<%BANNER%>

The Use of Calcium Carbonate Based Materials for Removing Dissolved Iron from Groundwater at Landfills

Permanent Link: http://ufdc.ufl.edu/UFE0042143/00001

Material Information

Title: The Use of Calcium Carbonate Based Materials for Removing Dissolved Iron from Groundwater at Landfills
Physical Description: 1 online resource (153 p.)
Language: english
Creator: WANG,YU
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: FERROUS -- GROUNDWATER -- IRON -- LANDFILL -- NOM -- PRB
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Elevated iron concentrations have been observed in the groundwater underneath several Florida landfill sites. An in situ method for remediating groundwater for iron, using a permeable reactive barrier comprised of calcium carbonate based materials (CCBMs), is examined as a potentially effective and economic treatment technique. Firstly, different CCBMs were tested for their ability to remove Fe (II). The removal of Fe (II) from the aqueous phase using different reactive materials (limestone, concrete, dolomite, marble, quartz sand, gypsum, and witherite) was investigated under anaerobic conditions. Limestone was found to have the best removal effectiveness, and the final Fe (II) concentration was reduced to 0.03mg/L. Based on observations from kinetic experiments, the removal process of Fe (II) by limestone appears to be a two-step process. The first step is rapid sorption of Fe (II) onto the CCBMs surfaces, and the second step is relatively slow co-precipitation of iron containing solids formed through various chemical reactions. In the second study, the effects of various environmental factors (i.e., pH, co-existing cations, and organic matter) on the removal reaction were investigated using laboratory batch studies. Solution pH has a slight effect on iron removal, with higher pH achieving better removal. Sodium and calcium have a considerable effect on the iron removal process by increasing the ionic strength of the solution. Manganese is a competing ion for iron for the adsorption sites on CCBM and can also be removed by limestone; therefore, the presence of manganese prohibits the iron removal and reduces the removal effectiveness. NOM was found to decreases Fe (II) uptake by limestone and reduce the removal effectiveness by complexing Fe (II) most likely through carboxyl group and thus mobilizing Fe (II) in the aqueous phase. The Fe (II) removal capacity of CCBMs was evaluated using a laboratory continuous flow column under anaerobic conditions. The loading capacity of limestone and crushed concrete were determined to be 4.06 g of iron per 1 kg limestone and 3.80 g of iron per 1 kg of crushed concrete. The removal process reduced the porosity of the columns by formation of precipitation on reactive materials by 3.6% and 3.4% for limestone and crushed concrete, respectively. Although porosity reduction has a negative impact on the PRB performance, the results suggest that this impact was negligible. Two pilot-scale PRBs were constructed and operated at a landfill site with elevated iron concentration. More than 95% of the Fe (II) was removed by limestone PRB, and more than 97% of Fe (II) by crushed concrete PRB. Neither PRB showed any indication of deteriorating performance over the 12-month monitoring period. Groundwater pH after passing through the limestone PRB stayed in the neutral range, while pH from the crushed concrete PRB was above 9, which may be a possible concern. In conclusion, the overall research shows that CCBMs are an effective material for PRB in removing dissolved iron from the groundwater.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by YU WANG.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Townsend, Timothy G.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-04-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042143:00001

Permanent Link: http://ufdc.ufl.edu/UFE0042143/00001

Material Information

Title: The Use of Calcium Carbonate Based Materials for Removing Dissolved Iron from Groundwater at Landfills
Physical Description: 1 online resource (153 p.)
Language: english
Creator: WANG,YU
Publisher: University of Florida
Place of Publication: Gainesville, Fla.
Publication Date: 2011

Subjects

Subjects / Keywords: FERROUS -- GROUNDWATER -- IRON -- LANDFILL -- NOM -- PRB
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre: Environmental Engineering Sciences thesis, Ph.D.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract: Elevated iron concentrations have been observed in the groundwater underneath several Florida landfill sites. An in situ method for remediating groundwater for iron, using a permeable reactive barrier comprised of calcium carbonate based materials (CCBMs), is examined as a potentially effective and economic treatment technique. Firstly, different CCBMs were tested for their ability to remove Fe (II). The removal of Fe (II) from the aqueous phase using different reactive materials (limestone, concrete, dolomite, marble, quartz sand, gypsum, and witherite) was investigated under anaerobic conditions. Limestone was found to have the best removal effectiveness, and the final Fe (II) concentration was reduced to 0.03mg/L. Based on observations from kinetic experiments, the removal process of Fe (II) by limestone appears to be a two-step process. The first step is rapid sorption of Fe (II) onto the CCBMs surfaces, and the second step is relatively slow co-precipitation of iron containing solids formed through various chemical reactions. In the second study, the effects of various environmental factors (i.e., pH, co-existing cations, and organic matter) on the removal reaction were investigated using laboratory batch studies. Solution pH has a slight effect on iron removal, with higher pH achieving better removal. Sodium and calcium have a considerable effect on the iron removal process by increasing the ionic strength of the solution. Manganese is a competing ion for iron for the adsorption sites on CCBM and can also be removed by limestone; therefore, the presence of manganese prohibits the iron removal and reduces the removal effectiveness. NOM was found to decreases Fe (II) uptake by limestone and reduce the removal effectiveness by complexing Fe (II) most likely through carboxyl group and thus mobilizing Fe (II) in the aqueous phase. The Fe (II) removal capacity of CCBMs was evaluated using a laboratory continuous flow column under anaerobic conditions. The loading capacity of limestone and crushed concrete were determined to be 4.06 g of iron per 1 kg limestone and 3.80 g of iron per 1 kg of crushed concrete. The removal process reduced the porosity of the columns by formation of precipitation on reactive materials by 3.6% and 3.4% for limestone and crushed concrete, respectively. Although porosity reduction has a negative impact on the PRB performance, the results suggest that this impact was negligible. Two pilot-scale PRBs were constructed and operated at a landfill site with elevated iron concentration. More than 95% of the Fe (II) was removed by limestone PRB, and more than 97% of Fe (II) by crushed concrete PRB. Neither PRB showed any indication of deteriorating performance over the 12-month monitoring period. Groundwater pH after passing through the limestone PRB stayed in the neutral range, while pH from the crushed concrete PRB was above 9, which may be a possible concern. In conclusion, the overall research shows that CCBMs are an effective material for PRB in removing dissolved iron from the groundwater.
General Note: In the series University of Florida Digital Collections.
General Note: Includes vita.
Bibliography: Includes bibliographical references.
Source of Description: Description based on online resource; title from PDF title page.
Source of Description: This bibliographic record is available under the Creative Commons CC0 public domain dedication. The University of Florida Libraries, as creator of this bibliographic record, has waived all rights to it worldwide under copyright law, including all related and neighboring rights, to the extent allowed by law.
Statement of Responsibility: by YU WANG.
Thesis: Thesis (Ph.D.)--University of Florida, 2011.
Local: Adviser: Townsend, Timothy G.
Electronic Access: RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-04-30

Record Information

Source Institution: UFRGP
Rights Management: Applicable rights reserved.
Classification: lcc - LD1780 2011
System ID: UFE0042143:00001


This item has the following downloads:


Full Text

PAGE 1

1 T HE USE OF CALCIUM CARBONATE BASED MATERIALS FOR REMOVING DISSOLVED IRON FROM GROUNDWATER AT LANDFILLS By YU WANG A DISSERTATION PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF TH E REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 201 1

PAGE 2

2 201 1 Yu Wang

PAGE 3

3 To my family

PAGE 4

4 ACKNOWLEDGMENTS I express my sincerest gratitude to my advisor, Dr. Timothy Townsend for his support, guidance, inspiration and humor throughout my graduate study at the University of Florida. I also thank the other members of my committee, Dr. Dean Rhue, Dr. Jean Claude Bonzongo, and Dr. Treavor Boyer, for their valuable advice, help, and support in the past several years. I thank Dr. William Harris Dr. Xinde Cao, Dr. Bin Gao and Dr. Pu Jin for their valuable advice, help, and support in my research. I would also like to acknowledge Escambia County Solid Waste Management Division and Hinkley Center for Solid and Hazardous Waste M anagement for their funding support. My thanks are extended to Sandra Jennings, Ron Hixon, and Devon Kenney for their assistance in my field pilot study. They helped me set up two permeable reactive barriers in field. Additionally, I would like to thank Dr Hwidong Kim, Dr. Jae Hac Ko, Dr. Qiyong Xu, and Dr. Brajesh Dubey for their help. I also extend my gratitude to my colleagues (Adrian Gale, Jianye Zhang, Shrawan Singh, Saraya Sikora, Jianxia Hou and Akua Oppoun) in the Solid and Hazards Waste M anagement research group for their field work help. Special thanks go to my wonderful parents for their tremendous love and support in my life. Finally, my heartfelt appreciation goes to my wife, Dr. Jie Gao, for her love, help, encouragement and accompaniment thr ough many years Without them, I could not have achieved this goal.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 8 LIST OF FIGURES ................................ ................................ ................................ .......... 9 LIST OF ABBREVIATIONS ................................ ................................ ........................... 13 ABSTRACT ................................ ................................ ................................ ................... 14 CHAPTER 1 INTRODUCTION ................................ ................................ ................................ .... 16 Background and Problem Statement ................................ ................................ ...... 16 Research Objective s ................................ ................................ ............................... 20 Research Approach ................................ ................................ ................................ 21 Organization of Dissertation ................................ ................................ .................... 22 2 COMPAR ISON OF PERMEABLE REACTIVE BARRIER MATERIALS FOR TREATING IRON CONTAMINATED GROUNDWATER ................................ ........ 25 Introduction ................................ ................................ ................................ ............. 25 Methods and Materials ................................ ................................ ............................ 27 Reactive Materials Collection and Preparation ................................ ................. 27 Removal Effectiveness of Materials ................................ ................................ 28 Reaction Kinetics of Materials ................................ ................................ .......... 28 Effects of Materials Size on Removal Effectiveness ................................ ......... 29 Chemical Anal ysis and Characterization of Removal Reaction Products ......... 29 Data Analysis ................................ ................................ ................................ ... 30 Results and Discussion ................................ ................................ ........................... 30 Removal Effectiveness of Different Materials ................................ ................... 30 Kinetics of Removal Reaction by Reactive Materials ................................ ....... 32 Effects of Materials Size on Removal Effectiveness ................................ ......... 34 Characterization of Precipitates on Reactive Material ................................ ...... 34 Summar y ................................ ................................ ................................ .......... 35 3 EFFECTS OF ENVIRONMENTAL FACTORS ON THE REMOVAL REACTION BETWEEN CALCIUM CARBONATE BAESD MATERIALS AND FE(II) ................. 48 Introducti on ................................ ................................ ................................ ............. 48 Method and Materials ................................ ................................ ............................. 49 Experimental Materials ................................ ................................ ..................... 49 Collecti on and Preparation of Natural Organic Matters ................................ .... 50

PAGE 6

6 pH Effect on Fe (II) Removal Reaction ................................ ............................. 51 Effects of Co existing Cations on Fe ( II) Removal Reaction ............................. 52 Effects of Organic Matter on Fe(II) Removal Reaction ................................ ..... 53 Data Analysis ................................ ................................ ................................ ... 53 Results and Discussion ................................ ................................ ........................... 54 Effects of pH on Fe (II) Removal Effectiveness ................................ ................ 54 Effects of Na + on F e (II) Removal Effectiveness ................................ ............... 55 Effects of Ca 2+ on Fe (II) Removal Effectiveness ................................ ............. 56 Effects of Mn 2+ on Fe (II) Removal Effectiven ess ................................ .............. 56 Effects of Natural Organic Matter on Fe (II) Removal Effectiveness ................ 57 Effects of Possible Field Conditions on Fe (II) Rem oval Effectiveness ............ 60 Summary ................................ ................................ ................................ ................ 61 4 REMOVAL OF FE (II) BY LABORATORY CALCIUM CARBONATE BASED MATERIALS PASSIVE REACTIVE COLUMN ................................ ........................ 77 Introduction ................................ ................................ ................................ ............. 77 Materials and Methods ................................ ................................ ............................ 78 Column Setup ................................ ................................ ................................ ... 78 Column Effluent Analysis ................................ ................................ .................. 79 Results and Discussion ................................ ................................ ........................... 81 Limestone Column ................................ ................................ ........................... 81 Crushed Concrete Column ................................ ................................ ............... 82 Loss of Porosity and Formation of Precipitates ................................ ................ 83 Longevity of Reactive Materials ................................ ................................ ........ 85 Summary ................................ ................................ ................................ ................ 86 5 APPLICATION OF CALCIUM CARBONATE BASED PERMEABLE REACTIVE BARRIERS FOR REMEDIATING IRON CONTAMINATED GROUNDWATER AT A LANDFILL SITE ................................ ................................ ........................... 100 Introduction ................................ ................................ ................................ ........... 100 Method and Materials ................................ ................................ ........................... 101 Site Description ................................ ................................ .............................. 101 Construction of Permeable Reactive Barriers and Monitoring Network .......... 102 Groundwater Sampling and Analysis ................................ ............................. 103 Bromide Tracer Test ................................ ................................ ....................... 103 Results and Discussions ................................ ................................ ....................... 104 Hydraulic Conditions of the PRB Area ................................ ............................ 104 Tracer Test ................................ ................................ ................................ ..... 108 Performances of PRBs ................................ ................................ ................... 109 Summary ................................ ................................ ................................ .............. 110 6 CONCLUSION ................................ ................................ ................................ ...... 121 Summary ................................ ................................ ................................ .............. 121 Conclusion ................................ ................................ ................................ ............ 12 3

PAGE 7

7 Future Work ................................ ................................ ................................ .......... 124 APPENDIX A ADDITIONAL PRBS INFORMATION ................................ ................................ .... 125 B ADDITIONAL GROUNDWATER DATA AND FIGURES ................................ ....... 127 LIST OF REFERENCES ................................ ................................ ............................. 146 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 153

PAGE 8

8 LIST OF TABLES Table page 2 1 Size and density of all tested reactive materials ................................ ................. 37 2 2 Metal cont ents and pH of all tested reactive materials ................................ ....... 38 2 3 First order reaction constants of removal reactions for reactive materials .......... 39 3 2 Linear regression equation and correlation coefficient between NOM concentration and final Fe (II) concentration ................................ ...................... 63 3 3 Physicochemical properties of natural organic matters (after ultrafiltra tion) ....... 64 4 1 Characteristics of reactive materials ................................ ................................ ... 87 4 2 Calcium and iron contents in the precipitates formed on reactive media surf ace. ................................ ................................ ................................ ............... 88 4 3 Porosity of columns before and after test (Limestone: columns a and b; Crushed concrete: columns c and d) ................................ ................................ .. 89 5 1 Groundw ater parameters in PRB area (April 2010) ................................ .......... 111 A 1 Details of groundwater monitoring wells installed for PRBs .............................. 126

PAGE 9

9 LIST OF FIGURES Figure page 1 1 Remediation of Fe (II) contaminant using a CCBM PRB. ................................ ... 24 2 1 Final Fe (II) concentrations and pH values of the mixtures of synthetic groundwater and reactive materials after 72 hours treatment. ........................... 40 2 2 Fe (II) concentrations measured from the mixture of synthetic groundwater and reactive materials over time. ................................ ................................ ....... 41 2 3 Fe (II) concentrations measured from the mixture of syntheti c groundwater and reactive materials in the first hour. ................................ ............................... 42 2 4 Fe (II) concen trations measured from the mixture of synthetic groundwater and reactive materials in the first hour (Logarithmic C vs. time). ........................ 43 2 5 Final Fe (II) concentrations and pH values for the mixture of synthetic groundwater and different sized reactive materials ................................ ............ 44 2 6 Change of Fe (II), calcium, and carbonate concentrations and pH values over time in limestone and synthetic groundwater system. ................................ ........ 45 2 7 SEM graphs of limestone surface before and after treatment. ........................... 46 2 8 XRD result of precipitates from the reaction between synthetic groundwater and lim estone. ................................ ................................ ................................ ... 47 3 1 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone under different pH conditions for 12 hours. ................................ ........ 65 3 2 Fe (II) concentrations measured over time in the mixture of synthetic groundwater and limestone under different pH conditions. ................................ 66 3 3 Fe (II) concentrations measured i n the mixture of synthetic groundwater and Limestone with different sodium concentrations for 12 hours. ............................ 67 3 4 Fe (II) concentrations measured over time in the mixture of synthetic groundwater and limestone with and without sodium ions. ................................ 68 3 5 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone with different calcium concentrations for 12 hours. ............................ 69 3 6 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone with different Mn concentrations for 12 hours ................................ ..... 70 3 7 Fe (II) concentrations measured over time inthe mixture of synthetic groundwater and limestone with and without Mn 2+ ................................ ........... 71

PAGE 10

10 3 8 Final Fe (II) concentrations after removal by limestone with different c oncentration of NOM (10 g limestone + 200 mL 50 mg/L Fe (II) + 10 mg/L C NOM, pH = 7) ................................ ................................ ................................ ..... 72 3 9 Fe (II) concentrations measur ed over time in the mixture of synthetic groundwater and limestone with and without NOMs ................................ ........... 73 3 10 Final Fe (II) concentrations after removal by limestone with different concentration of NOM (10 g limes tone + 200 mL 50 mg/L Fe (II) + 5, or 10, or 20 mg/L C NOM, pH = 7) ................................ ................................ .................... 74 3 11 Charge densities of three NOMs calculated by the titration result. The carge density at pH 7 and carboxyl acidity are estimated from these curves. .............. 75 3 12 C harge densities of three NOMs vs. Fe (II) concentration. ................................ 76 4 1 Schematic diagram of the experimental column setup. Water flows from bottom to top. ................................ ................................ ................................ ...... 90 4 2 Performance of limestone in column test: Fe (II) concentration versus pore volume. (Column a and b) The initial Fe (II) concentration in synthetic groundwater is 50 mg/L. ................................ ................................ ..................... 91 4 3 Performance of limestone in column test: pH versus pore volume. (Column a and b) pH drop to 6.4 after 290 pore volumes. ................................ ................... 92 4 4 Performance of limestone: Ca 2+ versus pore volume (a) and di stance (b). Calcium concentration drops with pore volume increasing and port distance (from the entrance) increasing. ................................ ................................ ........... 93 4 5 Performance of limestone: total carbonate concentration versus pore v olume (a) and distance (b). Total carbonate concentration drops with pore volume increasing and port distance (from the entrance) increasing. ............................. 94 4 6 Performance of crushed concrete column: Fe (I I) concentration versus pore volume. The initial Fe (II) concentration in synthetic groundwater is 50 mg/L. ... 95 4 7 Performance of crushed concrete column: pH versus pore volume. .................. 96 4 8 Correlation between pH and Fe (II) concentration in effluent samples ............... 97 4 9 Performance of crushed concrete: carbonate concentration versus p ore volume (a) and distance (b). Total carbonate concentration drops with pore volume increasing and port distance (from the entrance) increasing. ................. 98 4 10 Performance of crushed concrete: Ca 2+ versus pore volume (a) and distance (b). ................................ ................................ ................................ ...................... 99

PAGE 11

11 5 1 PRBs area map, showing locations of monitoring wells, existing monitoring wells (DW4S and MW2), and two PRBs location. ................................ ............. 112 5 2 Change in groundwater table elevation (above sea surface level) and precipitation over time ................................ ................................ ...................... 113 5 3 Difference of groundwater table elevation (Above sea surface level) in Limestone PRB (upper) and Crushed concrete PRB (lower). ........................... 114 5 4 Change of Fe (II) concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The Fe (II) concentration decreased after passing through PRBs.. ................................ ................................ ................................ 115 5 5 Change of total iron concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The total iron concentration decreased after passing through PRBs. ................................ ................................ ..................... 116 5 6 Change of Fe (II) concentration in Well DW4S and MW2 over time. DW4S was affected by PRBs since the Fe (II) concentration dropped. ....................... 117 5 7 Change of pH in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The pH increased after passing through PRBs. ....................... 118 5 8 Change of Mn concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The Mn concentratio n decreased after passing through PRBs. ................................ ................................ ................................ .. 119 5 9 Maximum bromide concentration observed in each monitoring wells (in mg/L. Numbers in parentheses are maximum Br concentration) Groundwater flow pa ssed through PRBs. ................................ ................................ ...................... 120 A 1 Location of active groundwater monitoring points at the Klondike landfill. (PRBs area was marked as EA 1 in map) ................................ ........................ 125 B 1 Change of dissolved oxygen level over time in background wells (Limestone PRB) ................................ ................................ ................................ ................. 127 B 2 Change of dissolved oxygen level over time in down gradient wells (Limestone PRB) ................................ ................................ .............................. 128 B 3 Change of dissolved oxygen level over time in PRB wells (Limestone PRB) ... 129 B 4 Change of dissolved oxygen level over time in backgro und wells (Crushed concrete PRB) ................................ ................................ ................................ .. 130 B 5 Change of dissolved oxygen level over time in down gradient wells ( Crushed concrete PRB) ................................ ................................ ................................ .. 131

PAGE 12

12 B 6 Change of dissolved oxygen level over time in PRB wells ( Crushed concrete PRB) ................................ ................................ ................................ ................. 132 B 7 Change of pxidation reduction potential (ORP) over time in background wells (Limestone PRB) ................................ ................................ .............................. 133 B 8 Change of oxidation reduction potential (ORP) over time in down gradient wells (Limestone PRB) ................................ ................................ ..................... 134 B 9 Change of oxidation reduction potential (ORP) over time in PRB wells (Limestone PRB) ................................ ................................ .............................. 135 B 10 Change of oxidation reduction potential (ORP) over time in background wells ( Crushed concrete PRB) ................................ ................................ .................. 136 B 11 Change of oxidation reduction potential (ORP) over time in down gradient wells ( Crushed concrete PRB) ................................ ................................ .......... 137 B 12 Change of oxidation reduction potential (OR P) over time in PRB wells ( Crushed concrete PRB) ................................ ................................ .................. 138 B 13 Change of conductivity over time in background wells (Limestone PRB) ......... 139 B 14 Change of conductivity over time in down gradient wells (Limestone PRB) ..... 140 B 15 Change of conductivity over time in PRB wells (Limestone PRB) .................... 141 B 16 Change of conductivity over time in background wells ( Crushed concrete PRB) ................................ ................................ ................................ ................. 142 B 17 Change of conductivity over time in down gradient wells ( Crushed concrete PRB) ................................ ................................ ................................ ................. 143 B 18 Change of conductivity over time in PRB wells ( Crushed concrete PRB) ......... 144 B 19 Change of bromide tracer concentration over time in monitored wells. (a) limestone PRB area and (b) crushed concrete PRB area. ............................... 145

PAGE 13

13 LIST OF ABBREVIATION S AMD Acidic mining d rainage CCBM Calcium carbonate based materials CKGO Cedar Key g roundwater organic matter D 10 Eff ective size, 10 percent of the sample is finer than this size GWCTL Groundwater cleanup target level NRRL New River Regional Landfill leachate organic matter PRB Permeable reactive barrier MSW Municipal solid waste NOM Natural Organic M atter SRHA Suwannee River Humic A cid TO C Total Organic C arbon XRD X R ay diffraction ZP Zeta Potential

PAGE 14

14 Abstract of Dissertation Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Docto r of Philosophy THE USE OF CALCIUM CARBONATE BASED MATERIALS FOR REMOVING DISSOLVED IRON FROM GROUNDWATER AT LANDFILLS By YU WANG May 201 1 Chair: Timothy Townsend Major: Environmental Engineering Elevated iron concentrations have been observed in the groundwater underneath several Florida landfill sites. An in situ method for remediating groundwater for iron, using a permeable reactive barrier compris ed of calcium carbonate based materials (CCBMs), is examined as a potentially effective and economic tr eatment technique. Firstly, different CCBMs were tested for their ability to remove Fe (II). The removal of Fe (II) from the aqueous phase using different reactive materials (limestone, concrete, dolomite, marble, quartz sand, gypsum, and witherite) was i nvestigated under anaerobic conditions. Limestone was found to have the best removal effectiveness and the final Fe (II) concentration was reduced to 0.03mg/L Based on observation s from kinetic experiments, the removal process of Fe (II) by limestone app ears to be a two step process The first step is rapid sorption of Fe (II) onto the CCBMs surfaces and the second step is relative ly slow co precipitation of iron containing solids formed through various chemical reactions In the second study, the effect s of various environmental factors (i.e., pH, co existing cations, and organic matter) on the removal reaction were investigated using laboratory batch studies. Solution pH has a slight effect on iron removal, with higher pH

PAGE 15

15 achieving better removal. Sodiu m and calcium have a considerable effect on the iron removal process by increasing the ionic strength of the solution. Manganese is a competing ion for iron for the adsorption sites on CCBM and can also be removed by limestone ; t herefore, the presence of m anganese prohibits the iron removal and reduces the removal effectiveness. NOM was found to decreases Fe (II) uptake by limestone and reduce the removal effectiveness by complexing Fe (II) most likely through carboxyl group and thus mobilizing Fe (II) in the aqueous phase. The Fe (II) removal capacity of CCBMs was evaluated using a laboratory continuous flow column under anaerobic conditions. The loading capacity of l imestone and crushed concrete were determined to be 4.06 g of iron per 1 kg limestone an d 3.80 g of iron per 1 kg of crushed concrete. The removal process reduced the porosity of the columns by formation of precipitation on reactive materials by 3.6% and 3.4% for limestone and crushed concrete, respectively. Although porosity reduction has a negative impact on the PRB performance, the results suggest that this impact was negligible T wo pilot scale PRBs were constructed and operated at a landfill site with elevated iron concentration More than 95% of the Fe (II) was removed by limestone PRB, and more than 97% of Fe (II) by crushed concrete PRB. Neither PRB showed an y indication of de teriorating performance over the 12 month monitoring period. G roundwater pH after pass ing through the limestone PRB stayed in the neutral range, while pH from the crushed concr ete PRB was above 9, which may be a possible concern In conclusion, the overall research shows that CCBMs are an effective material for PRB in removing dissolved iron from the groundwater.

PAGE 16

16 CHAPTER 1 INTRODUCTION Background and Problem Statem ent Elevated iron concentrations have been observed in groundwat er monitoring wells at several municipal solid w aste (MSW) landfill s in Florida. I mpacted wells have been identified at active lined MSW landfill u nits and closed unlined cells. I ron concentra tions in these monitoring wells have exceeded both Groundwater Cleanup Target Level (G W CTL) of 0.3 mg/L and a health based criterion of 4.2 mg/L often used by the Florida Department of Environment Protection (FDEP) (Gadagbui and Roberts, 2002). I ron contamination h as raised concern in these areas because high iron containing groundwater c an impair surface water resources and ecosystems. In addition, elevated iron concentrations in well water for human consumption can result in problems suc h as iro n precipitation and water rust in the water pipe (Parisio et al 2006) High iron concentration also has toxicity to humans and other animals and aqueous plants and can even cause death (Lepp, 1957). A lethal dose for adult is 100 g, and poisoning occur w ith an intake of 200 300 mg/kg body weight of child (Reilly, 2002). Landfill may affect the groundwater and vadose zone and contaminate groundwater. Heavy metals contamination was found in u nlined landfills sites due to the landfill leachate problem sin ce 1960s (Christensen et al. 2001). L ined l andfill s can lead to decreased oxygen levels in underlying aquifer, potentially causing anaerobic conditions which are favorable for s oil iron reductive dissolution to occur (Ponnamperuma, 1972; Lovley, 1991; We ber et al. 2006) In soil iron reductive dissolution, Fe (III) serves as an electron accepter for microorganisms deriving energy from organic chemicals (Lovley et al. 2004), resulting in Fe (II) formation from this

PAGE 17

17 reaction. In anaerobic groundwater, solu ble Fe (II) is the ma jor iron form present in the aqueous phase (Lovley, 1991) and it has also been found to be the dominant species released from nearby landfill soils under anaerobic conditions (d eLemos et al., 2006) Currently, the most commonly used m ethod for iron remediation in groundwater is ere contaminated groundwater is pumped from the aquifer and treated ex situ. Existing techniques for iron removal from ground water include aeration, softening, chlorination, ozonation, and fi ltration (Ellis et al., 2000) and aeration is the most popular one. In the aeration process, dissolved Fe (II) is oxidized and precipitates to form iron oxides (Mettler et al., 2001) which can then be removed via physical separation (Ellis et al., 2000) H owever, this technique can be costly and energy intensive. This motivated research is aiming at evaluating an alternative approach where in situ reactive media (i.e., a reactive wall built across the flow direction) is used for its ability to reduce dissol ved iron concentration C alcium carbonate based mineral (CCBM) is common ly used material in drinking water treatment (Tegethoff, 2001) and has been used more recently to treat groundwater ( Cravotta and Trahan, 1999; Mettler et al., 2001 ; Turner et al. 200 8 ) The use of CCBM for removing metals from drinking water and industrial wastewater has been found effective (Aziz and Smith 1992; Aziz et al., 2001) For example, CCBM has been used in a passive reactive system for acidic mine drainage (AMD) treatment in field (Sterner 1997; Indraratna et al., 2010) In th ese passive rea ctive systems, CCBM increase s the pH of AMD and form s precipitates to remove metals. Also, CCBM has the ability to adsorb and reacting with divalent metals (Aziz and Smith 1996) Sever al studies have reported the effective remov al of copper and manganese by CCBMs (Aziz

PAGE 18

18 et al. 2001; Aziz and Smith 1992; Aziz and Smith 1996; Thornton 1995) Aziz et al. (2001) further predicted that iron could also be removed using such a system T his technique is advantageous because CCBMs are inexpensive ( e.g., limestone is generally less than $50 per ton) compared to many other materials Therefore, in our study, CCBMs were used as the reactive material s A few studies have discusse d the reaction be tween Fe (II) and CCBM Wajon et al. (1985) speculated that Fe (II) adsorption and/or co precipitation on calcite took place on the calcite surface and calcium siderite CaFe(CO 3 ) 2 formation was found after the removal reaction C alcium siderite is a minera l which a calcium ion substitute the place of a ferrous ion. S iderite (a byproduct of the reaction between Fe (II) and carbonate) has a low solubility product (Ksp=3.2x10 11 ) and thus can potentially form the precipitates Possible reaction s between Fe (II ) and calcium carbonate can be summarized as follows: CaCO 3 Ca 2+ +CO 3 2 (1 1 ) Fe 2+ +CO 3 2 FeCO 3 ( 1 2) Fe 2+ + Ca 2+ + 2CO 3 2 CaFe (CO 3 ) 2 ( 1 3) Calcium carbonate can dissolve in water and release calcium an d carbonate ions (Reaction 1 1 ). Fe (II) alone can react with carbonate to form siderite (Reaction 1 2) or Fe (II) and calcium ions can react with carbonate to form calcium siderite (Reaction 1 3). Recently, Mettler et al. (2009) investigated the sorptio n mechanism of Fe (II) on the calcite surface (the most stable polymorph mineral of calcium carbonate), and revealed that the reaction between Fe (II) and calcite has two kinetically distinct stages, a fast initial step (less than one hour), followed by a slow uptake step (several days). The first

PAGE 19

19 step was later described as adsorption and the second step as incorporation by re crystallization or co pr ecipitation (Mettler et al., 20 0 9 ). Precipitates formed in the reaction between Fe (II) and calcium carbona te material can cause reduction in removal effectiveness and hydraulic conductivity as a result of the clogging of pore spaces (Mackenzie et al., 1999), which can be assessed by conducting column tests (Bilek et al. 2006, Kamolpornwijit et al 2003). Colu mn tests have been conducted with different materials for the removal of a wide variety of contaminants, including Zn and Mn (Komnitsas et al., 2004, Dikinya et al., 2008, Darbi et al., 2003, Golab et al., 2009). For example, the porosity of zero valent ir on column was reduced more than 50% by minerals formation (Bilek et al., 2006), and landfill leachate passing through soil column can reduce a 0.5 to 6.9 % of hydraulic conductivity due to the organic byproduct formation on soil particle surface ( Islam and Singhal, 2004 ). Therefore, laboratory column studies would be useful for examining the removal effectiveness of reactive material, the potential impact on media porosity, and the longevity of the media bed. P ermeable reactive barrier s (PRB) are consider ed one of the more effective technique s to remediate contaminated grou ndwater since it is an in situ approach (DOD, 2002; EPA, 2001). Figure 1 1 is a conceptual example of CCBM PRB Hundreds of PRBs have been constructed in N orth America to remediate groun dwater contamination (Henderson and Demond 2007) Different reactive materials have been used in PRB construction, including zero valent iron powder s (remove p olychlorinated organic compounds ) and iron oxides (remove arsenic and other heavy metals) (Hende rson and Demond 2007; Komnitsas et al. 2007). CCBM s have also been used

PAGE 20

20 in a few PRB sites for removal of various groundwater contaminants (Hedin et al. 1994) Limestone and crushed concretes media were used in an AMD PRB field application, and more tha n 90% of metals were removed by both materials. As such, considering the low cost of CCBM and its potential to precipitate iron, using a passive reactive media containing CCBM could be an ideal alternative technique for efficient and cost effective iron re moval which is critically needed. However, until now, such a system has never been used on remediation of Fe (II). The effectiveness of CCBMs for iron removal and potential interaction of Fe (II) with CCBMs remains to be understood. To better design the g roundwater treatment system and achieve the optimal treatment conditions, understanding the reaction mechanism of the reactions between Fe (II) and CCBM is very important since removal mechanism can provide necessary parameters and information for the fiel d installation and operation. Environmental factors (e.g., pH, co existing cations, and natural organic matter) may affect the aquatic reactions, so further investigation of these factors is also required. Research Objectives The overall purpose of this r esearch project is to evaluate the effectiveness of Fe (II) removal from groundwater using calcium carbonate (CaCO 3 ) and similar minerals (CCBM) and to provide information and parameters to design passive reactive barriers for implementation in the field. The first objective of this study wa s to compare the removal effectiveness of iron in the aqueous phase using eight different passive reactive materials in terms of t he reaction rate s, removal effectiveness, and removal mechanism. The second objective wa s to assess the effects of various environmental factors (i.e. pH, co existing cations, and organic matter) on the removal process

PAGE 21

21 The third objective wa s to examine the Fe (II) removal capacity of CCBM (limestone and crushed concrete), including the lo ading capacity of CCBMs and reduction in the porosity of packed reactive media induced by the formation of precipitates (reaction products). The fourth objective wa s to investigate the overall performance and treatment effectiveness of CCBMs comprised perm eable reactive barrier s in removing Fe (II) from iron contaminated groundwater in a small pilot field study. Research Approach Iron removal effectiveness in the aqueous phase using different passive reactive materials was compared in batch tests. Limeston e, crushed concrete, marble, witherite, dolomite, gypsum and quartz sand were used to investigate the reaction kinetics and final Fe (II) removal effectiveness. The reactions between reactive materials and Fe(II) solution were conducted in a serial reactio n time. All the filtrates of treated solution were analyzed for Fe (II) concentration, carbonate concentration, and pH values. The influence of various environmental factors on the Fe (II) removal reaction were investigated in batch tests. The tests were c onducted at pH condition of 4, 5, 6, 7, 8 and 9 The final Fe (II) concentration in aqueous phase and pH were recorded. Different sodium, manganese, and calcium concentration s were used in the study of effects of co existing ions The impacts of t hree dif ferent types of natural organi c matters (NOM) were examined including s urface water natural organic matter landfill leachate organic matter and groundwater natu ral organic matter The removal reaction between limestone and Fe (II) was studied with diffe rent concentration s of NOMs, and the final Fe (II) concentrations were determined to evaluate the removal effectiveness.

PAGE 22

22 Continuous flow column tests were performed to determine the total removal capacity of limestone and crushed concrete Fifty mg/L Fe ( II) solution was used as the synthetic groundwater and continuously injected into limestone or crus hed concrete packed columns. Fe (II) concentration and pH in the leachates were monitored during the whole tests. After the column tests, the reduction in co lumn porosity was estimated by gravity method In the field study, two PRBs (one limestone and the other recycled crushed concrete) were constructed at a closed unlined landfill exhibiting elevated iron c oncentration in groundwater. The PRBs were placed p erpendicular to groundwater flow. Twenty eight groundwater monitoring wells were installed up gradient of the PRB, down gradient of the PRB, and with in PRBs. Groundwater quality was analyzed throughout to evaluate the PRB performance. Organization of Diss ertation This PhD dissertation is presented in 6 chapters, including the present introductory chapter (Chapter 1). Chapter 2 evaluates the removal eff ectiveness of iron in the aqueous phase using different passive reactive materials under anaerobic conditi ons. The material with the best performance was used in subsequent studies. Chapter 3 assesses the effects of various environmental factors including pH, co existing cations, and natural organic matter, on the Fe (II) removal effectiveness by calcium carb onate Chapter 4 focuses on the Fe (II) removal capacity of CCBMs and investigates the reduction of porosity caused by the reaction products Chapter 5 studies the treatment eff ectiveness of two CCBMs comprising PRBs in removing Fe (II) from contaminated g roundwater in the field Chap ter 6 summarizes the findings, results, and future works from all the above research experiments. The literat ure cited as references are included

PAGE 23

23 at the end of this document. Appendix A and B contain additional information and figures of PRB research for C hapters 5

PAGE 24

24 Figure 1 1 R emediation of Fe (II) c ontaminant using a CCBM PRB Fe (III) is reduced to Fe (II) underneath landfill, and the dissolved Fe (II) can transport with groundwater flow. CCBM PRB can remove Fe (II) from the contaminated groundwater.

PAGE 25

25 CHAPTER 2 COMPARISON OF PERMEA BLE REACTIVE BARRIER MATERIALS FOR TREATI NG IRON CONTAMINATED GR OUNDWATER Introduction Human activities and natural phenomena have caused releases of Fe (II) into groundwater and surface water, creating environmental problems for humans and other living organisms (Rao et al. 2008). Iron c an enter surface water systems to form iron oxide precipitates which is a n issue of concern in drinking water treatmen t (Das et al., 2007) In U.S, the guideline level of iron in drinking water should be less than 0.3 mg/L (secondary drinking water standard) which is equal to the Groundwater Cleanup Target Level (GWCTL) suggested by US EPA. In anaerobic aquifer, iron oxides and other iron containing mine rals tend to be reduced from positive three charge state (Fe(III), ferric iron) to the positive two charge state (Fe(II), ferrous iron) (Lepp 1975). This reduction process is usually associated with microorganism activities (Lovley 1996). Compared to th e insoluble iron oxide minerals, Fe (II) is soluble and mobile and thus poses more risks to the surrounding environments and living organisms. Although le ss common than most other metal recently, elevated dissolved iron levels (> 20 mg/L) were obser ved in north Florida areas. E xcessive iron concentration in water can be harmful to humans and an effective and efficient technique is highly desired to solve this iron contamination problem. Conventional iron removal from groundwater is achieved by aeration fo llowed by filtration (Mettler et al. 2001, Das et al. 2007). Other ex situ treatment techniques have been developed to remove iron rich groundwater and reasonable removal effectiveness (more than 90% iron) has been achieved, such as adsorption oxidation, b io oxidation,

PAGE 26

26 and oxidation floc formation (Munter et al. 2005; Munter et al. 2008). However, all these techniques have the same problem: high cost due to the pumping and external treatment process. Therefore, an economical alternative is proposed herein using a permeable reactive barrier (PRB) for in situ removal of Fe (II). The selection of reactive material is very important. A good reactive material candidate should be cheap, effective, and good for long term use and environmental friendly (Golab et a l. 2009). A number of studies have investigated Fe (II) removal by different reactive media, including calcium carbonate based materials (CCBMs) (Das et al ., 2007, Sharma et al ., 1999). Sharma et al. (1999) found that limestone could effectively remove mo re than 90% of Fe (II) from groundwater, and Mettler et al. (2002) reported that calcite can remove more than 92% of Fe (II) from synthetic groundwater under anaerobic condition. These results suggest that CCBM s have the potential to remove Fe (II) as part of a PRB system. Moreover, CCBMs are substantially cheaper than many other reactive materials. However, there remains a lot to be understood for the Fe (II) removal by CCBMs, including the removal mechanisms, precipitation product, impacts of environmenta l factors, loading capacity, and most importantly, the in situ performance of the CCBMs comprising PRBs. These are the main objectives of this and the following studies. This chapter pertains to the comparison of the removal e ffectiveness of different pass ive reactive materials ( e.g. limestone, concrete, dolomite, quartz sand, gypsum, and witherite) for Fe (II) in the aqueous phase under anaerobic conditions

PAGE 27

27 Methods and Materials Reactive Materials Collection and Preparation In this study, a total of eigh t different materials were assessed. Two CCBM rocks (limestone and marble), two sources of recycled crushed concrete since crushed concrete have calcium carbonate content two carbonate minerals (dolomite and witherite), and two control materials (gypsum a nd quartz sand) were included. Limestone (Gainesville, FL, USA) and m arble (Wards Sci, Rochester, NY) represent CCBM Two recycled c rushed concrete samples wer e collected from two recycling facilities (Escambia Recycle Inc. (marked as concrete A), and Per dido landfill (marked as concrete B)). Dolomite (Wards Sci, Rochester, NY) and w itherite (BaCO 3 ) (Flabster Inc. UK) represent other minerals containing carbonate. G ypsum (Wards Sci, Rochester, NY) is a mineral with calcium ion in its structure withour car bonate. Q uartz sand (Fisher Science Inc Atlant a, GA, USA) served as a control (40 80 mesh). All minerals were crushed prior to use. All materials except quartz sand with a particle range 7.5 to 15 mm were used for the removal effectiveness test Table 2 1 presents characterization of all the materials. For limestone and crushed concrete A (they are better removal effectiveness materials) a n additional particle size effects test were conducted with series of sizes. The particles were separated in the foll owing size ranges: 25 mm~50 mm, 15mm~25 mm, 7.5 mm~15 mm, and 3 mm~7.5 mm. Prior to the experiment, all materials were washed with DI water, soaked in diluted nitric acid (pH=1) for 8 hours, rinsed again with DI water, and dried at 40 C. A ll materials wer e digested following digestion method ( EPA 3050B ) using concentrated trace metal grade nitric acid. Metal concentration s (Fe, Ca, Ba, Mn, Mg) w ere measured by ICP AES ( Thermo Electron Corporation, Trace Analyzer ) as per USEPA SW 846

PAGE 28

28 Method 6010B A ll the m aterials were measured for pH using EPA 9045 method which requires measuring 1:2 solid/water mixtures (100 g materials and 200 mL DI water) Removal Effectiveness of Materials A laboratory b atch experiment was conducted to test the removal effectiveness of Fe (II) The majority of the experiment ( with except ing of rotation and shaking part) was run in an anaerobic chamber, which is operated with 99.9% nitrogen gas, to keep an anoxic condition. A synthetic groundwater with 50 mg/L Fe (II) prepared by mi xing ferrous chloride (FeCl 2 Fisher Sci Inc.) with DI water (pre bubbled 24 hours with 99.9% nitrogen gas), was used for the experiment. The e xperimental solid water ratio was fixed at 1/20 (S olid/Liquid ratio ). Fifty g of each material and 1000 mL Fe Cl 2 solution containing was 50 mg/L Fe (II) were added together into a 1L bott le in the glove box The mixture was tumbled on a rotator (outside of the glove box) for 12 hours at 30 RPM. The mixture was then filtered through 0.45 m membrane (nylon 0.45 m filter paper, Fisher Sci. Inc.) in the glove box. Fe (II) concentrations were measured using the Hach spectrometer col orimetric method (Hach program 2150). Where dissolved Fe (II) ion reacts with 1, 10 phenanthroline forming a colored complex with the co lor change is proportional to the Fe (II) concentration. pH values were measured by portable pH meter (Acuma 209, Fisher Sci Inc.). During the test, Dissolved oxygen (DO) concentration was randomly monitored, and 63 samples were measured for D O Reaction Kinetics of Materials Six of the materials were used in the kinetics study since quartz sand and gypsum were found to have little removal effectiveness. 50 g of each material and 1000 mL 50 mg/L Fe (II) of FeCl 2 solution were added into a 1L bottle in the glove box. Seven

PAGE 29

29 batches were prepared for each material and shaken on a rotator at 30 RPM for seven different reaction time ( 10 min, 30 min, 1 hou r, 6 hours, 12 hours, 24 hours, and 72 hours ) in a glove box Each batch contains three replicates and each r eplicate was prepared by mixing 50 g of the material and 1000 mL 50 mg/L Fe (II) of FeCl 2 solution into a 1L bottle The samples were then filtered through a nylon 0.45 m filter paper. Fe (II) concentration and pH of the filtrate were measured for crushed concrete samples. Fe (II) concentration calcium concentration, carbonate concentration, and pH of the filtrate were measured for limestone samples. For carbonate analysis, the preserved samples were analyzed by a TOC analyzer (Tekmar Dorhman Phoenix 8000 TOC ( Total Organic Carbon) Analyzer) and the inorganic carbon content is determined by subtracting the total organic carbon content from the total carbon content. Calcium concentration s w ere measured by ICP AES ( Thermo Electron Corporation, Trace Analyzer ) as per USEPA SW 846 Method 6010B. Effects of Materials S ize on Removal Effectiveness Batch experiments using these materials with preselected sizes were conducted as described in the above section. A similar batch method with the removal effectiveness wa s used to test the removal effectiveness. Fe (II) concentration and pH of the filtrate were measured Chemical Analysis and Characterization of Removal Reaction Products The inorganic carbon content is determined by subtracting the total organic carbon con tent from the total carbon content. Calcium concentration was measured by ICP AES as per USEPA SW 846 Method 6010B. pH values were measured by pH meter (Acuma 209, Fisher Sci Inc.). The solid samples were examined by X ray diffraction ( XRD ) to study the pr operties of the precipitates formed on limestone and crushed

PAGE 30

30 concrete particle surface. T he zeta potential (ZP) of limestone with pH 6, 7, and 8 was measured using the EKA electrokinetic analyzer (Anton Paar). The EKA is used to determine the ZP of larger (noncolloidal) particles and is based on the streaming potential method. More detailed information about this methodology can be found at report by Bismarck ( 2004). Data Analysis All the experiments were run in triplicate, and data were illustrated as mean standard deviation (STD). Statistical analyses were performed using one way analysis p value of less than 0.05 was considered to be statistically significant. Results and Discussion Removal Effectiveness of Different Materials During the test, an average of 0.74 0.45 mg/L dissolved oxygen concentration was monitored. This average DO concentration is lower than the regular DO concentration in contaminated site (1.24 0.47 mg/L DO in Klondike landfill). Since the main purpose is the simulation of the groundwater condition s the relative low DO will not be considered as a factor. And in the control blank test, a n average 47.2 2.4 mg/L Fe (II) was recovered for 23 samples (in itial Fe (II) was 50 mg/L) This high recovery indicates the dissolved oxygen has less impact on the test. Table 2 2 summarizes the composition information of the tested materials. Final Fe (II) concentrations and pH values after treatment are shown in Fig ure 2 1. Final Fe (II) concentrations were significantly higher ( p <0. 0 01 ) for gypsum and quartz sand whereas gypsum was higher than quartz sand (p < 0.001) Although not significantly different, t he lowest Fe (II) concentrations were observed for limesto ne and crushed

PAGE 31

31 concrete treated samples. The Fe (II) concentrations of limestone treated samples after 24 hours were close to 0.03 mg/L, which is much below the GWCTL (0.3 mg/L). The pH values of limestone treated samples were close to 7, which is neutral and close to natural groundwater 6 8.5 (FDEP, 1992 ). This pH value is lower than the pH result groundwater for 28 days. The Fe (II) concentrations of crushed concrete treate d samples were approximately 0.05 mg/L, and pH values were around 9, which is also lower than Golab et al. (2006) reported pH value for concrete leachate (pH 10) The reason for higher pH in concrete treated samples is that crushed concrete is an alkaline mixture and contains calcium oxides and calcium hydroxides components. Compared to the 28 day reaction period in Golab et al. (2006) study, the reaction duration in the current study is much shorter, up to 72 hours, and this may explain the lower pH valu es in this work. After the 24 hour batch study, the final Fe (II) concentrations of Marble, dolomite, and witherite treated samples were determined to be 0.162 0.06, 0.231 0.05, 0.261 0.07 mg/L, respectively. The results suggest that these carbonate min erals can be used to treat iron contaminated groundwater; however, a barium concentration of 10.5 mg/L was noticed in witherite treated samples. As high barium level is a potential environmental concern (MCL of barium is 2 mg/L) (EPA, 2010) witherite was eliminated from further study. The final Fe (II) concentrations of gypsum and quartz sand treated samples were 31.2 and 17.2 mg/L, respectively which were significantly higher than other materials. This result is consistent with the study by Sharma et al. (1999), who found that quartz

PAGE 32

32 sand and gypsum have little cation sorption ability, and are not appropriate for long term treatment. Therefore, gypsum and quartz sand were not selected for further studies. Besides the use of CCBMs, a number of papers have reported Fe (II) sorption by other minerals (Luther et al. 1996; Nano and S trathmann, 2006 ; Willians et al. 2005). For example, Nano and S trathmann (2006) studied Fe (II) sorption on TiO 2 and found more than 80% Fe (II) sorbed in a pH 7 solution. Howev er, the removal capacity and removal reaction rate are important issue s for the selection of removal materials. The final Fe (II) concentration of CCMBs treated samples indicates CCMBs have better removal effectiveness. Kinetics of Removal Reaction by Reac tive Materials Six reactive materials (exclud ing quartz sand and gypsum due to lower removal effectiveness) were used to conduct the removal reaction kinetics test. The change of Fe (II) concentration over time was plotted in Figure 2 2. For better illustr ation, figure 2 3 shows the first hour removal reaction and it is apparent that more than 90% of Fe (II) was removed by the six reactive materials in the first h our. After 12 hours, the Fe (II) concentrations in the aqueous phase become stable. The Fe (II) concentration in limestone treated sample achieved the lowest value (0.031 mg/L) after 72 hours, and more than 99% Fe (II) was removed during the reaction time. The entire Fe (II) removal process cannot be described by a simple rate law, however, a first order kinetics can be applied to the first hour removal process. The first order reaction can be written as (2 1)

PAGE 33

33 i.e., the Fe (II) removal rate is a function of Fe (II) concentration. Figure 2 4 is plotted as InC vs. time. The calculated regression constants for the six materials are listed in Table 2 3 with the highest being 2. 75 hour 1 for limestone w hich also has the lowest final Fe (II) concentration after treatment. Crushed concretes A also have good reaction rate (2.36 hour 1 ) and relatively high removal effectiveness. The obvious difference of the removal rates in the fir st hour and afterwards suggests a possibly two step (fast and slow) removal process. As a matter of fact, this two step process has already been observed in the treatment of various metals (Mn 2+ : Lorens et al. 1981, Franklin et al. 1983, Cd 2+ Davis et al. 1987, Martin Garinn et al. 2003; Zn 2+ ,Zachara et al. 1988). Based on these studies, the removal process was found to compris e two kinetically steps, a fast initial step (1 2 hours), followed by a slow uptake lasting hours or days. Mettler et al. (20 09) reported a two phase uptake for Fe (II) removal by calcite, and interprete d the first step as adsorption and the second step as re crystallization o r co precipitatin This is probably the case for our study as we used the same materials and observed ve ry similar pattern. For the limestone removal reaction, the changes in Fe (II) concentration, calcium concentration, carbonate concentration, and pH were plotted in Figure 2 6. As shown in this figure, along with the decrease in Fe (II) concentration calc ium concentration increased quickly within the first hour, and then stabilized at 12 mg/L ( 0.3 mM ) in solution. Carbonate concentration increased rapidly in the first two hours, and stayed at a concentration around 42 mg/L ( 0.7 mM ) in solution. This result indicates that carbonate and calcium concentrations will increase after groundwater passing through

PAGE 34

34 PRB. They also suggest that carbonate concentration may be higher than calcium concentration. Effects of Materials Size on Removal Effectiveness The size o f the reactive materials is an important factor. Bigger size material has smaller surface area. Material size may affect the removal effectiveness and removal capacity in an engineering process because of the differen ce of surface area Surface area is clo sely associated with available adsorption sites and surface reactivity. The more surface area, the quicker Fe (II) will be adsorbed onto or interact with the materials. The best performing reactive materials, limestone and crushed concrete A were chosen fo r the following effectiveness test of different sized reactive materials. The final Fe (II) concentrations and pH values of treated samples were determined and presented in Figure 2 5. Final Fe (II) concentrations values were significantly different ( p <0.0 1 ) for the tested sizes It is clearly shown that final Fe (II) concentration increased with increasing particle size and large particles had lower Fe (II) removal effectiveness. This is probably because smaller sized particles usually have larger surface area leading to higher surface capacity for adsorption and reaction (Stumm and Morgan 1996). The final pH values were similar, suggesting that particle size of reactive materials affect pH of the treated samples. Therefore, size of the reactive materials can affect the longevity and effectiveness but cannot affect pH of groundwater in PRB implementation. Characterization of Precipitates on Reactive Material Yellow and brown precipitates formed on the surface of limestone reactive materials after one hour of the batch test. These precipitates continued to form until the completion of the batch test. After 3 day aging in anoxic condition, SEM examination of

PAGE 35

35 samples from the treated limestone particles showed an amorphous layer (made up by small colloids) co ating the surface, and the size of most of colloids is less than 1000 nm (Figure 2 7) The digestion analysis of this layer shows that it contains Ca (35%) and Fe (6.5%). XRD analysis of the precipitates identified CaFe(CO 3 ) 2 as the main crystal min eral in the products (Figure 2 8 ). This result is similar to the study by Wajon et al (1985) who reported that siderite and CaFe(CO 3 ) 2 are the main products in the reaction between calcium carbonate and Fe (II) and the possible reason is that siderite may not form crystal on limestone surface. We further measured the zeta potential of limestone powder surface before treatment and got 4.6, 5.1, 5.2, and 6.1 mV for pH 6, 7, 8, and 9, respectively. The weak nega tive charge in neutral pH range can also favor Fe (II) sorption on the limestone surface. Summary Eight different CCBMs were examined for their potential and effectiveness in removing Fe (II) from groundwater. The final Fe (II) concentration in treated sol ution followed: Limestone < concrete < marble < dolomite < witherite < sand < gypsum. Using the first order kinetics for the first hour reaction, limestone has the highest reaction rate. Although all carbonate materials were able to reduce Fe (II) concentr ation to below drinking water limit and GWCTL (0.3 mg/L), only limestone and crushed concrete are selected for later studies due to their superior performance and/or environmental benefits. The size effects of reactive materials on the removal effectivene ss were also determined. Smaller size material has better removal effectiveness because of larger surface area. I n terms of the removal mechanism, t he removal process of Fe (II) by limestone is considered as a two step reaction: t he first step being a rapi d sorption

PAGE 36

36 process and the second step being relative ly slow co precipitation. Overall, these findings indicate CCBMs has good potential to remove Fe (II), and limestone and crushed concrete are picked as reactive materials candidates for further PRB test in which we would attempt to understand the influence of environmental factors on the removal effectiveness and performance of these materials in field

PAGE 37

37 Table 2 1 Size and density of all tested reactive materials Reactive materials S ize range (mm) Avera ge Diameter (mm) Specific gravity Dry density (kg/m 3 ) Limestone 1 3.0 5.0 4.1 2.61 1.62 Limestone 2 7.0 10.0 8.5 2.61 1.54 Limestone 3 15.0 25.0 29.1 2.61 1.53 Limestone 4 40.0 50.0 43.5 2.61 1.51 Dolomite 7.0 10.0 8.7 2.85 1.63 Witherite 7.0 10. 0 8.5 3.04 1.69 Marble 7.0 10.0 8.4 2.74 1.46 Crushed concrete A 1 3.0 5.0 4.2 2.40 1.67 Crushed concrete A 2 7.0 10.0 8.5 2.40 1.31 Crushed concrete A 3 15.0 25.0 28.4 2.40 1.30 Crushed concrete A 4 40.0 50.0 42.5 2.40 1.26 Crushed concrete B 7.0 10.0 8.5 2.40 1.28 Gypsum 7.0 10.0 8.7 2.79 1.48 Quartz sand 0.3 0.5 0.43 2.32 1.86 Size range is determined by measuring 100 particles Average diameter is calculated by measuring 100 particles

PAGE 38

38 Table 2 2 Metal contents and pH of all tested reactive materials Materials Ca (%) Mg (%) Ba (%) Na (%) pH Limestone 36.2 1.21 6.62 Dolomite 25.6 17.2 7.3 Witherite 24.3 1.03 21.5 7.52 Marble 34.5 -6.98 Crushed concrete A 12.4 6.73 10.2 9.13 Crushed concrete B 11.5 3.45 12.3 9.34 Gypsum 30.2 -6.53 Quartz sand 6.46 Metal contents were calculated from the results of digested samples pH was measured by EPA 9425 method

PAGE 39

39 Table 2 3 F irst order reaction constants of removal reactions f or reactive materials K (hour 1 ) C o rrelation coefficient ( r 2 ) Limestone 2.75 0.91 Crushed c oncrete A 2.36 0.89 C rushed c oncrete B 1.92 0.9 Marble 1.82 0.91 Dolomite 1.32 0.82 Witherite 0.9 0 0.88 K is the first order reaction coefficient and r 2 is the linear correlat ion coefficient. Limestone has the fastest reaction in one hour.

PAGE 40

40 Figure 2 1 Final Fe (II) concentrations and pH values of the mixture s of synthetic groundwater and reactive materials after 72 hours treatment ( Error bars represent the standard deviation of three replicate samples. ) N umber s on the top of bars are final pH of solution. The top dash line is the initial Fe (II) concentration (50 mg/L), and lower dash line is Groundwater Cleanup Target level (0.3 mg/L)

PAGE 41

41 Figure 2 2 Fe (II) concentrations measured from the mixture of synthetic groundwater and reactive materials over time. (Each result point represent s the average value of triplicate samples ) The top dash line is the ini tial Fe (II) concentration (50 mg/L). Fe (II) concentration dropped quickly in first hour and kept declining for three days.

PAGE 42

42 Figure 2 3 Fe (II) concentrations measured from the mixture of synthetic groundwater and r eactive materials in the first hour. (Each result point represents the average value of triplicate samples ) The i nitial Fe (II) concentration is 50 mg/L.

PAGE 43

43 Figure 2 4 Fe (II) concentrations measured from the mixture o f synthetic groundwater and reactive materials in the first hour ( Logarithmic C vs. time). The initial Fe (II) concentration is 50 mg/L. The data points are fit to linear regression equations.

PAGE 44

44 Figure 2 5 Final Fe (I I) concentrations and pH values for the mixture of synthetic groundwater and different sized reactive materials ( Open bars represent limestone samples and grey bar s represent concrete A samples ; No. 1 is 3 mm~7.5 mm No. 2 is 7.5 mm~15 mm, No. 3 is 15 mm~25 mm, and No. 4 is 25 mm~50 mm). Final pH values are on the top of bars. Error bars represent the standard deviation of three replicate samples.

PAGE 45

45 Figure 2 6 Change of Fe (II) calcium, and carbonate concentration s and pH values over time in limestone and synthetic groundwater system The initial concentrations are 50 0, 0 mg/L for Fe (II), Ca (II) and carbonate, respectively Apparently, concentrations and pH values changed rapidly in the first few hours.

PAGE 46

46 Fi gure 2 7 SEM graphs of limestone surface before and after treatment. Left one is limestone particle surface before treatment (50000 time magnification) and right one is limestone particle surface after treatment (30000 time magnification)

PAGE 47

47 Figure 2 8 XRD result of precipitates f rom the reaction between synthetic groundwater and limestone. The precipitates were collected from limestone surface after the treatment. Calcite and calcium siderite were detected in crystal structure.

PAGE 48

48 CHAPTER 3 EFFECT S OF ENVIRONMENTAL FACTORS ON THE REMOVAL REACTION BETWEEN CALCIUM CARBONATE BAESD MATERIALS AND FE(II) Introduction In the previous study the removal ability of Fe (II) by c alciu m carbonate based materials (CCBMs) w as assess ed. The results showed that a ll CCBMs have the potential to remove Fe (II) from the synthetic iron contaminated groundwater As a matter of fact, i n the real groundwater treatment system many complex environmental conditions /factors are present and play an important role in the trea tment process. A number of studies have been conducted to evaluate the eff ects of different conditions on pollutant remova l (Masscheleyn et al. 1991; Jeon et al. 2005; Gom ari et al. 2006) In particular, in situ permeable reactive barriers (PRB) encount er various environmental conditions /factors in field that can prohibit or prom ote the removal reaction. M any environmental factors may affect the performance of CCB M PRBs for Fe (II) removal e. g., pH, co exi s ting cations, and natura l organic matters (N OM) being of particular importance Groundwater pH can affect the removal effectiveness of Fe (II) by minerals (Nano and S trathmann, 2006) with h igher pH prom o t ing metal ions sorption (Benjamin et al. 1981 a, 1981b ) and /or accelerat ing f errous hydroxides formation on mineral surface (pH > 9) (Stumm and Morgan 1996). Co existing catio ns are another important factor. Sodium common in landf ill leachate and landfill contaminated groundwater ( Kjeldsen et al., 2002; Keimowitz et al. 2005 a, 2005b ) can lead to reduction in the metal removal effectiveness by increasing the ionic strength ( Greene et al. 1987). Calcium ions can be naturally released from CCBM PRB and it m ay affect the removal reaction by preventing other cations from approaching the CCBM surface Manganese a contamina nt often associated with Fe (II) contamination (Jeon et al.

PAGE 49

49 2005) can also compet e with Fe (II) for adsorption sites on the CCBM surface (Aziz et al. 2001). A growing number of studies have evaluated the role of natural organic matter (NOM) in water treatment processes. Some research ha s already been performed on the sorption of NOM o n to CCBM s and the associated effects on metal sorption (Lee et al. 2005) It has been found that NOM can interact with metal ions through chelation and complexation and can thus cause a decrease in uptake capacity (Jeon et al 2005). NOM can also attach on to the mineral surface and inhibit the reaction between metals and CCBM s For example, NOM was found to restrain the reaction between Cu (II) ion a nd calcite with NOM reduc ing 3% of Cu (II) removal efficiency (Lee et al. 2005). This chapter focuses on the effects of various environmental factors (i.e. pH, co existing cations, and organic matters) on the reaction between CCBM and Fe (II) The stud y is divided into three objectives: the first objective is to understand the effect of pH on the removal reaction between Fe (II) and limestone material ; the second objective is to study how the co existing cations affect the interaction between Fe (II) an d limestone material and the last objective is to test the effects of NOM on the interaction between Fe (II) and limestone material. Method and Materials Experimental Materials The reactive material used in this study was limestone with a granular size of 7.0 10 mm in diameter (named Limestone 1 in Chapter 2). All experiments were conducted at room temperature (25 2 C). The composition of limestone is listed in table 2 2. O xygen free deionized water was prepared by pre purg ing the water with nitrogen gas f or at least half an hour (Airgas, 99.99% purity). FeCl 2 NaCl, MnCl 2 and CaCl 2 were

PAGE 50

50 used to simulate natural conditions. S tock solution ( 1000 mg/L Fe (II) ) was made with FeCl 2 and oxygen free DI water. 1000 mg/L Na M n and Ca stock solution s w ere made wi th oxygen free DI water and NaCl, MnCl 2 and CaCl 2 respectively Collection and Preparation of Natural Organic Matters Three NOM s were used in this study. The first one is concentrated natura l organic matter solution which is waste brine from a MIEX pro cess (Apell and Boyer, 2010), collected from a Florida groundwater treatment plant ( Cedar key groundwater treatment plant CKGO) The second organic matter is landfill leachate collected from a local MSW landfill ( New River Regional Landfill FL, NR R L). Th e last is Suwannee River H u mic A cid (SRHA, IHSS Standard ), which served as a control sample collected from surface water These three NOMs were chosen to investigate the effect of NOM on the removal efficiency of Fe (II) by limestone since they represent N OM in groundwater, landfill leachate, and surface water The o riginal CKGO and NRLL NOM solutions were pre filtered through a 0.45 m nylon filter ( Fisher Sci Inc.). The filtrat es w ere then desalte d by ultrafiltration through a m embrane with a nominal cuto ff molecular weight of 1000 Daltons wherein salts passed through and NOM molecule s remained (Bjelopavlic et al., 1999) The NOM molecules on membrane were flushed out by DI water, and the conductivity of this NOM solution was measured to be less than 100 S cm 1 The UV absorbance of each NOM solution was measured on a Perkin Elmer Lambda 800 spectrophotometer with 1 cm quartz cell at the wavelength of 254 nm Dissolve d organic carbon concentration s were measured on a Tekmar Dorhman Phoenix 8000 TOC (Tota l Organic Carbon) Analyzer. Specific UV absorbance

PAGE 51

51 (SUVA 254 ) was calculated by dividing the 254 nm UV absorbance by the TOC amount (Boyer and Singer, 2008). A modified titration test was used for each NOM sample to determine the charge density (Lin et a l 2005). A 100 mL stock solution containing 100 mg/L C of NOM was prepared for each NOM sample. The stock NOM solutions were purged with N 2 for 30 minutes and adjusted to pH 3.0 with 0.01N HCl. A 25 mL burrette filled with 0.04 M NaOH was used for titrati on. 0.1 mL of the NaOH titrant was added stepwise, and 1 minute was allowed to reach equilibrium. The added volume of NaOH and pH were recorded. Accumet 209 pH meter was used to monitor pH. The charge density of NOM was calculated as follows : w here [A ] is the net charge of NOM in equivalents per liter, V 0 is the initial sample volume and C is the concentration of NOM of sample (g as NOM/ L ) (Lin 2005). Carboxyl acidity which i s proportional to the number of carboxyl group s in NOM molecules was defined as the char ge density at pH 8 (Boyer and Singer, 2008). pH Effect on Fe (II) Removal Reaction To test the effects of pH, Fe (II) solution s with varying pH values were used as th e synthetic groundwater. The stock solution was spiked into buffer solution s (pH 4, 5, 6, 7, 8, 9, respectively). MES (2 (N morpholino) ethanesulfonic acid) and HEP E S (N 2 hydroxyethylpiperazine N' 2 ethanesulfonic acid) were used to make 0.01 M buffer sol ution of pH 5, 6, 7, 8, and 9 since these two organic chemical have no effects on the ionic strength and have very weak reactivity with metals ( Good et al. 1966 ) and pH 4 buffer solution was made by diluted 0.01 N HNO 3

PAGE 52

52 Fifty grams of limestone and 1000 mL of each Fe (II) solution (fixed pH) were added to a 1L bottle s in a gl ove box filled with N 2 The samples were shaken horizontal ly at 30 RPM for various reaction time (10 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, and 12 hours) After reaction the water phase was removed from the bottle and filtered through 0.4 m cellulose filter paper (Fisher Sci. Inc.) in the glove box The final Fe (II) concentration and pH of the filtrate were measured. All experiments were conducted in triplicate. Effects o f Co existing Cati ons on Fe (II) Removal Reaction Th e effect s of co existing ions (Na + Ca 2+ and Mn 2+ ) on Fe (II) removal by limestone were studied For each metal ion various levels of co existing cation concentrations were applied to simulate real grou ndwater condition s The i nitial pH of the samples was adjusted to 7 with diluted HNO 3 or NaOH solution. A g love box was used to prepare solution to achieve low dissolved oxygen condition. Appropriate amount s of FeCl 2 and NaCl stock solutions were mixed to yield 5, 10, 50, 100, and 200 mg/L Na and 50 mg/L Fe (II) A bottle with 1000 ml of Na/Fe solution and 50 g of limestone was mixed and shaken on a horizontal rotator ( Fisher model 341, 30 RPM). After a 12 hour rotation period, the Fe (II) concentration an d pH of the filtrate were measured. For 0 and 200 mg/L Na concentration samples, a kinetics test was performed with different reaction time (10 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, and 12 hours). Similarly, FeCl 2 and CaCl 2 stock solutions conta ining 10, 50, 100, and, 200 mg/L Ca (II) and 50 mg/L Fe(II ), and FeCl 2 and MnCl 2 stock solutions containing 5, 10, 50, and 100 mg/L Mn (II) and 50 mg/L Fe(II) were prepared respectively For 2 00 mg/ L Ca and 100 mg/L Mn concentration samples, a kinetics te st was performed with different

PAGE 53

53 reaction time (10 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, and 12 hours). All tests were run in triplicate. Effects of Organic Matter on Fe(II) Removal Reaction A 100 mg C/L stock solution was prepared for each NOM (C KGO, NRRL, and SRHA) During the NOM experiment, the pH of all solution s was adjusted to a pproximate 7 with diluted HNO 3 or NaOH solution. Tests using SRHA were performed in duplicate and small scale ( due to the limited amount of sample ) and NRLL and CKGO tests were performed in triplicate Four different procedures were employed for evaluating the effects of NOM on removal effectiveness of Fe (II) by limestone with different mixing order Table 3 1 describes the experiment steps of four conditions. In thi s study 200 ml of 10 mg C/L NOM solution, 50 mg/L Fe (II), and 10 g limestone were used. The third test condition was repeated for various NOM concentration s as Fe (II) and NOM are premixed under a real groundwater condition. Two NOM c oncentrations, 5 and 20 mg/L C, were used to test the effects of NOM concentrations on removal effectiveness. Also, a kinetics study (50 mg/L Fe (II) and 10 mg/L C NOM were premixed and shaken for 24 hours, and then mixed with limestone) was performed with different react ion time s for each NOM (10 minutes, 30 minutes, 1 hour, 2 hours, 6 hours, and 12 hours). Data Analysis All the experiments were run in triplicate, and data were illustrated as mean standard error (SE). Statistical analyses were performed using one way an alysis of value of less than 0.05 was considered to be statistically significant.

PAGE 54

54 Results and Discussion Effects of pH on Fe (II) Removal Effectiveness After 12 hours of r eaction time pH of each sample was measured and found to fall within 0.5 of the initial pH indicating minimal pH change during the reaction As shown in Figure 3 1, final Fe (II) concentration at pH 4 was significantly different (p<0.01) from other pHs pH 4 solution ha d the highest final Fe (II) content (4.50 0.57 mg/L ) whereas Fe (II) concentrations in other pH solution s were all below the Groundwater cleanup target level ( GWCTL ) of 0.3 mg/L (0.01 0.01 mg/L to 0.120.02 mg/L ) The se result s demons trate that lower pH significantly decreases the removal effectiveness. As theory is discussed in the earlier chapter, Fe (II) removal by minerals follows a two phase process (Nano and Strathmann, 2006) An initial phase explained by rapid sorption during t he first few hours is followed by a much slower uptake process. At lower pH the removal effectiveness was the lowest ( 10% Fe (II) remaining in water) due to the greater amount of H + in solution which can inhibit the reaction between Fe (II) ion s approach the limestone surface (Stumm and Morgan 1996). In general, limestone particle surface has a negative charge, and therefore, Fe (II) ions, which have positive charges, can be attract onto the limestone particles. In the lower pH conditions, the surface cha rge of limestone can shift toward less negative and reduce the attraction of Fe (II) ions. In the co precipitation step, lower pH means more cations (H + ) and they can inhibit the formation of iron precipitates. When pH increased to above 6, the removal eff ectiveness between pH 7, 8, and 9 conditions do not have significant difference (Figure 3 1; p<0.01). Figure 3 2 summarizes the Fe (II) removal kinetics under different pH conditions. Fe (II) concentrations in pH 6, 7, 8, and 9 solutions dropped quickly, with more than

PAGE 55

55 99% Fe (II) removed in the first hour. But in pH 4 solution, only 38% Fe (II) was removed in first hour, and 75% removal in pH 5 solution, indicating that Fe (II) removal efficiency decreased with pH. Effects of Na + on Fe (II) Removal Eff ectiveness The final Fe (II) concentrations in the mixtures with different Na + concentration s were significantly different ( p <0.01) rang ing from 0.027 0.01 mg/L to 0.32 0.08 mg/L ( Figure 3 3 ) The final pH values of all samples were within the range o f 7.00.5. Compare d to the control sample (no Na + and final Fe (II) was 0.027 mg/L), final Fe (II) concentration in the 200 mg/L Na sample was more than ten times higher (i.e., 0.32 mg/L ). Figure 3 4 shows change of Fe (II) concentration over time (with an d without Na + ). As can be seen f rom this figure, Fe (II) removal process was affected by the presence of Na + Na + has one positive charge, the possible reason of the effect is similar as the effects of H + In general, increasing Na + concentration causes an increase in the ionic strength. The possible ways in which ionic strength influences the removal of metals by mineral are 1) affecting the sorption process, and 2) affecting co precipitation process. The monovalent Na + ion can affect the interfacial pot ential and the activity of electrolyte ions and adsorption, and interfere the competition of electrolyte ions and adsorbing anions for the sorption sites (Jeon et al. 2005). S odium will not be bound or sorbed covalently by mineral surface (Stumm and Morga n 1996) but it will affect the mineral re crystallization process by inhibit ing the approach of other ions to the limestone surface This explain s why ionic strength ha s a n effect on the Fe (II) removal by limestone in the presence of Na +

PAGE 56

56 Effects of C a 2+ on Fe (II) Removal Effectiveness = 0.05, p <0.01, ANOVA analysis) ranging from 0.027 0.008 (control) to 0.42 0.05 mg/L of Fe ( 200 mg/L Ca 2 + ) ( Figure 3 5 ) As can be seen f rom figure 3 5, similar to Na + higher calcium concentration also cause d more free Fe (II) ions in solution possibly due to the same fact discussed above that h igher Ca 2+ can slightly inhibit the metal ion co precipitate or re crystallize process (Rouff et al. 2005). Since the ze ta pot ential of the limestone was 4.6 to 6.1 mV, Ca 2+ will act as counter ions on the surface and thus impede the pathway of Fe (II) to approach the limestone surface. Effects of Mn 2+ on Fe (II) Removal Effectiveness Fe (II) concentrations in the aqueous samples obtained after 12 hours from the tests were 0.0280.01, 0.0520.01, 0.1230.04, and 0.350.07 mg/L for the samples initially containing 5, 10, 50, and 100 mg Mn /L respectively (Figure 3 6) The final Fe (II) in different Mn concentration samples p <0.01, ANOVA analysis). Similar to Na + and Ca 2+ the result s suggests that Mn 2+ can also reduce the removal effectiveness of Fe (II) by limestone One possible reason is that Mn 2+ and Fe 2+ are competitors for CO 3 2 in the solution to form insoluble MnCO 3 and FeCO 3 minerals (Jensen et al. 2002). In addition, Mn 2+ can compete with Fe (II) for the adsorption site on calcite surface (Cave et al. 2005). Mn 2+ also has the ability to inhibit the dissolution of calcite su rface and thus affects the Fe (II) co precipitation on calcite surface (Vinson et al. 2007). Final Mn (II) concentrations after 12 hours were 0.230.04, 0.720.15, 1.210.23, 1.420.35 mg/L for the samples initially containing 5, 10, 50, and 100 mg Mn /L respectively, proving that Mn (II) is a competing ion for Fe (II) and that limestone can

PAGE 57

57 also be used to remove Mn (II) from Mn contaminated groundwater. As Mn contamination is often observed along with iron, limestone treatment would be a suitable metho d to remove Fe (II) and Mn (II) simultaneously. Effects of Natural Organic Matter on Fe (II) Removal Effectiveness Another important environmental factor for Fe (II) removal from groundwater is the content of natural organic matters (NOMs). In contamin ated groundwater, NOMs concentration ranges from 0 to 100 mg/L (Krasner et al., 1996; Jeon et al. 2005), and they can interact with metal ions by chelation and complexation, thus resulting in reduction in iron removal effectiveness (Jeon et al., 2005). F igure 3 8 shows the final Fe (II) concentration for each experimental procedure. Overall, final Fe (II) concentrations of with and without NOM samples were significantly different (p<0.01) and the Fe (II) concentration increased in the presence of NOM. Amo ng all the samples, only those with limestone and synthetic groundwater premixed achieved a final Fe (II) concentration less than 0.3 mg/L (GWCTL). For all the experiments, the lowest final Fe (II) concentration was obtained when limestone was premixed wi th Fe (II). This is probably because of the co precipitation of Fe (II) on limestone surface. Co precipitation of Fe (II) on limestone surface is an irreversible reaction, and therefore, release of Fe (II) into water phase will only take place after its up take by limestone. Similar result was reported by Mettler et al. (2009), who found that Fe (II) could hardly be uptaken by 1, 10 phenanthroline after reacting with calcite for 6 hours. After Fe (II) reacted with and incorporated in the limestone, it become s less likely for NOM to form complex with Fe (II), resulting in less dissolved Fe (II) in solution.

PAGE 58

58 Samples with limestone and NOM premixed had relatively higher Fe (II) in solution among four experimental procedures. Pre mixing of limestone and NOM cau sed adsorption of NOM on the limestone as the first reaction (Lin et al., 2005) and reduced the amount of available NOM for complexing with Fe (II). Samples with Fe (II) and NOM premixed had relatively higher Fe (II) in solution, too. This procedure repre sents field condition because Fe (II) is allowed to react with NOM before groundwater passing through the reactive materials. NOM can form soluble complex with cations through carboxylic acid moieties at neutral pH (Leenheer et al., 1998). The formation of dissolved NOM Fe (II) complexes may have two effects on the interaction of Fe (II) with limestone: (1) reduces the free Fe (II) activity and (2) changes the charge of NOM molecules and affects the adsorption of NOM onto limestone surface (Lin et al., 2005 ). Figure 3 9 shows the change of final Fe (II) concentration over time (NOM premixed with Fe (II) solution, with and without NOMs). According to this figure, Fe (II) removal process was only slightly affected by the presence of NOMs, and the Fe (II) chan ge curves have similar pattern for three different NOMs. This result shows that NOMs have little effect on the removal reaction kinetics, while having more impacts on the iron removal effectiveness. The reduction in the Fe (II) removal effectiveness as a function of NOM concentration for the three different types of NOM is presented in Figure 3 10. I t can be clearly seen that Fe (II) concentration increased with increasing amount of NOM, indicating that more NOM can mobilize more Fe (II) in the aqueous pha se. The linear regression relationship equation and r square between NOM concentration and final Fe

PAGE 59

59 (II) concentration were listed in Table 3 2. The ANOVA test shows that three linear equations are significantly different (p<0.01) Base on the linear regre ssion equation, every 1 mg C/L SRHA or CKGO NOM can keep about 0.05 mg/L Fe (II) in limestone Fe (II) solution at tested NOM concentration range, and 1m g C/L NRRL NOM can retain 0.032 mg/L Fe (II) in solution. SRHA ( surface water NOM) and CKGO (groundwate r NOM) showed higher ability to retain Fe (II) in water phase, and landfill leachate had lower ability among three NOMs. This difference is probably associated with the physico chemical properties of these NOMs. Table 3 3 lists the carboxyl acidity (define d as the charge density at pH 8), charge density at pH 7 (since pH is 7 in current study), and SUVA 254 of the tested NOMs. The order for the carboxyl acidity and charge density at 7 for the three NOMs are same: SRHA > CKGO > NRRL. This order is similar to th e effects of NOMs on Fe (II) removal reaction (SRHA=CKGO > NRRL). Carboxyl acidity represents the number of carboxyl groups in the NOM molecules (Boyer and Singer, 2008), which is an important property as Fe (II) can be complexed by carboxyl groups in NOMs ( Lee et al. (2005). Ritchie and Perdue (2003) reported that the charge density of SRHA is around 9.2 (pH 8), which is close to our result. Therefore, as a further attempt, correlation between the NOM charge density at pH 7 and retained Fe (II) concentration was studied. Figure 3 12 shows the relationship between Fe (II) concentration and charge density of NOMs. A linear regression equation was obtained (r = 0.9097): (3 1)

PAGE 60

60 Th is equation suggests that the Fe (II) removal effectiveness can be predicted by the charge density of NOM in solution, i.e., more carboxyl acidity, more dissolved Fe (II) in the water phase. Also, equation 3 1 can be transformed ( r = 0.9097): (3 2) This equation indicates that each Fe (II) ion requires 95 NOM carboxyl groups to form stable complexes. Therefore, final Fe (II) concentration can be calculated by the NOM charge den sity titration result. SUVA 254 (specific ultraviolet absorbance at 254 nm ) is calculated by dividing a sample's UV absorbance at 254 nm by the DOC (dissolved organic carbon, in mg/L), and then multiplying by 100. Park et al. (2007) reported SUVA 254 of SRHA being 3.98, which is close to our result (3.5). All three NOMs have a SUVA 254 less than 4, indicating that these NOMs have low aromatic carbon content and are influenced by microbial byproducts (Comstock et al., 2010). However, t here is no clear relation ship between SUVA and Fe (II) concentration in water phase indicating that SUVA 254 values cannot be used to predict the remaining Fe (II) concentration in water phase in the presence of NOM Effects of P ossible Field Conditions on Fe (II) Removal Effectiv eness In target contaminated groundwater sites, the different environmental factors all play roles in the remediation processes. In Kl o ndike landfill groundwater, pH is around 6.7, sodium concentration around 50 mg/L, manganese concentration 2 3 mg/L, ca lcium concentration around 10 mg/L, and NOM is around 10 20 mg C/L. According to our above results, under these conditions, pH and concentrations of Na, Ca and Mn

PAGE 61

61 should have no significant effects on the removal reaction whereas NOM is the only factor tha t may decrease the removal effectiveness. Summary Various environmental factors including pH, co existing cations, and natural organic matters, can change the water chemistry and have impacts on the Fe (II) removal by CCBMs. When investigating the pH eff ects, we found that l ower pH (pH = 4) could reduce the removal effectiveness of Fe (II) by limestone. In neutral pH range, the removal effectiveness was enhanced slightly by increasing pH. Na + has slight effect on the removal process due to the increase in ionic strength and Ca 2+ has similar effects as sodium. Mn 2+ can also be removed by limestone and is a competitive cation for Fe (II) as they compete for sorption sites on the limestone In addition, NOM decrease s Fe (II) uptake by limestone and thus reduc e s the removal effectiveness. C arboxyl group s in NOM molecules can form complex es with Fe (II) and retain more Fe (II) in the aqueous phase. On the other hand, NOM only slightly a ffects the Fe (II) removal reaction kinetics i.e., removal efficiency Overa ll, various environmental factors could affect the Fe (II) removal process by limestone. Co existing cations and NOM could be an issue as they are naturally present in groundwater and can mobile more Fe (II) in groundwater and therefore cause iron concent ration over GWCTL

PAGE 62

62 Table 3 1 Experimental steps for the NOM effect test Different conditions First step Second step (1) Limestone was mixed with NOM, aging for 24 hours Fe (II) was then spiked into the mixture (2) Limestone was mixed with Fe (II), ag ing for 24 hours NOM was then spiked into the mixture (3) NOM and Fe (II) were equilibrated for 24 hours Solution was then added to limestone (4) NOM and Fe (II) were added into a limestone system simultaneously This test is conducted with 50 g limesto ne particles and 1000 mL synthetic groundwater with 50 mg/L Fe (II). The pre mix step is 24 hours, and the followed reaction step is 12 hours.

PAGE 63

63 Table 3 2 Linear regression equation and correlation coefficient between NOM concentration and final Fe (II) co ncentration NOM Equation r 2 SRHA y = 0.051 8 x + 0.0802 0 .959 CKGP y = 0.0517x + 0.0322 0.9747 NRRL y = 0.0321x + 0.0082 0.99 In equation, y represents Fe (II) concentration, and x represent s the NOM concentration.

PAGE 64

64 Table 3 3 Physicochemical p roperties of natural organic matters (after ultrafiltration) NOM Carboxyl acidity (meq/g C) Charge density at pH 7 (meq/g C) SUVA 254 SRHA 9.6 9.4 3.5 CKGP 9.0 8.6 3.7 NRRL 8.3 8.0 3.2 Carboxyl acidity is the NOM charge density at pH 8. Charge density at pH 7 is the NOM charge density at pH 7.

PAGE 65

65 Figure 3 1 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone under different pH conditions for 12 hours ( The initial solution containe d 50 mg/L Fe (II) and 50 g/L limestone. Error bars represent the standard deviation of three replicate samples.

PAGE 66

66 Figure 3 2 Fe (II) concentrations measured over time in the mixture of synthetic groundwater and lim estone under different pH conditions The initial solution contained 50 mg/L Fe (II) and 50 g/L limestone. Fe (II) concentration dropped quickly in the first hour and kept declining in next three days. Error bars represent the standard deviation of three r eplicate samples.

PAGE 67

67 Figure 3 3 Fe (II) concentrations measured in the mixture of synthetic groundwater and Limestone with different sodium concentrations for 12 hours. The initial solution contained 50 mg/L Fe (II) an d 50 g/L limestone. Error bars represent the standard deviation of three replicate samples.

PAGE 68

68 Figure 3 4 Fe (II) concentrations measured over time in the mixture of synthetic groundwater and limestone with and without sodium ions The initial solution contained 50 mg/L Fe (II) and 50 g/L limestone. Fe (II) concentration dropped quickly in the first hour and kept declining in next three days. Error bars represent the standard deviation of three replicate samples.

PAGE 69

69 Figu re 3 5 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone with different calcium concentrations for 12 hours. The initial solution contained 50 mg/L Fe (II) and 50 g/L limestone. Error bars represent the standard deviation of three replicate samples.

PAGE 70

70 Figure 3 6 Fe (II) concentrations measured in the mixture of synthetic groundwater and limestone with different Mn concentrations for 12 hour s. The initial solution contained 50 mg/L Fe (II) and 50 g/L limestone. Increasing Mn concentration increase s the final Fe (II) concentration i.e. reduces the iron removal effectiveness. Error bars represent the standard deviation of three replicate sampl es.

PAGE 71

71 Figure 3 7 Fe (II) concentrations measured over time in the mixture of synthetic groundwater and limestone with and without Mn 2+ The initial solution contained 50 mg/L Fe (II) and 50 g/L limestone. Fe (II) con centration dropped quickly in the first hour and kept declining in next three days. Error bars represent the standard deviation of three replicate samples.

PAGE 72

72 Figure 3 8 Final Fe (II) concentration s after removal by limestone with different concentration of NOM (1 0 g limestone + 200 mL 50 mg/L Fe (II) + 10 mg/L C NOM pH = 7) Error bars represent the standard deviation of three replicate samples.

PAGE 73

73 Fig ure 3 9 Fe (II) concentrati ons measured over time in the mixture of synthetic groundwater and limestone with and without NOMs The initial solution contained 50 g/L limestone, 50 mg/L Fe (II) and 10 mg/L NOM. Fe (II) concentration dropped quickly in the first hour and kept declining in next three days. Error bars represent the standard deviation of three replicate samples.

PAGE 74

74 Figure 3 10 Final Fe (II) concentrations after removal by limestone with different concentration of NOM (1 0 g limestone + 200 mL 50 mg/L Fe (II) + 5, or 10, or 20 mg/L C NOM, pH = 7) NOM was premixed with Fe (II) solution Error bars represent the standard deviation of three replicate samples. The dash lines are linear regression lines

PAGE 75

75 Figure 3 11 Charge densities of three NOMs calculated by the titration result. The c h arge density at pH 7 and carboxyl acidity are estimated from these curves.

PAGE 76

76 Figure 3 12 Charge densities of three NOMs vs. Fe (II ) concentration.

PAGE 77

77 CHAPTER 4 REMOVAL OF FE (II) BY LABORATORY CALCIUM CARBONATE BASED MATERIALS PASSIVE REACTIVE COLUMN Introduction The use of calcium carbonate based materials (CCBM) for removing metals has been found effective in water and industrial w astewater treatment T his technique is particularly advantageous as the use of CCBMs such as limestone, is relatively inexpensive ( i.e., generally less than $50 per ton) compared to other techniques. More recently these materials have been used in the fie ld as reactive materials to treat contaminated groundwater (Mettler et al. 2001) For example, limestone was used in a passive reactive system for acidic mine drainage (AMD) treatment (Sterner et al. 1997), in which limestone increase d the pH of AMD and f orm ed precipitates to remove metals. Also, limestone has been reported for the effective remov al of copper and manganese due to its ability to adsorb and react with divalent metals (Aziz et al. 2001; Aziz et al. 1992; Aziz et al. 1996; Thornton et al. 1995) Although Aziz et al. (2001) also speculated that iron could be removed using such a system, no research has been conducted so far to investigate the effectiveness of CCBM for iron contamination, especially in field. Prior to the installation of PRB, there are a number of parameters to be determined, including the effectiveness and longevity of reactive materials, and these parameters are commonly obtained by column tests (Santomartino and Weber, 2007). Column tests have been conducted on different ma terials including CCBM to assess their effectiveness in removing various pollutants from groundwater, including heavy metal ions such as Zn, Mn, and Cr ( Van Nooten et al., 2008; Turner et al. 2008, Soler et al. 2008, Golab et al. 2009; Aziz et al., 1996; Komnitsas et al. 2004). Precipitates

PAGE 78

78 formation and loss of porosity were reported in several column studies (e.g., Islam et al. 2008, Dikinya et al. 2008) and they could negatively impact the performance of the reactive materials. Therefore, it is very im portant to monitor and evaluate these potential problems properly and estimate PRB effectiveness before field test. In Chapter 2 both limestone and crushed concrete showed potential to remove Fe (II) from groundwater in batch experiments This study aim s to estimat e the Fe (II) removal effectiveness and longevity of these two materials and the associated reduction of porosity using a dynamic flow column Materials and Methods Column Setup Column experiments were conducted at room temperature (25 2 C) to assess the removal of Fe (II) by limestone and crushed concrete PRB and determine the longevity of PRB. Information of these two reactive materials is listed in Table 4 1. The column set up is shown in Figure 4 1. In the sampling ports, straight tubing in serted to the center of the column was used to collect liquid samples and to avoid disturbing the flow. A peristaltic pump was used to inject synthetic groundwater from the bottom port to the top outlet, at approximately 5 ml/min. At this flow rate, the r etention time is around two hours for each column. Base on results in Chapter 2, limestone and crushed concrete can remove 99% Fe (II) concentration within this duration. An accelerated flow rate was selected to access the longevity of the reactive materia ls because it is comparable to that achievable in the field (Golab et al. 2009). Two columns were packed with limestone (Column a, b), and the other two columns with crushed concrete (Column c, d). Approximately 2.1 kg with an average 8.5 mm diameter limes tone and 1.94 kg with an average 8.5 mm diameter crushed concrete were packed into the columns,

PAGE 79

79 respectively. All materials were pre washed with 18 m 1 de i onized water. After the columns were set up, approximately two pore volumes of 18 m 1 demoniz ed water were then pumped through the column to remove the loose particles from the column. C olumn p ore volume s w ere calculated from the weight difference of dry and completely saturated column. I nitial porosity of the limestone and crushed concrete column s were estimated to be 0.430 and 0.445, respectively. A F eCl 2 solution, containing 50 mg/L Fe (II) and pre purged with nitrogen gas was used in the test as synthetic Fe (II) contaminated groundwater. After initiating the column operation, samples from eac h sampling port were collected every two hours for the first six hours, and then once every two days until the breakthrough point (breakthrough point was estimated in a preliminary test in which Fe (II) was only analyzed once per 24 pore volume). Samples w ere collected every three hour s for 12 hours after 288 pore volumes for limestone columns and 240 pore volumes for crushed concrete, respectively. pH and Fe (II) concentrations were analyzed for all samples. Calcium concentration s and carbonate concentrat ion were analyzed for 2, 48, 168, and 300 pore volume samples from Column a and b (limestone), and 2, 48, 144, 252 pore volume samples from Column c and d (crushed concrete). The mass of effluent sample was measured to adjust flow rate once per day. All sa mples were filtered before analysis. Dissolved oxygen in influent and effluent were checked daily to ensure the dissolved oxygen in relative ly low level (all DO concentrations were less than 1.5 mg/L, which is close to the DO of iron contaminated groundwat er). Column Effluent Analysis Fe (II) concentrations were measured using the Hach spectrometer col orimetric method (Hach program 2150). Where dissolved Fe (II) ion reacts with 1, 10

PAGE 80

80 phenanthroline forming a colored complex with the color change is proport ional to the Fe (II) concentration. For total carbon ate (H 2 CO 3 ,+HCO 3 + CO 3 2 ) analysis, the preserved samples were analyzed by a TOC analyzer (Tekmar Dorhman Phoenix 8000 TOC ( Total Organic Carbon) Analyzer) and the inorganic carbon content is determine d by subtracting the total organic carbon content from the total carbon content. Calcium concentration s w ere measured by ICP AES ( Thermo Electron Corporation, Trace Analyzer ) as per USEPA SW 846 Method 6010B. pH values were measured by a portable pH meter (Acuma 209, Fisher Sci Inc.). Once the column tests were completed, N 2 was purged slowly into the columns to drain water and dry the reactive material residues for 48 hours After 48 hours, there was no water come from the columns, then the porosity of pa cked limestone and crushed concrete was estimated again by gravimetric analysis. After the porosity test, columns were disassembled and solid samples were examined by XRD to study the properties of the precipitates formed on the limestone and crushed concr ete particle surface. Powders of precipitates were carefully removed from particle surface to analyze for the metal contents using the EPA 3050B digestion method. In addition, a sequencing batch test was performed the removal capacity of two materials. Bri efly, in the sequencing batch test, 500 mg/L e (II) solution was mixed with limestone or crushed concrete system and the water phase was removed after equilibrium. This process was repeated for seven times and the total Fe (II) removal amount is determined the removal capacity.

PAGE 81

81 Results and Discussion Limestone Column Fe (II) concentration in the effluent from t he limestone columns was continuously below 0.3 mg/L (GWCTL) until 290 pore volume s of synthetic groundwater had passed through the columns (Figure 4 2). Results from duplicate columns ( a ) and ( b ) are presented separately because they have slightly different breakthrough curve, probably due to the heterogeneity of the packed columns. After 291 pore volume, the effluent Fe (II) concentration rose rapid ly to a value close to that of the influent Fe (II) concentration. Over time, the lowe r part of the columns became less eff ective (Figure 4 2). The Fe (II) concentration from all lower port (7.6 cm) samples were above 0.3 mg/L, and it is clear from this ex perimental data that limestone treatability at lower port was exhausted after 130 pore volume s Middle port (15.3 cm) and higher port (22.9 cm) samples began to lose Fe (II) treatability after 160 pore volume s and 260 pore volume s respectively. Figure 4 3 shows the pH change. Before injection of the synthetic groundwater into columns, effluent pH was controlled by limestone. The dissolution of limestone caused an average pH of 7.0 which is close to limestone pH (6.82). After injection of the synthetic g roundwater, effluent pH dropped to 6.8 6.9 immediately, which was The limestone material has buffer ing capacity, which can buffer the synthetic groundwater pH. Before 288 pore volume s of synt hetic groundwater passed through the columns, the effluent pH was stable with a pH value around 6.8. After 288 pore volumes, the effluent pH dropped quickly to close to 6.3 the initial pH of the synthetic groundwater. The results indicate

PAGE 82

82 that the initial part of limestone gradually lost its buffer ing capacity but limestone particles w ere still able to remove Fe (II) The concentrations of calcium and carbonate ions leached from limestone increased slightly with distance along the columns from higher to lower ports, which is opposite to the pH trend (i.e., lower
PAGE 83

83 of pH change over time. Compared to the limestone samples, the pH and Fe (II) concentration changes of crushed concrete samples followed a same pattern. Figure 4 8 shows th e negative linear correlation between pH and Fe (II) concentrations in limestone and crushed concrete columns. The correlation coefficient s are 0.9756 and 0.8242 for crushed concrete and limestone, respectively, indicating that Fe (II) removal is closely r elated to pH change in CCMB columns. Increasing pH leads to the more reaction between Fe (II) precipitates formation in the CCBM treatment system. Figure 4 9 illustrates the total carbonate concentration over time in crushed concrete columns. Although con crete released less carbonate ions into water than limestone, the change of total carbonate concentration in crushed concrete columns is similar to that of limestone columns. Figure 4 10 shows the calcium concentration over time. Because of higher solubili ty of Ca(OH) 2 calcium concentrations in crushed concrete column effluent samples were higher than limestone columns. Golab et al. (2009) also found that crushed concrete released more calcium ion than oyster shell (the major component of oyster shell is c alcite). Loss of Porosity and Formation of Precipitates When the synthetic groundwater came into contact with limesto ne and crushed concrete, Fe (II) ions were removed and pH increased. Furthermore, precipitates formed in each column over time. In limeston e columns, yellow brown precipitates formed on crushed concrete surface. Precipitates formation is commonly observed on PRB reactive material surface and different minerals and pr ecipitates have been detected on CCBM surface. Wagon et al. (1984) observed calcium siderite formation on calcite surface. For precipitates in limestone columns, Komnitsas et al. (2004) reported

PAGE 84

84 formation of Goethite and Fe(OH) 3 precipitates on limestone f or the treatment of acid mine drainage (AMD). In addition, calcium siderite has been identified by XRD (similar results in Chapter 2). Unfortunately, XRD failed to identify the precipitates formed on crushed concrete. This is probably because of the amorph ous, poorly crystallized or fine grained nature of the minerals which is difficult to identify by XRD (Furukawa et al. 2002). As we discussed earlier, CaFe( CO 3 ) 2 and CaCO 3 are likely to form on CCBM surfaces. Table 4 2 presents the metal contents of preci pitates. Calcium content (34.2 %) is higher than iron content (11.2 %) in limestone precipitates and this may be attributed to calcite precipitation and mixing into the precipitates since calcite dissolving into solution and re precipitating on mineral sur face is commonly observed for CCBM (Cave and Talens Alesson, 2005). Iron content is relatively high (32.1%) in crushed concrete precipitates. That indicates ferrous minerals are the dominating species in the pr ecipitates. Results of the porosity tests are presented in Table 4 3. As a result of precipitate formation, the average porosity loss is 0.034 for limestone columns, and 0.032 for crushed concrete columns. Compared to other reported column studies (Phillips et al. 2000; Bilek et al., 2006), the poros ity loss in our test is relatively low. For example, Kamolpornwijit et al. (2003) reported that porosity of their ZVI column was reduced by up to 0.25 of porosity The loss of porosity can cause reduction of hydraulic conductivity, and clogging in PRB syst em (Islam et al. 2004). Based on our results, we can conclude that the formation of precipitates will result in only slight reduction in porosity and thus will have little impact on the PRB operation.

PAGE 85

85 Longevity of Reactive Materials For limestone columns, Fe (II) concentrations in the outlet samples exceeded 0.3 mg/L (GWCTL) after 291 pore volumes of synthetic groundwater. Total amount of Fe (II) removed by limestone can be calculated as: The total amount of Fe (II) removed by limestone is much less than that calculated based on iron removal from the sequencing batch test (i.e., 32.9 g Fe per kg limestone), which is conducted in a sequencing batch test. For c rushed conc rete, the outlet samples Fe (II) concentration exceeded 0.3 mg/L (GWCTL) after 240 pore volumes of synthetic groundwater. Similarly, the amount of Fe (II) removed by crushed concrete can be calculated as: This again is less than the amount removed in the sequencing batch test, which is 27.9 g Fe per kg crushed concrete. The possible reasons why col umn results were less than batch tests are: (1) bigger reactive materials were broken into smaller particles which have bigger surface reaction area during the shaking process, (2) bypass channel formation or preferential flow paths in column resulted from heterogeneity of the column (e.g., variation in particle size and shape), which can reduce the effective contact and reaction time of CCBM with Fe (II) solution (Kamolpornwijit et al., 2003), and ( 3 ) formation of slow movement water film on reactive mater ials surface, rendering the reactive material surface hard to access by Fe (II) (Horton and Hawkins, 1965).

PAGE 86

86 In the literature, Santomartino and Webb (2007) estimated the longevity of limestone columns to remove iron from AMD and got 28.3 g Fe precipitate d on 1 kg limestone. Komnitsas et al. (2004) reported iron removal from AMD by their limestone column at around 30 g Fe per kg limestone (calculated from their data). Both results are better than our current study. These discrepancies can be attributed to different column setup and different water characteristics. For example, our study used a much higher flow rate (5 mL/min vs. 2.75 mL/min in Santomartino and Webb (2007) and 2.3 mL/min in Komnitsas et al. 2004, respectively) so Fe (II) was given less time to interact with the reactive materials, resulting in less Fe (II) removal by unit weight material. Summary O verall two CCBMs removed Fe (II) from synthetic groundwater in the column system. The column tests have shown that both reactive materials are suc cessful in rem ediating Fe (II) from contaminated groundwater, achieving more than 99% and 99% iron removal for limestone and crushed concrete, respectively. Calcium siderite was possibly formed in limestone columns as precipitates. But this formation had i nsignificant impact on the system performance as porosity was only slightly reduced after the operation. In terms of longevity, before Fe (II) concentration of outlet sample increased above GWCTL, 4.06 g iron was removed by 1 kg limestone, and 3.80 g iron by 1 kg crushed concrete. These values are less than what we predicted from a sequencing batch test and may be the results of inhomogeneous particle size, p referential flow path and packing and formation of water films on the materials.

PAGE 87

87 Table 4 1 Charac teristics of reactive materials Ca content (%) Mg content (%) pH Average diameter (mm) Limestone 36.2 1.21 6.62 8.5 Crushed concrete 12.4 2.73 9.43 8.5 *Ca and Mg contents are calculated from digested sample. *Limestone and crushed concrete pH were me asured with EPA method 9045 Average diameter was calculated with 200 particles.

PAGE 88

88 Table 4 2 Calcium and iron contents in the precipitates formed on reactive media surface Limestone precipitates Crushed concrete precipitates Fe 11.2 % 32.1 Ca 34.2 % 13.4 *The precipitates were collected from column residues surface.

PAGE 89

89 Table 4 3 Porosity of columns before and after test (Limestone : columns a and b; C rushed concrete : columns c and d ) Initial porosity Final porosity Column a 0.432 0.405 Column b 0.430 0.393 Column c 0.440 0.410 Column d 0.450 0.412 *Initial porosity is the porosity before treatment, and final porosity is after the treatment.

PAGE 90

90 c Figure 4 1 Schematic diagram of the experimental column setup Water flows from bottom to top.

PAGE 91

91 Figure 4 2 P erformance of limestone in column test: Fe (II) concentration versus pore volume. (Column a and b) The initial Fe (II) concentration in synthetic groundwater is 50 mg/L.

PAGE 92

92 Figure 4 3 Performance of limestone in column test : pH versus pore volume. (Column a and b) pH drop to 6.4 after 290 pore volumes.

PAGE 93

93 Figure 4 4 Performance of limestone : Ca 2+ versus pore vol ume (a) and distance (b) Calcium concentration drops with pore volume increasing and port distance (from the entrance) increasing.

PAGE 94

94 Figure 4 5 Performance of limestone : total carbonate concentration versus pore volu me (a) and distance (b) Total c arbonate concentration drops with pore volume increasing and port distance (from the entrance) increasing.

PAGE 95

95 Figure 4 6 Performa nce of crushed concrete column : Fe (II) concentration ve rsus pore volume The initial Fe (II) concentration in synthetic groundwater is 50 mg/L.

PAGE 96

96 Figure 4 7 Performance of crushed concrete column: pH versus pore volume

PAGE 97

97 Figure 4 8 Correlation between pH and Fe (II) concentration in effluent samples

PAGE 98

98 Figure 4 9 Performance of crushed concrete: carbonate concentration versus pore volume (a) and distance (b) Total c arbonate concentration drop s with pore volume increasing and port distance (from the entrance) increasing.

PAGE 99

99 Figure 4 10 Performance of crushed concrete: Ca 2+ versus pore volume (a) and distance (b)

PAGE 100

100 CHAPTER 5 APPLICATION OF CALCIUM CA RBONATE BASED PERMEABLE REACTIVE BARRIERS FOR REMEDIATING IRON CONTAMINATED GROUNDWATER AT A LANDFILL SITE Introduction The extraction and treatment of contaminated ground water at landfill sites remains to be an extremely costly practice (Christensen et al., 2001 ) Some remediation techniques cost lots of energy and resource s but rarely achieve good results. Instead of remov ing the contaminated water from the subsurface for above ground treatment, recently, the emplacement of a permeable reactive barrier composed of calcium carbonated base materials (CCBMs) which intercepts the contaminated plume and transforms the contaminant to a solid form becomes more attractive. First, the in situ approach requires no above ground treatment facilities and the space ca n be returned to its original use. Second, there is no need for expensive above ground treatment, storage, transport, or disposal. Third, there are little or no operation and maintenance costs. Some PRBs have been constructed for contamination remediation ( Wilkin et al., 2009; Johnson et al., 2008 a, 2008b ; Lai et al., 2006; Puls et al., 1999; Indraratna et al., 2010; Beak and Wilkin, 2009 ; Slater and Binley, 2006 ) In Florida, elevated concentrations of iron have been observed in the groundwater monitoring wells at several solid waste disposal facilities. At some of these sites, the impacts of iron contamination on surface water systems have been noted as well. Permeable reactive barrier is therefore proposed as an effective and economical operational techn ique to remove dissolved iron from contaminated groundwater. However, although this technique has shown success for many pollutants including heavy metals (Indraratna et al., 2010) it has never been applied to iron contaminated

PAGE 101

101 groundwater. In addition, d ifferent from many other metals, iron has two oxidation states, Fe (II) and Fe (III), which may add further complications to the problem. In previous study l imestone and crushe d concrete showed the best removal effectiveness among all nine tested CCBMs an d were selected for later studies The author later showed i n Chapter 4 that both selected reactive materials demonstrated great performance in removing iron using a continuous flow column test with 4.06 g iron removed by 1 kg limestone and 3.80 g iron by 1 kg crushed concrete. In this chapter, the current study, for the first time, presents the field test with limestone and crushed concrete materials for iron remediation in groundwater addressing multiple objectives: 1) to evaluate whether a PRB system c omposed of CCBMs could remediate, in situ, groundwater contaminated with dissolved iron; and 2) to determine if results of prior laboratory studies were consistent with field observations and results. Method and Materials Site Description The research fiel d is located at 7219 Mobile Highway, Pensacola city, Florida, approximately 8 miles northwest of downtown Pensacola. It is a closed unlined sanitary landfill. Since closure of the landfill in 1982, routine groundwater monitoring has been conducted. High ir on concentration was first observed in 1987 in several groundwater monitoring wells and its concentration persistently exceeds 0.3 mg/L the groundwater cleanup target level (GWCTL) since then One of the existing monitoring wells DW4S (Figure 5 1 ) has hi gh dissolved iron concentration (20 50 mg/L) in groundwater for years. The area between DW4S and landfill was chosen as the target location for PRBs because both the groundwater level (0.5 1.0 meter below ground surface) and the surface aquifer (4. 5 met er s below ground) are shallow so that PRBs can easily reach

PAGE 102

102 aqui fer Hydraulic gradient at the water table was estimated using potentiometric surface interpreted from three point (DW4S, MW2, and PIW2) by manual groundwater elevation measurements (Devlin et al. 2003). The groundwater velocity for the study area was estimated to vary from 0.06 to 0.12 m/day (Geosyntec, 2004). Construction of Permeable Reactive Barriers and Monitoring Network Following the batch and column tests (in Chapter 2 and 4) for the removal effectiveness of reactive media two pilot scale PRBs were designed for installation in the test area in June and July 2009 The pilot PRBs were installed approximately 20 m hydraulically downgradient from the landfill boundary. They are 6 m long ( perpendicular to groundwater flow), 0.9 m wide (parallel to groundwater flow), and 4.5 m deep (cross the surface aquifer). The reactive barriers were installed over a week long using a dewatering pump and shear piles excavation method. The trenches were ba ckfilled with a 4 .2 m thick layer of reactive materials (from 4.5 m to 0.3 m below ground surface) and a 0.3 m thick layer of backfill soil s (from 0 to 0.3 m below ground surface). T op of the reactive materials is above the highest groundwater level (0.46m below ground surface) observed during the site study. One PRB (a) was filled with limestone (7 10 mm in diameter), and the other (b) filled with recycled crushed concrete (70 150 mm in diameter). The limestone PRB contains approximately 100 t of reactiv e materials with an estimated porosity of 0.5 and the crushed concrete PRB contains about 110 t of reactive materials with an estimated porosity of 0.55. A total of twenty eight groundwater monitoring wells were installed inside the trench, upgradient and downgradient of PRBs to evaluate reactive and hydraulic performance of the system. Figure 5 1 shows locations of the study area, PRBs and monitoring wells. All monitoring wells have a 1.5 m screen interval, and most of them

PAGE 103

103 were constructed using 5.08 cm PVC pipe s and screen s with a slot size of 0.051 cm. The upgradient and downgradient monitoring wells were installed at two depths: shallow (2.4 m deep, screen 0.9 2.4 m) and deep (4.2 m deep, screen 2.7 4.2 m). The wells in the trench have different dept h due to the construction restriction. Details of all wells are described in Appendix A Groundwater Sampling and Analysis Before taking groundwater samples, groundwater level was measured with a groundwater level meter (Heron PW 50 portable water meter) The groundwater samples were collected with a peristaltic pump following a purge at least three well volumes of ground water. T urbidity was less than 20 NTU for all collected groundwater samples with an average value of 10.2 NTU. pH, DO, ORP, Fe (II) conce ntration and conductivity were measured immediately in field. For metal contents samples were acidified immediately by nitric acid to pH less than 2 in the field and transported to the laboratory for analysis with ICP A ES. Bromide Tracer Test To further characterize groundwater flow through the PRBs, a tracer test was conducted 10 months after PRBs installation in April 2010. Approximately 100 L of DI water containing 5000 mg/L bromide was injected at AS2 and BS4 wells over a period of 1 hour. The amount of bromide tracer used here was determined following Johnson Both wells have a screen from 1 m to 2.5 m below ground. Groundwater samples were collected periodically from downgradient wells (AT12, AT13, AT14, AD1, AD5, AS6, AD7, AS8 AD9, AS10, AD11, BT12, BT 13, BT14, BD1, BD5, BS6, BD7, BS8, BD9, BD10, and Bd11) over 2 months. Bromide concentration was analyzed with Dionex OCS 1100.

PAGE 104

104 Results and Discussions Hydraulic Conditions of the PRB Area The groundwater level in PRB area was m onitored from June 2009 to May 2010, and the rainfall data was collected from a station close to the landfill site Results of the water elevation and rain fall change over time from selected wells are shown in Figure 5 2 From June 2009 to September 2009, a relative ly low water table was observed in the study area after the PRBs installation but the rainfall was relative high during that time. This is believed to be a result of 1) dewatering during the PRB installation, and/or 2) groundwater filling in the pore volume of PRBs when PRBs were first installed These two effects became negligible after September 2009 as steady groundwater flow has been observed from upgradient wells to downgradient wells of PRBs, as indicated by parallel water elevation in diff erent wells over time. This shows that the fluctuation in water elevation after achieving steady state was controlled primarily by the rainfall and evaporation Figure 5 3 illustrate s change in the water lever difference between upgradient well AD3 /BD3 tw o control wells (AD1 /BD1 and AD5 /BD5 ), and downgradient well AD7 /BD7 As can be seen from Figure 5 3 there is no significant change over time as the difference s are less than 0.1 inch. The change of water level of f our water table monitoring wells (AD1, A D5, BD1, BD5) have the same trend as the background wells (AD3 and BD3), suggesting no bypassing flow or blocking from either side of the PRBs. Iron Level in PRB A rea During the first year monitoring period, groundwater parameters were measured to evalua te the performance of PRBs. Table 5 1 presents selected groundwater parameters measured in background wells, down gradient wells, and PRBs wells in

PAGE 105

105 April 2010. Fe (II) concentration dropped from 23 mg/L in upgrad i ent to 2.0 mg/L in downgradient limestone P RB wells and 0.76 mg/L in downgradient concrete PRB wells. In addition, Fe (II) concentration was 0.39 mg/L in limestone PRB wells, and 0.3 mg/L in crushed concrete PRB wells. Although the Fe (II) concentration is still above the 0.3 mg/L GWCTL, it is less than 4.2 mg/L health risk based standard. Figure 5 3 (a) and (b) demonstrates that Fe (II) concentration in groundwater decreased dramatically after passing through the PRBs. Figure 5 4 shows the change of Fe (II) concentration of both PRBs over time. The se results show that limestone and concrete PRBs successfully remove d Fe (II) from groundwater. Before July 2009, the Fe (II) concentration was between 10 mg/L to 20 mg/L in all wells. After the installat ion of PRBs in July, the Fe (II) concentration s in P RBs and downgradient wells for both PRBs dropped to 1 to 2 mg/L due to the removal of Fe (II) by reactive materials, while the Fe (II) in upgradient wells increased after August 2009. This concentration rise was probably caused by 1) construction of wells that affected the water parameters, and 2) relative ly high rainfall in June and July 2009 that diluted groundwater. Since Fe (II) concentration is close to the total iron concentration under reducing condition, the total iron has a similar tre nd as Fe (II) (Figure 5 5 ). Similarly, Keimowitz et al. (2005) found that both Fe (II) and total iron concentrations in groundwater were the same when stud ying the groundwater geochemistry of a closed landfill in Maine and concluded the dominance o f Fe (II) in groundw ater. In this study, Fe(II) is also the dominant iron specie in groundwater. Fe (II) concentration s in two existing groundwater monitoring well s DW4S and MW2 w ere monitored before and after PRBs installation and the results are shown in Figure 5 6. It is apparent that Fe (II) was effectively removed in DW4S over 10 months.

PAGE 106

106 pH in PRB area Figur e 5 7 illustrates the change of groundwater pH values over time in the PRB area. The groundwater pH was not affected by the closed landfill, which remained in the ran ge of 6 .2 7.0 (averag ing 6.7 over the study period) in upgradient wells, which is close to the typical Fl orida groundwater pH (FDEP, 1992 ). For limestone PRB, pH inside limestone PRB was steady after the PRB installation (around 7.2). pH values of t he grou ndwater in downgradient wells slightly increased to around 6.9 after one and half months, later th a n Fe (II) removal in downgradient wells which occurred immediately after PRB installation pH values of different wells followed: upgradient < downgradient < PRB wells. Likewise, in another study using limestone passive barriers to treat acidic groundwater pH of downgradient wells was also found to be higher than that of the effluent (Hedin et al. 1994 ) The pH difference among groundwater up gradient of lim estone PRB, inside limestone PRB, and down gradient limestone PRB suggests that limestone materials not only remove the Fe (II) but also increase the groundwater pH. The groundwater inside crushed concrete PRB has consistently been alkaline (pH around 8.7 9 ) after P RB installation. As shown in figure 5 7 the pH difference between PRB wells and downgradient wells was fluctuating. In first three months, pH in PRB wells was higher th a n that in downgradient wells whereas a fter three months, pH in PRB wells be came lower. This phenomen on may be caused by the less amount of bypass groundwater flow and/ or crushed concrete mixed near monitoring wells due to the construction. Trend of the pH values in downgradient wells of concrete PRB in current study was very simi lar to a recent study using crushed concrete in the first year (Indraratna et al. 2010), although that study ha d a n acidic groundwater.

PAGE 107

107 Manganese in PRB area Manganese (Mn) levels w ere also monitored in all monitoring wells along with iron A relative ly hi gh Mn concentration was found in upgradient wells, which was about 2 mg/L, higher than the Groundwater cleanup target level of 0.05 mg/L. Figure 5 8 shows that Mn concentration in down gradient wells were substantially lower than the upgradient wells of bo th limestone and crushed concrete PRBs. This finding is consistent with the results reported by Aziz et al. (2002) who also noticed that limestone has the potential to remove Mn from contaminated water. Literature showed that elevated manganese concentrati on is usually associate d with iron concentration increa se in groundwater (Keimowitz et al. 2005). In Chapter 3, we concluded that Mn c ould inhibit the removal of Fe (II) by limestone material. However, th e relative ly low (2 mg/L) Mn concentration should no t be a n issue for Fe (II) removal in this study Other parameters in PRB area The DO concentrations in the entire PRBs area were relatively low and T able 5 1 lists DO measurements in April 2010. DO concentration s in PRBs area were around 1 2 mg/L, much low er than the normal surface aquifer DO level with an average of 5 mg/L (FDEP, 2008). Lower DO suggests the reducing condition of groundwater. Ap p endix figure s B 1 to 6 show the change of DO concentration s in all wells. The oxidation reduction potential ( ORP ) measurements of the area were less than 100 mV again, confirming the reducing condition The groundwater after passing through PRBs as can be seen in Figure s B 7 to 1 2 Figure B 13 to 18 shows the conductivity change over time. T he interesting finding is that conductivity in PRB wells is the lowest among all wells, probably due to the fact that limestone/crushed concrete can sorb and immobilize metal ions.

PAGE 108

108 In addition, t otal organic carbon was also analyzed for selected samples in April 2010 ( Table 5 1) TOC content stayed in the same range before and after groundwater pass ed through the PRBs. The maximum TOC concentrations reached 40 mg/L, higher than the Florida background groundwater TOC level of 2.2 to 14.1 mg/L (FDEP, 1992 ) w hich is possibly attributed to the organic containing leachate from this unlined landfill. Base on the results in Chapter 3, the relatively high organic matter concentration will reduce the removal effectiveness of Fe (II) and t his may be another reason fo r the relatively high residual Fe (II) concentrations in the down gradient groundwater. Tracer Test The trace test was conducted to confirm the groundwater flow through the PRBs. Over the 50 days monitoring period, bromide was detected inside PRB wells a nd in down gradient wells Figure 5 9 shows the maximum bromide tracer concentration detected in each well. For limestone PRB, the maximum Br concentrations were found in AS6, AD9, and AT12 wells, which are down gradient of the Br injection source AS2. How ever, b romide was not detected in AD1, AD5, AT14, AS8, and AD11 indicat ing only the perpendicular groundwater flow to the limestone PRB. Figure 5 8a shows the change of Br concentration over time. Although AT12 is the closest well to AS2, the maximum conc entration of Br was detected in well AS6 instead, because AS6 ha d the same screen depth as the injection well AS2. For crushed concrete PRB, similar result s were observed. M aximum Br concentrations were observed in BS8, BD11, and BT14 whereas no Br was det ected in BD1, BD5, BT14, BS8, and BD11. According to figure 5 8b, the maximum concentration of Br was detected in well BT14 for the same reason discussed above These results confi rmed the groundwater flow through the PRBs. The bromide concentration over t ime is shown in Ap p endix Figure B 19.

PAGE 109

109 Performances of PRBs Overall, both PRBs consistently removed Fe (II) for at least 12 months. The average Fe (II) concentrations of limestone and concrete PRBs are 23.2 10.3 mg/L and 24.6 9.6 mg/L in background wells,, 2.22 1.86 and 1.28 1.82 mg/L in down gradient wells, and 1.08 0.71 and 0.47 0.45 mg/L in PRB wells, respectively, showing substantial removal by the PRBs. Results from the PRB wells suggest that more than 95% Fe (II) have been removed by limestone PRB, and more than 98% Fe (II) by crushed concrete PRB. In addition, according to the results from downgradient wells, more than 90% Fe (II) has been removed by limestone PRB and more than 95% Fe (II) by crushed concrete PRB. As we discussed above, higher Fe (II) concentration in down gradient wells than PRB wells may be resulted from bypassing flow in the PRB since bypassing is a major concern for the PRB application (Hendoson and Demond, 2007). Furthermore, the presence of NOM can also negatively affect the Fe (I I) removal effectiveness and thus keep the iron concentration above GWCTL (0.3 mg/L) since NOM can form complexes with Fe (II) and retain Fe (II) in the aqueous phase. Although we showed in our earlier study (Chapter 3) that manganese can also act as a com peting ion for iron removal by CCBM, the really low Mn concentration measured in field suggests little impact of Mn on the PRB performance. Based on the iron removal effectiveness estimated by column tests in Chapter 4, we can also calculate the longevity of PRBs. Column test results in Chapter 3 suggest that 4.06 g iron was removed by 1 kg limestone, and 3.80 g iron by 1 kg crushed concrete. Base d on the estimated groundwater flow 0.1 m/day to 0.2 m/day in Klondike Landfill the groundwater retention time in PRBs is 4.5 9 days Therefore, the limestone PRB is estimated to work 4.5 9 day multiplied by 290 pore volumes, which is 3.6 7.1

PAGE 110

110 years, and the Crushed concrete 3.0 5.9 years (4.5 9 days multiplied by 240 pore volumes). These results indicate that both PRBs are good for long term treatment of iron contamination, while limestone PRB appears to have greater longevity. Periodical monitoring of Fe (II) concentrations in the studied sites is currently on going to validate the calculation and ensure the e ffectiveness of the systems. Summary Both limestone and crushed concrete PRBs performed very well in the first year of operation by effectively removing excessive Fe (II) from the target groundwater and improving the water quality. Moreover, neither PRB sh owed deteriorating performance over the 14 month monitoring period. In PRB wells, more than 95% Fe (II) have been removed by limestone PRB, and more than 98% Fe (II) by crushed concrete PRB, and in downgradient wells, more than 90% Fe (II) have been remove d by limestone PRB, and more than 95% Fe (II) by crushed concrete PRB. Although the downgradient iron concentrations were still higher than the 0.3 mg/L GWTCL, they all dropped to below 4.2 mg/L health risk based standard which is a substantial improvemen t from the influent. The groundwater bypassing flow in the PRBs and presence of high TOC are believed to be two leading reasons for the residual iron concentration in the downgradient wells. pH of the groundwater remained in the neutral range before and af ter the limestone PRB however, pH of crushed concrete PRB rose above 9, which may cause a concern or issue in the future

PAGE 111

111 Table 5 1 Groundwater parameters in PRB area (April 2010) Up gradient wells Down gradient wells of Limestone PRB In PRB wells of L imestone PRB Down gradient wells of Crushed concrete PRB In PRB wells of Crushed concrete PRB ORP (mV) 78 103 140 153 80 pH 6.67 6.90 7.16 9.2 8.79 DO (mg/L) 1.03 0.95 1.04 0.87 0.94 Alkalinity (mg/L as CaCO 3 ) 130 450 323 376 463 Fe(II) (mg/L) 23 .0 1.98 0.39 0.76 0.3 SO 4 2 (mg/L) 289 (AD3) 274 (AD7) 228 (AT13) 242 (BD7) 223 (BT13) Na + (mg/L) 17.5 15.3 15.4 21.3 19.3 Ca 2+ (mg/L) 9.2 32.1 30.1 32.1 29.8 Cl (mg/L) 22 (AD3) 20(AD7) 23 (AT13) 18 (BD7) 21 (BT13) TOC (Total organic carbon, mg/L) 24.5 13.2 12.6 18.9 15.3 Most of parameters are presented as average values, and SO 4 2 and Cl are presented as single well sample.

PAGE 112

112 Figure 5 1 PRBs area map, showing locations of monitoring wells, existing monitoring wells (DW4S and MW2), and two PR Bs location. The above map is plain view of PRB area, and the below map is a plain view of all constructed PRBs and wells.

PAGE 113

113 Figure 5 2 Change in g roundwater table elevation ( above sea surface level) and precipitation over time. Limestone PRB (upper) and Crushed concrete PRB (lower) (The precipitation data was collected from a station near Klondike landfill.)

PAGE 114

114 Figure 5 3 Difference of g roundwater table elevation (Above sea surf ace lev el) in Limestone PRB (upper) and Crushed concrete PRB (lower) The relative indicates the

PAGE 115

115 Figure 5 4 Change of Fe (II) concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The Fe (II) concentration decreased after passing through PRBs.

PAGE 116

116 Figure 5 5 Change of t o tal iron concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB. The total iron concentration decreased after passing through PRBs.

PAGE 117

117 Figure 5 6 Change of Fe (II) concentration in Well DW4S an d MW2 over time. DW4S was affected by PRBs since the Fe (II) concentration dropped.

PAGE 118

118 Figure 5 7 Change of pH in PRBs over time. (a) limestone PRB a nd (b) crushed concrete PRB The pH increased after passing through P RBs.

PAGE 119

119 Figure 5 8 Change of Mn concentration in PRBs over time. (a) limestone PRB and (b) crushed concrete PRB The Mn concentration decreased after passing through PRBs.

PAGE 120

120 Figur e 5 9 Maximum bromide concentration observed in each monitoring wells ( in mg/L. N umber s in parentheses are maximum Br concentration) Groundwater flow passed through PRBs.

PAGE 121

121 CHAPTER 6 CONCLUSION S ummary Elevated iron concentration s w ere detected in groundwater underneath Florida landfill si tes T here is a need to determine effective remediat ion strategy at those site s. A lthough permeable reactive barrier s (PRB) ha ve been widely used to tr eat contaminated groundwater, it has never been applied to remove Fe (II) prior to this study. Due to it s prove n success for other contaminants (e. g., heavy metals, PCBs) and relatively low cost, this technique was proposed for in situ iron treatment. T he first task wa s the selection of proper reactive materials. Since calcium carbonate based materials (CCB Ms) have the potential to remove dissolved iron from groundwater, various CCBMs were tested for their effectiveness to remove iron. The final Fe (II) concentration in all treated solution droppe d from the initial 50 mg/L to below 0.03 mg/L Compared to th e drinking water limit and GWCTL (0.3 mg/L), all carbonate materials are qualified as a reactive material. Among the tested materials, limestone showed the best performance in removing ferrous iron from Fe (II) contaminated groundwater (more than 99%) and was thus selected for use in further studies In addition, although slightly less effective (99%) crushed concrete was also selected because it is a recycled material which is sustainable and environmental friendly. The second experiment evaluated the eff ects of environmental factors on the iron removal process. Our results suggest that at a l ower pH (pH = 4) can reduce the removal effectiveness of Fe (II) by limestone whereas neutral pH range, the rem oval effectiveness is promoted. Both Na + and Ca 2+ have n egative impacts on the removal

PAGE 122

122 process because they can increase the ionic strength and have a positive charge Mn 2+ can be removed by limestone, so it is a competitive ion for the Fe (II) removal reaction Carboxyl groups in the NOM molecules can complex with Fe (II) in the solution, and therefore, the presence of NOM reduces Fe (II) uptake by limestone and inhibits the removal process The third experiment wa s designed to assessing the loading capacity and longevity of CCBMs using a continuous flow colum n test. The results demonstrate that both reactive materials are successful in remediating ferrous iron from contaminated groundwater. Being to the Fe (II) concentration s in the outlet of the column reaching the GWCTL, 4.06 g iron was removed per 1 kg lim estone, and 3.80 g iron per 1 kg crushed concrete. Both values are less than what we predicted from a sequencing batch test Precipitates formed during the removal reaction and calcium siderite was identified as a possible product In the final field test both limestone and crushed concrete PRBs improved water quality by effectively re moving of Fe (II). Neither PRB showed an indication of deteriorationg performance over the 14 month monitoring period. In wells, Fe (II) concentrations were reduced to less than 2 mg/L by limestone PRB, and less than 0.5 mg/L by crushed concrete PRB. I n the down gradient wells, Fe (II) concentrations were reduced to less than 1 mg/L by limestone PRB, and less than 0.5 mg/L b y crushed concrete PRB. pH of groundwater which pass ed through the limestone PRB stayed in neutral range, which the pH in crushed concrete PRB increased above 9, which maybe a concern or issue.

PAGE 123

123 Conclusion The four experiments described in detail in the previous chapters studied the use of CCBMs in PRB s to remove Fe (II) from contaminated groundwater. The following specific conclusions were reached: All CCBMs have the ability to remove Fe (II) from contaminated groundwater. Limestone and crushed concrete have the best removal effectiveness for remov ing Fe (I I) from synthetic groundwater. The pH of limestone treated groundwater remained in the neutral range (6 8.5) while that of crushed concrete treated groundwater became alkaline (>9) From a kinetic study, the removal process of Fe (II) by limestone appears to be a two step process, co nsisting of a rapid sorption step followed by a relative ly slow co precipitation. pH can affect removal effectiveness as acidic condition s inhibit the removal process. Na and Ca slightly inhibit the removal process and reduce th e removal effectiveness. Mn (II) competes with Fe (II) for the sorption sites on the reactive media NOM s mobilize Fe (II) in the water phase and keep the iron concentration above the cleanup target level. The final Fe (II) in solution can be predicted by the NOM charge density. Loading capacities of the limestone and crushed concrete materials are 4.06 g iron by 1 kg limestone and 3.80 g iron by 1 kg crushed concrete as determined by a column test. The p orosity of the columns packed by l imestone and crus hed concrete was only slightly reduced after treatment. Fe (II) concentration has been reduced to less tha n 2 mg/L by limestone PRB, and less than 1 mg/L by crushed concrete PRB. Overall, the CCBMs are a good reactive material in permeable reactive barrie r s (PRBs) for removing dissolved iron from contaminated groundwater. CCBMs PRB is a n effective and economic remediation method to remove dissolved iron from groundwater. l years. CCBMs PRB are suitable for use when the contaminated site has shallow

PAGE 124

124 groundwater level and when groundwater has low flow rate and low NOM content. In deep groundwater level area, the construction has the difficulty to approach the depth. Higher groundwater flow rate will decrease the groundwater retention time in PRBs, that required thicker reactive material layer. Higher NOM concentration will remain more Fe (II) in water phase to reduce the removal effectiveness. Future Work The results from t h is PhD research will help understand the environmental effects of limestone and crushed concrete permeable reactive barriers treatment system. In order to further increase the removal effectiveness and reduce the final Fe (II) concentration, a study using mixed reactive materials should be conducted to remove more Fe (II) from groundwater. A combination of limestone and crushed concrete could be studied for better removal effectiveness. Other reactive materials which can remove NOM from groundwater could al so be added into the mixed material. For better predict ing the longevity of permeable reactive materials, an extended monitoring period could help obtain the breakthrough time of PRBs in field

PAGE 125

125 APPENDIX A ADDITIONAL PRBS INFORMATION Figure A 1 Location of active gr oundwater monitoring points at the Klondike landfill. ( PRBs area was marked as EA 1 in map ) PIW 2 MW 2 DW 2I DW 4S MW 3 MW 3A DW 5S DW 5I MW 5A DW 6I DW 6S DW 6D MW 5 MW 6 MW 7A PIW 3 Monitoring Well (Deep) Monitoring Well ( Shallow) Monitoring Well (Intermediate) Feb. 2007 Iron conc. (mg/L) < 4.2 4.2 20 20 50 > 50 PIW 5 EA 1 EA 2 EA 3 (VZA ) (VZA ) (PRB) North Eleven Mile Creek

PAGE 126

126 Table A 1 Details of groundwater monitoring w ells installed for PRBs Well ID Depth (ft) Casing information Screen Info rmation Surface completion Well Diameter (inch) Material Screen Length (ft) Filter Pack (mesh) Slot Size Material AD1 14 2 PVC 5 0.01 20/30 PVC Stick up AS2 8 2 PVC 5 0.01 20/30 PVC Stick up AD3 14 2 PVC 5 0.01 20/30 PVC Stick up AS4 8 2 PVC 5 0.01 20/30 PVC Stick up AD5 14 2 PVC 5 0.01 20/30 PVC Stick up AS6 8 2 PVC 5 0.01 20/30 PVC Flush Mount AD7 14 2 PVC 5 0.01 20/30 PVC Stick up AS8R 8 2 PVC 5 0.01 20/30 PVC Stick up AD9 14 2 PVC 5 0.01 20/30 PVC Flush Mount AS10 8 2 PVC 5 0.01 20/30 PVC Stick up AD11 14 2 PVC 5 0.01 20/30 PVC Stick up AT12 12 2 HDPE 5 None Drilled holes HDPE Stick up AT13 12 2 HDPE 5 None Drilled holes HDPE Stick up AT14 12 2 HDPE 5 None Drilled holes HDPE Stick up BD1 14 2 PVC 5 0.01 20/30 PVC Stick up BS2 8 2 PVC 5 0.01 20/30 PVC Stick up BD3 14 2 PVC 5 0.01 20/30 PVC Stick up BS4 8 2 PVC 5 0.01 20/30 PVC Stick up BD5 14 2 PVC 5 0.01 20/30 PVC Stick up BS6R 8 2 PVC 5 0.01 20/30 PVC Stick up BD7R 14 2 PVC 5 0.01 20/30 PVC Stick up BS8R 8 2 PVC 5 0.01 20/30 PVC Stick up BD9 14 2 PVC 5 0.01 20/30 PVC Stick up BS10 8 2 PVC 5 0.01 20/30 PVC Stick up BD11 14 2 PVC 5 0.01 20/30 PVC Stick up BT12 6 2 HDPE 5 None Drilled holes HDPE Stick up BT13 9 2 HDPE 5 None Drilled holes HDPE Stick up BT14 11 2 HDPE 5 None Dr illed holes HDPE Stick up

PAGE 127

127 APENDIX B ADDITIONAL GROUNDWATER DATA AND FIGURES Figure B 1 Change of d issolved oxygen level over time in background wells ( L imestone PRB)

PAGE 128

128 Figure B 2 Change of d issolved oxygen level over time in down gradient wells ( L imestone PRB)

PAGE 129

129 Figure B 3 Change of d issolved oxygen level over time in PRB wells (L imestone PRB)

PAGE 130

130 Figure B 4 Change of d issolved oxygen level over time in background wells ( Crushed concrete PRB)

PAGE 131

131 Figure B 5 Change of d issolved oxygen level over time in down gradient wells ( Crushed concrete PRB)

PAGE 132

132 Figure B 6 Change of d issolved oxygen level over time in PRB wells ( Crushed concrete PRB)

PAGE 133

133 Figure B 7 Change of p xidation reduction potential (ORP) over time in background wells (Limestone PRB)

PAGE 134

134 Figure B 8 Change of o xidation reduction potential (ORP) over time in down gradient wells (Limestone PRB)

PAGE 135

135 Figure B 9 Change of o xidation reduction potential (ORP) over time in PRB wells (Li mestone PRB)

PAGE 136

136 Figure B 10 Change of o xidation reduction potential (ORP) over time in background wells ( Crushed concrete PRB)

PAGE 137

137 Figure B 11 Change of o xidation reduction potential (O RP) over time in down gradient wells ( Crushed concrete PRB)

PAGE 138

138 Figure B 12 Change of o xidation reduction potential (ORP) over time in PRB wells ( Crushed concrete PRB)

PAGE 139

139 Figure B 13 Ch ange of c onductivity over time in background wells (Limestone PRB)

PAGE 140

140 Figure B 14 Change of c onductivity over time in down gradient wells (Limestone PRB)

PAGE 141

141 Figure B 15 Change of c ond uctivity over time in PRB wells (Limestone PRB)

PAGE 142

142 Figure B 16 Change of c onductivity over time in background wells ( Crushed concrete PRB)

PAGE 143

143 Figure B 17 Change of c onductivity over ti me in down gradient wells ( Crushed concrete PRB)

PAGE 144

144 Figure B 18 Change of c onductivity over time in PRB wells ( Crushed concrete PRB)

PAGE 145

145 Figure B 19 Change of b romide tracer concentra tion over time in monitored wells (a ) Limestone PRB area and ( b ) crushed concrete PRB area

PAGE 146

146 LIST OF REFERENCES Apell, J. N., and Boyer, T. H. (2010). "Combined ion exchange treatment for removal of dissolved organic matter and hardness." Water Research, 44(8), 2419 2430. Aziz, H. A., Othman, N., Yusuff, M. S., Basri, D. R. H., Ashaari, F. A. H., Adlan, M. N., Othman, F., Johari, M., and Perwira, M. (2001). "Removal of copper from water using limestone filtration technique Determination of mechanism of r emoval." Environment International, 26(5 6), 395 399. Aziz, H. A., and Smith, P. G. (1992). "The Influence of Ph and Coarse Media on Manganese Precipitation from Water." Water Research, 26(6), 853 855. Aziz, H. A., and Smith, P. G. (1996). "Removal of mang anese from water using crushed dolomite filtration technique." Water Research, 30(2), 489 492. Beak, D. G., and Wilkin, R. T. (2009). "Performance of a zerovalent iron reactive barrier for the treatment of arsenic in groundwater: Part 2. Geochemical modeli ng and solid phase studies." Journal of Contaminant Hydrology, 106(1 2), 15 28. Benjamin, M. M., and Leckie, J. O. (1981a). "Competitive Adsorption of Cd, Cu, Zn, and Pb on Amorphous Iron Oxyhydroxide." Journal of Colloid and Interface Science, 83(2), 410 419. Benjamin, M. M., and Leckie, J. O. (1981b). "Multiple Site Adsorption of Cd, Cu, Zn, and Pb on Amorphous Iron Oxyhydroxide." Journal of Colloid and Interface Science, 79(1), 209 221. Bilek, F. (2006). "Column tests to enhance sulphide precipitation wi th liquid organic electron donators to remediate AMD influenced groundwater." Environmental Geology, 49(5), 674 683. Binley, A., Slater, L. D., Fukes, M., and Cassiani, G. (2005). "Relationship between spectral induced polarization and hydraulic properties of saturated and unsaturated sandstone." Water Resources Research, 41(12), Bismarck, A., Springer, J., Mohanty, A. K., Hinrichsen, G., and Khan, M. A. (2000). "Characterization of several modified jute fibers using zeta potential measurements." Colloid and Polymer Science, 278(3), 229 235. Bjelopavlic, M., Newcombe, G., and Hayes, R. (1999). "Adsorption of NOM onto activated carbon: Effect of surface charge, ionic strength, and pore volume distribution." Journal of Colloid and Interface Science, 210(2), 271 280. Boyer, T. H., and Singer, P. C. (2008). "Stoichiometry of removal of natural organic matter by ion exchange." Environmental Science & Technology, 42(2), 608 613.

PAGE 147

147 Cave, K., and Talens Alesson, F. I. (2005). "Comparative effect of Mn(II) and Fe(III ) as activators and inhibitors of the adsorption of other heavy metals on calcite." Colloids and Surfaces a Physicochemical and Engineering Aspects, 268(1 3), 19 23. Christensen, T. H., Kjeldsen, P., Bjerg, P. L., Jensen, D. L., Christensen, J. B., Baun, A ., Albrechtsen, H. J., and Heron, C. (2001). "Biogeochemistry of landfill leachate plumes." Applied Geochemistry, 16(7 8), 659 718. Chung, H. I., Kim, S. K., Lee, Y. S., and Yu, J. (2007). "Permeable reactive barrier using atomized slag material for treatm ent of contaminants from landfills." Geosciences Journal, 11(2), 137 145. Comstock, S. E. H., Boyer, T. H., Graf, K. C., and Townsend, T. G. (2010). "Effect of landfill characteristics on leachate organic matter properties and coagulation treatability." Ch emosphere, 81(7), 976 983. Cravotta, C. A. (2003). "Size and performance of anoxic limestone drains to neutralize acidic mine drainage." Journal of Environmental Quality, 32(4), 1277 1289. Darbi, A., Viraraghavan, T., Butler, R., and Corkal, D. (2003). "Co lumn studies on nitrate removal from potable water." Water Air and Soil Pollution, 150(1 4), 235 254. Das, B., Hazarika, P., Saikia, G., Kalita, H., Goswami, D. C., Das, H. B., Dube, S. N., and Dutta, R. K. (2007). "Removal of iron from groundwater by ash: A systematic study of a traditional method." Journal of Hazardous Materials, 141(3), 834 841. Davis, J. A., Fuller, C. C., and Cook, A. D. (1987). "A Model for Trace Metal Sorption Processes at the Calcite Surface Adsorption of Cd 2+ and Subsequent Soli d Solution Formation." Geochimica Et Cosmochimica Acta, 51(6), 1477 1490. deLemos, J. L., Bostick, B. C., Renshaw, C. E., Sturup, S., and Feng, X. H. (2006). "Landfill stimulated iron reduction and arsenic release at the Coakley Superfund Site (NH)." Envir onmental Science & Technology, 40(1), 67 73. Dikinya, O., Hinz, C., and Aylmore, G. (2008). "Decrease in hydraulic conductivity and particle release associated with self filtration in saturated soil columns." Geoderma, 146(1 2), 192 200. DOD. (2002). "Pass ive reactive barrier." Departmrnt of defense. Ellis, D., Bouchard, C., and Lantagne, G. (2000). "Removal of iron and manganese from groundwater by oxidation and microfiltration." Desalination, 130(3), 255 264. EPA. (2001). "A Citizen's Guide to Permeable R eactive Barriers." FDEP. (1992). "Background hydrogeochemistry." F lorida D epartment E nvironmental P rotestion

PAGE 148

148 Franklin, M. L., and Morse, J. W. (1983). "The Interaction of Manganese(Ii) with the Surface of Calcite in Dilute Solutions and Seawater." Marine Chemistry, 12(4), 241 254. Furukawa, Y., Kim, J. W., Watkins, J., and Wilkin, R. T. (2002). "Formation of ferrihydrite and associated iron corrosion products in permeable reactive barriers of zero valent iron." Environmental Science & Technology, 36(24), 5 469 5475. Gadagbui, B., and Roberts, S. (2002). "Development of Cleanup Target Levels (CTLs) For Chapter 62 777, F.A.C." Center for Environmental & Human Toxicology. University of Florida Geosyntec. (2006). "Feasibility evaluation klondike landfill." Tampa FL. Golab, A. N., Peterson, M. A., and Indraratna, B. (2006). "Selection of potential reactive materials for a permeable reactive barrier for remediating acidic groundwater in acid sulphate soil terrains." Quarterly Journal of Engineering Geology and Hyd rogeology, 39, 209 223. Golab, A. N., Peterson, M. A., and Indraratna, B. (2009). "Selection of permeable reactive barrier materials for treating acidic groundwater in acid sulphate soil terrains based on laboratory column tests." Environmental Earth Scien ces, 59(1), 241 254. Gomari, K. A. R., Hamouda, A. A., and Denoyel, R. (2006). "Influence of sulfate ions on the interaction between fatty acids and calcite surface." Colloids and Surfaces a Physicochemical and Engineering Aspects, 287(1 3), 29 35. Good, N E., Winget, G. D., Winter, W., Connolly, T. N., Izawa, S., and Singh, R. M. M. (1966). "Hydrogen Ion Buffers for Biological Research." Biochemistry, 5(2), 467 &. Hedin, R. S., Watzlaf, G. R., and Nairn, R. W. (1994). "Passive Treatment of Acid Mine Drain age with Limestone." Journal of Environmental Quality, 23(6), 1338 1345. Henderson, A. D., and Demond, A. H. (2007). "Long term performance of zero valent iron permeable reactive barriers: A critical review." Environmental Engineering Science, 24(4), 401 4 23. Indraratna, B., Regmi, G., Nghiem, L. D., and Golab, A. (2010). "Performance of a PRB for the Remediation of Acidic Groundwater in Acid Sulfate Soil Terrain." Journal of Geotechnical and Geoenvironmental Engineering, 136(7), 897 906. Islam, J., and Sin ghal, N. (2004). "A laboratory study of landfill leachate transport in soils." Water Research, 38(8), 2035 2042. Jensen, D. L., Boddum, J. K., Tjell, J. C., and Christensen, T. H. (2002). "The solubility of rhodochrosite (MnCO3) and siderite (FeCO3) in ana erobic aquatic environments." Applied Geochemistry, 17(4), 503 511.

PAGE 149

149 Jeon, B. H., Dempsey, B. A., Burgos, W. D., Barnett, M. O., and Roden, E. E. (2005). "Chemical reduction of U(VI) by Fe(II) at the solid water interface using natural and synthetic Fe(III) oxides." Environmental Science & Technology, 39(15), 5642 5649. Johnson, R. L., Thoms, R. B., Johnson, R. O., and Krug, T. (2008a). "Field evidence for flow reduction through a zero valent iron permeable reactive barrier." Ground Water Monitoring and Reme diation, 28(3), 47 55. Johnson, R. L., Thoms, R. B., Johnson, R. O., Nurmi, J. T., and Tratnyek, P. G. (2008b). "Mineral precipitation upgradient from a zero valent iron permeable reactive barrier." Ground Water Monitoring and Remediation, 28(3), 56 64. Ka molpornwijit, W., Liang, L., West, O. R., Moline, G. R., and Sullivan, A. B. (2003). "Preferential flow path development and its influence on long term PRB performance: column study." Journal of Contaminant Hydrology, 66(3 4), 161 178. Kamolpornwijit, W., Liang, L. Y., Moline, G. R., Hart, T., and West, O. R. (2004). "Identification and quantification of mineral precipitation in Fe 0 filings from a column study." Environmental Science & Technology, 38(21), 5757 5765. Keimowitz, A. R., Simpson, H. J., Stute, M., Datta, S., Chillrud, S. N., Ross, J., and Tsang, M. (2005a). "Naturally occurring arsenic: Mobilization at a landfill in Maine and implications for remediation." Applied Geochemistry, 20(11), 1985 2002. Keimowitz, A. R., Zheng, Y., Chillrud, S. N., Ma illoux, B., Jung, H. B., Stute, M., and Simpson, H. J. (2005b). "Arsenic redistribution between sediments and water near a highly contaminated source." Environmental Science & Technology, 39(22), 8606 8613. Kjeldsen, P., Barlaz, M. A., Rooker, A. P., Baun, A., Ledin, A., and Christensen, T. H. (2002). "Present and long term composition of MSW landfill leachate: A review." Critical Reviews in Environmental Science and Technology, 32(4), 297 336. Komnitsas, K., Bartzas, G., and Paspaliaris, I. (2004). "Effici ency of limestone and red mud barriers: laboratory column studies." Minerals Engineering, 17(2), 183 194. Krasner, S. W., Croue, J. P., Buffle, J., and Perdue, E. M. (1996). "Three approaches for characterizing NOM." Journal American Water Works Associatio n, 88(6), 66 79. Lai, K. C. K., Lo, I. M. C., Birkelund, V., and Kjeldsen, P. (2006). "Field monitoring of a permeable reactive barrier for removal of chlorinated organics." Journal of Environmental Engineering Asce, 132(2), 199 210.

PAGE 150

150 Lee, Y. J., Elzinga, E J., and Reeder, R. J. (2005). "Cu(II) adsorption at the calcite water interface in the presence of natural organic matter: kinetic studies and molecular scale characterization." Geochimica Et Cosmochimica Acta, 69(1), 49 61. Leenheer, J. A., Brown, G. K. MacCarthy, P., and Cabaniss, S. E. (1998). "Models of metal binding structures in fulvic acid from the Suwannee River, Georgia." Environmental Science & Technology, 32(16), 2410 2416. Lepp, H. (1975). Geochemistry of iron, Dowden, Hutchingon & Ross, Inc. Lin, Y. P., Singer, P. C., and Aiken, G. R. (2005). "Inhibition of calcite precipitation by natural organic material: Kinetics, mechanism, and thermodynamics." Environmental Science & Technology, 39(17), 6420 6428. Lorens, R. B. (1981). "Sr, Cd, Mn and Co Distribution Coefficients in Calcite as a Function of Calcite Precipitation Rate." Geochimica Et Cosmochimica Acta, 45(4), 553 561. Lovley, D. R. (1991). "Dissimilatory Fe(Iii) and Mn(Iv) Reduction." Microbiological Reviews, 55(2), 259 287. Lovley, D. R., Holmes, D. E., and Nevin, K. P. (2004). "Dissimilatory Fe(III) and Mn(IV) reduction." Advances in Microbial Physiology, Vol. 49, 49, 219 286. Luther, G. W., Shellenbarger, P. A., and Brendel, P. J. (1996). "Dissolved organic Fe(III) and Fe(II) complexes i n salt marsh porewaters." Geochimica Et Cosmochimica Acta, 60(6), 951 960. Mackenzie, P. D., Horney, D. P., and Sivavec, T. M. (1999). "Mineral precipitation and porosity losses in granular iron columns." Journal of Hazardous Materials, 68(1 2), 1 17. Mart in Garin, A., Van Cappellen, P., and Charlet, L. (2003). "Aqueous cadmium uptake by calcite: A stirred flow through reactor study." Geochimica Et Cosmochimica Acta, 67(15), 2763 2774. Masscheleyn, P. H., Delaune, R. D., and Patrick, W. H. (1991). "Effect o f Redox Potential and Ph on Arsenic Speciation and Solubility in a Contaminated Soil." Environmental Science & Technology, 25(8), 1414 1419. Mettler, S., Abdelmoula, M., Hoehn, E., Schoenenberger, R., Weidler, P., and von Gunten, U. (2001). "Characterizati on of iron and manganese precipitates from an in situ ground water treatment plant." Ground Water, 39(6), 921 930. Mettler, S., Wolthers, M., Charlet, L., and von Gunten, U. (2009). "Sorption and catalytic oxidation of Fe(II) at the surface of calcite." Ge ochimica Et Cosmochimica Acta, 73(7), 1826 1840.

PAGE 151

151 Munter, R., Ojaste, H., and Sutt, J. (2005). "Complexed iron removal from groundwater." Journal of Environmental Engineering Asce, 131(7), 1014 1020. Munter, R., Overbeck, P., and Sutt, J. (2008). "Which is the best oxidant for complexed iron removal from groundwater: The Kogalym case." Ozone Science & Engineering, 30(1), 73 80. Nano, G. V., and Strathmann, T. J. (2006). "Ferrous iron sorption by hydrous metal oxides." Journal of Colloid and Interface Science 297(2), 443 454. Parisio, S., Keimowitz, A. R., Simpson, H. J., Lent, A., and Blackman, V. (2006). "Arsenic rich iron floc deposits in seeps downgradient of solid waste landfills." Soil & Sediment Contamination, 15(5), 443 453. Phillips, D. H., Gu, B., W atson, D. B., Roh, Y., Liang, L., and Lee, S. Y. (2000). "Performance evaluation of a zerovalent iron reactive barrier: Mineralogical characteristics." Environmental Science & Technology, 34(19), 4169 4176. Ponnamperuma, F. N. (1972). "The chemistry of sub merged soils." Adv. Agron. 24, 29 96. Puls, R. W., Paul, C. J., and Powell, R. M. (1999). "The application of in situ permeable reactive (zero valent iron) barrier technology for the remediation of chromate contaminated groundwater: a field test." Applie d Geochemistry, 14(8), 989 1000. Rao, N. S. (2008). "Iron content in groundwaters of visakhapatnam environs, Andhra Pradesh, india." Environmental Monitoring and Assessment, 136(1 3), 437 447. Reilly, C. (2002). Metal contamination of food, Blackwell Scien ce Ltd, Malden, MA. Ritchie, J. D., and Perdue, E. M. (2003). "Proton binding study of standard and reference fulvic acids, humic acids, and natural organic matter." Geochimica Et Cosmochimica Acta, 67(1), 85 96. Rouff, A. A., Elzinga, E. J., Reeder, R. J. and Fisher, N. S. (2005). "Effect of aging on Pb(II) sorption at the calcite water interface." Geochimica Et Cosmochimica Acta, 69(10), A609 A609. Santomartino, S., and Webb, J. A. (2007). "Estimating the longevity of limestone drains in treating acid mi ne drainage containing high concentrations of iron." Applied Geochemistry, 22(11), 2344 2361. Sharma, S. K., Greetham, M. R., and Schippers, J. C. (1999). "Adsorption of iron(II) onto filter media." Journal of Water Supply Research and Technology Aqua, 48( 3), 84 91. Sterner, P. (1997). "A Laboratory Study of the Geochemistry of Anoxic Limestone Drains," West Virginia University.

PAGE 152

152 Stumm, W., and Morgan, J. J. (1996). Aquatic Chemistry, John Wiley & Sons, Inc., New York. Tegethoff, W. (2001). Calcium Carbonate : from the cretaceous period into the 21st century Birkhauser Verloag. Thornton, F. C. (1995). "Manganese Removal from Water Using Limestone Filled Tanks." Ecological Engineering, 4(1), 11 18. Turner, B. D., Binning, P. J., and Sloan, S. W. (2008). "A calc ite permeable reactive barrier for the remediation of Fluoride from spent potliner (SPL) contaminated groundwater." Journal of Contaminant Hydrology, 95(3 4), 110 120. Van Nooten, T., Diels, L., and Bastiaens, L. (2008). "Design of a Multifunctional Permea ble Reactive Barrier for the Treatment of Landfill Leachate Contamination: Laboratory Column Evaluation." Environmental Science & Technology, 42(23), 8890 8895. Van Nooten, T., Lieben, F., Dries, J., Pirard, E., Springael, D., and Bastiaens, L. (2007). "Im pact of microbial activities on the mineralogy and performance of column scale permeable reactive iron barriers operated under two different redox conditions." Environmental Science & Technology, 41(16), 5724 5730. Vinson, M. D., Arvidson, R. S., and Luttg e, A. (2007). "Kinetic inhibition of calcite (104) dissolution by aqueous manganese(II)." Journal of Crystal Growth, 307(1), 116 125. Wajon, J. E., Ho, G. E., and Murphy, P. J. (1985). "Rate of Precipitation of Ferrous Iron and Formation of Mixed Iron Calcium Carbonates by Naturally Occurring Carbonate Materials." Water Research, 19(7), 831 837. Weber, K. A., Achenbach, L. A., and Coates J. D. (2006). "Microorganisms pumping iron: anaerobic microbial iron oxidation and reduction." Nature Reviews Microbiology, 4(10), 752 764. Wilkin, R. T., Acree, S. D., Ross, R. R., Beak, D. G., and Lee, T. R. (2009). "Performance of a zerovalent iron re active barrier for the treatment of arsenic in groundwater: Part 1. Hydrogeochemical studies." Journal of Contaminant Hydrology, 106(1 2), 1 14. Williams, A. G. B., Gregory, K. B., Parkin, G. F., and Scherer, M. M. (2005). "Hexahydro 1,3,5 trinitro 1,3,5 t riazine transformation by biologically reduced ferrihydrite: Evolution of Fe mineralogy, surface area, and reaction rates." Environmental Science & Technology, 39(14), 5183 5189. Zachara, J. M., Kittrick, J. A., and Harsh, J. B. (1988). "The Mechanism of Z n 2+ Adsorption on Calcite." Geochimica Et Cosmochimica Acta, 52(9), 2281 2291.

PAGE 153

153 BIOGRAPHICAL SKETCH Yu Wang born in Shenyang city, Liaoning province, P. R. China, is the only child of Shusen Wang and Yingli Wan Yu Wang received his e nvironmental s cience and e ngineering from Nanjing University (Nanjing, P. R. China) in 1999 After three year study in Nanjing University graduate school, he got a Master of Science in E nvironmental Science in 2002. He served as a research scientist in Chi nese Research Academy of Environmental Sciences (Beijing, P. R. China) from 2002 to 2004. He was admitted as a graduate student in the Department of Civil Engineering at Auburn University where he got another Master of Science in Civil Engineering in 2006 Then he started his doctoral research in groundwater remediation in the Department of Environmental Engineering Sciences in the University of Florida under the direction of Dr. Timothy Townsend.