matrix was not directly characterized, the images suggest that it is a combination of plant materials, sequestered minerals, and products of microbial origin (EPS). There appeared to be diverse morphologies present in the biofilm microbial

PAGE 162

162 communit y and biofilm. F ilaments of varying diameters and shapes (round and flat cross sections) were observed in the biofilm, the larger of which where likely plant materials ( Raskin 2009; Bennett Stamper, 2009) Similarly, biofilm matrices had varying textures and surface areas Biofilm qualitative appearance was the primary tracked criteria during the SEM analyses and was characterized by the observed unique characteristics and surface area between samples. Full scale Biofilter The full scale biofilter media appeared to be encapsulated by the biofilm (Figure 8 1) throughout the study. Bacteria in the full scale biofilter were abundant with diverse cell morphologies. Bacilli (rod shaped), cocci (spherical), and coccibacilli (oval), and helical bacteria were obser ved in a variety of sizes. The highest levels of filamentous materials were observed in the full scale biofilter samples. The size and shape of the filaments likely indicate microbial stalks (Kirisits, 20 10 Personal communication with Mary Jo Kirisits, Assistant Professor at University of Texas Austin, various dates 2010. ), flagella (Xi, 2009) or filamentous bacteria (Raskin 2009) Overall, the Full scale filter (FSF) samples appeared consistent with each other and the biofilter control. All full scale filter SEMs exhibited an extensive (relative to surface area) heterogeneous biofilm matrix on the GAC with observable bacterial concentrations.

PAGE 163

163 Figure 81 Fullscale biofilter GAC media SEM micrograph Biofilter Control SEM micrographs of the pilot baseline control biofilter appeared consistent throughout the study (Figures 82 and 83). The baseline control biofilter also had a large number of filaments connected to microbial cells. The size and shape of these filaments suggest stalked bacteria (Kirisits, 20 10) Bac terial stalks may be an evolutionary adaptation to life in nutrient limited conditions (Madigan et al. 2009). The elongations of the stalks are potentially a response to nutrient limiting conditions (Wagner et al. 2006) The stalk is an extension of the cell that maintains the diffusion (nutrient uptake) capabilities of the cell proper. Thus, by extending stalk length, a Bacilli Coccus Biofilm matrix Biofilm filaments

PAGE 164

164 microorganism may increase its surface area with little net increase in cell volume (and the accompany ing metabolic requirements). Bacteria with increased surfaceto volume ratios maintain an increased ability to take up nutrients and expel wastes (Madigan et al. 2009). During instances of severe nutrient limitation, t erminal stalk lengths may exceed 30 m (Gonin et al. 2000). The prevalence of stalked bacteria and their corresponding stalk lengths on biofilter media SEMs appeared to coincide with influent nutrient conditions. The control biofilter also displayed an increased abundance of biofilm matri x r elative to the nutrient enhanced biofilter. Indeed, many bacteria observed on the media were difficult to differentiate due to their embedment in the biofilm Figure 82 Biofilter control media SEM micrograph: 2 weeks of pilot operation

PAGE 165

165 Figure 83 Biofilter control media SEM micrograph: 38 weeks of pilot operation Several clusters of small white granules were observed on the control biofilter media (Figure 84). The white appearance suggests that the clusters are mineral in nature rather than biological. These granules are seen in several of the samples but could not be fully characterized. However, a trial with energy dispersive X ray spectroscopy indicated potentially high amounts of sequestered phosphorous. Some research has shown phosphorus precipitation may be mediated by bacterial extracellular enzymes (Kerdachi and Roberts 1980). These precipitates may then chemically bind to the EPS. The excessive biofi lm material and potential sequestration of phosphorus suggests that the control biofilter biofilm responded to the PO4P limitation by producing higher levels of EPS. This EPS may then serve to assist with capture and storage of the limiting nutrient. Furt her investigation is necessary to fully characterize the biofilm and observed clusters. Biofilm matrix Biofilm filaments or stalked bacteria Bacilli

PAGE 166

166 Substrate Enhancement Studies SEM micrographs of media collected from the substrateenhanced biofilters generally showed extensive biofilm, regardless of the substrate tested. The biofilm matrices appeared more pervasive than that observed on the media samples collected from the control biofilter. Furthermore, micrographs taken from successive substrateenhanced biofilter media samples showed increased biofilm prevalenc e and decreased morphological diversity. Indeed, the matrix coated most of the bacteria within the biofilm, making cell differentiation challenging. Figures 85 and 86 illustrate the extent of the biofilm in the substrate enhanced filters. Figure 84 Biofilter control media SEM micrograph: white granules White granules

PAGE 167

167 Figure 85 Substrateenhanced, nutrient limited biofilter: MicroC substrate Figure 86 Substrateenhanced, nutrient limited biofilt er media SEM micrograph: ethanol substrate

PAGE 168

168 Nutrient Enhancement Studies The nutrient enhanced biofilter media samples showed the lowest prevalence of biofilm matrice s and filaments relative to all micrographs analyzed (Figure 87). Polymeric strands are vi sible between individual cells. However, the pervasive biofilm matrix observed on other samples was not seen. Furthermore, these samples showed the highest levels of cell morphological diversity and abundance. Substrateand nutrient enhanced biofilter media samples showed increased biofilm prevalence when operated with a N H4N limitation than with ammonium chloride supplementation (Figures 8 8 and 89). Indeed, biofilter media from the NH4N limited substrate and nutrient enhanced biofilter exhibited thick, ribbonlike filaments in a majority of the micrographs. These imag es suggest that both NH4N and PO4P may play an important role in the formation and morphology of biofilm materials. Figure 87 Nutrient enhanced biofilter media SEM micrograph: phosphoric acid

PAGE 169

169 Figure 88 Substrateand nutrient enhanced biofilter media SEM micrograph: ethanol substrate, phosphoric acid, ammonium chloride Figure 89 Substrateenhanced biofilter media SEM micrograph: ethanol substrate, phosphoric acid Cocci

PAGE 170

170 Oxidation Enhancement Studies The oxidant enhanced biofilt er media samples showed microbial abundance and morphological diversity including multiple sizes of bacilli, cocci, and filamentous bacterial morphologies. Stalked bacteria also had a high prevalence in these samples. However, the biofilm matri x w as not as significant as in the substrateenhanced biofilter media samples (Figure 810). These images suggest the hydrogen peroxide supplementation may change biofilter communities over a short period. Additional studies are necessary to fully characterize the effects of oxidation enhancement on biofilm formation. Summary of SEM M icrographs The studied enhancement strategies produced observable changes in biofilter media biofilm appearance when analyzed through SEM. Images taken suggest that biofilm and cellular morphologies may be influenced by nutrient limitations and subsequent supplementation. Table 81 summarizes the observations made through the SEM analyses during the study. Figure 810. Oxidant enhanced biofilter media SEM micrograph: hydrogen per oxide

PAGE 171

171 Table 8 1 Summary of observations made during SEM analysis of biofilter media Test Condition Observed qualitative element frequency relative to the biofilter control Condition Nutrient Ratio (added chemicals ) Cellular abundance Biofilm matrix prevalence Biofilm filaments Biofilter control 100:6:0 (none) Nutrient enhanced biofilter 100:10:1 (0. 02 mg/L PO4P) High Very low Low Substrateenhanced biofilter 100:2:3 (1 mg/L C [MicroC] ) Low to high Very high Low to high Substrate and nutrient enhanced biofilter 100:3:2 (1 mg/L C [ethanol] 0.04 mg/L PO4P) High Same High Substrate and nutrient enhanced biofilter 100:14:2 (1 mg/L C [ethanol] 0.1 NH4N, 0,04 mg/L PO 4 P) High Low Same Oxidant enhanced biofilter 100: 6 :0 (1 mg/L H2O2) Very high Low High Biofilm and cellular morphologies varied with substrate tested. However, all substrate enhanced biofilter samples exhibited high levels of Biofilm matrices Substrateand Nutrient Enhanced Biofilter operated with supplemental ammonium chloride feed (non nutrient limited) Plate Count Select biofilter media samples were surveyed for HPC. Overall, these results suggest that phosphorus is responsible for increasing the number of viable cells in the filter. Figure 8 11 illustrates relative HPC between the nutrient enhanced biofilter and the biofilter control media. The data included in Figure 811 represent a single sample set that was analyzed in triplicate. These data suggest that nutrient supplementation may increase the prevalence of viable heterotrophic bacteria on biofilter media (p = 0.005, p 0.05). The HPC data support the observed increases in biofilter activity and DOC removal during the nutrient enhancement study (Table 62 and Figure 816)

PAGE 172

172 Figure 81 1 Biofilter media HPC per m L of phosphate buffered saline media samples: biofilter control and nutrient enhanced biofilter Biofilm Formation Characterization The biofilm formation capacity was assessed using crystal violet (CV) assay as described by OToole and Kolter (1998). The CV assay is best used to identify populations or communities of microorganisms that have significantly increased biofilm formation. Since the CV assay results are heavily dependent on the initial cell concentration of the inoculum, the biofilm formation capacity of the cells from the filter samples was compared among inocula with similar cell concentrations (as determined by HPC). As shown in Figure 812, biofilter media biofilm formation potential was lower

PAGE 173

173 for the nutrient enhanced biofilter relative to the control for the one sample tested A ll substrateenhanced media showed higher biofilm formation potentials relative to the control biofilter (p = 0.01) Conversely, the s ubstrate and nutrient enhanced biofilter media showed similar biofilm formation potential to the control biofilter (Figure 8 13) (p = 0.14) These results suggest that nutrient limitations may drive biofilm formation potential in a biofilter with or without substrate supplementation. The biofilm formation potential results generally corresponded to relative filter hydraulic performance between the control, nutrient enhanced, substrateenhanced, and substrateand nutrient enhanced biofilters (Figures 51, 6 1, and 6 2). Figure 812. Relative biofilm formation potential between biofilter control and nutrient enhanced biofilter6 6 The error bars presented in all figures in this Chapter represent the standard deviation of the data set. This standard deviation accounts for operational variability (i.e. feed water conditions) and sampling/analytical error. 0.00 0.05 0.10 0.15 0.20 0.25 Absorbance with CFU Normalization Nutrient Enhanced Biofilter Control Biofilter Operating Conditions Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P

PAGE 174

174 Figure 81 3 Relative biofilm formation potential between biofilter control, substrateenhanced biofilter, and substrateand nutrient enhanced biofilter EPS Q uantification EPS was quantified in glucose equivalents. EPS concentrations were found to vary significantly during the study, even in control biofilter samples. Therefore, conclusive characterization of enhancement strategies was not possible. The variation in EPS is l ikely to multiple factors including lack of spatial homogeneity in the media samples (Kirisits, 2010), limited sampling frequency, and variability in sample preparation and hold time due to operator error and shipping methods. I n general, it was found that the EPS was lowest for the nutrient enhanced biofilter, relative to the control biofilter (Figure 8 13). Subst rate enhancement appeared to increase free and bound EPS concentrations while substrateand nutrient enhancement had little effect on EPS

PAGE 175

175 pro duction as compared to the control biofilter (Figure 815) The decreased presence of EPS in the nutrient enhance d and substrateand nutrient enhanced biofilter samples corresponded with decreased headloss (Figures 51, 6 1, and 62) relative to the control and substrateenhanced biofilters, respectively. The results suggest that supplementation of substrates increased the normalized (to HPC) production and quantity of biofilter EPS in nutrient limited conditions. However, both EPS production potential and the EPS concentration were reduced to levels found on the biofilter control media when the nutrient requirements were satisfied (Figures 813, 8 15). This data supports the work of Mauclaire et al. (2004) that identified EPS as a significant source of foul ing and decreased hydraulic conductivity in biological filters. Figure 81 4 Nutrient enhancement influences on biofilter media EPS relative to the control biofilter 0 1 2 3 4 5 6 7 8 9 Free EPS Bound EPS Glucose Equivalents (mg/L) Nutrient Enhanced Biofilter Control Biofilter Operating Conditions Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P

PAGE 176

176 Figure 81 5. Substrate enhancement greatly increased EPS concentrations under nutri ent limited conditions ATP Characterization Baseline Control ATP concentrations were measured on the biofilter media at the beginning and end of selected biofilter runs. All ATP data presented in the tables below represent the mean data of the enhanced biofilters and the biofilter control during parallel operation. Generally, ATP increased in the control biofilter media by approximately 50% from the start of a filter run to the end of a filter run. However, this increase varied from 0 to 115% through the study. These variations were likely due to uncharacterized fluctuations in feed water quality. Substrate Enhancement Studies The substrateenhanced biofilter yielded higher media ATP concentrations than the control biofilter ( Table 57). However, the subs trate enhanced biofilter media yielded

PAGE 177

177 lower ATP concentrations relative to the substrate and nutrient enhanced biofilter media (Figure 8 14) (p = 0.02, p 0.05) This observation was consistent for all substrates tested. One possible explanation for this observation is that biofilter microorganisms utilize d the excess available carbon in nutrient limited conditions (Table 56) to produce additional EPS relative to the control (Figure 8 15) This transformation does not produce new microbial cells, and thu s it has a lower impact on ATP concentrations. Nutrient Enhancement Studies ATP concentrations in biofilter media were monitored during all nutrient enhancement tests. Biofilter ATP concentrations were consistently higher in biofilters with nutrient suppl ementation. The nutrient enhanced biofilter showed 30% higher terminal (end of filter run) ATP concentrations relative to the biofilter control. The substrateand nutrient enhanced biofilter showed 70 to 250% (varied by substrate) higher terminal ATP conc entrations relative to the substrateenhanced biofilter. The increase in ATP correlated with higher DOC removals, suggesting that PO4P supplementation may enhance cell synthesis in PO4P limited conditions. Increased HPC counts on the nutrient enh anced bi ofilter media (Figure 811) support this hypothesis. Figure 814 illustrates effect of nutrient limitation on biofilter media ATP observed during parallel operation of (ethanol) substrateenhanced and (ethanol) substrateand nutrient enhanced biofiltration (single sample set). A 3week characterization (N=4) of the effects nutrient enhancement on biofilter media ATP concentrations is provided in Figure 815.

PAGE 178

178 Figure 81 6 Effects of nutrient supplementation on substrateenhanced biofilter media ATP concentrations Figure 817. Nutrient enhancement and nutrient and substrateenhancement ATP characterization Oxidant Enhancement Studies The pilot biofilter control was not operated in parallel to the oxidant enhanced biofilter. However, hydrogen perox ide supplementation did not decrease ATP 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 1,600,000 1,800,000 Control Biofilter Substrate Enhanced Biofilter Substrate and Nutrient Enhanced Biofilter Biofilter Media ATP Concentration ( pg/mL) Start of Filter Run End of Filter Run Operating Conditions Substrate and Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.07 mg/L as P Target Substrate Dose: Various, 1 mg/L as C Substrate Enhancement Target Substrate Dose: Various, 1 mg/L as C 0 500,000 1,000,000 1,500,000 2,000,000 2,500,000 3,000,000 Biofilter Control Nutrient Enhanced Biofilter Substrate and Nutrient Enhanced Biofilter Media ATP Concentrations (pg/ml of media) Start of Filter Run End of Filter Run Operating Conditions Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.02 mg/L as P Substrate and Nutrient Enhancement Target Nutrient Dose: Phosphoric Acid, 0.07 mg/L as P Target Substrate Dose: Ethanol, 1 mg/L as C

PAGE 179

179 concentrations to levels below the historical levels observed in the biofilter control. These results support the DOC removal data presented in Chapter 7, suggesting that 2 weeks of steady state hydrogen peroxide su pplementation may not negatively impact biological activity in a biofilter. As stated in Chapter 2, some microorganisms are capable of expressing catalase and other oxidoreductase enzymes to reduce peroxides to innocuous water and oxygen. Figure 816 illus trates ATP concentrations in the oxidant enhanced biofilter media relative to those collected during the previous month of biofilter control operation. Figure 81 8 Nutrient enhancement and nutrient and substrateenhancement ATP characterization T RFLP T RFLP was performed on select biofilter media samples to characterize relative shifts in community diversity/similarity. The ShannonWeaver index (SWI) and Srenson index (SI) were used to characterize sample diversity and sample set 0 200,000 400,000 600,000 800,000 1,000,000 1,200,000 1,400,000 Biofilter Control Oxidant Enhanced Biofilter Biofilter Media ATP Concentration ( pg/mL) Start of Filter Run End of Filter Run

PAGE 180

180 similarity, respectively. The SWI accounts for both species richness and evenness, with higher values suggesting higher community diversity. A value of 0 for the SI indicates that there are no OTUs in common between two samples (no similarity), while a value of 1 for the SI i ndicates that two samples have identical OTUs. The following bullets provide a summary on the test conditions evaluated: Test 1: Media from the nutrient enhancement validation study was compared against media collected from the same biofilter two week s p rior when it was operated under control conditions. Test 2: Substrateand nutrient enhanced biofilter (ethanol and phosphoric acid) before and after ammonium chloride supplementation. Test 3: Biofilter control and the oxidant enhanced biofilter (hydrogen peroxide) Test 4: Biofilter control 36 weeks of operation and 38 weeks of operation. Test 5: Biofilter control and the full scale biofilter Test 6: Substrateenhanced biofilter (ethanol) and the substrateand nutrient enhanced biofilter (ethanol and phosphoric acid) Table 82 summarizes the results from this analysis. Diversity did vary across each test and during long term operation of the control (SWI = 3.30 to 4.32 for all samples, 3.30 to 3.75 for biofilter control). According to the SWI, the highes t community diversity was present in the full scale filters. This increase in diversity might due to the introduction and proliferation of microbial populations in the openair filters that would not occur in the pilot system (e.g. blue green algae). Higher diversity was observed for the nutrient enhanced biofi lter and the biofilter control. The biofilter control media sample collected after 36 weeks of operation still showed 68% similarity in diversity relative to the fullscale biofilters. Decreased sim ilarit ies were observed when comparing the biofilter control against substrateenhanced and nutrient -

PAGE 181

181 enhanced biofilters. Furthermore, nutrient enhancement appeared to have the greatest effect on shifting microbial communities (i.e., lowest similarity with biofilter control) Table 8 2 Diversity and s imilarity i ndices for various biofilter media samples Test number Sample set A A diversity index (SWI) Sample set B B diversity index (SWI) A and B similarity index (SI) 1 Nutrient enhancement validation 3.40 Biofilter control (week 38) 3.30 0.46 2 Substrateand nutrient enhanced biofilter w/ammonium 4.08 Substrate and nutrient enhanced biofilter w/o ammonium 4.11 0.83 3 Biofilter control (week 38) 3.30 Oxidant enhanced biofilter 3.66 0.67 4 Biofilter control (week 3) 3.74 Biofilter control (week 36) 3.85 0.74 5 Biofilter control (week 36) 3.74 FSF 4.32 0.68 6 Substrateenhanced biofilter 3.37 Substrate and nutrient enhanced biofilter 3.97 0.61 Diversity index of each sample was determined by the SWI. Similarity between samples was determined by the SI using 0.5bp comparisons. Microbial community similarities also varied over test duration. The biofilter control showed 74% similarity in microbial community in weeks 3 and 36. Clone Libraries Clone libraries were performed on media collected from the control, nutrient enhanced validation, oxidant enhanced, and the C:N:P balanced substrateand nutrient enhanced biofilters. Specific hydraulic and water treatment observations from each biofilter during the time of sample collection include Biofilter control: Biofilter was operated under full scale/control conditions. Media sample was collected immediately before nutrient enhancement validation testing was performed. Water treatment and hydraulic performance was representative to data collected throughout the study. The observed C:N:P ratio at the time of sample collection was 100:7:0. Nutrient enhancement biofilter: Biofilter was operated with phosphoric acid supplementation. Media was collected af ter 1 month of steady state operation. Water treatment performance was representative data collected throughout the study. However, hydraulic performance was less than

PAGE 182

182 recorded means (terminal headloss of ~6 ft). The observed C:N:P ratio at the time of sam ple collection was 100:7:1. Substrate and nutrient enhanced biofilter: Biofilter was operated with ethanol, phosphoric acid, and ammonium chloride supplementation. Media was collected after 10 days of steady state operation. Water treatment and hydraulic performance was representative to data collected throughout the study. The observed C:N:P ratio at the time of sample collection was 100:14:2. Oxidant enhanced biofilter: Biofilter was operated with hydrogen peroxide supplementation. Media was collected af ter 10 days of steady state operation. Water treatment and hydraulic performance was representative to data collected throughout the study. The observed C:N:P ratio at the time of sample collection was 100:6:0. The clone libraries are provided in Tables 82, 8 3, 8 4, and 85. Identified genera were compared against the literature to identify whether they contained potential MIB, atrazine, or Mncycling bacteria. In addition, the clone libraries were exam in ed for organisms known to produce EPS under nutrient limited conditions. The clone library data presented in Tables 83 through 86 identifies communities that may support the potential functionalities described. However, testing did not seek to identify the genes required to express these functions Therefore, this discussion is speculative and is presented to solely to support the hydraulic and water treatment performance observed in the biofilters. The clone libraries from each biofilter included genera that contained MIB, atrazine, and Mn cycling bacteria (Ishida and Miyaji, 1992; Egashira et al. 1992; Stuck i et al. 199 5 ; Lauderdale, 2004; Kohl et al. 2006) However, all genera contain multiple strains that may or may not express reported functionality The substrateand nutrient enhanced biof ilter held the highest number Burkholderia clones. This genus contains known Mn oxidizing bacteria, supporting the water quality treatment observed in this biofilter.

PAGE 183

183 Table 83 Biofilter control clone library Phylum Class Order Family Genus No. of clones Potential functionality Atrazine degradation MIB and geosmin degradation Mn oxidation Proteobacteria AlphaProteobacteria Rhodospirillales Rhodospirillaceae Azospirillum 2 Rhizobiales Rhizobiaceae Agrobacterium 3 O Rhizobium 2 O Bradyrhizobiaceae Bradyrhizobium 11 Afipia 1 Mesorhizobium Mesorhizobium 1 Ochrobactrum Ochrobactrum 1 Sphingomonadales Sphingomonadaceae Novosphingobium 1 Beta Proteobacteria Rhodocyclales Rhodocyclaceae Sterolibacterium 1 Burkholderiales Burkholderiaceae Burkholderia 1 O Ideonella 1 Nitrosomonadales Nitrosomonadaceae Nitrospira 5 Gamma Proteobacteria Thiotrichales Thiotrichaceae Beggiatoa 2 Legionellales Legionellaceae Legionella longbeachae 5 Pseudomonadales Pseudomonadaceae Pseudomonas 3 O O Moraxellaceae Acinetobacter 3 Xanthomonadales Xanthomonadaceae Dokdonella 1 Chromatiales Ectothiorhodospiraceae Halorhodospira halophila 1 Methylococcales Methylococcaceae Methylocaldum 1 Delta Proteobacteria Desulfobacterales Desulfobacteraceae Desulfobacter 3 Desulfovibrionales Desulfovibrionaceae Desulfovibrio 4 Desulfuromonadales Geobacteraceae Geobacter 1 Epsilon Proteobacteria Nautiliales Nautiliaceae Nautilia 5 Acidobacteria Actinobacterida e Acidobacteriales Acidobacteriaceae Acidobacteria 6 Actinomycetales Micrococcacea Arthrobacter 1 O O Streptomycetaceae Streptomyces 1 O Nocardiaceae Rhodococcus 1 O O Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Flavobacterium 1 O Sphingobacteria Sphingobacteriales Chitinophagaceae Lacibacter 1 Firmicutes Bacilli Bacillales Staphylococcaceae Staphylococcus 1

PAGE 184

184 Table 84 Substrate and nutrient enhanced biofilter clone library (carbon limited) Phylum Class Order Family Genus No. of clones Potential functionality Atrazine degradation MIB and geosmin degradation Mn oxidation Proteobacteria AlphaProteobacteria Rhizobiales Rhizobiaceae Agrobacterium 1 O Bradyrhizobiaceae Bradyrhizobium 1 Rhodospirillales Acetobacteraceae Acetobacter 3 Rhodobacterales Rhodobacteraceae Rhodovulum 1 Beta Proteobacteria Burkholderiales Comamonadaceae Simplicispira 1 Variovorax 1 unclassified Sulfurovum 1 Burkholderiaceae Burkholderia 17 O Cupriavidus 2 Derxia gummosa Derxia 2 Rhodocyclales Rhodocyclacease Zoogloea 3 Neisseriales Neisseriaceae Aquaspirillum 2 unclassified Betaproteobacteria Ferrovum Ferrovum myxofaciens 3 Rhizobiales unclassified Methylocystis 1 Nitrosomonadaceae Nitrospiraceae Nitrosospira 1 Gamma Proteobacteria Chromatiales Ectothiorhodospiraceae Halorhodospira 2 Legionellales Legionellaceae Legionella 2 Pseudomonadales Pseudomonadaceae Pseudomonas 5 O O Acidithiobacillales Acidithiobacillaceae Acidithiobacillus 3 Thiotrichales Thiotrichaceae Beggiatoa 1 Xanthomonadales Xanthomonadaceae Rhodanobacter 1 Delta proteobacteria Desulfovibrionales Desulfovibrionales Desulfovibrio 1 Desulfuromonadales Geobacteraceae Geobacter 3 Desulfovibrionales Desulfovibrionales Desulfonatronum 2 Actinobacteria Actinobacteria Rhizobiales Bifidobacteriaceae Bifidobacterium 1 Actinobacteridae Actinomycetales Mycobacteriaceae Mycobacterium 1 O Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Flavobacterium 1 Firmicutes Bacilli Bacillales Bacillaceae Bacillus 1 O Clostridia Clostridiales Clostridia Clostridium 4

PAGE 185

185 Table 85 Oxidant enhanced biofilter clone library Phylum Class Order Family Genus No. of clones Potential functionality Atrazine degradation MIB and geosmin degradation Mn oxidation Proteobacteria AlphaProteobacteria Rhodospirillales Acetobacteraceae Acetobacter 3 Rhodospirillaceae Azospirillum 1 Rhizobiales Hyphomicrobiaceae Hyphomicrobium 1 O Methylobacteriaceae Methylobacerium 2 O Beijerinckiaceae Methylocapsa 1 Methylocella 1 Rhizobiaceae Rhizobium 1 O Bradyrhizobiaceae Rhodopseudomonas 1 O Bradyrhizobium 18 Caulobacterales Caulobacteraceae Brevundimonas 1 Sphingomonadales Sphingomonadaceae Sphingobium 3 Beta -Proteobacteria Rhodocyclales Rhodocyclaceae Zoogloea 1 Sterolibacterium 2 unclassified Betaproteobacteria Thiobacter 2 unclassified Betaproteobacteria Ferrovum 2 Gamma Proteobacteria Pseudomonadales Moraxellaceae Acinetobacter 2 Aeromonadales Aeromonadaceae Aeromonas caviae 1 Thiotrichales Thiotrichaceae Beggiatoa 1 Enterobacteriales Enterbacteriaceae Enterobacter 1 O Chromatiales Halothiobacillaceae Halothiobacillus 1 Pseudomonadaceae Pseudomonas 1 O O Enterobacteriaceae Enterobacteriaceae Serratia 2 Pseudomonadales Thiotrichaceae Thiothrix 1 Acidobacteria Actinobacteria Acidobacteriales Acidobacteriaceae Acidobacteria 1 Streptosporangiaceae Streptosporangineae 1 Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Flavisolibacter 2 Sphingobacteria Sphingobacteriales Sphingobacterium 3 Firmicutes Bacilli Bacillales Staphylococcaceae Staphylococcus 1

PAGE 186

186 Table 86 Ammonia limited, nutrient enhanced biofilter clone library Phylum Class Order Family Genus No. of clones P otential functionality Atrazine degradation MIB and geos min degradation Mn oxidation Proteobacteria AlphaProteobacteria Rhodospirillales Acetobacteraceae Acidiphilium 1 Sphingobacteriales Sphingomonadaceae Sphingomonas 4 Rhizobiales Bradyrhizobiaceae Afipia 2 Agromonas 1 Bradyrhizobium 19 Methylobacteriaceae Microbacterium 2 O Nordella Nordella 2 unclassified Rhizobiales 2 Rhodobacterales Rhodobacteraceae Rhodobacter 3 Beta Proteobacteria Burkholderiales Alcaligenaceae Achromobacter 1 Burkholderiaceae Burkholderia 2 O Rhodocyclales Rhodocyclaceae Azoarcus 1 Sterolibacterium 1 Neisseriales Neisseriaceae Chitiniphilus shinanonensis 1 Nitrosomonadales Nitrosomonadaceae Nitrospira 2 unclassified B etaproteobacteria Thiobacter 2 Gamma Proteobacteria Aeromonadales Aeromonadaceae Aeromonas 2 Pseudomonadales Moraxellaceae Acinetobacter 1 Pseudomonadaceae Pseudomonas 7 O O Enterobacteriaceae Enterobacteriaceae Serratia 1 Xanthomonadales Xanthomonadaceae Xanthomonas 1 Delta Proteobacteria Desulfovibrionales Desulfovibrionaceae Desulfovibrio 3 Desulfobacterales Desulfobacteraceae Desulfobacter 1 Geobacteraceae Geobacter 1 Lacibacter 1 Bdellovibrionales Bacteriovoracaceae Bacteriovorax 1 Actinobacteria Actinobacteridae Actinomycetales Pseudonocardiaceae Actinobispor a 1 Bacteroidetes Flavobacteria Flavobacteriales Flavobacteriaceae Flavobacterium 3 O Sphingobacteria Sphingobacteriales Chitinophagaceae Chitinophaga 1 Sphingobacteriales Chitinophagaceae Niastella 1 Sphingobacteriales Sinorhizobium 2 Sphingobacteria Sphingobacteriales Cyclobacteriaceae Aquiflexum 1 Firmicutes Bacilli Bacillales Bacillaceae Bacillus 1 O Clostridia Clostridiales Clostridia Clostridium 1

PAGE 187

187 Perhaps most significant observation in the clone libraries for these samples was the drastic reduction in the Bradyrhizobium population when the stoichiometric C:N:P ratio was implemented. Bradyrhizobium constituted only 1.5% of the clones in the stoichiometric C:N:P sample, but constituted 31% in the oxidant enhanced biofilter, 27% in the nutrient enhanced biofilter (with NH4N limitation), and 15 % in the biofilter control. Bradyrhizobium has been shown to increase EPS production under N limitation (Quelas et al. 2006) and related rhizobia have been shown to increase EPS production under PO4P limitation (Skorupska et al. 200 6). Figure 817 shows relative population sizes of Bradyrhizobium (blue) and Burkholderia (yellow) in clone libraries obtained from the substrateand nutrient enhanced biofilter (100:14:2 C:N:P) and the biofilter control (100:6:0) In addition, using the calculated length of the OTU corresponding to Bradyrhizobium T RFLP showed that the Bradyrhizobium OTU maintained between 10 and 15% abundance (Figure 818) in the biofilter control media throughout the study; however, during the nutrient enhancement validation study, Bradyrhizobium abundance decreased to approximately 1% in this biofilter. Figure 8 18 illustrates the observed shift in Bradyrhizobium in the biofilter with phosphoric acid supplementation. This shift correlated with improved filter hydraulic performance ( The prevalence of Bradyrhizobium in the oxidant enhanced biofilter is potentially due to a completive advantage provided by EPS production. Studies have shown EPS provides microbial resistance to oxidative stress ( Kir ly, 1998). Nutrient enhanced biofilter media samples with the lowest Bradyrhizobium abundance also maintained the lowest filter run terminal headloss.

PAGE 188

188 Figure 81 9 Comparison of biofilter media clone libraries under phosphorus limited and carbonlimited conditions Figure 820. Relative abundance (via T RFLP) of Bradyrhizobium befor e and after phosphoric acid supplementation (0.02 mg/L as P) 0 2 4 6 8 10 12 14 16 0 50 100 150 200 250 Relative Abundance of Bradyrhizobium (%) Days of Continuous Biofilter Operation Nutrient Enhancement Conditions Baseline Control Conditions 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Biofilter Control (PO4 P Limited) Substrate and Nutrient Enhanced (Carbon Limited) Relative Abundance Clostridium sp Bacillus Flavobacterium Mycobacterium Bifidobacterium Desulfonatronum Staphylococcus Lacibacter cauensis Rhodococcus Streptomyces Arthrobacter Acidobacteria Nautilia profundicola Geobacter sp. Desulfovibrio gigas Desulfobacter Methylocaldum Halorhodospira halophila Dokdonella Acinetobacter sp Rhodanobacter sp Beggiatoa sp Acidithiobacillus Pseudomonas Legionella longbeachae Beggiatoa sp. Nitrosospira sp Methylocystis sp Ferrovum myxofaciens Aquaspirillum Zoogloea sp. Derxia gummosa Cupriavidus Burkholderia Sulfurovum sp Variovorax sp Simplicispira Rhodovulum Acetobacter Sterolibacterium Novosphingobium sp. Ochrobactrum Mesorhizobium Bradyrhizobium Rhizobium sp Agrobacterium Azospirillum sp. 15% 1% 25% 1% C:N:P 100:6:0 C:N:P 100: 14 : 2 C:N:P 100: 6 : 0 C:N:P 100: 6 : 0 C:N:P 100: 6 : 1

PAGE 189

189 CHAPTER 9 FULL SCALE PROCESS INTEGR ATION ASSESSMENT AND ECONOMIC EVALUATION Objectives During pilot testing, the substrate enhancement, nutrient enhancement, and enzyme enhancement strateg ies all demonstrated some water treatment benefits. The enzyme and nutrient enhancement strategies also provided significant hydraulic benefits, making them particularly promising for full scale implementation. The oxidant enhancement strategy was evaluated only briefly, and additional testing is required to better understand process mechanisms and characterize long term performance. In addition, dosage sensitivity and optimization must still be performed because a 1 mg/L continuous dose of hydrogen peroxide may be cost prohibitive for many utilities (~$12/MG treated at a bulk cost of $1.5/lb). The chemical cost of oxidant enhancement strategy implementation would exceed the total cost for backwashing (Table 95). Therefore, the full scale process integration assessment and cost evaluation was conducted solely on the nutrient enhanced biofilter strategy. Design requirements and cost estimates were developed around the JKWTP and PBSWTP process configurations, capaci ties, and requirements (phosphorus limitation only). All costs are presented as $/MG produced. Process Integration Conceptual Design and Implementation Nutrient enhanced biofiltration can be implemented at the Arlington facilities with the installation of a phosphoric acid feed system that would satisfy the PO4P limitation through continuous feed at 20 g/L as P. The phosphoric feed system would include a peristaltic pump skid housing two pumps (1 duty, 1 standby) with variable frequency

PAGE 190

190 drives (VFDs) and supervisory control and data acquisition (SCADA) integration. The pumps would have flow pacing capabilities with automatic control from the raw water flow meters. The system would be located in each facilitys chemical room. The assumed phosphoric acid fee d rate would be 0.044 gal/MG treated (assuming 85% acid purity). At peak capacity (PBSWTP 72 mgd, JKWTP 97.5 mgd,), the feed rate would be 3.2 to 4.3 gallons per day of 85% phosphoric acid. The acid would be pumped through flexible tubing and injected into a carrier water line containing finished water. The carrier water system would convey approximately 10 gpm of carrier water (finished water + phosphoric acid) to a nozzled manifold located within a common filter influent channel. The manifold would be arraigned perpendicular to flow with chemical injected countercurrent to flow to improve chemical dispersement. Chemical storage would include two standard, self contained, 250gallon totes (1 duty, 1 standby) provided by the phosphoric acid supplier. Eac h tote would be capable of providing over 50 days of storage at peak flow. Empty totes would be replaced upon each chemical delivery. Approximately 75 to 150 ft of carrier piping would be required to convey phosphoric acid from the chemical room to the bio filter feed channel. Figure 91 illustrates a conceptual layout for the phosphoric acid feed system. Figure 91 Conceptual integration schematic for nutrient enhancement From Sedimentation Carrier Water Intermediate Ozone Contactor BW BWW Ozone Peristaltic Phosphoric Acid Feed Pumps 250 Gallon Phosphoric Acid Totes Filtration To Disinfection

PAGE 191

191 Process Monitoring The implementation of biofilter nutrient enhancement or any eng ineered biofiltration strategy requires not only the intentional design on operation of the biofilters as a biological system, but also intentional monitoring of the biological activity and health. Nutrient enhancement shows significant promise as a biofil tration operational strategy; however, not all water treatment plants should be treated equally. Monitoring is critical, as process optimization is an iterative approach that is unique to each water source and treatment train. Many analytical tools (including those discussed in this dissertation) are currently available that provide utilities with cost efficient and easily accessible onsite monitoring. Headloss monitoring and ATP analyses are two examples of real time tools that may indicate successful impl ementation of biofiltration enhancement or provide an early warning for system upset. It is recommended that utilities consider these and other methods for process monitoring to promote optimal biofiltration performance. Additional Considerations Biofilter nutrient enhancement may affect not only biofilter hydraulics and filtered water quality, but also distribution system stability. The results presented in Chapter 6 suggest that nutrient enhancement will not increase chloramine decay or DBP formation. Bio filter nutrient enhancement also provided significant reduction in DOC relative to the biofilter control, likely limiting regrowth potential. However, long term distribution system impacts of nutrient enhanced biofiltration remain unknown. Previous work ev aluating nutrient levels in finished waters supports both positive and negative impacts on distribution system regrowth and disinfectant stability (Chapter 2). Biofilter nutrient breakthrough may increase biological activity in distribution systems.

PAGE 192

192 Conver sely, a substratelimited condition in the distribution system may reduce EPS production and biofilm formation thus increasing microbial susceptibility to residual disinfectants and improving long term stability. Regardless, it is recommended that concerned utilities carefully monitor nutrient addition and breakthrough to prevent undesired concentrations in the distribution system. Although dosed concentrations are likely to be less than 0.1 mg/L as N or P, low level nutrient breakthrough may be of particul ar concern for areas held to strict numeric nutrient criteria limits on wastewater discharge permits. Cost Assessment Assumptions The assumptions used for the cost assessment are included in Table 91. Capital C ost The estimated capital cost for a phosphoric acid dosing system is $25,000, including $18,000 for the peristaltic pump skid and $7,000 for installation and the associated piping. The amortized production cost for the equipment is $0.07/MG. Table 9 1. Cost assessment assumptions Criteria Units Value Average annual facility production mgd 60 Discount Rate % 3 Estimated Contingency on Power and Chemical Costs % 25 Power cost $/kWh 0.1214 Equipment life Yr 20 Phosphoric acid feed requirement* mg/L as P 0.02 Phosphoric acid cost $/lb 0.72 85 % bulk solution Nutrient limitation is assumed for PBSWTP based on similar water quality to JKWTP P/C mass ratio of 0.026

PAGE 193

193 Operation and C hemical C ost The peristaltic pumps are operated at a maximum of 220 watts. Assuming a continuous maximum pow er draw, the normalized production cost for power is approximately $0.01/MG treated. The estimated chemical cost is $0.44/MG treated. The total operation and chemical production cost is $0.45/MG treated, or $0.56 with a 25% contingency. Total E stimated C os t for I mplementation The estimated production cost for nutrient enhanced biofiltration at the JKWTP and PBSWTP is $0.63/MG treated. Potential Net Costs a nd Cost Savings The increased hydraulic performance observed with the nutrient enhanced biofilters may lead to real cost savings during full scale implementation. The basis for this estimate is an assumed increase in biofilter run time. This increase in filter run time translates to a corresponding decrease in backwash frequency and all costs associated wit h backwashing (e.g., power and backwash wastewater retreatment). All backwash wastewater must be repumped to the head of the plant and retreated. Treatment costs vary between the JKWTP and PBSWTP. Therefore, the potential savings for nutrient enhancement are different for each plant. Tables 92, 9 3, and 94 shown below provide an estimate of the cost of filter backwashing developed for the JKWTP and PBSWTP. These estimates include the chemical and pumping costs for the recycled backwash water. Net cost and saving estimates were developed for biofilter runtime improvements of 5, 10, 15, and 20%

PAGE 194

194 Table 9 2. Backwash water production estimates Parameter Treatment plant PBSWTP JKWTP Total flow (MG) 23,470 12,858 Total # of filters washed 3,627 3,310 Filter backwash water per filter (MG) 0.363 0.233 Total backwash water produced and returned (MG) 1335 771 The flows (in MG) and filter numbers are based on information obtained from the annual summary report from October 2008 to September 2009 Low production rates at the JKWTP were attributed to plant shutdowns for an expansion project that occurred during the study. Table 9 3. Chemical costs ($/MG) to retreat backwash wastewater Parameter Treatment plant PBSWTP JKWTP Liquid oxygen (LOX) $15.92 $11.20 Alum $40.08 $23.08 Polymer $5.46 $5.05 Ozone generation $13.92 $22.12 Total cost $75.38 $61.45 The chemical cost numbers are based on information obtained from the annual summary report from October 2008 to September 2009 Table 9 4 Pumping costs to recycle backwash wastewater Parameter Treatment plant PBSWTP JKWTP 3 Backwash pump ($/MG) 43.68 Recycle pump ($/MG) 36.00 36.00 Total cost ($/MG) 79.68 36.00 The pumping costs are based on information obtained from the annual summary report from October 2008 to September 2009 Power rates are based on Arlingtons 2010 rate of $0.1214/kWh As shown in Table 95, cost savings may be realized by implementing a nutrient enhancement strategy for the biofiltration process if biofilter runs are extended by greater than 10% (~11 % reduction in backwash frequency required for savings at JKWTP). Arlington may achieve a net cost savings of approximately over $17,100 per year if the PBSWTP and JKWTP full scale biofilters respond to phosphoric acid

PAGE 195

195 supplementation similar to the pilot system (15% backwash frequency reduction). These saving would provide an implementation payback of less than two years. In addition, nutrient enhancement may also provide additional cost savings through extending the life of biofilter media (via decreased attrition) and mechanical equipment (underdrains/caps, pumps, etc). Table 9 5. Nutrient enhancement implementation net costs or savings Parameter Treatment plant PBSWTP JKWTP Backwash Cost Summary Chemical cost ($/MG) 75.38 61.45 Pumping cost ($/MG) 79.68 36.00 Total cost ($/MG) 155.06 97.45 Filter backwash flow (MG) 0.37 0.23 Total cost/filter backwash ($/BW) 57.37 22.41 Total backwash cost/year ($/yr) 208,088 74,189 Total backwash cost/MG treated ($/MG) 8.87 5.77 Projected savings from 5 % extended filter runs Backwash savings ($/MG) 0.44 0.29 Net savings from nutrient enhancement ($/MG) 0.19 (Cost) 0.34 (Cost) Potential annual savings at previous annual production rates ($) 3468 (Cost) 6205 (Cost) Projected savings from 10 % extended filter runs Backwash savings ($/MG) 0.89 0.56 Net savings from nutrient enhancement ($/MG) 0.26 0.07 (Cost) Potential annual savings at previous annual production rates ($) 4,745 1,278(Cost) Projected savings from 15 % extended filter runs Backwash savings ($/MG) 1.33 0.87 Net savings from nutrient enhancement ($/MG) 0.70 0.24 Potential annual savings at previous annual production rates ($) 12,775 4,380 Projected savings from 20 % extended filter runs Backwash savings ($/MG) 1.62 1.12 Net savings from nutrient enhancement ($/MG) 1.00 0.50 Potential annual savings at previous annual production rates ($) 18,068 8,943 Net savings incorporate estimated capital and operating costs for nutrient enhancement. All nutrient enhancement costs include 25 % contingency on power and chemical costs.

PAGE 196

196 CHAPTER 10 SUMMARY AND CONCLUSIONS New regulatory and technological developments are driving more utilities to consider the use of biofiltration to treat their drinking water. These developments include (1) the concern over the formation of disinfection by products (DBPs); (2) the emergence of ozone for taste, odor, and color control; (3) the increased awareness of how biological act ivity in distribution systems contributes to disinfectant demand, aesthetic upsets, and corrosion; and (4) the push for green technologies. Currently, drinking water biofiltration is largely operated as a passive process. Particle/turbidity removal and headloss drive the design and operation of biofiltration as they would conventional filtration. Thus, biofilter design parameters are typically limited to media configuration, backwash strategy, and loading rate. The biological removal of dissolved organic and inorganic contaminants is an anticipated benefit of biofiltration. However, common practice does not seek to enhance the bioactivity responsible for those mechanisms. Indeed, in an effort to improve filter productivity and minimize headloss, many utilities employ chlorinated or chloraminated water to the filter backwash or feed. However, this practice is to the detriment of biological activity and may be ineffective at removing a primary foulant of biofilters extracellular polymeric substances (EPS) pro duced by bacteria resident in the filter biofilm. EPS can occupy as much as 1,000 times the filter media void space as bacteria, playing a more significant role in both fouling and headloss (Mauclaire, 2004) An alternative approach is to move the practic e of biofiltration from a passive process designed and operated around conventional filtration objectives to an intentionally operated biological system, i.e. engineered biofiltration. Engineered

PAGE 197

197 biofiltration targets multiple water quality objectives while maintaining hydraulic performance. These benefits can be achieved simultaneously by providing specific conditions that promote the improved biological activity. Engineered biofiltration shifts the industry accepted paradigm so that the design and operation of biofilters should be driven not only by filtration but also by biological treatment objectives. Problem Statement and Hypothesis Drinking water biofilters at surface water treatment facilities commonly experience shortened run times; underdrain clogging, T&O breakthrough, and Mn breakthrough (Hunt, 2009; Zhu, 2010 Evaluate potential biof ilter enhancement strategies comprised of dosing low levels of common drinking water treatment chemicals at a feed point just upstream of a biofiltration process. These chemicals were added to provide substrate, nutrient, and/or oxidant optimization of the biofilter process influent. ). A pilot study was conducted at the City to evaluate methods for restoring and enhancing the performance of the biofiltration process. This evaluation entailed both a characterization and evaluation of biological activity in the filters and an examination of potential enhancement strategies. The hypothesis of this work is both water treatment and hydraulic performance of a biofilter can be improved by modifying influent conditions for enh anced biological activity. The purpose of this research was to identify strategies to enhance the biological activity in a biofilter without compromising productivity or particulate removal performance. Specific objectives included: Investigate of biological drinking water treatment process fundamentals (e.g., microbial ecology, bacterial metabolism, and contaminant removal mechanisms) to understand how DOC, MIB, geosmin, and Mn can be removed effectively in a single treatment step Zhu, I. Personal communication with Ivan Zhu, Senior Engineer at F.B. Leopold, ITT, on May 15, 2010.

PAGE 198

198 Biological clogging (filter headloss) can be minimized The ultimate goal of this work is to shift an industry accepted paradigm so that the design and operation of biofilters are driven not only by filtration but also by biological treatment objectives. Objectives The research included 10 months of biofiltration enhancement pilot scale testing at the JKWTP to evaluate methods for restoring and enhancing the performance of the Citys biofiltration process. This evaluation entailed both a characterization and evaluation of biological activity in the biofilters and an examination of potential enhancement strategies. The strategies tested were selected based on previously published literature and industry experience. The follow ing studies were performed to meet research objectives: Characterize the baseline performance of the JKWTP operating under existing conditions. This included an assessment of the systems ability to meet the Citys current treatment objectives; Evaluate pr imary substrate augmentation enhancement strategies; Evaluate nutrient augmentation enhancement strategies; Evaluate peroxide supplementation for augmenting the oxidative action and response of the biofiltration process; Identify and track the microbial co mmunities within the biofilters; Develop conceptual full scale design and operating parameters for the recommended modified biofiltration process and estimate the associated capital and production costs.

PAGE 199

199 Results Baseline Biofiltration Characterization The objective of this task was to characterize baseline water treatment performance, confirming similitude between pilot and full scale treatment and to provide control data for parallel tests occurring with other pilot filters. Similitude was characterized by operating a control biofilter under full scale operational conditions (i.e., no enhancement strategies were used for the control biofilter) from March 5, 2009 through October 1, 2009. These tests provided approximately seven months of data that confirmed and further elucidated the treatment capabilities of the existing process at the JKWTP. Monitored water quality parameters included DOC, total Mn, ammonianitrogen (NH4N), orthophosphatephosphorus (PO4P), MIB, and a suite of over 150 pharmaceuticals and pesticides. These water quality data demonstrated treatment performance similitude between the pilot and full scale filters (Tables 45, 4 6, 4 7, 4 8, 4 9). Measured water quality values between the control biofilter and the full scale biofilter were wi thin the standard deviations of their respective data sets. The study confirmed potential treatment deficiencies under full scale biofilter operating conditions. MIB breakthroughs of 11 ng/L and 25 ng/L were observed during simulated moderate (~30 ng/L) an d high load (~160 ng/L) conditions. In addition, Mn breakthrough of approximately 25 g/L was observed during simulated peak load testing (220 g/L). These results support the high frequency of T&O and (black/brown) colored water complaints received by the City (Hunt, 2009). The baseline biofiltration characterization also identified nutrient limitations in the full scale (and pilot) biofilter feed. During the course of 5months of steady state testing the bioavailable C:N:P ratio varied from 100:10:0 to 10 0:6:0. The absence of PO4P in the biofilter feed was consistent

PAGE 200

200 throughout the study, likely the result of the enhanced coagulation process. The observed PO4P limitation in the biofilter control was identified as a limiting factor for both hydraulic (15% higher terminal headloss) and water treatment performance (DOC, Mn, and MIB removal) relative to the nutrient enhanced biofilter (C:N:P of 100:10:1). The biofilter control filter media also contained approximately 30% less ATP. In addition, the microbial characterization of the biofilter control suggests increased EPS (free and bound), biofilm formation potential, and prevalence of biofilm matrices relative to the nutrient enhanced biofilter. 16S rRNA clone libraries were also developed for media samples c ollected from the biofilter control. These analyses found higher prevalence of microbial populations belonging to the genus Bradyrhizobium than observed in media collected from the nutrient enhanced biofilter operated under carbon limiting conditions (C:N: P of 100:14:2) (Figure 819). Select species within the genus Bradyrhizobium have been characterized to increase EPS production under nutrient limiting conditions (Quelas et al. 2006). These results suggest that nutrient limitation may not only increase biofilter EPS production (and thus headloss), but also select for populations responsible for producing excess amounts of EPS under limited conditions. The ability to produce additional EPS under nutrient limiting conditions might provides Bradyrhizobium a competitive advantage under normal biofilter operations (e.g., where additional EPS provides additional resistance to sloughing from abrasion and scouring during backwash) The findings of this study all suggest that the nutrient limited conditions may inhibit optimal water treatment performance while increasing filter headloss (and potentially

PAGE 201

201 underdrain fouling). These findings are supported by the work of Mauclaire et al. (2004), Nishijima et al. (1997), and Sang et al. (2003). Substrate Enhancement S tudies The objective of this study was to evaluate various primary substrate augmentation strategies for enhancing biofiltration performance. Increasing the concentration of primary substrate (through intermediate ozonation or primary substrate augmentation) in a biological treatment application may increase the rate of recalcitrant and trace organic compound degradation through secondary substrate metabolism and/or cometabolism. This study evaluated four primary substrates, including acetic acid, molasses, MicroC (a proprietary glycerinbased product produced by Environmental Operating Solutions, Inc, Bourne, Mass.), and ethanol. The substrates were evaluated individually through four successive oneto two month phases, each dosed at target of 1 mg/L as C. The l imited test durations and rapid succession of the substrates tests biased the results of these studies. The substrateenhanced biofilter likely retained biofilm matrix/community/population history after each substrate tested, influencing the effectiveness of successive tests. However, specific substrates, such as molasses, yielded immediate negative impacts on biofilter performance (Table 51), The substrateenhanced biofilters yielded 25 to 150% higher terminal headloss than the control biofilter during 18 hour filter runs (Table 51). However, sustainable (>1 week continuous operation) 18hour filter runs were not achieved with any substrate tested due to excessive headloss. Microbial analyses identified that substrateenhancement increased biofilter medi a ATP (7 to 30% Table 57), EPS (>180 % Table 8 15), and biofilm formation potential (>150% Table 813) relative to the control, indicating a higher degree of microbial activity and biofilm formation. The C:N:P ratio

PAGE 202

202 varied during these tests (Table 67) ; however, P and N limitations were persistent for all substrates tested. Interestingly, the substrateand nutrient enhanced biofilter showed higher ATP levels (>70% Table 816) than the substrateenhanced biofilter, yet terminal headloss, EPS, and biofi lm formation potential measurements remained near biofilter control levels. These microbial analyses suggest that headloss may be attributed to a higher production of EPS produced by bacteria and not to increased overall biological activity (as measured by ATP or HPC) Substrate augmentation strategies did not provide significant improvement of trace organics or metals removal relative to the control biofilter. Background DOC removal percentages were also comparable between the control and substrateenhance d biofilters when acetic acid and MicroC were used. The biofilter supplemented with molasses showed twice as much DOC breakthrough as the control biofilter, while the biofilter operated with ethanol supplementation showed an average of 50% higher background DOC removal relative to the control, though the differences in the mean breakthroughs were not shown to be statistically significant. Substrate enhancement provided no significant benefit to the pilot biofilter under the conditions and substrates tested Therefore, the strategies tested for substrateenhancement are not recommended. Possible strategy limitations that may have inhibited substrateenhanced biofilter performance include: Substrates tested do not support necessary enzymatic activity for seco ndary substrate or cometabolic process for trace organic degradation

PAGE 203

203 The supplemental bioavailable C increases biofilter nutrient deficiencies resulting in EPS production and headloss; therefore, limiting the steady state operational performance necessary for optimal substrate utilization. Nutrient Enhancement Studies The objective of this task was to evaluate various nutrient augmentation strategies for enhancing biofiltration performance. The primary focus of this study was to evaluate nutrient enhanceme nt by dosing 0.020 mg/L phosphoric acid as PO4P to satisfy the biofilter feed nutrient requirement for 0.8 mg/L of bioavailable DOC, achieving a 100:10:1 bioavailable C:N:P ratio. Phosphorus supplementation decreased biofilter terminal headloss by approxi mately 15% relative to the control (Table 61). This improvement could translate to longer filter run times and decreased backwash frequencies. Microbial characterizations suggest that biofilter nutrient enhancement (to achieve a 100:10:1 C:N:P ratio) can simultaneously decrease biofilm formation potential (>1400% ) and EPS media concentrations (>30% ), while increasing media ATP concentrations (>30 % ) (Figures 8 11, 812, 8 17) The observed decreases in EPS and biofilm formation potential are likely the caus e for the decrease in biofilter headloss. In addition, clone libraries indicated that pilot biofilter media operated under nutrient enhancement conditions had lower prevalence of organisms belonging to Bradyrhiz obium (<2 % ) than did biofilter media operated under nutrient limited conditions (where Bradyrhiz obium communities exceeded 10% ). Eliminating nutrient limitations appears to reduce the ability of biofilter populations to produce excess amounts of EPS. Communities relying on biofilm formation to persis t in drinking water biofilters may wash out under substratelimited conditions allowing for a different community of organisms to establish on filter

PAGE 204

204 media. These results suggest that the nutrient enhancement strategy may extend underdrain life by decreas ing the formation of EPS, which could be responsible for underdrain cap clogging, as experienced by Arlington and other utilities (Zhu, 2010). Phosphorus supplementation also enhanced biofilter water treatment performance. The increased media biological activity observed with nutrient enhancement corresponded with improved water treatment performance .The nutrient enhanced biofilter provided higher removals of DOC and MIB relative to the biofilter control. Biofilter nutrient enhancement improved DOC removal by over 75% relative to the biofilter control (Table 65). This result confirms previous work suggesting the nutrient limitation may inhibit optimal microbial substrate utilization and organics removal (Nishijima et al. 1997; Sang et al. 2003). MIB degradation follows secondary substrate metabolic pathways (Rittman, 1995); therefore, it is possible that a portion of the additional DOC removed served as primary substrate for the populations responsible for MIB transformation. Improved bio filter Mn removal was also observed under nutrient enhanced conditions. Although soluble Mn oxidation and removal may occur through physical/chemical processes in a biological filter (Kohl et al. 2006; Sahabi et al. 2009), it may also be oxidized by autotrophic bacteria (Diem and Stumm, 1984; Nealson, 1992; Sahabi et al. 2009). Phosphoric acid suppresses pH; therefore, its application should not contribute to the oxidation of Mn. However, complete Mn removal was observed in the nutrient enhanced biofilter under both highand moderateload conditions. Interestingly, clone library analyses on nutrient enhanced biofilter media indicated a relatively high prevalence of populations belonging to the genus Burkholderia relative to the biofilter control (> 25% t o ~1 % ) (Figure 8 19). The genus

PAGE 205

205 Burkholderia contains some species previously identified as Mnoxidizers (Kohl et al. 2006). These results suggest that nutrient limitation select against biofilter communities responsible for contaminant removal and improv ed water treatment. An additional nutrient enhancement evaluation sought to examine the effects of NH4N supplementation under N limited conditions. Ethanol and phosphoric acid were dosed to a biofilter at concentrations that created an artificial ammonia limitation (C:N:P equaled 100:3:2). This condition was maintained for 7 weeks and resulted in decreased hydraulic performance as measured by gradual increase in 18hr terminal headloss during the pilot filter runs (Table 62, Figure 6 2). The average term inal headloss for this biofilter was 9.9 feet (60% greater than that of the control), with many runs ending before 18hours due to a headloss trigger. Ammonium chloride was then dosed to eliminate the NH4N limitation. Consequently, average terminal headloss in the enhanced biofilter column decreased to 6 feet (56% improvement), indicating that ammonia limitations may be as deleterious to hydraulic performance as phosphorus limitations during biofilter operation (Figure 6 2). These results suggest that both N and P limitations may result in a loss in hydraulic performance in drinking water biofilters. The results of the nutrient enhancement study leads to the important conclusion that appropriate nutrient supplementation produces more biological activity but at the same time improves filter performance because the bacterial production of EPS is reduced with the removal of the nutrient limitation stress. Oxidant Enhancement Studies The objective of this task was to evaluate peroxide supplementation for augm enting the oxidative action and response of the biofiltration process. Many microorganisms express a class of enzymes, known as peroxidase family

PAGE 206

206 oxidoreductases, when exposed to hydrogen peroxide. The resulting reaction neutralizes the hydrogen peroxide, oxidizing organic (or inorganic) substrates. Peroxidases, a specific group of enzymes within the peroxidase family, release free radicals when reducing hydrogen peroxide (Schumb et al. 1955). These free radicals may then oxidize recalcitrant organic carbon, further reducing the concentration of DOC and trace organics (Klibanov et al. 1980; Huang et al. 2004).In addition, inactive microorganisms and/or their extracellular materials may also be oxidized by the hydrogen peroxide and free radicals (Neyens, 2003), thereby improving biofilter hydraulics. Preliminary testing of this strategy was conducted by dosing 1 mg/L of hydrogen peroxide to a pilot filter for two weeks. The test filter demonstrated 15% removal of filter feed DOC (50% greater than that obser ved with the control and full scale filters), and complete removal of Mn and MIB. The oxidant enhancement strategy also decreased terminal headloss to 2.2 feet, or 66% less than the control biofilter (Figure 7 1). ATP measurements were performed on the ox idant enhanced biofilter media to characterize biological activity relative to the control. The data showed that hydrogen peroxide supplementation did not decrease ATP concentrations in the biofilter media relative to the biofilter control (Figure 818) and these data suggest biofilter populations are resistant to the levels of hydrogen peroxide dosed in this study. However, the peroxide concentration was sufficient to drive significant improvements in filter hydraulic performance. These improvements are possibly due to microbially mediated peroxide oxidation process that destabilized EPS and other organic biofilter foulants sufficiently to be removed through mineralization or through filter backwash.

PAGE 207

207 Microbial Tracking The objective of this task was to c orrelate the microbial ecology and activity in the biofilters to hydraulic and treatment performance. Media samples were collected from the pilot and full scale biofilters after each process change and analyzed for microbial activity, community structure, and morphology. The microbial tracking task included the following analyses: SEM, HPC, biofilm formation potential, EPS quantification, T RFLP, and clone libraries. Scanning e lectron m icroscopy SEM provides a qualitative means to track general changes in biofilm appearance and cellular morphology. Biofilter media samples were sent to the USEPA ORD for SEM characterization. SEM analyses showed considerable differences in biofilm morphology among biofilters that employed different operational strategies. In g eneral, biofilms developed under nutrient limited conditions appeared to have a higher prevalence of a biofilm matrix relative to biofilter with nutrient enhancement ( Figures 8 2, 8 7). The observed presence of the biofilm matrix significantly increased i n samples collected from biofilter operated with substrate supplementation. In addition, the prevalence of stalked bacteria and their corresponding stalk lengths on biofilter media SEMs appeared to coincide with influent nutrient limiting conditions. The b acterial stalk is a morphological adaptation that allows these organisms to increase their surface areato volume ratio, improving nutrient diffusion through the cell wall, thus allowing the organism to be more competitive under nutrient limited conditions (Madigan et al. 2009). Conversely SEMs of the nutrient enhanced biofilter media showed a decrease in biofilm matrix, while exhibiting a higher degree of cellular morphological diversity and larger cells.

PAGE 208

208 Biofilter m edia HPC Select biofilter media samples were surveyed for HPC. Overall, these results suggest that phosphorus is responsible for increasing the number of viable cells in the filter (Figure 8 11). These data suggest that nutrient supplementation may increase the prevalence of viable heterotrophic bacteria on biofilter media. The HPC data support the observed improvements in biofilter activity and DOC removal during the nutrient enhancement study. Biofilm f ormation p otential The biofilter media biofilm formation potential was lower for the nutrient enhanced biofilter relative to the control for the one sample tested (Figure 812). All substrateenhanced biofilters showed higher biofilm formation potentials relative to the control conversely the substrateand nutrient enhance biofilter media show ed similar biofilm formation potential to the control biofilter (Figure 813). These results suggest that nutrient limitations may drive biofilm formation potential in a biofilter with or without substrate supplementation. The biofilm formation potential r esults generally corresponded to relative filter hydraulic performance between the control, nutrient enhanced, substrateenhanced, and substrateand nutrient enhanced biofilters EPS q uantification Measured biofilter media EPS was lowest for the nutrient enhanced biofilter, relative to the control biofilter (Figure 813). S ubstrateenhancement appeared to increase free and bound EPS concentrations, while substrateand nutrient enhancement had little effect. The decreased presence of EPS in the nutrient e nhanced and substrateand nutrient enhanced biofilter samples corresponded with decreased headloss relative to the control and substrate enhanced biofilters, respectively. The

PAGE 209

209 results suggest that supplementation of substrates increased the normalized (to HPC) production and quantity of biofilter EPS in nutrient limited conditions. However, both EPS production potential and the EPS concentration were reduced to biofilter control levels when the nutrient requirements were satisfied (Figures 813, 8 15). Aden osine t riphosphate m easurements ATP is a high energy molecule considered the energy currency of life. Higher concentrations of ATP are associated with increased metabolic activity. The ATP analyses provided a real time assessment of biofilter activity that correlated directly with DOC removal. Nutrient limited biofilters consistently showed lower concentrations of ATP in media samples than those that those operated under the same carbon load conditions and a bioavailable C:N:P balance of 100:10:1. When nutr ients were dosed to previously nutrient limited biofilters, biofilter ATP concentrations increased, as did DOC removal. Thus, nutrient enhanced biofilters simultaneously demonstrated (1) increased ATP concentrations, (2) improved effluent water quality, and (3) lower headloss trends. Interestingly, biofilters operated with 1 mg/L (as C) ethanol supplementation exhibited minor improvements to ATP concentrations relative to the improvements observed during phosphorus supplementation tests. 16S rRNA and c lone l ibraries The most significant observation in the clone libraries for these samples was the drastic reduction in the Bradyrhizobium population when the stoichiometric C:N:P ratio was met on the enhanced biofilter. Bradyrhizobium constituted only 1.5% of th e clones in the stoichiometric C:N:P sample, but constituted 15% under the baseline control conditions. Bradyrhizobium has been shown to increase EPS production under N -

PAGE 210

210 limitation and related rhizobia have been shown to increase EPS production under P limitation. Full s cale Process Integration Assessment and Economic Evaluation During pilot testing, the substrate enhancement, nutrient enhancement, and oxidant enhancement strategies all demonstrated some water treatment benefits. The oxidant and nutrient en hancement strategies also provided significant hydraulic benefits, making them particularly promising for full scale implementation. The oxidant enhancement strategy was evaluated only briefly, and additional testing is required to better understand proces s mechanisms and characterize long term performance. A full scale process integration assessment was performed for the nutrient enhancement strategy, as it was better characterized and validated than the oxidant enhancement strategy, and it was shown to be operationally and economically effective when used without other supplements. Process integration includes the installation of a chemical feed system capable of dosing orthophosphate and/or ammonia to the top of the biofilter feed channel, Chemical feed s ystem sizing is a function of biofilter production capacity and the degree of nutrient limitation. At the JKWTP, the filter feed has a typical orthophosphate limitation of 20 g/L as P. Less than 5 gallons per day of 85% phosphoric acid would mitigate this limitation, which means that a standard 250gallon tote would be capable of providing in excess of 50 days of storage for this facility. The estimated installed capital cost for a nutrient dosing system is $25,000, and the associated chemical cost estimat e is approximately $0.44 per MG treated. This chemical cost may be offset by increased biofilter production efficiency. Indeed, factoring in consumable nutrient costs (and contingency), Arlington Water Utilities could save over $17,100/yr in operating cost s if nutrient supplementation

PAGE 211

211 increased biofilter runs by 15% at both the JKWTP and PBSWTP. Implementing nutrient enhancement strategies may provide additional cost savings through extending the life of media and biofilter underdrains/caps. All savings realized with the hydraulic improvements of biofilter nutrient enhancement compliment the improved water treatment performance: enhanced DOC, Mn, and MIB removal. Utilities implementing biofilter nutrient enhancement should perform regular assessment of biofi lter fed nutrient and BOM levels. The costs associated with implementation will be dependant on the presence, concentration, and requirement for both PO4P and NH4N supplementation. Nutrient limitations (and corresponding requirements) may likely vary dur ing source water seasonal changes. Summary Currently, biofiltration is largely operated as a passive process in the water treatment industry. Particle/turbidity removal and headloss drive the design and operation of conventional filtration as well as biof iltration. Thus, biofilter design parameters are typically limited to media configuration, backwash strategy, and loading rate. The biological removal of dissolved organic and inorganic contaminants is an anticipated benefit of biofiltration. However, comm on design and operational practice does not seek to enhance the biological activity responsible for those mechanisms. Indeed, in an effort to improve filter productivity and minimize headloss, many utilities employ chlorinated backwashes and other biomass control strategies. The purpose of this research was to identify strategies to enhance the biological activity in a biofilter without compromising productivity or particulate removal performance. The ultimate goal of this work is to shift an industry accep ted paradigm so that the design and operation of biofilters are driven not only by filtration but also by

PAGE 212

212 biological treatment objectives. Strategies comprised dosing low levels of common drinking water treatment chemicals at a feed point just upstream of a biofiltration process. Of the strategies tested, nutrient enhancement and oxidant enhancement showed the most promise for drinking water biofilter applications. Substrateenhancement proved to be ineffective at providing either improved water treatment p erformance or reliable biofilter hydraulic operation. The oxidant enhancement strategy significantly improved filter hydraulic performance without compromising biological activity. However, the chemical cost for the dosage tested (1 mg/L) is impractical for full scale implementation ($12/MG). Therefore, additional optimization and validation are necessary to define operational parameters for full scale implementation. The nutrient enhancement strategy is elegant in its simplicity: operate a given biofiltrat ion process so that an approximate bioavailable C :N:P molar ratio of 100:10:1 is maintained. Associated water quality benefits may include improved biological treatment of organic carbon, Mn, and MIB. Associated hydraulic benefits may include lower termina l filter headloss and decreased media and underdrain clogging. Therefore, nutrient enhancement strategies may be applicable to any utility with existing or planned biofiltration facilities. The primary goal of this work was to move the practice of biofiltration from a passive process designed and operated around conventional filtration objectives to an intentionally operated biological system, i.e. engineered biofiltration. Engineered biofiltration targets multiple water quality objectives while m aintaining hydraulic performance.

PAGE 213

213 CHAPTER 11 FUTURE WORK This research identified unique and promising biofilter enhancement strategies that may provide both hydraulic and water treatment performance improvements to surface water utilities upon impl ementation. However, additional work further evaluating both fullscale performance and enhancedbiofilter microbial community functionality is warranted. Currently the City of Arlington is developing a work plan to perform full scale nutrient enhancement on select filters. The next phase of pilot research will be performed under Water Research Foundation Tailored Collaboration No. 4346. This research will validate the Engineered Biofiltration approach and include additional operational refinement of the hydrogen peroxide enhancement and nutrient enhancement strategies. Source water and seasonal variation, long term steady state performance, and distri bution system impacts will also be evaluated in this study. Specific research objectives should include further evaluation of the following: (1) refine operational parameter at other facilities, (2) elucidate mechanisms and mitigation strategies for underdrain clogging, (3) identify biofilter microbial communities and understand their role in EPS production and/or contaminant cycling, and (4) characterize how enhanced biofiltration affects distribution system health and water quality. This research evaluat ed enhanced biofiltration strategies at one facility with a relatively stable feed water quality. Therefore, further characterization would be beneficial to examine strategy effectiveness at facilities treating different source waters. The impacts of feed water temperature, nutrient loads, bioavailable organic carbon in different water matrices may have profound effects on enhanced biofilter performance.

PAGE 214

214 In addition, the effects of pretreatment process selection and optimization on the enhanced biofilter st rategies should be studied (e.g., ozonation, coagulationsedimentation, softening, etc.). The implementation of enhanced biofiltration in a pretreatment mode (i.e., direct filtration of raw water) also merits evaluation. Direct biofiltration would take adv antage of the relatively higher concentrations of nutrients available in the raw water, reducing or eliminated the need for supplementation. In addition, the DOC removal and other water quality enhancements achieved through direct biofiltration may decreas e operating costs for downstream processes, e.g. coagulant and polymer. Further study is necessary to characterize the mechanisms for underdrain clogging and evaluate the ability of the enhanced biofilter strategies to prevent and/or mitigate clogged underdrains. This evaluation will require long term biofilter operation (greater than 12 months) to adequately characterize and verify the performance of the underdrain caps under optimized steady state conditions. This task would evaluate (1) underdrain cap p erformance after steady state operation with biofilter enhancement strategies, and (2) clogged underdrain cap mitigation with biofilter enhancement strategies. These evaluations would include autopsies to reveal the extent and mode of underdrain cap cloggi ng. Autopsy evaluations would include surface analysis and biological assessment techniques such as identification and abundance of microbial, bulk characterization and quantification of biofilm EPS, and insitu characterization and imaging of biofilms. In addition, the underdrain caps would be analyzed for potential organic and inorganic accumulation by methods such as Fourier transform infrared

PAGE 215

215 (FTIR) spectroscopy and SEM imagery with energy dispersive spectroscopy (EDS). Permeability tests would also be conducted on all clogged underdrain caps. Additional biofilter microbial activity characterization is necessary to define the functionality of the present microbial communities and identify their roles in EPS production and contaminant cycling. This resea rch constructed clone libraries targeting the bacterial 16S rRNA gene for only a few pilot biofilters. It is likely that biofilters at different utilities have different microbial community structures. In addition, change in community structure over time ( e.g., seasonal variation) within each system is anticipated. The microbial community structure depends on source water quality, filter characteristics (e.g., filter media) and operational characteristics (e.g., backwash frequency and intensity, temperature). Therefore, it is unknown if the populations identified during this research are commonly present in other biofilters used for drinking water treatment. Thus, is it important to perform microbial community characterization on the biofilters for facilities interested in enhancement strategy implementation. Subsequent identification of protein coding genes through pyrosequencing and combined with reverse transcription quantitative PCR (RT qPCR) assays combined with reverse transcriptase (RT PCR) would then allow researchers to collected data on metabolism, function, and removal mechanisms for organisms of interest (Raskin, 2010) Additional microscopy studies may also present additional information on biofilter media microbial communities and biofilm charact eristics. TEM would be used on future samples to accurately quantify biofilm thickness a potential measure to determine relative microbial stress between biofilter operating conditions. Fluorescence i n situ h ybridization analyses would be used to support genetic analyses to obtain information

PAGE 216

216 on possible niche differentiation of the present microbial populations (Amann and Fuchs, 2008). Further microbial community analyses could lead researchers towards the identification of enzymes responsible for organi c and inorganic contaminant cycling in biofilters. Once identified, and understood, methods to further enhance specific enzyme production and activation may be developed. The effectiveness of drinking water biofiltration on trace organic removal, e.g. tast es and odors may be greatly enhanced if microorganisms could be stimulated to increase the production of specific enzymes, such as peroxidase. Further evaluation of biofilter hydrogen peroxide enhancement may elucidate the potential for this approach. The effects of enhanced biofiltration on distributions systems warrant additional investigation. Studies using annular reactors and pipe loops would be beneficial in characterizing improvement or degradation of effluent stability with respect regrow th, disinfectant decay, corrosion and disinfectant byproduct formation (for both regulated and nonregulated (DBPs). Furthermore, microbiological methods could be implemented to investigate the effect of disinfection on biofilter effluent microbial communi ty structure. Effluent samples from enhanced and control biofilters could be surveyed for microbial species that are resistant to disinfection. This information will be collected and compared to the existing body of knowledge to evaluate how the disinfecte d biofilter effluent may affect the distribution system.

PAGE 217

217 REFERENCES Ahmad, R ., Amirtharajah, A. Al Shawwa, A., Huck P 1998. Effects of Backwashing on Biological Filters. Journal AWWA 90(12):6273. Al Rifai, J., Gabelish, C., Schafer, A ., 2007. Occurrence of Pharmaceutically Active and Non Steroidal Estrogenic Compounds in three Different Wastewater Recycling Schemes in Australia. Chemosphere, 69 (5): 803815 Amann, R., Fuchs B 2008. Singlecell Identification in Microbial Communities by Improved Fluorescence in Situ Hybridization Techniques. Nature Reviews Microbiology 6:339348. A merican Water Works Association 1999. Water Quality and Treatment, 5th Edition. New York: McGraw Hill, Inc. Anderss on, A., Laurent, P., Kihn, A., Prvost, M., Servais P 2001. Impact of Temperature on Nitrification in Biological Activated Carbon (BAC) Filters Used for Drinking Water Treatment. Water Research, 35(12):29232934. Ang, Y. 2006. Development and Optimization of a FixedBed Bioreactor for the Removal of 2 Methylisoborneol and Geosmin in the Production of Drinking Water. Masters Thesis, University of Duisburg Essen, Duisburg, Germany. A ssociated Press (2008, March 10) Drugs found in drinking water USA Today Auriol, M. FilaliMeknassi, Y., Adams C Tyagi, D., Noguerol, T., Pena B 2007. Removal of Estrogenic Activity of Natural and Synthetic Hormones from a Municipal Wastewater: Efficiency of Horseradish Peroxidase and Laccase from Trametes Versicolor. Chemosphere, 17897698 (P, S, E, B, D). Baker N 1948. The quest for pure water. American Water Works Association, New York, NY. Becker, W ., OMelia C 1996. Ozone, Oxalic Acid, and Organic Matter Molecular Weight Effects on Coagulation. Ozone Science and Engineering, 18(4): 311324. Beech, I B. Zinkevich, V., Tapper, R., Gubner, R. (1997). Direct involvement of an extracellular complex produced by a marine sulfatereducing bacterium in deterioration of steel. Journal of Geomicrobiology 15: 121 134. Bennett Stamper, C. 2009. Personal communication with Christina Bennett Stamper, Research Assistant USEPA ORD on December 23, 2009.

PAGE 218

218 Berry, D., Xi, C., Raskin L. 2008. Effect of Growth Conditions on Inactivation of Escherichia coli with Monochloramine. Environmental Science & Technology DOI: 10.1 021/es8017545. Bouwer. E., Crow, P 1990. Biological processes in drinking water treatment. Journal AWWA 80(9): 8293. Briones, A ., Daugherty, B. Angenent, L Rausch, K Tumbleson, M Raskin L 2007. Microbial Diversity and Dynamics in Multi and S ingle Compartment anaerobic Bioreactors Processing SulfateRich Waste Streams. Environmental Microbiology. 9:93106. Brown, J 2006. Simultaneous Destruction of Multiple Drinking Water Contaminants Using Biological Filtration. In Proc. of the AWWA Annual Conference and Exhibition, San Antonio, Texas, June 1115. Brown, J 2007a. Biological Treatments of Drinking Water. The Bridge 37(4):3036. Brown, J 2007b. Biological Drinking Water Treatment: Benefiting from Bacteria. Paper presented at the National Academy of Engineering 2007 U.S. Frontiers of Engineering Symposium, Redlands, WA, September 26. Brown, J., Lauderdale, C 2009. Biofiltration of Organic Carbon, MIB, Geosmin, EDCs, and Membrane Foulants. WQTC Conference Proceedings Burger, M., Mercer, S., Shupe, G., Gagnon, G., 2008. Manganese removal during benchscale biofiltration Water Research 42:4733. Carlson, K., Amy G. 1998. BOM Removal during Biofiltration. Journal AWWA 90(12):4252. Carmichael, W 2001. Assessment of Blue Green Algal Toxins in Raw and Finished Drinking. Water. AWWA Research Foundation, Denver, CO. Casale, R., LeChevallier, M., Pontius, F. 2001. Review of Manganese Control and Related Manganese Issues. Denver: American Water Works Association (AWWA) Research Foundation and AWWA. Chae, S., Summers, S., Kim, S., Ahn H. 2006. Biofiltration of MIB and Geosmin under Varying Influent Conditions. Paper presented at the AWWA Water Quality and Technology Conference, November 59, Denver, CO. Chemical Wate r and Wastewater Treatment IX. 2007. (Proceedings, 12th Gothenburg Symposium of 2007, Ljubljana, Slovenia), pp 297 306.

PAGE 219

219 Chcharo, M. Chcharo, L 2008. RNA:DNA Ratio and Other Nucleic Acid Derived Indices in Marine Ecology. International Journal of Molec ular Sciences 9:14531471. Choi, Y., Li, X., Raskin, L., Morgenroth, E. 2007. Effect of Backwashing on the Removal of Perchlorate in Fixed Bed Biofilm Reactors, Water Research 41(9):19491959. Christensen, B., Characklis W 1990. Physical properties of biofilms. Biofilms. Edited by W. G. Characklis & K. C. Marshall. New York:Wiley. Christensen, B ., Naper, T ., Vollan, K., Bake, R. 1990. Biofilm removal by low concentration of hydrogen peroxide. Biofouling 2:165 175. Davidson, K ., Gilpin, L., Hart, M., Fouilland, E ., Calleja, I. Laurent C ., Leakey, R. 2007. The influence of the balance of i norganic and organic nitrogen on the trophic dynamics of microbial food webs. Limnology and Oceanography 52(5): 21472163 Decho, A 1990. Microbial exopolymer secretions in ocean environments: their role(s) in food webs and marine processes. Oceanographic Marine Biology Annual. Review 28: 73 153 DeWaters, J ., DiGiano F 1990. Influence of Ozonate Natural Organic Matter on Biodegradation of Micropollutant in GAC Bed. Journal AWWA 82(8):69. Diem, D., Stumm, W. 1984. Is Dissolved Mn2+ Being Oxidized by O2 in Absence of MnBacteria or Surface Catalysts? Geochimica et Cosmochimica Acta, 48:15711573. Duboi s, M., Gilles, K ., Hamilton J., Rebers P ., Smith F. 1956. Colorimetric Method for Determination of Sugars and Related Substances. A nalytical chemistry ,28(3): 350365. Dugan, N 1998. Determination of Biofiltration Performance, Estimation of Biofiltration Model Parameters and Validation of Model Performance at Pilot and Full Scale, and with Respect to pH, in Drinking Water Biofilters. M.S. thesis. Department of Civil and Environmental Engineering, University of Cincinnati, U.S. Dussert, B., Vanstone, G 1994. The biological activated carbon process for water purification. Water Engineering & Management 141(12): 22 24. Egashira, K., Ito, K., Yoshiy, Y. 1992. Removal of musty odor compound in drinking water by biological filter. Water Science and Technology 25 (2) 307 314.

PAGE 220

220 Egert, M., Friedrich M 2005. Post amplification Klenow fragment treatment alleviates PCR bias caused by parti ally single stranded amplicons. Journal of Microbiological Methods 61:6975. Eichler, S., Christen, R., Holtje C., Wesphal, P., Botel, J., Brettar, I., Mehling, A., Hofle, M. 2006. Composition and Dynamics of Bacterial Communities of a Drinking Water Supp ly System as Assessed by RNA and DNA Based 16S rRNA Gene Fingerprinting. Applied and Environmental Microbiology 72:3:18581872. Elhadi, S. Huck, P., Slawson, R 2003. Removal of Earthy/Musty Odours from Drinking Water by Biological Filtration: Temperat ure and Media Effects. In Proc. AWWA Water Quality Technology Conference (WQTC), Philadelphia, Pennsylvania, U.S. Elhadi S., Huck, P., Slawson, R 2006. Factors affecting the removal of geosmin and MIB in drinking water biofilters. Journal AWWA 98(8): 108 119. Emelko, M., Huck, P., Coffey, B., Smith, E. 2006. Effects of Media, Backwash, and Temperature on Full Scale Biological Filtration. Journal AWWA 98(12):6173. Escobar I. Randall A Taylor J. 2001. Bacterial growth in distribution systems: Effect of assimilable organic carbon and biodegradable dissolved organic carbon. Environmental Science & Technology 35(17): 34423447 Evans, P ., Optiz, E., Daniel, A., Schulz, C. 2009. Biological Drinking Water Treatment Perceptions and Actual Experiences in North America. Water Research Foundation Final Report, Denver, CO: Fang, W., Hu, J., Ong, S. 2009. Influence of phosphorus on biofilm formation in model driking water distribution systems. Journal of Appli ed Microbiology 106: 13281335. Fenchel, T., King, G., Blackburn, T. 2000. Bacterial Biogeochemistry. Academic Press, New York. p 20. Flemmin g H ., Wingender, J., Griebe, T., Mayer, C. 2000. Physicochemical Properties of Biofilms. Biofilms: Recent Advan ces in Their Study and Control. Overseas Publishers Association, Amsterdam. Flemming, H., Wingender J. 2001a. Relevance of microbial polymeric substances (EPSs) part I: structural and ecological aspects. Water Science Technology 43: 18. Flemming, H. Wingender J 2001b. Relevance of microbial polymeric substances (EPSs) part II: technical aspects. Water Science Technology 43: 9 16.

PAGE 221

221 Fonseca, A ., Summers, R., Hernandez M 2001. Comparative Measurements of Microbial Activity in Drinking Water Biof ilters. Water Research 35(16):38173824. Fuerhacker M Durauer A Jungbauer A. 2001. Adsorption isotherms of 17 betaestradiol on granular activated carbon (GAC). Chemosphere, 44(7): 15731579 Gillespie, C. 1925. Filtration Plant Census, Journal AWWA 14: 123142. Goldgrabe, J., Summers, R., Miltner R. 1993. Particle Removal and Head Loss Development in Biological Filters, Journal AWWA 85(12):94106. Gonin, M Quardokus E O'Donnol D Maddock J. ,Brun Y 2000. Regulation of stalk elongation by phosphate in Caulobacter crescentus. Journal of Bacteriology 182(2):33747 Greer L. Szalay A 2002. Imaging of light emission from the expression of luciferases in livin g cells and organisms: a review Luminescence, 17 (1): 43 74. Guay, C., Peldszus, S., Huck, P., McPhail, B., MosquedaJimenez D 2007 Removal of selected pharmaceutically active compounds and endocrine disrupting substances in surface water by membrane filtration with biological filtration pretreatment. In: Chemic al Water and Wastewater Treatment X, (Proceedings, 12th Gothenburg Symposium of 2007, Ljubljana, Slovenia): 297306. Ho, L., Hoefel, D., Bock, F., Saint, C., Newcombe G. 2007. Biodegradation rates of 2methylisoborneol (MIB) and geosmin through sand fil ters and in bioreactors, Chemosphere, 66:2210. Howe, K., Clark M. 2002. Coagulation Pretreatment for Membrane Filtration. Water Research Foundation Final Report, Denver, CO. Hozalski, R. Bouwer, E 2001. NonSteady State Simulation of BOM Removal in Drinking Water Biofilters: Applications and Full Scale Validation. Water Research, 35(1):211223. Huang, Q., Pinto, R., Burlingame, D., Tang, J., Weber, W 2004. Enhanced Removal of Natural Organic Matter via PeroxidaseMediated Oxidative Coupling. IWA P ublishing 4(4):3340. Huang, Q., Weber W 2000. Interactions of Soil Derived Dissolved Organic Matter with Phenol in PeroxidaseCatalyzed Oxidative Coupling Reactions. Environmental Science and Technology 38(1):338 344, 2004. Huck, P ., Coffey, B., Amirtharajah, A., Bouwer E 2000. Optimizing Filtration in Biological Filters. Water Research Foundation Final Report, Denver, CO:

PAGE 222

222 Huck, P., Finch, G., Hrudey, S., Peppler, M., Amirtharajah, A., Bouwer, E., Albritton, W., Gammie L 1998. Design of Biological Processes for Organics Control. Water Research Foundation Final Report, Denver, CO: Huisman, L., Wood, W 1974. Slow Sand Filtration. WHO, Geneva, Switzerland. Hunt, J. Personal communication with Julia Hunt, Arlington Public Works Director, on Ma rch 15, 2009. Ishida, H., Miyaji, Y. 1992. Biodegradation of 2 methylisoborneol by oligotrophic bacterium isolated from a eutrophied lake. Water Science Technology 25 (2) : 269276. Izaguirre, R., Wolfe, L., Means, E (1988) Bacterial degradation of 2 methylisoborneol. Water Science and Technology 20 (8/9) : 205210. Johnson, G. 1914. Present Day Water Filtration Practice. Municipal Journal 36 (22): 782784. Juhna, T., Rubulis J 2004. Problem of DOC Removal during Biological Treatment of Surface Water with a High Amount of Humic Substances. Water Science and Technology: Water Supply 4(4):183187. Juhna, T., Melin E. 2006. Ozonation and Biofiltration in Water Treatment Operational Status and Optimization Issues. Techneau Report D.5.3.1 B. Kawamura, S. 2000. Design and Operation of Water Treatment Facilities Second Edition, John Wiley & Sons, NY. Kim, K., Logan B.E. 2000. FixedBed Bioreactor Treating Perchlorate Contaminated Waters. Environmental Engineering Science, 17(5): 257 265. Kirly, Z., El Zahaby, H., Klement Z. 1997. Role of Extracellular Polysaccharide (EPS) Slime of Plant Pathogenic Bacteria in Protecting Cells to Reactive Oxygen Species Journal of Phytopathology Journal of Phytopathology, 145 (23) 5968. Klibanov, A., Alberti, B., Morris, E., Felshin, L. 1980. Enzymatic Removal of Toxic Phenols and Anilines from Wastewaters. Journal of Applied Biochemistry 2:414 421. Knocke, W. Van Benschoten, J., Kearney, M Soborski, A. Reckhow D . 1990. Alternative Oxidants for the Removal of Soluble Iron and Manganese. Water Research Foundation Final Report, Denver, CO:

PAGE 223

223 Kobayashi, H., Rittmann B 1982. Microbial Removal of Hazardous Organic Chemicals. Environmental Science and Technology 16:1 70A. Kohl, P. Medlar S. 2006. Occurrence of Manganese in Drinking Water and Manganese Control. Water Research Foundation Final Report, Denver, CO: Krasner S Sclimenti, M Coffey B. 1993. Testing biologically active filters for removing aldehydes formed during ozonation. Journal AWWA 85(5):62 71. Lauderdale, C., Lindner A.. 2004. Characterization of a Methylisoborneol Degrading Mixed Culture Derived from Natural Waters, Water Research pp. 41354142 Lauderdale, C. Brown J. 2005. Biological Treatment of Tastes, Odors, and Other Trace Organic Contaminants. Paper presented at the Florida Sec tion of the American Water Works Association Annual Conference, Orlando, Florida, November 2830, 2005. Lauderdale, C. Brown J. 2007. A Novel Biological Treatment Approach for the Treatment of Algal Metabolites. Poster presented the AWWA Annual Confere nce and Exhibition, Ontario, Canada, June 2428. LeChevallier M ., Schulz, W., Lee R 1991. Bacterial Nutrients in Drinking Water. Applied and Environmental Microbiology 57(3):857862. LeChevallier, M ., Becker, W., Schorr, P., Lee R 1992. Evaluating the Performance of Biologically Active Filters. Journal AWWA 84(4):136146. Li, X., Upadhyaya, G., Yuen, W., Brown, J., Morgenroth, E., and Raskin, L 2009. Changes in Microbial Community Structure and Function of Drinking Water Treatment Bioreactors Upon Phosphorus Addition, Water Research Submitted. Lim, M. Snyder, S., Sedlak D. 2008. Use of Biodegradable Organic Carbon to Assess the Potential Transformation of Wastewater Derived Contaminants in Surface Waters. Water Research, 42:29432952. Liu, J ., Liu, C. Edwards, E., Liss, S. 2006. Effect of phosphorus limitation on microbial floc structure and gene expression in activated sludge. Water Science and Technology 54(1) 247 255. Liu, X., R.M. Slawson, and P.M. Huck. 2000. Drinking Water Biofiltrati on: Assessing Key Factors and Improving Process Evaluation. In Proceedings of the AWWA Annual Conference and Exhibition, Denver, CO. Liu, X., Huck, P., Slawson, R 2001. Factors Affecting Drinking Water Biofiltration. Journal AWWA 93(12):90101.

PAGE 224

224 LuminU ltra. 2008. Products catalogue. Lundgren, B Grimvall, A., Savehed, R. 1988. Formation and Removal of Off flavor. Water Science & Technology 20:8/9:245. Lutgarde, R., 2009. Personal communication with Lutgarde Raskin, Professor at University of Michigan, September 24, 2009. Madigan, M., Martinko, J., Dunlap, P., Clark, D. 2009. Brock Biology of Microorganisms. 12th Editiion, Pearson/Benjamin Cummings, New York. Mallevialle, J., Suffet, M 1987. Identification and Treatment of Tastes and Odors in Drinking Water. AWWARF and AWWA, Denver, CO. Manem, J ., Rittman B 1992. The Effects of Fluctuations in Biodegradable Organic Matter on Nitrification Filters. Journal AWWA 84(4):147157. Marda, S., Kim, D., Gordy, J., Pierpont, S., Gianatasio, J., A mirtharajah, A., Kim. J. 2008. Plant Conversion Experience: Ozone BAC Process Installation and Disinfectant Residual Control. Journal AWWA 100(4):117128. Mauclaire, L., Schurmann, A., Thullner, M., Gammeter S., Zeyer, J., 2004. Sand filtration in a water treatment plant: biological parameters responsible for clogging Journal of Water Supply: Research and Technology AQUA 53 (2) 93108. Metcalf and Eddy, Inc. 2002. Wastewater Engineering: Treatment, Disposal and Reuse. 3rd Edition, McGraw Hill Inc., New Y ork. Meyer, K. Summers, R., Westerhoff, P., Metz D 2005. Biofiltration for Geosmin and MIB Removal. In Proc. American Water Works Annual Conference, San Francisco, CA., U.S. Miller, R. 1986. Oxidation of Cell Wall Polysaccharides: A Potential Mechani sm for Cell Wall Breakdown in Plants. Biochemical and Biophysical Research Communications 141(1):238244. Miltner, R. R.S. Summers, and J.Z. Wang. 1995. Biofiltration Performance: Part 2, Effect of Backwashing. Journal AWWA 87(12):6470. Moll, D., Sum mers, R., Fonseca, A., Matheis W. 1999. Impact of Temperature on Drinking Water Biofilter Performance and Microbial Community Structure. Environmental Science and Technology 33:23772382. Morgenroth, E. 2008a. Biofilm Systems. Henze, M., van Loosdrecht, M. C. M., Ekama, G., and Brdjanovic, Damir. Biological Wastewater Treatment Principles, Modelling, and Design. (Chapter 18). IWA Publishing, London.

PAGE 225

225 Morgenroth, E. 2008b. Modelling Biofim Systems. In: Henze,M., van Loosdrecht, M.C.M., Ekama,G. and Brdjanovic,D. (eds.), Biological Wastewater Treatment Principles, Modelling, and Design, IWA Publishing, London. Najm, I., Kennedy, M., Naylor W.. 2005. Lignite versus bituminous GAC for biofiltration a case study. Journal of AWWA 97(1), 94 101. Namkung, E., Rittmann B 1987a. Estimating Volatile Organic Compound Emissions from Publicly Owned Treatment Works. Journal Water Pollution Control Federation, 59:670. Namkung, E., Rittmann B 1987b. Removal of Taste and Odor Causing Compounds by Biofi lms Grown on Humic Substances. Journal of AWWA 79(7):107. Nealson, K. 1992. The ManganeseOxidizing Bacteria, pp. 23102320. In The Prokaryotes. 2nd edition Volume 3. New York: Springer Verlag N erenberg, R., Rittmann B. Soucie, W 2000. Ozone/biofiltr ation for removing MIB and geosmin. Journal of American Water Works Association, 92:12, 85100. Neyens, E., Baeyens, J., Weemaes, M De Heyder B.. 2002. Advanced biosolids treatment using H2O2oxidation. Environmental Engineering Sciences 19 (1):27 35 Nielsen, K., Boye M 2005. Real Time Quantitative Reverse TranscriptionPCR Analysis of Expression Stability of Actinobacillus pleuropneumoniae Housekeeping Genes during In Vitro Growth under IronDepleted Conditions. Applied and Environmental Microbiology 71(6):29492954. Nielsen, P. Jahn, A., Palmgren, R. 1997. Conceptual model for production and composition of exopolymers in biofilms. Water Science and Technology 36(1): 1119. Niquette P., Prvost, M., Servais, P., Beaudet, J., Coallier, J., Lafrance, P 1998. Shutdown of BAC Filters: Effects on Water Quality. Journal of AWWA, 90(12):5361. Nishijim a, W., Shoto, E., Okada M. 1997. Improvement of biodegradation of organic substance by addition of phosphorus in biological activated carbon. Wate r Science and Technology (36):12 pp 251 257 Norton, C. LeChevallier M 2000. A pilot Study of Bacteriological Population Changes Through Potable Water Treatment and Distribution. Applied and Environmental Microbiology 66:268 276.

PAGE 226

226 Nouvion, N ., Block J., Faup, G 1987. Effect of biomass quantity and activity on TOC removal in a fixedbed reactor. Water Research, 21(1): 35 40. Oerther, D., Pernthaler, J., Schramm, A., Amann, R., Raskin L 2000. Monitoring Precursor 16S rRNAs of Acinetobacter spp. in A ctivated Sludge Wastewater Treatment Systems. Appl ied and Environmental Microbiol ogy 66 (5):21542165. OToole, G., Kolter R 1998. Initiation of biofilm formation in Pseudomonas fluorescens WCS365 proceeds via multiple, convergent signaling pathways: a genetic analysis. Molecular Microbiology 28:449 461. Pard i eck, D., Bouwer, E ., Stone, A 1992. Hydrogen peroxide use to increase oxidant capacity for in situ bioremediation of contaminated soils and aquifers: A review. Journal of Contaminant Hydrology 9:221 242. Pernitsky, D. Finch, G., Huck P 1995. Disinfection Kinetics of Heterotrophic Plate Count Bacteria in Biologically Treated Potable Water. Water Research, 29(5):12351241 Prevost, M., Coallier, J., Mailly J 1995. Removal of Various Biodegr adable Organic Compounds by First and Second Stage Filtration. Proc. 12th Ozone World Congress, Lille, France. Price, M., Enos, R. Hook, A., Hermanowicz, S 1993. Evaluation of ozone/biological treatment for disinfection byproducts control and biologically stable water. Ozone: Science and Engineering. 15: 95 130. Priester, J., Ho rst, A., Van De Werfhorst, L., Saleta J., Mertes L., Holden, P. 2006. Enhanced visualization of microbial biofilms by staining and environmental scanning electron mic roscopy. Journal of Microbiological Methods 68:557587. Quelas, J. Lpez Garca, J., Casabuono, A., Althabegoiti, M., Mongiardini, E., Prez Gimnez, J., Couto, A., Lodeiro, A. 2006 Effects of N starvation and C source on Bradyrhizobium japonicum exopolysaccharide production and composition, and bacterial infectivity to soybean roots Arch Microbiology 186:2:119. Ralebitso T ., Senior E. van Verseveld, H 2004. Microbial aspects of atrazine degradation in natural environments. Biodegradation 13: 11. Raymond, M., Hozalski, A., Goel, S., Edw ard, J. 1995. TOC Removal in Biological Filters. Journal AWWA 87(12): 4054 Rinck Pfeiffer, S., Ragusa, S., Sztajnbok, P., Vandevelde, T. 2000. Interrelationships between biological, chemical, and physical proces ses as an analog to clogging in aquifer storage and recovery (ASR) wells. Water Research, 34(7): 21102118.

PAGE 227

227 Rittmann, B 1995. Transformation of organic micropollutants by biological processes. The Handbookof Environmental Chemistry, Quality, and Treatment of Drinking Water (J. Hrubec, editor), 5(B), 31. Rittmann, B., McCarty, P 2001. Environmental Biotechnology McGraw Hill, Boston, Massachusetts. Ryu, J. Beuchat L 2004. Biofilm Formation by Escherichia coli 0157:H7 on Stainless Steel: Effect of Exopolysaccharide and Curli Production on Its Resistance to Chlorine Applied and Environmental Microbiology 71(1): 247254. Ryu J., Kim, H., Beuchat L. 2004. Attachment and Biofilm Formation by Escherichia coli O157:H7 on Stainless Steel as Influenced by Exopolysaccharide Production, Nutrient Availability, and Temperature. Journal of Food Protection, 67(10):212331. Sahabi, D Takeda, M. Suzuki I., Koizumi, J. 2009. Removal of Mn2+ from water by aged biofilter media: The role of cataly tic oxides layers Journal of Bios cience and Bioengineering, 107 ( 2 ): 151 157 Sang J., Zhang, X Li L., Wang, Z 2003. Improvement of Organics Removal by BioCeramic Filtration of Raw Water with Addition of Phosphorus. Water Research, 37(19):47114718. Scholz M ., Martin R 1997. Ecological equilibrium on biological activated carbon. Water Research, 31(12): 2959 2968. Schumb, W., Stratterfield, C ., Wentworth, R. 1955. Hydrogen peroxide. Van Nostrand Reinhold, New York, N. Y. Sekar, R. Nair, K., Rao, V., Venugopalan, V. 2002. Nutrient dynamics and successional changes in a lentic freshwater biofilm Freshwater Biology 47( 10 ): 18931907 Servais, P., Billen, G., Bouillot, P., 1994. Biological colonization of granular activated carbon filters in drinking water treatment. Journal of Environmental Engineering, 120 (4), 888 899. Siering, P., Ghiorse, W 1997. Development and appl ication of 16S rRNA targeted probes for detection of ironand manganeseoxidizing sheathed bacteria in environmental samples. Applied Environmental Microbiology 63:2:644. Simpson, D ., 2008. Biofilm processes in biologically active carbon water purificat ion. Water Research 42 (2008) 2839 2848.

PAGE 228

228 Skorupska, A., Janczarek, M. Marczak, A. Mazur, Krl J 2006. Rhizobial exopolysaccharides: genetic control and symbiotic functions. Microbial Cell Factories 5:7. Snyder E., Pleus, R., Snyder, S 2005. Pharmaceuticals and EDCs in the US water industry An update. Journal AWWA 97(11): 3236 Snyder, S., Wert, E ., Lei, H., Westerhoff, P., Yoon, Y. 2007 Removal of EDCs and pharmaceuticals in drinking and reuse treatment processes. Final Report (#91188), Water Research Foundation Final Report, Denver, CO: Snyder, S ., Westerhoff, P., Yoon, Y., Sedlak, D 2003, Pharmaceuticals, personal care products, and endocrine disruptors in water: Implications for the water industry. Environmental Engineering Science, 20:449469. Stratton, R., Namkung, E., Rittmann B 1983. Biodegradation of Trace Organic Compounds by Biofilms on Porous Medium. Journal AWWA 75:463. S tucki G., Yu, C., B aumgartner T., Gonzalez V alero, J.1995. Microbial atrazin e mineralisation under c arbon limited and denitrifying conditions Water Research, 29 (1): 291296 Suffet, I., Ho, J., Chou, D., Khiari, D., Mallevialle, J., Kawczynski, E. 1995. Advances in Taste and Odor Treatment and Control. American Water Works Association. Denver, CO Sutherland, I., 1977. Surface Carbohydrates of the Prokaryotic Cell. Academic Press, New York. Sutherland, I 2001. Exopolysaccharides in biofilms, flocs and related structures. Water Science and Technology 43 (6): 77 86 Swertf eger, J 1996. Biological Degradation of Ozone Byproducts During Drinking Water Treatment. M.S. thesis, Department of Civil and Environmental Engineering, University of Cincinnati, U.S. Tchobanoglous, G., Burton F 1991. Wastewater Engineering: Treatment, Disposal and Reuse. 3rd Edition, McGraw Hill Inc., New York. Thurman, E., Ferrer, I., Zweigenbaum, J ., 2006, Automated screening of 600 pesticides in food by LC/TOF MS using a molecular feature database search: Agilent Application Note 59895496. Trudgill, P ., 1984. Microbial degradation of the alicyclic ring: structural relationships and metabolic pathways. In: Gibson, D.T. (Ed.), Microbial Degradation of Organic Compounds. Marcel Dekker Inc., New York, USA, pp. 131 180.

PAGE 229

229 Tyupalo, N., Dneprovkii Y 1981. Studying the Effects of Ozone with Iron (II) Ions in Aqueous Solutions. Zhurnal Neogranicheskoi Khimi, 26(3): 664667. U nited States EPA. 1990. Technologies for Upgrading Existing or Designing New Drinking Water Treatment Facilities. EPA/625/489/023. Cincinnati, OH. Unitie d States EPA 1991. Site Survey Characterization for Sub Surface Remediation, EPA/625/R 91/026, Office of Research and Development, U.S. EPA, Washington D.C. United States EPA 2004. Drinking Water Health Advisory for Manganese. EPA 822R 04003 Office of Water (4304T) Health and Ecological Criteria Division, U.S. EPA, Washington, DC Urfer, D., Huck, P., Booth, S., Coffey B 1997. Biological Filtration for BOM and Particle Removal: A Critical Review. Journal AWWA 89(12):8398. Urfer, D., Huck P., Gagnon G 1999. Modeling enhanced coagulation to improve ozone disinfection Journal AWWA 91(3): 59 73 Vokes, C 2007. Impact of Ozone and Biological Filtration on Water Quality Parameters in Arlington, Texas. Ozone Science and Engineering 29:2 61271. Volk C., LeChevallier M. 1999. Impacts of the Reduction of Nutrient Levels on Bacterial Water Quality in Distribution Systems. Applied and Environmental Microbiology 65(11): 49574966. Walling, C. 1975. Fenton's Reagent Revisited. Accounts of Chem. Research, 8:125131 Wang, J. Summers, R. Miltner R. 1995. Biofiltration Performance: Part 1, Relationship to Biomass. Journal AWWA 87(11):5563. Wani A., Surakasi V ., Siddharth J ., Raghavan R., Patole, M ., Ranade D., Shouche Y 2006. Molecular analyses of microbial diversity associated with the Lonar soda lake in India: an impact crater in a basalt area. Research Microbiology 157: 928 937. Wanner, O., Eberl, H., Morgenroth, E., Noguera, D., Picioreanu, C., Rittmann, B., van L oosdrecht. M. 2006. Mathematical Modeling of Biofilms. IWA Publishing, London, UK. Scientific and Technical Report Series Report No. 18. Water Environment Federation. 2009. Manual of Practice No. 8: Design of Wastewater Treatment Plants, Updated Edition in Development.

PAGE 230

230 Water Environment Federation. 1998. Manual of Practice No. 8: Design of Wastewater Treatment Plants, 4th Edition. Wawrik B., Kerkhof, L. J., Kukor, J., Zylstra, G. 2005. Effect of different carbon sources on community compostition of bac terial enrichments from soil. Applied and Environmental Microbiology, 71: 67766783. Wert E., Neemann J., Rexing D., Zegers, R 2008. Biofiltration for removal of BOM and residual ammonia following control of bromate formation. Water Research, 42: 372378 Westerhoff, P., Summers, R. Chowdhury, Z., Kommineni S. 2005. OzoneEnhanced Biofiltration for Geosmin and MIB Removal. Water Research Foundation Final Report. Denver, CO. Wright, H. 1995. Characterization of Soybean Peroxidase for the Treatment of P henolic Wastewaters. Masters Thesis, Mcgill University, Montreal. Xi, C. 2009. Personal communication with Chuanwu Xi, Professor at University of Michigan, September 26, 2009. Zappia, L., Warton, B., Alessandrino, M., Scott, D., Wylie, J., Heitz, A Hiller, B Masters, D Nolan, P. Thiel, P Kagi, R Joll, C., Franzmann, P 2007. Pilot Scale Testing of Biofilter Post Treatment of MIEX Treated Water. Journal of Water Supply: Research and Technology AQUA, 56(4):217232. Zhang S., Huck P 1996a. Removal of AOC in biological water treatment processes: a kinetic modeling approach. Water Research, 30(5),11951207. Zhang S., Huck P 1996b. Biological Water Treatment: A Kinetic Modeling Approach. Water Research 30(2):456464.

PAGE 231

231 BIOGRAPHICAL SKETCH Chance Lauderdale is third generation graduate of the University Florida. He completed his Bachelor of Science and Master of Science degree in the Department of Environmental Engineering Sciences in 20 01 and 2004, respectively. He received his Doctor of Philosophy from the University of Florida in the summer of 2011. Chance has been employed as an engineer with Carollo Engineers since 2004. A majority of Chances work has focused on process development for biological drinking water treatment.