Remediation of Dissolved Iron from Groundwater at a Closed Landfill Using Air Sparging and Calcium Carbonate Based Perme...

MISSING IMAGE

Material Information

Title:
Remediation of Dissolved Iron from Groundwater at a Closed Landfill Using Air Sparging and Calcium Carbonate Based Permeable Reactive Barriers
Physical Description:
1 online resource (146 p.)
Language:
english
Creator:
Sikora, Saraya Q
Publisher:
University of Florida
Place of Publication:
Gainesville, Fla.
Publication Date:

Thesis/Dissertation Information

Degree:
Master's ( M.E.)
Degree Grantor:
University of Florida
Degree Disciplines:
Environmental Engineering Sciences
Committee Chair:
Townsend, Timothy G
Committee Members:
Annable, Michael D
Bonzongo, Jean-Claud

Subjects

Subjects / Keywords:
aeration -- concrete -- contamination -- groundwater -- iron -- landfill -- limestone -- manganese -- sparging
Environmental Engineering Sciences -- Dissertations, Academic -- UF
Genre:
Environmental Engineering Sciences thesis, M.E.
bibliography   ( marcgt )
theses   ( marcgt )
government publication (state, provincial, terriorial, dependent)   ( marcgt )
born-digital   ( sobekcm )
Electronic Thesis or Dissertation

Notes

Abstract:
High concentrations of iron (Fe(II)) and manganese (Mn(II)) reductively dissolvedfrom soil minerals have been detected in groundwater monitoring wells surrounding many Florida Landfills (GCTLs of 0.3 and 0.05 mg/L for Fe and Mn,respectively). Two in situ technologies, air sparging and permeable reactive barriers (PRBs) were utilized for the purpose of remediation at a closed, unlined landfill (Klondike Landfill) in Florida, USA. The goal of aeration was to oxidize Fe and Mn to their respective, immobile forms, +3 and +4. Two calcium carbonate based PRBs were evaluated after Fe and Mn sorption capacity began to diminish in the third year of groundwater treatment and results were compared to laboratory studies. Barrier materials were excavated for thepurpose of characterization.    Air sparging is a treatment technology commonly applied to known contaminant masses which can be removed by aerobic degradation or volatilization. In the current study air sparging had limited success in aquifer reaeration at three distinct sites downgradient of the Klondike Landfill. Vadose zone aeration (VZA) and shallow air sparging were employed from August through December 2011 with limited success; impacts were seen in only 3 of 9 wells during shallow air sparging and only 2 of 17 wells at vadose zone aeration areas with effects seen only after system shutdown for VZA. Dissolved iron levels fell in select monitoring wells radially surrounding air injection wells, however in wells as close as 8 feet to injection wells a uniform decrease in dissolved iron levels was not consistently observed. Additionally, in wells affected positively by air sparging (average FeTOT concentration observed in affected monitoring wells throughout the study was 1.40 mg/L compared to a background of 15.38 mg/L, corresponding with anaverage Mn concentration of 0.60 mg/L compared to a background level of 0.27 mg/L, changes of 90% and -121% change from initial background monitoring (background concentrations in wells affected by initial VZA were counted before VZA initiation to give a true background reading). Reference wells located well outside the radius of influence of air sparge areas showed little variation,with no trending in any direction, in FeTOT (64.98 ± 6.20 mg/L and16.32 ± 1.58 mg/L in two wells screened at different depths) and Mn (2.14 ± 0.28 mg/L and 1.59 ± 0.18 mg/L) concentrations from August 2011 through late April 2012, indicating all effects seen are the result of air injection activities at study sites and not a natural site phenomenon. Based on these results air sparging may be recommended for use in intercepting plumes of dissolved Fe,though consideration should be made for potentially reducible Mn.    Two permeable reactive barriers were employed downgradient of a closed landfill for Fe/Mn treatment. Excavation of the barriers did not allow researchers to characterize Fe/Mn adsorbed from groundwater passage through the barrier, likely due to soil buildup in the barrier. No trend was observed in iron and manganese content in samples collected from above the water table and at different depths throughout the barriers. In field experiments related to Fe(II) adsorption capacity no trend was clear as to remaining Fe(II) adsorption and depth in the trench. No method was successful in separating groundwater adsorbed metals vs. soil metals adhered to the surface of reactive materials.
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 Saraya Q Sikora.
Thesis:
Thesis (M.E.)--University of Florida, 2012.
Local:
Adviser: Townsend, Timothy G.
Electronic Access:
RESTRICTED TO UF STUDENTS, STAFF, FACULTY, AND ON-CAMPUS USE UNTIL 2013-08-31

Record Information

Source Institution:
UFRGP
Rights Management:
Applicable rights reserved.
Classification:
lcc - LD1780 2012
System ID:
UFE0044774:00001


This item is only available as the following downloads:


Full Text

PAGE 1

1 REMEDIATION OF DISSOLVED IRON FROM GROUNDWATER AT A CLOSED LANDFILL USING AIR SPARGING AND CALCIUM CARBONATE BASED PERMEABLE REACTIVE BARRIERS By SARAYA SIKORA A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF ENGINEERING UNIVERSITY OF FLORIDA 2012

PAGE 2

2 2012 Saraya Sikora

PAGE 3

3 To all who have inspired by intellectual interest in science and environmental engineering and public h ealth issues

PAGE 4

4 ACKNOWLEDGMENTS I would like to acknowledge my academic advisor, Dr. Timothy Townsend, for the support and guidance throughout my graduate degree at University of Florida. I would also like to thank the other members of my advisory commit tee, Dr. Mike Annable and Dr. Jean Claude Bonzongo. I thank Dr. Hwidong Kim, Dr. Willie Harris, and Dr. Jianye Zhang for their invaluable assistance in laboratory analysis technique. I am grateful to the Escambia County Solid Waste Management Division and the Hinkley Center for Solid and Hazardous Waste Management for their assistance in funding the research that went into this work. From the Escambia County Solid Waste Division I specifically want to mention Ron Hixson, Pat Johnson, Jim Howes, Brent Sch neider, Bobby Ellis, and especially Doyle Butler for his significant contribution to the Klondike Landfill Pilot Study through communicating with blower vendors, site electrician, surveyors, FDEP, arranging for site maintenance and checking on and servicin g the blowers. I am grateful for the assistance of my colleagues for their help in field work and laboratory procedures Priya Hrenko, Yu Wang, Max Krause, Wesley Oehmig, and Nicholas Yonezawa. For inspiring my initial interest in the sciences I would like to thank Mr. Bill Hausmann, and my parents Ed and Teresa Sikora. Finally, my sincerest appreciation goes to my fianc, Daniel Pleasant, for his help, patience, encouragement, and especially love.

PAGE 5

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS ................................ ................................ ................................ .. 4 LIST OF TABLES ................................ ................................ ................................ ............ 7 LIST OF FI GURES ................................ ................................ ................................ .......... 8 LIST OF ABBREVIATIONS ................................ ................................ ........................... 11 ABSTRACT ................................ ................................ ................................ ................... 12 CHAPTER 1 INTRODUCT ION ................................ ................................ ................................ .... 15 Background and Problem Statement ................................ ................................ ...... 15 Research Objectives ................................ ................................ ............................... 20 Research Approach ................................ ................................ ................................ 21 Organization of Thesis ................................ ................................ ............................ 22 2 REMEDIATION OF DISSOLVED IRON AND MANGANESE USING AIR SPARGING AND VADOSE ZON E AERATION AT A CLOSED LANDFILL ............ 24 Introductory Remarks ................................ ................................ .............................. 24 Methodology ................................ ................................ ................................ ........... 27 Site Description ................................ ................................ ................................ 27 Construction and Design of Air Injection System and Monitoring Network ....... 29 Groundwater Sampling and Analysis ................................ ............................... 32 Air Injection Cycle Monitoring ................................ ................................ ........... 33 Results and Discussion ................................ ................................ ........................... 33 Hydrologic Conditions ................................ ................................ ....................... 33 System Operation ................................ ................................ ............................. 36 Groundwater Chemistry ................................ ................................ .................... 39 VZA and shallow AS s ystems ................................ ................................ .... 40 Iron levels at air sparge sites ................................ ................................ ..... 42 Manganese levels at air sparge sites ................................ ......................... 44 ORP and dissolved o xygen ................................ ................................ ........ 46 pH at air sparge sites ................................ ................................ ................. 49 Total organic carb on ................................ ................................ .................. 51 Major Observations and Conclusions ................................ ............................... 52 Cost Analysis ................................ ................................ ................................ .... 53 Sum mary ................................ ................................ ................................ ................ 53

PAGE 6

6 3 END OF LIFE ASSESMENT OF TWO CALCIUM CARBONATE BASED PERMEABLE REACTIVE BARRIERS FOR REMEDIATION OF IRON CONTAMINATED GROUNDWATER AT A CLOSED LANDFILL ........................... 75 Introductory Remarks ................................ ................................ .............................. 75 Methods and Materials ................................ ................................ ............................ 78 Site Description ................................ ................................ ................................ 78 Groundwater Sampling and Analysis ................................ ............................... 78 Composition of Original Reactive Materials ................................ ...................... 79 Rema ining Iron Adsorption Capacity ................................ ................................ 80 Reactive Media Collection and Preparation ................................ ...................... 80 Determination of Adsorbed Iron and Manganese ................................ ............. 81 Characterization of Particle Precipitates ................................ ........................... 82 Results and Discussion ................................ ................................ ........................... 84 Groundwater Monitoring Data ................................ ................................ .......... 84 Hydraulic d ata ................................ ................................ ............................ 84 Groundwater c hemistry ................................ ................................ .............. 84 Composition of PRB Materials ................................ ................................ .......... 88 Remaining Iron Adsorption Capacity ................................ ................................ 88 Leaching Tests ................................ ................................ ................................ 90 Characterization of Precipitates of Reactive Material ................................ ....... 92 Comparison of PRB Performance to Laboratory Predictions ............................ 94 Summary ................................ ................................ ................................ ................ 95 4 CONCLUSION ................................ ................................ ................................ ...... 108 Summary ................................ ................................ ................................ .............. 108 Concluding Remarks ................................ ................................ ............................. 109 Future Work ................................ ................................ ................................ .......... 110 APPENDIX A AIR SPARGING: ADDITIONAL INFORMATION ................................ .................. 112 B PRBS: ADDITIONAL INFORMATION ................................ ................................ ... 134 LIST OF REFERENCES ................................ ................................ ............................. 138 BIOGRAPHICAL SKETCH ................................ ................................ .......................... 146

PAGE 7

7 LIST OF TABLES Table page 2 1 Well details ................................ ................................ ................................ ......... 55 2 2 Operational schedule of air injection systems in 2012 ................................ ........ 55 3 1 Influent groundwater parameters from May 2011 through February 2012 ......... 97 3 2 Effluent groundwater parameters from May 2011 through February 2012 ......... 97 3 3 Elemental composition of PRB materials by mass ................................ ............. 97 A 1 Operational schedule for VZA and A S systems operating in 2011 ................... 112 A 2 Vadose zone aeration and air sparge groundwater data pre 2011 activities ................................ .......... 112 A 3 Vadose zone aeration and air sparge groundwater data during 2011 activities (monitoring events from Aug. ................................ ......................... 112 A 4 Additional groundwater da ta from post shutdown monitoring on July 19 th and 20 th 2012 ................................ ................................ ................................ .......... 112 B 1 Monitoring well details ................................ ................................ ...................... 135 B 2 Results of initial laborato ry leaching tests on nanopure rinsed limestone, (3 M HCl, L/S ratio of 1.66:1) ................................ ................................ .................... 136 B 3 Iron and manganese concentrations in nanopure rinse water used to clean initial samples (L/S ratio of 1.6 6 :1). ................................ ................................ .. 136 B 4 Results of laboratory leaching tests on limestone, fluid concentration (0.5 M HCl leaching fluid) ................................ ................................ ............................ 136 B 5 Results of laboratory leaching tests on limestone, calculated average mass removal ................................ ................................ ................................ ............. 136 B 6 Results of laboratory leaching tests on concrete (0.5 M HCl leaching fluid) ..... 137 B 7 Results of laboratory leaching tests on concrete, calculated average mass removal ................................ ................................ ................................ ............. 137

PAGE 8

8 LIST OF FIGURES Figure page 1 1 Conceptual schematic of air sparging as applied to an iron/manganese plume in groundwater ................................ ................................ ......................... 23 1 2 Conceptual schematic of permeable reactive barrier application to iron /manganese con taminated groundwater ................................ ....................... 23 2 1 Air sparge site locations relative to the Klondike Landfill and existing compliance wells ................................ ................................ ................................ 56 2 2 So uth air sparge site layout ................................ ................................ ................ 57 2 3 Central air sparge site layout ................................ ................................ .............. 58 2 4 North air sparge site layout ................................ ................................ ................. 59 2 5 South site water surface elevation ma ps, values represent ft NAVD .................. 60 2 6 Typical injection pressure vs. time at the north study site ................................ ... 61 2 7 South site groundwater mounding durin g sparge cycles ................................ .... 62 2 8 Central site pressure in soil gas monitoring wells during air sparge injecti on cycles (all units inches of water) ................................ ................................ ......... 63 2 9 South site total iron vs. time ................................ ................................ ................ 64 2 10 Central site total iron vs. time ................................ ................................ ............. 65 2 11 North site total iron vs. time ................................ ................................ ................ 65 2 12 South site average total iron concentration (mg/L) from Mar. ............ 66 2 13 Central site average total iron concentration (mg/L) from Mar. .......... 67 2 14 North site average total iron concentration (mg/L) from Ma r. ............. 67 2 15 South site average total iron concentration (mg/L) measured during sampling on July 19, 2012, 49 days post shut down ................................ .......................... 68 2 16 Central site average total iron concentration (mg/L) measured during sampling on July 19, 2012, 49 days post shut down ................................ .......... 69 2 17 North site average total iron concentration (m g/L) measured during sampling on July 20, 2012, 50 days post shut down ................................ .......................... 70

PAGE 9

9 2 18 South site manganese vs. time ................................ ................................ ........... 70 2 19 Central sit e manganese vs. time ................................ ................................ ........ 71 2 20 North site manganese vs. time ................................ ................................ ........... 72 2 21 Total manganese vs. total iron in wells affected by air sparg ing (Avg. Fe TOT < 4.2 mg/L during study period) ................................ ................................ ............. 73 2 22 ORP vs. dissolved iron in wells affected by air sparging (Avg. Fe TOT < 4.2 mg/L during study period) ................................ ................................ ................... 74 3 1 PRB Area Maps ................................ ................................ ................................ 99 3 2 Groundwater Table Elevation Difference ( NAVD) in the PRBs ........................ 100 3 3 Cha nge in total iron levels in PRB monitori ng wells over time ......................... 101 3 4 Change in total manganese levels in PR B monitoring wells over time ............ 102 3 5 Total iron, manganese, and sodium levels in co mpliance wells for reference .. 103 3 6 pH Change in PRB monitoring wells ove r time ................................ ................. 104 3 7 SEM Photographs of limestone particl e surface ................................ .............. 105 3 8 Particle analysis of an unwashed limestone particle subjected to Fe(II) adsorption at 50 mg/L ................................ ................................ ....................... 107 A 1 South sparge site. ................................ ................................ ............................. 113 A 2 Central sparge site ................................ ................................ ........................... 113 A 3 North sparge site ................................ ................................ .............................. 114 A 4 South site water surface elevation July 19, 2012, values represent ft NAVD. .. 115 A 5 Central site water surface elevation m a p, values represent ft NAVD ............... 116 A 6 Central site water surface elevation map July 19, 2012, values represent ft NAVD. ................................ ................................ ................................ .............. 117 A 7 No rth site water surface elevation m ap, values represent ft NAVD .................. 118 A 8 North site water surface elevation map July 19, 2012, values represent ft NAVD. ................................ ................................ ................................ .............. 119 A 9 Typical injection pressure vs. time at south study site ................................ ...... 120

PAGE 10

10 A 10 Typical injection pressure vs. time at central study site ................................ .... 121 A 11 Central site groundwater mounding during 6 hour sparge cycles ..................... 122 A 12 North site groundwater mounding during 6 hour sparge cycles ........................ 123 A 13 Central site pressure in soil gas monitoring wells during air sparge injection cycles (all units inches of water) ................................ ................................ ....... 124 A 14 North site pressure i n soil gas monitoring wells during air sparge injection cycles (all units inches of water) ................................ ................................ ....... 125 A 15 North site pressure in soil gas monitoring wells during air sparge injection cycles (all uni ts inches of water) ................................ ................................ ....... 126 A 16 South site ORP vs. time ................................ ................................ ................... 127 A 17 Central site ORP vs. time ................................ ................................ ................. 127 A 18 North site ORP vs. time ................................ ................................ .................... 128 A 19 South site dissolved oxygen vs. time ................................ ................................ 129 A 20 Central site dis solved oxygen vs. time ................................ .............................. 130 A 21 North site dissolved oxygen vs. time ................................ ................................ 131 A 22 pH vs. time, affected wells are considered those w ith average Fe TOT < 4.2 mg/L throughout the study after air sparge initiation ................................ ......... 132 A 23 TOC vs. total iron in reference well AS2 ................................ ........................... 133 B 1 Klondike Landfill map with locations of active groundwater monitoring wells, PRB area identified ................................ ................................ ........................... 134

PAGE 11

11 LIST OF ABBREVIATION S AS Air s parge Bls Below land surface CCBM Calcium carbonate based material GCTL Groundw ater cleanup target l evel HBRG Health b ased r isk g uideline MCL Maximum contaminant l evel NAVD North American vertical d atum NOM Natural organic m atter PQL Practical quantification l imit PRB Permeable reactive b arrier RfDo Oral reference d ose ROI Ra dius of i nfluence TOC Total organic c arbon VZA Vadose zone aeration XRD X Ray d iffraction WSE Water surface e levation ZVI Zero valent i ron

PAGE 12

12 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Engineering REMEDIATION OF DISSOLVED IRON FROM GROUNDWATER AT A CLOSED LANDFILL USING AIR SPARGING AND CALCIUM CARBONATE BASED PERMEABLE REACTIVE BARRIERS By Saraya Sikora August 2012 Chair: T imothy Townsend Major: Environmental Engineering Sciences High concentrations of iron (Fe(II)) and manganese (Mn(II)) reductively dissolved from soil minerals have been detected in groundwater monitoring wells surrounding many Florida Landfills (GCTLs of 0.3 and 0.05 mg/L for Fe and Mn, respectively) Two in situ technologies, air sparging and permeable reactive barriers (PRBs) were utilized for the purpose of remediation at a closed, unlined landfill (Klondike Landfill) in Florida, USA. The goal of aerati on was to oxidize Fe and Mn to their respective, immobile forms, +3 and +4. Two calcium carbonate based PRBs were evaluated after Fe and Mn sorption capacity began to diminish in the third year of groundwater treatment and results were compared to laborato ry studies. Barrier materials were excavated for the purpose of characterization. Air sparging is a treatment technology commonly applied to known contaminant masses which can be removed by aerobic degradation or volatilization. In the current study air sp arging had limited success in aquifer reaeration at three distinct sites downgradient of the Klondike Landfill. Vadose zone aeration (VZA) and sh allow air sparging were employed from August through December 2011 with limited succe ss;

PAGE 13

13 impacts were seen in o nly 3 of 9 wells during shallow air sparging and only 2 of 17 wells at vadose zone aeration areas with effects seen only after system shutdown for VZA Dissolved iron levels fell in select monitoring wells radially surrounding air injection wells, however in wells as close as 8 feet to injection wells a uniform decrease in dissolved iron levels was not consistently observed. Additionally, in wells affected positively by air sparging (average Fe TOT < 4.2 mg/L the health based risk guideline, after air sparg e commencement) rising manganese concentrations were observed, indicating that the redox potential of the water is moving from an iron reducing environment to a manganese reducing environment. Average Fe TOT concentration observed in affected monitoring wel ls throughout the study was 1.40 mg/L compared to a background of 15.38 mg/L, corresponding with an average Mn concentration of 0.60 mg/L compared to a background le vel of 0.27 mg/L, changes of 90 % and 121% change from initial background monitoring (backg round concentrations in wells affected by initial VZA were counted before VZA initiation to give a true background reading). Reference wells located well outside the radius of influence of air sparge areas showed little variation, with no trending in any d irection, in Fe TOT (64.98 6.20 mg/L and 16.32 1.58 mg/L in two wells screened at different depths) and Mn (2.14 0.28 mg/L and 1.59 0.18 mg/L) concentrations from August 2011 through late April 2012, indicating all effects seen are the result of air injection act ivities at study sites and not a natural site phenomenon. Based on these results air sparging may be recommended for use in interc epting plumes of dissolved Fe, though consideration should be made for potentially reducible Mn

PAGE 14

14 Two permeable r eactive barriers were employed downgradient of a closed landfill for Fe/Mn treatment. Excavation of the barriers did not allow researchers to characterize Fe/Mn adsorbed from groundwater passage through the barrier, likely due to soil buildup in the barrie r. No trend was observed in iron and manganese content in samples collected from above the water table and at different depths throughout the barriers. In field experiments related to Fe(II) adsorption capacity no trend was clear as to remaining Fe(II) ads orption and depth in the trench. No method was successful in separating groundwater adsorbed metals vs. soil metals adhered to the surface of reactive materials.

PAGE 15

15 CHAPTER 1 INTRODUCTION Background and Problem Statement High concentrations of iron and mang anese have been observed in groundwater monitoring wells surrounding many MSW landfills in Florida and elsewhere. These metals were originally thought to originate from landfill leachate plumes. However current research lends to the theory that redox shift s in the aquifer cause these metals to dissolve from naturally occurring metal oxides present in the soil matrix (Wang et al., 2011). Ratios of commonly occurring leachate constituents, such as sodium to iron are also used to confirm this theory. Many soi ls contain an abundance of iron (III) oxide minerals, commonly ferrihydrite (Fe(OH) 3 ), goethite FeOOH) akageneite FeOOH) lepidocrocite FeOOH) hematite Fe 2 O 3 Fe 2 O 3 ) (Heron et al., 1994). These minerals give the soils their color, and allow other minerals to sorb to their surfaces. Ma et al. (1997) reported typical ir study is in Entisol dominated area) of 960 and 33 mg/kg dry, respectively, making iron the most abundant metal on a mass/mass basis present in entisol soils. A USGS survey revealed U.S average i ron and manganese levels of 14,000 and 260 mg/kg dry for the Eastern United States (Shacklette and Boerngen, 1984). At unlined MSW landfills the introduction of organic matter into shallow aquifers by landfill leachate leads to redox reactions n ecessary for the breakdown of the organic matter into its ultimate products of carbon dioxide methane, and water. In order for these redox reactions to take place an electron acceptor must be utilized. In cases where O 2 is not available to fulfill this ro le iron oxides (as well as manganese oxides)

PAGE 16

16 can be utilized by groups of bacteria as electron acceptors for consumption of organic matter ; for example the reductive dissolution of magnetite or hematite 2( Fe 2 O 3 ) s + C 6 H 12 O 6 + e 4Fe(II) + 6(CO 2 ) + 12H + In shallow, uncontaminated aquifers, the persistence of aerobic conditions causes iron to primarily occur as iron (III) oxides and hydroxides (Schwertmann and Taylor, 1977; Heron et al., 1994; Heron and Christensen, 1995). At neutral pH redox half reactio ns involving manganese and iron oxides persist from the pE range of approximately +9 to 1 (ORP values of +531 mV to 59 mV at 25 C) and approximately 1 to 10 (ORP values of 59 mV to 590 mV), respectively (Stumm and Morgan, 1996). Many lined landfill s in the state have experienced the same effects, indicating the simple disturbance of the atmosphere soil oxygen exchange may be sufficient to produce reductive dissolution of iron and manganese. Iron oxide reduction is considered beneficial in many cases because it decreases the distance possibly hazardous organic compounds may travel, sometimes referred to in this capacity as a redox buffer (Heron and Christensen, 1995). Conversely iron oxidation has been utilized for groundwater treatment as well; zero valent iron (ZVI) is one of the most common reactive materials used in permeable reactive barrier technology, where contaminated water passively flows through treatment media. ZVI has been used in the field to intercept groundwater contaminated with chrom ate (Puls et al., 1999), acid mine drainage (Benner et al., 1999), chlorinated aliphatic hydrocarbons (McMahon et al., 2005), molybdenum and uranium (Morrison et al., 2006). Lab scale studies conducted have demonstrated the ability of ZVI to effectively re move uranium, technetium, and molybdenum oxides (Cantrell et al., 199 5). The redox sensitive nature of iron and its relative high abundance in soil causes it to play a large role in

PAGE 17

17 groundwater chemistry as it is often the most abundan t potential electron acceptor in soils at landfills (Lyngkilde and Christensen, 1992). The process of reductive dissolution can introduce arsenic as well as dissolved iron and manganese to groundwater, as arsenic binds well to iron oxide minerals (Cummings et al., 1999; Lack ovic et al., 1999; de Lemos et al., 2006). Increases in dissolved arsenic levels have been strongly correlated (> 80%) to degradation of organic contaminants (de Lemos et al., 2006). Cummings et al. (1999) observed arsenic mobilization by action of dissimi latory iron reducing bacteria (Shewanella alga BrY) from crystalline ferric arsenate. Average Florida soil concentrations of arsenic is 2.4 mg/kg dry soil with a range of 0.01 to 6.1 mg/kg dry soil observed (Ma et al., 1997). In a USGS survey average arsen ic levels in soils were found to be 4.8 mg/kg dry soil for the Eastern U.S. (Shacklette and Boerngen, 1984). Arsenic levels at the Klondike landfill and other landfills in the Florida panhandle as well as Florida peninsula regularly exceed the regulatory t hreshold. The treatment technology evaluated within this chapter has the potential to remediate arsenic from contaminated waters through reoxidation of dissolved iron and thus readsorption of arsenic minerals to iron oxides. As a large portion of Floridia ns rely on groundwater for drinking (approximately 90% of the roughly 16 million residents) ( US EPA, 2011 ) and thus the state provides for strict regulation on groundwater contamination. Additionally, the water table in Florida is often present at relativel y shallow depths and because of this, groundwater contaminants often make their way into surface waters; groundwater discharges can provide up to 80 percent of surface water flow (FDEP, 2011). Iron in surface waters can quickly oxidize, causing red floc fo rmation which impairs water quality by increasing

PAGE 18

18 turbidity, decreasing dissolved oxygen, and contributing iron reducing bacteria which leave sheen on the water surface. Iron and manganese present in water supply wells to homes causes staining of laundry, sinks, toilets, and gives the water a metallic taste, even if the levels are not high enough to be considered harmful to human health. It is for these reasons the secondary drinking water standards (based on aesthetics of water quality) of 0.3 mg/L and 0.0 5 mg/L for iron and manganese, respectively, are employed as groundwater cleanup target levels (GCTLs) for Florida. Considering health effects which may arise from consumption of water, rather than aesthetics, the health based risk guidelines (HBRG) of 4.2 and 0.33 mg/L were derived using a n oral reference dose ( RfDo ) of 0.6 and 0.047 mg/kg/day for iron and manganese, respectively (Gadagbui and Roberts, 2002). High concentrations of iron and manganese in ground water are not unique to Florida. O ther southern states, such as Alabama, Arkansas, Georgia, North Carolina, and South Carolina to name a few have experienced the same difficulties with iron and mang anese presence in groundwat er. State documents acknowledge their presence, as well as the presence of iro n and manganese reducing bacteria are not a health risk, but an aesthetic nuisance (Langston and Tacker, 1989; Christenbury, 1990; Hariston, and Stribling, 1995; Herman, 1996). Residents that rely on we ll water often have to treat the water or are forced t o deal with reddish brown staining of laundry, pipes, sinks, clogged pipes, and bad tasting water (Herman, 1996). Langston and Tacker (1989) advise Arkansas residents that Fe and Mn removal are pH dependent, with iron removal at a pH of 7.5 or above and m anganese removal at pH 8.5 or above for private residents using oxidation for contaminant removal. Northern states, such as Wisconsin,

PAGE 19

19 Minnesota, and New Hampshire have also published literature on the iron and manganese problem (Plowman, 1989; Machmeier, 1990; Shaw et al., 1990). Landfills are required to remediate these contaminants on a much larger scale than an individual homeowner, and thus the costs are greater and there is a need for cost effective, in situ remediation strategies. Landfill operators removal; this can be costly and in the case of iron/manganese contamination, when the source is the soil underneath the landfill, this may be a futile effort. Oxidation by chemical means, air or pur e oxygen injection ex situ has been shown to produce solid precipitates which can then be filtered out of water (Theis and Singer, 1976; Ellis et al., 2000). Air sparging is a technology which has the capability to return an aquifer to aerobic conditions t hrough the forced injection of air (in some cases pure oxygen or chemical oxidants are used) through vertical or horizontal wells in situ (Marley et al., 1995; Bass et al., 2000). Sparging is typically more successful in sandy formations over clay, and has been successfully utilized in cases of petroleum hydrocarbons and chlorinated solvents spills (Johnson et al., 1993; Chao and Ong, 1995; Lundegard and LaBreque, 1995; Rhodes et al., 19 95; Johnson et al., 2001). See F igure 1 1 for an illustration of air sp arge technology as it is to be applied in the current study. In the recent past permeable reactive barrier (PR B) technology has become proven and effective technology, desirable fo r their relatively low o peration and maintenance costs (USEPA, 2001; Henders on and Demond, 2007). Barriers can consist of reactive media tailored for specific contaminant removal, and are most commonly composed of zero valent iron (for removal of chlorinated organics or other

PAGE 20

20 heavy metals such as chromate and uranium) and iron ox ides (arsenic and other heavy metal removal) (Puls et al., 1999; Morrison et al., 2006; Henderson and Demond, 2007). Calcium carbonate based materials (CCBMs) have been shown to be successful in the remediation of divalent metals in laboratory and column s tudies (Aziz and Smith, 1992; Aziz et al., 2001 Mettler et al., 2009). CCBMs such as limestone are typically inexpensive, costing less than $50 per ton, and the use of materials such as recycled crushed concrete could be considered usage of a waste product for desi red purposes (Wang, 2011). See F igure 1 2 for a conceptual illustration of PRB technology in use. Research Objectives The overall purpose of this research project is to evaluate two diverse in situ groundwater remediation techniques for dissolved iron and manganese removal downgradient of disposed waste at a closed, unlined landfill. The first technique investigated was air sparging, the forced injection of ambient air into groundwater with the purpose of aquifer reparation. The purpose of this res earch was to characterize air flow into the sandy aquifer for pulsed sparging and to determine the effects of air injection on groundwater chemistry (particularly dissolved concentrations of iron/manganese) Two permeable reactive barriers compose d of cal cium carbonate minerals remediated groundwater dissolved iron and manganese successfully as part of a previous dissertation (Wang, 2011). The current study continued groundwater monitoring after peak Fe/Mn removal efficiency had passed The barriers were t hen examined ex situ for the following purposes: to determine if remaining Fe adsorption capacity could be correlated to in situ data if particle adsorbed Fe and Mn could be determined via acid leaching, and particle characterization of excavated particle s by

PAGE 21

21 XRD, SEM and EDS Results from the study were then compared with laboratory testing conducted prior to barrier installation (Wang, 2011). The knowledge that the source of the iron in the groundwater is the soil makes a conventional pump and treat syst em impractical. Therefore in situ technologies offer the greatest promise at remediation of naturally occurring metals which become contaminants under given redox conditions. Research Approach The efficacy of air sparging on dissolved iron and manganese l evels in groundwater was evaluated using positive displacement blowers to pump ambient air into the surficial aquifer. Three air sparge sites were constructed downgradient of a closed landfill. Air sparge wells were designed and displacement blowers were s elected based on guid elines and commonly used methods available in the current literature (Hudak, 2000; Leeson et al., 2002; Marley et al., 1995). Manifold piping was constructed to inject air into two wells either perpendicular or parallel to groundwater flow and included pressure gauges and flow meters. Networks of six, eight, or nine groundwater monitoring wells were utilized for water quality monitoring throughout the study, depending on the site. Soil gas monitoring wells were utilized at two sites to determine air flow pathways nearest to the injection wells. Groundwater quality was analyzed throughout the study to evaluate technology performance. Two permeable reactive barriers (one composed of limestone and the other of crushed concrete) installed in June and July 2009 began losing their efficacy in remediating groundwater dissolved iron to below the health based guideline of 4.2 mg/L between October 2010 and May 2011 (Wang, 2011) Groundwater quality analysis of the material was examined from May 201 1 to May 2012 upon which time the barriers

PAGE 22

22 were excavated by mechanical means and samples were taken from variable depths and experiments for remaining iron adsorption capacity as well as total iron and manganese adsorbed were conducted. Organization of Th esis This thesis is presented in 4 chapters, including the current chapter (Chapter 1). Chapter 2 presents findings of three field air sparging experiments occurring downgradient of a closed landfill and evaluates this technology as it relates to the disso lved contaminants iron and manganese. Chapter 3 is an end of life assessment of two functional permeable reactive barriers based on groundwater monitoring data, excavation and subsequent iron removal capacity experiments performed in the field and laborato ry, and reactive media analysis. Chapter 4 summarizes all findings, results, and potential for future work based on conducted research experiments. Subsequently, references for all cite d works are included. Appendix A contain s additional information and fi gures relating to air sparge research as d etailed in Chapter 2. Appendix B contain s additional information and figures relating to permeable reactive barrier research for Chapter 3.

PAGE 23

23 Figure 1 1. Conceptual schematic of air sparg ing as applied to an iron/manganese plume in groundwater Figure 1 2. Conceptual schematic of permeable reactive barrier application to iron /manganese contaminated groundwater

PAGE 24

24 CHAPTER 2 REMEDIATION OF DISSO LVED IRON AND MANGAN ESE USING AIR SPARGI NG AND VADOSE ZONE AERATION AT A CLOSED LANDFILL Introduct ory Remarks High concentrations of iron and manganese have persisted at the Klondike Landfill in Escambia County Florida since 1987. Ratios of iron to normal leachate constituents in the groundwater (such as sodium) are significantly higher than in leachate, indicating that the origin of the dissolved iron present in the groundwater is not the landfill leachate, as dilution of iron and sodium would be the same, but reductively dissolved iron from soil. Kj eldsen and Christophersen (2000) reported iron to sodium ratios in a leachate survey from old landfills in Denmark to be on average 2.76 Na/ Fe (m/m). Leachate collected from an operating landfill in Winter Haven, Florida, a lined landfill, contained an av erage of 607 232 mg Na/mg Fe. Ratios of Fe/Na (m/m) in downgradient wells of the unlined Klondike Landfill averaged 1.05 0.11 in February 2012, strongly indicating that the source of the iron is the soil. Pump and treat systems have been employed to re move ferrous iron from groundwater. The extraction and treatment of groundwater can be extremely costly. Pump and treat systems typically work well for areas where a known contaminant mass was released; however when the source of the regulatory exceedance s is the natural site soils pump and treat may not be effective. Natural attenuation has been cited as a viable treatment option for metals in leachate polluted aquifers (Christensen et al., 2001), however given regulatory pressure many landfill owners and operators are forced to employ pump and treat as a means to remove dissolved metals from groundwater. The distance required for natural attenuation can vary based on the reducing capacity of leachate plumes. In a leachate polluted aquifer Heron and

PAGE 25

2 5 Christ ensen, 1995, observed an iron/manganese reducing zone from approximately 75 to 300 m from the landfill boundary along the central flow line; a nitrate reducing zone (Fe <1 mg/L) persisted from approximately 340 m to greater than 400 meters from the landfil l boundary. Christensen et al. (1994) in an investigation of landfill leachate pollution, found leachate impacts (including redox effects, such as iron reductive dissolution) in groundwater generally not exceeding 1000 meters from the landfill boundary. Ai r sparging and vadose zone aeration are in situ techniques to increase aerobic degradation of contaminants by injecting air into the groundwater or unsaturated zones of the soil and are typically most successful in sandy aquifers (Murray et al., 2000). Tho ugh preferential flow pathways commonly develop, using high air flow rates (6 to 20 scfm) can help break through lower permeability areas and promote more uniform air channel network (Baker et al., 2005; Johnson et al., 2001). These techniques have been su ccessfully utilized for petroleum hydrocarbon and organic contaminants such as volatile organic compounds (Baker et al., 1995; Chao and Ong, 1995; Marley et al., 1995; Rhodes et al., 1995). In many cases soil vapor extraction is employed to prevent volatil ized contaminant transport out of the range of air injection (Martinson et al., 1995). Based on current literature, the most effective technique for determining radius of influence of an air sparging system is increases in dissolved oxygen (Marley et al., 1995). The contaminants of concern, iron and manganese are mineralized and nonvolatile in their oxidized states, Fe( I II) and Mn(IV), thus soil vapor extraction was not used in this study.

PAGE 26

26 Initially, vadose zone aeration was used at two of the study sites and a shallow air sparge system was used at one site for a 5 month period prior to conversion to deeper air sparging. During this period the air sparge and VZA blowers operated on a 12 hour diel cycle. Groundwater and soil gas monitoring during this experi ment showed that air sparging had potential to reduce dissolved iron and manganese concentrations, while vadose zone aeration did not. Due to drought in the Pensacola area the water table dropped approximately 1.82 feet from June 2010, when the first grou ndwater monitoring event occurred and initial injection well construction was completed to July 2011, when blowers and manifold piping were placed at the site; this, essentially rendering the initial air sparge sys tem used a partial vadose zone aeration sy stem. Results from the initial study indicated that a fully aerated vadose zone has no significant effect ( p <0.05) in decreasing dissolved groundwater iron concentrations during operation. From background monitoring in July 2011 to the last sampling event in November 2011 Fe(II) levels measured in the field rose at the central site from an average of 22.32 7.14 to 30.1 3.3 mg/L; at the north site the mean Fe(II) level also rose, from 14.87 5.19 to 24.4 10.5 mg/L. Interestingly, upon background monit oring for the deeper air sparging study, it was revealed that dissolved iron concentrations had dropped significantly from the last monitoring during vadose zone aeration air injection. At the central study site the average Fe(II) from 9 site monitoring we lls concentration measured in the field dropped to 10.92 6.11 mg/L and at the North site from 8 monitoring wells to 8.69 5.85 mg/L, drastically reduced from values previously discussed, though still above the GCTL and health based risk level. Also dem onstrated was the efficacy of air sparging (even a

PAGE 27

27 shallow system top of injection well screen not fully submerged ) to impact dissolved iron concentrations along the lines of subsurface air flow. Fe(II) concentration decreased during shallow air sparging in monitoring well S 3 at the south site from 19.7 to 5.9 mg/L, an approximate 70% improvement from August to November, 2011 Methodology Site Description The Klondike landfill, located at 7219 Mobile Highway, Pensacola, Florida, approximately 8 miles nor thwest of downtown Pensacola served as the research site. It is a closed, unlined sanitary landfill operated from 1976 to 1982 as a temporary facility for domestic and commercial refuse (Class II landfill). Because it is unlined, with no leachate collectio n system, a comparison of leachate to groundwater chemistry is not possible. Figure 2 1 is a site map. Routine groundw ater monitoring commenced in 198 6. High iron concentrations per sistently exceeding 0.3 mg/L (FL G CTL) were observed in several monitoring wells immediately Additionally manganese and arsenic above Fl GCTL levels are consistently detected in surface water and groundwater monitoring wells (Geosyntec, 2005). Average groundwater velocity in the shallow aquifer (+16 to 8 ft NAVD ) was calculate d to be 0.39 linear ft/day, flowing from east to west towards Eleven M ile creek. Slug tests were conducted during May 2012 and were interpreting in accordance with methods described by Bouwer and Rice (1976) for determining hydraulic conductivity of uncon fined aquifers with completely (typical of the south air sparge site) or partially submerged well screens (typical of the central and north air sparge sites). Aquifer depth at each well was calculated based on survey data for each well and site hydrogeolo gic data (Geosyntec, 2005). Using data from six monitoring wells at the south site the average calculated hydraulic conductivity was

PAGE 28

28 26.43 ft/day with a median of 21.49 ft/day. Calculated hydraulic conductivities at the central and north sites were 27.4 7 11.47 and 22.27 16.23 ft/day, respectively. A gradient term was then applied based on water table elevation measurements in February 2012 Annual Groundwater Monitoring Report for wells on either side of the landfill. Well PIW 3 was used for an upgra dient point and wells MW 6 and MW 7AR were used for downgradient points. Interestingly, water surface elevations from monitoring events at each site did not show a consistent trend of an east west groundwater gradient. Calculated groundwater velocities for the south, central, and north sites were 0.11, 0.10, and 0.14 ft/day, respectively. The three study sites used in this experiment are located between the landfill and eleven mile creek, named the south site, central site, and north site. The south site wa s initially used for air sparging at a shallow depth (10 ft bls, 5 foot screen)). The central and north sites were initially used for vadose zone aeration with injection well configurations that vary from one another; injection wells at the central site ar e in line perpendicular to groundwater flow and wells at the north site are in line parallel to groundwater flow. Routine groundwater monitoring data, included in the 2011 Fall Semi Annual Klondike Landfill Report (consisting of data collected on August 4 2011) revealed gr oundwater flow persisting from east to w est towards Eleven Mile Creek. In background and compliance monitoring wells fully screened in the shallow aquifer zone (+16 to 8 ft NAVD) as designated by Geosyntec Consultants in a 2005 feasibil ity evaluation. ORP values in upgradient wells were 126.3 1.34 (relatively oxidizing) and in downgradient wells were 50.1 30.7, indicative of a shift in redox potential

PAGE 29

29 downgradient of the landfill. In the shallow aquifer iron levels up and downgradie nt of the landfill exceeded the regulatory limit of 0.3 mg/L; manganese levels were only exceeded downgradient of the landfill. Iron in upgradient wells PIW 2 and PIW 3 measured 7.57 and 0.789 mg/L, respectively and in downgradient wells averaged 29.85 1 5.77 mg/L. Manganese levels in the background wells were below the regulatory MCL (4.68 0.757ug/L) while the in the shallow compliance wells levels observed were above the regulatory MCL (272 113ug/L). ORP, iron, and manganese levels observed in Fall 2 011 are in line with historical measurements. Construction and Design of Air Injection System and Monitoring Network University of Florida researchers completed well installation for 2 vadose zone aeration systems and 1 air sparging system in June 2009 (3 sites total). All sites were placed between the landfill and the surface water body known as eleven mile creek. A site map including groundwater flow direction and site location is included as Figure 2 1. Well installation included 2 air injection wells a t each site and a network of groundwater and soil gas (in the case of the two VZA areas) monitoring wells around the injection wells. Figures 2 2 through 2 4 detail well configuration at each site. See Table A 1 for the operational schedule of VZA and shal low AS blower systems. Current research indicates that the desired depth of the top of a sparge well screen below the zone of treatment is approximately 2 meters (6.56 ft). The outlet pressure of the 3 hp blowers (VZA blowers) in their original configurati on thus had to be modified in order to accommodate this setup. The sheaves connected to the blower side of each VZA blower was resized to a smaller diameter (approximately 15% as recommended by the manufacturer) in order to increase RPM output. This allowe d VZA b lowers to produce a pressure of approximately 10 psi, enough to allow injection wells

PAGE 30

30 to be drilled to the desired depth and also increasing the temperature rise of injected air. Water levels in the close vicinity of the air sparge injection wells d ropped from ~3.79 ft bls in June 2010 to ~6.43 ft bls from August October 2011. This hydrologic change effectively converted the air sparge injection wells into dual air sparge and vadose zone aeration wells. Pressure drop calculations were performed for m inor and friction losses along galvanized and pvc piping assuming an air temperature of 60 F (to be conservative). To calculate the allowed depth of well screen below the water table a design method by Leeson et al. (2002) was utilized, assuming a packing annulus of clean sand and a formation consisting of clayey sand based on the soil texture report (Geosyntec, 2006). Based on these calculations as well as commonly used air sparge well parameters (screen length, diameter) the wells were sized (Marley et al 1995). Details of injection wells as constructed can be found in T able A 1. Due to hydrologic conditions and manifold piping lengths at the north site air could be injected deeper beneath the water table than at the central site. The goal of the origina l experimental design sought to examine 2 identical systems in different configurations with respect to the groundwater flow (perpendicular and parallel arrangements), so the injection well at the central site was placed 1 ft deeper than at the north site in order to inject air at approximately the same elevation with respect to the height of the water table. Three positive displacement blowers were purchased from Enviro Equipment Inc. based on their ability to provide air at a desired flow rate and pressu re range based on commonly utilized values for sandy soils (Marley et al., 1995). All blowers were skid

PAGE 31

31 mounted and were designed for or modified to run on the single phase power available at the site. A 5 hp blower (Roots 32 URAI) was purchased and placed at the (initial) air sparging site and two 3 hp blowers (Roots 22 URAI) were placed at the two (initial) VZA sites in June 2011. All blowers were equipped with 24 hour cycle timers. Once blowers were placed at the site, manifold piping to connect blower o utlets to previously constructed injection wells was constructed from galvanized steel and PVC piping. Galvanized steel was chosen for the piping nearest to the blower outlet due to air temperature rise from the blower and its high heat resistance as compa red to the PVC piping. Steel piping began at the blower outlet, ran along the ground and split to provide flow to two wells. See Appendix A for site photos. Pressure and flow gauges were placed in line with flow for each injection well and used to adjust a ir flow for equal injection into each well and for periodic monitoring of pressure changes over injection cycles. Blowers were connected to previously constructed power lines in August 2011. Blowers operated using pulsed sparging. In several studies this has shown to increase bulk mixing of aquifer water mass through the formation (causing mounding) and collapse of air channels in the saturated aquifer (Johnson et al., 1993; Boersma et al., 1995; Bass et al., 2001; Yang et al., 2005). Roots 22 URAI 3 hp bl owers (used at the central and north sites) were set for two daily six hour cycles to occur during the coolest hours of the day (8 pm 2 am & 4 am 10 am). After blower modification to produce a higher pressure airflow temperature rise was an issue. This selection allowed blowers to operate during the cooler hours of the day to minimize overheating of PVC manifold piping due to extremely high daytime temperatures (occasionally reaching 100 F) during the Florida summer months. The Roots 32 URAI 5 hp blowe rs were set

PAGE 32

32 for three daily six hour cycles as they were able to supply air easily at the expected injection pressure (8 pm 2 am & 4 am 10 am and 12 pm 6 pm). A total of twenty five groundwater monitoring wells were installed up and downgradient of i njection we lls for system evaluation. See F igures 2 1 through 2 4 for locations of the study areas, injection systems, and monitoring wells. All groundwater monitoring wells have a 2.5 ft screen interval, constructed of 2 inch PVC, and screens with slot si ze of 0.5. Soil gas monitoring wells were constructed 5 ft bls in the vadose zone and present only in the two original vadose zone aeration areas (central and north site). When monitored during vadose zone aeration it was revealed the air injection adequat ely reaerated the zone to atmospheric levels. The wells were used during the air sparging study to help determine air flow pathways in the subsurface. Detai ls of all wells in included in T able A 1. Groundwater Sampling and Analysis Prior to groundwater sam pling at a study site the depth to groundwater was measured in all site wells with a water level meter (Solinst Model 101). Groundwater samples were collected with a peristaltic pump (Geotech Series II) following a purge of at least one well volume and the n stabilization of pH, DO, turbidity, temperature, and specific conductivity over three consecutive reading as specified in Florida Statutes 2200. ORP (Oakton ORPTestr10) and Fe(II) (Hach Ferrover) levels were both measured in the field after parameter sta bilization occurred. Florida form FD 9000 24 was used for field data collection. Groundwater samples to be analyzed for total metal contents were preserved immediately after field collection using nitric acid to a pH of less than 2, immediately placed on i ce, and then transported to the laboratory for analysis by inductively couple plasma atomic emission spectroscopy ( ICP AES ) All

PAGE 33

33 samples were analyzed in duplicate and appropriate blanks and calibration checks were included in each analysis. Select wells w ere periodically measured for total organic carbon (Shimadzu TOC V CPH ). Because air injection can cause excess turbidity in wells all sampling was conducted with blower systems turned off. In addition to monitoring all wells at air sparge study sites, two reference wells located well outside the potential influence of the sparge blowers were sampled with screen intervals of 0.17 to 6.17 and 6.31 to 11.31 ft NAVD. Average screen intervals in groundwater monitoring wells at the south, central, and north study sites were 8.18 to 10.68, 12.44 to 14.94, and 11.98 to 10.68 ft NAVD. Statistical analyses were performed using one way analysis of variance (ANOVA) followed by a t test considering a p value of less 0.05 was considered statistically significant. Air Inje ction Cycle Monitoring During the study a complete six hour air injection cycle was examined for injection pressure, airflow, pressure in soil gas wells (in the case of the central and north site), and groundwater mounding during the first air injection cy cle, two weeks later, and every four weeks subsequently. Soil gas pressure was measured using a Dwyer 477 digital manometer. The purpose of these measurements was to determine radius of influence of airflow and preferential airflow pathways which may affec t dissolved contaminant concentrations. Results and Discussion Hydrologic Conditions Using calculated groundwater velocities based on site slug tests and groundwater elevation data from regulatory monitoring wells the approximate distance groundwater moved downgradient in the study period (106 days). Reddy and Adams (2000)

PAGE 34

34 observed reduced groundwater flow in air sparge areas using air flow rates of 2.5 L/min and 4.75 L/min (much lower than used in the current study; air sparge zones remained radial and was not affected by groundwater flow in coarse sand at a hydraulic gradient of 0.011 (g radient of the Klondike Landfill is approximately 0.004). In light of this research the time which the blowers were operational was subtracted from the time groundwater was allowed to flow. Approximate flow distance during the stud y from February 15th to June 1 st was approximately from 8.61 to 9.02 ft for the south site, and 5.3 0 and 7.42 ft, for the central, and north sites, respectively. Estimated groundwater flow from Jun e 1 st through July 19 th (final groundwater monitoring event) based on slug test data was 5.13, 4.90, and 6.87 ft for the south, central and north sites, respectively; meaning that the distance between linear sets of wells (i.e. N 4 to N 1, 8.3 ft) would n ot have been traversed. Using values calculated by Geosyntec (2004) flow distances would vary between 9.8 and 19.11 feet at each site, indicating this distance would have been traversed. Given depth to water measurements taken over the course of the study the average depth below the water table of the top of the sparge well screen was 6.74, 5.35, 5.07 ft at the south, central, and north sites, respectively. These values are within the most commonly found values of 5 10 ft reported in a review study by Ma rley et al. (1995). Johnson et al. (1993) in a survey of air sparging research, reported depths as low as three feet (Marley et al., 1990; Marley 1991) and up to forty feet below the water table ( Marley et al., 1990). The general flow of groundwater flow towards eleven mile creek is well established; however micro variations at individual sites may cause groundwater flow to

PAGE 35

35 favor one well or another. Figure s 2 5 and A 4 through A 8 display water surface elevation (WSE) maps based on depth to water measure ments as well as survey data for well top of casings on Feb 14 th (prior to sparging) and July 19 th after sparging had ceased. The overall rise in water level occurred actually only occurred between May and July; water table dropped an average of 0.82 ft b etween monitoring on Feb. 15 th and May 24 th Heavy rains in the region caused the water table to rise approximately 1.58 and 1.59 ft at the central and north sites, respectively, based on water level measurements between the last sampling event of the spar ge period (May 24, 2012) and sampling after system shutdown which occurred on July 19 th and 20 th Data from the south site indicate a 2.24 ft rise, but is likely skewed due to rain occurring prior to and during sampling on July 19 th Given the relatively shallow gradient at the landfill 0.004 ft/ft, the accuracy of the depth to water meter (0.01 ft), and the unknown error associated with surveying, this map may not be able to definitively show correct elevation distances which drive flow South site data reveal a similar WSE of all wells with no clear flow pathways towards any direction. Due to the greater distance between wells closer to the landfill boundary and downgradient wells, the central and north sites offer a better opportunity to deduce groundwa ter flow direction. At the central site data indicate groundwater flow from east to west, as expected, with a possible component of northwest tendency moving water towards C 1 as the low spot, though the elevation differences are very slight. At the north site sampling in February, 2012 do not even indicate an east to west flow through the site (Figure A 5) Data from May 2012 (Figure A 5 ) indicate a definite east to west flow ; potentially towards the north west; these observations coincide with

PAGE 36

36 data from Ju ly, 2012 (Figure A 6) Analysis of the north site indicates groundwater flow east to west, as expected, and potentially northwest towards well N 1 as the low spot. Groundwater directionality, as deduced by water surface elevations, seems to flow as establi shed based on past observations (Geosyntec, 2004). System Operation Initial activation of blower systems occurred during the 2011 vadose zone aeration and shallow air sparging studies ; see Table A 1 for operational details during this period Ball valves were used to equalize air flow injection per well at two of the study sites, to approximately 25 scfm (approx. 707 L/min); due to the natural heterogeneity of soils flow tended to favor one injection well over the other, equalizing pressure rather than flo w Mechanical and electrical difficulties prevented the Roots URAI 32 5 hp blower from operating consistently during operation; however groundwater monitoring was still conducted throughout the study period. At the central and north sparge site s Roots URAI 22 3 hp blowers operated without incident Table A 2 details operational schedules of systems at each site. Components of manifold piping systems were monitored for temperature using an infrared thermometer over the first injection cycle on February 15 th The average outlet piping temperature observed midway through an injection cycle (3 4 hours after startup) at the central and north sites (3 hp blowers) was 136.1 F with an outside temperature of 58 F. PVC piping closer to injection wells measured 100.1 a nd PVC hose measured 94.0 F. Schedule 40 PVC used is rated for use up to 140 F (Harvel, 2012) Figure s 2 6, A 9, and A 10 present injection pressure measured from gauges installed in manifold piping. Injection pressure peaked immediately after blower star tup and began declining instantaneously. After approximately one to two hours of blower

PAGE 37

37 operation injection pressure leveled off, indicating formation of airflow pathways offering little resistance to air flow. The high initial injection pressure observed during every system operation monitoring event indicates bulk mixing occurring during off times. The sites would likely benefit from a greater frequency of pulsed sparge periods throughout 2 hours. A comparison of these graphs shows similar trends, with f low preferentially favoring one injection well over the other at the south site. Regardless, the trend of a negative exponential curve is consistent. Air must form channels in the subsurface, once those channels are formed and air flow remains constant pre ssure drops off substantially. Other researchers have observed similar results in studies utilizing air sparging for remediation of petroleum hydrocarbons (Lundegar d and La Breque, 1995). Ground water mounding is displayed in F igures 2 7, A 11, and A 12. Mou nding indicates a driving force for contaminants, which can be a design concern when off site migration of contaminants is of major concern. Lundegard and LaBreque (1995) and Lundegard (1995) found mounding to be a transient phenomenon and radially symmetr ic around a single sparge well. They observed the water table dropping in monitoring wells closest to the sparge well. While some researchers have utilized mounding as an indicator of ROI, Lundegard and LaBreque (1995) utilized electrical resistance tomogr aphy to characterize air flow during air sparging in a sandy aquifer and observed that radius of influence conclusions drawn based on groundwater mounding had the potential to overstate radius of influence estimates from 2 to 8 times the actual ROI.

PAGE 38

38 Soil gas pressure readings taken at the central and north sites revealed injected air breakthrough to above the water table at distances as short 2.75 ft from the nearest injection well at the central site see Figure 2 8 Closer proximity to sparge wells seeme d to correlate with higher soil gas pressure observed at the central site. At the north site well NG 5 appeared to be too close to the sparge well (2.92 ft to the nearest) and in 3 of 4 monitoring events had one of the lowest pressures recorded. See Figure s A 13 through A 15 for visual soil gas pressure data. Lundegard and LaBreque (1995) observed soil gas pressure influence as far as 23 meters (75.5 ft) from a sparge point with the top of the well screen located 4.6 m below the water table. Given the desir e to treat as large an aquifer area as possible and the short circuiting observed, a deeper well screen would likely have resulted in a greater ROI observed at Klondike Landfill study sites. No established criteria exist for d efinitive determination of th e i deal flow pulsing scenario for a given site. Similarly Lundegard and Anderson (1993) observed that the time required for a given system to establish steady state air injection behavior can vary from hours to years. Typical parameters measured to indicat e radius of influence include soil gas pressure, soil gas composition, dissolved oxygen, groundwater mounding, and helium tracer test results (Lundegard and LaBreque, 1995; Marley et al., 1995) During the initial air flow phase a greater ROI exists becaus e preferential pathways to the water table and vadose zone have yet to develop (Lundegard and Anderson, 1993). The best indicator of ROI observed at the Klondike Landfill was a rain ev ent during monitoring in early M arch where, at the south sparge site, bu bbling could be seen near wells S 3, S 4, S 5, and S 6 ; approximately 22 feet from a sparge point at

PAGE 39

39 well S 3 At steady state the hydraulic pressure outside the region of air flow is balanced by the air pressure. The shallowest depth below the water table utilized by Lundegard and Anderson was 10 ft; based on their research the duration of transient behavior decreased as injection depth decreased. Limited mounding measurements taken in all radial monitoring wells during experimentation at the shallow injec tion interval were insufficient to allow for a conclusion with regard to transient and steady state air flow in the shallow aquifer. Groundwater Chemistry During the study period groundwater was sampled from Feb 2012 through May 2012 when sparging activit ies at the site ended While not considered parameters of interest for this study, they assist in site characterization. Measurements for turbidity, conductivity, and temperature were recorded as required. Average turbidity observed in monitoring wells was 8.08 6.53 12.7 16.7, and 6.63 4.26 NTUs at the south, central and north sparge sites respectively Turbidity was sometime observed in the form of red iron oxide particles floating in collected samples in wells affected by air sparging, though the ir presence did not commonly push the turbidity above 20 NTUs. Average conductivity of water sampled was 1138 143 903.2 181.63, 509.7 137.0 S/cm Average temperature was 20.6 1.87, 22.32 2.11, 21.1 2.11 C The difference in temperature at ea ch site can be attributed to whether sampling occurred in the morning or afternoon. No temporal variation was observed for these parameters other than temperature While dissolved oxygen and ORP are temperature dependant, the narrow range of temperatures ( approximately 18 25 C) observed in addition to the focus on comparing wells affected and unaffected by air sparging caused the author

PAGE 40

40 Calcium, a rsenic and sodium were measured in groundwater samples during ICP AES analysis. B ackground concentrations for arsenic were 0.07 0.02, 0.04 0.02, and 0.02 0.01 mg/L at the south, central, and north sites, respectively. Analysis of results revealed no sample at any time measuring greater than 10 g/L (GCTL) for arsenic after air spa rge commencement during the sparging study period from Feb. Jun 1 st Ba ckground results are slightly less than Keimowitz et al. (2005) observed As approximately 0.3 mg/L downgradient of a closed landfill; De Lemos et al. (2006) observed concentrations of 13.8 g/L at another closed landfill. Calcium and sodium, indicators of leachate presence were present in variable quantities at the sites. Average calcium concentrations observed at the sout h, central and north sites were 81.82 19.25 161.74 21.51 and 77.68 31.65 mg/L, respectively. Average sodium concentration s observed were 61.87 5.99, 9.80 2.06 and 5.27 1.50 mg/L, respectively. VZA a nd shallow AS s y s tems No remedial impact on dissolved iron was observed during VZA in the parallel confi guration at the central site. Average total iron concentration seen at the site increased from 27.49 8.93 to 29.72 4.60 mg/L after approximately 4 months of blower operation Remedial impact of dissolved iron was seen in only well N 7 in the perpendicu lar configuration during vadose zone aeration at the north site, though the effect was short lived, most likely due to the drop in water table at the site (approximately 0.59 feet from Aug. to Nov. 2011), resulting in air injection even farther from the wa ter table. After 4 months of vadose zone aeration at the north site average total iron in all wells raised slightly from 18.24 2.34 to 21.32 5.65 mg/L.

PAGE 41

41 At the s outh site shallow air sparging (10 ft injection well, 5 ft screen) was successful in decreas ing dissolved iron concentrations in several monitoring wells. The well screen was only partially submerged during air injection (water table averaged 6.54 ft bls during air injection), not in line with what current research suggests as optimal conditions for air sparging (Marley et al., 1995; Fields et al., 2002). Just prior to system startup monitoring in Aug. 2011 revealed ferrous iron concentrations of 17.3 and 19.7 mg/L in wells S 3 and S 6, respectively. Average total iron concentrations observed from Sept. 3.21 and 4.19 3.42 mg/L, for S 3 and S 6, respectively. Well N 8 appeared to have been particularly impacted by vadose zone aeration which t ook place prior to air sparging. P rior to vadose zone a eration, in August 2011 the average Mn concentration was 0.35 0.01 mg/L, slightly above the HBRG determined by Gadag bui and Roberts (2003). Interestingly from the time vadose zone aeration ended to the background monitoring prior to air sparging comme ncement average total iron levels at the central and north sites decreased to 16.64 and 9.58 mg/L, in each case, a significant decrease ( p <0.05). One potential explanation for this observation may be that in December, January, and early February when tempe ratures are the lowest, cold air sank through pores loosened in the soil from VZA and oxidized some dissolved iron. Overall vadose zone aeration was unsuccessful and not recommended for changing the redox status of groundwater to remediate dissolved iron ; it relies on downward diffusion of air which is heated by blowers, this phenomenon does not occur and is against the general scientific understanding of subsurface air injection (Johnson et al. 1995; Marley et al., 1995; Fields et al., 2002)

PAGE 42

42 Iron levels at air sparge sites Background measurements taken on February 14 th 2012 revealed average total iron concentrations of 16.64 7.23 mg/L. Higher values were not observed to correlate with proximity to the landfill boundary. Lyngkilde and Christensen (1992) observed ferrogenic zone persisting as far as 350 m from a landfill boundary. In the redox mapping downgradient of a closed landfill, Christensen et al. (2000) found ferrous iron as high as 269 mg/L 63 meters from a landfill boundary with ferrogenic zones persisting to approximately 220 m downgradient of the closed landfill. Average difference between field measured ferrous iron measurements and total iron as measured with ICP AES was only 0.453 mg/L (Fe TOT > Fe(II)). Lyngkilde and Christensen (1992) repor ted iron concentrations 1.5 39.0 (average 8.88) in a ferrogenic zone. During the three and a half month monitoring period (Feb 15 th Jun 1 st ) when blowers operated at regular daily intervals (except at the south sparge site), groundwater parameters were measured to evaluate the success of air sparging at each site. Average t otal iron levels during the sparge study period were significantly decreased from background levels ( p <0.05) in all groundwater monitoring wells at the south site, 4 of 9 wells at the central site, and 4 of eight wells at the north site. Given the effectiveness of initial AS and VZA experiments the background was redefined in some cases, so statistical significance to a true background concentration could be performed. Certain wells ex hibited decreases below the HBCTL but then dramatic bounce backs; in well C 3 initial measured iron concentrations were 15.88 mg/L, dropped to its lowest value of 4.09 mg/L in mid march, then measured 20.02 mg/L in late May. These effects were also seen in wells C 6, C 7, C 8, and to a lesser extent in wells C 9 and N 5.

PAGE 43

43 Reductions in total iron to below the GCTL of 0.3 mg/L were only observed in wells S 6, C 1, C 4, C 5, and N 7, though for a limited number of monitoring events. Reductions to below 4.2 mg/ L For regulatory purposes these wells will be considered affected wells, henceforth. See Figures 2 12 through 2 14 for a plan view of the sites and average total iron levels measured throughout the air injection period (Feb. 15 th Jun. 1 st ). Based on the significant reduction in dissolved iron levels from the start of the experiment to the end of sparging at the central sparge site it can be deduced that preferential flow pathways exist towards monitoring wells S 3 and S 6 at the south site, C 1, C 4, C 5 and C 9 at the central site, toward wells N 4, N 6, N 7, and N 8 at the north site Given the hig h air injection flow rate, to have seen such pre ferential flow pathways develop was unexpected, particularly at the south study site (Fields et al, 2002). Based on these observations a standard radius of influence solely based on distance from sparge wells cannot be concluded. In a document on air sparging design the USEPA (1994) recommends that air sparging for NAPL removal at a site will not be successful if ferrous iron measured at the site is greater than 20 mg/L; the iron will oxidize (desired in this experiment) and on oxidation can reduce soil pore space and reduce permeability. Johnson (1998) noted that iron precipitation can be accounted for as an im pediment to contaminant removal in air sparging. Other authors (Carter and Clark, 1995; Johnson et al., 1995; Schrauf and Pennington, 1995) have noted the presence of high iron (and manganese) only as a nuisance for remediation of more deleterious contamin ants such as BTEX chemicals, trichloroethylene Pres sure backup

PAGE 44

44 in the system (see F igures 2 6, A 9, and A 10 ) was not observed at the Klondike sites, potentially due to the high flow rates used. Monitoring occurring 49 and 50 days post blower shutdown r evealed persistent low iron levels in wells S 3 and S 6 at the south site. See Figures 2 15 through 2 17 for post shutdown total iron levels measured by ICP AES. Iron levels in wells S 4 and S 5 rose significantly to near background levels and above 4.2 mg /L. It is possible that upon sampling on July 19 2012 water had not traveled from the zone of wells S 4 and S 5 and impacts in wells S 1 and S 2 could be observed in the future. At the central site post shutdown monitoring revealed iron levels returning t o background in 3 of 4 affected wells (C 1, C 4, and C 5). If water flowed from C 5 to well C 2 in the 49 days between shutdown and sampling it may have been diluted with high Fe(II) water. Well C 3 exhibited a significant drop in total iron, from 20.02 on May 24 th to 1.42 mg/L on July 19 th Based on estimated groundwater flow the author was not able to see a clear movement of low iron water to wells farther downgradient. Water from well N 4 at the north site may have traveled towards N 1 downgradient; Fe TO T at N 4 rose from 0.56 to 9.90 mg/L from May July, in N 1 the level dropped from 11.41 to 7.16 mg/L. Iron levels rose in wells N 7 but not in N 8, closer to the landfill boundary. Soil gas pressure data indicates that high vertical conductivity exists in the aquifer and air flow in the aquifer may have allowed directly injected air to breakthrough to the water table quickly, thus limiting iron oxidation effectiveness and radius of influence. Manganese levels at air sparge sites Figures 2 18 through 2 2 0 detail observed total manganese concentrations in groundwater monitoring wells throughout the study period. Monitoring data during system operation period (Feb 15 th June 1 st ) reveals a significant ( p <0.05) reduction in

PAGE 45

45 dissolved manganese to below the health based guideline (0.33 mg/L) from background levels to the last monitoring event durin g system operation only in well C 4. However in a greater number of wells increases in dissolved manganese concentrations were observed. Of importance, at the nort h air sparge site the highest manganese levels were observed in the two wells which consistently had the lowest dissolved iron concentrations, N 7 and N 8. In impacted wells N 5, N 6 and N 7 manganese levels rose significantly ( p <0.05) from background moni toring to above the GCTL and health based guideline (N 6 and N 7) at the last monitoring event before blower shutdown. D uring air sparging experiments Mn levels at well N 8 averaged 2.17 1.27. In impacted well C 5 at the central sparge site the same tre nd was observed (lower iron, higher manganese) at a significant level ( p <0.05). Similarly at the south sparge site well S 6 was significantly affect by air sparging in terms of iron levels and conversely the average Mn concentration observed rose from 0.23 0.01 to 2.08 0.04 mg/L. See Figure 2 21 for the correlation. ORP values observed in this well correspond to Mn reducing zones for the given pH (approximately 6.81 throughout the study period after air sparge commencement). Wells positively impacted by aeration (decreased iron levels, raised ORP values) did not surpass the Mn reducing range outlined by Stumm and Morgan (1996). Stumm and Morgan (1996) stated that oxygen saturated water should have a redox potential as high as 710 mV. The greatest redox potentials observed at air sparge sites were in wells N 7 and N 8 at the north sparge site with average ORP values of 131 40 and 251 70 mV after air sparge commencement and before system shutdown. Based on the

PAGE 46

46 results of this study, it was not possible to raise the redox potential into the aerobic zone and above the manganese reducing zone. Ellis et al. (2000) were able to remove dissolved iron with air injection; however manganese oxidation was only shown to be possible with chemical oxidant (KMnO 4 ) ad dition. Iron is a dominant element in soil at the Klondike landfill and given the observation by many researchers (Lovley, 1991; Lyngkilde and Christensen, 1992; Ludvigsen et al., 1998; Christensen et al., 2000) that the reduction of Mn and Fe occur togeth er and that Fe reducers can outcompete for electron acceptors in soil, leaving reserves of reducible Mn prior to air sparging. Post shutdown groundwater monitoring revealed no consistent trends in Mn levels at the south site. At the central site the Mn con centration in groundwater from well C 4 returned to background levels. At the north site in well N iron level, a Mn level of 13.45 mg/L was observed. Mn concentration in well N 7 returned to pre sparging level (0.33 mg/L), whic h was coupled with an increase in Fe. Wells N 2, N 3, and N 4 saw increases in Mn concentration from previous monitoring which were significant ( p <0.05). Air sparging at the Klondike landfill may have had the capability of raising redox potential from iro n reducing to manganese reducing zones and causing significant increases in dissolved manganese observed in monitoring wells. O RP and dissolved o xygen Oxidation reduction potential (ORP) correlated well to dissolved iron concentrations. Higher ORP values correlated to lower dissolved iron concentrations, as detailed in F igure 2 22 In eleven air sparge affected wells (average < 4.2 mg/L Fe TOT observed throughout the study period) the average ORP observed after blower startup was +6, +43, and +101 mV at the south, central, and north study sites, respectively.

PAGE 47

47 Compared to average values at unaffected wells, 86, 38 and 48 mV, respectively. The redox range of iron reduction to Fe(II) (approx. 59 to 590 mV for pH=7) was surpassed in several monitoring well s in the study. Interestingly, in a comparison of commonly measured parameters at air sparge sites to find radius of influence ORP was not cited (Marley et al., 1995). Redox potentials observed in manganogenic and ferrogenic zones at landfills by Lyn g kild e and Christensen (1995) ranged from 83 to +409 and +20 to +319. De Lemos et al. (2006) observed high Fe/As (350 ppm) in a batch experiment corresponding to an Eh of approximately 450 mV; in a field study at a landfill measurements averaging approximatel y 100 mV were noted. In a laboratory experiment on the associations of selenium at under variable redox conditions, Peters et al. (1997) found air sparging incapable of raising the redox potential of estuarine sediment slurry above 350 mV Post shutdow n monitoring revealed ORP values of 33, 77, and 28 mV at the south, central, and north sites, respectively. All affected wells exhibited a decrease in ORP from monitoring during air injection, as expected. See F igures A 16 through A 18 for detailed depi ction of ORP values changing with time in each groundwater monitoring well. Lundegard and Labreque (1995) found dissolved oxygen to be a semi accurate indicator of radius of influence during a shallow air sparging experiment. The authors plotted were able to characterize dissolved oxyge n behavior in one well through the duration of air sparging. A period of monitoring was conducted during blower operation at the south sparge site on January 12, 2012, prior to the full experimental startup. During this trial dissolved oxygen levels only increased significantly from sampling the

PAGE 48

48 day prior in well S 6, where preferential flow pathways are known to exist. Wells S 7, S 8, and S 9 erupted with flowing water and would likely have exhibited near saturation dissolved oxygen levels had sampling been feasible. The wells were later abandoned to avoid airflow short circuiting. Dissolved oxygen levels never entered the maximum achievable oxygen concentration range of 8 10 mg/L possible for air saturated water (Fields et al., 2002), even in affected wells. See Figures A 19 through A 21 for DO vs. time plots. Background dissolved oxygen concentrations measured just prior to air sparge start up in February 2012 were 0.51 0.03, 0.77 0.15, and 1.06 0.26 mg/L, for the so uth, central, and north sites, respectively. Average dissolved oxygen concentration observed in affected wells vs. unaffected wells throughout the study period after blower startup were compared. The difference was statistically insignificant ( p <0.05) at t he central site. At the south and north sites the difference was statistically significant (average DO in affected wells 1.32 and 2.60 mg/L). Increased DO levels do not appear to be necessary for Fe(II) oxidation and concentration decrease to below 4.2 mg/ L, so for the purpose of this study were not considered for determination of effective ROI. Post shutdown monitoring revealed a significant ( p <0.05) drop in DO in all monitoring wells. It is possible that because of the high demand for oxygen for organic matter breakdown, all injection oxygen contained in air was immediately utilized and never persisted in the free dissolved form. In research on landfill sites experiencing similar reductive dissolution issues Bjerg et al. (1995) and Grossman et al. (2002) observed leachate and groundwater with no detectable DO; Nanny and Ratasuk (2002) surveyed three impacted landfills with DO levels ranging from 0.0 to 1.10 mg/L in leachate. Heron

PAGE 49

49 and Christensen (1995) observed DO < 1 mg/L in an Fe/Mn reducing plume downg radient of a landfill. Dissolved oxygen concentrations possible through in jection of pure oxygen rather than air, range from 40 50 mg/L (Fields et al., 2002). ORP levels and dissolved oxygen concentration in reference wells monitored throughout the stud y decreased slightly towards a more reduced state. ORP values shifted from an average of 38 mV at the study commencement (Feb 2012) to 66.5 mV (May, 2012). Average ORP values observed in the shallow and deep well were 69 and 41 mV, respectively; corres ponding average DO observed were 0.78 and 0.79 mg/L. Average DO values shifted from an average of 0.49 (Feb. 2012) to 0.79 mg/L (May 2012), well in the range of DO values observed by other researchers at iron contaminated landfill sites. pH at air sparge sites Background monitoring at the sparge sites revealed average pH levels of 6.81 0.21, 6.55 0.08, 6.48 0.27. The higher standard deviations observed at the south and north sites are most likely the result of impacts seen from the initial shallow ai r sparging at the south site and vadose zone aeration at the north site During the sparge study pH in affected wells (6.76) at the south site was significantly greater than in unaffected (6.58); the same trend was noted at the central site between affected (6.36) and unaffected wells (6.30), though to a lesser extent. The opposite trend was observed at the north site between affected (5.79) and unaffected wells (6.36). Although these trends were observed, not all affected or unaffected wells at a particular site had lesser or greater pH value s than their counterparts; see F igure A 22 for a graphical depiction of pH over time at sparge sites. No research was found in the literature regarding pH and air sparging, indicating that no relationship has been establ ished.

PAGE 50

50 Interestingly at the north sparge site the pH in wells N 7 and N 8 dropped below 6 (pH remained above 6 in all other wells) and averaged 5.81 and 4.75, respectively, during the sparge study. Background pH in these wells measured in Aug 2011 reveal ed measurements of 6.26 and 6.27, respectively; not significantly different ( p <0.05) than other monitoring wells at that site. Well N 8 is much closer to landfilled waste than other monitoring wells and the pH could be impacted by organic acids leaching fr om the landfill. The significant decrease in pH in wells N 7 and N 8 to well below the EPA recommended range (6.5 8.5) may cause concern for air sparging at future iron contaminated sites. Post shutdown monitoring revealed pH rising back to background le vels (6.45) in well N 7 and rising to 5.07 in well N 8, perhaps indicating sparging was the cause of the decreased. pH in reference wells did not change to a great extent through the study and ranged from 6.16 to Overall average pH observed in the shallow and deep reference well was 6.46 and 6.31, respectively. In a 5 month air sparging remediation project for petroleum hydrocarbons, Aivallioti and Gidarakos (2008) observed no substantial changes in pH, only minor variations. Lyngkilde and Christensen (1992 ) observed average pH values of 6.38 and 7.18 in ferrogenic and manganogenic redox zones downgradient of the Vejen Landfill in Denmark, respectively; average pH in the aerobic redox zone was 5.27. Nanny and Ratasuk (2002) observed landfill leachate with pH of 6.6, 6.8, and 7.1. pH of landfill leachate generally increases over time (going from the acid forming to methanogenic phase), and the Klondike landfill would be considered mature, in the methanogenic phase, with a leachate pH most likely around 7.5 9 (Ehrig, 1983).

PAGE 51

51 USEPA (2012) recommends pH values of 6.5 8.5 for freshwater; however Florida Department of Environmental Protection (1992) published results of background hydrogeochemistry in the state. The northwest district (containing the Klondike Lan dfill) reported a median pH value of 4.9 for the surficial aquifer, the lowest of any state district (next closest 5.6 for the Suwannee river district). This is due to the low buffering capacity quartz dominated soil in the Florida panhandle allowing for c arbonic and organic acids to dominate the pH (FDEP, 1992). Total organic carbon Total organic carbon levels observed in groundwater samples taken from select wells (2 from the south site and 3 from the central and north sites) during sampling revealed av erage TOC background concentrations of 24.11 4.68, 11.93 4.59, and 9.33 3.30 mg/L at the south, central, and north sites, respectively. It was suspected that wells affected by air injection in terms of lessened iron concentrations would also tend to have lower TOC concentrations. TOC in the reference well declined slightly from samples Feb to April 2012, as did total iron concentration measured; interestingly the linear regression of this data gave an r squared value of 0.999 (see F igure A 23 ), confir ming the hypothesis. However in all three of the study areas the trend was not clear and no linear relationship could be established. In affected wells C 5 and N 8 the highest TOC concentration did correlate to the highest total iron concentration and vice versa. Theis and Singer (1974) found higher molecular weight compounds more strongly bonded with Fe(II) and thus oxidation by chemical means was more difficult. Lyngkilde and Christensen (1992) observed nonvolatile organic carbon concentrations of 1.2 71.9 mg/L (avg. 6.43) in ferrogenic redox zones and 1.3 7.5 mg/L (avg. 2.36) in

PAGE 52

52 manganogenic zones. Calace et al. (2001) observed older landfill leachate containing higher molecular weight compounds which had a strong affinity for cadmium and copper. Ma jor Observations and Conclusions Results from system operation data taken during pulsed air injection, coupled with groundwater chemistry indicate air sparging does have the ability to reduce dissolved iron concentrations in the aquifer. Iron and manganese oxide reduction have been observed to occur concurrently (Champ et al., 1979; Lindberg and Runnels, 1984; Lovley, 1991; Lyngkilde and Christensen, 1992) The observation in this study that lower concentrations of iron correlated with higher concentrations of manganese may indicate that iron reducing bacteria outcompete manganese reducing bacteria at background Eh levels During air sparging the death of iron reducing bacteria may afford manganese reducers the ability to flourish. Consistent with observat io ns of other air sparge sites (Lundegard and LaBreque, 1995) radial continuity of air flow around sparge points was not observed; for example at the central site strong breakthrough to the water table was observed approximately 5 feet from the injection w ell 1 to the west, corresponding to decreased iron levels seen in monitoring wells to the east see Figures 2 8 and 2 13 The opposite effect was noted in injection well 2 at the central site, higher soil gas pressure to the east, lower iron concentrations to the west. This indicates a greater affected linear zone to the east of injectio n well 1 and vice versa. The flow of groundwater should gradually expand the zone of influence towards eleven mile creek; however in the absence of periodic blower operatio n, the continued addition of organic matter from the landfill to the aquifer and abundance of iron oxides

PAGE 53

53 will likely cause an increase in dissolved iron concentrations in the aquifer at study sites over time. Cost Analysis In addition to the capital costs of aeration blowers, injection wells, and manifold piping, the necessity of electrical power to run blower systems contributes significantly to the project cost. The average power cost to run the central or south site in a mont h where blowers operate d 12 hrs per day in 2 6 hour cycles was $137.53. Due to blower breakdowns estimation for the south site was not possible. Installation of 15 ft injection wells ran approximately $400 per well. Mobilization costs for the drilling rig to access the relative ly remote site ran approximately $350 per job. Field s et al. (2002) estimated a cost of 0.01$/pound of oxygen supplied for air injection, with the cost rising to 0.10$/pound of oxygen for pure liquid oxygen injection (max DO 40 to 40 mg/L); Interestingly h ydrogen peroxide addition costs approximately $10/pound of oxygen supplied though with no increase in achievable DO concentration. In an analysis of 32 sites utilizing pump and treat technology for superfund cleanup the USEPA (2001 b ) average capital cost w as 4.9 million dollars with a staggering $770,000/year of O & M costs. Summary Air sparging took place at 3 sites downgradient of a closed landfill for a 3.5 month study period in 2012 for the purpose of reducing dissolved iron and manganese concentration s in surrounding groundwater monitoring wells. Decreased iron was achieved in most monitorin g points from background levels; however dissolved manganese increased in wells with the most strongly reduced iron concentrations. No arsenic was detected in any m onitoring well after project startup. Injected air appeared

PAGE 54

54 to breakthrough to the water table within a few feet of the sparge point, as observed at sites with soil gas monitoring wells. The limited ROI observed may be due to several factors, the h eat impa rted to the injected air by passage of air through the blower systems will cause the air to become less dense and thus have a more difficult time penetrating farther from the sparge point for a large ROI effect additionally the ROI could have probably bee n expanded with the use of deeper sparge wells (average top of well screen 5.72 feet above the water table in this study). If the aerated zone remains oxidizing for Fe(II) an extended zone of influence may be observed with groundwater flow. ; this was obser ved to a limited extent, at the north site during post shutdown groundwater monitoring. While aquifer reareation m ay not be feasible for large scale use due to its limited potential to decontaminate a large area in sandy soil other oxidation techniques, s uch as in situ chemical oxidation may be experimented with for potential to solve the Fe/Mn contamination problem. The ability of a short period of air sparging to oxygenate an area for a long period of time needs to be evaluated. If the air sparge blower were made to be portable and travel from well to well along a transect of land dow ngradient of a landfill it may be feasible. However the potential rise in dissolved Mn groundwater concentration accompanying lower Fe concentrations must be examined and ev aluated.

PAGE 55

55 Table 2 1. Well d etails Casing Information Screen Information Study Site Well ID Depth bls (ft) Well Diameter (inches) Material Screen Length Filter Pack (mesh) Slot Size South S 1 through S 6 10 2 PVC 2.5 20/30 0.01 South SIW 1 & SIW 2 15 2 PVC 3 10/20 0.02 Central C 1 through C 9 8 2 PVC 2.5 20/30 0.01 Central CG 1 through CG 10 5 2 PVC 2.5 20/30 0.01 Central CIW 1 & CIW 2 15 2 PVC 3 10/20 0.02 North N 1 through N 8 8 2 PVC 2.5 20/30 0.01 North NG 1 through NG 6 5 2 PVC 2.5 20/3 0 0.01 North NIW 1 & NIW 2 14 2 PVC 3 10/20 0.02 Table 2 2. Operational schedule of air injection s ystems in 2012 Study Site Dates Operated Total Approximate Hours South February 14th March 3 8, May 17th June 1st 2012 576 666 Central and North F ebruary 14th June 1st, 2012 1272

PAGE 56

56 Figure 2 1. Air sparge site locations r elative to the Klondike La ndfill and existing compliance w ells

PAGE 57

57 Figure 2 2. South air sparge site l ayout

PAGE 58

58 Figure 2 3. Central air sparge s i te l ayout

PAGE 59

59 Figure 2 4. North air sparge site l ayout

PAGE 60

60 Figure 2 5. South site water surface elevation maps, values represent ft NAVD A) February 14, 2012 B) May 24, 2012

PAGE 61

61 Figure 2 6 Typical injection pressure vs. time at the north study s ite

PAGE 62

62 Figure 2 7 South site groundwater mounding during s parg e c ycles A) February 15 2012 B) May 25, 2012

PAGE 63

63 Figure 2 8 Central site p ressur e in soil gas monitoring wells during air sparge injection c ycles (all units inches of wa ter) A) March 2, 2012 B) May 3, 2012

PAGE 64

64 Figure 2 9 South site t otal i ron vs. t ime

PAGE 65

65 Figure 2 10 Central site total i ron vs. time Figure 2 11 North site total i ron vs. time

PAGE 66

66 Figure 2 1 2 South site a verage total iron concentration (mg/L) from Mar

PAGE 67

67 Figure 2 1 3 Central site average total iron concentration (mg/L) from Mar. Figure 2 1 4 North site average total iron co ncentration (mg/L) from Mar.

PAGE 68

68 Figure 2 1 5 South site average total iron concentration (mg/L) measured during sampling on July 19, 2012, 49 days post shut down

PAGE 69

69 Figure 2 1 6 Central site average total iron concentration (mg/L) measured during sampling on July 19, 2012, 49 days post shut down

PAGE 70

70 Figure 2 1 7 North site average total iron concentration (mg/L) measured during sampling on July 20, 2012, 50 days post shut down F igure 2 1 8 South site m anganese vs. time

PAGE 71

71 Figure 2 1 9 Central site m anganese vs. time

PAGE 72

72 Figure 2 20 Nort h site m anganese vs. time

PAGE 73

73 Figure 2 21 Total manganese vs. total iron in wells affected by air s parging (Avg. Fe TOT < 4.2 mg/L during study period)

PAGE 74

74 Figure 2 22 ORP vs. d i ssolved iron in wells affected by air s parging (Avg. Fe TOT < 4.2 mg/L during study period)

PAGE 75

75 CHAPTER 3 END OF LIFE ASSESMENT OF TWO CAL CIUM CARBONATE BASED PERMEABLE REACTIVE BARRIERS FO R REMEDIATION OF IRO N CONTAMINATED GROUNDWATER AT A CLO SED LANDFILL Introductory Remarks The Klondike Landfill, located in Escambia County, Florida, is a closed Class I MSW landfill wh ich operated from 1976 until 1982 as an unlined landfill. In 1987 the county began detecting levels of iron and manganese much greater than regulatory limits in groundwater monitoring wells downgradient of the landfill at which point semi annual monitoring commenced. Iron in the groundwater is predominantly present in the mobile ferrous form (Fe(II)). It is strongly suggested that the source of the iron is the soil; in anaerobic aquifers naturally occurr ing iron oxides and other iron containing minerals are reduced from Fe(III) (ferric iron) to soluble Fe(II) and manganese is reduced from Mn(IV) to Mn(II) (Lepp, 1975; Nicholson et al., 1983; Lovley, 1991; Bjerg et al., 1995; Heron and Christensen, 1995; Rao et al., 2008; Wang et al., 2011). Waste disposal si tes, particularly unlined ones, have been demonstrated to create anaerobic aquifer conditions, in turn causing the reductive dissolution of iron and manganese soil minerals; in some cases this causes arsenic release, due to its tendency to be bound onto ir on soil minerals (Keimowitz et al., 2005; De Lemos et al., 2006; Ghosh et al., 2006; Parisio et al., 2006; Di Palma and Mecozzi, 2010; Wang et al., 2011). Though limestone has not been shown to remove arsenic, iron oxides have (Joshi and Chaudhuri, 1996; T hirunavukkarasu, et al., 2003; Gibert et al., 2010). If the limestone PRBs could precipitate a layer of iron oxides they could, in turn remove arsenic from the groundwater.

PAGE 76

76 Calcium carbonate based materials have shown great promise in removal capacity of d ivalent and trivalent metal cations at approximately 90% efficacy, more typical contaminant species, specifically manganese (Aziz and Smith, 1992; Aziz and Smith, 1996), copper (Aziz et al., 2001; Aziz et al., 2008), cadmium, chromium (Cr(III)), lead, nick el, and zinc (Aziz et al., 2008). Based on a current literature review CCBM use for ferrous iron remediation has only been demonstrated in both batch (Mettler et al., 2009; Wang, 2011) and column (Wang, 2011) laboratory scale studies prior to the Klondike Landfill PRBs experiment (Wang, 2011). Wang (2011) conducted a pilot scale PRB experiment at the Klondike Landfill. The current study is an extension of that study and includes data collected as part of the initial pilot project. Two pilot scale permeable reactive barriers (PRBs) were installed 20 meters downgradient of the landfill boundary. Barriers were installed in June and July of 2009. One PRB consists of limestone particles (7 10 mm diameter) and the other of crushed concrete (70 150 mm diameter) Dimensions of each barrier are ap proximately 20 ft long (perpendicular to known groundwater flow ), 15 ft deep, and 3 ft wide Groundwater monitoring wells were installed up and downgradient of the barriers, as well as on the sides of the barrier and in t he barrier trench to ensure groundwater is not circumventing barriers as it flows. Upgradient and downgradient monitoring wells were installed at two depths, 8 ft (shallow) and 14 ft ( deep), each with a 5 ft screen. Permeable reactive barrier technology is currently considered a proven technology ( Cantrell et al., 1995; Henderson and Demond, 2007 ). In a comparison of PRBs vs. conventional pump and treat technology for groundwater treatment at United States superfund cleanup sites, USEPA (2001b) observed the average cost of PRB technology

PAGE 77

77 to average $730,000 (capital cost) whereas typical pump and treat capital costs alone were $ 4.9 million, plus an average of $ 770,000 for operation and maintenance costs per year. Ferrous, total iron, and total manganese conc entrations observed in downgradient monitoring wells of both PRBs began to rise significantly ( p <0.05) between monitoring events in late 2010 (October) and mid 2011 (May). Quarterly monitoring then continued for a period of one year. Prior to 2011 average total iron and manganese concentrations measured in downgradient wells of the limestone PRB was 2.23 1.39 and 0.42 0.37 mg/L, respectively. Post 2011 downgradient total iron and manganese levels measured 11.87 7.92 and 0.38 0.20 mg/L. Similar resu lts were observed in the downgradient wells of the crushed concrete PRB. The high standard deviation is mainly due to the difference in metal concentrations between shallow and deep monitoring wells. See T able 3 1 for a comparison of metal concentrations a nd other groundwater parameters measured in deep and shallow downgradient monitoring wells. The objectives of the current study are to track the performance of the PRBs in their third year of operation, determine if groundwater flow direction shifted durin g operation. Additionally after excavation of the barriers and sample collection from variable depths, the study aimed to determine if remaining adsorption capacity could be measured ex situ and correlated to in situ performance, of adsorbed iron and manga nese could be determined via leaching tests, and identify iron minerals on collected particles.

PAGE 78

78 Methods and Materials Site Description The research site is located within the 126 acre Klond ike Landfill (refer to chapter 2 for complete site description). Tw o pilot scale permeable reactive barriers were installed during June and July 2009 in the vicinity of compliance monitoring well DW 4S. Groundwater monitoring data revealed decreasing effectiveness of barriers in May, 2011. See F igure s 3 3 and 3 4 for a co mparison of groundwater monitoring data. Prior to this monitoring event average dissolved iron concentrations of below 4.2 mg/L were observed in downgradient monitoring wells. Geosyntec (2004) estimated groundwater velocity at the site to vary between 0.20 to 0.39 ft/day in the study area. Three distinct hydrogeologic units are present at the site, all study wells and the PRBs are present in the shallow unit which goes from +16 to 8 ft NAVD and consists of clayey sand with clean sand outcroppings. The area where the PRBs site (towards the northern boundary of the site) tends to be more clayey and afte r precipitation events water could be observed ponded on the soil surface. In order to study the PRBs and their diminishing efficacy as well as to determin e the cause it was first important to determine if groundwater flow reversed directions or was simply circumventing the reactive trench, potentially due to clogging, the water surface elevation of wells on the sides of the trench was compared with the wate r surface elevation in upgradient wells. Survey data of top of casing (TOC) elevation as well as depth to water measurements were used for analysis. Groundwater Sampling and Analysis Prior to groundwater sampling at a study site the depth to groundwater w as measured in all site wells with a water level meter (Solinst Model 101). Groundwater

PAGE 79

79 samples were collected with a peristaltic pump (Geotech Series II) following a purge of at least one well volume and then stabilization of pH, DO, turbidity, temperatur e, and specific conductivity over three consecutive reading as specified in Florida Statutes 2200. ORP (Oakton ORPTestr10) and Fe(II) (Hach Ferrover) levels were both measured in the field after parameter stabilization occurred. Florida form FD 9000 24 was used for field data collection. Groundwater samples to be analyzed for total metal contents were preserved immediately after field collection using nitric acid to a pH of less than 2, immediately placed on ice, and then transported to the laboratory for a nalysis by inductively coupled plasma atomic emission spectroscopy (ICP AES). All samples were analyzed in duplicate and appropriate blanks and calibration checks were included in each analysis. Statistical analyses were performed using one way analysis of variance (ANOVA) followed by a t test considering a p value of less 0.05 was considered statistically significant. Composition of Original Reactive Materials Samples of original limestone and crushed concrete were crushed with a ball mill (SPEX 8000M Mix er/Mill) until all sample passed through a 0.425 mm sieve, as not to preferentially differentiate the sample based on hardness. Samples were then digested according to EPA method for total metals analysis. After digestion the extract was then analyzed with ICP AES. The purpose of this analysis was to determine if iron was present in the original material because leaching tests to remove adsorbed iron and manganese may also dissolve the original material and the presence of these metals in the CCBMs should b e accounted for.

PAGE 80

80 Remaining Iron Adsorption Capacity During barrier excavation samples were gathered and submerged in a solution of 30 mg/L Fe(II) produced from ferrous chloride tetrahydrate salt. A 1000 ppm Fe(II) solution was prepared in the laboratory th e day prior to excavation and bubbled with Nitrogen (N 2 ) gas to prevent oxidation. This stock solution was then combined with nanopure water (18.2 Mohm/cm) in the field for a total liquid volume of 100 mL for limestone experiments and a variable recorded v olume for crushed concrete experiments; due to the high degree of variability between crushed concrete particle sizes and shapes the amount of liquid needed to submerge the particles varied. The 30 mg/L solutions were then combined with a measured mass of reactive materials. Due to sample were measured using the portable Hach FerroVer method. The solutions were allowed one hour of reaction time before analysis of pH and s olution ferrous iron. Due to the small size of limestone particles and thus the inability to draw samples directly from the pore water interior, slight agitation of the samples by hand were performed prior to analysis to ensure that all solution water cont acted the particles prior to analysis. To ensure ferrous iron oxidation did not play a role in the decrease of ferrous iron measured control samples consisting only of Fe(II) solution were utilized. Reactive Media Collection and Preparation In this study r eactive media from a crushed concrete and a limestone permeable reactive barrier were excavated and experiments were performed on them to determine remaining iron adsorption capacity, total iron, manganese, and arsenic adsorbed to the surface of particles, and character of surface minerals.

PAGE 81

81 An excavator was used to remove permeable reactive barrier materials from variable depths. Two gallon size bags of material were collected from each depth; one was immediately placed on ice for transport back to the lab oratory and the other was utilized for field experiments to determine remaining ferrous iron adsorption capacity. All experiments and analysis collected on barrier materials were also conducted on original crushed concrete and limestone particles left over from barrier installation. Determination of Adsorbed Iron and Manganese Laboratory batch experiments were performed to determine the quantity of iron and manganese sorbed to the surface of limestone and crushed concrete particles. Preliminary experiment s were conducted to determine whether rinsing the particle surface would cause a loss of sorbed iron. Original limestone particles were rinsed with nanopure water (18.2 MOhm/cm), then soaked in dilute nitric acid (pH=1) for 8 hours. Particles were then rin sed again with nanopure water, and oven dried at 80 C. Aliquots of 25 g of material were then submerged in 500 mL of a 50 mg/L Fe(II) solution prepared from FeCl 2 salt and bubbled with N 2 gas to achieve a DO concentration of 1 0.05 mg/L. Solutions were t hen rotated at 30 RPM for 1 hour. The mixture was then filtered and one particle was collected from each bottle for SEM/EDS examination. One half of the particles were rinsed three times with nanopure water (washed particles) while the other half were not (unwashed particles). HCl extraction was performed on both sets of particles. A 0.5 M HCl extraction to determine adsorbed iron and manganese was modified from Heron et al. (1994); an extraction recommended for its speed and ability for estimation of amorp hous iron(II), iron(II), FeS and FeCO 3 (expected to be present in the samples). An experimental liquid to solid ratio of 1/10 (m/m) was used for all limestone

PAGE 82

82 extraction experiments. 25 g of material and 250 mL of 0.5 M HCl were combined in 500 mL HDPE bo ttles. Samples were rotated for 24 hours at a speed of 30 RPM. The samples were removed and 100 mL of the mixture was filtered through a 0.45 m filter and preserved with trace metal grade nitric acid to pH < 2 for ICP AES analysis. All leaching experiment s on limestone were performed in triplicate; due to the lack of material and difficulty working with larger size material, experiments conducted on crushed concrete was performed singularly. Characterization of Particle Precipitates X ray diffraction (XRD) was utilized to determine the presence of iron minerals which may have formed over the life of the PRBs. Preparation for XRD analysis was performed on washed and unwashed crushed samples, as well as on surface particles collected from unwashed samples. Or iginal limestone and crushed concrete, samples collected for 18 24 inches below ground surface (never submerged in the water table), and 9 feet below ground (submerged throughout the life of the PRBs) were selected for XRD analysis. Materials to be crushed were first triple rinsed with nanopure water (18.2 MOhm/cm), then oven dried at 80 C for 24 2 hrs. Once cooled the materials were crushed in a ball mill (SPEX 8000M Mixer/Mill) until all sample passed through a 0.425 mm sieve, as not to preferentially d ifferentiate the sample based on hardness. An ultrasonic bath was used to remove surface particulates from unwashed materials for additional XRD analysis. In the case of the relatively uniform limestone particles 200 grams of material and 250 mL of nanopu re water (18.2 Mohm/cm) were combined. The crushed concrete particles were each weighed separately, measured with a tape measure in three dimensions for calculation of approximate surface area. Nanopure water was then added until particles were submerged a nd the volume was

PAGE 83

83 recorded. Mixtures of particles and nanopure water were then placed in an ultrasonic bath (Fisher Sci, FS60) for 10 minutes to remove surface precipitates. Due to the porous nature of the concrete the outer layer of precipitates was more recalcitrant and surface precipitates, respectively. The remaining mixture was then f iltered through a 0.45 m membrane (nylon 0.45 m, FisherSci, Inc.). The resulting liquid was then preserved with trace metal grade nitric acid to pH < 2 for metal analysis by ICP AES. The material remaining on the filter was then scraped into a ceramic cr ucible and oven dried at 80 C for 24 hours, then stored in a desiccator until analysis. In order to determine total metal contents of the surface layers a separate series of ultrasonic baths were performed, the remaining mixture was filtered (0.45 m), t he material collected on the filter was collected, dried in an oven at 80 C, and digested according to EPA method 3050B for total metals analysis. After digestion the extract was then analyzed with ICP AES. Scanning electron microscope (SEM) images coupled with energy dispersive spectroscopy (EDS) elemental analysis was used on select samples. From Fe(II) laboratory adsorption experiments an original, clean particle, a washed particle and an unwashed particle were analyzed. Original limestone and crushed co ncrete and field analysis required samples to be dry, so prior to analysis all samples were oven dried for a minimum of 24 hours at 80 C and stored in a desiccator until anal ysis.

PAGE 84

84 Results and Discussion Groundwater Monitoring Data Hydraulic d ata Groundwater table elevation data revealed groundwater flow direction did not change throughout the course of the study; there is also evidence that the groundwater did not begin to by pass the PRBs. Well BD3 was damaged sometime in early 2011, and thus well BS2 was used for comparison to surrounding wells. Because both shallow and deep monitoring wells are screened in the shallow, unconfined aquifer, the differential screen depth will n ot affect comparison. Figure 3 2 confirms a positive differential between upgradient wells (AD3 and BS2), wells surrounding the PRBS (AD1 AD5, BD1, and BD5), and downgradient wells (AD7 and BD7), indicating groundwater flow persisting from east to west dur ing the course of the study. Average groundwater elevation in the PRB trench during the study period was 11.22 ft NAVD (2.41 ft bls) at the limestone PRB and 11.36 ft NAVD (2.88 ft bls) at the crushed concrete PRB. See Tables 3 1 and 3 2 for a comparison o f up and downgradient parameters in deep and shallow monitoring wells. Groundwater c hemistry A significantly higher concentration of dissolved iron and manganese was observed in shallow upgradient monitoring wells over deep, while ORP and DO measurements w ere similar and not significantly different (Table 3 1) In a comparison of deep and shallow downgradient wells (Table 3 2) it was observed that Fe TOT and Fe(II) concentrations in shallow downgradient wells of the crushed concrete PRB averaged < 4.2 mg/L d uring the study period; also in those wells, the greatest rise from initial pH was observed, with low ORP values. In the downgradient wells of the

PAGE 85

85 limestone PRB lower Fe concentrations were observed in shallow monitoring wells, this is interesting, because deep upgradient wells have significantly lower iron concentrations ( p <0.05). Arsenic concentrations were significantly less than upgradient wells in shallow screened downgradient wells of both PRBs, but not in deep screened downgradient wells during the s tudy period (May 2011 Feb 2012). pH increase downgradient of PRB trenches was observed and likely due to the presence of carbonates dissolved from reactive materials. The crushed concrete PRB caused a greater increase in pH observed (average of 8.69 in d owngradient wells in the first year of operation ) compared to the limestone PRB. Data from the current study indicated the low pH of well BD9 (avg. 6.48) correlated with the greatest observed downgradient total iron concentration (32.19 mg/L) This may indicate groundwater short circuiting around the PRB towards well BD9. The same trend was observed in well BD7, in the middle of the PRB, while not in wells BS6 and BS8 which surround well BD7, unlikely the result of short circuiting In laboratory column experiments for Fe(II) treatment with concrete and limestone pH decreased as Fe(II) concentration increased in the effluent (Wang, 2011), the same trend observed in the field. In laboratory experiments, t he predicted longevity of the limestone barrier was estimated to be 3.6 7.1 years and the crushed concrete PRB was 3.0 5.9 yrs (Wang, 2011) These predictions were based on groundwater flow ve locity of 0.1 0.2 m/day and a Fe(II) removal capacity based on column test results of 4.06 g/kg and 3.80 g/kg for limestone and crushed concrete, respectively. Column tests utilized synthetic groundwater (50 mg/L Fe(II) from FeCl 2 salt). Wang (2011) investigated competing

PAGE 86

86 cations, such as sodium and manganese, reduced iron removal capacity of limestone only to at (same particle size used in PRBs) in batch experiments carried out over 12 hours with liquid to solid ratios of 1 g/20 mL, DO < 1.0 mg/L, initial pH of 7. It was observed that the final average Fe(II) level only fell above 0.3 mg /L (GCTL) at Na and Mn concentrations of 200 and 100 mg/L, respectively (Wang, 2011); final average Fe(II) levels above 4.2 mg/L were not observed even at these high levels of competing cations. Mn concent rations observed at the site were significantly les s in downgradient wells (average of in upgradient wells of 1.45 and 0.99 mg/L at the limestone and crushed concrete PRB, respectively) over the course of the study. Average sodium levels measured in February 2012 revealed upgradient levels of 7.97 and 13.1 9 mg/L at the limestone and crushed concrete PRB, respectively; concentrations which would not affect Fe(II) removal based on previous data. Natural organic matter can also interfere with iron removal through complexation and chelation (Theis and Singer, 1974). In the same study natural organic matter was used at a concentration of 10 mg C/L premixed with the 50 mg/L stock Fe(II) solution and was shown to affect Fe(II) removal by limestone significantly ( p <0.05); final Fe(II) concentrations observed avera ge approximately 0.35, 0.65, and 0.75 mg/L in the presence of landfill leachate, suwannee river humic acid, and drinking water treatment sludge. Based on routine groundwater monitoring data from Fall 2010, total organic carbon levels found groundwater at t he Klondike landfill from monitoring well MW 3 averaged 9.0 mg/L. In background monitoring wells TOC was not detected at significant levels (lower than the PQL of instrumentation). Analysis of samples collected from well

PAGE 87

87 shallow upgradient monitoring wells during 2012 revealed an average TOC concentration of 28.36 7.22 mg/L. Henderson and Demond (2007) investigated zero valent iron permeable reactive barriers (the most common PRB material) and found that the most common cause of PRB failure reported was i mperfect hydraulic site characterization; decreases in barrier permeability was not likely the cause of PRB failure in many cases, rather decreases in ZVI reactivity. Additionally, pH, Eh, calcium/alkalinity concentration (cause clogging), and residence ti me were found to be major factors in PRB failure (Morrison et al., 2006; Henderson and Demond, 2007). Clogging near the influent of permeable reactive barriers often plays a role in failure (Li et al., 2005; Li et al., 2006; Henderson and Demond, 2007) and the addition of more porous materials, such as pumice to ZVI have been shown to increase porosity in the influent, prolonging PRB lifespan (Moraci and Calabro, 2010). The decrease in PRB efficacy may have been due to soil buildup on the influent of the PR Bs causing Fe and Mn adsorption sites to diminish more quickly than anticipated based on laboratory data. Based on literature hydraulic changes would not likely be observed concurrent with diminishing PRB efficacy (lower reactivity of limestone and crushed concrete). Given field data collected it does not appear that a shift in groundwater flow direction caused an increase in dissolved iron concentrations due to the continuity in pH increase in identical wells observed from the current study to the previou s one (Wang, 2011). The low groundwater gradient does and margin of error in depth to water measurements and survey data does not allow for interpretation of groundwater flow movement change on such a microscale. Data from FDEP records indicates that

PAGE 88

88 groun dwater flow shifted slightly from east to west to east to south west, directing flow away from compliance well DW 4S, which may have accounted for the increasing iron concentration in compliance well DW 4S. This shift however, was not evident in the micros cale of the PRBs. Composition of PRB Materials See table 3 3 for a detailed account of PRB material compositions. Limestone and crushed concrete samples contained measureable amounts of iron. The iron concentration calculated for crushed concrete (3665 m g/kg dry) was greater than reported soil concentrations for Florida entisols (1,200 mg/kg dry) (Ma et al., 1997). Manganese was present in both samples in small amounts; approximately 28.42 and 68.22 mg/kg for limestone and crushed concrete, respectively. The presence of iron and manganese in the reactive materials makes quantification of adsorbed iron from groundwater not possible if rock is dissolved during leaching tests. Limestone samples contained high amounts of calcium and magnesium; approximately 47 .52 and 7.5 % by mass, respectively. Arsenic was detected at trace amounts in both samples. Elements measured by ICP AES constituted approximately 55.3 and 15.9 % of limestone and crushed concrete, by mass, respectively. The potassium and sodium concentrat ions observed, particularly in crushed concrete may indicate the presence of feldspars. Remaining Iron Adsorption Capacity Iron adsorption capacity experiments performed in the field revealed samples taken from all depths throughout each PRB still had the ability to remove ferrous iron. Limestone and crushed concrete control samples revealed average Fe(II) recoveries of 111.9% and 109.1% after one hour reaction time, indicating Fe(II) oxidation was a nonfactor in field experiments. The greatest observed Fe (II) mass removal by limestone

PAGE 89

89 was 0.021 0.0006 mg/g achieved by original limestone, however the greatest Fe(II) concrete experiments the greatest mass removal observed wa s 0.028 mg/g by the original material while the greatest removal efficiency of 95.70% was achieved by the samples collected from 9 to 11 ft bls. This discrepancy between mass and percent removal is due to the larger size particles available from the deeper samples collected. The average concrete particle mass collected from 9 to 11 ft bls was 637.6 grams while the average concrete particle mass available for the original material was 376.8 grams. Wang (2011) found Fe(II) removal >99% by limestone and crush ed concrete samples in laboratory batch experiments over 72 hours from an initial concentration of 50 mg/L. Kinetics experiments revealed that the majority of Fe(II) removal (at least 98%) occurred within the first hour of treatment. Average initial solut ion pH observed in limestone and crushed concrete experiments were 5.93 0.55 and 5.60 0.26, slightly less than the average influent water pH observed throughout the s tudy. Final average solution pH in samples which contained reactive media were 7.39 0.64 and 7.16 1.27 in the limestone and crushed concrete batch experiments, respectively. In laboratory batch experiments Wang (2011) observed final solution pHs of 6.89, 9.28, and 8.75 for one limestone sample, and two crushed concrete samples, respecti vely. These results show that the experiments performed do not have the ability to produce an accurate determination of additional Fe(II) removal capacity. It is suspected that the slight agitation required for particle contact with all liquid resulted in the creation of new particle surface area from removal of dirt build up through slight agitation; this

PAGE 90

90 would explain the disparity between high iron levels observed in monitoring wells and the achievement of low iron levels through batch experiments. Leac hing Tests Initial leaching tests on limestone samples wa shed with nanopure water revealed statistically insignificant ( p <0.05) differences in total iron and manganese content in see Table B 2 However, in nanopure rin se water samples which were preserved after sample washing significantly higher iron and manganese levels were observed in rinse water from the 3. These results lead the author to conduct further leaching experiments on unwashed samples, as not to rinse away iron and manganese precipitates collected on reactive media throughout the life of the PRBs. Leaching tests (0.5 M HCl) on limestone particles excavated from the PRB revealed no significant difference ( p <0.05) in the quantity of iron leached from samples collected from any depth in the PRB. See Tables B 3 and B 4 for data. Triplicate samples performed on the limestone particles revealed standard deviations in leached iron concentrations often exceeding the concentrations thems elves, indicating poor repeatability in the leaching tests. While the initial water rinses showed greater amounts may have been purely coincidental. The greatest obse rved average iron concentration in leaching fluid was 0.51 0.28 mg /L, corresponding to only 5 14 2 mg/ k g Fe(II) removal by reactive material. Wang (2011) calculated a removal capacity of limestone of 4.06 mg/g reactive material based on column tests with synthetic groundwater. Given the expected mass balance removal rates of 0.604 and 0.431 mg/g in the shallow zone of the limestone and crushed

PAGE 91

91 concrete PRBs, respectively, the leaching tests did not extract near the expected mass of iron for the given mass of reactive material. Data from crushed concrete leaching tests revealed relatively low iron concentrations; the greatest calculated mass removal (mg/ k g) to 3 decimal places was 1.543 mg/kg, present in the original material sample Table B 6 and B 7 displa y crushed concrete leaching test data. Manganese was present in leaching fluid; however like iron the difference between samples taken at variable depths did not show a clear trend, though significant differences between depths were noted occasionally, wh en compared to samples below the water table had lower concentrations of Mn leached. It is likely that the porewater. The greatest Mn and 0.006 (8 Mn levels of 0.047 and 0.013 mg/g in the shallow and deep zones of the limestone PRB and 0.045 and 0.007 mg/g in the shallow and deep zones of the crushed concrete PRB. The control batch experiments where limestone particles were subjected to Fe(II) adsorption and then to the 0.5 M HCl leaching procedure did not produce any appreciable dissolved iron in either washed or unwashed samples (average concentration observed for both sets of samples were below detection limit). A red color and iron oxide particles were observed in samples prior to and after HCl addition and 24 hour rotation, indicating the in ability of the acid to dissolve the iron. pH of control samples with no limestone was 5.31 0.09, pH of samples combined with limestone prior to acid leaching were 7.64 0.17.

PAGE 92

92 Given the small quantity of iron obse rved in leaching fluid relative to manganese and the knowledge that influent water contains significantly higher quantities of iron to manganese it seems likely that leaching tests were insufficient in removin g adsorbed iron from particles. Characterizatio n of Precipitates of Reactive Material Digestion of the surface layer of precipitates removed with an ultrasonic bath from limestone and crushed concrete particles revealed mixed results; for crushed concrete samples a clear increase can be seen in element al composition from material never submerged (2 ft bls) to subm erged (9 ft bls) material. See Table 3 4 for elemental composition of the layer on a dry weight basis. Comparing these results with average concentrations of metals in all Florida soils (Ma et al., 1997), Fe appears to be enriched significantly in all precipitates. Manganese concentrations seen in all precipitate samples from the PRBs are significantly greater than those observed by Ma et al. (1997) for all Florida soil types, but not out of the range of values observed. It is probable that the iron and manganese seen in precipitates are the result of soil washed onto PRB material in addition to iron and manganese adsorbed by the PRBs during their lifespan. X Ray diffraction of crushed limesto ne samples (washed and unwashed) identified only calcite, dolomite, quartz, and possibly cristobalite minerals. The hardness of the limestone most likely prohibited substantial penetration of precipitates into the particle structure, so crushing the partic le amounts to dilution of any precipitates which formed on the particles surface, where interaction with groundwater took place. Surface precipitates from original limestone and crushed concrete revealed calcite, quartz and dolomite; additionally original limestone precipitates contained kaolinite. Mica and

PAGE 93

93 hydroxyinterlayered mineral, common soil minerals, were found on samples which were contained in the PRBs. No iron minerals were detected. SEM and EDS characterization of washed and unwashed limestone particles after the control 50 mg/L Fe(II) adsorption batch test revealed discernable iron concentrations at 30,000x magnification and iron at a weight percent at 26.16% on an unwashed particle; see F igure 3 7. Conversely the same analysis on a washed par ticle revealed only 2.02% iron with no discernable concentrations. The confirmed iron concentrations resembled ferric oxide particles confirmed at a similar magnification (UNSW, 2012). These results are significant because they implicate that rinsing the p articles has a substantial effect on the amount of adsorbed iron. No iron was detected in an EDS scan of a limestone particle untreated with Fe(II) solution. SEM examination of the surface of unwashed limestone particles captured magnified images of the re ddish orange surface layer. In F igure 3 6 it can be seen that the surface of the limestone particle collected from 9 feet bls appears more cluttered with a layer of colloids coating the surface than the surface of the particle collected from 2 feet bls (never submerged in the water table). SEM examination of the crushed concrete surfaces did not produce images in which displayed a noticeable variation from original material, material above the water table, and material 9 feet bls. EDS analysis detected i ron in all samples collected from the PRB, in greater quantities in unwashed samples; no difference in iron peak area was observed between samples collected from was dete cted by EDS on any sample.

PAGE 94

94 Wajon et al. (1985) observed siderite and calcium siderite formation on the surface of calcite after exposure to wastewater containing iron sulphite in a carbon dioxide environment. Wang (2011) observed yellow and brown precipita tes formed on limestone particle surfaces and brown and black precipitates formed on crushed concrete surfaces after a column experiment utilizing 50 mg/L Fe(II) influent. The high amount of dirt and mud on the surface of excavated materials made determina tion of the color of actual precipitates difficult. Comparison of PRB Performance to Laboratory Predictions Assuming negligible porosity loss throughout the lifespan of the PRBs approximately 18.3 and 8.63 kg of Fe and 1.43 and 0.401 kg Mn were removed fr om the shallow (3 8 ft bls) and deep (9 14 ft bls) zones of the limestone trench, respectively. Approximately 14.4 and 9.74 kg Fe and 1.35 and 0.213 kg Mn were removed from the shallow and deep zones of the crushed concrete trench. These masses were ca lculated utilizing the average downgradient shallow and deep zone concentrations over the PRB lifespan. In the shallow zone these values equate to 0.604 and 0.431 g/kg remove of Fe for limestone and crushed concrete, respectively. In the deep zone mass rem oval rates of 0.285 and 0.292 g/kg were observed. In column experiments Wang, 2011 calculated a removal capacity of 4.06 and 3.80 g/kg for ferrous iron reacting with limestone and crushed concrete. The lower values observed in the field are most likely due to buildup of dirt and other natural materials on reactive materials, which may block adsorption sites for dissolved iron. Also, substantially larger particles were utilized in the field PRB (diameter varying between 70 and 150 mm vs average diameter of 8 .5 mm used in laboratory columns). Additionally, the reactive barriers were not allowed time enough to

PAGE 95

95 reach full breakthrough (equal influent and effluent concentration), so it is likely that mass removal rates would be greater had more time been allotted for treatment. Summary Limestone and crushed concrete PRBs performed well for the first two years of operation, removing dissolv ed iron from influent water at efficiency rate s of approximately 90% and 95% in the limestone and crushed concrete PRBs, respec tively. In the last year of operation the efficiency rate dropped significantly to an average of 64 and 61% in the limestone and crushed concrete PRBs, respectively. Downgradient iron levels in the last year of operation averaged > 4.2 mg/L the health base d guideline, prompting an investigation into the barriers for a n analysis of the breakdown of PRBs. Given the extreme sensitivity of permeable reactive barrier material to slight agitation in removal ability for ferrous iron (experiments performed in the f ield), extreme care is needed to extract materials from the barrier without disturbing the surface layer of precipitates. Interestingly the crushed concrete PRB remained only slightly less effective as the limestone PRB with much less surface area. SEM an d EDS examination of washed and unwashed limestone particles from Fe(II) adsorption batch experiments showed that washing the surface can cause a substantial loss of adsorbed iron. Digestion of the surface layer of precipitates removed via ultrasonic bath from the limestone and crushed concrete samples collected from the PRB trench from consistently above the water table (2 feet bls) and consistently below (9 feet bls) gave more confounding results. Iron oxides in soil above the water table had no chance of reductive dissolution vs. soil washed into the barrier below the water table. Results from limestone samples showed that precipitate iron concentration may not be a direct result of groundwater iron adsorption, given the greater concentration of iron in

PAGE 96

96 p at the Klondike Landfill most likely accounts for this difference, as it washed into the PRB trench during the period of time the PRBs remained at the site. The diffic ulty in differentiating between the iron adsorbed onto particles solely from groundwater treatment rather than soil contact has not been demonstrate d to be possible in this study; washing the particles may wash off iron precipitates as shown by SEM and EDS and not washing the particles allows differential amounts of soil present on particles to confound results.

PAGE 97

97 Table 3 1. Influent groundwater p arameters from May 2011 through February 2012 Parameter LS UG Shallow LS UG Deep CC UG Shallow CC UG Deep Fe(I I) (mg/L) 30.7 16.2 15.16 10.3 24.8 12.8 23.4 14.2 Fe Total (mg/L) 49.87 15.60 19.76 8.44 32.76 6.28 26.50 5.98 Mn Total (mg/L) 2.08 0.47 1.03 0.44 1.52 0.27 0.65 0.19 Arsenic (mg/L) 0.094 0.053 0.070 0.059 0.078 0.056 0.098 0.070 Calcium (mg/L) 109.4 20.59 64.03 23.72 80.15 26.83 75.89 17.33 Sodium* (mg/L) 1.66 0.70 0.39 0.02 1.54 0.11 0.79 0.11 pH 6.30 0.11 6.27 0.18 6.31 0.13 6.33 0.09 DO (mg/L) 0.89 0.23 0.97 0.17 0.91 0.17 0.86 0.20 ORP (mV) 55 27 58 22 64 28 61 22 *Excludes monitoring during August, 2011 due to equipment breakdown Only a function of monitoring from February 24 th and 25 th 2012 Table 3 2. Effluent groundwater p arameters from May 2011 through F ebruary 2012 Parameter LS D G Shallow LS D G Deep CC D G Shallow CC D G Deep Fe(II) (mg/L) 6.01 4.20 10.6 7.62 0.96 1.33 14.2 13.1 Fe Total (mg/L) 9.77 4.80 13.97 9.98 2.76 1.94 20.34 10.73 Mn Total (mg/L) 0.45 0.22 0.33 0.17 0.18 0.21 0.65 0.29 Arsenic (mg/L) 0.03 0.04 0.05 0.05 0.03 0.03 0.07 0.06 Calcium (mg/L) 61.0 33.0 56.2 21.3 43.4 13.1 69.5 18.5 Sodium (mg/L) 0.38 0.12 0.16 0.13 0.14 0.10 0.46 0.41 pH 6.67 0.12 6.53 0.39 8.90 1.65 6.97 1.10 DO (mg/L) 1.06 0.34 0.91 0.22 0.99 0.33 0.91 0.25 ORP (mV) 60 33 71 23 161 54 96 39 Fe(II) data e xcludes monitoring during August, 2011 due to equipment breakdown Sodium is o nly a function of monitoring from February 24 th and 25 th 2012 ORP data e xcludes monitoring during May 2011 due to equipment breakdown Table 3 3. Elemental composition of PRB materials by mass Sample/ Element Al (mg/kg) As (mg/kg) Ca (%) Fe (mg/kg) K (mg/kg) Mg (mg/kg) Mn (mg/kg) Na (mg/kg) Limestone 229.6 1.28 0 47.52 984.5 173.7 75860 28.42 116 Crushed Concrete 4879 2.041 14.74 3665 803.4 1500 68.22 273.8 All concentration expressed as an average value on a dry weight basis Calcium and magnesium concentrations for limestone and calcium concentrations for crus hed concrete are based on samples diluted 20x after digestion due to instrumentation limitations for limestone.

PAGE 98

98 Table 3 4 Elemental composition of precipitates extracted from PRB materials via ultrasonic bath and filtration, materials c ollected May 1 st a nd 2 nd 2012 Sample/ Element Ca (%) Mg (mg/kg ) Al (mg/kg ) Na (mg/kg ) Fe (mg/kg ) Mn (mg/kg ) As (mg/kg ) 2' bls Limestone 17.71 7,623 12,420 126.6 10,070 160.0 2.62 9' bls Limestone 19.68 8,131 6,948 136.0 7,216 182.6 1.85 2' bls Crushed Concrete 0.697 381.5 9,200 145.0 7,449 103.6 0.82 9' bls Crushed Concrete 1.863 866.2 11,750 187.7 13,630 108.0 1.57 All concentrations expressed on a dry weight basis

PAGE 99

99 Figure 3 1. PRB Area Map s A) General plan view with m onitoring well locations and existing com pliance wells (MW 3 and DW 4S ) B) Layout of the t wo PRBs with detailed nomenclature and dimensions

PAGE 100

100 Figure 3 2. Groundwater Table Elev ation Difference (NAVD) in the PRBs A) Limestone B) C rushed c oncrete (d ue to damage t o well BD3 and lack of data collected on other upgradient wells prior to 2011, data based on well BS2 as an upgradient reference point is displayed )

PAGE 101

101 Figure 3 3. Change in total iron l ev els in PRB monitoring wells ov er time A) L imestone B) Crushed Concrete

PAGE 102

102 Figure 3 4. Change in total manganese l evels in PRB monitoring wells over time A) Limestone B) C rushed concrete

PAGE 103

103 Figure 3 5 Total iron, man ganese and sodium levels in compliance wells for reference A) MW 3 scree ned from 0.11 to 5.11 ft NAVD B) DW 4S screene d from 6.44 to 3.56 ft NAVD

PAGE 104

104 Figure 3 6 pH Cha nge in PRB monitoring wells over time A) Limesto ne B) C rushed concrete

PAGE 105

105 Figure 3 7. SEM Photograph s of limestone particle surface photos courtesy of Michael Kessler and the Major Analytical Instrumentation Center at the University of Florida A) O riginal limestone particle B ) A limestone particle fr ground (nev er submerged in the water table C ) A limestone particle collected ground during trench excavation

PAGE 106

106 Figure 3 7. Continued

PAGE 107

107 Figure 3 8 Particle analysis of an unwashed limestone particle subjected to Fe(II) adsorpt ion at 50 mg/L A ) SEM Photograph photo courtesy of Michael Kessler and the Major Analytical Instrumentation Center at the University of Florida B ) EDS Report.

PAGE 108

108 CHAPTER 4 CONCLUSION Summary Elevated dissolved iron and m anganese concentrations will like ly persist in monitoring wells surrounding Florida landfills. Air sparging and permeable reactive barriers are effective techniques and have been utilized successfully to remove more hazardous contaminants threatening groundwater quality near landfills and other contaminated sites (Marley et al., 1995; Henderson and Demond, 2007; Indraratna et al., 2010). Their employment at sites contaminated with naturally occurring Fe(II) and Mn(IV) is novel. Air sparging downgradient of a closed landfill was able to si gnificantly decrease iron concentrations however in some cases manganese concentrations rose concurrently; the technique was not effective enough to fully aerate water in a surrounding radius to the aerobic zone (> 710 mV) Permeable reactive barrier techn ology, however, was effective in remediation of dissolved iron and manganese, though the crushed concrete barrier raised the pH of groundwater to outside the desired range for aquatic life, 6.5 9 (EPA, 2012) in some wells monitored. I f surface discharges are downgradient of a PRB this may be an issue of concern. Given the high cost of in and ex situ remediation of these contaminants and their natural origin natural attenuation of these pollutants may be the best strategy for site management. Pump and t reat systems to remove iron often use reaeration which occurs naturally in water bodies a given plume of gr oundwater intercepts. If a water body is being threatened by redox sensitive metals potentially an aeration trench could be dug to intercept groundwa ter and provide iron and manganese removal. Ferric iron

PAGE 109

109 oxides could then be periodically removed mechanically. The high cost of aeration blowers, power requirements, and operation and maintenance issues may mitigate the use of aeration for iron remediatio n at closed landfills. Aeration and PRBs are more successfully applied when a known contaminant mas s which can be captured exists In the case of reductively dissolved contaminants their persistence makes finding a permanent remediation solution a continui ng challenge. Conclu ding Remarks This thesis studied two technologies, air sparging and permeable reactive barriers for dissolved iron and manganese removal from groundwater. The following specific conclusions were reached: Air sparging was able to remedia te dissolved iron in a non uniform pattern surrounding injection wells. Decreased iron concentrations correlated to increased manganese concentrations at sparge study sites. Breakthrough of injected air to the subsurface at a short distance from a sparge p oint correlated with decrease zone of influence in the corresponding direction. PRBs were effective in removing both dissolved iron and manganese to below the health based risk guideline for between one and two years of operation. Slight agitation of sampl es removed from trenches was sufficient to create new adsorption sites on reactive material. Soil iron and adsorbed iron on reactive materials were not able to be distinguished.

PAGE 110

110 0.5 M HCl leaching test revealed adsorbed Fe concentrations inconsistent with mass balance calculations based on field data. Future Work The results of this thesis will aid in the understanding of in situ technology for removal of dissolved iron and manganese. The efficacy of air sparging should be studied under controlled laborato ry conditions where control of variables can be maintained. Extended monitoring of air sparge study sites as well as points downgradient of these sites should be conducted to determine the long term potential of using this technology for a set period of ti me. Shorter periods of pulsed air sparging should also be investigated, as the high air flow rates used in this study most likely did not allow for groundwater flow during sparging, lengthening the amount of time downgradient wells could expect to see posi tive changes in dissolved iron and manganese levels. If a zone of oxidation persists an oxidation curtain could be utilized with multiple sparge points spanning a landfill with a mobile blower system to inject air for short periods of time at each point. V adose zone aeration may have potential for success if utilized at the heart of the problem, the anaerobic zone directly underneath the landfill. The effect of air sparging on the size fraction of organic matter may be investigated, given the ambiguous resu lts observed in this study when only examining total organic carbon. Experimentation on another closed landfill in the Florida panhandle revealed average oxygen concentrations of 3.5 % and carbon dioxide levels of 12.5% v/v in soil gas monitoring wells dire ctly beneath disposed waste. If a zone of influence were enacted dir ectly where percolating organic rich liquid (if the landfill is unlined) for the breakdown of the organic matter it contains before the organic matter can reach the

PAGE 111

111 aquifer and reduce and bind with ferric and ferrous iron, respectively, potential for success is possible. Permeable reactive barriers offer a temporary solution to the prob lem of reductive dissolution; the problem will likely continue until the soil is exhausted of all reducabl e Fe/Mn. F uture work could be conducted with the aim of improving longevity of available materials or creating permeable reactive barriers composed of materials which can easily be removed and either replaced or regenerated, collected contaminants disposed of, and then replaced. The ability of iron and manganese oxides, as well as arsenic bound to these minerals, to be reductively dissolved from a landfill site (in the presence of leachate plume and in the presence of a lined landfill) should be studied in advance of landfill placement with variable methods. In some cases, ammonia persists as a contaminant along with the redox sensitive metals. The remedial strategies employed in this research should be looked for their ability to remove this contaminant as well. This data could then be compared to field data to determine the best method for predicting reductive dissolution of contaminants. Work should also be done to determine a method of estimating a zone of natural attenuation, given its low cost, and the n compared to observations in the field. If a zone natural attenuation can be accurately estimated landfills could be placed to allow for this zone to be well beyond any consumptive use wells or water bodies which may be negatively impacted by reductively dissolved metals

PAGE 112

112 APPENDIX A AIR SPARGING: ADDITI ONAL INFORMATION Table A 1. Operational schedule for VZA and AS systems operating in 2011 Study Site Dates Operated Daily Schedule Total Approximate Hours South (Shallow AS) and Central (VZA) Aug. 16 th 27 th 10 pm 10 am 132 North (VZA) Aug. 16 th 22 nd Aug. 29 th Sept. 2 nd 10 pm 10 am 132 South (Shallow AS) and Central (VZA) A ug. 27 th Dec. 22 nd 6 am 12 pm, 2 pm 8 pm, 10 pm 4 am 2124 North (VZA) S ept. 2 nd Dec. 22 nd 6 am 12 pm, 2 pm 8 p m, 10 pm 4 am 1998 Table A 2. Vadose zone aeration and air sparge groundwater data pre 2011 activities (monitoring events from O ct. through Aug. Site/ Parameter Fe(II) (mg/L) Total Mn (mg/L) DO (mg/L) pH ORP (mV) Turbidity (NTUs) South 16.6 6.43 0. 43 0.29 0.82 0.39 6.42 0.16 68 16 9.65 13.0 Central 18.9 8.93 0.43 0.06 0.82 0.41 6.35 0.07 69 10 5.46 4.85 North 15.8 7.09 0.32 0.20 0.78 0.43 6.36 0.12 52 18 5.83 5.27 Total Mn data is only reflective of s and only reflects values in select wells Table A 3. Vadose zone aeration and air sparge groundwater data during 2011 activities (monitoring events from Aug. Site/ Parameter Total Fe (mg/L) Total Mn (mg/L) Dissolved Oxygen (mg/L) pH ORP (mV) Turbidity (NTUs) South 16.15 11.58 0.53 0.35 0.84 0.16 6.44 0.11 51 25 3.51 3.21 Central 26.84 5.47 0.51 0.39 0.92 0.24 6.36 0.09 64 13 10.24 16.8 North 18.18 7.08 0.46 0.64 0.95 0.19 6.31 0.14 35 27 5.76 4.40 Table A 4. Additional groundwater data from post shutdown monitoring on July 19 th and 20 th 2012 Site/ Parameter Fe(II) (mg/L) Temperature (C) pH Conductivity (S/cm) Turbidity (NTUs) South 10.4 15.3 23.7 0.15 6.48 0.08 1370 274 4.03 2.33 Central 16.2 11.8 24.8 0.10 6.34 0.11 923 287 6.19 3.45 North 8.04 4.51 24.8 0.6 6.26 0.47 462 89 6.40 4.46

PAGE 113

113 Figure A 1. South sparge s ite photo courtesy of Saraya Sikora Figure A 2. Central sparge s ite photo courtesy of Saraya Sikora

PAGE 114

114 Figure A 3. North s pa rge s ite photo courtesy of Saraya Sikora

PAGE 115

115 Figure A 4. South site water surface elevation July 19, 2012, values represent ft NAVD.

PAGE 116

116 Figure A 5. Central site water surface elevation map, values represent ft NAVD A) February 14, 2012 B) May 24, 2012

PAGE 117

117 Figure A 6. Central site water surface elevation map July 19, 2012, values represent ft NAVD

PAGE 118

118 Figure A 7. North site water surface elevation map, values represent ft NAVD A) February 14, 2012 B) May 24, 2012

PAGE 119

119 Figure A 8. North site water surface elevation map July 19, 2012, values represent ft NAVD

PAGE 120

120 Figure A 9 Typical injection pressure vs. time at south study s ite

PAGE 121

121 Figure A 10 Typ ical injection p res sure vs. time at central study s ite

PAGE 122

122 Figure A 11 Central site groundwater mounding during 6 hour sparge c ycles A) February 15 2012, B) May 3, 2012

PAGE 123

123 Figure A 12 North site groundwater mounding during 6 hour sparge c ycles A) February 15, 2012 B) May 25 2012

PAGE 124

124 Figure A 13 Central site pressure in soil gas monitoring w el ls during air sparge injection c ycles (all units inches of water) A) February 15 2012 B) M arch 3 0 2012

PAGE 125

125 Figure A 14 North site pressure in soil gas monitoring wells during air sparge inject ion cycles A) February 15, 2012 B) March 2, 2012 C) April 6, 2012

PAGE 126

126 Figure A 15 North site pressure in soil gas monitoring wells during air sparge injec tion cycles A) May 3, 2012 B) May 25, 2012

PAGE 127

127 Figure A 1 6 South site ORP vs t ime Figure A 1 7 Central site ORP vs t ime

PAGE 128

128 Figure A 1 8 North site ORP vs. t ime

PAGE 129

129 Figure A 1 9 South s ite dissolved oxygen vs. t ime

PAGE 130

130 Figure A 20 Cen tral site dissolved oxygen vs. t ime

PAGE 131

131 Figure A 21 N ort h site dissolved oxygen vs. t ime

PAGE 132

132 Figure A 22 pH vs. time, a ffected wells are considered those with average Fe TOT < 4.2 mg/L throughout the study after air sparge initiation

PAGE 133

133 Fi gure A 23 TOC vs. total iron in reference w ell AS2

PAGE 134

134 APPENDIX B PRBS: ADDITIONAL INFORMATI ON Figure B 1 Klondike Landfill map with l ocations of active groundwater monitoring wells PRB area identified

PAGE 135

13 5 Table B 1. Monitoring well d etails Casing Information Screen Information PRB Well ID Depth bls (ft) Well Diameter (inches) Material Screen Length Filter Pack (mesh) Slot Size LS AD1 14 2 PVC 5 20/30 0.01 LS AS2 8 2 PVC 5 20/30 0.01 LS AD3 14 2 PVC 5 20/30 0.01 LS AS4 8 2 PVC 5 20/30 0. 01 LS AD5 14 2 PVC 5 20/30 0.01 LS AS6 8 2 PVC 5 20/30 0.01 LS AD7 14 2 PVC 5 20/30 0.01 LS AS8 8 2 PVC 5 20/30 0.01 LS AD9 14 2 PVC 5 20/30 0.01 LS AS10 8 2 PVC 5 20/30 0.01 LS AD11 14 2 PVC 5 20/30 0.01 LS AT12 12 2 HDPE 5 Drilled Holes None LS AT13 12 2 HDPE 5 Drilled Holes None LS AT14 12 2 HDPE 5 Drilled Holes None CC BD1 14 2 PVC 5 20/30 0.01 CC BS2 8 2 PVC 5 20/30 0.01 CC BD3 14 2 PVC 5 20/30 0.01 CC BS4 8 2 PVC 5 20/30 0.01 CC BD5 14 2 PVC 5 20/30 0.01 CC BS6 8 2 PVC 5 20/30 0.01 CC BD7 14 2 PVC 5 20/30 0.01 CC BS8 8 2 PVC 5 20/30 0.01 CC BD9 14 2 PVC 5 20/30 0.01 CC BS10 8 2 PVC 5 20/30 0.01 CC BD11 14 2 PVC 5 20/30 0.01 CC BT12 6 2 HDPE 5 Drilled Holes None CC BT13 9 2 HDPE 5 Drilled Holes None CC BT14 11 2 HDPE 5 Drilled Ho les None

PAGE 136

136 Table B 2. Results of initial laboratory leaching tests on nanopure rinsed limestone, (3 M HCl L/S ratio of 1.66:1 ) Sample Fe (mg/L) Fe (mg/ k g) Mn (mg/L) Mn (mg/ k g) Original Limestone 0.32 0.05 0. 5 3 0.42 0.03 0. 69 0.52 0.23 0. 85 1.00 0.38 1.67 0.58 0.16 0. 97 1.05 0.09 1.74 Table B 3. Iron and manganese concentrations in nanopure rinse water used to clean initial samples ( L/S ratio of 1.66 :1) Sample Fe (mg/L) Fe (mg/ k g) Mn (mg/L) Mn (mg/ k g) Original Limestone 0.0 9 0.03 0.53 0.02 0.03 0.69 0.40 0.02 0.85 0.05 0.00 1.67 0.63 0.02 0.97 0.02 0.00 1.74 Table B 4 Results of laboratory leaching tests on limestone fluid concentration (0.5 M HCl leaching fluid) Sample Fe (mg/L) Mg (mg/L) Mn (mg/L) Na (mg/L) Original Limestone 0.21 0.12 488 28.8 0.46 0.02 2.35 0.18 0.31 0.24 481 37.0 0.78 0.10 2.23 0.10 0.24 0.18 455 55.1 0.55 0.05 2.26 0.27 0.50 0.42 463 36. 5 0.79 0.06 1.93 0.20 0.51 0.28 432 39. 1 0.70 0.08 2.05 0.21 0.23 0.11 495 37.4 0.78 0.03 2.22 0.26 0.38 0.28 420 29.4 0.65 0.06 2.37 0.17 0.15 0.02 466 42.1 0.74 0.06 2.21 0.16 Table B 5 Results of laboratory leaching tests on limestone, calculated average mass removal Sample Fe (mg/ k g ) Mg (mg/ k g ) Mn (mg/ k g ) Na (mg/ k g ) Original Limestone 2.074 4 887 4.624 23.47 3.11 2 4812 7.777 22.28 2.3 69 4552 5.534 22.64 4.99 1 4 634 7.884 19.27 5.14 2 4 322 7.049 20.48 2.3 15 4 950 7.754 22.17 3. 796 4 195 6.517 23.75 1.54 3 4 663 7.356 22.15

PAGE 137

137 Table B 6 Results of laboratory leaching tests on concrete (0.5 M HCl leaching fluid) Sample Fe (mg/L ) Mg (mg/L ) Mn (mg/L ) Na (mg/L ) Original Crushed Concrete 0.20 0.01 77.03 0.58 0.55 0.01 16.35 0.04 CC 0.15 0.01 6.21 0.06 0.00 0.00 6.67 0.11 7 0.08 0.01 106.8 1.44 0.0 0 0.00 4.99 0.28 8 0.17 0.01 98.25 2.26 4.51 0.07 8.64 0.50 0.17 0.02 1.30 0.01 0.00 0.00 13.91 0.25 0.19 0.01 87.44 0.09 2.89 0.00 13.21 0.07 Table B 7 Results of laboratory leaching tests on concr ete, calculated average mass removal Sample Fe (mg/ k g ) Mg (mg/ k g ) Mn (mg/ k g ) Na (mg/ k g ) Original Crushed Concrete 0. 2132 83.07 0.5873 17.63 0 .1573 6.693 0.0042 7.193 7 0. 0898 115.2 0.0001 5.381 8 0. 1814 105.9 4.863 9.315 0. 1854 1.406 0.00 03 15.00 0. 2033 94.29 3.121 14.25

PAGE 138

138 LIST OF REFERENCES Aivalioti MV, Gidarakos EL. In well air sparging efficiency in remediation the aquifer of a petroleum refinery site. J Environ Eng 2008;7:71 82. Aziz HA, Othman N, Yusugg MS, Ba sri DRH, Ashaari FAH, Adlan MN, Othman F, Johari M. Removal of copper from water using limestone filtration technique determination of mechanism of removal. Environ Int 2001;26:395 399. Aziz HA, Smith PG. The influence of pH and coarse media on manganese p recipitation from water. Water Resour 1992;26:853 855. Aziz HA, Smith PG. Removal of manganese from water using crushed dolomite filtration technique. Water Resour 1996;2:489 492. Aziz HA, Adlan MN, Ariffin KS. Heavy metals (Cd, Pb, Zn, Ni, Cu, and Cr(III) ) removal from water in Malaysia: Post treatment by high quality limestone. Bioresource Technol 2008;99(6):1578 1583. Baker RS, Hayes ME, Frisbie SH. Evidence of preferential vapor flow during in situ air sparging. In: Hinchee RE, Miller RN, Johnson PC, ed itors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p.63 73. Bass DH, Hastings NA, Brown RA. Performance of air sparging systems: a review of case studies. J Hazard Mater 2000;72(2 3):101 119. Ben ner SG, Blowers DW, Gould WD, Herbert RB, Ptacek CJ. Geochemistry of a permeable reactive barrier for metals and acid mine drainage. Environ Sci Technol 1999;33(16):2793 2799. Bjerg PL, Rugge K, Pederson JK, Christensen TH. Distribution of redox sensitive groundwater quality parameters downgradient of a landfill (Grindsted, Denmark). Environ Sci Technol 1995;29:1387 1394. Boersma PM, Diontek KR, Newman PAB. Sparging effectiveness for groundwater restoration, In: Hinchee RE, Miller RN, Johnson PC, editors. I n situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 39 46. Bouwer H, Rice RC. A slug test for determining hydraulic conductivity of unconfined aquifers with completely or partially penetrating wells. W ater Resour Res 1976;13(3):423 428. Calace N, Liberatori A, Petronio BM, Pietroletti M. Characteristics of different molecular weight fractions of organic matter in landfill leachate and their role in soil sorption of heavy metals. Environ Pollut 2001:113( 3):331 339.

PAGE 139

139 Cantrell KJ, Kaplan DI, Wietsma TW. Zero valent iron for the in situ remediation of selected metals in groundwater. J Hazard Mater 1995;42(2):201 212. Carter SR, Clark JE. Oxygen enhanced in situ bioremediation in a sand and gravel aquifer. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 551 558. Champ DR, Gulens J, Jackson RE. Oxidation reduction sequences in ground water flow systems. Ca n J Earth Sci 1979;16:12 23. Chao K, Ong SK. Air Sparging: effects of VOCs and Soil Properties on VOC volatilization. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battell e Press; 1995. p. 441 446. Christenbury JH, Plowman FT, Wagenet L, Lemley A. Iron and Manganese. Clemson, South Carolina: South Carolina Cooperative Extension Serv ice, Clemson University; 1990. Christensen TH, Kjeldsen P, Albrechtsen H, Heron G, Nielsen PH Bjerg PL, et al. Attenuation of landfill leachate pollutants in aquifers. Environ Sci Technol 1994;24(2):119 202. Christensen TH, Bjerg PL, Banwart SA, Jakobsen R, Heron G, Albrechtsen H. Characterization of redox conditions in groungwater contaminant pl umes. J Contam Hydrol 2000;45(3 4):165 241. Christensen TH, Kjeldsen P, Bjerg PL, Jensen DL, Christensen JB, Baun A, et al. Biogeochemistry of landfill leachate plumes. Appl Geochem 2001;16:659 718. De Lemos JL, Bostick BC, Renshaw CR, Sturup S, Feng X. La ndfill stimulated iron reduction and arsenic release at the Coakley superfund site (NH). Environ Sc. Technol 2006;40:67 73. Di Palma L, Mecozzi R. Batch and column tests of metal mobilization in soil impacted by landfill leachate. Waste Manage 2010;30:1594 1599. Ehrig HJ. Quality and quantitiy of sanitary landfill leachate. Waste Manage 1983;1:53 68. Ellis D, Bouchard C, Lantagne G. Removal of iron and manganese from groundwater by oxidation and microfiltration. Desal 2000;130:255 264. FDEP. ndwater quality monitoring program: Background hydrogeochemistry. Tallahassee, Florida: Florida Department of Environmental Protection; 1992.

PAGE 140

140 FDEP. FS 2200 Groundwater Sampling. DEP SOP 001/01. Tallahasse, Florida: Florida Department of Environmental Prote ction; 2008. Available at: http://publicfiles.dep.state.fl.us/dear/sas/sopdoc/2008sops/fs2200.pdf Fields K, Gibbs J, Condit W, Leeson A, Wickramanayake G. Air Sparging: A Project Managers Guide. Columbus Ohio: Battelle Press; 2002. Gadagbui B, Roberts S. Development of Cleanup Target Levels (CTLs) for Chapter 62 777, F.A.C. Center for Environmental and Human To xicology. University of Florida; 2002. Geosyntec Consultants. Feasibility Evaluation Klondike Landfill. Prepared for Escambia County Department of Solid Waste Management. Tampa, Florida; 2005a. Geosyntec Consultants. Iron Assessment Report. Prepared for Escambia County Department of Solid Waste Management. Tampa, Fl orida; 2005b. Geosyntec Consults. Water quality monitoring plan. Prepared for Escambia County Department of Solid Waste Management. Tampa Florida; 2006. Ghosh A., Mukibi M, Saez AE, Ela WP. Leaching of arsenic from granular ferric hydroxide residues under mature landfill conditions. Environ Sci Technol 2006;40(19):6060 6075. Gibert O, de Pablo J, Cortina JL, Ayora C. In situ removal of arsenic from groundwater by using permeable reactive barriers of organic matter/limestone/zero valent iron mixtures. Envi ron Geochem Health 2010;32:373 378. Grossman EL, Cifuentes LA, Cozzarelli IM. Anaerobic methane oxidation in a landfill leachate plume. Environ Sci Technol 2002;36:2436 2442. Hariston JE, Stribling L. Typical contaminants and problems Iron and Manganese Professional paper ANR 790 2.3.3. Alabama A & M and Auburn Universities ; 1995. Available at: http://www.aces.edu/pubs/docs/A/ANR 0790/WQ2.3.3.pdf Harvel, GF. PVC Pipe Schedule 40. Easton, Pennsylvania; 2012. Available at: http://www.harvel.com/technical support center/product specifications/pvc pipe schedule 40 Henderson AD, Demond A H. Long term performance of zero valent iron permeable reactive barriers: a critical review. Environ Eng Sci 2007;24(4):401 423. Herman GM. Iron and Manganese in Household Water Publication number: HE 394 North Carolina Cooperative Extension service ; 199 6. Available at: http://www.bae.ncsu.edu/programs/extension/publicat/wqwm/he394.html

PAGE 141

141 Heron G, Crouzet C, Bourg ACM, Christensen TH. Speciation of Fe(II) and Fe(III) in Con taminated aquifer sediments using chemical extraction techniques. Environ Sci Technol 1994;28, 1698 1705 Heron G, Christensen TH. Impact of Sediment bound Iron on Redox Buffering in a Landfill Leachate Polluted A quifer. Environ Sci Technol 1995 ;29:187 192 Hudak PF. Principles of Hydrogeology. B oca Raton, Fl: CRC Press; 2000. Johnson PC, Johnson RL, Neaville C, Hansen EE, Stearns SM, et al. Do conventional monitoring practices indicate in situ air sparging performance? In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 1 20. Johnson PC. Asessment of the contributions of volatilization and biodegradation to in situ air sparging performance. Environ Sci Tech nol 1998;32:276 281. Johnson PC, Johnson RL, Bruce CL, Leeson A. Advances in in situ air sparging/biosparging. Biorem J 2001;5(4):251 266. Johnson RL, Johnson PC, McWhorter BB, Hinchee RE, Goodman I. An overview of in situ air sparging. Ground Water Monit Rev 1993;13(4):127 135. Johnston CD, Rayner JL, Briegel RD. Effectiveness of in situ air sparging for removing NAPL gasoline from a sandy aquifer near Perth, Western Australia. J Contam Hydrol 2002;59:87 111. Joshi A, Chaudhuri M. Removal of Arsenic from g round water by iron oxide coated sand. J Environ Eng 1996;122:769 771. Keimowitz AR, Simpson HJ, Stute M, Datta S, Chillrud SN, Ross J, et al. Naturally occurring arsenic mobilization at a landfill in Maine and implications for remediation. Appl. Geochem 2 005;20:1985 2002. Kjeldsen P, Christophersen M. Composition of leachate from old landfills in Denmark. Waste Manage Res 2001;19(3):249 256. Lackovic JA, Nikolaidis NP, Dobbs GM. Redox sensitive mobility of arsenic in proximity to a municipal landfill. In: Nikolaidis N, Erkey C, Smets BF, editors. Hazardous and Industrial Wastes Lancaster, Pennsylvania: Technomic Publishing; 1999. p. 451 459. Leeson, A., Johnson, P.C., Johnson, R.L., Vogel, C.M., Hinchee, R.E., Marley, M., Peargin, T., Bruce, C.L., Amerson, I.L., Coonfare, C.T., Gillespie, R.D., and D.B. McWhorter. 2002. Air Sparging Design Paradigm. Battelle Press Columbus, Ohio Lepp H. Geochemistry of iron. Dowden, Stroudsburg: Hutchinson Ross Publishing; 1975.

PAGE 142

142 Li L, Benson CH, Lawson EM. Impact of mineral fouling on hydraulic behavior of permeable reactive barriers. Ground Water 2005;43(4):582 596. Li L, Benson CH, Lawson EM. Modeling porosity reductions caused by mineral fouling in continuous wall permeable react ive barriers. J Contam Hydrol 200 6;83(1 2): 89 121. Lindberg RD, Runnels DD. Ground water redox reactions: an analysis of equilibrium state applied to Eh measurements and geochemical modeling. Science 1984;225:925 927. Lovley DR. Dissimilatory Fe(III) and Mn(IV) reduction. Microbiol Rev 1991;55(2):2 59 287. Ludvigsen L, Albrechtsen HJ, Heron G, Bjerg PL, Christensen TH. Anaerobic microbial redox processes in a landfill leachate contaminated aquifer (Grindsted, Denmark). J Contam Hydrol 1998;33(3 4):273 291. Lundegard PD, Anderson G. Multiphase numeric al simulation of air sparging performance. Ground Water 1993;34(3):451 460. Lundegard PD. Air Sparging: Much ado about mounding. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Oh io: Battelle Press; 1995. p. 21 30. Lundegard PD, LaBrecque D. Air sparging in a sandy aquifer (Florence, Oregon, U.S.A.): Actual and apparent radius of influence. J Contam Hydrol 1995;19:1 27 Lyngkilde J, Christensen TH. Redox zones of a landfill leachate plume (Vejen, Denmark). J Contam Hydrol 1992;10:273 289. Ma LQ, Tan F, Harris WG. Concentrations and distributions of eleven metals in Florida soils. J Environ Qual 1997;26(3):769 775. Machmeier RE. Iron in Drinking Water, AG FO 1318. Minnesota Cooperativ e Extension Service. Univer sity of Minnesota, St. Paul, MN; 1990. Marley MC, Walsh MT, Nangeroni PE. A case study on the application of air sparging as a complementary technology to vapor extraction at a gasoline spill site in Rhode Island. Hydrocarbon con taminated soils and groundwater conference. Amherst, Massachusetts; 1990. Marley MC. Air sparging in conjunction with vapor extraction for source removal at VOC spill sites. Fifth National Outdoor Action conference on aquifer restoration, groundwater moni toring, and geophysical methods National Water Well Association; Las Vegas, Nevada; 1991.

PAGE 143

143 Marley MC, Bruell CJ, Hopkins HH. Air Sparging Technology: A Practice Update. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, biovent ing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 31 37. Martinson M, Linck J, Manz C, Petrofske T. Optimized air sparging coupled with soil vapor extraction to remediate groundwater. In: Hinchee RE, Miller RN, Johnson PC, editors. In si tu aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 203 214. McMahon PB, Dennehy KF, Sandstrom MW. Hydraulic and geochemical performance of a permeable reactive barrier containing zero valent iron, Denver Federal Center. Ground Water 1999;37(3):396 404. Mettler S, Abdelmoula M, Hoehn E, Schoenenberger R, Weidler P, von Gunten U. Characterization of iron and manganese precipitates from an in situ groundwater treatment plant. Ground Water 39(6):921 930. Mettl er S, Wolthers M, Charlet L, von Gunten U. Sorption and catalytic oxidation of Fe(II) at the surface of calcite. J Geochem Soc Meteorit Soc 2009;73:1826 1840. Moraci N, Calabro PS. Heavy metals removal and hydraulic performance in zero valent iron/pumice p ermeable reactive barriers. J Environ Mangage 2010;91(11):2336 2341. Morrison SJ, Mushovic PS, Niesen PL. Early breakthrough of molybdenum and uranium in a permeable reactive barrier. Environ Sci Technol 2006;40(6):2018 2024. Murray WA, Lunardini RC, Ullo FJ, Davidson ME. Site 5 air sparging pilot test, Naval Air Station Cecil Field, Jacksonville, Florida. J Haz Mat 2000;72:121 145. Nanny MA, Ratasuk N. Characterization and comparison of hydrophobic neutral and hydrophobic acid dissolved organic carbon isol ated from three municipal landfill leachates. Water Res 2002;36:1572. Nicholson RV, Cherry JA, Reardon EJ. Migration of contaminants in groundwater at a landfill: A case study. J Hydrol 1983;63:131 176. Parisio S, Keimowitz AR, Simpson HJ, Lent A, Blackman V. Arsenic rich iron floc deposits in seeps downgradient of solid waste landfills. Soil Sediment Contam 2006;15(5):443 453. Peters GM, Mahar WA, Barford JP, Gomes VG. Selenium associations in estuarian sediments: Redox effects. Water Air Soil Pollut 1997; 99(1 4):275 282. Plowman FT. Iron and M anganese, Water Quality Fact Sheet 5. New Hampshire Cooperative Extension Service. Universi ty of New Hampshire: Durham, New Hampshire; 1989.

PAGE 144

144 Puls RW, Paul CJ, Powell, RM. The application of in situ permeable reactive (zero valent iron) barrier technology for the remediation of chromate contaminated groundwater: a field test. Applied Geochem 1999;14(8):989 1000. Rao NS. Iron content in groundwaters of Visakhapatnam environs, Andhra Pradesh, India. Environ Monit Assess 2008;139(1 3):437 447. Reddy KR, Adams JA. Effect of groundwater flow on remediation of dissolved phase VOC contamination using air sparging. J Haz Mat 2000;72:147 165. Rhodes DK, Burke GK, Smith N, Clark D. (1995). In situ diesel fuel bioremediation: a ca se history. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 441 446. Schrauf TW, Pennington LH. Design and application of an alternative groundwater sparging technology. In: Hinchee RE, Miller RN, Johnson PC, editors. In situ aeration: air sparging, bioventing, and related processes. Columbus, Ohio: Battelle Press; 1995. p. 145 158. Schwertmann U, Taylor RM. Iron oxides. In: Dianuer RC, editor. Minera ls in Soil Environments. Soil Science Society of America: Madison, Wisconsin; 1977. p. 145 180. Shacklette HT, Boerngen JG. Professional Paper 1270. Element Concentrations in soils and other surficial materials of the conterminous United States. United Sta tes Geological Survey: Washington DC; 1984. Available at: http://pubs.usgs.gov/pp/1270/pdf/PP1270_508.pdf Shaw BH, Peterson JO. Improving your drinking water quality. G3378. Wisconsi n Cooperati ve Extension Service Unive rsity of Wisconsin: Madison, WI; 1990. Theis TL, Singer PC. Complexation of Iron(II) by organic matter and it's effect on Iron(II) oxygenation. Environ Sci Technol 1974; 8(6):569 573. Thirunavukkarasu OS, Viraraghavan T, Subraman ian KS. Arsenic removal from drinking water using iron oxide coated sand. Water Air Soil Pollut 2003;142(1 4):95 111. Tyson A, Harrison K. Water quality for private water systems. Georgi a Cooperative Extension Service at t he Un iversity of Georgia: Athens, Georgia; 1990. UNSW Scanning electron microscopy. University of New South Wales: New South Wales, Australia; 2012. Available at: http://www.membrane.unsw.edu.au/autopsy/sem.asp USEPA. Air Spa rging. United States Environmental Protection Agency: Washington DC; 1994. Available at: http://www.epa.gov/oust/pubs/tum_ch7.pdf

PAGE 145

145 USEPA A citizens guide to permeable reactive barriers. United States Environmental Protection Agency: Washington DC; 2001 a Available at: http://www.epa.gov/ superfund/community/pdfs/suppmaterials/treatme nttech/permeable_reactive_barri ers.pdf USEPA. Cost analysis for selected groundwater cleanup projects: Pump and treat systems and permeable reactive barriers. United States Environmental Protection Agency: Washington DC; 2001b. Available at: http://www.epa.gov/ tio/download/remed/542r00013.pdf USEPA National Reco mmended Water Quality Criteria. United States Environmental Protection Agency: Washington DC; 2012. http://water.epa.gov/scitech/swguidance/standards/criteria/current/index.cfm#C Wajon JE, Ho GE, Murphy PJ. Rate of precipitation of ferrous iron and formation of mixed iron calcium carbo nates by naturally occurring carbonate materials. Water Res 1985;19:831 837. Wang, Y. PhD Dissertation. The use of calcium carbonate based materials for removing dissolved iron from groundwater at l andfills. University of Florida: Gainesville, Florida; 201 1. Wang Y, Sikora S, Kim H, Dubey B, Townsend T. Mobilization of iron and arsenic from soil by construction and demolition debris landfill leachate. Waste Manage 2011;32(5):925 932. Yang X, Beckmann D, Fiorenza S, Niedermeier C. Field study of pulsed air sparging for remediation of petroleum hydrocarbon contaminated soil and groundwater. Environ Sci Technol 2005;39(18): 7279 7286

PAGE 146

146 BIOGRAPHICAL SKETCH Saraya Sikora, born in Rockledge, Florida is the older of two children. She grew up in Merritt Island, F lorida and graduated from Merritt Island High School in 2005. She Florida in December 2009 as well as an Engineering Intern (EI) certification. Upon graduation she immediate ly commenced graduate school January of 2010 in the field of solid and hazardous waste under Dr. Timothy Townsend, also at University of Florida. After graduation she plans to work in the field of environmental engineering and work towards a Professiona l Engineering license.