USF Libraries
USF Digital Collections

Constructed wetland/filter basin system as a prospective pre-treatment option for aquifer storage and recovery and a pot...

MISSING IMAGE

Material Information

Title:
Constructed wetland/filter basin system as a prospective pre-treatment option for aquifer storage and recovery and a potential remedy for elevated arsenic
Physical Description:
Book
Language:
English
Creator:
Lazareva, Olesya
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Florida
Constructed wetland
Wastewater treatment
Clay settling area
Geogenic arsenic
Isotope mass-balance
Reactive transport modeling
Dissertations, Academic -- Geology -- Masters -- USF   ( lcsh )
Genre:
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The efficiency to improve the water quality of industrial and municipal wastewater in a constructed wetland/filter basin treatment system was investigated. The wetland system was constructed in a closed phosphate mine used for clay settling and sand tailings in Polk County, Florida. During 18-months of monitoring the chemical/microbiological composition of treated wetland water remained relatively constant, despite significant seasonal variations in temperature, rainfall and humidity. The following changes in water quality between input and output were observed: substantial decrease of water temperature (up to 10 °C), reduction of As, SO4, F, Cl, NO3, NO2, Br, Na, K, Ca, and Mg, change in pH from 9 to 6.5-7, increase of H2S (up to 1060 &#956;g/L), and a change from positive to negative ORP. There were no exceedances of the primary drinking water standards, volatile organic compounds, synthetic organic compounds, and radionuclides, but a number of exceedances for the secondary drinking water standards (Al, F, Fe, Mn, color, odor, total dissolved solids, and foaming agents). The concentration of fecal and total coliform bacteria in the wetland water was high, butsubsequently reduced during filtration in the filter basin from 30 - 730 and 1000 - 7000 count/100 mL to < 2 and < 100 count/ 100 mL, respectively. To resolve the complex hydrogeological conditions a combined isotope/chemical mass-balance approach was applied. The results were the following: (1) the composition of water in the wetland varied throughout the period of the study; (2) a change in isotopic composition along the wetland flow path; (3) the wetland contained mainly wastewater (88 - 100 %) during normal pumping operations; however, hurricanes and inconsistent pumping added low conductivity water directly and triggered enhanced groundwater inflow into the wetland of up to 78 %; (4) the composition of water in monitor wells was mostly groundwater dominated; however periodically seepage from a water body to the north was detected; and (5) seepage from adjacent water bodies into the wetland was not identified during operation, which would indicate a potential water loss from the wetland. To test if the wetland system could be a prospective pre-treatment option for water used in aquifer storage and recovery (ASR) scenarios, a set of bench-scale leaching experiments was carried out using rocks from the Avon Park Formation, the Suwannee Limestone and the Ocala Limestone. Since As in the Floridan Aquifer was mainly present as an impurity in the mineral pyrite the elevated iron and sulfide concentrations in the wetland water were thought to prevent pyrite dissolution. The experiments which covered a range of redox conditions showed that the amount of As released from the aquifer matrix was not perfectly correlated with the bulk rock As concentration, nor the redox state of the water. The following important results were obtained: (1) the highest concentration of As was leached from the Avon Park Formation and the lowest - from the Suwannee Limestone, although the Ocala Limestone had the lowest bulk rock As; (2)minor to no As was released using native Floridan groundwater; (3) Tampa tap water, which chemically and physically resembled the ASR injection water, caused the As leaching of up to 27 &#956;g/L, which was higher than the As drinking water standard; (4) the wetland and filter basin waters caused the highest release of As (up to 68 &#956;g/L), which was unexpected because those water types were less oxygenated than Tampa tap water and thus should be less aggressive; (5) the in-situ filtration of the wetland water through a 0.2 &#956;m membrane resulted in a reduction of As from 30 &#956;g/L to 16 &#956;g/L; and (5) the UV treatment significantly reduced both fecal and total coliform bacteria, but facilitated the increase of DO in initial waters, a change from negative to positive ORP, and the increase of As concentration in leachates. The experiments confirmed that perturbations of native aquifer conditions caused the release of As from the Floridan aquifer matrix, although the reaction may not be as simple as the dissolution of pyrite by oxygen, but additionally governed by a complex set of factors including the ORP of the system, SO42-/S2, Fe3+/Fe2+, dissolved organic carbon and microbial activity. In addition, the trend of As leaching could be governed by a set of factors, such as the porosity and permeability of the aquifer matrix influencing the rate and degree of free water saturation, amount of pyrite to be exposed to the preferential water flow paths, limited surface reactivity of pyrite with favored reactions on fractured mineral surfaces, the concentration and the selective leaching of As from individual pyrite crystals. To characterize and verify the geochemical processes in the column experiments, the Geochemist's Workbench reactive transport models (React and X1t) were developed. Results from the models correlated well to those from the column experiments andconfirmed the following: (1) the water-rock reaction between the aquifer matrix and native groundwater was favorable for pyrite stability preventing the release of As into solution; (2) the injection of oxidizing surface water into reducing native groundwater caused a change in redox potential of the system thus promoting the dissolution of pyrite, and (3) 1D reactive transport model of water-rock reaction between the aquifer matrix and surface water indicated a diverse behavior of As along the column, such as the oxidative dissolution of pyrite, mobilization and simultaneous sorption of As onto neo-formed HFO, followed by the reductive dissolution of HFO and secondary release of adsorbed As, and the potential non-oxidative dissolution of pyrite contributing the additional source of As to the solution.
Thesis:
Dissertation (PHD)--University of South Florida, 2010.
Bibliography:
Includes bibliographical references.
System Details:
Mode of access: World Wide Web.
System Details:
System requirements: World Wide Web browser and PDF reader.
Statement of Responsibility:
by Olesya Lazareva.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains X pages.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
usfldc doi - E14-SFE0004550
usfldc handle - e14.4550
System ID:
SFS0027865:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Constructed Wetland/Filter Basin System as a Prospective Pre Treatment Option for Aquifer Storage and Recovery and a Potential Remedy for Elevated Arsenic b y Olesya Lazareva A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Geology College of Arts and Sciences University of South Florida Major Professor: Thomas Pichler, Ph.D. Mark Stewart, Ph.D. Mark C. Rains, Ph.D. Gregory Druschel, Ph.D. Jonathan Arthur, Ph.D., P.G. Date of Approval: June 11, 2010 Keywords: Florida c onstructed wetland w astewater treatment c lay settling area geogenic a rsenic isotope mass balance, reactive transport modeling Copyright 2010 Olesya Lazareva

PAGE 2

ACKNOWLEDGEMENTS First of all I would like to gratefully acknowledge the help, support and encouragement of Dr. Thomas Pichler His advising and contribution to my improvement as a geoc hemist cannot be underestimated I am very thankful for his endles s effort to provide invaluable assistance with the development of this research and the re vision of the Dissertation. Thank you to my dissertation committee, Dr s Mark Stewart Mark C. Rains, Gregory Druschel and Jonathan Arthur for reviewing the manuscr ipt and providing very useful comments and valuable suggestions. This project was funded through a grant from the F IPR and SWFWMD with the assistance of Progress Energy Florida to a collaboration of Dr. Thomas Pichler and Schreuder Inc. I would like to tha nk Peter Schreuder, Dana Gaydos and staff from the Hines Energy Complex for the provided core material, rainfall data and the assistance during field sampling Mr. Wagner is thanked for his help during the USF groundwater sampling. Dr. Jonathan Wynn is tha nked for his assistance with the analysis and interpretation of the isotope data. Dr. Thorsten Dittmar is thanked for the analysis of dissolved organic carbon and total nitrogen. I would like to greatly appreciate Mr. Greg Jones for his considerable help usin Finally I am grateful to my family and friends Without their help, constant support, patience, advice and inspiration this work would never happen.

PAGE 3

i TABLE OF CONTENTS LIST OF TABLES iv LIST OF FIGURES vi ABSTRACT xi CHAPTER ONE: INTRODUCTION 1 1.1. Constructe d wetland for water reclamation 2 1.2. Constructed wetland for aquifer storage and recovery (ASR) 4 1.3. Review of arsenic occurrence and toxicity 8 1.4. Description of the study area 12 1.5. Geology 18 1.6. Objectives 24 1.7. Arrangement of D issertation 25 CHAPTER TWO: LONG TERM PERFORMANCE OF A CONSTRUCTED WETLAND/FILTER BASIN SYSTEM TREATING WASTEWATER IN COMPLEX HYDROG EOLOGICAL SETTINGS 27 2.1. Introduction 27 2.2. Methods and sampling procedure 2 8 2.2.1. Field and laboratory analysis 31 2.3. Results 36 2.3.1. Monitor wells 36 2.3.2. Precipitation measurements 36 2.3.3. Surface water level measurements 42 2.3.4. Evaluation of water quality along the wetland flow path 42 2.3.4.1. Field measurements 45 2.3.4.2. Laboratory analysis 49 2.3.5. Wetland/filter basin water quality monitoring 50 2.3.5.1. Primary, secondary drinking water standards, volatile organic compounds, synthetic organic comp ounds, and radionuclides 52 2.3.5.2. Fecal and total coliform 52 2.3.6. Isotopic composition 59 2.3.6.1. Wetland water 60 2.3.6.2. Monitor wells SA 8 and N 15 62

PAGE 4

ii 2.4. Discussion 66 2.4.1. Behavior of arsenic 71 2.4.2. Evaluation of wetland performance using isotopic mass balance approach 77 2.4.2.1. Water composition in the wetland 77 2.4.2.2. Water composition in monitor wells 81 2.5. Conclusions 92 CHAPTER THREE: BENCH SCALE LEACHING EXPERIMENTS TO INVESTIGATE GEOGENIC ARSENIC CONTAMINATION IN THE FLORIDAN AQUIFER 9 5 3.1. Introduction 95 3.2. Methods 98 3.2.1. Rock preparation and analysis 98 3.2.2. Water collection and analysis 100 3.2.3. Experimental Setup 103 3.3. Results 103 3.3.1. Rock chemical composition 107 3.3.2. Leaching experiments 111 3.3.2.1. Suwannee Limestone 111 3.3.2.2. Ocala Limestone 123 3.3.2.3. Avon Park Formation 126 3.4. Discussion 14 1 3.4.1. Importance of dissolved organic carbon for arsenic mobilization 148 3.5. Conclusions 149 CHAPTER FOUR: GEOCHEMICAL REACTIVE TRANSPORT MODELING 15 2 4.1. Introduction 15 2 4.2. Methods 15 3 4.3. Results and Discussion 156 4.4. Conclusions 17 4 CHAPTER FIVE: SUMMARY AND CONCLUSIONS 176 REFERENCES 180 APPENDIX A. AN 18 MONTH PERFORMANCE STUDY OF THE WETLAND/FILTER BASIN TREATMENT SYSTEM 20 3 APPENDIX B. CHEMICAL AND ISOTOPIC COMPOSITION OF WATERS U SED FOR THE MASS BALANCE APPROACH 216

PAGE 5

iii APPENDIX C. CHEMICAL AND MINERALOGICAL ANALYSIS OF THE SAMPLES FROM THE SUWANNEE LIMESTONE (IN MG/KG). 223 APPENDIX D. CHEMICAL AND MINERALOGICAL ANALYSIS OF THE SAMPLES FROM THE OCALA LIMESTONE (IN MG/KG). 226 APPENDIX E. CHEMICAL AND MINERALOGICAL ANALYSIS OF THE SAMPLES FROM THE AVON PARK FORMATION (IN MG/KG). 229 APPENDIX F. DAT A FROM THE BENCH SCALE LEACHING EXPERIMENTS INCLUDING THE COMPOSITION OF ORIGINAL WATERS AND RECOVERED LEACHATES. 231 APPENDIX G. GEOCHEMICAL REACTIVE TRANSPORT MODELING DATA 246 ABOUT THE AUTHOR E N D P A G E

PAGE 6

iv LIST OF TABLES Table 1. Maximum, minimum, arithmetic mean and standard deviation of analyzed parameters in the cooling pond, wetland pump and surface, filter basin south and north pumps 3 7 Table 2. Maximum, minimum, arithmetic mean and standard deviation of analyzed parameters in monitor wells MW 1 to MW 6 and water 39 Table 3. Change in water composition from the cooling pond (CP) to the wetland from pump (WP) during the dry (March 2007) and rainy (September 2007) seasons 46 Table 4. The wetland performance evaluated with the percent removal of each analyzed parameter 53 Table 5. Analysis of Primary Dr inking Water Standards (PDWS) 55 Table 6. Analysis of Secondary Drinking Water Standards (SDWS) 56 Table 7. Analysis of synt hetic organic compounds (SOC) 5 7 Table 8. Analysis of Volatile Organic Compounds (VOC) and Radionuclides 5 8 Table 9. Minimum, maximum, mean, median and standard deviation values of 63 Table 10. Chemical and isotopic data of waters used for the mass balance approach 83 Table 11. Chemical and isotopic data of the monitor wells (MW 1 to MW 6) used for the mass balance approach 87 Table 12. Maximum, minimum, mean and standard deviation of As concentrations in mg/kg for the Suwannee Limestone, Ocala Limestone and Avon Park Formation 109 Table 13. List of rock samples selected for the leaching experiments 112

PAGE 7

v Table 14. Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Suwannee Limestone (136 m to 139 m) using different types of water 11 6 Table 15. Initial water (before injection) and leachate recovered from 1 m columns filled with the Suwannee Limestone (171 to 174 m) us ing different types of water 119 Table 16. Initial water (before injection) and leachate quality recovered from 0.3 m columns filled with the Ocala Limestone (195 to 199 m) us ing different types of water 124 Table 17. Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Avo n Park Formation (255 m to 257 m) using different types of water 128 Table 18. Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Avon Park Formation (260 m to 261 m) us ing different types of water 138

PAGE 8

vi LIST OF FIGURES Figure 1. Stability diagram of pyrite and Fe(OH)3 in water at 25C and 1 atmosphere total pressure 7 Figure 2. Map of the study area located at the Hines Energy complex (Polk County, Central Florida) including water transfer system from the cooling pond to the U shaped constructed wetland and filter basin treatment system 13 Figure 3. Schematic concept diagram of the constructed wetland/filter basin treat ment system 14 Figure 4. Photographs of the constructed wetland treatment system 15 Figure 5. Photographs of the filter basin system 17 Figure 6. Lithostratigraphic and hydrogeologic units of the study area 19 Figure 7. Geological cross sections of southwestern Florida 20 Figure 8. Water sampling from the wetland pumping station an d wetland surface (top), and monitor wells (bottom ) 30 Figure 9. Evaluation of water quality along the wetland flow path 32 Figure 10. Water sampling and analysis along the wetland flow path 33 Figure 11. (A) Average values of SO 4 Fe, F, and As and (B) Average and variance of Na estimated by ANOVA at MW 1 to MW 6, N 15 and SA 8 41 Figure 12. Precipitation hydrograph recorded at the study area from May 1, 2006 to October 31, 2007, covering one dry and two rainy seasons 43 Figure 13. Water leve l elevations at the CP, WS, N 15 and SA 8 above National Geodetic Vertical Datum (NGVD) 44

PAGE 9

vii Figure 14. (A). Evaluation of T, pH, ORP, DO, H2S, SO4, conductivity, and As along the wetland flow path during the dry (March 2 007) and rainy (September 2007) sea sons; (B): Distribution of Na and Cl along t he wetland flow path and at the MWs during the dry (March 2007) and rainy (September 2007) seasons 4 7 Figure 15. Distribution of pH, ORP, and temperature (T) in the CP input and WP output waters 5 1 Figure 16. Fecal and total coliform bacteria detected at the CP, WP, FBS and at the FBN. FBS/FBN had the lowest fecal and total coliform confirming the crucial rol e 61 land (A) and, SA 8 and N 15 ( B) 64 65 monitor wells 67 Figure 20. Mass fluxes of Na and Cl in the CP and WP, and the calculated percent re moval of Na and Cl from the wetland 70 Figure 21. Variation of SO4/Cl in the cooling pond pump (CP) and the wetland pump (WP) as a function of time 72 Figure 22. Distribution of Fe at the WP, FBS and FBN with time 7 4 Figure 23. Mass fluxes of As in the CP and WP 75 Figure 24. (A) Map of the study area including water transfer system from the cooling pond to the U shaped constructed wetland; (B) Cross section along transect 1 79 Fig u re 25. Four end members for the WP used for the mass balance approach 82 Figure 26. Three end members for the monitor wells MW 1 to MW 3 (A), and MW 4 to MW 6 (B) used for the mass balance approach 8 6 Figure 27. The calculated mass of each end member in the wetland using an isotope/chemical mass balance approa ch 89 Figure 28. Evaluation of wetland surface water isotopic composition along the flow path 93

PAGE 10

viii Figure 29. Concentrations of As and Ca through time for the ASR Punta Gorda cycle during aquifer storage and recovery (ASR) operation in Charlotte County, southern Florid a 97 Figure 30. Collection of the wetland water (WW) from a depth of 0.5 m using a peristaltic pump (top).Bench scale column leaching experiments using WW (bottom) 102 Figure 31. Bench scale leaching experiments using wetland water (WW) treated with UV (top) and drinking water (DW) (bottom) 104 Figure 32. Photomicrograph and the scanning electron micrograph of framboidal and euhedral pyrites (shown by arrows) found in the Suwannee Limestone (136 m to 139 m) 106 Figure 33. (A) Photomicrograph and (B) the scanning electron micrograph of framboidal pyrites found in the Ocala Limestone (195 m to199 m); (C) EDS spectra confirming the presence of pyrite 107 Figure 34. Photomicrograph and the scanning electron micrograph of framboi dal and euhedral pyrites found in the Avon Park Formation (255 m to 257 m) 108 Figure 35. The calculated amount of pyrite and measured bulk rock As concentration in the Suwannee and Ocala Limestones, and the Avon Park Formation 113 Figure 36. Correlation between calculated amount of pyrite and measured bulk rock As concentration for the analyzed rock intervals with (top) and without the outlier (bottom) 114 Figure 37. As, SO4, pH, and ORP in leachates recovered from the Suwannee Limestone (136 m to 139 m) using 0.3 m columns 117 Figure 38. As, SO4, pH, and ORP in leachates recovered from the Suwannee Limestone (171 m to 176 m) using 1 m columns 12 0 Figure 39. Distribution of Ca versus pH (top) and As versus pH (bottom) in leachates recovered from the Suwannee Limestone (171 m to 176 m) using 1 m columns 121 Figure 40. As, SO4, pH, and ORP in leachates recovered from the Ocala Limestone (195 to 199 m) using 0.3 m columns 125

PAGE 11

ix Figure 41. As, SO4, pH, and ORP in leachates recovered from the Avon Park Formation (255 m to 257 m) using 0.3 m columns 129 Figure 42. Dissolved organic carbon (DOC) and total nitrogen (TN) in the groundwater (GW), wetland (WW) and filter basin water (BW) before and after UV 130 Figure 43. (A) Dissolved oxygen (DO) and oxidation reduction potential (ORP); and (B) Fecal and total coliform of the wetland (WW) and filter basin water (BW) before and after UV 131 Figure 44. Arsenic (As) in the initial water and first leachate recovered from the Avon Park Formation (255 m to 257 m) using wetland (WW) and filter basin water (BW) before and after UV, and drinking water (DW) 132 Figure 45. Arsenic, pH and ORP in leachates recovered from the Avon Park Formation (255 m to 257 m) using drinking water (0.3 m and 0.5 m columns) for a period of four months at three weeks interval 133 Figure 46. Distribution of SO4, As species and As total concentrations in leachates recovered from the Avon Park Formation (255 m to 257 m) from the consecutive injection of filter basin (BW) and drinking (DW) waters into a 0.3 m column 135 Figure 47. As, SO4, pH, and ORP in leachates recovered from the Avon Park Formation (260 m to 261 m) using 0.3 m columns 139 Figure 48. Distribution of Ca versus pH (top) and As versus pH (bottom) in leachates recovered from the Avon Park Formation (260 mto 261 m) using 0.3 m columns 140 Figure 49. Distribution of maximum As (top) and SO4 (bottom) concentrations in leachates recovered from each interval (with the subtraction of initial concentration) 14 2 Figur e 50. Distribution of ORP and As in leachates recovered during a different phases of the bench scale leaching experiments 145 Figure 51. Distribu tion of Eh (A) and pH (B) during simulated injections of model (1) 157 Figure 52 Distribution of As in fluid (A) and As species (A) during simulated injections of model (1) 158

PAGE 12

x Figure 53 Distribut ion of Eh (A) and pH (B) during si mul ated injections of model (2) 160 Figure 5 4 Distribut ion of As in fluid (A) and As species (B) during si mulated injections of model (2 ) 16 1 Figure 5 5 Distribution of Eh (A) and pH (B) during si mulated injections of model (3) 163 Figure 5 6 Distribut ion of As in fluid (A) and As species (B) during si mulated injections of model (3) 164 Figure 57 Distribution of mineral saturation states (A), Eh (B),and pH (C) along the column during 5 hours of simulated injections of model (4) 16 6 Figure 58 Distribution of As concentration in fluid along the column with di scharge rate of 0.01cm/sec (A) and 0.005 cm/sec (B ) during 5 hours of si mulated injections of model (4) 167 Figure 59 Distribution of As species along the column during 1 hour (A), 5 hours (B), and 10 hours (C) of simulated injections of model (4) 168 Figure 60 Distribution of As species (A), Eh (B), and pH (C) along the column during 1 hour of si mulated injections of model (4) 170 Figure 61. Distribution of pH (A) and Eh (B) during simulated injections of model (4) 172 Figure 62 Distribution of As in fluid (A) and As species (B) during simulated injections of model (4) 173

PAGE 13

xi CONSTRUCTED WETLAND/FILTER BASIN SYSTEM AS A PROSPECTIVE PRE TREATMENT OPTION FOR AQUIFER STORAGE AND RECOVERY AND A POTENTIAL REMEDY FOR ELEVATED ARSENIC OLESYA LAZAREVA ABSTRACT T he efficiency to improve the water quality of industrial and municipal wastewater in a constructed wetland/filter basin treatment system was investigated The wetland system was constructed in a closed phosphate mine used for clay settling and sand tailings in Polk County, Florida. During 18 month s of monitoring the chemical/microbiological composition of treated wetland water remained relatively constant despite significant seasonal variations in temperature, rainfall and humidity The following changes in water quality b etween input and output were observed : substantial decrease of water temperature (up to 10 C), reduction of As, SO 4 F, Cl, NO 3 NO 2 Br, Na, K, Ca, and Mg, change in pH from 9 to 6.5 7, increase of H 2 S (up to 1060 a change from positive to negative ORP. There were no exceed a nces of the primary drinking water standards, volatile organic compounds, synthetic organic compounds, and radionuclides, but a number of exceed a nces for the secondary drinking water standards (Al, F, Fe, Mn, color, odor, total dissolved solids, and foaming agents). The concentration of fecal and total coliform bacteria in the wetland water was high, but

PAGE 14

xii subsequently reduced during filtration in the filter basin from 30 730 and 1000 7000 count/100 mL to < 2 and < 100 c ount/ 100 mL, respectively. To resolve the complex hydrogeological conditions a combined isotope/chemical mass balance approach was applied. The results were the following: (1) the composition of water in the wetland varied throughout the period of the study; (2) a change in isotopic composition along the wetland flow path; (3) the wetland contained mainly wastewater (88 100 %) during normal pumping operations; however, hurricanes and inconsistent pumping added low conductivity water directly and trigg ered enhanced groundwater inflow into the wetland of up to 78 %; (4) the composition of water in monitor wells was mostly groundwater dominated; however periodically seepage from a water body to the north was detected ; and (5) seepage from adjacent water b odies into the wetland was not identified during operation which would indicate a potential water loss from the wetland. To test if the wetland system could be a prospective pre treatment option for water used in aquifer storage and recovery (ASR) scenari os, a set of bench scale leaching experiments was carried out using rocks from the Avon Park Formation, the Suwannee Limestone and the Ocala Limes tone. Since As in the Floridan A quifer wa s mainly present as an impurity in the mineral pyrite the elevated ir on and sulfide concentrations in the wetland water were thought to prevent pyrite dissolution. The experiments which covered a range of redox conditions showed that the amount of As released from the aquifer matrix was not perfectly correlated with the bulk rock As concentration nor the redox state of the water The following important results were obtained: (1) t he highest concentration of As was leached from the Avon Park Formation and the lowest from the Suwannee Limestone, although the Ocala Limes to ne had the lowest bulk rock As; (2 )

PAGE 15

xiii m inor to no As was released us ing native Floridan groundwater; (3 ) Tampa tap water, which chemically and physically resembled the ASR injection water, caused the As leaching of up to 27 g/L, which was higher than the As drinking water standard; (4 ) the w etland and filter basin waters caused the highest release of As (up to 68 g/L), which was unexpected because those water types were less oxygenated than Tampa tap water and th us should be less aggressive; (5 ) the i n situ filtra tion of the wetland water through a 0.2 m membrane resulted in a reduction of As from 30 g/L to 16 g/L ; and (5) the UV treatment significantly reduced both fecal and total coliform bacteria, but facilitated th e increase of DO in initial wat ers, a change from negative to positive ORP and the increase of As concentration in leachates. The experiments confirmed that perturbations of native aquifer conditions caused the r elease of As from the Floridan a quifer matrix, although the reaction may not be as simple as the dissolution of pyrite by oxygen, but additionally governed by a complex set of factors including the ORP of the system, SO 4 2 /S 2 Fe 3+ /Fe 2+ dissolved organic carbon and microbial activity. I n addition, the trend of As leaching could be governed by a set of factors, such as the porosity and permeability of the aquifer matrix influencing the rate and degree of free water saturation, amount of pyrite to be exposed to the preferential water flow paths limited surface reactivity of pyri te with favored reactions on fractured mineral surfaces, the concentration and the selective leaching of As from individual pyrite crystals To characterize and verify the geochemical processes in the column experiments the Geochemist's Workbench reactive transport models (React and X1t) were developed Results from the models correlated well to those from the column experiments and

PAGE 16

xiv confirmed the following: (1) the water rock reaction between the aquifer matrix and native groundwater was favorable for pyrite stability preventing t he release of As into solution; (2) t he injection of oxidizing surface water into reducing native groundwater caused a change in redox potential of the system thus promoting the disso lution of pyrite, and (3) 1D reactive transport model of water rock reaction between the aquifer matrix and surface water indicated a diverse behavior of As along the column, such as the o xidative dissolution of pyrite, mobilization and simultaneous sorption of As onto neo formed HFO, followed by the r eductive dissolution of HFO and s econdary release of adsorbed As, and the p otential non oxidative dissolution of pyrite contributing the additional source of As to t he solution.

PAGE 17

1 C HAPTER ONE INTRODUCTION The sustainability of water resources is rapidly becoming to be a major concern worldwide. Ujang (2009) characterized water sustainability as the capability of an ecosystem to maintain ecological functions, processes, productiv ity and biodiversity of water resources for future generations. At the present time, about 35 % of anthropogenic water use is considered unsustainable due to environmental contamination, unsanitary conditions, progressively depleting ground and surface wa ter resources ( Clarke and King, 2004 ). Access to clean and safe drinking water has developed into a luxury for more than one billion people (Watkins, 2006). Particularly, the excessive groundwater pumping from the Floridan Aquifer, due to growing water dem ands in highly populated areas of the Florida (Jones and Owen, 2001), causes lowering of the water table, thus affecting spring flows and lake levels, saltwater intrusion along the coast, and land subsidence (Peck et al., 2005). Yearly, about 3x10 12 L of g roundwater in Florida is withdrawn by pumping (Scott, 2001a). Therefore, the reduction of water consumption as well as the development of water reuse, reclamation or recycling technologies could provide more sustainable alternative to the extensive groundw ater consumption (Alley et al., 1999) The development of constructed wetlands for water treatment is a cost

PAGE 18

2 effective process, which provide water quality suitable for reuse and thus may become an alternative supply of clean water (House et al., 1999). 1.1. Co nstructed wetland for water reclamation Phosphate mining in central Florida annually disturbs about 15 25 km 2 of land through generation of clay settling areas (CSA), open mine pits, phosphogypsum stacks, and tailing sand deposits (FIPR). Of these the CSA and tailing sand deposits are of particular interest to the wetland based water treatment approach. Generally, constructed wetlands (CWs) are defined as artificial wastewater treatment systems composed of a shallow basin filled with substrate, such as soil or gravel, and planted with vegetation tolerant of saturated soil conditions (EPA, 2000; Davis, 1995). The use of CWs as a cost effective method for wastewater or stormwater treatment was initiated in 1904 in Australia and became prevalent in the Unit ed States during the early 1970s (Cole, 1998). Natural processes in the wetland remove organic, inorganic and microbiological contaminants ( Carruthers, 2002; Vrhovsek et al., 1996 ). Vascular plants play a particularly important role, stabilizing substrates while enhancing permeability, reducing water velocities and thus allow the settling of suspended solids, using nutrients, carbon, and trace elements for plant stem and root systems, and transporting gases between the sediments and atmosphere (Liu et al., 2007; Butler and Dewedar, 1991). In addition, photosynthesis by algae increases the concentration of dissolved oxygen affecting nutrient and metal reactions (Davis, 1995). But more importantly, the stems and roots of plants provide the necessary surface ar ea for growth and adhesion of microorganisms, which facilitate the decomposition of organic material and the recycling of nutrients

PAGE 19

3 (Martin and Moshiri, 1994). Previous studies demonstrated that CWs were advantageous treatment systems for the remediation o f acid mine drainage (AMD) due to comparably low cost and maintenance (Stottmeister et al., 2006; Woulds and Ngwenya, 2004; Braun et al., 2003). In general, contaminated mine water is highly acidic and rich in sulfate and metals such as Al, As, Cd, Cu, Fe, Pb, Mn, Ni and Zn (Stottmeister et al., 2006; Ritcey, 1989). The AMD treatment mechanisms in CWs may include metal adsorption on soil matter, accumulation into below and above ground plant tissues, and microbial mediated co precipitation or volatilizatio n of particular metalloids (Stottmeister et al., 2006; Jackob and Otte, 2003). The reclamation of wastewater and phosphate mining lands using CW technology could prove to be especially important as Florida law requires reclamation of previously mined phosp hate lands into wildlife habitat and watershed enhancement (i.e., lakes, wetlands, pasture, and agricultural lands). Wetlands are known to provide excellent opportunities for environmental improvement and restoration through (1) wastewater treatment; (2) w ater quality improvement and encouragement of water reuse; (3) fish and wildlife habitat; (4) passive recreation, such as bird watching and photography; (5) active recreation such as hunting; (6) flood storage, and (7) resynchronization of storm rainfall and surface runoff (Davis, 2005; EPA, 1993). Studies showed that e xtensive groundwater pumping from the Floridan Aquifer causes saltwater intrusion along the coast and lowering of groundwater levels (Peck et al., 2005). Therefore, a significant purpose of CWs i n Florida metropolitan areas could be the generation of water that meets drinking water standards to supplement rivers and streams and to satisfy public, industrial, and agricultural water demands. The combination of tailing sand deposits and CSA, whi ch are

PAGE 20

4 post mining features of phosphate mining, are ideal for the construction of wetland based water treatment systems. The CSA provides an excellent area for the wetland to develop, while the sand from the tailing sand deposits can be used for filtratio n prior to extracting the water from the system. With some adaptation this type of water treatment should be applicable to mining sites potentially worldwide. 1.2. Constructed wetland for aquifer storage and recovery (ASR) One additional application for centra l Florida could be to use the wetland water to recharge the Floridan Aquifer System, since this water is reducing and thus physicochemically similar to native Floridan groundwater (Lazareva and Pichler, 2010). The injection of wetland water could be essent ial to the future of aquifer storage and recovery (ASR) as a means of water management in Florida. The principle behind ASR is the storage of treated surplus surface water in a confined aquifer during rainy seasons followed by its recovery during times of need (Arthur et al., 2005; Arthur et al., 2001). The ASR system prevents water loss due to evaporation, which is normally associated with surface water reservoirs and is rapidly developing into the alternative for withdrawing Floridan Aquifer groundwater ( Jones and Owen, 2001). In addition, ASR plays a considerable role in environmental restoration, such as the Comprehensive Everglades Restoration Program (CERP) (Arthur et al., 2007). The CERP is an ecosystem restoration/water supply project developing up t o 333 ASR wells to restore the Everglades and to supply growing water demands in Florida. Currently, 34 ASR facilities in Florida are in operation and 46 are under construction (FLDEP, 2008).

PAGE 21

5 The ASR storage zone includes confined permeable zones of the U pper Floridan Aquifer System such as the Avon Park Formation, Suwannee and Ocala Limestones, and part of the Hawthorn Group ( Missimer, 1997; Scott, 1990; Scott, 1988; Miller, 1986) Generally, native groundwater in these zones is under reducing conditions (Sprinkle, 1989). In order to meet federal regulation, the water to be injected must be treated to meet primary drinking water standards (Arthur et al., 2001). As part of the microbial treatment ozonation is part of the process, causing the water to becom e extremely oxygen rich. As a result, the injection of treated surface waters into reducing native groundwater causes the oxidative dissolution of pyrite (FeS 2 ), and the mobilization of arsenic (As) with concentrations in recovered water of up to 130 g/L ( Arthur et al., 2005; Arthur et al., 2001). The Maximum Contaminant Level for As in drinking water, established by US EPA on January 23, 2006, is 10 g/L. The injection water and native groundwater have < 2 g/L and < 1 g/L of As, respectively ( Jones and Pichler, 2007 ; Arthur et al., 2001). Studies showed that As in the Florida subsurface is mostly associated with pyrite as a substitute element for sulfur in the FeS 2 structure (Lazareva and Pichler, 2007; Price and Pichler, 2006). The enrichment of As in p yrite in the Avon Park Formation, Suwannee Limestone, and the Hawthorn Group was up to 8260, 11200, and 5820 mg/kg, respectively (Dippold, 2009; Lazareva and Pichler, 2007; Price and Pichler, 2006). Generally, the oxidation of pyrite by O 2 acts as a source of acid, sulfate, iron and arsenic, and can be described by three steps ( Evangelou, 1995) : (1) FeS 2 + 7/2O 2 + H 2 2+ +2SO 4 2 +2H + Fe 2+ can be further oxidized to Fe 3+ which hydrolyzes into hydrous ferric oxides (HFO displayed as Fe(OH) 3 ) to discharge extra amount of acid into the environment (Figure 1):

PAGE 22

6 (2) Fe 2+ + 1/4O 2 +H + Fe 3+ + 1/2H 2 O and (3) Fe 3+ + 3H 2 3 + 3H + Consequently, As can be re sorbed onto HFO if the conditions remain sufficiently oxygenated to promote HFO stability, but release d back to native groundwater under reducing conditions. The additional factors affecting the mobility of As may include the following: input and native groundwater chemistry, aquifer matrix chemistry/mineralogy, site hydrogeology, injection water rock cont act time, and amount of cycle tests (Arthur et al., 2005). Poole (2009) reported about the proposed methods of dechlorination that could reduce the oxidation reduction potential level and deoxygenation that could decrease the dissolved oxygen in the inject ion water used for ASR in central Florida. The cycle tests showed the positive results for reduction of As leaching with the dechlorination system only. In order to prevent the oxidation of pyrite and mobilization of As during ASR, the injection of wetland water should be considered. Discharge from a wetland could be the ASR water of choice, because it is often more reducing with high sulfide and low dissolved oxygen levels, which are favorable for the stability of pyrite (Lazareva and Pichler, 2010 ; Jones and Pichler, 2007). Therefore, in depth knowledge and understanding of As mobilization from the Floridan Aquifer could be very important to forecast As behavior during anthropogenic physico chemical changes in Florida and potentially worldwide.

PAGE 23

7 Figure 1. Stability diagram of pyrite and Fe(OH) 3 in water at 25C and 1 atmosphere total pressure ( Modified from Evangelou, 1995). Oxidizing Conditions Reducing Conditions

PAGE 24

8 1.3. Review of arsenic occurrence and toxicity Arsenic is a well known carcinogen that can lead to a wide range of health problems, such as atherosclerosis, kidney, bladder, colon, skin and lung cancer (Smith et al., 2009; Huang et al., 2008; Pu et al., 2007; Steinmaus et al., 2006; Wu et al., 2006; Chen et al., 2003; Smith et al., 2000; Smith et al., 1992). Over t he past several decades As contamination from both natural and anthropogenic sources became one of the most important health issues in many countries across the world (Nriagu, 2002; Thornton, 1999). The elevated concentrations of As in surface or groundwa ter were discovered in China, India, Pakistan, Taiwan, Bangladesh, Vietnam, Thailand, Chile, Mongolia, Mexico, Argentina, Canada, and the United States ( Brammer and Ravenscroft, 2009; Nriagu, 2002). About 150 million people worldwide are exposed to elevate d As contamination concentration in their drinking water (Ravenscroft et al., 2009). A multitude of scientific s tudies showed that 30 million people in Bangladesh and 6 million inhabitants in West Bengal (India) are chronically exposed to groundwater poiso ning with up to 3200 g/L of As ( Nordstrom 2002; Smith et al., 2000; Dhar et al, 1998; Khan et al., 1997). The average abundance of As in the upper continental crust significantly varies from 1.5 mg/kg (Taylor and McLennan, 1985) to 4.8 mg/kg (average data from Sims et al., 1990; Gao et al., 1998), commonly increasing in igneous and sedimentary rocks, such as coal and shale deposits (Smedley and Kinniburgh, 2002). The average concentrations for shale and river mud are 12.4 and 8.4 mg/kg, respectively (Govin daraju, 1994). The average As for the limestone geostandard GSR 6 is reported as 2.6 mg/kg (Baur and Onishi, 1969). Arsenic containing pyrite (FeS 2 ) is perhaps the most common mineral

PAGE 25

9 source of As (Nordstrom 2002). Huerta Diaz and Morse (1990) detected As concentrations in marine sedimentary pyrites of as much as 9300 mg/kg. Thomas and Sanders (1998) reported that framboidal pyrites contained up to 1000 mg/kg of As. Studies about its occurrence in the Floridan subsurface determined that As is mostly associa ted with pyrite as a substitute element for sulfur in the FeS 2 structure (Lazareva and Pichler, 2007; Price and Pichler, 2006). The enrichment of As in pyrite in the Avon Park Formation, Suwannee Limestone, and the Hawthorn Group was up to 8260, 11200, and 5820 mg/kg, respectively (Dippold, 2009; Lazareva and Pichler, 2007; Price and Pichler, 2006). Therefore, sedimentary pyrite can be a significant sink and source of As. Smedley and Kinniburgh (2002) reported that calcium phosphate or apatite can contain u p to 1000 mg/kg of As Soil constituents, such as clays and organic substances, can easily interact with heavy metals such as As via ion exchange or surface adsorption (Manning and Goldberg, 1997; Evangelou, 1995). Clays readily adsorb As because of the ox ide like character of the edges of its grains (Claesson and Fagerberg, 2003). Clays and organic substances have very small particle size, which therefore result in a large surface area per unit volume and ability to adsorb As. Moreover, the potential of h umic substances to complex with heavy metals is due to the existence of oxygen containing functional groups such as carboxyl (COOH), hydroxyl (Oh), and carbonyl (C=O) (Evangelou, 1995). It has been shown that As integrates into sediments by co precipitatio n with hydrous ferric oxides (HFO), or is adsorbed onto extremely high surface area of precipitated HFO (Hongshao and Stanforth, 2001; Pichler et al., 1999; Hinkle and Polette, 1999; Evangelou, 1995; Bowell, 1994; Chao and Theobald, 1976).

PAGE 26

10 The cycling of As between different valence states and chemical species in natural waters and soils/rocks is greatly affected by both abiotic and biotic reactions (Inskeep et al., 2002). The essential processes that control As cycling between the solid and aqueous phases are oxidation/reduction, dissolution/precipitation, adsorption/desorption, and bi ological alterations ( Stollenwerk, 2003 ; Inskeep et al., 2002; Hering and Kneebone 2002 ; Zobrist et al., 2000 ) Arsenic adsorption and desorption reactions are influenced by changes in pH, redox potential and the presence of competing anions. Solid phase precipitation and dissolution reactions are primarily controlled by pH, redox state and chemical composition (saturation) (Hinkle and Polette, 1999). Behavior, fate, bioavail ability, and toxicity of As in the environment vary significantly depending on the chemical species in which As occurs. In the aqueous (mobile) environment such as surface water and groundwater As exists in III, +III, and +V oxidation states (Hering and K neebone, 2002; Stottmeister et al., 2006). The most common aquatic forms are the trivalent arsenite (As (III)) and the pentavalent arsenate (As (V)). Organic forms of As include monomethylarsonic (MMA) and dimethylarsinic (DMA) acids and mostly present in seafood. While the inorganic forms are highly toxic, organic species are far less toxic (Le, 2002). It is considered that As (III) is more mobile and more toxic to biota and plants than As (V) (Inskeep et al., 2002). Generally, As (V) is thermodynamically favore d in oxic surface waters at Eh greater than 100 m V at pH 8 and greater than Eh 300 mV at pH 4 (Inskeep et al., 2002). Below these redox potentials As (III) is stable either as the H 3 AsO 3 As S complexes, or As (III) solid phases such as As 2 S 3 (Insk eep et al., 2002). At the same time, the coexistence of both species is common (Kuhn and Sigg, 1993; Anderson and Bruland, 1991; Mok and Wai, 1990). Important solid phases of As

PAGE 27

11 (V) or As (III) are the Fe, Mn and Ca arsenates, and As (III) sulfides such as orpiment (As 2 S 3 ), and realgar (AsS). Arsenopyrite (FeAsS) is a significant source of As formed under reducing environments (Inskeep et al., 2002). Generally, As (V) displays a strong affinity for most metal hydroxides and clay minerals, forming surface co mplexes. In contrast, As (III) is selective, demonstrating a strong affinity for iron hydroxides (Inskeep et al., 2002). Both As (III) and As (V) form analogous surface complexes on goethite. Numerous studies reported about the oxidation of As (III) by man ganese oxides in waters and lake sediments (Deschamps et al., 2003; Chiu and Hering, 2000). The reduction of As (V) to As (III) is typically observed under microaerobic to anoxic conditions such as in flooded soils, sediments, and aquifers (Masscheleyn at al., 1991). Under reducing conditions, there are two potential pathways for a reductive dissolution of sorbed As: (1) As can be released from the solid (i.e., Fe(OH) 3 or Al(OH) 3 ) through reduction to arsenite, or (2) through direct substrate degradation (i .e., Fe(OH) 3 ) (Inskeep et al., 2002). Rochette et al. (2000) reported about the reduction of As (V) by H 2 S at a low pH. Microorganisms are able to facilitate the reduction of As (V) to As (III) and the oxidation of As (III) to As (V) (Salmassi et al., 2002 ; Jones et al., 2000; Zobrist et al., 2000; Ahmann et al., 1994). A wide variety of microorganisms can reduce aqueous As (V) /oxidize As(III) or adsorbed on Fe(OH) 3 or Al(OH) 3 as a detoxifying mechanism or a source of energy ( Silver et al., 2002; Oremland e t al., 2002; Macur et al., 2001; Zobrist et al., 2000 ; Ghosh et al., 1999; Phillips and Taylor, 1976; Osborne and Ehrlich, 1976 ). Particularly, an anaerobic microorganism ( Sulfurospirillum barnesii ) is able of both reductive dissolution of Fe (III) oxides and reduction of As (V) to As (III) (Zobrist et al., 2000). High As in the Bangladesh groundwater had been derived from reductive

PAGE 28

12 dissolution of arsenic rich HFO occurring as a coating on sedimentary grains (Inskeep et al., 2002; Nickson et al., 2000). Typ ically, the highest concentrations of As in groundwater are found in aquifers with areas high in organic material where greater microbial activity would result in elevated rates of the Fe (III) oxide reductive dissolution (Inskeep et al., 2002). 1.4. Descripti on of the study area The constructed wetland /filter basin (CW/FB) treatment system was located on a clay settling area and tailing sand deposits at the Hines Energy Complex, Polk County, est ablished in 1999 and used for the experimental treatment of industrial wastewater from the Hines Energy electric power generating plant (cooling water), tertiary treated effluent from the city of Bartow, as well as rain and excess surface water. The surfac e flow wetland was approximately 1 500 m long, 10 m wide, ranged in water depth from 0.5 to 2 m and was constructed in a U shape. The length width ratio of the wetland was higher than maximum suggested value of 1:1 to 1:2 (EPA, 2000) due to specific topogr aphic settings of the study area. The area of the wetland was about 12 250 m 2 (Figure 4). The wetland was not lined and the substrate consisted of clay matrix with the decomposing organic matter. Wetland vegetation was allowed to evolve naturally (i.e., no t planned and planted) due to comparably high costs and maintenance. It consisted of both native Floridian and non native species such as water lettuce ( Pistia stratiotes ), carolina willow ( Salix caroliniana ) brazilian pepper ( Schinus terebinthifolius ), w ater fern ( Salvinia ), baby's tears ( Micranthemum umbrosum) cattail ( Typha spp ), willow ( Salix spp. ), and

PAGE 29

13 Figure 2. Map of the study area located at the Hines Energy complex (Polk County, Central Florida) including water transfer system from the cooling pond to the U shaped constructed wetland and filter basin treatment system. Note: MW 1 to MW 6 monitor wells; N 15 and SA 8 water cropping areas to the north and south of the wetland; FBN and FBS filter basin north and south pumps.

PAGE 30

14 Figure 3. Schematic concept diagram of the constructed wetland/filter basin treatment system. Note: Schmutzdecke biological film formed on the sand surface.

PAGE 31

15 Figure 4. Photographs of the constructed wetland treatment system. Cooling Pond Wetland

PAGE 32

16 common duckweed ( Lemna minor ). The wetland was surrounded by two bodies of water: N 15 to the north and SA 8 to the south. These previously mined and reclaimed phosphate lands are now a water cropping system to capture, store and reuse stormwater (PEF, 2004). T he 6 000 m 2 filter basin (FB) constructed on tailing sands was used to improve the efficiency and reliability of the treatment system. The depth of the FB (i.e., sand bed) was 4 m and the walls and bottom of the FB were lined with a polyethylene cover. The most impo rtant feature of the FB was the development of a biologically active 2 cm) (Figure s 3, 5). This reddish brown slimy biofilm acted as a fine filter of solid particles (mechanical fil tration) and a zone of biological action providing the degradation of soluble organics and the potential elimination of pathogens, color and odor contaminants in water (Muhammad et al., 1997; Huisman and Wood, 1974). Water from the Hines Energy Complex coo ling pond (CP) was pumped into the wetland at different rates depending on the season. During the rainy seasons of 2006 and 2007, the hydraulic loading rates into the wetland were around 0.33 and 0.45 L/day/m 2 respectively. During the dry season 2006, it was about 0.61 L/day/m 2 The residence time of water in the wetland was approximately 120 days. At the end of the flow path the water was either pumped back into the CP to control water levels in the wetland, or discharged onto the filter basin surface. Fo llowing filtration through the fine sand the water was collected in a series of pipes.

PAGE 33

17 Figure 5. Photographs of the filter basin system. Note: Highlighted in blue treated wetland water discharged onto the filter basin surface forming a biofilm Filter Basin South and Nor th Side

PAGE 34

18 1.5. Geology The research area is located in Polk County, a part of the Southwest Florida Water Management District (SWFWMD) (Figure 2). Here the hydrogeological framework can be subdivided into three discrete units from the bottom u pward: the Floridan Aquifer System (FAS), the Intermediate Aquifer System (IAS) or Intermediate Confining Unit, and the Surficial Aquifer System (SAS) (Figure 6) (Miller, 1986). The FAS underlies a region of approximately 259 000 km 2 throughout southern A labama, southeastern Georgia, southern South Carolina, and the entire Florida peninsula; it is one of the most prolific and extensive aquifers in the world (Budd and Vacher, 2004; Scott, 1992). The Paleocene to Lower Miocene FAS consists of the following s tratigraphic units (from oldest to youngest): The Oldsmar/Cedar Keys, and Avon Park Formations, the Ocala and Suwannee Limestones (Figure 7). In some areas, the Tampa Member and the lower part of the Arcadia Formation of the Hawthorn Group are included in the upper part of the FAS where it comprises permeable carbonate lenses (Scott, 1991). The FAS is a vertically continuous sequence of carbonate rocks with typically high porosity and permeability which is subd ivided into the Lower Floridan A quifer (LFA), t he Middle confinin g unit, and the Upper Floridan A quifer (UFA) based on the hydrologic properties of the lithologic units (Miller, 1986). The LFA system is comprised of the Oldsmar Formation and the upper the Cedar Keys Formation (Randazzo and Jones, 1997) The confining unit separatin g the Upper and Lower Floridan A quifers consists of a very fine grained (micritic) limestone, clay or dolomite filled with anhydrite in the pore space (USGS, 1999). Typically, the base of the UFA is identified by the presence of anhydrite beds, which are considered to be the top of the Cedar Keys

PAGE 35

19 Figure 6. Lithostratigraphic and hydrogeologic units of the study area (Modified from SWFWMD Report, 2000; Scott, 1989).

PAGE 36

20 Figure 7. Geological cross sections of southwestern Florida ( Modified from Swancar and Hutchinson, 1995 ).

PAGE 37

21 Formation. The upper sections of the Ocala Limestone and the Avon Park Formation have the highest transmissivity throughout the UFA due to secondary porosity and permeability caused by fr acturing and dissolution processes (Randazzo and Jones, 1997; Miller, 1986; Gilboy, 1985; Stringfield, 1966). Generally, the porosity levels throughout the FAS range from 10 to 50 % (Budd and Vacher, 2004). The Middle Eocene Avon Park Formation is principa lly composed of interbedded fossiliferous limestone and dolostone with the distinctive dark brown fine crystalline dolomite unit containing a gypsiferous wackestone mudstone towards the base of Formation (Arthur, 2008; Miller, 1986; Gilboy, 1985; Randazzo and Jones, 1997; Cander, 1991; Chen, 1965). The thickness of the Avon Park Formation varies from 305 m to 488 m (1000 ft to 1600 ft) (Miller, 1986). It is unconformably underlied by the Paleocene to Lower Eocene Oldsmar Formation (Randazzo and Jones, 1997) The limestone contains a variable amount of organic rich laminations and marine grass fossil beds (Dippold, 2009; Budd and Vacher, 2004). In addition, minor pyrite, chert, and quartz were identified (Dippold, 2009; Arthur et al., 2008). The amount of gyp sum and anhydrite interbedded in the dolomite increases with depth. The dolomite can occur highly fractured and sucrosic in texture (Arthur et al., 2008). Generally, the Avon Park Formation has interparticle porosity as well as dissolution channels with co nduit type permeability zones (Cander, 1991). The Upper Eocene Ocala Limestone overlies the Avon Park Formation and varies in thickness from 30 m to 152 m (100 ft to 500 ft) (Gilboy, 1985). It can be subdivided into lower and upper sections based on the li thological composition (Miller, 1986). The lower part consists of a white, poorly to moderately indurated and partially dolomitized

PAGE 38

22 fossiliferous grainstone and packestone (Miller, 1986). The upper portion is composed of a white, poorly to well indurated, fossiliferous grainstone, packestone and wackestone. Minor amounts of quartz and pyrite were also observed. The present fossils are foraminifers, echinoids, bryozoans, mollusks and rare vertebrates (Scott, 1991). As noted above, the Ocala Limestone has ver y high transmissivity which is an important unit for the UFA and potentially for ASR procedure (Miller, 1986). The Oligocene Suwannee Limestone reaches a thickness between 50 m and 100 m (164 ft to 328 ft) and is principally composed of wackestone to pelle tal and foraminiferal grainstone with minor phosphate quartz sand and clays (Williams et al., 2002; Green et al., 1995; Miller, 1986; Gilboy, 1985). In addition, minor amounts of pyrite, organic material, and chert nodules are present (Green et al., 1995; Miller, 1986). Generally, the limestone has intergranular and high moldic porosity zones, which are particularly important for ASR procedure (Price and Pichler, 2006; Miller, 1986). The Upper Oligocene to Lower Pliocene IAS is the predominantly composed o f the Hawthorn Group, which is subdivided into a lower section comprising the undifferentiated Arcadia Formation, Tampa and Nocatee Members of the Arcadia Formation and the upper section of the Peace River Formation (Scott, 1988). The Arcadia Formation unc onformably overlies the Oligocene Suwannee Limestone and achieves its maximum thickness of about 183 m (600 ft) in the Okeechobee Basin (Scott, 1988). It was initially interpreted to be primarily Lower Miocene (Scott, 1988), but is now recognized to be Low er Oligocene to Middle Miocene (Brewster Wingard et al., 1997; Missimer, 1997). The undifferentiated Arcadia Formation is principally composed of variable amount of siliciclastics within carbonate matrix. Typically, the carbonates are

PAGE 39

23 fossiliferous, yellow ish gray to light greenish gray to light brown, micro to finely crystalline limestones and dolostones with variable amount of quartz sands, gray to greenish gray clays, chert, phosphate material, and pyrite (Lazareva and Pichler, 2007; Scott, 1988). The c lays are palygorskite, sepiolite, illite/smectite mixed layer, and insignificant amounts of kaolinite (Compton et al., 1993). Thin beds of sand and clay are widespread and molds and casts of mollusks are commonly found in the dolostones. The lithology of t he Upper Oligocene to Lower Miocene Tampa Member of the Arcadia Formation varies from a white to yellow gray marine wackestone to packestone containing variable amounts of dolostone, clay, quartz sand, and minor phosphate material (Wingard et al., 1993; Sc ott, 1988). Thin carbonate, quartz sand and clay beds are commonly observed throughout the Tampa Member. The Tampa Member is typically a poor to well indurated limestone with intergranular or moldic porosity containing variable amounts of fossils such as m ollusks, corals, and foraminifera (Scott, 1988). The contact between the undifferentiated Arcadia Formation and Tampa Member in central Florida is sporadically marked by a chert layer at the top of the Tampa Member (Gilboy, 1985). The Upper Oligocene to Lo wer Miocene Nocatee Member forms the base of the Arcadia Formation in the southern area of SWFWMD. It contains the highest amount of siliciclastic material compared to the entire Arcadia Formation (Scott, 1988). The Middle Miocene to Lower Pliocene Peace R iver Formation, currently being exploited for the phosphate ore, unconformably overlies the Arcadia Formation and reaches a maximum thickness of 46 m to 61 m (150 ft to 200 ft) in the Okeechobee Basin (Scott, 1990). It is composed of gray to greenish gray clays (palygorskite, sepiolite, illite/smectite mixed layer), sandy clays, and carbonates with variable amount of phosphate sand and gravel,

PAGE 40

24 iron oxides, and pyrite (Lazareva and Pichler, 2007; Green at al., 1995). The carbonates are comprised of interbedd ed limestones, dolostones, and a siliciclastic component dominates the Peace River Formation (Scott, 1988). The water yielding beds of carbonate lenses in the IAS lie between clayey confining units. As a result, the water in the IAS is under principally co nfined conditions excluding local areas, where the upper confining unit is missing and the system is in direct hydraulic contact with the overlying SAS. The IAS is the major source of water supply in Charlotte and Sarasota Counties, where the underlying FA S is deeply buried and holds only brackish water (SWFWMD, 2000). The Upper Pliocene to Pleistocene SAS is generally comprised of unconsolidated to poorly indurated clastic deposits such as sands, sandy clays, phosphorite and some well indurated carbonate rocks. It is primarily unconfined with some semi confined or locally confined sections (Gilboy, 1985). Generally, water moves downward from the SAS through the upper confining unit of the IAS subsequently following short flow paths and discharging to surf ace drainage area. On the other hand, some water penetrates downward through the lower confining unit of the IAS to recharge the underlying FAS (USGS, 1999). 1.6. Objectives This dissertation was part of a 2 er from a natural treatment system will cause dissolution of pyrite in the Floridan Aquifer 03 153R). The project was funded through a grant from the Florida Institute of Phosphate Research (FIPR) and the Southwest Fl orida Water Management District (SWFWMD) with assistance of Progress Energy Florida to

PAGE 41

25 Mr. Peter Schreuder (Schreuder Inc.) and Dr. Thomas Pichler (University of South Florida). The purpose of the project was to improve the operation of the sand filtration system and to examine the hypothesis that anoxic water from the wetland and sand filter system would cause less dissolution of pyrite (and release of As) in the Floridan Aquifer limestone matrix, compared to highly oxygenated surface waters, when injected into recharge wells. Of the overall research project, my part was: 1) To evaluate the performance of the wetland for 18 month in complex hydrogeological settings using a combined isotope/chemical mass balance approach; 2) To consider the possible injection of the treated wetland water into the Upper Floridan Aquifer as the alternative source water for ASR ; 3) To simulate the ASR process with a series of bench scale leaching experiments; 4) To incorporate the experimental data into coupled reactive transport models u 1.7. Arr angement of D issertation The dissertation is composed of five chapters. The first chapter introduces the reader to the sustainability of water resources around the world and importance of constructed we tlands for water reclamation and reuse with a possible injection into the Floridan Aquifer. In addition, the occurrence and toxicity of arsenic are reviewed, and a detailed description of geology and hydrogeology of site area is given. In order to avoid re dundancy, the information given in the first chapter was omitted from all subsequent

PAGE 42

26 term performance of a constructed wetland/filter a special issue of Chemical Geology (Lazareva and Pichler, 2010) and submitted to Ecological Engineering Bench scale leaching experiments to investigate geogenic arsenic contamination in Floridan Aquifer accounts for the bench scale leaching experiments to simulate water rock interaction between different types of water and Floridan Aquifer matrices and examine the pyrite dissolution potential and mobilization of geogenic As under a range of redox condi tions. Publication of this chapter in Environmental Science and Technology is planned for later results of reactive transport modeling using The results from this part will be published together with results from the previous chapter. The fifth chapter summarizes and synthesizes the scientific contributions of the previous chapters.

PAGE 43

27 CHAPTER TWO LONG TERM PERFORMANCE OF A CONSTRUCTED WETLAN D/FILTER FASIN SYSTEM TREATING WASTEWATER IN COMPLEX HYDROGEOLOGICAL SETTINGS 2.1. Introduction Phosphate mining in central Florida is widely distributed and annually disturbs about 15 25 km 2 of land through generation of clay settling areas, open mine pi ts, and tailing sand deposits (FIPR). Florida law requires reclamation of previously mined phosphate lands as wildlife habitat and watershed enhancement The reclamation of wastewater and phosphate mining lands using constructed wetland technology is very important in Florida, providing an excellent opportunity for environmental improvement and restoration (EPA, 1993). Studies showed that i ncreased groundwater pumping from the Floridan Aquifer causes saltwater intrusion along the coast, and lowering of grou ndwater levels, thus affecting spring flows and lake levels (Peck et al., 2005). Therefore, a significant purpose of constructed wetland in Florida metropolitan areas could be the generation of water that meet s drinking water standards to supplement rivers and streams and to satisfy public, industrial, and agricultural water demands.

PAGE 44

28 Investigation of the performance and water balance of a constructed wetland in complex hydrogeological conditions can be very challenging. At the same time, isotopic compositi on of oxygen and hydrogen can provide important information about the source of water, i.e., the actual H 2 O molecules rather than dissolved constituents. Applications range across the whole spectrum of hydrogeological and hydrological studies, including hy drothermal systems (e.g., McCarthy et al., 2005; Pichler, 2005a), aquifer recharge (e.g., Gonfiantini et al., 1998), groundwater surface water interaction (e.g., Baskaran et al., 2009), contamination studies (e.g., Pichler, 2005b) and the water cycle (e. g., Craig, 1961). Delineating the source of water is particularly important for evaluating and managing this resource in areas with limited supply. The quantification of groundwater inflow into wetlands and lakes in Florida, for example, is an important co mponent in the water balance equation (Sacks, 2002). Groundwater input can differ significantly depending on the topographic settings, type of soil, depth to bedrock, vegetation, fractures, climate, and the anthropogenic activity, affecting water levels an d quality (Kendall and Coplen, 2001). Estimating this term can be very complicated, but quantification is possible using a chemical and isotope mass balance method (Sacks, 2002; Winter, 1995; Krabbenhoft et al., 1990; Stauffer, 1985). The isotope mass bala nce method for estimating the water balance of lakes and wetlands has been used extensively (e.g., Hunt et al., 1998, Yehdegho et al., 1997; Michaels, 1995; Krabbenhoft at al., 1994; Dincer, 1968) and successfully applied in central Florida (Sacks, 2002). 2.2. Methods and sampling procedure

PAGE 45

29 The description of the study area was comprehensively described in first chapter and therefore excluded from this c hapter. In order to evaluate the performance of a constructed wetland /filter basin treatment system und er a variety of climatic conditions, the monitoring was carried out for 18 months. Water sampling began in April 2006 and finished in October 2007. Bi monthly water samples were obtained from the effluent discharge (EF), cooling pond (CP), wetland pumping station (WP) and wetland surface (WS), ta iling sand filter basin north (FBN) and south (FBS) pumping stations, and water bodies to the north and south of the wetland, N 15 and SA 8 (surface water) (Figure 2). The WP sample was collected from a submerged su mp located at depth of 2 m and the WS sample surface water from the same location (Figure 8) Overall, 244 samples were collected and analyzed. In addition, water samples from the CP, WP, FBS, and FBN were collected for primary, secondary drinking water standards (PDWS/SDWS), volatile organic compounds (VOC), synthetic organic compounds (SOC), radionuclides (RAD), total and fecal coliform bacteria with assistance of Schreuder Inc. PDWS and SDWS were collected monthly, and VOC, SOC and RAD were collected q uarterly. In total, 98 samples were collected for the PDWS/SDWS, and 48 samples for VOC, SOC, and RAD. A total of 88 samples were collected bi monthly for fecal and total coliform To evaluate possible groundwater input into and water leaking out of the wetland 6 monitor wells were installed along the flow path and sampled monthly. This was necessary to separate wetland induced changes in water chemistry from those due to dilution by Floridan groundwater or seepage from the water bodies to the north and south, N 15 and SA 8 (Figure 2 ). In total, 121 samples from MWs were collected and analyzed (Figure 8).

PAGE 46

30 Figure 8 Water sampling from the wetland pumping station and wetland surface (right), and monitor wells (left).

PAGE 47

31 In order to investigate and ev aluate the change of water chemistry along the wetland flow path samples were collected within the surface water at depths of 0 and 0.5 m using a peristaltic pump. During the dry season (March 19 20, 2007), water samples were collected at 17 stations and during the rainy season (September 24 25, 2007) at 11 stations along the flow path of the wetland (Figure s 9 10 ). 2.2.1. Field and laboratory analysis The water samples were analyzed immediately in the field for pH, temperature (T), oxidation reduction potential (ORP) and conductivity using a Myron L UltrameterTM. The meter was calibrated in the field with known buffer solutions. The dissolved oxygen (DO) concentration was determined using a HACH HQ40d Meter with a LDO (luminescent dissolved oxygen) pro be. The LDO sensor was calibrated following the manufacturer's specifications. The concentration of ferrous iron (Fe(II)) was determined with a CHEMets Colorimetric field kit with a color chart, and sulfide (H 2 S) was analyzed using a Methylene Blue Method on a HACH DR 2400 portable separated into two 30 mL HDPE bottles. One aliquot was used for the determination of major anion concentrations and stable isotopes The other sample was acidified to 1% HNO 3 for the analysis of major cations and arsenic. In addition, the set of unfiltered and unacidified samples from the CP, WP and FBS/FBN was collected and analyzed for fecal and total coliform bacteria from September, 2006 to Septemb er, 2007 The samples were stored in a cooler, transported and analyzed at the Center for Water and Environmental Analysis, University of South Florida, Tampa. Calcium (Ca),

PAGE 48

32 Figure 9 Evaluation of water quality along the wetland flow path. Note: WP wetland pump; FBN and FBS filter basin north and south pumps; Dashed arrow wetland flow path; Solid squares and empty circles wetland water samples taken on March 2007 and September 2007, respectively. 1 4 2 3 5 6 7 8 9 10 11 12 13 1 2 3 4 (1) (2) (3) (4) (4) (3) (2) (1) N 15 SA 8 N Cooling Pond WP FBN FBS 100 m

PAGE 49

33 Figure 10. Water sampling and analysis along the wetland flow path.

PAGE 50

34 iron (Fe), magnesium (Mg), manganese (Mn), silica (Si), sodium (Na), potassium (K), and strontium (Sr) were measured on a Perkin Elmer Optima 2000 DV inductively coupled plasma optical emission spectrometer (ICP OES). Fluori de (F), chloride (Cl), sulfate (SO 4 ), nitrate (NO 3 ), nitrite (NO 2 ), bromide (Br), and phosphate (PO 4 ) were determined by ion chromatography (IC) using a Dionex ICS 2000 system. Total arsenic (As) concentrations were measured on a PSA 10.055 Millenium Excal ibur hydride generation atomic fluorescence spectrometer (HG AFS) In preparation for HG AFS analysis, 10 mL of the sample solution was treated with 30% concentrated HCl, 2% saturated potassium iodide (KI) solution, and deionised water (DI) with a final di lution volume of 50 mL (5:1). This special treatment of the samples causes the reduction of As (V) to As (III) prior to the formation of the arsenic hydride (AsH 3 ) via addition of sodium tetraborohydride (NaBH 4 ). The AsH 3 generator is based on the reaction of sodium borohydride, hydrochloric acid and the sample. Formation of hydrogen free radical: NaBH 4 + 3H 2 O + HCl = H 3 BO 3 + 8H + NaCl Formation of the volatile arsenic hydride (gas) 8H + As 3+ = AsH 3 + H 2 excess In a consequence, the AsH 3 is atomized in a hydrogen flame and concentrations are determined by fluorescence spectrometry. The accuracy and precision of the measurements were verified through the use of internal and external standards, indicating a precision and accuracy better than 5%. Stable isot opes of oxygen and hydrogen were determined at the University of South Florida stable isotope laboratory using a Finnegan Delta V 3 keV Isotope Ratio

PAGE 51

35 Mass Spectrometer and a Gasbench II preparation device. For oxygen isotope determination, 200 L of sample was equilibrated with 0.3 % CO 2 in He mixture for 24 hours at 25 C. Prior to hydrogen isotope analysis, 20 mL of sample were reacted for 24 hours with 1 g of Cu wire to remove dissolved sulfides and to reduce the effects of H 2 S gas on the Pt catalyst dur ing the H 2 equilibration. Following this treatment, 200 L of sample was equilibrated with a 1 % H 2 and He mixture for 10 min. Both oxygen and hydrogen isotope measurements were compared to the Vienna Standard Mean Ocean Water (VSMOW) standard and reported 18 18 Precipitation measurements (amount of rainfall) were performed by the staff from the Hines Energy Complex weather station. Surface water level measurements were done by Schreuder Inc. and meteorological data was available from the nearby Frostproof Station of the Florida Automated Weather Network (FAWN). Measurements of fecal and total coliform bacteria PDWS, SDWS, VOC, SOC and RAD were carried out at the So uthern Analytical Laboratory of Florida using the Standard Operation Procedures outline d by Ch. 62 160, F.A.C (DEP, 2008) 2.3. Results 2.3.1. Monitor wells The lithology of monitor well (MW) cores along the wetland was uniform and sediments were compose d of light tan to brown poorly to well sorted fine sands or silts with occasional grey clay nodules. The brown color was due to the presence of organic

PAGE 52

36 material. After installation, each MW was developed using a small submersible pump to recharge debris fr ee water and sampled monthly for the duration of the study. The chemical composition of water collected from the MWs was different from the cooling pond (CP), wetland pump (WP), and the wetland surface (WS), with values in the MWs either higher or lower ( Appendix A, Tables 1 and 2). The chemical composition of the MWs was much closer in composition to sites SA 8 and N 15, and groundwater from the Intermediate Aquifer System (GW) (Sacks and Tihansky, 1996). In Figure 11 A the average concentrations of SO 4 F As, and Fe in monitor wells (MW 1 to MW 6) are compared. MWs were arranged according to the wetland flow path. MW 6 had the highest As (up to 9 g/L) and Fe (up to 50 mg/L) concentrations among all MWs. The analyses of water samples from the MWs showed t hat there was little to no leakage from the sites N 15 and SA 8 into the wetland treatment system. The analysis of variance (ANOVA) of the conservative tracer Na was applied to examine the difference in water chemi stry between MWs, SA 8 and N 15. For the f irst group of samples (MW 4 to MW 6 and SA 8), the F ratio (30.0) was significantly larger than the F critical value (2.7) indicating a statistically significant difference within the data set. For the second group of samples (MW 1 to MW 3 and N 15), the F ratio and F critical value were 7.5 and 2.7, respectively. The variance for MW 3 had the highest value (361.0) potentially demonstrating the highest influence from N 15 (Figure 11B). 2.3.2. Precipitation measurements Due to periodic variations in precip itation and temperature in the study area two different seasons can be distinguished: the dry season from November to April, and the

PAGE 53

37 Table 1. Maximum, minimum, arithmetic mean and standard deviation of analyzed par ameters in the cooling pond, wetland pump and surface, filter basin south and north pumps.

PAGE 54

38 Table 1. Continued

PAGE 55

39 Table 2. Maximum, minimum, arithmetic mean and standard deviation of analyzed parameters in monit or wells MW 1 to MW 6 and water. No te: MWs arranged according to the wetland flow path.

PAGE 56

40 Table 2. Continued.

PAGE 57

41 Figure 11 (A) Average values of SO 4 Fe, F, and As and (B) Average and variance of Na estimated by ANOVA at MW 1 to MW 6, N 15 and SA 8. Note: MW 1, 2, 3, 6, 5 to MW 4 monitor wells arranged according to the wetland flow path; GW* Floridan groundwater from the Intermediate Aquifer System (well ROMP 45 from Sacks and Tihansky, 1996).

PAGE 58

42 rainy or wet season from May to October. Daily precipitation was recorded at the Hines Energy Complex weather station from May 1, 2006 to October 31, 2007, covering one dry and two rainy seasons (Figure 12 ). The highest levels of rainfall (up to 123 mm) were detected during the major hurricane e vents (Ernesto and Alberto) in June and August 2006. The mean monthly rainfall ranged from 0 to 11 mm. Total seasonal precipitation during the dry season of 2006 was 324 mm, while the rainy seasons of 2006 and 2007 had between 917 and 782 mm, respectively. 2.3.3. Surface water level measurements The monitoring of surface water level elevations above National Geodetic Vertical Datum (NGVD) showed that the SA 8 had a higher level (51.6 52.5 m) compared to the N 15 (49.9 50.7 m) and the cooling pond (CP ) (49.0 49.3 m) (Figure 13 ). The elevation of wetland water surface (WS) ranged from 49.7 to 50.8 m with the lower levels in April May 2007. Records showed that at the end of April the CP pump was turned off and the WS was lowered approximately 1 m for maintenance purposes. After one month the CP pump was turned back on and the wetland treatment system became again operational. 2.3.4. Evaluation of water quality along the wetland flow path Evaluation of the wetland during the dry (March 19 20, 2007) an d rainy (September 24 25, 2007) seasons was important to understand the consistency and reliability of the treatment system in time and space (Figure 9). The wetland transect showed a distinct pattern of change in water chemistry along the flow path from t he input

PAGE 59

43 Figure 12 Precipitation hydrograph recorded at the study area from May 1, 2006 to October 31, 2007, covering one dry and two rainy seasons. Note: major hurricane events (Ernesto and Alberto) in June and August 2006.

PAGE 60

44 Figure 1 3 Water level elevations at the CP, WS, N 15 and SA 8 above National Geodetic Vertical Datum (NGVD). Note: CP cooling pond; N 15 and SA 8 water bodies to the north and south of the wetl and; WS wetland water surface.

PAGE 61

45 cooling pond (CP) to the output wetland water from pump (WP) (Table 3, Figure 14). 2.3. 4 .1. Field measurements Temperature (T). Generally, during the dry and rainy seasons temperature in the wetland was < 20 and 25C, respectively. However during the dry season, it reached up to 24.8 C (500 800 m) on the wetland surface. pH The pH values were reduced along the wetland flow path from 8.9 to about 7.0, but reached up to 8.7 on the wetland surface (500 700 m and 1400 m) during the dry season The formation of organic acids in the wetland was likely responsible for this drastic change in pH. Oxidation Reduction Potential (ORP). Generally, ORP was negati ve in the wetland indicating more reducing conditions. At the same time during the dry season algal blooms the ORP level at 300 700 m on t he wetland surface and at a depth of 0.5 m reached +254 and +217 mV, respectively. Sulfide (H 2 S). During the dry season the concentration of H 2 S was up to 900 g/L and did not show the trend along the wetland. In contrast, during the rainy season H 2 S reac hed up to 2545 g/L increasing along the wetland flow path. Dissolved Oxygen (DO). After entering the wetland, DO along the wetland flow path was generally reduced to < 0.5 mg/L. During the dry season it reached up to 8.7 mg/L on wetland surface (200 700 m of the wetland flow path). The concentration of DO at the WP during the dry and rainy seasons was up to 6.0 and 3.3 mg/L, respectively. This was likely due to trapping of oxygen during the pumping procedure. Ferrous Iron (Fe(II)). Concentration of Fe(II) at the WP was up to 0.2 mg/L but was not

PAGE 62

46 Table 3. Change in water composition from the cooling pond (CP) to the wetland from pump (WP) during the dry (March 2007) and rainy (September 2007) seasons.

PAGE 63

47 Figure 14 Continued. (A)

PAGE 64

48 Figure 14 ( A ) Eva luation of T, pH, ORP, DO, H 2 S, SO4, conductivity, and As along the wetland flow path during the dry (March 2007) and rainy (September 2007) seasons; ( B ) : Distribution of Na and Cl along the wetland flow path and at the MWs during the dry (March 2007) and rainy (September 2007) seasons Note: CP cooling pond pump; WP wetland pump; MW monitor wells arranged according to the wetland flow path; Elevated pH, DO, T, positive ORP, and low H 2 S during the dry season were caused by algal blooms.

PAGE 65

49 present in t he CP, which was likely due to high DO levels. 2.3.4 .2. Laboratory analysis Anions. During the dry and rainy seasons the concentrations of most anions decreased along the wetland flow path (Table 3). The F, Cl, and SO 4 levels gradually decreased from 3.1 to 2.8 mg/L, 118 to 107 mg/L, and 58 to 25 mg/L, respectively. The concentration of NO 2 and NO 3 was mostly below detection at the WP and the CP. During the dry season the concentration of Br and PO 4 at the WP was similar to the concentration at the CP. In the contrast, during the rainy season the level of Br was 1.5 times lower but PO 4 was 1. 7 times higher in the wetland. Cations. Generally, the behaviour of most cations had a comparable pattern. During the dry season, the concentration of Na, Mg, K, Ca, and Sr in the wetland was close to the CP. In contrast, during the rainy season the concentration of these cations in the wetland was around 1.5 times less than in the CP. This behaviour could be caused by a dilution effect or due to minor groundwater infl ow. The concentration of Mn and Fe (total) was mostly below detection. The CP water was virtually Si free. At the same time, during the dry and rainy seasons Si at the WP was 0.5 and 2 mg/L, respectively. Arsenic (As) During both seasons the concentration of As significantly decreased along the wetland flow path from 2.3 to < 0.2 g/L. The concentration of conservative tracer elements Na and Cl at the monitor wells (MWs), sampled on April 3 and October 3, 2007, was considerably lower than in the wetland w ater and CP ( Figure 14 B).

PAGE 66

50 2. 3.5 Wetland/filter basin water quality monitoring The major purpose of this study was to investigate the possibility of using a constructed wetland for the wastewater treatment in an area used during phosphate mining for clay settling. The results derived from the field and laboratory analyses are summarized in Table 1. The distribution of pH, ORP, and T in the cooling pond (CP) as the input and wetland (WP) as the output waters is demonstrated in Figure 15 The 18 month monit oring showed a s ignificant cha nge in pH from about 9 to 6.5 7, negative ORP confirming the reducing conditions of the wetland treatment system and a substantial decrease of water temperature (up to 10 C during rainy seasons) Generally, the wetland water T plotted closely to the reported average air T obtained from the Frostproof Station in Polk County (FAWN) (Appendix A) The performance of the wetland was evaluated by mass flux of analytes removed from or contributed to the wetland water (Table 4). The m ass fluxes were calculated as concentration flow rate = mass/time. The flow rates used for calculation varied depending on the season between 5012, 6757 (rainy seasons 2006 and 2007, respectively) and 9255 L/day (dry season 2006). The percent removal of each parameter was calculated as ((mass flux CP mass flux WP)/mass flux CP)*100 with and without the rainfall dilution factor over the study area (Table 4). The average amount of rainfall over the wetland area varied from 1995, 1697 (rainy seasons 2006 a nd 2007,respectively) to 745 L/day (dry season 2006).The corrected concentration with the rainfall dilution factor was calculated as concentration *(flow rate/(flow rate rainfall)) and subsequently used for the mass fluxes and percent removal. Along the w etland flow path most of the monitored constituents were removed

PAGE 67

51 Figure 15 Distribution of pH, ORP, and temperature (T) in the CP input and WP output waters. Note: CP cooling pond pump; WP wetland pump; Average air T measured at elevation of 0 .6 m was obtained from the Frostproof Station in Polk County (FAWN).

PAGE 68

52 from the CP with the exception of Si, PO 4 and Fe. During the rainy season 2007 (May July), however, the wetland contributed mass to the water due to technical difficulties. In addition, during the dry season, the wetland contributed mass in form of K (up to 40 %), likely from plant stem and root systems (Fisher, 1971). 2.3.5.1 Primary, secondary drinking water standards, volatile organic compounds, synthetic organic compounds, and radi onuclides. The results for primary, secondary drinking water standards (PDWS/SDWS), volatile organic compounds (VOC), synthetic organic compounds (SOC), and radionuc lides (RAD) are demonstrated in Tables 5 8 The analysis showed no exceeda nces of the PDWS, VOC, SOC or RAD compounds. At the same time, there were several exceeda nces for the SDWS for the following chemical constituents: were aluminum (7/49), fluoride (55/74), iron (22/74), manganese (10/74), color (56/74), odor (57/74), total dissolved solids (13/49) and foaming agents (1/49). 2.3.5.2. Fecal and total coliform Previous studies demonstrated the capability of constructed wetlands to reduce pathogenic microorganisms in wastewater (e.g., Hill and Sobsey, 2001; Neralla and Weaver, 2000). Removal ef ficiency was > 90% for total coliform and > 80% for fecal streptococcus (Kadlec and Knight 1996). Fecal and total coliform in the CP ranged between <1 370 count/100mL and <1 2000 count/100mL, respectively (Figures 16 A B). The maximum contaminant level (MCL) for total coliform (including fecal coliform) is that no more than 5.0 % of the

PAGE 69

53 Table 4. The wetland performance evaluated with the percent removal of each analyzed parameter.

PAGE 70

54 Table 4 Continued. Note: percentage included the rainfall dilution factor over the site area; Positive values wetland removes mass and negative wetland contributes mass to the water that flows through it; percent removal of total S was calculated from a sum of S(VI ) and S( II); Most of NO 2 and NO 3 was not detected in input and output waters; Fe was not detected in the CP.

PAGE 71

55 Table 5. Analysis of Primary Drinking Water Standards (PDWS).

PAGE 72

56 Table 6. Analysis of Secondary Drinking Water Standards (SDWS).

PAGE 73

57 Table 7. Analysis of synthetic organic compounds (SOC).

PAGE 74

58 Table 8. Analysis of Volatile Organic Compounds (VOC) and Radionuclides

PAGE 75

59 samples should detect total coliform in one month (US EPA Drinking Water Standards). Therefore, in this study, coliform bacteria must be < 1 count/100mL. The highest fecal and total coliform levels were detected in those samples collected at the end of the wetland flow path (WP). Levels were 30 730 count/100mL fecal and 1000 7000 count/100mL for total colif orm bacteria. This maybe caused by feces from fish, reptiles, amphibians, insects, and birds which were abundant in the wetland. In contrast, fecal and total coliform at the FBS/FBN were generally < 2 count/100mL (except one sample of 29 count/100mL) and < 100 count/100mL, respectively. These results clearly demonstrate providing mechanical filtration, degradation of soluble organics and elimination of pathogens, color and odor contaminants (Muhammad et al., 1997; Huisman and Wood, 1974). 2.3.6. Isotopic composition input into 6 monitor wells installed along the wetland flow path; (2) to differentiate potential sources of water in the wetland, and (3) to understand the possible factors control were used in combination with Na data. This approach proved necessary due to the complex hydrogeological framework at the site, which resulted in six possible sources of wate r to the wetland: (1) groundwater, (2) municipal effluent, (3) industrial waste water, (4) seepage from a water body to the north, (5) seepage from a water body to the south and (6) surface runoff/rain.

PAGE 76

60 2.3.6.1. Wetland water During the 18 month period o 18 O composition of wastewater in the cooling pond (CP) was relatively constant and ranged from 20.2 to 29.0 Table 9 Appendix B ). The isotopic composition of treated effluent (EF) discharging into the CP ranged from 8.3 to 1.67 to 18 O. These values were close to the Upper Floridan groundwater line reported by Swancar and Hutchinson (1995). In contrast to the CP, the isotopic composition of wetland water collected fr om a pump (WP) at the end of the flow path 18 O values at the WP ranged from 10.6 to 29.3 18 O of the wetland water from surface (WS) were very similar to the WP and ra nged from respectively. The most depleted isotope values at the WP corresponded to hurricanes Ernesto and Alberto, in June and September 2006. 18 O at the WP with the t 18 O 0.6 (R 2 = 0.98). Kendall and Coplen (2001) (R 2 = 0.96). The regression equation for the Upper Floridan groundwater line published by Swanca 18 O + 1.5 (R 2 = 0.97). The distribution of 18 O for the WP is very similar to the reported water lines but slightly lower in slope and intercept. The lower slope and intercept was caused by the pumping of the CP 18 O 0.97 (R 2 18 O at the FBS and

PAGE 77

61 Figure 16. Fecal and total coliform bacteria detected at the CP, WP, FBS and at the FBN. FBS/FBN had the lowest fecal and total coliform confirming the crucial role of the Note: CP cooling pond pump; WP wetland pump; FBS and FBN filter basin south and north pumps; Concentrations displayed in logarit hmic scale; Time ranged from Sept., 2006 to Sept., 2007. (A) (B)

PAGE 78

62 FBN was compatible with the WP T he trendline equation s for the FBS and FBN were = 4.95 18 O 0.04 (R 2 = 0.97) and = 4.89 18 O 0.23 (R 2 = 0.96), respectively ( Figure 1 8 ). Generally, the offset to the right from the local meteoric water line reflects the influence of evaporation (Gat, 1996; Craig and Gordon, 1965). Evaporation from an 18 O in the remaining water 18 humidity conditions (Kendall and McDonnell, 2006; Craig and Gordon, 196 5). As a result the slope of the meteoric water line drops below 8. In addition to humidity, the slope of evaporation loss depends on a number of environmental factors such as solar radiation, temperature and wind speed (e.g., Clark and Fritz, 1997). The a verage relative humidity of the study area from April 2006 to October 2007 was 74.6 % (FAWN). At this the slope of a local meteoric water line should be close to 5 (Kendall and Coplen, 2001). 2.3.6.2. Monitor wells, SA 8 and N 15 To evaluate possible se epage of water from N 15 and SA 8 into the wetland, 6 18 O in MW 1, MW 2 and MW 3 ranged from 4.19 to 4.95 Appendix B). For MW 18 O values had a highly defined peak in June August 2007 as a result of pumping operation. The CP pump was turned off in April and May for maintenance purposes and restarted in June, which caused the change of isotopi c signature in MW 1 due to its proximity to the 18 O values in MW 4, MW 5 and MW 6 showed less variation and ranged from 27.8 to 4.63 to (A)

PAGE 79

63 Table 9. Minimum, maximum, mean, median an 18 O and Note: CP cooling pond; EF effluent; WP wetland water from pump; WS wetland water from surface; MW 1 to MW 6 monitor wells arranged according to th e wetland flow path; SA 8 and N 15 water bodies to the north and south of the wetland; FBS and FBN filter basin south and north pumping stations, respectively.

PAGE 80

64 Figure 17 8 and N 15 (B). Note : WP wetland water from pump; WS wetland water from surface; CP cooling pond; EF effluent; N 15 and SA 8 water bodies to the north and south of the wetland; Local meteoric water line (LMWL) is from Kendall and Coplen (2001). Global meteoric water line (GMWL) is from Craig (1961). Upper Floridan Aquifer groundwater line (GWL) is from Swancar and Hutchinson (1995).

PAGE 81

65 Figure 1 8 pumps (FBN and FBS). Note: Local meteoric water line (LMWL) is from Kendall and Coplen (2001).

PAGE 82

66 18 O in the Intermediate Aquifer (Sacks and Tihansky, 1996) (Figure 19 A 18 O values in N 15 varied between 1.9 to 18 O in SA 8 were from 10.4 to 3 had the most distinctive isotopic composition compared to the groundwater value and was probably more influenced by the seepage from N 15 or WP (Figure 19). 2.4. Discussion During the 18 month period of monitoring, the constructed wetland (CW) demonstrated promising treatment efficiency for the remed iation of wastewater. The study showed a significant wetland cooling effect on water temperature with up to a 10 C difference between the input cooling pond (CP) and output wetland water from pump (WP) (Figure 15). Seasonal fluctuations of the wetland water temperature could influence the processes of microbial transformation (Kadlec, 1999). Kadlec (2006) studied the surface flow wetlands (Tres Rios, Arizona) for temperature and energy flows. The author reported that wetland water temperature had a ten dency to approach the mean air temperature depending on humidity. In this study, temperature cycles for WP and air demonstrated comparable distribution with the correlation coefficient R 2 of 0.83 (Figure 15). Treatment surface flow CW can have one or two t hermal regions depending on the water residence time (Kadlec, 2006). When the residence time is > 5 days, the wetland can be divided into 2 zones: accommodation and balance zones (Kadlec, 2006). In an accommodation region, the temperature and evapotranspir ation profiles are initially steep due to adjustments to weather conditions. In a balance zone, temperature in the wetland

PAGE 83

67 Figure 19 Note: MW 1 to MW 6 monitor wells arranged according to the wetland flow path, GW* Floridan groundwater from the Intermediate Aquifer (well ROMP 45, Sacks and Tihansk y, 1996), RW* ra in water ( Kish et al., 2009); N 15 and SA 8 water bodies to the north and south of the wetland; WP wetland water from pump. Upper Floridan groundwater line (GWL) is from Swancar and Hutchinson (1995). (A) (B) (B) (A)

PAGE 84

68 approaches to the balance level, i.e. adjusts to th e existing meteorological conditions (Kadlec, 2006). This finding was very useful for evaluating temperature along the CW flow path. Generally, the water temperature was gradually reduced along the wetland and could be divided into 2 zones with the bound ar y line around 1300 m (Figure 14 A). The cooling effect of the wetland could probably reduce evaporation losses. Meteorological monitoring, conducted during the study, demonstrated the dominance of 3 tropical seasons: one dry and two rainy seasons (Figure 1 2 ). Therefore, it is very important to evaluate the performance of the treatment system in response to seasonal changes such as heavy rainfall and drought events. Generally, the behavior of most cations had a comparable pattern. During the dry season, the c oncentration of Na, Mg, K, Ca, and Sr in the wetland was close to the CP. In contrast, during the rainy season the concentration of these cations in the wetland was around 1.5 times less than in the CP. This behavior could be caused by dilution of the wetl and water by rainfall. Elevated pH, DO, T, positive ORP of the wetland surface and low H 2 S during the dry season were generally associated with algal blo oms during spring time (Figure 14 A). Sawyer and McCarthy (1978) reported that in shallow ponds during d aylight algae use CO 2 for photosynthesis and release O 2 increasing the pH and DO levels as the carbonate bicarbonate equilibrium is destabilized. During night time hours this process is generally reversed when algae and plants stop producing O 2 but start u sing the available oxygen. In order to quantitatively establish the efficiency and the possible groundwater input into the wetland, the Na and Cl mass fluxes in the CP and WP were examined (Table 4, Figure 20 ). Both curves showed a comparable distribution as well as a significant change throughout the duration of the study. During the first rainy season

PAGE 85

69 (April October 2006), the mass fluxes of Na and Cl at the WP were around 60 % less than in the CP. These reduced mass fluxes were due to heavy rainfall dur ing two hurricanes, which added low conductivity water directly and triggered enhanced groundwater input into the wetland (Criss and Winston, 2003). During the dry season (November 2006 April 2007), the percentage removal of Na and Cl from the wetland wa s close to 0 % indicating the normal operation of the wetland treatment system, i.e. the balance between input and output. This allowed to estimate the residence time of water in the wetland to about 120 days. During the second rainy season (April Octobe r 2007), the percent removal of Na and Cl from the wetland varied significantly from 52 to 4 %. At the end of April, the CP pump was turned off and the wetland surface was lowered approximately 1 m for maintenance purposes (Figure 20). From May 2007, the wetland treatment system became again operational. Therefore, during this time a considerable mass flux of Na and Cl accumulated in the wetland could be caused by mechanical issues as well as high evaporation. As soon as the maintenance problems were fixed the percentage removal of Na and Cl from the wetland was close to 0 %. Figure 21 was plotted to assess the variation of a weight concentration SO 4 /Cl in the CP and WP as a function of time. The concentration of SO 4 changed through microbiological reducti on which is greatly influenced by seasonal temperature (Urban et al., 1994). High temperatures during summer and fall accelerate microbial activity in decomposing organic material, thus producing higher levels of H 2 S (Armannsson, 1999). The plot demonstrat ed a variability of SO 4 /Cl ratio in the wetland throughout the period the study with the lower levels during the rainy seasons 2006 and 2007 (< 0.15 and < 0.3, respectively) and the highest (up to 0.6) during the dry season 2006 (Figure 21). The

PAGE 86

70 Figur e 20 Mass fluxes of Na and Cl in the CP and WP, and the calculated percent removal of Na and Cl from the wetland. Note: CP cooling pond pump; WP wetland pump; WS water level wetland surface water level; NGVD National Geodetic Vertical Datum; Calcul ation included the rainfall dilution factor over the site area; Positive values wetland removes mass and negative wetland contributes mass to the water that flows through it.

PAGE 87

71 monitoring of the treatment system frequently showed higher concentrations of total Fe at the filter basin south (FBS) and north (FBN) pumps compared to the WP (Figure 22). During the dry season the average Fe level at the FBS was 0.34 mg/L, while at the WP and FBN it was 0.26 and 0.14 mg/L, respectively. In contrast, during the ra iny seasons of 2006 and 2007, the average Fe at the FBS was up to 1.20 mg/L, while at the WP and FBN it was 0.20 and 0.42 mg/L, respectively. The distribution of Fe concentration at the WP, FBS and FBN with time demonstrated lower Fe at the WP and FBN comp ared to the FBS. At the same time, concentrations of Fe at the WP, which was being applied to the surface of the filter basin (FB), remained relatively constant, while Fe at the FBS fluctuated quite significantly. Moreover, substantially higher Fe levels a t the FBS were detected particularly during the rainy seasons potentially indicating an outside source of water to the FB containing high Fe. This is possible due to the migration of the surficial groundwater into the FB from the surrounding area. Previous studies reported that the concentration of Fe in the surficial groundwater of the study area could be about 12 mg/L (ROMP 57A; Sacks and Tihansky, 1996). Field observations confirmed that the levels of the surficial groundwater outside the FB increased du ring heavy rainfall therefore increasing the hydraulic gradient around the FB. This fact was supported by visual observation of iron mud build up clogging flow meters and pipes of the FB, especially at the FBS. 2.4.1. Behavior of arsenic Arsenic (As) is a n element of great interest in Florida principally to aquifer storage and recovery (ASR) (Arthur et al., 2005). Studies showed that geogenic As was

PAGE 88

72 Figure 21 Variation of SO4/Cl in the cooling pond pump (CP) and the wetland pump (WP) as a funct ion of time. Note: Values are in mg/L.

PAGE 89

73 found in the carbonate Floridan Aquifer mostly associated with pyrite (Lazareva and Pichler, 2007; Price and Pichler, 2006). Pyrite is thought to dissolve and release As during ASR due to the injection of ozone treate d oxidizing surface waters into reducing native groundwater. As a result, concentrations of As in recovered water were up to 130 g/L (Arthur et al., 2005), far above the 10 g/L drinking water standard (DWS) for As (EPA). The type of constructed wetland/f ilter basin treatment system studied here could become a new alternative to treat wastewater in Florida and beyond. Water treatment processes in the wetland may consist of metal accumulation into vegetation, adsorption on soil particles, precipitation or c o precipitation caused by microbial activity (Stottmeister et al., 2006; Jacob and Otte, 2003; Stoltz and Greger, 2002; Dushenko et al., 1995). Following treatment through a CW the water would be in reducing conditions with high sulfide and low oxygen lev els which are favorable for the stability of pyrite (e.g., Jones and Pichler, 2007). Water of this composition would be physically and chemically similar to native groundwater and should not cause the dissolution of pyrite and leaching of As during ASR. T he pH, DO and temperature gradient from the CP to the wetland could be important for the behaviour of heavy metals such the redox sensitive element As (Stottmeister et al., 2006). Concentrations of As in the CP (up to 5 g/L) were considerably reduced at t he WP and the FBS/FBN (< 2 g/L) (Table 1). The mass loading of As in the CP and the wetland was calculated to understand the removal effectiveness of the wetland over time (Figure 23). The calculation was done with the rainfall dilution factor over the we tland study area. Generally, As in the wetland was reduced to 40 95 % of the original except on 07/03/08 it was 2 times higher than in the

PAGE 90

74 Figure 22 Distribution of Fe at the WP, FBS and FBN with time Note: WP wetland pump; FBS and FBN fil ter basin south and north pumps.

PAGE 91

75 Figure 23. Mass fluxes of As in the CP and WP. Note: CP cooling pond pump; WP wetland pump; Calculation included the rainfall dilution factor over the site area.

PAGE 92

76 CP, which could be due to As containing herbi cides and insecticides used at the Hines Energy Complex. Seasonal variations of As in the treatment system showed lower concentrations during summer and fall and higher during winter and early spring. In the summer fall seasons, when plants and algae gro w they uptake and immobilize nutrients such as phosphorous and nitrogen into their biomass. Because of the high affinity between arsenate [As (V)] and phosphate, plants easily incorporate As (V) into their cells (Catarecha et al., 2007). In contrast, the r educed photosynthesis and decay of plants during winter early spring facilitates the decrease of DO in water and releases nutrients back to the water column such as P and As. During certain seasons, plants and other wetland species could convert inorgani c nutrients to organic compounds resulting in a net export of nutrients from the wetland (Devito and Dillon, 1993). In addition to plant accumulation, the sandy clay based sediment rich in organic material could be an important sink for As retention. Bud dhawong et al. (2005) constructed a bench scale experimental wetland in Grosskaya Beuna area (Germany) to simulate the treatment of acid mine drainage. The authors reported that the highest efficiency of water treatment was achieved in the constructed wetl and with the combined planted gravel/soil system. This type of wetland maintained the essential conditions (i.e., pH, redox, surface area) for As binding rather than those wetlands which were constructed exclusively using vegetation or soil matrix (Stottme ister et al., 2006). At the same time, As can be co precipitated with iron oxides caused by root oxygen transport into rhizosphere. This transfer stimulates a significant metal buildup at the sediment root boundary forming iron plaques on the plants roots (Colmer, 2003). In additio n, elevated concentrations of HS /H 2 S in the constructed wetland could develop favorable conditions

PAGE 93

77 for coagulation or coprecipitation of As with S 2 Langner et al. (1999) suggested that the formation of As(III) S 2 phases coul d be an important sink of As (III) under reduced conditions such as wetlands. Also, the authors reported rapid reduction of As (V) and SO 4 2 to As (III) and S 2 using controlled wetland chambers. Additional contributions of As (V) and SO 4 2 could cause the formation of an amorphous As 2 S 3 or Fe AsS phases. 2.4.2. Evaluation of wetland performance using isotopic mass balance approach The concentrations of Na in conjunction with 18 O values were used in the isotope/chemical mass balance approach to differentiate potential sources of water in the wetland and monitor wells and to understand factors controlling the flow of seepage water into the wetland. The concentration of Cl, whic h can also be used as a tracer, was not used due to the possible non conservative behavior of Cl in wetland waters. Varner et al. (1999) reported that wetland can be a substantial source of methyl chloride (CH 3 Cl) emission to the atmosphere. The flux of th is compound is biologically mediated and greatly depends on temp erature and vegetation density. Also, Cl can be incorporated into plant tissues (Alloway, 1992) or adsorbed onto soil or mineral surface through a non specific adsorption (Katou et al., 1996; Altman, 1994). 2.4.2. 1. Water composition in the wetland The composition of wetland water can be described by a mixture of four possible sources such as: (1) Floridan groundwater from the Intermediate Aquifer System ( ROMP 45 from Sacks and Tihansky, 1996) (2) cooling pond water, (3) water body to the north (N 15), and (4) to the south (SA 8) of the wetland (Figure 24 ). The groundwater and

PAGE 94

78 had a tendency to approach the Up per Floridan groundwater line (Figure 19 A) ( Kish et al., 2009; Kendall and Coplen, 2001; Swancar and Hutchinson, 1995). When these water sources are considered as end members to the wetland water (i.e., have unique chemical composition compared to the mi xture), it is possible to estimate mixing proportions between them waters using a mass balance approach (e.g., Doctor et al., 2006; Christophersen and Hooper, 1992; Clark and Fritz, 1997). For this study, the following linear mass balance equations were ap plied to describe the wetland water: (A) Northern side of the wetland This three component mass 18 O and Na) and three equations to determine each variable. Assuming that the wetland water was a result of mixing ground water (GW), cooling pond water (CP) and N 15, the following equation could be used to assess the individual contributions from each source: (1) Making a substitution of equation (1) into the isotopic mass 18 O, (2) and substitution of equation (1) into the chemical mass balance equation for Na (3) followed by a combination of the equations 1, 2, and 3 to determine the proportion or mass of each water source in the wetland wate r :

PAGE 95

79 Figure 24 (A) Map of the study area including water transfer system from the cooling pond to the U shaped constructed wetland; (B) Cross section along transect 1 Note water sampling locations: MW 1 to MW 6 monitor wells; WP wetland water from pump; WS wetland water from surface; N 15 and SA 8 water bodies to the north and south of the wetland. Arrows represent possible groundwater flow paths. MW 5 MW 2 A Surficial Aquifer Confining Clay Unit Intermed iate Aquifer Constructed Wetland SA 8 N 15 B N 15 SA 8 N Cooling Pond WP 100 m MW 6 MW 3 MW 5 MW 4 MW 2 MW 1 B A WS (A) (B) EF

PAGE 96

80 (A) Southern side of the wetland The mass balance equations for the southern side of the wetland were analogous to the equations above with the substitution of N 15 for SA 8: Subscripts for the m CPn m CPs m GWn m GWs m N 15s and m SA 8n indicate the mass or percentage of each end member in the wetland water calculated for the northern ( n ) and southern ( s ) sides. The wetland mass balance was calculated using average values of 18 O and Na from the CP, N 15 and SA 8.

PAGE 97

81 Evaluation of the final percentage of four water sources (GW, CP, N 15, and SA 8) present in the wetland was calculated as an average value of waters from the northern and southern sides and was based on the followi ng (Figure 2 5 ) : Therefore, ; Similarly to the equation above, ; and 2.4.2.2 Water composition in monitor wells Similarly to the wetland wat er described above, the water composition in the monitor wells (MWs) was a mixture of three possible sources (Figures 24, 26 Table 11): MW 1, 2 and 3 : (1) groundwater (GW), (2) wetland water (WP), and (3) N 15. MW 4, 5 and 6 : (1) groundwater (GW), (2) wetl and water (WP), and (3) SA 8. Based on these assumptions the following linear mass balance equations were applied to evaluate the proportions of different waters present in the monitoring wells: MW 1 to MW 3 :

PAGE 98

82 Figure 25 Four end members for t he WP used for the mass balance approach. Note: WP wetland water from pump; CP cooling pond; GW Floridan groundwater from the Intermediate Aquifer System (well ROMP 45 from Sacks and Tihansky, 1996); N 15 and SA 8 water bodies to the north and sou th of the wetland; Error bars standard deviation.

PAGE 99

83 Table 10. Chemical and isotopic data of waters used for the mass balance approach. Note: CP cooling pond; SA 8 and N 15 water bodies to the no rth and south of the wetland; WP wetland water from pump; GW groundwater from the Intermediate Aquifer System (well ROMP 45 from Sacks and Tihansky, 1996) ; (*) values in mg/L; m mass or percentage of each end member in the WP.

PAGE 100

84 MW 4 to MW 6 : The ma ss balance equations for MW 4 to MW 6 were similar to the above with the substitution of N 15 for SA 8.

PAGE 101

85 The final calculated composition of the wetland water from pumping s tation (WP) is shown in Figure 27 It clearly demonstrates the changes in the WP quality throughout the duration of the study reflecting the influence of dry/rainy seasons and pumping operations. At the beginning of the monitoring, the wetland consisted of a mix of waters from the SA 8, N 15, GW, and CP. During the period of April July 2006 the composition of water at the northern side of the wetland had the following changes: input of N 15 decreased from 100 to 39 %, GW increased from 7 to 35 %, and the CP increased up to 26 %. At the same t ime, the composition of water at the southern side of the wetland showed that the contribution from SA 8 dropped from 63 to 19 %, GW increased from 16 to 46 %, and the CP increased to 35 %. However, between August and September 2006 the CP inflow dropped t o 22 % bu t the GW input increased up to 7 8 %. According to the operational data, before the end of September 2006 the pumping of the CP water into the wetland had maintenance and power issues and was periodically turned off. In addition, the considerable G W inflow was impacted by heavy rainfall during two hurricanes in June early September 2006 which added low conductivity water directly and triggered enhanced groundwater input into the wetland (Criss and Winston, 2003). Later, during the dry season ( Novem ber April 2006 ) the WP contained mostly the CP water (88 100 %) and minor inflows of GW (< 12 %) that were caused by short

PAGE 102

86 Figure 26. Three end members for the monitor wells MW 1 to MW 3 (A), and MW 4 to MW 6 (B) used for the mass bal ance approach. Note: WP wetland water from pump; CP cooling pond; GW Floridan groundwater from the Intermediate Aquifer System (well ROMP 45 from Sacks and Tihansky, 1996); N 15 and SA 8 water bodies to the north and south of the wetland; Error ba rs standard deviation. (A ) (B )

PAGE 103

87 Table 11. Chemical and isotopic data of the monitor wells (MW 1 to MW 6) used for the mass balance approach. Note: (*) values in mg/L; m mass or of each end member in MW.

PAGE 104

88 mechani cal problems. During the second rainy season (May October 2007) the CP water inflow remained high (87 100 %). However, the maintenance/power issues and a significant rainfall in August October caused the GW seepage to increase up to 13 %. Sacks (2002 ) used the isotope mass balance method for estimat ing the water balance of lakes in central Florida. The author reported that the majority of lakes in upland areas of Polk and Highlands Counties had from medium to high groundwater inflows with 25 50 % an d > 50 % of total inflow, respectively. The inflows depended on topography, humidity, air and lake surface temperatures, lake depth, distance downward to the Upper Floridan A quifer (thickness of the Surficial Aquifer and Intermediate Confining Unit) and fr action of wetlands. Evaluation of the wetland surface water (WS) along the flow path showed a distinct pattern of change in isotopic signature from the input CP to the output WP (Figure 28 ). The wetland transect was done three times with the interval o f about three weeks. First transect was performed at the beginning of the study on April 24, 2006. The 18 the wetland were relatively heavy, ranging from 0.33 to 3.16 % he beginning of the wetland 18 O and were depleted by 3.88 and 17.3 second transect on May 15, 2006 showed a slow isotopic depletion along the wetland flow path where 18 respectively. According to the mass balance calculation explained above, the wetland water was composed of 50 % of N 15, 37 % of SA 8, 8 % GW, and 5 % of the CP waters (Figure 27). T he third transect, performed after the hurricane Alberto (June 27, 2 006), demonstrated even higher depletion of 18 0.02 to 3.25 and

PAGE 105

89 Figure 27. The calculated mass of each end member in the wetland using an isotope/chemical mass balance approach. Note: WP wetland water from pump; CP cooling pond; GW Floridan groundwater fr om the Intermediate Aquifer System (well ROMP 45 from Sacks and Tihansky, 1996); N 15 and SA 8 water bodies to the north and south of the wetland; Error bars standard deviation.

PAGE 106

90 3.1 to 14.5 balance estimation showed that the wetland water consisted of 41 % of GW, 28 % of N 15, 18 % of CP, and 13 % the SA 8 waters. Therefore, the decrease in isotopic fractionation of 1.91 18 O and 13.2 for coincides with the drastic GW input into the wetland after a strong hurricane. At the same time, it is interesting to note that distinctive depletion of 18 was evident between 1300 and 2700 m but reverse b etween 0 and 1300 m of the wetland flow path. The calculated proportions of three end members in MW 1 to MW 6 estimated by the isotope/chemical mass balan ce approach are shown in Table 11 The composition of MW 1 to MW 3 was substantially induced by the s eepage from the N 15 with the highest levels at MW 2 and MW 3 (up to 91 %). The estimated GW inflow varied from 0 to 100 % decreasing from MW 1 to MW 3. The highest input from the WP (up to 100 %) was to MW 3 in August October 2007. The composition of MW 4 to MW 6 was mostly controlled by the GW inflow (40 100%). The estimated seepage from the SA 8 was < 28 % with increasing inflow from MW 5 to MW 4. The highest inflow from the WP (< 42 %) was to MW 6 (except one MW 4 sample from 10/30/08). The influenc e of each end member on the composition of MWs could be caused by several factors such as (1) lithologic settings of the study area causing a variability in porosity and permeability; (2) total depth of MWs; (4) hydrologic gradient; (5) proximity of N 15 a nd SA 8 to the wetland ; and (6) periodic variations in precipitation (dry/rainy season). The lithology of MWs cores along the wetland was relatively uniform and sediments were composed of light tan to brown poorly to well sorted fine sands or silts with occ asional grey clay nodules. The brown color was due to the presence of organic

PAGE 107

91 Figure 28. Evaluation of wetland surface water isotopic composition along the flow path. Note: CP cooling pond; WP wetland water from pump.

PAGE 108

92 material. Generally, clay content was higher at MW 4 to MW 6 compared to MW 1 to MW 3. At the same time, of all wells MW 3 had more medium fine sands at the top 1.5 m. The monitoring of water level elevations showed that the SA 8 had a higher level compared to N 1 5, CP, and the wetland (Figure 13 ). However, the isotope/chemical mass balance method showed that the seepage from SA 8 was substantially lower than from N 15. The higher inflow of N 15 into MWs could be due to a closer location to the wetland compared to SA 8. In addition, MW 1 to MW 3 were shallower than MW 4 to MW 6 and were less affected by the GW inflow but more affected by the N 15 dilution. The total depths of MW 1 to MW 3, and MW 4 to MW 6 were 3 and 4 m, respectively. As a result, the deeper wells co uld generally have a higher input of GW. Despite the fact that the inputs from N 15 and SA 8 were detected in MWs, they were not identified in the wetland once the treatment system became fully operational potentially indicating a water loss from the wetla nd. 2.5. Conclusions 1) Evaluation of the wetland performance during dry and rainy seasons was important to assess the reliability of the treatment system in time and space. The wetland transect showed a distinct pattern of change in water chemistry along t he flow path from the input cooling pond (CP) to output wetland water from pump (WP). 2) The study showed the following changes in water quality from the cooling pond (CP) to the constructed wetland/filter basin treatment system (CW/FB): 1. Substantial decr ease of water temperature (up to 10 C); 2. Significant change in pH from about 9 to 6.5 7;

PAGE 109

93 3. Negative ORP confirming the reducing conditions of the treatment system; 4. Substantial increase of H 2 variations of As in the wetland showed lower concentrations during summer fall and higher during winter early spring. 6. Substantial reduction of SO 4 F, Cl, NO 3 NO 2 Br, Na, K, Ca, and Mg. 7. No e xceeda nces of th e primary drinking water standards volatile organic compounds, synthetic organic compounds, an d radionuclides. Several exceeda nces for the secondary drinking water standards, such as Al, F, Fe, Mn, color, odor, total dissolved solids, and foaming agents. 8 Reduction of fecal and total coliform at the FBS/FBN from 30 730 and 1000 7000 count/100 mL to < 2 and < 100 count/ 100 mL, respectively. These results clearly 3) 18 O) in combination with geochemical data were useful tools to discriminate major sources of water in the constructed wetland and monitor wells (MWs). 4) The composition of water in MWs was determined to be ground water dominated. However, water in MW 1 to MW 3 was substantially induced by the seepage from N 15. 5) The possible factors controlling the fluid mixing in the monitoring wells (MWs) can be: (1) lithologic settings of the study area causing a variability in p orosity and permeability; (2) total depth of MWs; (4) hydrologic gradient; (5) proximity of the

PAGE 110

94 water bodies to the north and south (N 15 and SA 8) of the wetland; and (6) periodic variations in precipitation (dry/rainy season). 6) The composition of water i n the wetland varied throughout the period of the study. During the first 6 months of monitoring, the wetland water was composed of a mix of waters from SA 8, N 15, GW, and CP. The water quality was impacted by major hurricanes in June early September 2006 and inconsistent pumping operation due to maintenance/power issues. Once pumping operations stabilized and without the influence of hurricanes, the wetland water was mostly composed of the CP water (88 100 %) and minor inflow from GW caused by occasiona l rainfall and short mechanical problems. T he performance of the wetland/filter basin treatment system showed a great potential to improve the water quality of industrial and m unicipal wastewater. Despite significant seasonal variations with respect to tem perature, rainfall and hu midity, the chemical/microbiological composition of treated water remained relatively constant.

PAGE 111

95 CHAPTER THREE BENCH SCALE LEACHING EXPERIMENTS TO INVESTIGATE GEOGENIC ARSENIC CONTAMINATION IN THE FLORIDAN AQUIFER 3.1. Introduc tion During the past 20 to 30 years numerous occurrences of elevated arsenic (As) concentrations in groundwater were reported. With a few exceptions the source of As was geogenic, i.e., naturally occurring in the aquifer matrix. The release of As from the aquifer, however, was generally caused by anthropogenic perturbations of the phy sicochemical conditions, especially a redox change. There are a multitude of scientific publications addressing this issue, including several excellent reviews (e.g., Amini et al., 2008; Ferguson and Gavis, 1972; Korte and Fernando, 1991; McNeill et al., 2002; Smedley and Kinniburgh, 2002). This type of As contamination is a public health issue world wide. In particular the catastrophic problems in Bangladesh and West Bengal hav e been front page stories in newspapers and scientific journals (e.g., Ahmed et al., 2006). Despite more than 20 years of ongoing research into the As problem in Bangladesh there is no consensus about the geochemical mechanisms of As mobilization. The diss olution of hydrous ferric oxide (HFO) is called upon, as well as the oxidation of pyrite (e.g., Polizzotto et al., 2005; Swartz et al., 2004; van Geen et al., 2008). Nevertheless, it is becoming clear that the problem of As release is a complex problem of

PAGE 112

96 agriculture, hydrogeology, microbiology and inorganic geochemical processes, which reaches far beyond the simple approach of oxic vs. anoxic conditions (Dhar et al., 2008; Stute et al., 2007; Zheng et al., 2005). Similar to the problem outlined for Bangla desh, there are many other locations where similar geogenic As contamination exists; independent of the aquifer matrix, whether fluvial sediments, marine shale or carbonate rocks. Surprisingly little is known about geogenic As contamination in limestone/ca rbonate aquifers. A thorough literature search provided only 5 references (Armienta and Segovia, 2008; Gbadebo, 2005; Romero et al., 2004; Simsek et al., 2008; Vesper and White, 2003). Limestones, although excellent aquifers, are problematic because due to karstification, contaminants can be easily transported over large distances, posing a threat to public and private water supplies (e.g., Ducci et al., 2008; Katz, 2004; Kovacova and Malik, 2007; McMahon et al., 2008; Obeidat et al., 2008; Zhou and Beck, 2 008). Groundwater can flow through conduits so that there is little opportunity for filtration or sorption of contaminants onto aquifer material. Thus it is important to assess and understand geogenic As contamination in limestone aquifers, where anthropog enic perturbations could cause the release of As from the limestone matrix at a large distance from the area where eventually elevated As values occur. Arsenic in the Floridan Aquifer, an extensive carbonate aquifer, is mostly associated with pyrite as a substitute for sulfur in the FeS 2 structure (Lazareva and Pichler, 2007; Price and Pichler, 2006). Thus, a change in the redox conditions from anoxic to oxic could cause the dissolution of pyrite and the mobilization of As (Figure29). This scenario is the current working hypothesis to explain elevated As

PAGE 113

97 Figure 29. Concentrations of As and Ca through time for the ASR Punta Gorda cycle during aquifer storage and recovery (ASR) operation in Charlotte County, southern Florida (from Arthur et al., 2002 ).

PAGE 114

98 concentration observed during aquifer storage and recovery (ASR) operation s in Florida The principle behind ASR is the storage of treated surplus surface water in a confined aquifer during rainy seasons followed by its recovery during times of need During ASR cycle tests, the injection of oxygenated surface water into reducing native groundwater causes a transformation of the redox environment, oxidative dissolution of pyrite, and the release of As with values in recovered water of up to 130 g/L (Ar thur et al., 2005; Arthur et al., 2002; Williams et al., 2002; Arthur et al., 2001). The objective of this investigation was to carry out a series of bench scale leaching experiments to study the mobilization of geogenic As under a range of redox conditio ns. 3.2. Methods 3.2.1. Rock preparation and analysis Rock cuttings were collected during well installation in central Florida (Polk County). Stratigraphically, the cuttings represent the Suwannee Limestone, Ocala Limestone and the Avon Park Formation. Un til time of analysis they were stored under a nitrogen atmosphere at the Center for Water and Environmental Analysis, University of South Florida (USF). Later, the cuttings were dried and carefully examined with a 10x hand lens as well as a stereo microsc ope for the possibility of post drilling oxidation of redox sensitive minerals, such as pyrite. Generally 2 random samples and 2 samples of special interest were taken per interval. Samples of special interest represented single chips of rocks that were de emed to have higher As concentrations, such as, clays, pyrite, etc. Several samples were selected for a more focused mineralogical analysis using a

PAGE 115

99 Hitachi S 3500N scanning electron microscope (SEM) with energy dispersive spectrometry (EDS) at the College of Marine Science, USF. To determine bulk chemical composition the samples were powdered and homogenized in an agate mortar. To avoid cross contamination, the mortar and pestle were cleaned with pure silica sand and de ionized water (DI) water between samp les. Approximately 0.5 0.01 g of powdered sample was weighed into a digestion vessel and 10 mL of a 3:1 mixture of hydrochloric (HCl) and nitric acid (HNO 3 ) was added. The digestion vessel was immediately capped with a reflux cup to trap any arsine gas. The sample was heated in a HotBlock TM to a temperature of 95C for 30 min. After digestion, the solution was cooled, diluted to 50 mL (5:1) with DI, centrifuged for 15 min, filtered out with FilterMate TM to remove any suspended solids (silicates), and stor ed until analysis. Calcium, Fe, Mg, Mn, S, P, Si and Al were determined using a Perkin Elmer Optima 2000 DV inductively coupled plasma optical emission spectrometer (ICP OES). Sample preparation consisted of the dilution of 2 mL of the stored sample sol ution with DI into a 10 mL centrifuge tube. The sample dilution (5:1) was needed due to the high acid concentration. Bulk As concentration was measured on a PSA 10.055 Millennium Excalibur hydride generation atomic fluorescence spectrometer (HG AFS). The a ccuracy and precision of the measurements were better than 5% verified through the use of internal and external standards. Acid blanks for digestion and HG AFS did not reveal any detectable As. Background signal drift monitor was less than 2%. In total 206 rock samples were dissolved and subsequently analyzed (Appendix C, D, and E). After the chemical analysis, rock cuttings from the following sampling intervals were chosen for the bench scale column experiments based on results explained below in

PAGE 116

100 section 3.3.1.: Suwannee Limestone ( 136 m to 139 m, and 171 m to 174 m), Ocala Limestone (195 m to 199 m), and the Avon Park Formation (255 m to 257 m; and 260 m to 261 m) (Table 13). 3.2.2. Water collection and analysis During the experiments 7 different types o f water were used: (1) Tampa drinking water; (2) Native Floridan groundwater; (3) Native Floridan groundwater saturated with air; (4) Wetland water; (5) Wetland water exposed to UV; (6) Water from the filter basin, and (7) Water from the filter basin treat ed with UV. A complete chemical analysis is provided in Appendix F. Drinking water was directly collected from a tap in the laboratory after extensive flushing. That water was used as an analogue for ASR water (injectate), because ASR injectate has to be treated to meet all drinking water standards and requirements. Thus, tap water and ASR injectate are chemically and physically similar, i.e., pH, T, high dissolved oxygen and high oxidation potential level. Groundwater was collected from a well located on the USF Tampa campus and used as a baseline for comparison with the other waters. The production zone of that well was from 103.9 m to 184.4 m. Groundwater was sampled carefully into 9.5 L amber carboys filled with N 2 to avoid contact with atmospheric oxyg en and the resulting change in redox. The carboys were equipped with quick disconnect closures, which have internal valves that automatically close when a fitting was removed. Immediately after collection, the water was transported to the laboratory for th e leaching tests. To evaluate As mobilization in

PAGE 117

101 the presence of O 2 the same groundwater was saturated with air prior to injection into the columns. Wetland water was obtained from the wetland/filter basin treatment system discussed in the second chapter. It was collected directly from the wetland from a depth of 0.5 m using a peristaltic pump to ensure uniform water flow (Figure 30). Similarly to groundwater sampling, wetland water was stored in 9.5 L amber carboys treated with N 2 prior to collection. Afte r collection, carboys were immediately transported to the laboratory for the leaching tests. The experiments were divided into a series of tests with wetland water, as collected and treated with ultra violet light (UV) prior to injection (Figure 30). The a pplication of UV is of importance in commercial water treatment to kill bacteria. Thus it was deemed important to investigate how UV treatment would affect the dissolution of pyrite and the associated mobilization of As. Filter basin water was collected fr om the filter basin pump at the same location as the wetland water. The experiment was divided into a series of tests with the untreated and the UV treated filter basin water. In contrast to the wetland water, water from the filter basin was exposed to UV directly in the field. Following the sampling procedure, the water samples were analyzed as described in the second chapter. The accuracy and precision of the measurements were verified through the use of internal and external standards, indicating a prec ision and accuracy better than 5%. In addition, the concentration of dissolved organic carbon (DOC) and total N were measured by Dr. Thorsten Dittmar at the Department of Oceanography, Florida State University.

PAGE 118

102 Figure 30. Collection of the wetla nd water (WW) from a depth of 0.5 m using a peristaltic pump (top).Bench scale column leaching experiments using WW (bottom). Note: leachate samples on the bottom are indicated by arrows. N 2 WW WW

PAGE 119

103 3.2.3. Experimental setup The leaching experiments were performed i n standing PVC columns of 0.019 m in diameter, 0.3 m and 1 m in length using the different water compositions and aquifer matrices described above. Prior to each experiment, rock cuttings were dried (to prevent oxidation of pyrite), ground and homogenized in agate grinder to a size of coarse sand to increase the surface area, but at the same time to avoid clogging of tubing lines. After packing, the columns were flush ed with nitrogen for about 24 hours to eliminate any oxygen present in the pore space. The source water was percolated into the columns against gravity entering from the bottom, thus allowing for a more uniform flow through the column and a complete saturation of the rock. To achieve a flow rate of 2 mL/min and to avoid any contact with atmosphe ric oxygen Watson Marlow multichannel peristaltic pumps were used (Figures 30 and 31). A membrane filter was not used to ensure an undisturbed water flow through the columns. The leachate samples were collected from the top of the columns at certain interv als (every 60 mL) and subsequently analyzed for physical and chemical parameters described in second chapter. The column porosity was determined as the ratio of the weight of water filled the column to the total weight of water and sediment in the column subtracting the weight of the column as n = mass (water)/ (mass (water + sediment)) 100 %. 3.3. Results 3.3.1. Rock chemical composition Geochemical and mineralogical analysis of core material for this study corresponded to the reported studies of the Suwannee and Ocala Limestones, and the

PAGE 120

104 Figure 31. Bench scale leaching experiments using wetland water (WW) treated with UV (top) and drinking water (DW) (bottom). Note: yellow color in leachates was due to pyrite oxidation (indicated by arrows ). N 2 UV WW DW DW

PAGE 121

105 Avon Park Formation (Dippold, 2009; Price and Pichler, 2006; Scott, 1990; Scott 1988; Miller, 1986). Complete lithologic descriptions of all core samples were listed in Appendix C, D, and E. The majority of the analyzed samples from the Suwannee Limesto ne were composed of pure white limestone with a minor percentage of dolomite, clay, chert, quartz, iron oxides, phosphate grains, and pyrite. Generally, the limestone had intergranular and high moldic porosity zones. The mineralogical composition of the Oc ala Limestone consisted of white dolomitized fossiliferous limestone, poorly to moderately indurated with minor quartz, iron oxides, phosphate, clays and pyrite. Porosity was moldic and intergranular. The majority of samples from the Avon Park Formation we re dolostone dominated with a variable percentage of limestone, minor quartz, pyrite, phosphate, iron oxides, dark brown clays, and gypsum. Commonly, the dark brown dolostone was highly fractured and sucrosic in texture. The SEM analysis of the samples sel ected for the column leaching experiments confirmed the presence of euhedral and framboidal pyrite crystals with unoxidized surfaces (Figures 32 34). Bulk rock chemical composition conducted on the ICP OES and HG AFS for Ca, Fe, Mg, Si, S, P, Al, Na, K, and As of all samples were presented in Appendix C, D, and E. Minimum, maximum, mean, and standard deviation of As concentrations were listed in Table 12. The mean included both interval samples and samples of the special interest and, therefore, should be used with caution. Based on those data it seemed that the Ocala Limestone may be the preferred aquifer storage and recovery (ASR) storage zone (not considering any hydraulics) with the lowest potential to contaminate groundwater with As. The injection of oxygen rich surface water would mobilize As from all three units, but the amount released from the Ocala should be comparatively less. However, the

PAGE 122

106 Figure 32. Photomicrograph and the scanning electron micrograph of framboidal and euhedral py rites (shown by arrows) found in the Suwannee Limestone (136 m to 139 m). Note: Bar scale in lower right is 10 m. 1 mm

PAGE 123

107 Figure 33. (A) Photomicrograph and (B) the scanning electron micrograph of framboidal pyrites found in the Ocala Lim estone (195 m to 199 m); (C) EDS spectra confirming the presence of pyrite. Note: Bar scale for the SEM image in lower right is 50 m. 1 mm (A) (B) (C) 1 mm

PAGE 124

108 Figure 34. Photomicrograph and the scanning electron micrograph of framboidal and euhedral pyrites found in t he Avon Park Formation (255 m to 257 m). Note: Bar scale in lower right is 10 m. 1 mm

PAGE 125

109 Table 12. Maximum, minimum, mean and standard deviation of As concentrations in mg/kg for the Suwannee Limestone, Ocala Limestone and Avon Park Formation Suwannee Limestone Ocala Limestone Avon Park Fm. Minimum 0.02 0.02 0.64 Maximum 16.15 14.74 14.19 Mean 2.96 1.49 2.87 Standard Deviation 4.18 2.93 2.98 (N) 104 67 35

PAGE 126

110 average concentration may be a meaningless measure of As availability if its location/asso ciation within the aquifer matrix was not known. For example, if As was bound by hydrous ferric oxide (ferrihydrite) then only reducing conditions would cause its release into the groundwater. Alternatively if As would be present as impurities in pyrite, o nly oxidizing conditions could cause its release. Since pyrite in these rocks wa s generally of very small size a statistical technique was applied to determine whether pyrite was present in appreciable amounts and if that pyrite had the potential to be a s ource of As. The only elements that showed an appreciable degree of correlation with As were Fe, S and Al. Furthermore those elements also showed a good correlation to each other indicating that they occur together in the aquifer matrix. Since there was no known mineral that contains both Al and S the correlation of the two can only be explained by the coexistence of two discrete minerals. The correlation of Fe and S can be attributed to the presence of pyrite and the correlation of Fe and As to the presenc e of glauconite, a Fe bearing clay mineral. All of those minerals were previously described in the Suwannee Limestone, Ocala Limestone and Avon Park Formation. This indicated that As was almost exclusively associated with pyrite and clay minerals in all th ree rock units which in turn allowed to compare their averages. After the bulk rock chemical and mineralogical analysis of all samples, the following sampling intervals were chosen for the bench scale column experiments based on a variable amount of pyrite clays and concentration of As: Suwannee Limestone ( 136 m to 139 m, and 171 m to 174 m), Ocala Limestone (195 m to 199 m), and the Avon Park Formation (255 m to 257 m; and 260 m to 261 m) (Table 13). Figure 35 demonstrated the distribution of bulk rock As concentration and the calculated amount of pyrite for each

PAGE 127

111 selected interval. Based on a strong linear correlation between S and Fe in those samples, it was conceivable that Fe and S was controlled largely by the presence of pyrite. The weight percent of pyrite can be determined based on molar ratio of Fe to S: Fe (mg/kg) = S (mg/kg) 55.85/ (32.07 *2) FeS 2 (wt %) = (Fe (mg/kg) + S (mg/kg))/1000*(100/1000) All S was assumed to be associated with pyrite. At the same time, the amount of pyrite in the Avon P ark samples was calculated from the bulk Fe data due to the presence of gypsum (Table 13). The distribution of bulk rock As concentration demonstrated a strong correlation with the calculated amount of pyrite (R 2 = 0.92) except the 260 m 261 m interval f rom the Avon Park Formation (Figure 36). This result can be explained by the overestimation of pyrite constituent or existence of additional potential source of As such as dark brown clays (Table 12) (Lazareva and Pichler, 2007) 3.3.2. Leaching experimen ts All data from the bench scale leaching experiments, including the composition of original waters and recovered leachates were presented in Appendix F. 3.3.2.1. Suwannee Limestone (136 m to 139 m) The lithological composition of the Suwannee Limestone fr om the 136 m to 139 m interval consisted of limestone with minor percentage of quartz, pyrite, clays, and phosphate grains (Table 13). During the leaching experiments, drinking water (DW), wetland water (WW), untreated and air saturated groundwater (GW), a s well as untreated (BW) and UV treated filter basin waters (BW UV) were injected into 0.3 m columns

PAGE 128

112 Table 13. List of rock samples selected for the leaching experiments. Note: ls limestone; ds dolostone; p pyrite; c clay; Q qu artz; gy gypsum; ph phosphate; l.b light brown; d.b dark brown; Sample(1), (2) interval (bulk rock) analysis; Sample(a), (b) sample of special interest

PAGE 129

113 Figure 35. The calculated amount of pyrite and measured bulk rock As concentration in the Suwannee and Ocala Limestones, and the Avon Park Formation.

PAGE 130

114 Figure 36. Correlation between calculated amount of pyrite and measured bulk rock As concentration for the analyzed rock intervals with (top) and without the outlier (bottom). Note: the outlier (in circle) data from the Avon Formation (260 m to 261 m).

PAGE 131

115 (Appendix F, Table 14, Figure 37). The leaching columns contained from 113 g to 150 g of limestone with an average porosity of 24.5 %. The study showed that the injection of BW UV and BW caused the highest release of As from the aquifer matrix with concentrations of up to 6.1 g/L and 4.6 g/L, respectively. The UV treatment changed the redox potential of the BW from reducing to oxidizing indicated by the increase of dissolved oxygen (DO) positive oxidation reduction potential (ORP), and reduction of Fe(II), which in turn facilitated the oxidation of pyrite (Table 14). In total, about 32 g of As were recovered from a column using the BW and 24 g of As from the injection of BW UV (Figu re 37). The WW was injected for 2 days and d uring the first 4 hours of experiment, the concentration of As was up to 4.6 g/L. The pH values changed from 6.8 to 7.4 and the ORP varied from 143 mV to 38 mV. Generally, the ORP level was negative during the first day of the study confirming the reducing conditions of the system. At the end of first day the leaching column was closed overnight and the experiment was restarted next morning. On the second day of experiment, the concentration of As was up to 5. 3 g/L and the ORP was up to 108 mV. Overall, 60 g of As were r ecovered using the WW. The injection of native GW from the Floridan Aquifer caused minor to no release of As from the limestone matrix. The concentration of As in first leachate was 1.5 g/L a nd decreased to < 0.4 g/L thereafter. After the injection, pH and OPR values in first leachate decreased from 7 to 6.8 and from 120 mV to 156 mV, respectively. The first peak of As could be due to a possible micro oxidation of pyrite surface and formation of a fer rous sulfate phase, such as szomolnokite (FeSO 4 2 O) prior to the experiment

PAGE 132

116 (Costagliola et al., 1997). That phase containing As would dissolve during the injection of reducing Table 14. Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Suwannee Limestone ( 136 m to 139 m ) using different types of water.

PAGE 133

117 Figure 37. As, SO 4 pH, and ORP in leachates recovered from the Suwannee Limestone (136 m to 139 m) using 0.3 m columns. Note: DW: drinking water; WW: wetland water; GW: groundwater; GW w/air: groundwater saturated with air; BW and BW UV: untreated and UV treated filter basin water.

PAGE 134

118 groundwater, resulting in a possible As mobilization. Generally during the experime nt, the ORP was negative, confirming the reducing conditions of GW preserved in the amber carboy. The introduction of the same GW treated with air had a significant effect on the oxidation of pyrite with As values up to 3 times higher than the untreated GW clearly indicating the important role of DO. Overall, 27 g of As were recovered in leachates after the injection of aerated GW. The introduction of DW caused about 2 times less leaching of As (11 g) than the GW saturated with air, which could be due to 5.5 times lower concentration of DO detected in the DW. In a summary, the concentration of As in leachates arranged from highest to lowest was arranged as: BW UV > BW > GW/air > WW >DW > GW (171 m to 174 m) The lithological composition of the Suwannee Li mestone from the 171 m to 174 m interval consisted of limestone and dolostone mixed with minor quartz, pyrite, clays, and phosphate grains (Table 13). During the leaching experiments, drinking (DW), wetland (WW), and groundwater (GW) were injected into 1 m columns (Appendix F, Table 15, Figure 38). These experiments were performed to compare the leaching of As from a different interval of the same stratigraphic unit using longer column. The columns contained from 440.4 g to 441.0 g of sediment with an avera ge porosity of 22.6 %. The study confirmed that the injection of WW caused the highest release of As with concentrations of up to 19.4 g/L exceeding the As drinking water standard (DWS) of 10 39). During the first 8 hours of experiment, the concentration of As was increasing with time to 14.8 g/L. The pH level changed from

PAGE 135

119 Table 15. Initial water (before injection) and leachate recovered from 1 m columns filled w ith the Suwannee Limestone ( 171 to 174 m ) using different types of water.

PAGE 136

120 Figure 38. As, SO 4 pH, and ORP in leachates recovered from the Suwannee Limestone (171 m to 176 m) using 1 m columns. Not e: DW: drinking water; WW: wetland water; GW: groundwater; DWS: Drinking Water Standard..

PAGE 137

121 Figure 39. Distribution of Ca versus pH (top) and As versus pH (bottom) in leachates recovered from the Suwannee Limestone (171 m to 176 m) usi ng 1 m columns. Note: DW: drinking water; WW: wetland water; GW: groundwater.

PAGE 138

122 6.8 to 8.1. The ORP in first leachate increased from 143 V to 96 mV and decreased back to 106 mV, confirming the reducing conditions of the WW preserved in the amber carboy. At the end of first day the leaching column was closed overnight and the experiment was restarted next morning. On the second day of experiment, the first and second leachates had higher pH (up to 8.7), positive ORP and the As values < 4g/L, possibly indic ating As sorption to neo formed HFO and co deposition in limestone matrix. S ubsequently, pH level dropped to 7.6 and As concentration in leachates increased to up to 20 g/L. Previous studies showed that arsenite exhibit s a strong affinity and great sorpti on capacity for ferrihydrite and goethite (Manning et al., 1998). This process is pH dependent with an adsorption envelope at pH 8 to 9 (Manning et al., 1998) (Figure 39). In total, about 186 g of As were recovered using the WW injectant. The injection o f DW demonstrated the leaching of As from the column with concentrations of up to 8.1 g/L. The ORP in first leachate declined from 223 mV to 107 mV due to consumption of oxygen for pyrite oxidation. Overall, about 63 g of As were recovered from a column using DW. Native GW from the Floridan Aquifer was injected into a column for 2 days. The experiment was not interrupted overnight; however the injection rate was changed from 2 to 0.2 mL/min. Groundwater did not cause the oxidation of pyrite and leaching of As. Moreover, the concentration of As in leachates was 2.5 times less than the initial water (< 0.4 g/L). After the injection, pH values in leachates increased from 7.2 to 11.5 due to the presence of dolomite ( Figure 39 ). Generally, the ORP was negativ e, confirming the reducing conditions of GW preserved in the amber carboy.

PAGE 139

123 In a summary, the concentration of As in leachates arranged from highest to lowest was arranged as: WW > DW > GW 3.3.2.2. Ocala Limestone (195 m to 199 m) The lithological composi tion of the Ocala Limestone from the 195 m to 199 m interval consisted of a white dolomitized fossiliferous limestone, poorly to moderately indurated with minor quartz, and pyrite (Table 13). During the leaching experiments, drinking water (DW), wetland wa ter (WW), groundwater (GW), untreated (BW) and UV treated filter basin waters (BW UV) were injected into 0.3 m columns (Appendix F, Table 16, Figure 40). The columns contained from 118.5 to 128.7 g of sediment with an average porosity of 25.9 %. The study demonstrated that the injection of BW and BW UV caused the highest release of As with concentrations of up to 35.6 g/L and 23.6 g/L, respectively. In total, about 163 g of As were recovered from the BW and 128 g of As from the BW UV (Figure 40). The leaching of As from the injection of WW was up to 23.6 g/L behaving similarly to the BW UV injectant. Even though the oxidation reduction potential of the system remained under reducing conditions, the mobilization of As from the aquifer matrix was higher than from the DW. Overall, 134 g of As was leached during the injection of WW. The concentration of As in leachates from the injection of DW was up to 17.1 g/L. Overall, about 50 g of As were recovered using the DW The injection of GW from the Florida n Aquifer showed that As was up to 2.8 g/L in the first leachate and dropped to < 0.1 g/L at the end of experiment. Similarly to

PAGE 140

124 Table 16. Initial water (before injection) and leachate quality recovered from 0.3 m columns filled with the Ocala Limestone ( 195 to 199 m ) using different types of water.

PAGE 141

125 Figure 40. As, SO 4 pH, and ORP in leachates recovered from t he Ocala Limestone (195 to 199 m) using 0.3 m columns. Note: DW: drinking water; WW: wetl and water; GW: groundwater; BW and BW UV: untreated and UV treated filter basin waters

PAGE 142

126 the column filled with the Suwannee Limestone (136 m to 139 m), t he first peak of As could be due to a possible micro oxidation of pyrite surface and formation of a ferrous sulfa te phase, such as szomolnokite (FeSO 4 2 O) prior to the experiment (Costagliola et al., 1997) resulting in a potential release of As. The pH values in all leachates ranged from 6.9 to 7.5. Generally during the experiment, the ORP was negative, confirming t he reducing conditions of GW preserved in the amber carboy. In a summary, the concentration of As in leachates arranged from highest to lowest was arranged as: BW > BW UV = WW > DW > GW 3.3.2.3. Avon Park Formation (255 m to 257 m) The lithological compo sition of the Avon Park Formation from the 255 m to 257 m interval consisted of limestone sucrosic dolostone matrix with minor mineral phases, such as pyrite, gypsum and quartz (Table 13). During the leaching experiments, drinking water (DW), untreated ( WW) and UV treated wetland water (WW UV), untreated (GW) and aerated groundwater (GW/air), untreated (BW) and UV treated filter basin waters (BW UV) were injected into 0.3 m columns (Appendix F, Table 17, Figure 41). The columns contained from 157.3 g to 1 86.4 g of sediment with average porosity of 16.8 %. Similarly to the Ocala and Suwannee Limestones (136 m to 139 m interval), the injection of BW and BW UV caused the highest release of As with concentrations of up to 65.5 g/L and 56.2 g/L, respectively (Figure 41). Those values were 6.5 and 5.6 times higher than the As maximum contaminant level for DW (EPA, 2009). In total, about 246 g and 174 g of As were recovered using the BW and BW UV, respectively.

PAGE 143

127 The injection of WW caused the mobilization of A s to up to 47.1 g/L. At the same time, the introduction of WW UV resulted in As leaching to up to 35.3 g/L. Overall, a bout 198 g and 131 g of As were leached out from the columns during the injection of WW and WW UV. The pH values in all leachates rang ed from 6.4 to 7.5. During both experiments, the ORP of the system changed from reducing to oxidizing. Figures 42 44 demonstrated the effect of the UV radiation on the concentration of As, DO, ORP, dissolved organic carbon (DOC), total nitrogen (N) and t otal/fecal coliform. Generally, total and fecal coliform were present in the WW and BW. After UV treatment, however, fecal coliform was absent, but total coliform was 1 ct/100 mL in both types of water. The DO and ORP significantly increased after UV treat ment due to formation of oxygen radicals and other by pr oducts such as trace ozone ( Masschelein, 2002 ) changing the redox state of water from reducing to oxidizing As a result, the level of As in leachate increased after UV treatment. The concentration of DOC did not change. In addition, the injection of DW into two columns of 0.3 m and 0.5 m was performed for a period of 4 months at a 3 week sampling interval. The water was injected in the columns on July 18, 2006 for about 5 hours. Analyses showed that As in first leachates was up to 26.2 g/L (DW 0.3) and 35.3 g/L (DW 0.5), respectively (Figure 45). Distribution of As in both columns revealed a similar pattern. The concentration of SO 4 increased dramatically from 58 mg/L to 679 mg/L (DW 0.3) and 1051 m g/L (DW 0.5), respectively. The pH levels increased from 6.9 to 7.7 (DW 0.3) and 7.9 (DW 0.5), while ORP declined from 223 mV to 175 mV (DW 0.3) and 148 mV (DW 0.5) due to consumption of oxygen for pyrite oxidation. At the end of the day, the experiments w ere stopped and the columns saturated with the DW were sealed. Subsequently, the tests were

PAGE 144

128 Table 17 Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Avon Park Formation ( 255 m to 257 m ) using different types of water.

PAGE 145

129 Figure 41. As, SO 4 pH, and ORP in leachates recovered from the Avon Park Formation (255 m to 257 m) using 0.3 m columns. Note: DW: drinking water; WW and WW UV: wetland water untreated and treated with UV; GW and GW/air: groundwater untreated and treated with air; BW and BW UV: filter basin water untreated and treated with UV.

PAGE 146

130 Figure 42. Dissolved organic carbon (DOC) and total nitrogen (TN) in the groundwater (GW), we tland (WW) and filter basin water (BW) before and after UV. Note: GW* data for the Upper Floridan Aquifer ROMP 45 from Sacks and Tihansky (1996 ).

PAGE 147

131 Figure 43. (A) Dissolved oxygen (DO) and oxidation reduction potential (ORP); and (B) Fecal and total coliform of the wetland (WW) and filter basin water (BW) before and after UV. Note: diagram (B) clearly represents a great value of the filter basin treatment. (A) (B)

PAGE 148

132 Figure 44. Arsenic (As) in the initial water and first leachate recovered from the A von Park Formation (255 m to 257 m) using wetland (WW) and filter basin water (BW) before and after UV, and drinking water (DW).

PAGE 149

133 Figure 45. Arsenic, pH and ORP in leachates recovered from the Avon Park Formation (255 m to 257 m) usin g drinking water (0.3 m and 0.5 m columns) for a period of four months at three weeks interval.

PAGE 150

134 restarted on July 21, August 9 and 29, September 19, October 3 and 17, November 2 and 20. Generally, the leachates taken at the beginning of each experiment ha d the highest peaks of As and lowest ORP due to the extended time of water rock interaction in the columns. After 4 months of experiments, the concentration of As in both columns decreased to < 10 g/L. In total, about 750 g of As was recovered from a 0.3 column and 815 g of As from a 0.5 m column. Therefore, based on the calculated molar ratio As/FeS 2 the amount of FeS 2 dissolved from 0.3 and 0.5 m columns was 121.4 mg/kg and 131.8 mg/kg, respectively. These values are about 6 times less than total am ount of pyrite present in the columns (Table 13) The significant difference could be due to As co precipitation with HFO formed during the water rock interaction, which were visually observed during the experiments. In addition to the experiments above, t he 4 day injection sequence of the BW and DW was performed i n a 0.3 m column. This procedure was necessary to evaluate and compare the concentration of total As and arsenite/arsenate ( As ( III)/ As ( V)) species from the same column by two types of water. Du ring the 1 day injection of BW, the concentration of As in varied from 59.3 g/L to 23.8 g/L (Figure 46). The concentration of SO 4 increased from 44 mg/L to 757 mg/L and subsequently dropped to 80 mg/L. Similar distribution of SO 4 and As confirmed the dis solution of pyrite. During the 2 day injection of DW, the concentration of As changed from 66.8 g/L to 13.8 g/L. During the 3 day injection of BW, As values ranged from 35.4 g/L to 16.7 g/L. During the 4 day injection of DW, the concentration of As cha nged from 28.9 g/L to 7.7 g/L. The speciation of As showed that it was predominantly present as As(V) which is common under oxidizing conditions. This experiment clearly showed that As concentration in

PAGE 151

135 Figure 46. Distribution of SO 4 As specie s and As total concentrations in leachates recovered from the Avon Park Formation (255 m to 257 m) from the consecutive injection of filter basin (BW) and drinking (DW) waters into a 0.3 m column.

PAGE 152

136 leachates from BW was generally higher than from DW even th ough it was initially more reducing. In total, about 593 g of As were recovered from the column. The injection of GW from the Floridan Aquifer demonstrated that As was up to 4.3 g/L in the first leachate and dropped to 0.1 g/L at the end of experiment ( Figure 41). Similarly to the experiments above, t he first peak could be due to a possible micro oxidation of pyrite surface to a ferrous sulfate phase before the experiment resulting in release of As. The introduction of the same GW saturated with air had a tremendou s effect on the oxidation of pyrite with As values 10 times higher than the untreated water clearly indication the important role of DO. In a summary, the concentration of As in leachates arranged from highest to lowest was arranged as: BW > BW UV > WW > G W air > WW UV > DW > GW (260 m to 261 m) The lithological composition of the Avon Park Formation from the 260 m to 261 m interval was dolostone dominated with a presence of limestone, minor framboidal pyrite, and gypsum (Table 13). During the leaching ex periments, drinking (DW), wetland (WW), filter basin (BW), and groundwaters (GW) were injected into 0.3 m columns (Appendix F, Table 18, Figure 47). The columns contained from 146.6 g to 183.1 g of sediment with average porosity of 18.2 %. Similarly to the Suwannee Limestone (interval 171 m to 174 m), the highest concentration of As was detected after the injection of WW ranging between 28.8 g/L to 66.9 g/L after about 8 hours of experiment The leaching of As from the column was progressive throughout t he experiment (Figure 47 48). The pH changed from 6.6 to 7.5. The OPR ranged between 214 mV to 90 mV confirming

PAGE 153

137 the reducing conditions of the system. Overall, 452 g of As were recovered from the WW injectant. To evaluate the potential difference in m obilization of As, WW was filtered in situ through a 0.2 m membrane filter in order to eliminate particulate matter and reduce or remove present microorganisms. The experiment showed that i n contrast to the unfiltered WW, filtration resulted in a reductio n of As from 30.0 g/L to 16.4 g/L (Figure 47). These results clearly demonstrated the important role of microorganisms on the stability and surface reactivity of pyrite. Overall, about 91 g of As were recovered from this experiment. The injection of DW caused the mobilization of As of up to 9.0 g/L, which was 7.5 9 times lower than from the WW. The concentration of SO 4 in le achates from DW increased to 1199 mg/L. The ORP in first leachates declined from 320 mV to 1 mV likely due to the oxidation of p yrite and formation of HFO which were visually observed during the experiments. Overall, 55 g of As were recovered from th e DW The leaching of As from the BW showed a comparable behavior to the injection of DW. Generally, the concentration of As in leach ates was < 8 g/L, while the concentration of SO 4 was up to 1553 mg/L. However, after about 6 hours of the experiment the concentration of As reached up to 29.1 g/L. This could be due to trapping of oxygen during the water pumping and the following oxidat ion of water in carboys. In total, about 124 g of As were leached during the injection of BW. In a contrast to the WW, the injection of DW, BW, and GW caused the pH levels to increase from 6.5 to 10.8 due to the dissolution of dolomite (Figure 48). As ex pected, the groundwater fr om the Floridan Aquifer caused minor to no mobilization of As from the aquifer matrix. The concentration of As in first leachate was

PAGE 154

138 Table 18. Initial water (before injection) and leachate recovered from 0.3 m columns filled with the Avon Park Formation ( 260 m to 261 m ) using different types of water.

PAGE 155

139 Figure 47. As, SO 4 pH, and ORP in leachates recovered from the Avon Park Formation (260 m to 261 m) using 0.3 m columns. Note : DW: drinking water; WW: wetland water; WW_filt_0.2um: wetland water filtered in situ through 0.2 um membrane; GW: groundwater; BW: filter basin water.

PAGE 156

140 Figure 48. Distribution of Ca versus pH (top) and As versus pH (bottom) in leacha tes recovered from the Avon Park Formation (260 m to 261 m) using 0.3 m columns. Note: DW: drinking water; WW: wetland water; GW: groundwater; BW: filter basin water; Original water composition and first recovered leachates were highlighted in red and blac k; respectively.

PAGE 157

141 1.4 g/L and dropped to < 0.5 g/L within 5 hours of the experiment. In a summary, the concentration of As in leachates arranged from highest to lowest was arranged as: WW > WW_filt_0.2um > DW > BW > GW 3.4. Discussion A multitude of sc ientific publications reported that the oxidative dissolution of pyrite was a principal mechanism responsible for the release of arsenic (As) during aquifer storage and recovery (ASR) (Arthur et al., 2007; Arthur et al., 2005; Pichler et al., 2004; Arthur et al., 2002; Williams et al., 2002; Arthur et al., 2001; de Ruiter and Stuyfzand, 1998). The bench scale leaching tests of this study were intended to simulate the ASR process in a laboratory and provide insights important to the future of ASR in Florida and potentially worldwide. The experiments showed that the amount of As released from the Upper Floridan Aquifer (UFA) was not perfectly correlated with the bulk rock As concentration (Figures 35 and 49). The highest level of As was leached out from the Av on Park Formation and the lowest from the Suwannee Limestone, although the Ocala Limestone had the lowest bulk rock As. This conclusion corresponded to the work by Arthur et al. (2007) who performed bench scale leaching experiments and the sequential ext raction of As, Mo, and Sb from the lithostr atigrafic units of the UFA. This study demonstrated that the highest fraction of As was extracted from the Hawthorn Group, following by the Suwannee Limestone, Avon Park Formation, and the Ocala Limestone. Most of the As in all units was associated with sulfides (i.e., pyrites), although, organic matter, hydrous ferric oxides (HFO), and carbonates could also contribute to the bulk rock As (Arthur et al., 2007). The trend of As leaching from the

PAGE 158

142 Figure 49. Dis tribution of maximum As (top) and SO 4 (bottom) concentrations in leachates recovered from each interval (with the subtraction of initial concentration) Note: GW: groundwater; DW: drinking water; WW: wetland water: BW: basin water

PAGE 159

143 aquifer matrix could be g overned by combination of several factors, such as the porosity and permeability of the aquifer influencing the rate and degree of free water saturation, amount of pyrite to be exposed to the preferential flow paths during water rock interactions, surface reactivity of pyrite, concentration of As and selective As leaching from individual pyrite crystals. The additional factors affecting the mobility of As during ASR may include the chemistry of input and native groundwater, aquifer matrix chemistry/mineralo gy, site hydrogeology, injection water rock contact time, and the amount of cycle tests (Arthur et al., 2005). The Eh pH diagram of the mineral stability in the Fe S O system shows that pyrite is not stable under oxidizing conditions whereas Fe(OH) 3 is not stable under reducing subsurface environment (Evangelou, 1995) (Figure 1). Generally, the oxidation of pyrite by O 2 acts as a source of acid, sulfate, iron and arsenic, and can be illustrated by three steps ( Evangelou, 1995 ): (1) FeS 2 + 7/2O 2 + H 2 2+ +2SO 4 2 +2H + Fe 2+ can be further oxidized to Fe 3+ which hydrolyzes into HFO (displayed as Fe(OH) 3 ) to discharge extra amount of acid into the environment: (2) Fe 2+ + 1/4O 2 +H + Fe 3+ + 1/2H 2 O and (3) Fe 3+ + 3H 2 3 + 3H + Consequently, As can be re sorbed onto HFO if the conditions remain sufficiently oxygenated to promote HFO stability, but released back to native groundwater under reducing conditions. As demonstrated above, the dissolution of pyrite in the limestone matrix is influenced by the oxi dants of the reacting water. Native Floridan groundwater from the

PAGE 160

144 USF well was used as a baseline reference for comparison with other types of water and caused no to minor As mobilization during all leaching experiments. Generally, the concentration of As was slightly elevated in the first pore volume leachates and subsequently declined to < 1 g/L. This could be due to micro oxidation of pyrite surface and the possible formation of a thin film ferrous sulfate before the experiment (Pratesi and Cipriani, 2000) Cost agliola et al. (1997) investigated the deterioration of pyrite surfaces based on specimens from the Mineralogical Museum of the University of Florence. The authors determined that the alteration mineral present along the pyrite surface fractures and crysta l borders was mainly in a form of szomolnokite (FeSO 4 2 O). Therefore, the injection of reducing groundwater could facilitate a minor release of As a ssociated with a ferrous sulfate phase At the same time, several studies s howed that the expos ure of pyrite /arsenopyrite surfaces to the atmosphere d eveloped m icro oxidation products such as i ron oxy hydroxide with minor poly sul f ide and thiosul f ate pr oducts ( Eggleston et al. 1996 ; Nesbitt et al., 1995 ; Nes bitt and Muir 1 994 ) SEM analysis of the samples selected for the leaching experiments s howed that pyrite surface s w ere not oxidized ( Figures 32 34). T he bench scale leaching tests performed by Arthur et al. (2007) on more than 40 rock samples from the UFA revealed that the injection of low dissolved oxygen (DO) native groundwater caused the decrease of ORP and mobilization of As (up to 34 g/L ), Sb, and Mo (Figure 50). During the in jection of low DO surface waters (SW), the concentration of As in leachates reached up to 25 g/L. However, the introduction of high DO SW facilitated a sharp decrease of As in leachates, possibly due to As sorption and co precipitation with a solid or col loidal phase of HFO (Arthur et al., 2007; Lu et al., 2005). Tampa drinking water, which chemically and physically resembles

PAGE 161

145 Figure 50. Distribution of ORP and As in leachates recovered during a different phases of the bench scale leaching experiments (Modified from Arthur et al., 2007). Note: LDO NGW: low dissolved oxygen native groundwater from cored interval; LDO SW and HDO SW: low and high dissolved oxygen treated surface water from nearby canal; 1 and 2: Hawthorn Group core samples (166 m and 194 m bls, respectively); 3: Suwannee Limestone (262 m bls).

PAGE 162

146 the injection water used in ASR operation, caused the leaching of As of up to 27 g/L which was higher than the current As drinking water standard of 10 g/L (EPA, 2009) (Figure 49) One of the most important objectives of this study was to consider using the wetland water to recharge the UFA as the ASR water of choice However, the leaching experiments using the filter basin and wetla nd waters demonstrated the highest degree of As mobilization (up t o 68 g/L) from the aquifer matrices (Figure 49). This was unexpected because those types of water were less oxygenated than Tampa drinking water and thus should be less aggressive. Also, the leaching of As using the wetland water was noticeably correlated with depth (R 2 = 0.99) (Figure 49). As expected, the highest concentration of SO 4 was observed in leachates recovered from the Avon Park Formation due to occurrence of gypsum. It was interesting to notice that the injection of Floridan groundwater saturat ed with air (high DO) caused as much As release as the filter basin and drinking waters. The breakthrough curves for As and SO 4 were compatible for the majority of leaching tests. The distribution of As and SO 4 in the leachates versus time demonstrated ini tially high levels followed by a considerably rapid de cline before reaching relative ly steady state conditions of mineral di ssolution after about 3 hours. Walker et al. (2006) reported that the sudden reduction of arsenopyrite oxidation within the first 15 hours of experiment was governed by limited surface oxidation and preferential reactions on fractured mineral surfaces (Borda et al., 2004). Taking into consideration a molar ratio of Fe to S associated with pyrite (1:2) the concentration of Fe in leacha tes was substantially lower or even undetectable compared to S (as sulfate) demonstrating that most of Fe was attenuated within a column and aquifer matrix (Appendix F). Most likely, the released Fe consequently oxidized to HFO and

PAGE 163

147 precipitated either on t he column walls or on the pyrite surface (Walker et al., 2006; Nesbitt, 1995; Richardson and Vaughan, 1989 ). Therefore, due to high sorption capacity the released As could be strongly attracted and subsequently adsorbed onto neo formed HFO and co deposited in the column (Arthur et al., 2007; Lu et al., 2005; Hongshao and Stanforth, 2001; Nickson et al., 2000; Pichler et al., 1999; Hinkle and Polette, 1999; Manning et al., 1998; Evangelou, 1995; Bowell, 1994; Chao and Theobald, 1976 ). The ultraviolet (UV) h ad different effects on the leaching of As depending on the aquifer matrix. Originally, the UV treatment system was installed in the field and laboratory to reliably remove microorganisms present in the filter basin and wetland waters and to understand its influence on the ORP of water potentially causing the dissolution of pyrite and mobilization of As. It has been shown that UV can reduce the amount of odor causing agents, which primarily targets and destroys compounds such as ammonia and hydrogen sulfide creating an oxic environment. At the same time, t he UV effect may potentially create hydroxyl radicals and other by products such as trace ozone, and change the redox state of water ( Masschelein, 2002 ). Hydrogen peroxide ( H 2 O 2 ) in combination with UV ligh t decomposes to hydroxyl radical ( ) through a Haber Weiss mechanism as 2 + H 2 O 2 + O 2 ( Druschel et al., 2004; Pettigara et al., 2002; Watts et al., 1999 ; Stumm and Morgan, 1996). According to Masschelein (2002), resulted from UV radiation have both reducing and oxidizing properties. The reducing properties are due to t he following dissociation: OH = O + H + which are attributed to the oxygen mono ion. In addition, the reducing properties of OH can establish reactions in oxidation of, for example, ferrous iron (Fe(II)): Fe(II) + OH = (Fe(III) OH ). The author emphas izes that there are no precise standard methods for the

PAGE 164

148 detection and determination of OH radicals under conditions applicable to the processes of drinking water treatment, concluding that the lifetime of hydroxyl radical s is in a range of nanoseconds. Pr evious study showed that H 2 O 2 and increase the rate of pyrite oxidation (McKibben and Barnes, 1986). It was reported that s ulfite can be rapidly oxidize d i n the presence of H 2 O 2 and OH with rate constants on the order of 10 5 M 1 sec 1 and 5x 10 9 L m ol 1 sec 1 respectively (Ermakov et al., 1997; Huie and Neta, 1987). The current study showed that UV treatment significantly reduced both fecal and total coliform bacteria, but facilitated the increase of dissolved oxygen in initial waters, change in ORP from the reducing to oxidizing conditions, and subsequently, the increase of As concentration in leachates. However, the experiment with the Avon Park Formation (interval 255 m to 257 m) showed the opposite results, where As levels from the injection of t he filter basin and wetland waters treated with UV were less than those from untreated waters (Figure 44). 3.4.1. Importance of dissolved organic carbon for arsenic mobilization The concentration of dissolved organic carbon (DOC) in fresh waters usually r anges from 0.1 to 20 mg/L and reaches higher values in wetlands (Bauer and Blodau, 2006; Volk et al., 2002). The reported DOC in the Everglades varied between 23 and 136 mg/L (Aiken, 2002). The DOC in the wetland and filter b asin waters of the study area w ere up to 18.5 and 15.3 mg/L, respectiv ely. Those values were about 1 0 times higher than the Upper Floridan groundwater (from Sacks and Tihansky, 1996 ) The elevated DOC in the wetland is inferred to be the result of excretion and decomposition of organism s including fish, reptiles, amphibians, insects, birds, bacteria as well as vascular

PAGE 165

149 plants and algae (Volk et al., 2002). The concentration of total nitrogen (TN) in the filter basin and wetland waters was up to 1.5 and 1.6 mg/L, respectively. In a contra st, the TN in the Upper Floridan groundwater, reported by Sacks and Tihansky (1996 ), was < 0.002 mg/L (Figure 42). The elevated TN in the wetland was governed by dissolved/particulate inorganic and organic nitrogen coming from phytoplankton, bacteria and a quatic life. Several studies demonstrated that the DOC is very important in the mobilization, redox transformation and solubility of As in aquatic systems through the formation of strong complexes (Buschmann et al., 2006; Buschmann et al., 2005; Warwick e t al., 2005; Saada et al., 2003; Redman et al., 2002; Kaiser et al., 1997 ). The DOC may induce As methylation and become a source of energy for Fe SO 4 2 and humic reducing microorganisms resulting in reduction of Fe oxides and metal sulfide precipitation (Tufano and Fendorf, 2008; Huang and Matzner, 2006 ; Koretsky et al., 2003; Roden et al., 2000 ). Therefore, the elevated DOC in the wetland and filter basin waters could greatly influence on the release of As from the Floridan Aquifer matrix Bauer and Blod au (2006) reported that DOC could influence desorption and redox transformations of As bound to soils, sediments, and Fe oxides with the mobilization to up to 53.3 % of the total sorbed As The authors emphasized that the principal mechanism for As mobiliz ation from the solid phases was the competition for sorption sites between organic anions and As. The cycling of DOC and Fe in Bangladesh Aquifers was associated with elevated As concentrations (Mladenov et al., 2010; Anawar et al., 2003; Harvey and Swartz 2002). In addition, Aguilar et al. (2007) demonstrated that the highest concentration of As extracted from soils was leached by oxalic oxalate to up to 24 36 % of the total As. Lalvani et al. (1996) suggested that the dissolution of pyrite was enhanced

PAGE 166

150 by the oxalic acid abundantly present in many plants. Leaching experiments with the Avon Park Formation aquifer matrix (interval 260 m to 261 m) evidently demonstrated the crucial role of microorganisms on the stability and surface reactivity of pyrite. T he experiment showed that i n contrast to the unfiltered wetland water, in situ filtration through a 0.2 m membrane filter resulted in a 2 fold reduction of As leaching ranging from 16 g/L to 30 g/L. 3.5. Conclusions The bench scale leaching experiments to investigate the mobilization of geogenic arsenic (As) in the Floridan Aquifer under a range of redox conditions showed that: 1) The amount of As released from the aquifer matrix was not perfectly correlated with the bulk rock As concentration. The highest level of As was leached out from the Avon Park Formation and the lowest from the Suwannee Limestone, although the Ocala Limestone had the lowest bulk rock As. 2) Little to n o As release using native Floridan groundwater. 3) Tampa drinking water, which chemic ally and physically resembles the ASR injection water, caused the leaching of As of up to 27 g/L which was higher than the current As drinking water standard. 4) Wetland and filter basin waters showed the highest release of As (up to 68 g/L), which was unex pected because the wetland water was much less oxygenated than Tampa drinking water and thus should be less aggressive.

PAGE 167

151 5) UV treatment significantly reduced both fecal and total coliform bacteria, but facilitated change in oxidation reduction potential (ORP ) from the reducing to oxidizing conditions, and subsequently, the increase of As concentration in leachates. 6) Dissolved organic carbon (DOC) in the wetla nd and filter basin was 1 0 times higher than the native Floridan groundwater which was likely very impo rtant in the mobilization, redox transformation and solubility of As in aquatic systems; 7) The leaching of As from the aquifer matrix could be governed by combination of factors, such as the porosity and permeability of the aquifer influencing the rate and d egree of free water saturation, amount of pyrite to be exposed to the preferential flow paths during water rock interactions, limited surface reactivity of pyrite, favored reactions on fractured mineral surfaces, concentration of As and selective As leachi ng from individual pyrite crystals. Overall, the results demonstrated above could be very important to forecast As behavior during anthropogenic physico chemical changes such as the aquifer storage and recovery procedure. The bench scale leaching experime nts showed that perturbations of native aquifer conditions caused the release of As from the Floridan Aquifer matrix, although the reaction may not be as simple as the dissolution of pyrite by oxygen, but additionally governed by a complex set of factors i ncluding the oxidation reduction potential of the system, SO 4 2 /S 2 Fe 3+ /Fe 2+ dissolved organic carbon and microbial activity. In addition, the oxidative dissolution of pyrite caused coprecipitation of released As with neo formed hydrous ferric oxides wh ich could become an additional source of As for the native groundwater under reducing conditions.

PAGE 168

152 CHAPTER FOUR GEOCHEMICAL REACTIVE TRANSPORT MODELING 4.1. Introduction Aquifer storage and recovery (ASR) is the storage of treated surplus surface water in a confined aquifer during rainy seasons followed by its recovery during times of need (Arthur et al., 2001). The ASR procedure was widely supported around the world to meet increasing water demands and to provide more sustainable alternative to the ext ensive groundwater consumption (Alley et al., 1999). However, an impediment to ASR development in Florida, Australia, Denmark, and the Netherlands was due to the mobilization of arsenic (As) from the aquifer matrix (Jones and Pichler, 2007; Vanderzalm et a l., 2007; Mirecki, 2006; Arthur et al., 2005; Mikecki 2004; Stuyfzand and Timmer, 1999; Stuyfzand, 1998) Particularly, the injection of oxygenated surface water into reducing native Floridan groundwater caused a transformation of the redox environment, ox idative dissolution of pyrite, and the release of As with values in recovered water of up to 130 g/L (Arthur et al., 2005; Williams et al., 2002). For the past several years, increasing efforts were devoted to gain a better understanding of As behavior in the subsurface environment implementing numerical modeling and 1D flow systems ( Postma et al., 2007; Stollenwerk et al., 2007; Moldovan and Hendry, 2005; Appelo and de Vet, 2003; Appelo et al., 2002; Dzombak and Morel,

PAGE 169

153 1990). In addition, Wallis et al. (2 009) developed conceptual and reactive transport m odels to evaluate the impact of chemical, physical, and biochemical factors on the mobilization of As under field scale conditions in South West Netherlands. The major objective of this investigation was i ncorporate t he results of the bench scale leaching experiments demonstrated in the third chapter into fully coupled reactive transport models using the Geochemist's Workbench software (Bethke, 2006; Bethke, 1998). The modeling of fluid rock interactions wa s important to characterize the geochemical processes in the columns, to understand and to verify if the model could predict the results of simulated injections. Particularly, geochemical modeling was essential to examine the aquifer redox conditions, the stability of pyrite, and the behavior of As during injections of treated ASR water into the Floridan Aquifer. 4.2. Methods The chemical reactions and chemical evolution of the liquid phase were modeled using the Geochemist's Workbench (GWB) Professional (version 6.0.5) including the React and X1t modules (Bethke, 2006a; Bethke, 2006b). The React program models the equilibrium states of aqueous species in a fluid, the state of fluid saturation with respect to minerals, as well as the fugacity of dissolved gases (Bethke, 2006a). It can consider a number of mineral interactions, kinetic constrains on r eactions, and mixing scenarios. At the same time, X1t allows constructing a 1D reactive transport model composed of a groundwater flow an d transport models coup led to the React chemical reaction model (Bethke, 2006b). The transport model takes in consideration the movement of groundwater and chemical species dissolved in it by hydrodynamic dispersion, molecular

PAGE 170

154 diffusion, and advection (Bethke, 2006). The model c an be constructed for open or closed environmental systems with a particular primary or secondary fluid and the initial reactant chemical composition (i.e., aquifer matrix). It can trace several types of reaction paths, such as the titration, flow through model, flush model, and picking up the results of a run (Bethke, 2006 a ). D uring the titration the program adds a small amount of one or more reactants (i.e., minerals) to the system and recalculates the equilibrium states over the path of the simu lation. At the same time, a flow through model is useful to trace the fluid reac tion as it migrates through an aquifer preventing the minerals from dissolution once they precipitate d In contrast to a previous model, during the flush model the fluid, added as a reactant, displace s or flushes existing fluid from the system. Finally, the React allows an entire syst em, fluid or the mineral fraction and implementing those results as a new equilibrium system for a new simulation (Beth ke, 2006a) This type of the model was especially important for the present study to simulate the water rock interaction between the native groundwater and aquifer matrix with the subsequent injection of surface water resembling the ASR scenarios. The geoc hemical modeling was based on the composition of drinking water from the City of Tampa and the bulk rock chemical analysis of the Avon Park Formation (255 m to 257 m) described in the third chapter (Appendix F and G). City of Tampa d rinking water was used as an analogue for ASR surface water (injectate), because ASR injectate has to be treated to meet all drinking water standards and requirements. Therefore, both the drinking water and ASR injectate were chemically and physically similar, i.e., pH, T, high concentration of dissolved oxygen (DO) and high oxidation reduction potential level (ORP). Due to the complexity of the carbonate system in the leaching columns, such as

PAGE 171

155 the progressive transformation of aquifer fluid and rock saturation state s t he geoche mical modeling was divided into several case studies Particularly, it was important to develop the end member models representing theoretical extremes of the column experiments in terms of an open or closed system i.e. the available DO w hic h wa s crucial to the stability of pyrite and the mobilization of As. Generally, t he simulation was set for 5 hours The volume of reacting fluid was 60 mL to reflect the conditions of leaching experiments described in the third chapter (Appendix G). During the geochemic al modeling, the the mineral precipitation during simulation. species formed during the simulation. Prior to the simul ations, thermodynamic data for arsenopyrite (AsFeS) in the thermo.dat database (Wolery et al., 1986 ) the redox couple HS /SO 4 2 and constrain Eh of the system This way the product of arsenopyrite dissolution wa s set to one As(OH) 4 specie s instead of the redox couple AsH 3 (aq)/As(OH) 4 The arsenopyrite was added to the model as a source of As based on As/FeS 2 molar ratio calculated from the bulk rock analysis (Table 12). The dissolution reaction of arsenopyrite was modified from (1) to (2): AsFeS + 2 H 2 O + 0.5 H = Fe ++ + 0.5 As(OH) 4 + 0.5 AsH 3 + HS (1) AsFeS + 2.5 H 2 O + 0.375 H + + 0.375 SO 4 -= Fe ++ + As(OH) 4 + 1.375 HS (2) In addition to the thermodynamic database t h ermo.dat, the FeOH.dat database, includ ed in the GWB package was implemented to account for As sorption to hydrous FeOH.dat

PAGE 172

156 included the surface complexation constants, surface site density, and the specific surface area of the HFO species (Dzombak and Morel, 1990). 4.3 Results and D iscussion T he geochemical modeling was divided into a set of case studies explained below. The groundwater, surface water, and rock compositio n for each scenario is given in Appendices F and G (1) Water rock interaction between the aquifer matrix and surface water (Closed System) The major hypothesis of this model was to describe the water rock interaction between the aquifer matrix and drinking water where the water was reacted directly wit h rock (i.e., reactant) using the React code. These conditions were similar to the bench scale leaching experiments. During this simulation, t he system was considered to be closed, i.e. the source of O 2 was limited. Based on the bulk rock chemical analysis the aquifer matrix was composed of 258 g of calcite (CaCO 3 ), 35 g of dolomite (CaMgCO 3 ), 10 g of gypsum (CaSO 4 ), 0.2 g of pyrite (FeS 2 ), and 0.002 g of arsenopyrite (AsFeS) The geochemical modeling showed that about 0.6 mg of pyrite and 0.006 mg of ars enopyrite were dissolved over 5 hours The pH of the system increased from 6.9 to 7.3 (Figure 51 B ). The redox changed from the oxidizing to reducing conditions by 4.5 hours indicated by Eh. It d ecreased from +800 mV to 150 mV (Figure 51A) The increasing concentration of Ca 2+ indicated the dissolution of limestone during water rock interactions. About 110 g/L of As was sorbed onto HFO via weak bonding (FeOHAsO 4 2 ) until 4.5 hours of the simulation but released back to solution (70 ug/L) as As(OH) 3 throug h the reductive dissolution of HFO (Figure 52A) At the same time, the

PAGE 173

157 Figure 51. Distribution of Eh (A) and pH (B) during simulated injections of model (1). (A ) (B )

PAGE 174

158 Figure 5 2. Distribution of As in fluid (A) and As species (B) during simulated injec tions of model (1). (A ) (B )

PAGE 175

159 concentration of total As in the fluid was up to 45 g/L (Figure 52B) This type of the model was compatible to the results of the leaching experiments demonstrated in third chapter (Figure 45). (2) Water rock interaction between the a quifer matrix and surface water (Open System) The main hypothesis of this model was to describe the water rock interaction between the aquifer matrix and drinking water where the water was reacted directly with rock but it a contrast to model (1) the sys tem was considered to be open, i.e. the source of O 2 was fixed and essentially infinite (i.e., did not decrease) This step was important to evaluate how much pyrite would react and the amount of As would be le ached out from the aquifer matrix, if O 2 were not limited The composition of the aquifer matrix was following: 258 g of calcite (CaCO 3 ), 35 g of dolomite (CaMgCO 3 ), 10 g of gypsum (CaSO 4 ), 0.2 g of pyrite (FeS 2 ), and 0.002 g of arsenopyrite (AsFeS) and was based on the bulk rock chemical analysis. Th is geochemical modeling exercise demonstrated that similarly to model (1) about 0.6 mg of pyrite and 0.006 mg of arsenopyrite were dissolved over 5 hours. T he pH of the system increased from 6.9 to 7.3 at the end of reaction progress (Figure 53B). In contr ast to a model (1), Eh remained positive indicating oxidizing environment (Figure 53A) Although, the same amount of pyrite and arsenopyrite reacted during the simulations (0.6 mg and 0.006 mg, respectively), the concentration of As in the fl uid was only 0 .3 ug/L (Figure 54A) About 125 g/L of As was sorbed onto HFO via weak bonding (FeOHAsO 4 2 and FeHAsO 4 ) (Figure 54B) Therefore, the type of model did not reflect the results of leaching experiments de monstrated in third chapter.

PAGE 176

160 Figure 53 Distri but ion of Eh (A) and pH (B) during simulated injections of model (2). (A ) (B )

PAGE 177

161 Figure 5 4. Distribution of As in fluid (A) and As species (B) during simulated injections of model (2). (A ) (B )

PAGE 178

162 (3) Water rock interaction between the aquifer matrix and surface water (Ope n System; high amount of pyrite) The major hypothesis of a model (3) was to describe the water rock interaction of the system that would release the same amount of As as in a model (1) by increasing the amount of pyrite/arsenopyrite but remaining open (f ixed source of O 2 ) as a model (2) Here, the aquifer matrix was composed of 258 g of calcite (CaCO 3 ), 35 g of dolomite (CaMgCO 3 ), 10 g of gypsum (CaSO 4 ), and the amount of pyrite (FeS 2 ) and arsenopyrite (AsFeS) was modified to 0.7 g and 0.007 g respective ly. According to geochemical simulations, Eh of the system remained positive indicating oxidizing environment (Figure 55A) T he pH of the system dropped from 6.9 to 5.5 contradicting with the experimental data (Figures 55 B and 41). Similarly to the model (1), the concentration of total As in the fluid was up to 45 g/L (Figu re 56 ). In order to achieve this level of As concentration, approximately 700 mg of pyrite needed to be dissolved from the column According to the bench scale leaching tests the Avon Park Formation (255 m to 257 m) contained about 800 mg/kg of pyrite (Table 13) or 200 mg in the column. Therefore, this type of model did not agree with the results of leaching experiments demonstrated in the third chapter and should be considered with cau tion. (4) 1D reaction transport model of water rock interaction between the aquifer matrix and surface water (Closed System) The X1t code was used to model reactive transport in one dimension (Bethke, 2006b). This model was important to closely simulat e the results of bench scale experiments in time and space and to evaluate the geochemical processes in the leaching

PAGE 179

163 Figure 55 Distribution of Eh (A) and pH (B) during simulated injections of model (3). (A ) (B )

PAGE 180

164 Figure 56 Distribution of As in fluid (A) and As species (B) during simulated injections of model (3) (A ) ( B )

PAGE 181

165 column such as t he stability of pyrite and HFO, and the mobility of As The transport model was constructed for a closed system ( i.e. the source of O 2 was limited ), with the same composition of initial and inlet water (drinking water) and the following aquifer matrix composition: 258 g of calcite (CaCO 3 ), 35 g of dolomite (CaMgCO 3 ), 10 g of gypsum (CaSO 4 ), 0.2 g of pyrite (FeS 2 ), and 0.002 g of arsenopyrite (AsFeS) (the same as model s (1) and (2) ) (Appendix G). The simulation column dimensions were 50 cm (Length) by 2 cm (Width). Based on the leaching experiments, the discharge rate of fluid in the column was 0.01 cm 3 /cm 2 sec, and the porosity was 15 %. This geochemical modeling exercise showe d that about 0.06 mg/cm 3 and 0.00006 gm/cm 3 of pyrite and arsenopyrite were dissolved in the column over 5 hours. T he pH of the system varied from 6.9 (initial fluid) to 7.4. At the same time, t he highest pH was detected at 2.5 cm and gradually decreased to about 7.2 at 50 cm (Figure 57 C ). The concentration of As in the fluid reached up to 65 g/L at the end of the column (Figure 58 A ). Figure 59 demonstrated interesting behavior of As species along the column. According to the model, the oxidative dissolutio n of pyrite /arsenopyrite depletion of O 2 and As sorption onto newly formed HFO via weak bonding (FeOHAsO 4 2 FeHAsO 4 FeH 2 AsO 4 and FeH 2 AsO 3 ) was between 0 cm and 22 cm (Figure 59B) However, Eh of the system changed rapidly from the oxidizing to reduci ng (+700 to 150 mV) between 22 cm and 27 cm (Figure 57B) The model showed a sharp decrease of adsorbed As and steep increase of aqueous As concentration around 24 cm which was governed by the reductive dissolution of HFO. Ideally, the slope of absorbed a nd mobilized As should be the same but it was lower for the released As (Figure 5 9 ). The column interval between 27 cm and 50 cm showed that As(OH) 3 and As(OH) 4 species were dominant in the

PAGE 182

166 Figure 57 Distribution of mineral saturation states (A) Eh (B) and pH (C) along the column during 5 hours of simulated injections of model (4). (A) (B ) (C )

PAGE 183

167 Figure 58 Distribution of As concentration in fluid along the column with discharge rate of 0.01cm/sec (top) and 0.005 cm/sec (bottom) during 5 hours of simula ted injections of model (4). (A ) (B )

PAGE 184

168 Figure 59 Distribution of As species along the column during 1 hour (A) 5 hours (B) and 10 hours (C) of simulated injections of model (4). (A) (B ) (C )

PAGE 185

169 column and the system was sa turated with respect to pyrite (Figures 57A and 5 9B ). This pattern could be due to a possible non oxidative dissolution of pyrite or arsenopyrite driven by pH. The decrease of discharge rate from 0.01 cm 3 /cm 2 sec to 0.005 cm 3 /cm 2 sec (or 1 mL/min) demonstrated the reduced size of the reaction front of so rbed As species between 0 cm and 7 cm. In contrast, the concentration of As in the fluid increased from 65 g/L to 150 g/L (Figure 58B) In addition to 5 hours, injection times of 1 hour and 10 hours were investigated (Figures 59A and C) According to th ose simulations, the concentration of As in leachate varied from 350 g/L (for 1 hour) to 0.2 g/L (for 10 hours). Figure 59 clearly showed a broader distribution of sorbed As species during the 10 hour simulation compared to 1 hour. This tendency could be explained by a longer fluid flushing through aquifer matrix which supplied O 2 for oxidative dissolution of pyrite. As mentioned above, the elevated concentrations of As during 1 hour simulation could be governed by pH which drives nonoxidative dissolution of pyrite or arsenopyrite. Figure 60 showed the distribution of As species negative Eh and pH along the column support ing this hypothesis. 5) 1D reaction transport model of water rock interaction between the aquifer matrix, groundwater and surface wate r (Closed System) The present geochemical model, based on the model (4), was significantly modified to potentially reflect the actual water rock interact ions occurring during the injection of treated ASR water into the Floridan Aquifer. The composition o f the aquifer matrix was the same as in model (4). In contrast, during this simulation the groundwater was equilibrated initially with the aquifer matrix. The second simulation was set for 5

PAGE 186

170 Figure 60 Distribution of As species (A), Eh (B), and pH (C) along the column during 1 hour of simulated injections of model (4). (A ) (B ) (C )

PAGE 187

171 hours rock system used above as a new equilibrium system followed by a new simulation with the injection of drinking water This way the simulation of ASR could be comparable to field tests. According to geochemical simulations, the starting solution (groundwater + rock) was under reducing conditions and was favorable for pyrite stability preventing the release of As into solution. Subsequently, the starting fluid was displaced by the oxic fluid ( drinking water) which was indicated by increase in Eh fr om 150 mV to +750 mV (Figure 61A ). Particularly, during the first 20 minutes of the simulation, pyrite/ arseno pyrite was oxidized releasing large amount s of As giving an initial As pulse ( Figure 62A ). Subsequently, HFO were newly formed adsorbing most the dissolved As (Figure 62B) About 0.004 % of pyrite was dissolved during 5 hours of simulation resulting in the concentration of As of up to 135 g/L (Fi gure 62A). The pH of the system increased from 7.1 to 7.4. The highest pH was detected at 2.5 cm and gradually decreased to a bout 7.3 at 50 cm (Figure 61B). The increasing concentration of Ca 2+ indicated the dissolution of limestone during water rock inter actions. At the end of simulations, about 24 pore volumes were displaced which was equal to 1440 mL. Overall, the geochemical models correlated well to the results from the column leaching experiments and clearly showed that the injection of oxidizing surf ace waters into reducing native groundwater caused a change in redox potential of the system and thus promoted the dissolution of pyrite and mobilization of As Particularly, models (1) and (4) constructed for closed systems were probably the most reliable to characterize the geoch emical processes in the columns. In contrast, the models (2) and (3) developed for open systems showed that most of the pyrite was dissolved and almost all As was sorbed

PAGE 188

172 Figure 61 Distribution of pH (A) and Eh (B) during simulated injections of model (4). (A ) (B )

PAGE 189

173 Figure 62 Distribution of As in fluid (A) and As species (B) during simulated injections of model (4). (A ) (B )

PAGE 190

174 onto HFO via weak bonding and only 0.3 ug/L was present in the fluid (for model 2). Model (3) showed that appr oximately 700 mg of pyrite needed to be dissolved from the column to achieve the level of As concentration compatible to the leaching experiments. Those results did not agree with the bulk rock chemical composition demonstrated in the third chapter This study confirmed the previous model of Jones and Pichler (2007), which predicted the instability of pyrite in the Suwannee Limestone that would result in the leaching of As into storage zone water. Reactive transport modeling by Wallis et al. (2009) confirm ed that the behavior of As from the aquifer matrix in the Netherlands was governed by dissolution of arsenopyrite, which was stoichiometrically connected to pyrite oxidation, transformatio n of As(III) to As(V), followed by As adsorption on neo precipitated HFO via surface complexation 4.4. Conclusions The geochemical reactive transport modeling has shown to be a valuable tool for providing important information about water rock interactions in the Floridan Aquifer. The combination of bench scale leaching experiments and modeling could become essential to forecast redox conditions of the aquifer matrices during operation and development of aquifer storage and recovery systems in Florida and potentially worldwide. The results from the current models correla ted well to those from the column experiments and clearly showed the following: 1) Fluid rock reaction between the aquifer matrix and native groundwater was favorable for pyrite stability preventing the release of As into solution.

PAGE 191

175 2) The injection of oxidizing surface water into reducing native groundwater caused a change in redox potential of the system thus promoting the dissolution of pyrite and precipitation of newly formed HFO. 3) The evaluation of open vs. closed systems (fixed vs. limited O 2 ) illustrated t he following results: Closed system: the model was compatib le to the leaching experiments. The concentration of As in the leachate was to 45 g/L Open system: the model did not agree with the leaching experiments. The concentration of As in the leachate w as only 0.3 ug/L and a bout 125 g/L of As was sorbed onto HFO via weak bonding 4) 1D geochemical model of water rock reaction between the aquifer matrix and drinking water showed a remarkable behavior of As along the column, such as: Oxidative dissolution of pyrite, mobilization and sorption of As onto neo formed HFO; Reductive dissolution of HFO and secondary release of adsorbed As; Potential non oxidative dissolution of pyrite contributing to the additional source As in the solution.

PAGE 192

176 CHAPTER FIVE SUMMAR Y AND CONCLUSIONS The reclamation of wastewater and phosphate mining lands using constructed wetland/filter basin (CW/FB) treatment technology could prove to be especially important as Florida law requires reclamation of previously mined phosphate lands i nto wildlife habitat and wa tershed enhancement. In addition, the CW/FB system may provide water that meets drinking water standards to supplement lakes and rivers, satisfy industrial, agricultural and domestic water supply demands, and to provide more sust ainable alternative to extensive groundwater consumption. This study investigated the efficiency of the CW/FB treatment system to improve the water quality of industrial and municipal wastewater. The system was constructed in closed phosphate mines used fo r clay settling and sand tailings in Polk County, Florida An 18 month performance study of the CW/FB showed that despite of significant seasonal variations with respect to temperature, rainfall and humidity, the chemical/microbiological composition of tre ated water remained relatively constant. The study showed the following changes in water quality between the input and output: s ubstantial decrease of water temperature (up to 10 C), reduction of As, SO 4 F, Cl, NO 3 NO 2 Br, Na, K, Ca, and Mg, cha nge in pH from about 9 to 6.5 7, increase of H 2 S (up to and negative oxidation reduction potential confirming the reducing

PAGE 193

1 77 con ditions of the treatment system. There were no exceeda nces of the primary drinking water standards, volatile organic compound s, synthetic organic compounds, and radion uclides, but a number of exceeda nces for the secondary drinking water standards ( Al, F, Fe, Mn, color, odor, total dissolved solids, and foaming agents ) The concentration of fecal and total coliform bacteria at th e filter basin was reduced from 30 730 and 1000 7000 count/100 mL to < 2 and < 100 count/ 100 mL, respectively. The se results clearly demonstrate d A combined isotope/chemical mass balance approach to evaluat e the performance of the wetland in complex hydrogeological settings demonstrated the following: (1) c omposition of water in the wetland varied throughou t the period of the study; (2) d epletion of isotopic composition along the wetland flow path; (3) w etla nd was mostly compos ed of waste water (88 100 %) during normal pumping operations; however, hurricanes and inconsistent pumping added low conductivity water directly and triggered enhanced groundwater inflow into the wetland of up to 78 %; (4) c omposition of water in monitor wells was mostly groundwater dominated; however it was periodically induced by the seepage from a water body to the north; and the (5) s eepage from water bodies surrounding the wetland were not identified in the wetland water once the system became operational potentially indicating a water loss from the wetland. Of particula r interest in this study were t he bench scale leaching experiments to investigate the mobilization of geogenic arsenic (As) in the Floridan Aquifer under a range o f redox conditions They showed that t he amount of As released from the aquifer matrix was not perfectly correlated with the bulk rock As concentration. The highest level of As was leached out from the Avon Park Formation and the lowest from the

PAGE 194

178 Suwannee Limestone, although the Ocala Limestone had the lowest bulk rock As. Minor to n o As was released using native Floridan groundwater. Tampa drinking water, which che mically and physically resembled the ASR injection water, caused the leaching of As of up to 27 g/L which was higher than the current As drinking water standard. Wetland and filter basin waters showed the highest release of As (up to 68 g/L), which was unexpected because the wetland water was much less oxygenated than Tampa drinking water and t hus should be less aggressive. Overall, the se results could be very important to forecast As behavior during anthropogenic physico chemical changes such as the aquifer storage and recovery procedure. T he experiments confirmed that perturbations of native aquifer conditions caused the release of As from the Floridan Aquifer matrix, although the reaction may not be as simple as the dissolution of pyrite by oxygen, but additionally governed by a complex set of factors including the oxidation reduction potenti al of the system, SO 4 2 /S 2 Fe 3+ /Fe 2+ dissolved organic carbon and microbial activity. T he oxidative dissolution of pyrite caused coprecipitation of released As with neo formed hydrous ferric oxides (HFO) which could become an additional source of As for the native groundwater under reducing conditions. Moreover, the following set of parameters could have a fundamental role in the mobilization of As, such as the porosity and permeability of the aquifer influencing the rate and degree of free water saturati on, amount of pyrite to be exposed to the preferential flow paths during water rock interactions, limited surface reactivity of pyrite, favored reactions on fractured mineral surfaces, concentration of As and selective As leaching from individual pyrite cr ystals

PAGE 195

179 T he results of the geochemical reactive transport model ing using the Geochemist's Workbench (React and X1t) correlated well to those from the column experiments. The mo del ing clearly showed that fluid rock interaction between the aquifer matrix an d natural groundwater was favorable for pyrite stability preventing the release of As into solution. In contrast, the injection of oxidizing surface water into reducing native groundwater caused a change in redox potential of the system thus promoting the dissolution of pyrite. The 1D geochemical model simulating a water rock reaction between the aquifer matrix and surface water demonstrated the following processes along the column : oxidative dissolution of pyrite, mobilization and simultaneous sorption of As onto neo formed HFO, followed by the reductive dissolution of HFO and secondary release of adsorbed As, and a potential non oxidative dissolution of pyrite contributing the additional source As in to the solution.

PAGE 196

180 REFERENCES Aguilar, J., Dorronsoro, C, Fernandez, F., Fernndez, J., Garca, I., Martn, F., Sierra, M., and Simn, M., 2007. Arsenic contamination in soils affected by a pyrite mine spill (Aznalcollar, SW Spain). Water Air Soil Pollution 180, 271 281. Ahmann, D.L., Krumholz, R.L. and More l, F., 1994. Microbe grows by reducing arsenic. Nature 371, 750. Ahmed, M.F., Ahuja, S., Alauddin, M., Hug, S.J., Lloyd, J.R., Pfaff, A., Pichler, T., Saltikov, C., Stute, M., and van Geen, A., 2006. Ensuring safe drinking water in Bangladesh, Science. A iken, G., 2002. Everglades water chemistry surface water DOC, pore water DOC and SUVA data. U. S. Geological Survey, Greater Everglades Science Program: Place Based Studies. Alley, W.M., Reilly, T.E. and Franke, O.L., 1999. Sustainability of Ground Wate r Resources. U.S. Geological Survey Circular 1186. Alloway, B.J., 1992. Soil processes and the behavior of metals. In Heavy metals in soils (ed., B.J. Alloway). Blackie Academic and professional, 11 37. Altman, S.J., 1994. Non conservative behavior of br omide, chloride and nitrate during a groundwater tracer test near the saturated/unsaturated interface. GSA Conference 26(6), 24 27. Amini, M., Abbaspour, K.C., Berg, M., Winkel, L., Hug, S.J., Hoehn, E., Yang, H., and Johnson, C.A., 2008. Statistical mode ling of global geogenic arsenic contamination in groundwater. Environmental Science and Technology 42, 3669 3675. Anawar, H.M., Akai, J., Komaki, K., and Terao, H., 2003 Geochemical occurrence of arsenic in groundwater of Bangladesh: sources and mobilizat ion processes. Journal of Geochemical Exploration 77, 109 131.

PAGE 197

181 Anderson, L.C.D. and Bruland, K.W., 1991. Biogeochemistry of arsenic in natural waters: the importance of methylated species. Environmental Science and Technology 25(3), 420 427. Appelo, C.A. J. and de Vet, W.W.J.M., 2003. Modeling in situ iron removal from groundwater with trace elements such as As. In Arsenic in Groundwater, Welch, A.H. and Stollenwerk, K.G., Eds. Kluwer Academic, Boston, 381 401. Appelo, C.A.J., Van der Weiden, M.J.J., Tour nassat, C., Charlet, L., 2002. Surface complexation of ferrous iron and carbonate on ferrihydrite and the mobilization of arsenic. Environmental Science & Technology 36 (14), 3096 3103. rnational Symposium on the Geochemistry of the Earth Surface, Reykjavk, Iceland, 574. Armienta, M.A., and Segovia, N., 2008. Arsenic and fluoride in the groundwater of Mexico. Environmental Geochemistry and Health 30, 345 353. Arthur, J., Fischler, C., Kromhout, C., Clayton, J., Kelley, G.M., Lee, R.A., Li, L., O'Sullivan, M., Green, R., and C., W., 2008. Hydrogeologic framework of the Southwest Florida Water Management District. F.G.S Bulletin 68, 47. Arthur, J.D., Dabous, A.A., and Fischler, C., 2007, Aquifer storage and recovery in Florida: Geochemical assessment of potential storage zones, in: Fox, P., ed., Management of Aquifer Recharge for Sustainability, Acacia Publishing, Phoenix, Arizona, 185 197. Arthur, J.D., Cowart, J.B, and Dabous, A.A., 20 05. Water rock geochemical consideration for aquifer storage and recovery: Florida case studies. In: Tsang, C.F., Apps, J.A. (Eds.), Underground Injection Science and Technology. In: Developments in Water Science, 52. Elsevier, 327 339. Arthur, J.D., Dabo us, A.A., and Cowart, J.B., 2002. Mobilization of arsenic and other trace elements during aquifer storage and recovery, southwest Florida. U.S. Geological Survey Open File Report 02 89. Arthur, J.D., Dabous, A.A., and Cowart, J.B., 2001. Florida aquifer s torage and recovery geochemical study: Year three progress report. Florida Geological Survey Open File Report 83.

PAGE 198

182 Baskaran, S., Ransley, T., Brodie, R.S., and Baker, P., 2009, Investigating groundwater river interactions using environmental tracer. Austra lian Journal of Earth Sciences 56, 13 19. Bauer, M. and Blodau, C., 2006. Mobilization of arsenic by dissolved organic matter from iron oxides, soils and sediments. Science of the Total Environment 354(2 3), 179 190. Bauer, M., Fulda, B. and Blodau, C., 2008. Groundwater derived arsenic in high carbonate wetland soils: Sources, sinks, and mobility. Science of the Total Environment 401, 109 120. Baur, W.H. and Onishi, B.H., 1969. Arsenic. In: K.H. Wedepohl (Ed.), Handbook of Geochemistry. Springer Verlag, Berlin, A1 33. guide uide to R eact and Gtplot University of Illinois. Bethke, C. M., 2006 b modeling gui Illinois. Bowell, R.J., 1994. Sorption of arsenic by iron oxides and oxyhydroxides in soils. Applied Geochemistr y 9(3), 279 286. Brammer, H. and Ravenscroft, P., 2009. Arsenic in groundwater: a threat to sustainable agriculture in south and south east Asia. Environment International 35, 647 654. Braun, M., Barley, B. and Wood, H., 2003. Minewater treatment. Techno logy, Application and Policy, IWA Publishing. Brewster Wingard, G.L., Scott, T.M., Edwards, L.E., Weedman, S.D., Simmons, K.R., 1997. Reinterpretation of the peninsular Florida Oligocene; an integrated stratigraphic approach. Sedimentary Geology 108 (1 4) 207 228. Budd, D., and Vacher, H.L., 2004. Matrix permeability of the confined Floridan Aquifer, Florida, USA. Hydrogeology Journal 12, 531 549. Buddhawong, S., Kuschk, P., Mattusch, J., Wiessner, A., and Stottmeister, U., 2005. Removal of arsenic and zinc using different laboratory model wetland systems. Engineering in Life Sciences 5, 247 252.

PAGE 199

183 Buschmann J., Canonica S., Lindauer U., Hug S. J. and Sigg L., 2005. Photoirradiation of dissolved humic acid induces arsenic (III) oxidation. Environmental Sc ience and Technology 39 (24), 9541 9546. Buschmann J., Kappeler A., Lindauer U., Kistler D., Berg M. and Sigg L., 2006. Arsenite and arsenate binding to dissolved humic acids: Influence of pH, type of humic acid, and aluminum. Environmental Science and Te chnology 40 (19), 6015 6020. Butler, J.E. and Dewedar, A., 1991. Gravel bed hydroponic sewage treatment performance, design and potential for crop production. International Conference on Environmental Pollution, Lisbon, A29/1 A29/8. Cander, H. S., 199 1. Dolomitization and water rock interaction in the Middle Eocene Avon Park Formation, Floridan aquifer. Ph.D. Dissertation. The University of Texas, Austin, Texas. Carruthers, S., 2002. Passive wetland system for intensive horticulture. Practical Hydropo nics and Greenhouses 62. Catarecha, P., Segura, M.D., Franco Ponce, B., Lanza, M., Solano, R., Paz Ares, J., and Leyva, A., 2007. A mutant of the arabidopsis phosphate transporter PHT1;1 displays enhanced arsenic accumulation. The P lant Cell 19, 1123 1133. Chao, T.T., and Theobald, P.K., 1976. The significance of secondary iron and manganese oxides in geochemical exploration. Economic Geology and the Bulletin of the Society of Economic Geologists 71(8) 1560 1569. Chen, Y.C, Guo, Y. L., Su, H.J., Hsueh, Y.M., and Smith, T.J. 2003. Arsenic methylation and skin cancer risk in southwestern Taiwan. Journal of Occupational and Environmental Medicine 45, 241 248. Chen, C.S., 1965. The regional lithostratigraphic analysis of Paleocene and E ocene rocks of Florida. F.G Survey, 45. Chiu, V.Q. and Hering, J.G., 2000. Arsenic adsorption and oxidation at manganite surfaces. 1. Method for simultaneous determination of adsorbed and dissolved arsenic species. Environmental Science and Technology 34 (10), 2029 2034. Christophersen, N. and Hooper, R.P., 1992. Multivariate analysis of stream water chemical data: the use of principal components analysis for the end member mixing problem. Water Resources Research 28(1), 99 107.

PAGE 200

184 Claesson, M. and Fagerber g, J., 2003. Arsenic in groundwater of Santiago del Estero, Argentina; sources, mobilization controls and remediation with natural materials. Minor field studies, 62. Clark, I.D. and Fritz, P., 1997. Environmental isotopes in hydrogeology. CRC, 1 ed.,352. Clarke, R. and King,J., 2004. The Atlas of Water. London: Eartscan. Cole, S., 1998. The emergence of treatment wetlands. Environmental Science and Technology 32(9), 218 223. Colmer, T.D, 2003. Long distance transport of gases in plants: a perspective on internal aeration and radial oxygen loss from roots. Plant, Cell and Environment 26, 17 36. Compton, J. S., Hodell, D. A., Garrido, J. R. and Mallinson, D. J., 1993. Origin and age of phosphorite from the south central Florida Platform: Relation of ph osphogenesis to sea level fluctuations and d13C excursions. Geochimica et Cosmochimica Acta 57, 131 146. Costagliola, P., Cipriani, C., Manganelli Del Fa, C., 1997. Pyrite oxidation: protection using synthetic resins. European Journal of Mineralogy 9, 16 7 174. Craig, H. and Gordon, L., 1965. Deuterium and oxygen 18 variations in the ocean and the marine atmosphere. In Tongiorgi, E., ed, Proceedings of a Conference on Stable Isotopes in Oceanographic Studies and Paleotemperatures. Spoleto, Italy, 9 130. Craig, H., 1961. Isotopic variations in meteoric waters: Science 133, 1702 1703. Criss, R.E. and Winston, W.E., 2003. Hydrograph for small basins following intense storms. Geophysics Research Letters 30 (6), 1314 1318. Criss, R.E. and Winston, W.E., 2003 Hydrograph for small basins following intense storms. Geophysics Research Letters 30 (6), 1314 1318. Davis, L., 1995. A guide to creating wetlands for agricultural wastewater, domestic wastewater, coal mine drainage and stormwater in the Mid Atlantic Re gion 1, general considerations. USDA NRCS, U.S. EPA Region III, and PA. Department of Environmental Resources. de Ruiter, H., and Stuyfzand, P.J., 1998. An experiment on well recharge of oxic water into an anoxic aquifer, in Peters, J.H., ed. Artificial r echarge of groundwater: Nieuwegein, Netherlands, 299 304.

PAGE 201

185 Deschamps, E., Ciminelli, V.S.T., Weidler, P.G. and Ramos, A.Y., 2003. Arsenic sorption onto soils enriched in Mn and Fe minerals. Clays and Clay Minerals 51(2), 197 204. Devito, K. J. and Dillon, P. J. 1993. Importance of runoff and winter anoxia to the P and N dynamics of a beaver pond. Canadian Journal of Fisheries and Aquatic Sciences 50, 2222 2234. Dhar, R.K., Zheng, Y., Stute, M., van Geen, A., Cheng, Z., Shanewaz, M., Shamsudduha, M., Hoque M.A., Rahman, M.W., and Ahmed, K.M., 2008. Temporal variability of groundwater chemistry in shallow and deep aquifers of Araihazar, Bangladesh: Journal of Contaminant Hydrology 99, 97 111. Dhar, R.K., Biswas, B.K., Samanta, G., Mandal, B.K., Chowdhury, R.T., Chanda, C.R., Basu, G., Chakraborti, D., Roy, S., Kabir, S., 1998. Groundwater arsenic contamination and sufferings of people in Bangladesh may be the biggest arsenic calamity in the world. In: Proceedings of the International Conference on Arsenic P ollution of Groundwater in Bangladesh: Causes, Effects and Remedies, Dhaka, Bangladesh. Dhaka, India: Dhaka Community Hospital, 86 87. Dincer, T., 1968. The use of oxygen 18 and deuterium concentrations in the water balance of lakes. Water Resources Rese arch 4, 1289 1306. Dippold, A., 2009. Detailed geochemical investigation of the mineralogic associations of arsenic and antimony within the Avon Park Formation, central Florida: Implications for aquifer storage and recovery. Master Thesis, University of S outh Florida, Tampa, 138. Doctor, D.H., Alexander, E., Petric, M., Kogovsek, J., Urbanc, J., Lojen, S. and Stichler, S., 2006. Quantification of karst aquifer discharge components during storm events through end member mixing analysis using natural chemis try and stable isotopes as tracers. Hydrogeology Journal 14 (7), 1171 1191. Druschel, G.K., Hamers, R.J., Luther, G.W.III and Banfield, J.F., 2004. Kinetics and m echanism of t rithionate and tetrathionate oxidation at low pH by h ydroxyl r adicals Aquatic Geochemistry 00, 1 20. Kluwer Academic Publish ers. Ducci, D., de Masi, G., and delli Priscoli, G., 2008. Contamination risk of the Alburni Karst System (Southern Italy): Engineering Geology 99, 109 120.

PAGE 202

186 Dushenko, W.T., Bright, D.A. and Reimer, K.J., 1995 Arsenic bio accumulation and toxicity in aquatic macrophytes exposed to gold mine effluent: relationships with environmental partitioning, metal uptake and nutrients. Aquatic Botany 50, 141 158. Dzombak, D. A. and Morel, F. M. M., 1990. Surface complexa tion modeling: hydrous ferric oxide. Wiley Interscience: NewYork, Eggleston, C.M., Ehrhardt, J.J., and Stumm, W., 1996. Surface structural controls on pyrite oxidation kinetics: an XPS UPS, STM, and modeling study. American Mineralogist 81, 1036 1056 EPA (US Environmental Protection Agency), 2000. Constructed wetlands treatment of municipal wastewaters. EPA 625 R 99 010. EPA (US Environmental Protection Agency), 1993. Constructed wetlands for wastewater treatment and wildlife habitat, 17 case studies. EP A 832 R 93 005. EPA (US Environmental Protection Agency). 2009. National primary drinking water regulations. EPA 816 F 09 0004, Website: www.epa.gov/safewater. Ermakov A. N., Poskr ebyshev G. A. and Purmal A. P., 1997. Sulfite oxidation: The state of the art of the problem. Kinetics and Catalysis 38, 295 308. Evangelou, V.P. (Bill), 1995. Pyrite oxidation and its control, CRC Press, Inc., Corporate Blvd., N.W., Boca Raton, Florida, 293. Fankenberger, Jr.W.T, 2002. Environmental chemistry of arsenic. Mar cel Dekker, Inc., 391. FAWN (Florida Automated Weather Network). Weather database for the State of Florida. Website: http://fawn.ifas.ufl.edu/. Ferguson, J.F., and Gavis, J., 1972. A review of the arsenic cycle in natural waters. Water Research Pergamon Press 6, 1259 1274. FIPR (Florida Institute of Phosphate Research). Website: www.fipr.state.fl.us Fisher, R.A., 1971. Role of potassium in stomatal opening in the leaf of Vicia faba. Plant Physiology 47, 555 558. FLDEP (Florida Department of Environmen tal Protection), 2008. Underground Injection Control Volume 2008, Tallahassee.

PAGE 203

187 Galloway, J.N, Knap, A.H., Church, T.M., 1983. The Composition of Western Atlantic Precipitation Using Shipboard Collectors. Journal of Geophysical Research 88 (C15), 10859 10864. Gao S., Luo, T. C., Zhang B. R., Zhang, H. F., Han, Y. W., Hu, Y. K., and Zhao, Z. D, 1998. Chemical composition of the continental crust as revealed by studies in east China. Geochimica et Cosmochimica Acta 62, 1959 1975. Gat, J.R., 1996. Oxygen and hydrogen isotopes in the hydrologic cycle. Ann. Rev. Earth Planet. Sci. 24, 225 262. Gbadebo, A.M., 2005. Arsenic pollution in aquifers located within limestone areas of Ogun State, Nigeria. Natural Arsenic in Groundwater: Occurrence, Remediation an d Management, 85 92. Ghosh, M., Shen, J., and Rosen, B.P.,1999. Pathways of As (III) detoxification in Saccharomyces cerevisiae. Proceedings of the National Academy of Sciences USA 96, 5001 5006. Gilboy, A.E., 1985. Hydrogeology of the Southwest Florida Water Management District. Regional Analysis Section Technical Report 85 01. Gonfiantini, R., Frhlich, K., Araguas Araguas, L., and Rozanski, K., 1998. Isotopes in Groundwater Hydrology, in Kendall, C., and McDonnell, J.J., eds., Isotope Tracers in Catc hment Hydrology, Amsterdam, Elsevier, 203 238. Govindaraju, K., 1994. Compilation of working values and sample description for 383 geostandards. Geostandards Newsletter 18, 1 158. Grafe M., Eick M. J., Grossl P. R. and Saunders A. M., 2002. Adsorption of arsenate and arsenite on ferrihydrite in the presence and absence of dissolved organic carbon. Journal of Environmental Quality 31(4), 1115 1123. Green, R.C., Arthur J.D., DeWitt D., 1995. Lithostratigraphic and hydrostratigraphic cross sections through Pinellas and Hillsborough counties, Southwest Florida, Open File Report Florida Geological Survey, Report: 61, 26. Harvey, C.F. and Swartz, C.H., 2002. Arsenic mobility and groundwater extraction in Bangladesh. Science 298, 1602 1606. Hering, J.G. and Kneebone, P.E., 2002. Biogeochemical controls on arsenic occurrence and mobility in water supplies. In: Frankenberger W, ed.: Environmental Chemistry of Arsenic, Marcel Dekker, New York. 7, 155 182.

PAGE 204

188 Hill, V. R. and Sobsey, M. D, 2001. Removal of Salmonell a and microbial indicators in constructed wetlands treating swine wastewater. Water Science and Technology 44(11 12), 215 222. Hislop, K. A. and Bolt on, J.R., 1999. The Photochemical generation of hydroxyl radicals Technology 33 (18), 3119 3126. Hinkle, S.R. and Polette, D.J., 1999. Arsenic of the Willamette Basin, Oregon. Geochemistry of Arsenic. USGS Water Resources Investigations Report 98 4205, 28. Hongshao, Z. and Stanforth, R., 2001. Competitive adsorption of phosphate and arsenate on goethite. Environmental Science and Technology 35, 4753 4757. House, C.H., Bergmann, B. A., Stom p, A. M. and Frederick, D. J., 1999. Combining constructed wetlands and aquatic and soil filters for reclamation and reuse of water. Ecological Engineering 12(1 2), 27 38. Huang, Y.K., Huang, Y.L., Hsueh, Y.M., Yang, M.H., and Wu, M.M., 2008. Arsenic expo sure, urinary arsenic speciation, and the incidence of urothelial carcinoma: a twelve year follow up study. Cancer Causes Control 19, 829 839. Huang, J. H. and Matzner, E., 2006. Dynamics of organic and inorganic arsenic in the solution phase of an acidic fen in Germany. Geochimica et Cosmochimica Acta 70(8), 2023 2033. Huerta Diaz, M.A. and Morse J. W., 1990. Geochemistry of trace metals associated with sedimentary pyrite from anoxic marine sediments. Eos, Transactions, American Geophysical Union, 71(2), 98. Huie, R. E. and P. Neta, 1987. Rate constants for some oxidations of S(IV) by radicals in aqueous solutions. Atmos. Environ. 21, 1743 1447. Huisman, L. and Wood, W.E., 1974. Slow Sand Filtration. World Health Organization, Geneva, Switzerland, 44. Hunt, R.J., Bullen, T.D., Krabbenhoft, D.P, and Kendall, C., 1998. Using stable isotopes of water and strontium to investigate the hydrology of a natural and a constructed wetland. Ground Water 36(3), 434 443.

PAGE 205

189 Inskeep, W.P., McDermott, T.R., and Fendorf, S., 2002. Arsenic (V)/(III) cycling in soils and natural waters: Chemical and microbiological processes. In: Frankenberger, Jr., W. T., ed.: Environmental Chemistry of Arsenic. Marcel Dekker, Inc., 391. Jacob, D.L. and Otte, M.L., 2003. Conflicting proce sses in the wetland plant rhizosphere: metal retention or mobilization? Water, Air and Soil Pollution 3(1), 91 104. Jones, G.W. and Pichler, T., 2007. Relationship between pyrite stability and arsenic mobility during aquifer storage and recovery in South west Central Florida. Environmental Science and Technology 41(3), 723 730. Jones, S.K. and Owen, R.S., 2001, Regional Water Supply Plan. Southwest Florida Water Management District Report, Brooksville Florida. Kadlec, R.H., 2006. Water temperature and ev apotranspiration in surface flow wetlands in hot arid climate. Ecological Engineering 26, 328 340. Kadlec, R.H., and Knight, R.L., 1996. Treatment Wetlands. Lewis Publishers, 893. Kaiser K., Guggenberger G., and Zech, W., 1997. Dissolved organic matter sorption on subsoils and minerals studied by 13C NMR and DRIFT spectroscopy. European Journal of Soil Science 48, 301 310. Katou, H., Clothier, B.E. and Green, S.R., 1996. Anion transport involving competitive adsorption during transient water flow in an andisol. Soil Science Society of America Journal 60 (5), 1368 1375. Katz, B.G., 2004. Sources of nitrate contamination and age of water in large karstic springs of Florida. Environmental Geology 46, 689 706. Kendall, C. and McDonnell, J. J, 2006. Isoto pe tracers in catchment hydrology. Elsevier, 839. Kendall, C. and Coplen, T.B., 2001. Distribution of oxygen 18 and deuterium in river waters across the Unites States. Hydrological Processes 15, 1363 1393. Khan, A.W., Ahmad, S.A., and Salim Ullah Sayed, M.H., 1997. Arsenic contamination in groundwater and its effect on human health with particular reference to Bangladesh. Journal of Preventive and Social Medicine, 16 (1), 65 73. Kish, G.R., Stringer, C.E., Stewart, M.T., and Rains, M.C., 2009. A geochem ical mass balance method for base flow separation, upper Hillsborough River watershed, west central Florida. USGS Scientific Investigations Report.

PAGE 206

190 Koretsky, C.M., Moore, C.M., Lowe, K.L., Meile, C., DiChristina, T.J., and Van Cappellen, P., 2003. Seasona l oscillation of microbial iron and sulfate reduction in saltmarsh sed iments (Sapelo Island, GA, USA). Biogeochemistry 64, 179 203. Korte, N.E., and Fernando, Q., 1991. A review of arsenic (III) in groundwater. Critical Reviews in Environmental Control 21 1 36. Kovacova, E., and Malik, P., 2007. Groundwater vulnerability of the karst fissure hydrogeological structure of south facing slopes of the Nizke Tatry Mts., Slovakia. Acta Carsologica 36, 261 268. Krabbenhoft, D.P, Bowser, C.J., Kendall, C. and Ga t, J.R., 1994. Use of oxygen 18 and deuterium to assess the hydrology of groundwater lake systems. In Baker, L.A., ed., Environmental Chemistry of Lakes and Reservoirs. Washington, American Chemical Society, 67 90. Krabbenhoft, D. P., Bowser, C., Anderso n, M.J.P, and Valley, J.W., 1990. Estimating groundwater exchange with lakes 1. The stable isotope mass balance method. Water Resource Research 26 (10), 2445 2453. Kuhn, A. and Sigg, L., 1993. Arsenic cycling in eutrophic Lake Greifen, Switzerland: Influe nce of seasonal redox processes. Limnology and Oceanography 38, 1052 1059. Lalvani, S.B., Zhang, G., and Lalvani, L.S., 1996. Coal pyrite passivation due to humic acids and lignin treatment. Fuel Science and Technology International 14 (9), 1291 1313. La ngner, H.W., Allen, W.C., Hook, P.B., Stein, O.R., and Inskeep, W.P., 1999. Reduction and solid phase control of dissolved arsenic in constructed wetlands. In: Abstracts, Soil Science Society of America Annual Meeting, Salt Lake City, UT. Lazareva, O. and Pichler, T., 2010. Long term performance of a constructed wetland/filter basin system treating wastewater, Central Florida. Chemical Geology 269, 137 152. Lazareva, O. and Pichler, T., June 2009. Isotope/chemical mass balance estimation of water inflo w into a constructed wetland. Goldschmidt Conference on Geochemistry, Davos, Switzerland.

PAGE 207

191 Lazareva, O. and Pichler, T., 2007. Naturally occurring arsenic in the Miocene Hawthorn Group, southwestern Florida: Potential implication for phosphate mining. App lied Geochemistry 22, 953 973. Le, X.C, 2002. Arsenic speciation in the environment and humans. In: Frankenberger W, ed.: Environmental Chemistry of Arsenic, Marcel Dekker, New York. 7, 155 182. Liu, J., Dong, Y., Xu., H., Wang, D., and Xu., J., 2007. Ac cumulation of Cd, Pb and Zn by 19 wetland plant species in constructed wetland. Journal of Hazardous Materials 147, 947 953. Lu, L., Wang, R., Chen, F., Xue, J., Zhang, P., and Lu, J., 2005. Element mobility during pyrite weathering: implications for acid and heavy metal pollution at mining implicated sites. Environmental Geology 49 (1), 82 89. Lombi E., Sletten R.S., and Wenzel, W.W., 2000. Sequentially extracted arsenic from different size fractions of contaminated soils. Water Air Soil Pollution 124, 319 332. Macur, R.E., Wheeler, J.T., McDermott, T.R. and Inskeep, W.P., 2001. Microbial populations associated with the reduction and enhanced mobilization of arsenic in mine tailings. Environmental Science and Technology 35 (18), 3676 3682. Mann, C.J. a nd Wetzel, R.G., 1995. Dissolved organic carbon and its utilization in a riverine wetland ecosystem. Biogeochemistry 31, 99 120. Manning, J.A. and Goldberg, S., 19987. Adsorption and stability of arsenic (III) at the clay mineral water interface. Environ mental Science and Technology 31, 2005 2011. Martin, C.M. and Moshiri, G. A., 1994. Nutrient reduction in an in series constructed wetland system treating landfill leachate. Water Science Technology 29(4), 267 272. Masschelein, W. J., 2002. Ultraviolet l ight in water and wastewater sanitation. Lewis Publishers, 192. Masscheleyn, P.H., Dlaune, R.D., and Patrick, W.H. Jr., 1991. Effect of redox potential and pH on arsenic speciation and solubility in a contaminated soil. Environmental Science and Technolog y 25, 1414 1418.

PAGE 208

192 Mays, P. A. and Edwards, G. S., 2001. Comparison of heavy metal accumulation in a natural wetland and constructed wetlands receiving acid mine drainage. Ecological Engineering 16, 487 500. McCarthy, K.T., Pichler, T., and Price, R.E., 20 05, Geochemistry of Champagne Hot Springs Shallow Hydrothermal Vent Field and Associated Sediments, Dominica, Lesser Antilles: Chemical Geology 224, 55 68. M cKibben M. A. and Barnes H. L.,1986. Oxidation of pyrite in low temperature acidic solutions: Rate laws and surface textures. Geochimic a et Cosmochimica Acta 50, 1509 1520. McMahon, P.B., Bohlke, J.K., Kauffman, L.J., Kipp, K.L., Landon, M.K., Crandall, C.A., Burow, K.R., and Brown, C.J., 2008. Source and transport controls on the movement of nitrate to public supply wells in selected principal aquifers of the United States. Water Resources Research 44, 17. McNeill, L.S., Hsiao wen, C., and Edwards, M., 2002. Aspects of arsenic chemistry in relation to occurrence, health and treatment. In Frankenberge r, W.T., Jr, ed., Environmental chemistry of arsenic: New York, Marcel Dekker, 141 154. Michaels, S.S., 1995. Regional analysis of lakes, groundwaters and precipitation, Miller, J.A., 1986. Hydrogeologic framewo rk of the Floridan aquifer system in Florida and in parts of Georgia, South Carolina, and Alabama. USGS Professional Paper, Report P 1403 B, B1 B91. Mirecki, J.E., 2006. Geochemical models of water quality changes during aquifer storage and recovery (ASR) cycle tests, Phase I: Geochemical models using existing data. US Army Corps of Engineers, Engineers Research and Development Center, TR 06 8. Mirecki, J.E., 2004. Water quality changes during cycle tests at aquifer storage and recovery (ASR) systems in s outh Florida. US Army Corps of Engineers, Engineers Research and Development Center, TR 04 8. Missimer, T.M., 1997. Late Oligocene to Pliocene sea level history of the southern Florida Platform based on sequence stratigraphy. GSA 29 (3), 59.

PAGE 209

193 Mladenov, N ., Zheng, Y., Miller, M.P., Nemergut, D. R., Legg, T., Simone, B., Hageman, C., Rahman, M.M., Ahmed, K.M., and McKnight, D.M., 2010. Dissolved Organic Matter Sources and Consequences for Iron and Arsenic Mobilization in Bangladesh Aquifers. Environmental S cience and Technology 44, 123 128. Mok, W. M. and Wai, C. M., 1990. Distribution and mobilization of arsenic species in the Coeur D'Alene River, Idaho. Environmental Science and Technology, 24, 102 108. Moldovan, B.J. and Hendry, M.J., 2005. Characterizi ng and quantifying controls on arsenic solubility over a pH range of 1 11 in a uranium mill scale experiment. Environmental Science and Technology 39 (13), 4913 4920. Muhammad, N., Parr, J., Smith, M.D., and Wheatley, A.D., 1997. Removal of heavy metals b y slow sand filtration. Proceedings of the 23rd WEDC International Conference on Water Supply and Sanitation, Durban, South Africa, 167 170. Neralla, S. and Weaver, R.W., 2000. Improvement of domestic wastewater quality by subsurface flow constructed wetl ands. Bioresource Technology 75, 19 25. Nesbitt, H.W. and Muir, I. J., and Prawn, A. R. 1995 Oxidation of arsenopyrite by air and air saturated, distilled water, and implications for mechanism of oxidation Geochi mica et Cosmochimica Acta 59 ( 9 ) 1773 1786 Nesbitt, H.W. and Muir, I. J., 1994. X ray photoelectron spectroscopic study of a pristine pyrite surface reacted with water vapour and air Geochi mica et Cosmochimica Acta 58 ( 21 ), 4667 4679. Nickson R.T., McArthu r J. M., Ravenscroft, P., Burgess W. G., and Ahmed K.M., 2000. Mechanism of arsenic release to groundwater, Bangladesh and West Bengal. Applied Geochemistry 15(4), 403 413. Nordstrom, D.K., 2002. Worldwide occurrences of arsenic in ground water. Science, 296, 2143 2145. Nriagu, J.O., 2002. Arsenic poisoning through the ages. In: Frankenberger, Jr., W. T., ed.: Environmental Chemistry of Arsenic. Marcel Dekker, Inc., 391. Obeidat, M.M., Ahmad, F.Y., Hamouri, N.A., Massadeh, A.M., and Athamneh, F.S., 2008 Assessment of nitrate contamination of karst springs, Bani Kanana, northern Jordan. Revista Mexicana De Ciencias Geologicas 25, 426 437.

PAGE 210

194 Oremland, R.S., Newman, D.K., Kail, B.W. and Stolz, J.F., 2002. Bacterial respiration of arsenate and its significa nce in the environment. In: Frankenberger W, ed.: Environmental Chemistry of Arsenic, Marcel Dekker, New York. 7, 155 182. Osborne, F.H. and Ehrlich, H.L., 1976. Oxidation of arsenite by a soil isolate of Alcaligenes. Journal of Applied Bacteriology 41, 2 95 305. Peck, M.F., McFadden, K.W., and Leeth, D.C., 2005. Effects of decreased ground water withdrawal on ground water levels and chloride concentrations in Camden County, Georgia, and ground water levels in Nassau County, Florida, from September 2001 t o May 2003. USGS Report 2004 5295, 36. Pettigara B. R., Bl ough N. V., and Mignerey A. C., 2002. Mechanisms of hydrogen peroxide decomposition in soils. Environmental Science and Technology 36, 639 645. PEF (Progress Energy Florida), 2004. Current Lines, DEF 001 04. Website: www.progress energy.com Phillips, S.E. and Taylor, M.L., 1976. Oxidation of arsenite to arsenate by Alcaligenes faecalis. Applied Environmental Microbiology 32, 392 399. Pich ler, T. and Lazareva, O., 2010. Using mine sites to const ruct wetlands for sustainable water treatment: An example from central Florida. American Chemical Society National Meeting & Exposition, San Francisco, California. Pichler, T., 2005a. Stable and radiogenic isotopes as tracers for the origin, mixing and su bsurface history of fluids in shallow water hydrothermal systems. Journal of Volcanology and Geothermal Research 139, 211 226. ) as a tracer for battery acid (H2SO4) contamination. Environmental Geology 47, 215 224. Pichler, T., Arthur, J., Price, R., and Jones, G., 2004, The arsenic problem during aquifer storage and recovery (ASR). Goldschmidt Geochemistry conference: Processes in Geochemistry, Forces, Fluxes, and Structure; Copenhagen, Denmark. Pichler, T., Hendry, J. and Hall, G.E.M., 2001. The Mineralogy of arsenic in uranium mine tailings at the Rabbit Lake in pit facility, Northern Saskatchewan, Canada. Environmental Geology 40(4 5), 495 506.

PAGE 211

195 Pichler, T., Veizer, J., and Hall, G.E.M., 1999. Natural input of arsenic into a coral reef ecosystem by hydrothermal fluids and its removal by Fe(III) oxyhydroxides. Environmental Science and Technology 33 (9), 1373 1378. Polizzotto, M.L., Harvey, C.F., Sutton, S.R., and Fendorf, S., 2005. Processes conducive to the release and transport of arsenic into aquifers of Bangladesh. Proceedings National Academy of Sciences of the USA. Poole, B.R., 2009. Solving the ASR Arsenic Problem: Pilot Tested Approach Shows Promising Results. AWRA Conference, Seattle, Washington. Postma, D., Larsen, F., Minh Hue, N.T., Duc, M.T., Viet, P.H., Nhan, P.Q., Jessen, S., 2007. Arsenic in g roundwater of the Red River floodplain, Vietnam : Controlling geochemical processes and reactive transport modeling. Geochimica et Cosmochimica Acta 71 (21), 5054 5071. Pratesi, G. and Cipriani, C., 2000. Selective depth analyses of the alteration product s of bornite, chalcopyrite and pyrite performed by XPS, AES, RBS European Journal of Mineralogy 12,397 409 Price, R.E. and Pichler T., 2006. Abundance and mineralogical association of arsenic in the Suwannee Limestone (Florida): Implications for arsenic release during water rock interaction. Chemical Geology 228(1 3), 44 56. Pu, Y.S., Yang, S.M., Huang, Y.K., Chung, C.J., and Huang, S.K., 2007. Urinary arsenic profile affects the risk of urothelial carcinoma even at low arsenic exposure. Toxicology and Applied Pharmacology 218, 99 106. Pyne, R.D.G., 2005. Aquifer Storage and Recovery, A guide to groundwater recharge through wells. Gainesville, Florida, ASR Systems Publications. Randazzo, A. and Jones, D.S., 1997. The geology of Florida. University Pres s of Florida; 1st ed., 400. Ravenscroft, P., Brammer, H., and Richards, K.S., 2009. Arsenic pollution: a global synthesis. Wiley Blackwell, U.K. Redman, A.D., Macalady, D.L., and Ahmann, D., 2002. Natural organic matter affects arsenic speciation and sor ption onto hematite. Environmental Science and Technology 36, 2889 2896. Reese, R.S., 2002. Inventory and review of aquifer storage and recovery in southern Florida. USGS Water Resources Investigations Report 02 4036.

PAGE 212

196 Richardson, S. And Vaughan, D.J., 19 89. Arsenopyrite: a spectroscopic investigation of altered surfaces. Mineralogical Magazine 53, 223 229. Ritcey, G. M., 1989. Tailings management: problems and solutions in the mining industry. Elsevier, 969. Ritter,K., Aiken, G.R., Ranville, J.F., Bau er, M. and Macalady, D.L., 2006. Evidence for Fe(III). Environmental Science and Technology 40(17), 5380 5387. Rochette, E.A., Bostick, G.Li., and Fendorf, S., 2000. Kinetics of arsenate reduc tion by dissolved sulfide. Environmental Science and Technology 34, 4714 4720. Roden, E.E., Urrutia, M.M., and Mann, C.J., 2000. Bacterial reductive dissolution of crystalline Fe(III) oxides in continu ous flow column reactions. Applied Environmental Micro biology 66, 1061 1065. Romero, F.M., Armienta, M.A., and Carrillo Chavez, A., 2004. Arsenic sorption by carbonate rich aquifer material, a control on arsenic mobility at Zimapan, Mexico. Archives of Environmental Contamination and Toxicology 47, 1 13. Sa ada, A., Breeze, D., Crouzet, C., Cornu, S., and Baranger, P., 2003. Adsorption of arsenic (V) on kaolinite and on kaolinite humic acid complexes. Role of humic acid nitrogen groups. Chemosphere 51, 757 763. Sacks, L.A., 2002. Estimating ground water infl ow to lakes in Central Florida using the isotope mass balance approach. USGS Water Recourses Investigation Report 02 4192. Sacks, L.A., Swancar, A. and Lee, T.M., 1998. Estimating groundwater exchange with lakes using water budget and chemical mass balan ce approaches for ten lakes in ridge areas of Polk and Highlands Counties, Florida. USGS Water Recourses Investigation Report 98 4133, 51. Sacks, L.A. and Tihansky, A.B., 1996. Geochemical and isotopic composition of ground water, with Emphasis on source s of sulfate, in the Upper Floridan Aquifer and Intermediate Aquifer System in Southwest Florida. USGS Report 96 4146. Salmassi, T.M., Venkateswaren, K., Satomi, M., Nealson, K.H., Newman, D.K. and Hering, J.G., 2002. Oxidation of arsenite by Agrobacteriu m albertimagni, AOL15, sp. nov., isolated from Hot Creek, California. Geomicrobiology 19, 53 66.

PAGE 213

197 Scott, T. M., Campbell, K. M., Rupert, F. R., Arthur, J. D., Missimer, T. M., Lloyd, J. M., Yon, J. W., and Duncan, J. G., 2001. Geologic Map of the State of Florida. Florida Geological Survey & Florida Department of Environmental Protection, Map Series 146. Scott, T. M., 2001a. Water sustainability in Florida responsibilities. GSA Annual Meeting, Boston MA, Paper No. 83 0. Scott, T., 1992. A geological overview of Florida, in Survey, F.G., ed. Volume Open File Report 50, Florida Geological Survey. Scott, T.M., 1990. The lithostratigraphy of the Hawthorn Group of peninsular Florida. FGS Open File Report 36, 12. Scott, T. M., 1988. The lithostratigraphy of the Hawthorn Group (Miocene) of Florida. FGS Bulletin 59, 148. Sikdar, P.K., and Sahu, P., 2009. Understanding wetland sub surface hydrology using geologic and isotopic signatures: Hydrology and Earth System Sciences 13, 1313 1323. Silver, S., Phung, L.T., and Rosen, B.P., 2002. Arsenic metabolism: resistance, reduction, and oxidation. In: Frankenberger, Jr., W. T., ed.: Environmental Chemistry of Arsenic. Marcel Dekker, Inc., 391. Sims, K. W. W., Newsom, H. E., and Gl adney, E. S., 1990. Chemical fractionation Sb, W, and M., and In Origin of the Earth (eds. H. E. Newsom, J. H. Jones, and J. H. Newson). Oxford University Press, Oxford, 291 317. S imsek, C., Elci, A., Gunduz, O., and Erdogan, B., 2008. Hydrogeological and hydrogeochemical characterization of a karstic mountain region. Environmental Geology 54, 291 308. Smedley, P. L., and Kinniburgh, D.C., 2002. A review of the source, behavior and distribution of arsenic in natural waters. Applied Geochemistry 17 (5), 517 568. Smith A.H., Ercumen, A., Yuan, Y., and Steinmaus, C.M., 2009. Increased lung cancer risks are similar whether arsenic is ingested or inhaled. Journal of Exposure Science and Environmental Epidemiology 19, 343 348. Smith, A.H., Lingas, E.O., Rahman, M., 2000. Contamination of drinking water by arsenic in Bangladesh: a public health emergency. Bulletin of the World Health Organization, 78 (9), 1093 1103.

PAGE 214

198 Smith, A. H., Hopenha yn Rich, C., Bates, M. N., Goeden, H. M., Hertz Picciotto, I., Duggan, H. M., Wood, R., Kosnett, M. J., and Smith, M. T., 1992. Cancer risks from arsenic in drinking water. Environmental Health Perspectives 97, 259 267. Sprinkle, C.L., 1989. Geochemistry of the Floridan aquifer system in Florida and in parts of Georgia, South Carolina, and Alabama. USGS Professional Paper 1403 I, 105. Stauffer, R.E., 1985. Use of solute tracers released by weathering to estimate groundwa ter inflow to seepage lakes. Environmental Science and Technology 19, 405 411. Steinmaus, C., Bates, M.N., Yuan, Y., Kalman, D., and Atallah, R., 2006. Arsenic methylation and bladder cancer risk in case control studies in Argentina and the United States. Journal of Occupational and Environmental Medicine 48, 478 488. Stollenwerk, K.G., Breit, G.N., Welch, A.H., Yount, J.C., Whitney, J.W., Foster, A.L., Uddin, M. N., Majumder, R.K., Ahmed, N., 2007. Arsenic attenuation by oxidized aquifer sediments in Ba ngladesh. Science of the Total Environment 379 (2 3), 133 150. Stollenwerk, K.G., 2003. Geochemical processes controlling transport of arsenic in groundwater: a review of adsorption. In Welch, A.H. and Stollenwerk, K.G. Arsenic in ground water: Geochemist ry and occurrence. Kluwer Academic Publishers, Boston, 475. Stolz, M. and Greger, M., 2002. Accumulation properties of As, Cd, Cu, Pb and Zn by four wetland plant species growing on submerged mine tailings. Environmental and Experimental Botany 47, 271 28 0. Stottmeister, U., Buddhawong, S., Kuschk, P., Wiessner, A., and Mattusch, J., 2006. Constructed wetlands and their performance for treatment of water contaminated with arsenic and heavy metals. In: Soil and Water Pollution Monitoring, Protection and Re mediation, NATO Science Series 69, 417 432. Stringfield, V.T., 1966. Artesian water in Tertiary limestone in the southeastern States. U.S. Geologic Survey Professional Paper 517, 32.

PAGE 215

199 Stuben, D., Berner, Z., Chandrasekharam, D., and Karnakar, J., 2003. Ar senic enrichment in ground water of west Bengal, India: Geochemical evidence for mobilization of As under reducing conditions. Applied Geochemistry 18, 1417 1434. Stumm W. and Morgan J. J., 1996. Aquatic chemistry: Chemical equilibria and rates in n atura l w aters. Third edition, Wiley & Sons, New York, 1022 Stute, M., Zheng, Y., Schlosser, P., Horneman, A., Dhar, R.K., Datta, S., Hoque, M.A., Seddique, A.A., Shamsudduha, M., Ahmed, K.M., and van Geen, A., 2007. Hydrological control of As concentrations i n Bangladesh groundwater. Water Resources Research 43. Stuyfzand, P.J. and Timmer, H., 1999. Deep well injection at the Langerak and Nieuwegein sites in the Netherlands: chemical reactions and their modeling. Kiwa SWE 96.006. Nieuwegein, Netherlands Stu yfzand, P.J., 1998. Quality changes upon injection into anoxic aquifers in the Netherlands: Evaluation of 11 experiments. In Peters, J.H. et al., Eds., Artificial Recharge of Groundwater, A.A. Balkema, Rotterdam, Netherlands, 283 291. Swancar, A. and Hutc hinson, C.B., 1995, Chemical and isotopic composition and potential for contamination of water in the Upper Floridan Aquifer, West Central Florida, 1986 89. USGS Water Supply Paper 2409. Swartz, C.H., Blute, N.K., Badruzzman, B., Ali, A., Brabander, D., J ay, J., Besancon, J., Islam, S., Hemond, H.F., and Harvey, C.F., 2004. Mobility of arsenic in a Bangladesh aquifer: inferences from geochemical profiles, leaching data, and mineralogical characterization: Geochimica et Cosmochimica Acta 68, 4539 4557. SW FWMD, 2000. Aquifer characteristics within the Southwest Florida Water Management District. Report 99 1, 123. Taylor, S.R. and McLennan, S.M., 1985. The continental crust; composition and evolution. Geoscience Texts. Blackwell Scientific Publications, Oxf ord, 312. Thomas, R.C., and Sanders, J.A., 1998. Arsenic coprecipitation in low temperature pyrites: implications for bioremediation via sulfate reducing bacteria. Geological Society of America Abstracts with Program, 30(7), A 58.

PAGE 216

200 Thornton, I., 1999. Ars enic in the global environment: Looking toward the millennium. In Chappell, W.R., Abernathy, C.O., Caldron, R.L., eds. Arsenic Exposure and Health Effects. Amsterdam, Elsevier, 1 7. Tufano, K. J. and Fendorf, S., 2008. Confounding impacts of iron reducti on on arsenic retention. Environmental Science and Technology 42, 4777 4783. Ujang, Z., 2009. Water Sustainability & Strategic Initiatives: A proposed water industry roadmap 2010 2020. Malaysia Water Conference and Exhibition, PWTC. Urban, N.R, Brezonik P.L, Baker, L.A, Sherman, L.A, 1994. Sulfate reduction and diffusion in sediments of Little Rock Lake, Wisconsin. Limnology and Oceanography 39, 797 815. USGS, 1999. Ground water atlas of the United States, Alabama, Florida, Georgia, South Carolina. HA 730 G, Reston, Va. The Survey. Vanderzalm, J., Dillon, P., Le Gal La Salle, C., 2007. Arsenic mobility under variable redox conditions induced during ASR. In Management of Aquifer Recharge for Sustainability, Fox, P., Ed. Acadia Publishing, Phoenix, Arizo na, 211 257. van Geen, A., Radloff, K., Aziz, Z., Cheng, Z., Huq, M.R., Ahmed, K.M., Weinman, B., Goodbred, S., Jung, H.B., Zheng, Y., Berg, M., Trang, P.T.K., Charlet, L., Metral, J., Tisserand, D., Guillot, S., Chakraborty, S., Gajurel, A.P., and Upre ti, B.N., 2008. Comparison of arsenic concentrations in simultaneously collected groundwater and aquifer particles from Bangladesh, India, Vietnam, and Nepal. Applied Geochemistry 23, 3244 3251. Varner, R.K., Crill, P.M. and Talbot, R.W., 1999. Wetlands: a potentially significant source of atmospheric methyl bromide and methyl chloride. Geophysical Research Letters 26 (16), 2433 2436. Vesper, D.J., and White, W.B., 2003. Metal transport to karst springs during storm flow: an example from Fort Campbell, Ke ntucky/Tennessee, USA: Journal of Hydrology 276, 20 36. Viers, J., Dupre, B., Braun, J J., Freydier, R., Greenberg, S., Ngoupayou, J.N., and Nkamdjou, L.S., 2001. Evidence for non conservative behavior of chloride in humid tropical environments. Aquatic G eochemistry 7, 127 154.

PAGE 217

201 Volk C., Wood, L., Johnson, B., Robinson, J., Zhu H. W. and Kaplan L., 2002. Monitoring dissolved organic carbon in surface and drinking waters Journal of Environmental Monitoring, 4, 43 47, DOI: 10.1039/b107768f Vrhovsek D., K ukanja V., and Bulc T, 1996. Constructed wetland (CW) for industrial waste water treatment, Water Research 30(10), 2287 2292. Walker, F.P., Schreiber, M.E. and Rimstidt, J.D., 2006. Kinetics of arsenopyrite oxidative dissolution by oxygen. Geochimica et C osmochimica Acta 70, 1668 1676. Wallis, I., Prommer, H., Post, V., Stuyf zand, P. And Simmons, C., 2009. Numerical evaluation of arsenic mobilisation during deepwell injection of aerobic groundwater into a pyritic aquifer Goldschmidt Conference on Geochem istry, Davos, Switzerland. Environmental Chemistry 2, 119 124. Watkins, K., 2006. Human Development Report. Published for the United Nations Development Programme (UNDP), 4 22. Webb, D.A., 1939. The Sodium and Potassium Content of Sea Water. Journal of Experimental Biology 16, 178 183. Welch, A.H. and Stollenwerk, K.G., 2003. Arsenic in ground water: Geochemistry and occurrence. Kluwer Academic Publishers, Boston, 475. Wil liams, H., Cowart, J.B., Arthur, J.D., 2002. Florida aquifer storage and recovery geochemical study, Southwest Florida. Year One and Year Two Progress Report. Florida Geological Survey, Tallahassee, 100. Winter, T.C., 1995. Hydrological processes and the water budget of lakes. In Lerman, A., Imboden, D., and Gat, J., Physics and chemistry of lakes: Berlin, Springer Verlag, 37 62. Wolery et al. 1994. T hermo.dat database based on data file data0.3245r46. Reformatted t as the default. Rockware, Golden, Colorado. Woulds, C. and Ngwenya, B.T., 2004. Geochemical processes governing the performance of a constructed wetland treating acid mine drainage, Central Scotland. Applied Geochemistry 19, 1773 1783.

PAGE 218

202 Wu, M.M., Chiou, H.Y., Hs ueh, Y.M., Hong, C.T., and Su, C.L., 2006. Effect of plasma homocysteine level and urinary monomethylarsonic acid on the risk of arsenic associated carotid atherosclerosis. Toxicology and Applied Pharmacology 216, 168 175. Yehdegho, B., Rozanski, K., Zoje r, H. and Stichler, W., 1997. Interaction of dredging lakes with the adjacent groundwater field: An isotope study. Journal of Hydrology 192, 247 270. Zheng, Y., van Geen, A., Stute, M., Dhar, R., Mo, Z., Cheng, Z., Horneman, A., Gavrieli, I., Simpson, H.J ., Versteeg, R., Steckler, M., Grazioli Venier, A., Goodbred, S., Shahnewaz, M., Shamsudduha, M., Hoque, M.A., and Ahmed, K.M.. 2005, Geochemical and hydrogeological contrasts between shallow and deeper aquifers in two villages of Araihazar, Bangladesh: Im plications for deeper aquifers as drinking water sources. Geochimica et Cosmochimica Acta 69, 5203 5218. Zhou, W.F., and Beck, B.F., 2008. Management and mitigation of sinkholes on karst lands: an overview of practical applications: Environmental Geology 55, 837 851. Zobrist, J., Dowdle, P. R., Davis, J. A. and Oremland, R. S., 2000. Mobilization of arsenite by dissimilatory reduction of adsorbed arsenate. Environmental Science and Technology 34, 4747 4753.

PAGE 219

203 APPEN DIX A. AN 18 MONTH PERFORMANCE STUDY OF THE WETLAND/FILTER BASIN TREATMENT SYSTEM

PAGE 220

204

PAGE 221

205

PAGE 222

206

PAGE 223

207

PAGE 224

208

PAGE 225

209

PAGE 226

210

PAGE 227

211

PAGE 228

212

PAGE 229

213

PAGE 230

214

PAGE 231

215

PAGE 232

216 APPENDIX B. CHEMICAL AND ISOTOPI C COMPOSITION OF WAT ERS USE D FOR THE MASS BALANCE APPROACH

PAGE 233

217 Note: CP cooling pond; SA 8 and N 15 water bodies to the north and south of the wetland; WP wetland water from pump; GW groundwater (based on well ROMP 45 data from Sacks and T ihansky, 1996) ; (*) values in mg/L; m mass or percentage of each end member in the WP.

PAGE 234

218 Note: MW monitor wells; (*) values in mg/L; m mass or percentage of each end member in MW; WP wetland wat er from pump; GW groundwater (based on well ROMP 45 data from Sacks and Tihansky, 1996); SA 8 and N 15 water bodies to the north and south of the wetland.

PAGE 235

219

PAGE 236

220

PAGE 237

221

PAGE 238

222

PAGE 239

223 APPENDIX C. CH EMICAL AND MINERALOG IC AL ANALYSI S OF THE SAMPLES FROM THE SUWANNEE LIMESTONE ( IN MG/KG)

PAGE 240

224

PAGE 241

225

PAGE 242

226 APPENDIX D. CH EMICAL AND MINERALOG ICAL ANALYSI S OF THE SAMPLES FRO M THE OCALA LIMESTONE ( IN MG/KG)

PAGE 243

227

PAGE 244

228

PAGE 245

229 APPENDIX E. CHEMICAL AND MINERALOGICAL ANALYS I S OF THE SAMPLES FRO M THE AVON PARK FORMATION ( IN MG/KG)

PAGE 246

230

PAGE 247

231 APPENDIX F. DATA FROM THE BENCH SC ALE LEACHING EXPERIM ENTS INCLUDING THE COMPOSITION OF ORIGI NAL WATERS AND RECOVERED LEACHATES

PAGE 248

232

PAGE 249

233

PAGE 250

234

PAGE 251

235

PAGE 252

236

PAGE 253

237

PAGE 254

238

PAGE 255

239

PAGE 256

240

PAGE 257

241

PAGE 258

242

PAGE 259

243

PAGE 260

244

PAGE 261

245

PAGE 262

246 APPENDIX G GEOCHEMICAL R EACTIVE TRANSPORT MO DELING DATA

PAGE 263

247 (1) Water rock interaction between the aquifer matrix and surface water (Closed System) Note: reactants 0.003 times

PAGE 264

248 (2) Water rock interaction b etween the aquifer matrix and surface water (Open System) Note: reactants times 0.003 ; activity of O 2 (aq) wa s fixed

PAGE 265

249 (3) Water rock interaction between the aquifer matrix and surface water (Open System; high amount o f pyrite) Note: reactants 1 times; activity of O 2 (aq) wa s fixed

PAGE 266

250 (4) 1D reaction transport model of water rock interaction between the aquifer matrix and surface water (Closed System) Note: Reaction: from 0 hours to 5 hours Reactants 0.003 times Domain: 50 cm long and divided into 10 nodes along x Domain: 2 cm wide (along y) Discharge: 0.01 cm/sec log Permeability (darcy) = ( 5) + (15)*Porosity Dispersivity: domain length/100 Diffusion coeffi cient (cm2/sec): 1e 6 Initial porosity: 0.15

PAGE 267

251 5) 1D reaction transport model of water rock interaction between the aquifer matrix, groundwater and surface water (Closed System) Note: Reaction: from 0 hours to 5 hours Reactants 0.008 times Domain: 50 cm long and divided into 10 nodes along x Domain: 1.9 cm wide (along y) Discharge: 0.01 cm/sec log Permeability (darcy) = ( 5) + (15)*Porosity Dispersivity: domain length/100 Diffusion coefficient (cm2/sec): 1e 6 Initial porosity: 0 .15

PAGE 268

ABOUT THE AUTHOR Olesya Lazareva earned a B.S. Degree in Geology, Advanced D egree onosov Moscow State University (Moscow, Russia), a M.S. Degree in Environmental Geochemistry and Ph .D. in Geology from at the University of South Florida (Tampa, USA) During the graduate studies (2002 2010), she was a Teaching and Research A ssociate collaborating with the FIPR, SWFWMD EPA, FGS, and FDEP, and the University of Bremen. She was a L ab M anager in the Center for Water and Environmental Analysis at the University of South Florida from 2006 to 2008. Her main research i nterests include hydrogeology, environmental, aqueous, and isotope geochemistry, as well as the sustainability and environmen tal quality of water resources. She participated in the international conferences on geochemistry (i.e., Goldschmidt, AGU, and ACS ) and ha d peer reviewed publications in the Applied Ge ochemistry and Chemical Geology.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam 22 Ka 4500
controlfield tag 007 cr-bnu---uuuuu
008 s2010 flu s 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0004550
035
(OCoLC)
040
FHM
c FHM
049
FHMM
090
XX9999 (Online)
1 100
Lazareva, Olesya.
0 245
Constructed wetland/filter basin system as a prospective pre-treatment option for aquifer storage and recovery and a potential remedy for elevated arsenic
h [electronic resource] /
by Olesya Lazareva.
260
[Tampa, Fla] :
b University of South Florida,
2010.
500
Title from PDF of title page.
Document formatted into pages; contains X pages.
502
Dissertation (PHD)--University of South Florida, 2010.
504
Includes bibliographical references.
516
Text (Electronic thesis) in PDF format.
538
Mode of access: World Wide Web.
System requirements: World Wide Web browser and PDF reader.
3 520
ABSTRACT: The efficiency to improve the water quality of industrial and municipal wastewater in a constructed wetland/filter basin treatment system was investigated. The wetland system was constructed in a closed phosphate mine used for clay settling and sand tailings in Polk County, Florida. During 18-months of monitoring the chemical/microbiological composition of treated wetland water remained relatively constant, despite significant seasonal variations in temperature, rainfall and humidity. The following changes in water quality between input and output were observed: substantial decrease of water temperature (up to 10 C), reduction of As, SO4, F, Cl, NO3, NO2, Br, Na, K, Ca, and Mg, change in pH from 9 to 6.5-7, increase of H2S (up to 1060 μg/L), and a change from positive to negative ORP. There were no exceedances of the primary drinking water standards, volatile organic compounds, synthetic organic compounds, and radionuclides, but a number of exceedances for the secondary drinking water standards (Al, F, Fe, Mn, color, odor, total dissolved solids, and foaming agents). The concentration of fecal and total coliform bacteria in the wetland water was high, butsubsequently reduced during filtration in the filter basin from 30 730 and 1000 7000 count/100 mL to < 2 and < 100 count/ 100 mL, respectively. To resolve the complex hydrogeological conditions a combined isotope/chemical mass-balance approach was applied. The results were the following: (1) the composition of water in the wetland varied throughout the period of the study; (2) a change in isotopic composition along the wetland flow path; (3) the wetland contained mainly wastewater (88 100 %) during normal pumping operations; however, hurricanes and inconsistent pumping added low conductivity water directly and triggered enhanced groundwater inflow into the wetland of up to 78 %; (4) the composition of water in monitor wells was mostly groundwater dominated; however periodically seepage from a water body to the north was detected; and (5) seepage from adjacent water bodies into the wetland was not identified during operation, which would indicate a potential water loss from the wetland. To test if the wetland system could be a prospective pre-treatment option for water used in aquifer storage and recovery (ASR) scenarios, a set of bench-scale leaching experiments was carried out using rocks from the Avon Park Formation, the Suwannee Limestone and the Ocala Limestone. Since As in the Floridan Aquifer was mainly present as an impurity in the mineral pyrite the elevated iron and sulfide concentrations in the wetland water were thought to prevent pyrite dissolution. The experiments which covered a range of redox conditions showed that the amount of As released from the aquifer matrix was not perfectly correlated with the bulk rock As concentration, nor the redox state of the water. The following important results were obtained: (1) the highest concentration of As was leached from the Avon Park Formation and the lowest from the Suwannee Limestone, although the Ocala Limestone had the lowest bulk rock As; (2)minor to no As was released using native Floridan groundwater; (3) Tampa tap water, which chemically and physically resembled the ASR injection water, caused the As leaching of up to 27 μg/L, which was higher than the As drinking water standard; (4) the wetland and filter basin waters caused the highest release of As (up to 68 μg/L), which was unexpected because those water types were less oxygenated than Tampa tap water and thus should be less aggressive; (5) the in-situ filtration of the wetland water through a 0.2 μm membrane resulted in a reduction of As from 30 μg/L to 16 μg/L; and (5) the UV treatment significantly reduced both fecal and total coliform bacteria, but facilitated the increase of DO in initial waters, a change from negative to positive ORP, and the increase of As concentration in leachates. The experiments confirmed that perturbations of native aquifer conditions caused the release of As from the Floridan aquifer matrix, although the reaction may not be as simple as the dissolution of pyrite by oxygen, but additionally governed by a complex set of factors including the ORP of the system, SO42-/S2, Fe3+/Fe2+, dissolved organic carbon and microbial activity. In addition, the trend of As leaching could be governed by a set of factors, such as the porosity and permeability of the aquifer matrix influencing the rate and degree of free water saturation, amount of pyrite to be exposed to the preferential water flow paths, limited surface reactivity of pyrite with favored reactions on fractured mineral surfaces, the concentration and the selective leaching of As from individual pyrite crystals. To characterize and verify the geochemical processes in the column experiments, the Geochemist's Workbench reactive transport models (React and X1t) were developed. Results from the models correlated well to those from the column experiments andconfirmed the following: (1) the water-rock reaction between the aquifer matrix and native groundwater was favorable for pyrite stability preventing the release of As into solution; (2) the injection of oxidizing surface water into reducing native groundwater caused a change in redox potential of the system thus promoting the dissolution of pyrite, and (3) 1D reactive transport model of water-rock reaction between the aquifer matrix and surface water indicated a diverse behavior of As along the column, such as the oxidative dissolution of pyrite, mobilization and simultaneous sorption of As onto neo-formed HFO, followed by the reductive dissolution of HFO and secondary release of adsorbed As, and the potential non-oxidative dissolution of pyrite contributing the additional source of As to the solution.
590
Advisor: Thomas Pichler, Ph.D.
653
Florida
Constructed wetland
Wastewater treatment
Clay settling area
Geogenic arsenic
Isotope mass-balance
Reactive transport modeling
690
Dissertations, Academic
z USF
x Geology
Masters.
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.4550