UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Aqueous and mineralogical analysis of arsenic in the reduced, circumneutral groundwaters and sediments… Bolton, Mark 2004

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_2004-0376.pdf [ 22.34MB ]
Metadata
JSON: 831-1.0052461.json
JSON-LD: 831-1.0052461-ld.json
RDF/XML (Pretty): 831-1.0052461-rdf.xml
RDF/JSON: 831-1.0052461-rdf.json
Turtle: 831-1.0052461-turtle.txt
N-Triples: 831-1.0052461-rdf-ntriples.txt
Original Record: 831-1.0052461-source.json
Full Text
831-1.0052461-fulltext.txt
Citation
831-1.0052461.ris

Full Text

A Q U E O U S A N D M I N E R A L O G I C A L A N A L Y S I S O F A R S E N I C IN T H E R E D U C E D , C I R C U M N E U T R A L G R O U N D W A T E R S A N D S E D I M E N T S O F T H E F R A S E R R I V E R D E L T A , BRITISH C O L U M B I A by Mark Bolton B.Sc. Simon Fraser University, 1999 A THESIS S U B M I T T E D IN P A R T I A L F U L F I L L M E N T O F T H E R E Q U I R E M E N T S F O R T H E D E G R E E O F M A S T E R O F S C I E N C E . . . in the Department of Earth and Ocean Sciences We accept this thesis as conforming to the required standard U N I V E R S I T Y O F BRITISH C O L U M B I A September 2004 © Mark Bolton, 2004 FACULTY OF G R A D U A T E STUDIES THE UNIVERSITY OF BRITISH C O L U M B I A Library Authorization In presenting this thesis in partial fulfillment of the requirements for an advanced degree at the University of British Columbia, I agree that the Library shall make it freely available for reference and study. I further agree that permission for extensive copying of this thesis for scholarly purposes may be granted by the head of my department or by his or her representatives. It is understood that copying or publication of this thesis for financial gain shall not be allowed without my written permission. Mark Bolton 10/09/2004 Name of Author (please print) Date (dd/mm/yyyy) Title of Thesis: Aqueous and mineralogical analysis of arsenic in the reduced, circumneutral groundwaters and sediments of the Fraser River delta, British Columbia Degree: M.Sc. Year: 2004 Department of Earth and Ocean Sciences The University of British Columbia Vancouver, BC Canada grad. ubc.ca/forms/?f orm I D=TH S page 1 of 1 last updated: 14-Sep-04 A B S T R A C T A n aqueous and mineralogical investigation of the groundwaters and sediments of the Fraser River delta, B C is presented. Groundwater and sediment samples were collected from two sites. The D N D site is located in an upland area, approximately 3 km upgradient from the north arm of the Fraser River. The Kidd2 site is located adjacent to the north arm of the Fraser River. A saline water wedge extends inland from the bed of the Fraser River to the subsurface beneath the Kidd2 site. Concentrations of dissolved arsenic (As) are relatively low at the D N D site, while concentrations of approximately 30 u.g/L dissolved As are present within the saline water wedge, at the Kidd2 site. The aqueous geochemical data suggest that dissolved As is generated by the reduction of hydrous ferric oxide (HFO) minerals. The data also suggest that the elevated concentrations of dissolved As are correlated to the mass loading that occurs as dense saline water infiltrates from the base of the Fraser River into the adjacent aquifer. Sediment samples from both sites were analysed using a sequential extraction procedure (SEP) that was designed to target specific pools of solid-phase As. Results of the SEP analyses indicate that solid-phase As is present in the sediments at trace concentrations. However, processes such as sorption and the precipitation of arsenical sulphide minerals do not appear to be sufficient to mitigate the relatively elevated concentrations of dissolved As at the Kidd2 site. If concentrations of dissolved As at the Kidd2 site were to remain constant at 30 ug/L, the time required to flush the solid-phase As that is associated with H F O minerals from the sediments is estimated to be over 24,750 years. It would take significantly longer for the HFO-associated solid-phase As to be flushed from the upland areas of the delta, where mass loading of reactants to the groundwater is significantly lower. The results of this investigation indicate that a disturbance to the geochemistry of the Fraser River delta has the potential to generate significant concentrations of dissolved As, possibly as high as those encountered in other deltaic environments such as Bangladesh. ii TABLE OF CONTENTS Abstract ii List of Tables vi List of Figures vii Acknowledgements ix 1.0 INTRODUCTION 1 1.1 P U R P O S E A N D O B J E C T I V E S 3 2.0 CASE STUDY: THE BENGAL BASIN 5 2.1 D E P O S I T I O N A L H I S T O R Y 8 2 . 2 E A R L Y S E D I M E N T D I A G E N E S I S 9 2.2 .1 Initial oxic conditions 9 2 .2 .1 .1 Arsenic sorption 11 2 . 2 . 2 Reducing conditions 1 4 2 .2 .2 .1 Iron cycling 1 6 2 . 2 . 2 . 2 Carbon cycling 1 8 2 . 2 . 2 . 3 Sulphur cycling 1 9 2 . 2 . 2 . 4 Arsenic cycling 2 1 2 . 2 . 2 . 4 . 1 Arsenic release 2 1 2 . 2 . 2 . 4 . 2 Arsenic mobility 2 4 2 . 2 . 2 . 4 . 3 Arsenic sequestration 2 4 2 .3 G R O U N D W A T E R H Y D R O L O G Y O F T H E B E N G A L B A S I N 2 6 2 . 4 I D E N T I F I C A T I O N O F A R S E N I C C O N T A M I N A T I O N I N T H E B E N G A L B A S I N 2 8 2 .5 P R O P O S E D M E C H A N I S M S F O R A S C O N T A M I N A T I O N I N T H E B E N G A L B A S I N 3 0 2 .5 .1 Oxidation of pyrite 3 0 2 . 5 . 2 Competitive anion exchange: phosphate and arsenic 3 1 2 .5 .3 Reductive dissolution of H F O minerals 3 2 2 . 5 . 4 Competitive anion exchange: bicarbonate and arsenic 3 4 3.0 FRASER RIVER DELTA 35 iii 3.1 B A S E M E N T G E O L O G Y 35 3.2 D E P O S I T I O N A L H I S T O R Y O F T H E F R A S E R R I V E R D E L T A 36 3.3 M I N E R A L O G Y A N D A Q U E O U S G E O C H E M I S T R Y 38 3.4 G R O U N D W A T E R H Y D R O L O G Y 44 4.0 FIELD SITE DESCRIPTION 46 4.1 D E P A R T M E N T O F N A T I O N A L D E F E N C E ( D N D ) SITE 46 4.2 K I D D 2 F I E L D S I T E 47 5.0 METHODOLOGY 49 5.1 S A M P L E C O L L E C T I O N 49 5.1.1 Water samples 49 5.1.2 Sediment samples 51 5.1.2.1 D N D site 51 5.1.2.2 Kidd2 site 52 5.2 S A M P L E A N A L Y S E S 53 5.2.1 Water samples 53 5.2.1.1 Field analyses 53 5.2.1.2 Laboratory analyses 53 5.2.1.2.1 D N D site 53 5.2.1.2.2 Kidd2site 54 5.2.2 Sediment samples 54 5.2.2.1 Polished thin sections 54 5.2.2.2 Sequential extractions 55 5.2.2.2.1 Sequential extraction procedure 55 5.2.2.2.2 Analyses of extractant solutions 57 5.2.2.3 Analysis of total solid phase arsenic in sediments 57 5.2.2.4 Analysis of total sulphur in sediments 57 6.0 RESULTS AND DISCUSSION 62 6.1 G R O U N D W A T E R A N A L Y S E S 62 6.1.1 D N D site 62 iv 6.1.1.1 Interpretati on 71 6.1.1.2 Arsenic field kit results 86 6.1.1.3 Arsenic speciation results 87 6.1.2 Kidd2s i te 89 6.1.2.1 Interpretation 94 6.2 S E D I M E N T A N A L Y S E S 105 6.2.1 Mineralogical analyses 106 6.2.2 Solid phase analyses 106 6.2.2.1 Interpretation 112 7.0 S U M M A R Y A N D C O N C L U S I O N S 119 7.1 C O N C E P T U A L M O D E L 120 7.2 R E C O M M E N D A T I O N S 123 7.3 I M P L I C A T I O N S 125 8.0 R E F E R E N C E S 127 Appendix A : Borehole logs for the Kidd2 site 135 Appendix B: Selected photographs 141 Appendix C: Sequential extractions ;. 149 Appendix D: Sample calculations for alkalinity titration analyses 157 Appendix E : Certificate for certified reference material T M D A - 5 4 . 3 161 Appendix F: Sample calculations 164 v L I S T O F T A B L E S Table 5.1: Summary of publications for Sequential Extraction Procedures for As-bearing sediments 58 Table 5.2: Sequential Extraction Procedures for As-bearing sediments for investigations in Bangladesh 60 Table 5.3: Sequential Extraction Procedures for As- and Fe-bearing sediment cores from D N D and Kidd2 sites 61 Table 6.1: Field parameters, concentrations of anions and D O C in profiles PI and P2, D N D site 63 Table 6.2: Concentrations of cations in profiles PI and P2, D N D site 66 Table 6.3: Calculated SI values for relevant minerals in profile PI , D N D site 78 Table 6.4: Calculated SI values for relevant minerals in profile P2, D N D site 86 Table 6.5: Comparison of concentrations of cations in samples with and without As speciation 88 Table 6.6: Field parameters, concentrations of anions and D O C in profile P3, Kidd2 site 90 Table 6.7: Concentrations of cations in profile P3, Kidd2 site 92 Table 6.8: Calculated SI values for relevant minerals in profile P3, Kidd2 site 105 Table 6.9: Concentrations of solid-phase As, Fe and S in sediment cores C I , D N D site andC3, Kidd2 site I l l vi L I S T O F F I G U R E S Figure 2.1: Map of Bangladesh 5 Figure 2.2: Typical cross-section throughout Bangladesh 6 Figure 2.3: Typical vertical profile in Bangladesh 7 Figure 3.1: Location map of the Fraser River delta 35 Figure 3.2: Evolution of the Fraser River delta 37 Figure 3.3: Surficial geology of the present-day Fraser River delta 39 Figure 3.4: Location map of the Municipality of Mission 43 Figure 3.5: Results of water well search for Fraser River delta 44 Figure 4.1: Locations of the D N D and Kidd2 field sites 46 Figure 4.2: Aerial photograph of the D N D site 47 Figure 4.3: Aerial photograph of the Kidd2 site 48 Figure 6.1: Concentrations of cations with depth in profile PI , D N D site 73 Figure 6.2: Concentrations of D O C and H C 0 3 " with depth in profile PI , D N D site 74 Figure 6.3: Concentrations of As and NrLi + with'depth in profile PI , D N D site 75 Figure 6.4: The relationship of As to other species in profile PI , D N D site 76 Figure 6.5: The relationship of HCO3" to other species in profile PI , D N D site 77 Figure 6.6: Concentrations of cations with depth in profile P2, D N D site 80 Figure 6.7: Concentrations of D O C and H C 0 3 " with depth in profile P2, D N D site 81 Figure 6.8: Concentrations of As and N F L * with depth in profile P2, D N D site 82 Figure 6.9: The relationship of As to other species in profile P2, D N D site 83 Figure 6.10: The relationship of HCO3" to other species in profile P2, D N D site 84 vii Figure 6.11: Concentrations of cations with depth in profile P3, Kidd2 site 95 Figure 6.12: Concentrations of D O C and H C 0 3 " with depth in profile P3, Kidd2 site ... 96 Figure 6.13: Concentrations of As, N H 4 + and PO4 3" with depth in profile P3, Kidd2 site 97 Figure 6.14: Concentrations of As and Cl" with depth in profile P3, Kidd2 site 98 Figure 6.15: The relationship of As to other species in profile P3, Kidd2 site 99 Figure 6.16: The relationship of HCO3" to other species in profile P3, Kidd2 site 100 Figure 6.17: Cross-section illustrating conceptual model of groundwater flow at the Kidd2 site 102 Figure 6.18: Photomicrograph of sediments in thin section C l -41 , D N D site 108 Figure 6.19: Photomicrograph of sediments in thin section C3-41, Kidd2 site 108 Figure 6.20: Concentrations of solid-phase As in sediment in core C l , D N D site 113 Figure 6.21: Concentrations of solid-phase As in sediment in core C3, Kidd2 site 114 Figure 7.1: Conceptual model of arsenic geochemistry at the Kidd2 site 121 viii A C K N O W L E D G E M E N T S I would like to thank my supervisor, Dr. Roger Beckie for his advice during the completion of this thesis, and the opportunity to travel to Bangladesh to participate in research activities at the Munshiganj field site. Thanks also to Dr. Leslie Smith and Dr. Ul i Mayer for their participation as members of the review committee. The funding for this thesis was provided by the National Sciences and Engineering and Research Council of Canada (NSERC) in the form of an operating grant to Roger Beckie. In addition, I would like to thank the following people for their individual contributions to this thesis: • Maureen Soon and Bert Mueller (UBC), provided assistance and advice with a number of the analytical procedures and instruments. • Mario Bianchin, PhD student at U B C , developed a significant portion of the field equipment and procedures used for this investigation. He provided valuable assistance in helping plan the fieldwork and operate the equipment in the field. • Graduate student Craig Thompson was very generous with his time and effort, as he assisted with the majority of the fieldwork for this investigation. • The friendly staff at T im H o r t o n s ® provided the litres of coffee and dozens of doughnuts necessary for Craig to perform at his maximum efficiency. • Graduate students Tilman Roschinski and Paulo Herrera also provided assistance in the field. • Fellow graduate student Karin Wagner offered advice and assistance with a number of analytical procedures and instruments. Finally, thanks to Brenda for support, encouragement and lunch money while I completed my degree. ix 1.0 I N T R O D U C T I O N Arsenic (As) is ubiquitous in nature and has been estimated to be the 20 t h most abundant element in the earth's crust (Cullen et al. 1989). Arsenic is distributed throughout the atmosphere, earth materials such as soil and rock, and in natural waters in rivers, lakes, oceans, and groundwater. Sources of As contamination can be broadly categorised as being either anthropogenic or a result of natural geochemical processes. In the past 10 to 15 years, the identification of large-scale, naturally occurring As contamination in groundwater resources has led to the conclusion that As contamination of drinking water wells is by far the largest As contamination problem affecting human and ecosystem health (Smedley et al. 2002). As an aqueous species, As occurs as negatively charged or neutral species (Korte et al. 1991). The main As species in natural waters are the inorganic species arsenite [As(ILI)] and arsenate [As(V)] (Watt et al. 2003). Thermodynamic processes suggest that As(V) is present in oxic environments and As(III) is present in anoxic environments. However, numerous authors such as Cullen and Reimer (1989), Korte and Fernando (1991), and Watt and Le (2003) report that As(III) and As(V) have been detected simultaneously in both oxidising and reducing environments. Thus, the speciation of As is a function of not only redox potential and pH, but also other geochemical processes such as kinetic reaction rates, the sorption potential of the sediments, and biologically mediated reactions. Due to these complicating factors, thermodynamic equilibrium is rarely achieved in many natural, low-temperature systems. As a result, As(ffl) and As(V) may coexist in both oxic and anoxic environments. The toxicity of As is determined, in part, by its speciation (Watt et al. 2003). Inorganic arsenical species are generally more toxic than organic species, with As(ffl) considered to be more toxic to humans and other animals than As(V) [(Morton et al. 1994), (Oremland et al. 2003)]. Korte and Fernando (1991) report that As(III) is 25 to 60 times more toxic than As(V), as As(III) reacts with sulfhydryl groups, thereby increasing its residence time in the body. However, during the detoxification process, the body reduces As(V) to As(III) and then transforms the As(III) into methylated forms for excretion in the urine, skin, hair and fingernails (Masud 2000). Some research also suggests that trivalent 1 methylated species are also acutely toxic to the human body (Styblo et al. 2000). Thus, the fact that the body oxidises and reduces inorganic As (Morton et al. 1994), coupled with the suggestion that organic As species are also toxic, has lead some researchers to suggest that the difference in toxicity between As(III) and As(V) is relatively minor, and the ingestion of either species is harmful to humans [(Naqvi et al. 1994), (Yamauchi et al. 1994)]. Although the specific mechanisms by which As affects bodily functions have not been completely identified, research suggests that As inhibits energy-linked transformations in the body and inhibits cell proliferation for regeneration (Morton et al. 1994). Arsenic has been identified as a toxin that promotes carcinogenic reactions such as the inhibition of D N A replication, repair and cell division (Masud 2000), the inhibition of D N A methylation, interference with signal transduction pathways, and the promotion of chronic stimulation of growth factors (Mass et al. 2001). Thus, although As has not been specifically identified as a carcinogenic substance, it does compromise the body's ability to fight cancer. Once ingested, As can accumulate in the organs and tissues of the body. Masud (2000) reports that an estimated 40-60 % of ingested As is retained in the body. The rate of bioaccumulation is directly proportional to the concentrations of As that are ingested (Cullen et al.-1989), and numerous authors report that there is. a direct relationship between the concentrations of ingested inorganic As and incidences of human health effects [(Chen et al. 1994), (Naqvi et al. 1994)]. Human health concerns in affected people include cardiac arrhythmia and vascular system damage, such as the characteristic blackfoot disease, which is caused by a lack of circulation in the extremities followed by the formation of gangrene (Naqvi et al. 1994). Other effects include cellular damage in the gastrointestinal tract, hematological effects such as leukopenia and anemia, and dermal effects such as hyperkeratosis and hyperpigmentation. The most common health concerns in affected people in Bangladesh are melanosis (93.5 %), keratosis (68.3 %), hyperkeratosis (37.6 %) and dipigmentation leuco-melanosis (39.1 %) as well as other diseases such as cancers, and liver disease (Masud 2000). However, unlike many other skin cancers, As-related skin cancers are not limited to areas of U V radiation exposure. 2 Many patients that have As-related skin cancers also exhibit increased incidences of internal cancers such as lung, liver, bladder and prostate cancer (Naqvi et al. 1994). Given the toxic effects of As, standards and guidelines have been developed for the maximum allowable concentrations of As in drinking water. The World Health Organisation (WHO) has published a recommended maximum for As in drinking water of 10 jxg/L (Rahman et al. 2002). Due to the widespread As contamination in many drinking water wells in the Bengal Basin, the Bangladeshi government has published a less stringent recommended limit of 50 p.g/L. However, it is estimated that over 1 million people in Bangladesh could be suffering from arsenocosis, and the incidence of cancer is in the tens of thousands (Swartz et al. 2003). 1.1 Purpose and objectives Smedley and Kinniburgh (2002) identify two geochemical conditions that can cause the release of naturally occurring As to groundwater on a large scale. The first condition is the development of high p H (i.e. >8.5) conditions in arid or semi-arid, closed groundwater systems. The second trigger is the development of reducing conditions at circumneutral p H values in deltaic and floodplain sediments. The highest documented occurrences of naturally-occurring, widespread As contamination in drinking water sourced from groundwater wells are in the Bengal Basin, in Bangladesh and India (West Bengal). However, the global extent of As contamination in similar deltaic aquifers is not yet known and preliminary results suggest that concentrations of dissolved As in drinking water wells in Taiwan, Vietnam, Cambodia and Nepal could be similar in magnitude to those recorded in Bangladesh (Adamsen et al. 2002). The purpose of this thesis is to investigate the occurrence of dissolved As in the reduced, circumneutral groundwaters in the deltaic sediments of the Fraser River delta, British Columbia. The majority of the current research involving the geochemistry of As in reduced, circumneutral groundwaters is based on the Bengal Basin. Therefore, prior to addressing the Fraser River delta, a discussion of the Bengal Basin as a case study for naturally-occurring As in groundwater is presented in Chapter 2. In particular, Chapter 2 3 summarises the depositional history of the Bengal Basin and identifies the early sediment diagenesis and geochemical processes that are relevant to the cycling of dissolved As and As-influencing species. Chapter 2 also describes the groundwater hydrology of the Bengal Basin and provides a brief summary of how the As contamination problem was first identified in the drinking water. Chapter 2 concludes with a brief discussion of the mechanisms that have been proposed to explain the occurrence of the As contamination in the Bengal Basin. Chapter 3 presents a summary of the geology and the depositional history of the Fraser River delta, with a direct comparison to the Bengal Basin. In particular, the mineralogy and aqueous geochemistry, and the groundwater hydrology of the Fraser River delta is discussed. Chapter 4 provides a description of the two field sites that were explored. Chapter 5 then discusses the field and analytical methodologies that were utilised during this investigation.- Chapter 6 presents the results of the investigation and provides an interpretation of the data, with detailed discussions on the geochemistry of both the groundwater and sediments. The geochemistry at each of the two field sites is then compared, and potential geochemical mechanisms are discussed. Conclusions and recommendations are then presented in Chapter 7. 4 2.0 C A S E S T U D Y : T H E B E N G A L B A S I N Bangladesh, which constitutes the majority of the Bengal Basin, is located at the confluence of three major river systems: the Ganges, Brahmaputra and Meghna river systems (see Figure 2.1). N e p a l Dm«iiJbC-N 7 I n d i a l: , / ^ n t a h o r . B o 9 - a * Mariarganj . Sylhftt' }^ ~" [SL .Nawabgan! MyrftanswTflJi'*Gourif*jf - / ^ ^Raisn«r» Bangladesh -HabaW ^ u ™ ; e n . i n . b t o . India Ce/SBa|ari M - NMyanrrowi 9 » . i (Burma) Figure 2.1: M a p of Bang ladesh , source : www.ndi.org/worldwide /as ia /bang ladesh .asp The basin is characterised as exhibiting low relief and experiencing rapid sedimentation rates, on the order of 1 cm/y (Beckie 2002). A generalised cross-section of the stratigraphy observed in Bangladesh is provided in Figure 2.2. The stratigraphy is generally characterised as Pleistocene sediments with some overlying Holocene deposits located in paleochannels. Series of intermittent peat lenses are also located in the Bangladeshi stratigraphy. The Pleistocene sediments are generally orange, indicating that they are oxidised, whereas the Holocene sediments are grey, suggesting that they are quite reduced (Acharyya et al. 2000). The transmissivity of the Holocene sediments ranges from 1000 to 5000 m 2 /d, and is generally higher than that of the Pleistocene sediments (Ravenscroft et al. 2002). Thus, wells in Bangladesh have generally been completed in Holocene sediments. However, the elevated concentrations of dissolved As are generally associated with wells that are completed in the Holocene sediments. 5 Tig. I. Simplified geological sect km Ihuniyji N K Bangladesh. For location o f section A B C , sec F ig , A. Figure 2.2: Typical cross-section throughout Bangladesh. Note: grey area represents Holocene sediments, and the white area represents Pleistocene sediments [source: Ravenscroft et al. (2002)] Figure 2.3 provides a detailed vertical profile of a field site in the Munshiganj district of Bangladesh, where Harvey et al. (2002) completed a number of field investigations. The vertical profile that is presented in Figure 2.3 is consistent with the general trends that are observed in Holocene sediments throughout Bangladesh. Specifically, the Holocene sands are capped with a low permeability silt or clay, creating a leaky-to-confined aquifer system (Ravenscroft et al. 2002). The concentration of solid-phase As is generally less than 2 u,g/g, in the range of 1-6 ug/g. The distribution of solid-phase As is relatively constant throughout the profile and does not exhibit the peak that is characteristic of dissolved As. Solid-phase As and Fe are poorly related, and the solid-phase sulphur is negatively correlated to the concentrations of dissolved As. The solid-phase As is present as sorbed on solids, coprecipitated in solids, and incorporated into silicates and crystalline sulphides (Harvey et al. 2002). In general, the finer grained aquifers with fine sand and silt tend to exhibit the highest As concentrations in groundwater (Acharyya et al. 2000). 6 B. Solid Arsenic Depth so C. Sulfur D. Ammonia & Calcium E. DIC, DOC & CH 4 OSiWai* (mMJ * * * * * * <i»M t 10) B A s fuMl > DK (mM) I 10 100 0 1 2 3 4 5 6 7 4 5 6 7 8 0 10 11 r- : f "i A y Total nHd S *. vP * 0 D » ".. * M«trmne o * A ^ • • «3 "35/ Solid A » A \ A \ Jo \ / i b 1 • l l ! A DOC WC no ''As Extraction Targ*tt •'A'aakJyadswted • SirongtyadsorUil E3 Carbonate* AVS. •mwpnous ondos amorphous As • 'Silicate*, A»/>i • C17sL.1l lint SuHhdM pynte o i a a 4 B e 7 • M M ua-'fl 0-1 1 10 0 1 2 0.2 0.6 1.0 1 • Total Softd Sulfur f.uM-'rt a CateiimMmM) GDOCfmM) / * Acid volatile Sulfide A Methane [mM Figure 2.3: Typical vertical profile in Bangladesh, source: Harvey et al. (2002) A number of different dissimilatory-arsenate reducing prokaryote (DARP) species from Bangladeshi sediment samples have been isolated and identified (Oremland et al. 2003). As the concentration of As changes, certain D A R P species are more efficient at reducing As than others. Also, the D A R P species can also utilise a variety of electron donors to reduce As. Thus, the sediment in Bangladesh is characterised as containing a microbial population with the metabolic diversity necessary to interact with a number of chemical species such as sulphur, iron and nitrate to reduce As(V) to As(III) at varying concentrations. The following section in this chapter summarises the depositional history of the Bengal Basin. In particular, the patterns of sedimentation and the associated geomorphological processes are discussed. Early sediment diagenesis and the associated geochemical processes of the Bengal Basin are then presented in Section 2.2, followed by a discussion of the groundwater hydrology of the Bengal Basin in Section 2.3. Section 2.4 provides a brief summary of how the As contamination problem in the Bengal Basin was first identified and what complications have arisen in attempting to characterise the nature of the problem. This chapter concludes with a discussion of the mechanisms that have been proposed to explain how the elevated concentrations of dissolved As developed. 7 2.1 Depositional history Approximately 18,000 years before present (18 ka BP), in the late Pleistocene-early Holocene, sea levels in the vicinity of the present-day Bengal Basin approached a low of approximately 135 m below present day sea level (Acharyya et al. 2000). At this time, the valleys of the proto-Meghna and Ganges-Brahmaputra rivers incised deeply into the existing Pleistocene sediments, and sands and gravels were deposited at the base of the rivers. Due to the presence of a thick unsaturated zone and steep hydraulic gradients, the existing Pleistocene sediments were exposed to extensive subaerial weathering. Coupled with the fact that they had previously been exposed to long periods of groundwater flushing, the Pleistocene sediments were relatively oxidised and weathered (Kinniburgh et al. 2001). During the subsequent rise in sea level between 18 - 12 ka BP, rapid sedimentation occurred as fluvial and fluvial-deltaic, fining upward sequences of sand and some silt and clay were deposited in the valleys of the proto-Meghna and Ganges-Brahmaputra (Acharyya et al. 2000). A temporary regression in sea levels between 12 -10 ka B P caused the recently deposited sediments to undergo a slight degree of oxidation and weathering. Following a rise from 10 - 9.5 ka BP, the sea level remained relatively stable until 7 k a BP. During this time, the lower parts of the proto-Meghna and Ganges-Brahmaputra developed into marshes and lagoons. Silt and mud units that were rich in organic matter and interbedded with sand lenses from small tributaries, were deposited in the marshes and lagoons. From 7 - 5.5 ka BP, sea levels rose rapidly to higher than present-day levels, as tidal mangrove invaded the southern portions of the delta and the river valleys. During this period, extensive marine and fresh-water peat developed. After 5.5 ka BP, the sea level dropped, causing a marine regression and the migration of the shoreline to present day locations. The residual salinity that is presently observed in Holocene sediments adjacent to the coast indicates that the sediments have undergone incomplete flushing (Ravenscroft et al. 2002). The patterns of sedimentation that occurred in the Bengal Basin deposited a variety of upwards fining sequences such as braided river deposits and floodplain deposits (Nickson et al. 2000). As such, there is very little lateral continuity in the stratigraphy in Bangladesh, other than the classification of the sediments as either older Pleistocene 8 deposits, or younger Holocene sediments. The Holocene sediments are capped with a mostly continuous, low permeability unit that is comprised of silt and clay. As will be discussed in following sections, the presence of this low permeability cap is critical to the creation and maintenance of reducing conditions in the Holocene aquifers. 2.2 Early sediment diagenesis Early sediment diagenesis is a term that refers to the physiochemical changes that take place within sediments and the surrounding porewater during and after burial, generally during the first million years (Berner 1971). In sedimentary environments, authigenic minerals persist in the subsurface because they crystallise into more stable phases. However, these processes occur at very slow rates due to kinetic limitations and many geochemical processes would not occur without microbial mediation. At the time of deposition, the porosity of a sedimentary unit is a function of mineralogy. Clay minerals maintain an open structure due to electrostatic forces, whereas sand grains tend to pack together in a more geometric pattern (Berner 1971). During and after deposition, the Holocene sediments in the Bengal Basin experienced physical compaction, causing a loss of water from underlying sediments as the overlying sediments were deposited. As a result, the porosity in clay units decreased significantly as they compacted, whereas the sandy units experienced very little change in porosity. Thus, it is expected that the Holocene sediments in the Bengal Basin, which are characterised as mainly sandy units, have undergone light to moderate compaction. 2.2.1 Initial oxic conditions When minerals that contain ferrous iron [Fe(II)] such as pyrite (FeS2) undergo weathering and oxidation, Fe(II) is released into solution. Once in the dissolved form in an oxic environment, Fe(II) is oxidised to ferric iron [Fe(III)]. In solution, the Fe(III) undergoes rapid hydrolysis (neutralisation) and forms hydrous ferric oxide (HFO) minerals such as ferrihydrite [Fe(OH)3.nH20]. Ferrihydrite precipitates on pre-existing sediments, or remains in suspended form as colloids that are less than 5 urn in size (Langmuir 1997). When the aqueous solution is exposed to increases in p H or ionic strength, colloids fall 9 out of solution and form precipitates. The point of zero net proton charge (PZNPC) of ferrihydrite is at a p H of approximately 8.6 (Takamatsu et al. 1985). Below this pH, the surface of the mineral is OH" deficient, thereby creating a preference for anions in solution (Langmuir 1997). However, as the p H of the solution increases and approaches 8.6, the charge on the surface of the ferrihydrite approaches zero and the colloids in solution no longer repel each other. As a result, the ferrihydrite colloids settle out of solution and form precipitate on pre-existing sediments. Ferrihydrite colloids are also destabilised by increases in ionic strength (Langmuir 1997). Similar to when the P Z N P C is approached, when the ionic strength of a solution increases, ferrihydrite colloids no longer repel each other and they tend to flocculate and settle out of solution, forming precipitate on pre-existing grains. Ferrihydrite is precipitated as a poorly crystalline, metastable phase (Alpers et al. 1994). Once it precipitates, and prior to the onset of reducing conditions, ferrihydrite undergoes hydration and dehydration processes to form iron oxyhydroxides such as goethite (FeOOH) and iron oxides such as hematite (Fe20a), respectively (Langmuir 1997). However, goethite and hematite have slow growth kinetics at natural surface temperatures. Since the conversion of ferrihydrite to goethite or hematite is kinetically limited, iron hydroxides and oxides exist in various degrees of crystallinity within a sediment and change in mineralogy from soil to soil within similar redox and p H conditions (Masscheleyn et al. 1991). Thus, the suite of iron oxide, hydroxide, and oxyhydroxide minerals are generally referred to as hydrous ferric oxide (HFO) minerals. In general, hydroxide minerals such as goethite are more stable in saturated, low temperature conditions than oxide minerals such as hematite (Langmuir 1997). Mineral weathering patterns and fluvial transport and deposition processes played a critical role in the development of the geochemistry of the present-day Bengal Basin. Sediments from weathered rocks and minerals in the Himalayan Mountains were transported down to the Bengal Basin, via the Meghna, Ganges and Brahmaputra river systems (Ravenscroft et al. 2002). The sediments were coated in varying amounts of ferrihydrite precipitate. These sediments were rapidly deposited, as sediments accumulated in the extensive tidal mangroves and estuaries that had developed in the 10 Bengal Basin during much of the Holocene (Acharyya et al. 2000). At the saltwater interface, where river water mixed with the marine water in the estuaries, ferrihydrite colloids flocculated, creating additional ferrihydrite precipitate on the existing sediments (Howard et al. 1988). 2.2.1.1 Arsenic sorption The electrical double layer theory is applicable to the colloid chemistry of ferrihydrite colloids and ferrihydrite-coated sediment particles (Berner 1980). Type II electrical double layers consist of ions that are adsorbed at the surface of the mineral, causing it to become charged. These adsorbed mineral precipitates are the potential-determining ions and constitute the fixed layer of the mineral. Balancing counterions are attracted to the fixed layer and form the oppositely charged mobile layer. Together, these two layers constitute the electrical double layer of a colloid. There are two opposing forces in the mobile layer: the attraction to the oppositely charged fixed layer, and the concentration gradient that develops between the high concentration of ions in the mobile layer at the surface of the colloid, and the low concentration of ions in solution. The point of zero net proton charge (PZNPC) is the p H at which the overall charge of the fixed layer and the overall charge of the mobile layer are equal. As discussed in Section 2.1.2, the surface of ferrihydrite is positively charged at a p H below 8.6. Accordingly, oxyanions such as As(V) adsorb onto H F O minerals such as ferrihydrite according to equation 2.1: H F O + H A s 0 4 2 " H F O = H A s 0 4 + 2 H 2 0 (2.1) The adsorptive capacity of ferrihydrite is extremely high and adsorption continues to take place at high surface concentrations of anions such as As(V) (Pierce et al. 1982). Ferrihydrite is described as a "highly hydrated structure that is permeable to hydrated ions", where the "the ions are free to diffuse throughout the structure and are not restricted to external surface sites such as the case with more crystalline solids" (Pierce et al. 1982). Pierce and Moore (1982) and Fuller et al. (1993) report that rates of As(V) sorption on ferrihydrite are initially very fast, at a rate that decreases with time. Thus, dissolved As(V) is initially rapidly adsorbed at the surface of the ferrihydrite. This rapid 11 adsorption is followed by a slow diffusion process in which the As(V) anions migrate into the inter-particle porosity of the aggregated ferrihydrite particles (Fuller et al. 1993). Thus, as concentrations of dissolved As increase, there is a rapid initial change in concentrations of adsorbed arsenic on the surfaces of the ferrihydrite, followed by a slower diffusion controlled process. Ferrihydrite has a high adsorptive capacity for As because it is deposited as fine-grained clusters with a specific surface area of up to 600 m 2 /g (Smedley et al. 2002). As the ferrihydrite crystallises into more ordered crystalline H F O minerals such as goethite (FeOOH), the surface area of the mineral decreases to as low as 150 m 2 /g or less. Thus, the ageing of ferrihydrite into more crystalline minerals significantly reduces its adsorptive capacity. Results from de Vitre et al. (1991) confirm that the adsorption of As on ferrihydrite is much higher than on more crystalline H F O minerals such as goethite. Fuller et al. (1993) also note that the ageing of pure ferrihydrite prior to sorption experiments caused a decrease in the resulting concentrations of adsorbed As(V). This was attributed to crystal growth, which causes a reduction in the number of adsorption sites and slower rates of As(V) diffusion into the more crystalline mineral. Results of laboratory experiments indicate that As(V) has a stronger affinity for ferrihydrite than As(III). However, As(III) sorption is faster at high arsenic concentrations and pH, while As(V) sorption is relatively faster at low arsenic concentrations and p H (Raven et al. 1998). Pierce and Moore (1982) note that As(V) sorbs to a greater extent than As(III) in oxic waters, at the low concentrations observed in natural waters, and Fuller et al. (1993) also note that the adsorption of As(V) by ferrihydrite increased with decreasing p H . Although both As(V) and As(III) form inner-sphere surface complexes with ferrihydrite, As(V) has a higher affinity for the mineral (Manning et al. 1997). In their adsorption experiments, Fuller et al. (1993) found that the presence of dissolved carbonate species did not affect the initial sorption and subsequent release of As(V) on ferrihydrite. Similarly, the addition of phosphate (PO43") or sulphate (SO42") to the equilibrating solution had very little effect on the desorption of arsenic on ferrihydrite 12 (Pierce et al. 1982). Laboratory results compare well with observations from complex natural systems, as As adsorption on ferrihydrite in the presence of other competing species has been observed in natural lacustrine environments (Belzile et al. 1990). However, in their lab experiments Pierce and Moore (1982) also noted that significantly less As was adsorbed to ferrihydrite that had already been equilibrated with sulphate- or phosphate-containing solutions, causing a significant number of the adsorption sites to be already occupied. Jackson et al. (2000) also note that phosphate is a competing ligand that is effective at preventing the readsorption of As on more crystalline H F O minerals such as goethite after the As has been displaced during the dissolution of more amorphous H F O minerals. Howard et al. (1988) note that As adsorption can also be limited when As is blocked from binding sites by dissolved organic matter. In addition to H F O minerals, significant amounts of As(V) adsorb to the "oxide-like" edges of clays (Lin et al. 2000). Korte and Fernando (1991) state that iron oxyhydroxides are important for As sorption at high As concentrations, while clay content is more significant at low As concentrations. This coincides with Acharyya et al. (2000) who note that the finer grained aquifers with fine sand and silt tend to exhibit the highest As concentrations in groundwater in Bangladesh. Therefore, in addition to providing small particle sized sediments upon which significant amounts of ferrihydrite can precipitate, the edges of clay minerals can also adsorb significant amounts of As. In summary, when the Bengal Basin was deposited during the Holocene, significant amounts of sediment accumulated. The source of this sediment was the Himalayan mountains, to the north of the Bengal Basin. Physical and chemical weathering caused the rocks and minerals in the mountains created a massive sediment load that was transported to the Bengal Basin by the Ganges and Brahmaputra rivers. As the rocks and minerals in the Himalayas weathered, Fe- and As- containing minerals underwent oxidation, producing ferric [Fe(III)] iron and arsenate [As(V)] in solution. Ferric iron tends to precipitate out of solution as a ferrihydrite precipitate. Therefore, the sediments in the Ganges and Brahmaputra rivers were coated with varying amounts of ferrihydrite precipitate. In this oxic, fluvial environment, the As(V), which was present at low concentrations in the river water, rapidly adsorbed onto the ferrihydrite precipitates. 13 Once the river water discharged into the estuaries and tidal mangroves of the Bengal Basin, more ferrihydrite flocculated, providing additional fresh sorption sites to which As(V) adsorbed. Arsenate [As(V)] also accumulated in fine grained muds and organic deposits. This process of As sequestration in estuaries is illustrated by Howard et al. (1984) who observe that the flocculation of "iron-containing micro-colloids" (ferrihydrite) is an important process in the removal of As from solution in the Beaulieu estuary. Following deposition, some of the ferrihydrite aged into more crystalline minerals such as goethite, causing a certain amount of the As(V) to desorb. However, the ageing of ferrihydrite would have been decreased considerably due to the presence of As(V) anions within the inter-particle porosity of the aggregated particles (Aggett et al. 1986). Therefore, a portion of the ferrihydrite remains in the amorphous, fine-grained form. 2.2.2 Reducing conditions Throughout most of the Bengal Basin, the Holocene sands are generally capped with a low permeability silt or clay, creating a leaky-to-confined aquifer system (Ravenscroft et al. 2002). The present-day reducing conditions in the Holocene aquifers were created due to the presence of a high water table and the confining silts and clays that impede the entry of dissolved oxygen to the underlying aquifers (Acharyya et al. 2000). In reducing environments, the dissolved oxygen is exhausted and the oxidation of dissolved organic carbon (DOC) will proceed via a series of terminal electron accepting processes (TEAPs) (Lovley et al. 1995). In general, the sequence of anaerobic terminal electron acceptors in saturated sedimentary environments, in order of decreasing energy yield are: nitrate (N03~), manganese [Mn(IV)], iron [Fe(III)], sulphate (S04 2"), and carbon dioxide (C0 2 ) which reduces to methane. Although these redox reactions may achieve steady state, they are not in thermodynamic equilibrium in most natural environments (Postma et al. 1996). These processes are microbially mediated reactions, which must be catalysed in order to proceed at any significant rate. Nitrogen is found in organic matter as dissolved species and gases (Appelo et al. 1993). During nitrification, bacteria oxidise amines from organic matter to nitrate (NO3") and, to 14 a lessor degree, nitrite (NOV). The resulting NO3" is released into the adjacent groundwater. Denitrification occurs as the dissolved nitrate (NO3") is reduced to nitrogen gas (N 2) (Appelo et al. 1993) via: NO3" + 6H+ +5e" => Vi N 2 + 3 H 2 0 (2.2) And, to a much lessor degree, N 2 can further reduce to ammonium (NFL*) via: Vi N 2 + 4 H + +3e" => NH4 + (2.3) The reduction of Mn(IV), Fe(III), SO42" and C 0 2 includes preliminary fermentation steps that convert the organic carbon into simpler forms such as acetate and other short-chain fatty acids (Lovley et al. 1995). This fermentation step is the rate-limiting process for the lower energy yield T E A P s . Lovley and Chapelle (1995) note that the energy yield from the reduction of Mn(IV) and Fe(III) is very similar, and therefore would occur simultaneously in the substrate. However, a distinction between zones of Mn(IV) and Fe(III) reduction are observed in many natural settings. Microbial populations that are able to reduce both Mn(IV) and Fe(III) have been identified. Once in solution, the Fe(JJ) produced from the reductive dissolution of H F O minerals is re-oxidised by these microbial populations back to Fe(III) by the reductive dissolution of M n oxides (Lovley et al. 1995). Thus, there is no net reduction of Fe(III) until the Mn(IV) is reduced. Therefore, M n oxides play a minor role in the cycling of anions such as As because Mn(IV) will be reduced in an anoxic zone prior to the reduction of Fe(III). Distinct redox zones are not usually observed in many reducing environments. This is especially true in zones of Fe(III) and SO42" reduction. Ferric iron [Fe(UI)] reduction is a relatively slow process when minerals are involved because the co-ordination environment of the iron mineral must change, as the Fe(III) at the mineral surface forms chemical bonds with reactants and multiple metal-oxygen bonds in the mineral crystal must be broken in order to release Fe(II) into solution (Christensen et al. 2000). This dissolution process is kinetically limited, and microbial mediation is required for the process to occur at a significant rate. Also, the reductive dissolution of more crystalline 15 HFO minerals such as goethite is much slower than for more amorphous HFO minerals such as ferrihydrite. Unstable HFO minerals such as ferrihydrite will undergo reduction with no associated SO42" reduction because the reduction of these less stable minerals 2 2 yields higher energy than S O 4 " reduction. However, S O 4 " will.undergo reduction in the presence of more stable HFO minerals such as hematite and goethite because the energy yield from S O 4 2 " reduction is equal to, or greater than, the energy yield from the more stable HFO minerals. Therefore, depending upon the stability of the HFO minerals present within the subsurface, S O 4 2 " may or may not undergo reduction (Postma et al. 1996). Postma and Jackobsen (1996) also note that as the pH of the porewater increases, then S O 4 2 " reduction becomes more energetically favourable than stable ferric mineral reduction. Thus, the broad distribution of different HFO minerals present in the subsurface creates a broad redox zone in which simultaneous Fe(III) and S O 4 " reduction may occur. The reducing conditions in many aquifers in the Bengal Basin were created as the low permeability cap impeded the flow of O2 into the underlying Holocene aquifers. In the absence of aerobic processes, other species are reduced as bacteria oxidise the DOC that is sourced from the decomposition of organic matter. In the vicinity of the elevated concentrations of dissolved As, the concentrations of O2 and NO3" are generally non-2- •+* detectable, the concentrations of S O 4 " are low, and the concentrations of N H 4 and methane are elevated [(Harvey et al. 2002), (Anawar et al. 2003)]. Also, dissolved Fe is present predominantly in the reduced form as Fe(II). The influence of reducing conditions on iron, carbon, sulphur, and arsenic cycles is discussed in detail in the following sections. 2.2.2.1 Iron cycling As discussed in the preceding section, the distribution of HFO minerals present in the subsurface ranges from amorphous ferrihydrite to more crystalline goethite. In the absence of higher energy yielding TEAPs, the oxidation of organic matter (CH2O) will be accompanied by the reductive dissolution of HFO minerals via the following generalised equation (Nickson et al. 2000): 16 4 F e O O H + C H 2 0 + 7 H 2 C 0 3 => 4 F e / + + 8 H C O 3 " + 6 H 2 0 (2.4) The dissolution rate of H F O minerals is kinetically limited and is dependent upon the degree to which the minerals have crystallised Therefore, the reductive dissolution of H F O minerals is a slow process in which the more amorphous ferrihydrite will undergo reductive dissolution first, followed by the reduction of less energetically favourable, more crystalline mineral phases such as goethite (Christensen et al. 2000). Agget and Roberts (1986) observe that the dissolution rates of H F O minerals decreases considerably if the minerals are deposited as a precipitate on pre-existing grains as opposed to deposited in the crystalline form. Therefore, the more amorphous minerals in the Holocene sediments in Bangladesh will be the first minerals to undergo reductive dissolution in the presence of organic carbon oxidation, followed by the reduction of more crystalline minerals. Analytical results from contaminated drinking water wells in Bangladesh indicate that the concentrations of dissolved SO42" are very low [(Harvey et al. 2002), (Anawar et al. 2003)]. This suggests that SO42" and H F O minerals are undergoing simultaneous reduction. In order for this to happen, some of the H F O minerals that are present in the subsurface are expected to be the more stable, crystalline forms. As suggested in equation 2.4, as H F O minerals undergo reductive dissolution, the concentrations of dissolved F e 2 + and bicarbonate ( H C O 3 ) in the porewater increase. Depending upon the concentrations of F e 2 + and HCO3", the porewater may become oversaturated with respect to siderite (FeCOs) and precipitate out of solution according to equation 2.5: F e 2 + + HCO3" <^ F e C 0 3 + H + (2.5) The relationship between F e 2 + and HCO3" in Bangladeshi groundwater is generally poor (Ravenscroft et al. 2002). This may be in part because, in addition to the reductive dissolution of H F O minerals, HCO3" is also produced from the reduction of NO3" and SO42", controlled by the precipitation and dissolution of a number of different mineral phases, and influenced by the p H of the porewater. The processes that influence the concentrations of HCO3" in a reducing environment are discussed in more detail in Section 2.2.2.2. 17 In the presence of significant concentrations of HS", Fe may also be sequestered in the precipitation of sulphide minerals such as pyrite: F e 2 + + 2HS" <=> F e S 2 + F T + 2e- (2.6) The precipitation of pyrite and its influence on As is discussed in more detail with respect to sulfur cycling in Section 2.2.2.3 of this thesis. However, in general, the concentrations of F e 2 + and S0 4 2 " are negatively correlated in Bangladeshi groundwaters that exhibit elevated concentrations of As. This suggests that sulphide minerals are precipitating in the vicinity of high concentrations of dissolved As (Ravenscroft et al. 2002). 2.2.2.2 Carbon cycling Bicarbonate (HCOY) in groundwater is generated by a number of geochemical processes including the chemical weathering of minerals, the dissolution of sedimentary carbonates, the reductive dissolution of H F O minerals and species such as NO3" and SO4 microbial respiration, and during the natural buffering of a porewater to changes in pH. Depending upon the groundwater conditions, HCO3" will remain in solution as HCO3", protonate to form carbonic acid ( H 2 C 0 3 ) or deprotonate to form carbonate (CO3 2 ) : HCO3" + H + <=> H 2 C 0 3 (2.7) HCO3" <=> H+ + C O 3 2 (2.8) Thus, in low p H solutions, H 2 C 0 3 will be the dominant species and in high p H solutions CO3 2 " will be the dominant species. In the. circumneutral groundwaters of Bangladesh, HCO3" is the dominant dissolved inorganic carbon species. Organic matter ( C H 2 0 ) is thermodynamically unstable and oxidises in the presence of oxygen via the following generalised equation: R C H 2 0 + 0 2 => C 0 2 + H 2 0 + R (2.9) where R is an organic residue that does not react and remains after the redox reaction (Leventhal 1982). In the circumneutral groundwaters of Bangladesh, the C 0 2 that is produced in equation 2.9 and other redox reactions will speciate into HCO3" via equation 2.10: 18 C 0 2 + H 2 0 <^> H C 0 3 " + H + (2.10) As discussed in Section 2.2.2, in the absence of 0 2 , organic carbon will oxidise anaerobically via the reduction of NO3", Mn(IV), Fe(III) and SO42", and methanogenesis. In the highly reducing groundwaters in Bangladesh, Fe(III) and SO4 2" reduction, and, to some degree, methanogenesis, are occurring. In circumneutral groundwaters, all of these processes produce HCO3". Concentrations of HCO3" are also regulated by mineral dissolution and precipitation reactions. When a solution becomes saturated with respect to HCO3", and depending upon the concentrations of certain cations in solution, a number of minerals may precipitate out of solution. Minerals such as siderite (FeCCh) and calcite (CaCOs) may precipitate according to equations 2.5 and 2.11, respectively: C a 2 + + HCO3" <^> C a C 0 3 + r T (2.11) However, Appelo and Postma (1993) note that the precipitation of calcite can be strongly inhibited by phosphate. Thus, in aqueous systems with significant concentrations of phosphate, the solution may achieve and maintain supersaturation with respect to calcite. Given the different sources and sinks of HCO3", it is difficult to identify the mechanisms and processes that influence the concentrations of HCO3" in the Holocene aquifers in Bangladesh. In general, the concentrations of HCO3" and As are positively correlated. 2.2.2.3 Sulphur cycling Sulphate (SO4 2 ) can be introduced into groundwater by infiltrating rainwater, due to the mixing of fresh water with brackish water, the oxidation of pyrite in shallow, oxic zones, the application of fertilisers, and the weathering of sulphate-bearing minerals such as gypsum (CaS0 4 -2H 2 0) , anhydrite (CaS0 4 ) (Goldhaber et al. 1974). In reducing environments, the fundamental stages in sedimentary pyrite formation include the reduction of SO42" to HS", the reaction of dissolved HS" with iron and iron minerals to form non-crystalline iron monosulfides (FeS) such as mackinawite and greigite, and the transformation of iron monosulfides into sedimentary pyrite (FeS 2) (Berner 1970). 19 H F O minerals are the most important source of sedimentary iron in the formation of iron sulfides in natural sedimentary environments (Morse et al. 1987). As discussed in Section 2.2.2, depending upon the stability of the H F O minerals present, Fe(IJJ) and SO4 2 " reduction can occur simultaneously in the subsurface (Postma et al. 1996). In the presence of organic carbon ( C H 2 O ) , SO4 2 " reduces via equation 2.12: R C H 2 0 + SO4 2 " => HS" + H C O 3 " + R (2.12) However, the reduction of SO4 2 " will not occur at natural, shallow subsurface temperatures in the absence of microbial catalysis (Goldhaber et al. 1974). Levanthal (1982) and Morse et al. (1987) report that SO4 2 " reduction is mediated by Desulfovibrio and Desulfotomaculum organisms. These microorganisms are not only capable of reducing SO4 2 ", but also tolerant of heavy metals and the high sulphide concentrations that are characteristic of reducing environments (Goldhaber et al. 1974). Once in solution, dissolved HS" will react with available F e 2 + and H F O minerals to form non-crystalline iron monosulphides (FeS) such as mackinawite (Berner 1970): F e 2 + + HS" <=> FeS + F T (2.13) These components of the sulphides in a sediment suite are referred to as the acid-volatile sulphides (AVS) . In the absence of organic matter, H F O minerals can also undergo reductive dissolution in the presence of HS" [(Berner 1970), (Goldhaber et al. 1974)]. The products of this redox reaction include monosulphides and elemental sulphur: 2 F e O O H + 3HS" <^> 2FeS + S° + H 2 0 + 30H" (2.14) The resulting monosulphides are very fine grained and thermodynamically metastable in natural sedimentary conditions (Morse et al. 1987). Thus, during early sedimentary diagenesis, monosulphides convert into pyrite according to equation 2.15: FeS + S ° < ^ F e S 2 (2.15) The rate of pyrite formation is generally much slower than the precipitation of iron monosulphides (Berner 1970). However, the rate of monosulphide and pyrite 20 precipitation varies for different sedimentary environments, as they are affected by the efficiency of the native microbial populations, the availability of organic matter and SO42" and the availability and form of the H F O minerals present. The reductive dissolution of H F O minerals proceeds as two phases: an initial rapid phase in which surface reduction occurs, followed by a dissolution phase from which most of the F e 2 + is produced (Pyzik et al. 1 9 8 1 ) . The rate of dissolution follows equation 2 . 1 6 : ([(reduction Fe) I dt = k A p e o o H S T ° 5 0H+° 5 ( 2 . 1 6 ) where k is the rate constant, A is the specific surface area of the H F O mineral, S T is the total molar sulphide concentration, and an+ is the hydrogen ion activity. Thus, finer grained, amorphous ferrihydrite with a greater surface area will undergo dissolution prior to the more crystalline H F O minerals such as goethite. If the rate of H F O reductive dissolution that produces Fe(II) is quite slow, then the ratio of FeS to F e S 2 in the sediments could also be quite low (Goldhaber et al. 1 9 7 4 ) . Therefore, in the presence of sufficient SO42" and organic matter, pyrite formation can be iron limited, even in clastic-dominated sediments (Dean et al. 1 9 8 9 ) . In general, concentrations of Fe and SO4 " in Bangladeshi groundwaters are negatively correlated, further suggesting that pyrite precipitation in the vicinity of the elevated dissolved As concentrations (Ravenscroft et al. 2 0 0 2 ) . 2.2.2.4 Arsenic cycling 2.2.2.4.1 Arsenic release As discussed in Section 2 . 2 . 1 . 1 , As is sorbed onto H F O minerals such as ferrihydrite and goethite during sedimentary deposition. When H F O minerals undergo reductive dissolution (equation 2 . 4 ) , adsorbed As is released into solution (Ravenscroft et al. 2 0 0 2 ) . Numerous authors have reported that the processes of As adsorption onto H F O minerals and desorption at the onset of H F O reductive dissolution are the mechanism by which As is cycled in lacustrine environments [(Aggett et al. 1 9 8 8 ) , (Aurillo et al. 1 9 9 4 ) , (Belzile et al. 1 9 9 0 ) , (Seyler et al. 1 9 8 9 ) , (Spliethoff et al. 1 9 9 5 ) ] . Correlation between As and Fe profiles in ocean environments also suggest that As is released to porewaters at the same 2 1 time as iron, as FIFO minerals undergo reductive dissolution [(Edenborn et al. 1986), (Sullivan et al. 1996)]. Smedley and Kinniburg (2002) suggest that in reducing environments, adsorbed As(V) may reduce to As(III), causing a certain amount of As desorption in the absence of H F O dissolution. However, laboratory results indicate that under anaerobic conditions, adsorbed As(V) does not undergo reduction, but remains adsorbed onto H F O minerals (Langner et al. 2000). Also, Cummings et al. (1999) report that in laboratory experiments, the dissimilatory Fe(III)-reducing bacterium (DIRB) "Shewanella alga B r Y " mobilised As(V) during H F O reduction without reducing it. These results suggest that the introduction of As(III) to porewater is a two-step process by which As(V) is first desorbed, possibly during the reductive dissolution of H F O minerals, followed by the reduction of As(V) to As(III) by an As-reducing bacteria, or due to chemical reduction (i.e. HS"). Thus, these studies suggest that, in the absence of the reduction of H F O , the reduction of adsorbed As(V) to As(III) is not a mechanism by which significant amounts of As are released into solution. Anawar et al. (2003) identify HCO3" as one of the most effective natural leaching agents in groundwater regimes. Laboratory experiments have confirmed that the adsorption of HCO3" on H F O minerals is very strong (Zachara et al. 1987), as carbonate species cover the majority of the sorption sites on H F O minerals, thereby limiting the sorption of other anions such as As (van Geen et al. 1994).' Zachara et al. (1987) and van Geen et al. (1994) performed laboratory experiments on the adsorption of carbonate on hematite and ferrihydrite, respectively. Comparison of their results suggests that the binding of carbonate species is comparable on a per site basis for different H F O minerals. Laboratory results from K i m et al. (2000) indicate that the release of As from sulphide minerals is strongly related to the concentrations of HCO3" in the leaching solution (Kim et al. 2000). The authors of the study suggest that As-carbonate complexes form, thereby releasing As from sulphide minerals and transporting the As into solution. They suggest that these As-carbonate complexes are incorrectly identified in anion exchange studies, leading to the incorrect identification of As species in solution. Lee and Nriagu (2003) performed laboratory experiments in which CO2 gas was bubbled through an anoxic arsenic trioxide solution. The analytical results indicate that As-carbonate complexes had 22 formed, confirming K i m et al.'s (2000) suggestion that As-carbonate complexes are an effective mechanism by which As can be leached from arsenical sulphides. These As-carbonate complexes, which are thought to include As (C03)2~ , A s ( C 0 3 ) ( O H ) 2 \ A s ( C 0 3 ) 2 ( O H ) 2 \ and A s ( C 0 3 ) + , are considered to be stable in solution. Although the carbonation of arsenian sulphide minerals is a potential mechanism for As to be leached into groundwater, the concentration of such sulphides is quite low in the aquifers of Bangladesh (Ravenscroft et al. 2002). However, Appelo et al. (2002) propose that As-carbonate complexes could potentially cause As that is adsorbed on H F O minerals to leach into solution. Results from laboratory experiments indicate that a solution with H C 0 3 " (NaHC03 solution) efficiently extracted As from Bangladeshi sediments (Anawar et al. 2003). This could be due to the formation of As-carbonate complexes on the surface of the iron hydroxide mineral, suggesting that although the adsorbed As is nonexchangeable, it is capable of undergoing rapid ligand exchange. Thus, once an As-carbonate complex is formed, the As is less strongly adsorbed to the H F O minerals, and leaches into solution. Although thermodynamic processes indicate that As(V) is present in oxic environments and As(III) is present in anoxic environments, both species have been detected in both oxidising and reducing environments (Watt et al. 2003). At circumneutral p H and under reducing conditions, As(V) is present predominantly as H A S O 4 " and the As(III) is present predominantly as H3ASO3. Thus, the reduction of As(V) to As(III) in circumneutral waters follows equation 2.17: H A s 0 4 2 " + 4 H + + 2e-<^> H 3 A s 0 3 + H 2 0 (2.17) In the absence of microbial mediation, the reduction of As(V) to As(III) is a relatively slow process (Seyler 1989). Even with microbial catalysis, adsorption can significantly reduce the rate of As speciation in natural systems (Korte et al. 1991). As discussed above, during the reductive dissolution of H F O minerals, adsorbed As(V) is released into solution prior to it being reduced (Cummings et al. 1999). Therefore, as H F O minerals undergo reductive dissolution, As(V) can be released into solution quickly, followed by the relatively slower process of As(V) reduction. The accumulation of As(V) in anoxic 23 environments has been documented in the anoxic hypolimnion layers of lakes in the Aberjona Watershed, Massachusetts (Aurillo et al. 1994). 2.2.2.4.2 Arsenic mobility Once in solution, As(III) is more mobile than As(V) (Watt et al. 2003). Due to the different sorption potential of As(III) and As(V) for H F O minerals, differential scavenging of As occurs in the subsurface (Spliethoff et al. 1995). As(III) is a neutral, o ' uncharged molecule ( H 3 A S O 3 ) at the p H of most natural groundwaters (Manning et al. 1997). Therefore, As(III) is less strongly adsorbed on most mineral surfaces and more mobile than the negatively charged As(V) oxyanions. Results of laboratory experiments confirm that As(V) and As(III) do in fact exhibit different adsorption isotherms and therefore, travel at different velocities (Gulens et al. 1973). Harvey et al. (2002) characterise the sorption of As in the Bangladeshi subsurface as following a Langmuir isotherm, and Smedley and Kinniburg (2002) characterise the transport of As as following a non-linear adsorption isotherm. However, in reducing environments at an approximate p H of 8, the velocity at which As(III) and As(V) move through sand columns is much slower due to adsorptive influences (Gulens et al. 1973). If the rate at which As(V) is released into solution during the reductive dissolution of H F O minerals is greater than the rate at which As(V) is reduced to As(III), then As(III) may migrate along the flowpath and As(V) may persist in the vicinity of the H F O minerals. This chromatographic separation may partially account for the variation in As:Fe ratios that are observed in different areas of Bangladesh, as the relationship between Fe and As at the source may become uncorrelated due to differential transport. However, a p H of 7 is optimum for As(III) adsorption (Pierce et al. 1982). Therefore, in the circumneutral groundwaters of Bangladesh, the separation of As(ffl) and As(V) along the flowpath may be limited. 2.2.2.4.3 Arsenic sequestration As discussed in Section 2.2.1, H F O minerals were deposited under oxic conditions as colloids and precipitates in the Bengal Basin. When the H F O minerals undergo reductive 24 dissolution, a certain amount of freshly exposed H F O would provide fresh sites available for anion adsorption. New adsorption sites would also be created on the surface of H F O minerals if As were leached from the mineral surface due to the formation of As-carbonate complexes. Smedley and Kinniburg (2002) suggest that low Ka values, rather than high absolute concentrations of As in the sediment, are responsible for the elevated concentrations of dissolved As in reducing groundwater. In sediments with high Kd values, it is expected that As that has been released to groundwater will be readsorbed onto the fresh H F O sorption sites. In addition to sorption mechanisms, As can also be sequestered during the precipitation of arsenian sulphide minerals. As discussed in Sections 2.2.2.1 and 2.2.2.3, when significant concentrations of HS" and F e 2 + accumulate, a solution may become saturated with respect to pyrite (FeS2) (equation 2.6). Numerous authors have detailed the mechanisms by which Fe, S and As are sequestered in ocean environments due to the precipitation of pyrite [(Belzile et al. 1986), (Berner 1970), (Dean et al. 1989), (Goldhaber et al. 1974)]. The concentration of As is generally enriched in pyrite relative to bulk sediments due to the enrichment of the metastable monosulphides (FeS) that are the precursors to pyrite (Wilkin et al. 2002). In the Laurentian Trough of the St. Lawrence seaway, As is present in the lattice sites of authigenic pyrite with a Fe:As ratio of 103 (Belzile et al. 1986). However, the rate at which amorphous iron monosulphides react with elemental sulphur to form pyrite is on the scale of years in natural, sedimentary environments. In addition to being included into the lattice of pyrite minerals, As species form outer sphere sorption complexes on sulphides (Farquhar et al. 2002). As(V) is strongly partitioned to iron monosulphides (FeS) such as mackinawite and pyrite (FeSi) at pH>5 (Bostick et al. 2002). However, Farquhar et al. (2002) note that As adsorption on, and coprecipitation with, mackinawite is more efficient than pyrite. Outer sphere complexation with As creates a different phase on the surface of mackinawite. This phase is most likely a poorly crystalline arsenian sulphide precipitate such as orpiment ( A S 2 S 3 ) (Farquhar et al. 2002). Bostick et al (2002) also note an arsenian sulphide precipitate ( A S 2 S 3 ) on pyrite, but found that it occurs at the adsorption edges, and only 25 form in highly sulphidic environments. In slightly sulphidic environments, As(V) sorption on FeS and FeS2 results in the formation of a FeAsS-like surface precipitate. The results from Farquhar et al. (2002) and Bostick et al. (2002) suggest that although As is adsorbed on, and coprecipitated with, FeS and FeS2, relatively low amounts of As are sequestered during these processes. When a solution becomes supersaturated with respect to arsenic and sulphide, orpiment (As2S3) and realgar (AsS) may also precipitate out of solution according to equations 2.18 and 2.19 (Kim et al. 2000): 2 H 3 A s 0 3 + 3 H + + 3HS" => A s 2 S 3 + 6 H 2 0 (2.18) 2As2S 3 + 2 H + => 4AsS + 2HS" (2.19) However, the precipitation of orpiment and realgar only occur under acidic conditions [(Cullen et al. 1989), (Kim et al. 2000), (Watt et al. 2003)] and at high S:As ratios in the order of 20:1 (Rochette et al. 2000). Therefore, in the circumneutral groundwaters of Bangladesh which are characterised as having a low S:As ratio due to a relatively low mass loading of SO42", the precipitation of arsenian sulphide minerals is not expected to occur. In addition to Fe oxide and sulphide minerals, both As(ffl) and As(V) can complex with humic acid amine groups in organic-rich environments (Keon et al. 2001). Thus, As cycling in reducing environments can be complicated'by a number of complexation and mineral dissolution and precipitation processes. 2.3 Groundwater hydrology of the Bengal Basin Although many authors have explored the aqueous geochemistry of the Holocene aquifers in the Bengal Basin, the literature provides very little detail on the transport processes that are occurring in the area. In general, much of the southern portion of the basin is flooded during the monsoon months, and the groundwater table is relatively shallow during the warmer months (Beckie 2002). The Bengal Basin is characterised as 26 having low topographic relief, and as a result, the water table exhibits low hydraulic gradients. The aquifer units within the substrate are poorly connected. Isotopic evidence suggests that the groundwater in Bangladesh has a long residence time, with modern groundwater generally observed only in the top 20 to 40 m, while deeper groundwater ranges in age from 2 - 12 ka BP, suggesting that the Holocene sediments have experienced relatively little flushing since deposition [(Harvey et al. 2002), (Kinniburgh et al. 2001), (Smedley et al. 2002)]. This is supported by Ravenscroft et al. (2002), who note that the residual salinity observed in the Holocene sediments indicates insufficient flushing. Thus, if the rate at which geochemical processes are releasing As to the groundwater is rapid in comparison to the rate at which groundwater migrates along the flowpath, As would tend to accumulate in source areas, as is observed in the Bangladeshi profile. The relatively high pumping rates associated with irrigation wells can create strong vertical gradients. Harvey et al. (2002) suggest that such gradients could mobilise young organic carbon from ground surface and transport it to depth. The resulting oxidation of the organic matter at depth could induce the reductive dissolution of H F O minerals, thereby releasing adsorbed As into solution. Acharyya et al. (1997) also suggest that the extraction of groundwater for irrigation may be linked to the As contamination in Bangladesh. However, they note that these suggestions are mere speculation in the absence of sufficient data. In contrast, Ravenscroft et al. (2002) discount irrigation pumping as influencing the concentrations of dissolved As in Bangladesh. Van Geen et al. (2003) suggest that concentrations of dissolved As might slowly increase as a tube well ages. The authors report a small but statistically significant increase in As concentrations in tube wells that is estimated at a rate of 16 ± 2 u,g/L per decade of well age. This trend could be the result of patchy As distributions in the subsurface, as wells draw water in from surrounding contaminated areas, or the development of vertical flow paths that transport young organic matter from the surface, thereby driving the reductive dissolution of H F O minerals and releasing the associated As. Therefore, the authors 27 suggest that baseline levels in sensitive areas need to be established in order to investigate the problem further. 2.4 Identification of arsenic contamination in the Bengal Basin Prior to the 1970's, most of Bangladesh's rural population consumed water from surface water sources and hand-dug wells that were contaminated with human and agricultural wastes (Bearak 2002). As a result, thousands of people died each year due to water borne diseases such as hepatitis, cholera and diarrhoea. In response to this health concern, international aid organisations sponsored the installation of groundwater wells to provide Bangladeshi citizens with a clean source of drinking water, and a water supply for irrigation during the dry months. These wells proved to be free from the water borne diseases that were prevalent in surface water sources. However, no analyses were performed for concentrations of dissolved metals. In the 1990's, health officials began noting high incidences of arsenicosis and other arsenic-related diseases in Bangladesh. Concurrently, the British Geological Survey detected elevated concentrations of dissolved As in groundwater wells, and subsequently, - uncovered a massive As contamination problem in the Bengal Basin (Beckie 2002). This widespread problem has been described by some health officials as the worst mass poisoning of a population in history (Smith et al. 2000). In efforts to quantify the scale of the As contamination problem in Bangladesh and implement a safe drinking water campaign for its citizens, a number of different aid and academic organisations have conducted field programs to identify and label contaminated wells. A number of different As testing field kits have been used in Bangladesh and West Bengal. These kits include the Merck kit, the N I P S O M (National Institute of Preventative and Social Medicine) kit, and to a lessor degree, the G P L (General Pharmaceuticals Limited) kit, the H A C H kit, the Arsenator field kit and the A I I H & P H (All India Institute of Hygiene and Public Health) kit (Rahman et al. 2002). Based on the results of these field programs, many wells in Bangladesh have been characterised as safe (i.e. less than the Bangladeshi As standard of < 50 ug/L) and painted green, or unsafe 28 (>50 | ig /L dissolved As) and painted red. However, in their study of the reliability of different As field testing, kits, Rahman et al. (2002) collected numerous samples from wells that had already been labelled as safe or unsafe. The water samples were analysed with more accurate analytical laboratory methods such as the flow injection hydride generation atomic adsorption spectrometry (FI -HG-AAS) technique. The laboratory analytical results indicated many false negatives and false positives, suggesting that the field kits are qualitative at best. Thus, the authors conclude that although the As testing field kits provide a relatively inexpensive method for classifying drinking water wells, the health costs associated with false classifications far outweigh the usefulness of the field kits. As discussed in Chapter 1 of this thesis, the speciation of dissolved As indicates the geochemical conditions in which the As is present, and has implications for the toxicity of the As. In the late 1980's, the scientific community started to become aware of the fact that As(III) is more prevalent in groundwater than what had previously been reported (Korte et al. 1991). As a result of better sampling techniques, researchers began to develop more effective protocols for sample preservation once groundwater samples were collected and exposed to oxidising conditions. A review of the literature by Korte and Fernando (1991) suggested that the best method of sample preservation had been filtration with a 0.45 um filter and acidifying to p H 2. However, many researchers agree that the best method is to speciate As in the field. Huang and Dasgupta (1999) present a field-based method for As speciation that includes a portable stripping voltammetric instrument. The water sample is first acidified to p H 1-2. Arsenite [As(III)] concentrations are determined with a gold film electrode at a deposition potential of -0.2 V (vs. A g / A g C l in 1 M HC1). A l l of the As is then oxidised with the addition of an oxidant to the water sample and the total As is determined by a deposition potential of -1.6 V . The original As(V) concentration is then calculated by difference (Huang et al. 1999). The method is fairly complicated, requires extra equipment, and must be run off a notebook computer. The cartridge method proposed by Le et al. (2000) offers the advantage that it is a simple, cheap and reliable method for the extraction of As(ffl) and As(V) in the field. This method involves pumping the water sample through a 0.45 urn 29 filter for dissolved species, then a silica-based strong anion exchange cartridge. As(V) sorbs to the anion cartridge and the As(III) passes through the cartridge due to the differences in sorption potential between the two species. The resulting water sample is analysed for concentrations of As(III). The cartridge is eluted with H C L and the eluent solution is analysed for As(V) (Le et al. 2000). 2.5 Proposed mechanisms for As contamination in the Bengal Basin Since the As contamination problem was first identified, a number of hypotheses have been proposed in an attempt to explain the mechanisms by which the elevated concentrations of dissolved As have developed. Hypotheses that have been put forward by the scientific community include: the release of As due to the oxidation of As-bearing pyrite (Mandal et al. 1998), the displacement of As from sorption sites on aquifer minerals due to competitive exchange of phosphate (PO43) anions sourced from fertilisers (Acharyya et al. 2000), and the release of As that is associated with the reductive dissolution of iron oxyhydroxides (HFO minerals) and oxidation of organic carbon (Nickson et al. 2000). In addition to these three hypotheses, researchers have also suggested that As-carbonate complexes could potentially cause As that is adsorbed to H F O minerals to leach into solution [(Anawar et al: 2003), (Appelo et al. 2002)]. 2.5.1 Oxidation of pyrite As discussed in Section 2.2.2.4.3, As is sequestered during the precipitation of pyrite. This has led some researchers to suggest that the oxidation of pyrite, which would release the sequestered As, is the mechanism by which the As contamination has developed in Bangladeshi groundwaters (Mandal et al. 1998). However, this suggestion has been discounted for a number of reasons. Primarily, the reaction for the oxidative dissolution of pyrite follows equation 2.20 (Ravenscroft et al. 2002): 2FeS 2 + 7 0 2 +2H 2 0 => 2Fe 2 + + 4H+ + 4S0 4 2 " (2.20) 30 Therefore, the dissolution of pyrite creates Fe:SC*42" at a molar ratio of 0.5. Analytical 2 2 results of Bangladeshi groundwaters indicate that Fe and SO4 ", and As and SO4 " are negatively correlated (Anawar et al. 2003). Ravenscroft et al. (2002) also note that the rate of pyrite oxidation is greatest under oxic conditions. Therefore, if the oxidation of arsenian pyrite were responsible for the release of As into groundwater in Bangladesh, one would expect to see the highest concentrations of As in the shallow, dug wells which are the most oxic. This is not observed in Bangladesh. Rather, the elevated concentrations of dissolved As are located zones that are highly reduced (McArthur et al. 2001). Furthermore, Nickson et al. (2000) note that the abundance of pyrite is very low in Bangladeshi sediments, ranging from 0.02 % to 0.3 %. Thus, pyrite is not present in concentrations high enough to provide a significant source of As to the groundwater. This has led researchers to conclude that pyrite is a sink for, rather than a source of, As in the Bangladeshi aquifers. 2.5.2 Competitive anion exchange: phosphate and arsenic Acharyya et al. (2000) suggest that the elevated concentrations of As in Bangladeshi groundwaters are a result of competitive anion exchange between As and the P0 4 3 " that is sourced from fertilisers. Arsenate [As(V)] shares similar chemical characteristics with P 0 4 3 \ suggesting that As(V) and PO4 3 are easily exchangeable on HFO minerals (Smedley et al. 2002). The displacement of adsorbed As(V) with PO4 3" has been demonstrated in estuarine sediments (Manning et al. 1997). However, the results of this study suggest that soluble and readily exchangeable As was a minor fraction (<5 %) of the total solid-phase As in the estuarine sediments. Anawar et al. (2003) performed sequential extraction experiments on Bangladeshi sediments. The results of their experiments showed that a relatively low percentage of As in the sediments is mobile, suggesting that the application of phosphorous-containing fertilisers does not have an effect on the mobilisation of As. Laboratory results have also estimated P043":As partition ratios in the order of 5000 (Ravenscroft et al. 2002). Thus, the upper limit of 5 mg/L for concentrations of PO4 3" in Bangladeshi groundwater would only cause concentrations of dissolved As in the order of 2 ug/L. 31 McArthur et al. (2001) also note that waters containing significant concentrations of As, P or Fe contain concentrations of H C O y that exceed 200mg/L, suggesting that the waters are fairly evolved. However, if competitive anion exchange between As and PO4 " was the mechanism driving the As contamination, relatively younger waters that are contaminated with fertilisers would contain the highest concentrations of P and As. Thus, McArthur et al. (2001) suggest that NHV" and P0 4 3 " in the groundwater in the contaminated zone are byproducts of the reductive dissolution of buried peat deposits and, to a lessor degree, from contamination from latrines and from the desorption of minor amounts of PO43" from H F O minerals. However, as discussed in Section 2.2.1.1, PO43" is effective at preventing the readsorption of As on H F O minerals once the As has been displaced. Therefore, although competitive anion exchange between As and PO43" may not be a direct mechanism by which the elevated concentrations of dissolved As have formed in Bangladeshi groundwaters, the presence of PO43" in the groundwater may be an indirect factor. 2.5.3 Reductive dissolution of H F O minerals As discussed in Section 2.2.2.1, in the presence of organic carbon, and in the absence of other, higher energy yield oxidants, H F O minerals such as ferrihydrite and goethite will undergo reductive dissolution. Many authors suggest that the reduction of H F O minerals, and the associated release of adsorbed As, is the mechanism by which the elevated concentrations of dissolved As have developed in Bangladesh [(Anawar et al. 2003), (McArthur et al. 2001), (Nickson et al. 2000), (Ravenscroft et al. 2002)]. Nickson et al. (2000) report that, based on the estimated amount of diagenetically available Fe in Bangladeshi sediments, the concentration of diagenetically available As could generate concentrations of dissolved As as high as 289-517 ppm, suggesting that H F O minerals could be a significant source of As. Van Geen et al. (2003) note that under reducing conditions, a loss of only approximately 0.1 ug/g of As in the solid phase can result in the enrichment of the surrounding groundwater by up to 200 jxg/L dissolved As. This is an insignificant loss from the As content of the sediments in the Bengal Basin, which range in As concentration from 1-6 ug/g. Also, As and Fe are present in the sediments at Fe:As 32 molar ratios that vary from 1500 to 6000 (McArthur et al. 2001). McArthur et al. (2001) suggest that this variation is due to the variable nature of As to Fe on H F O minerals, as the two elements are sequestered at different rates during the various precipitation and ageing processes of H F O minerals. Harvey et al. (2002) also determined that, at their field site in Munshiganj, solid-phase As was found to be positively correlated with Fe, as the As was adsorbed to, and coprecipitated with Fe in solids, and incorporated into silicates and crystalline sulfides. The results also indicate that less than 6 % of the Fe(UI) was in the form of amorphous iron oxyhydroxides (FeOOH), suggesting that if ferrihydrite controls As mobility, it does so in small quantities. However, the idea that the introduction of organic carbon causes the microbially mediated reductive dissolution of F e O O H and the associated release of As and, upon consumption of the organic carbon, the subsequent resorption of As on fresh sorption sites, is supported by the results of field redox manipulation experiments that were conducted by Harvey et al. (2002). The correlation between dissolved As and Fe in Bangladeshi groundwater is poor, although the correlation between As and H C 0 3 " is better. Ravenscroft et al. (2002) propose that the reductive dissolution hypothesis still has some merit, and the relationship between Fe and As, and Fe and HCO3" are not clear due to the influence of other geochemical processes such as the reduction of NO3" and SO42", mineral precipitation and dissolution, and readsorption. Nickson et al. (2000) suggest that once H F O minerals undergo reductive dissolution, the F e 2 + does not remain in solution and precipitates as siderite (FeCOs). The original Fe to As relationship could also be masked by the migration of As(III) from the source. In addition to the relationship between Fe and As, there is also uncertainty associated with the source of the organic carbon that is proposed to drive the H F O mineral dissolution. The presence of NH4 + and CH4 indicates that the degradation of organic matter is occurring in the Bangladeshi subsurface. Anawar et al. (2003) note that the correlation between dissolved As and dissolved organic carbon (DOC) suggests that organic matter plays a significant role in controlling As transport. Ravenscroft et al. 33 (2002) note that the organic carbon that is disseminated throughout the aquifer matrix is oxidised, and the carbon contained in overbank deposits is entrapped in low permeability clays and silts. Thus, Ravenscroft et al. (2002) suggest that neither is a significant source of organic carbon. Rather, the authors suggest that the source of the organic carbon that drives redox processes in the subsurface is the series of peat lenses that are intermittently located in the Bangladeshi stratigraphy. However, Harvey et al. (2002) used 1 4 C and 8 1 3 C isotope analyses to determine that recent biogeochemical reactions at their field site in Munshiganj were driven by relatively young organic carbon that was transported by the recent inflow of groundwater. The results of the isotope analyses indicate that the D O C in the groundwater is 3000-5000 years old, which is similar to the ages of the peat lenses identified at other field sites. In contrast, the DIC is dated at a much younger age of approximately 700 years BP. Methane, an oxidative byproduct of carbon metabolism, in the water samples was also dated at 700 years BP. The authors suggest that the organic carbon that drives the redox processes at the As peak is young organic carbon that is transported from the surface in infiltrating water. Further, Harvey et al. (2002) note that, as with other field sites in Bangladesh, the peat lenses that have been proposed as the redox driver were not present in the subsurface at their field site in Munshiganj. However, van Geen et al. (2003) note that thin peat lenses are not easily detected when analysing the cuttings from many drilling rigs such as air rotary rigs. Thus, thin peat lenses may in fact be present at some locations, even though they are not reported in borehole logs. • 2.5.4 Competitive anion exchange: bicarbonate and arsenic As discussed in Section 2.2.2.4.1, the results of K i m et al (2000) and Lee and Nriagu (2003) suggest that arsenic carbonate complexes exist. The laboratory results from Anawar et al. (2003) indicate that HCO3" extracts arsenic from Bangladeshi sediments efficiently. The positive correlation between As and HCO3" in Bangladeshi groundwaters supports the idea that HCO3" may cause As to be displaced from the surfaces of H F O minerals. However, no other evidence supports this proposed mechanism. 34 3.0 F R A S E R R I V E R D E L T A The Fraser River Delta is located in the Lower Mainland area, on the southwest coast of British Columbia. The delta is mainly comprised of the Municipality of Delta, which includes Ladner, and the City of Richmond (see Figure 3.1). Figure 3.1: Location map of the Fraser River delta, note: the delta includes the City of Richmond and the Municipality of Delta (source: www.bctravel.com/ van/van.gif) 3.1 Basement geology The basement geology in the vicinity of the Fraser River delta is composed of clastic sedimentary rocks that were sourced from the Coastal Mountains to the north and the Cascade Mountains to the southeast, and range in age from Late Cretaceous to Tertiary (Clague 1998). The Coastal Mountains are composed mainly of Jurassic and Cretaceous aged granitic rocks, and the Cascade Range is mainly comprised of Devonian to Cretaceous aged volcanic and sedimentary rocks, with a small granitic fraction (Monger et al. 1994). This basement geology is overlain by thick sequences of Quaternary sediments. The pattern of deposition during the Quaternary was strongly influenced by glacial processes during the Pleistocene and the effects of changing sea levels during the Holocene (Clague et al. 1998). 35 3.2 Depositional history of the Fraser River delta As with the Bengal Basin, changes in sea level had a profound effect on sedimentation patterns as the Fraser River delta was deposited during the Holocene. During much of the Pleistocene, the glaciers and ice fields that formed the Cordilleran Ice Sheet covered southwestern British Columbia (Clague et al. 1998). The maximum extent of the last glacial period was approximately 15000 years before present (15 ka BP), at which time the glacier overlying the present day Fraser River lowlands is estimated to have been over 2 km thick. During this time, the Fraser River lowlands experienced glacial erosion and glacial till units were deposited along the base of the glaciers. The glaciation of the Fraser River lowland area ended approximately 11 ka BP. As the glaciers retreated, the sediments from the ablating glaciers were rapidly deposited as a series of sedimentary successions that are up ,to several hundred meters thick. As these sedimentary successions were deposited, the underlying glacial till units were eroded by fluvial and coastal erosion. Thus, the distribution of glacial till that underlies the Quaternary sedimentary successions is thin and patchy." Approximately 8-10 ka BP, the Fraser River floodplain rapidly prograded from its mouth, which was located near present-day New Westminster, depositing a delta into the Georgia basin. Since that time, the Fraser River delta has continued to expand towards the west and southwest (Clague et al. 1998). The rate at which the Fraser River delta prograded was affected by changes in the relative sea-level. The sea-level has risen approximately 12m in the last 9000 years, with the majority of that rise occurring during the period from 9 - 2.5 ka B P (Williams et al. 1988). The vertical accretion and lateral progradation of the Fraser River delta has continued over the past 9000 years at rates ranging from 2.4 mm/y - 5.3 mm/y and 1 m/y - 6.5 m/y, respectively, depending upon the rate at which the sea-level was rising (Williams et al. 1988). Unlike the Bengal Basin, the sediment supply to the Fraser River delta was sufficient for the rate of deposition to keep pace with the rate of sea-level rise. Thus, rises in sea-level did not permit the development and inland invasion of the extensive tidal marshes and lagoons that existed in the Bengal Basin. Therefore, other than the thick bog deposits that cap large portions of the Fraser River delta, there are relatively few peat lenses or other 36 sources of organic carbon within the delta stratigraphy. Unlike the proto-Meghna and proto-Ganges Brahmaputra systems, the Fraser River delta was not confined to eroded river valleys. Rather, as the Fraser River delta expanded, it blanketed the majority of the lowlands area with young sediments. The evolution of the Fraser River delta at 10 ka BP, 5 ka BP, and at present is illustrated in Figure 3.2. Figure 3.2: Evolut ion of the Fraser River del ta, note: yel low = Ho locene sed iments , grey = pre-Ho locene landmass [source: C lague , 1998)] The deltaic sediments that were deposited as the Fraser River delta migrated towards the Georgia basin are up to 300 m thick (Clague et al. 1998). The stratigraphy is characterised as consisting of three separate units: deep marine deltaic bottomsets are overlain by deltaic foresets, which are in turn overlain by subaqueous to subaerially deposited topset sediments (Clague et al. 1998). The bottomset unit consists of clayey silts that were deposited in deep water marine environments. As the Fraser River delta migrated towards the Georgia basin, the foreset unit was deposited as thick sequences of sands and silts were deposited along the delta foreslope. The topset unit, which ranges from 10 - 30m in thickness, can be further catagorised into four major lithologic units (Williams et al. 1988). The subaqueous clays and silts that comprise the base of the topset unit are overlain by well-sorted medium to coarse sands that rarely contain wood fragments, and otherwise, lack significant organic matter. These distributary-channel sands were deposited under relatively high-energy fluvial conditions and subject to wave and tidal modification. The channel sands are overlain by interbedded sands and silts, which generally fine upwards. Traces of plant remains and wood fragments are scattered throughout the interbedded sand and silt unit, which is interpreted to have been deposited in mid-tidal-flat environments (Williams et al. 1988). This sand and silt unit, which 37 ranges in thickness from 2-6m in most areas, grades into an organic-rich silt that contains organic matter both finely disseminated throughout the unit and concentrated in horizontal laminations. This organic silt unit, which ranges in thickness from l-10m, was deposited in low energy environments such as tidal marshes, and as overbank deposits in fresh and brackish water marshes during flooding events (Clague et al. 1983). Patchy organic peat bog deposits cap the topset unit. The bog deposits range in thickness from l-4m in the central part of the Fraser River delta, and taper off towards the western edge of the delta. Analyses of the organic fragments within the peat unit suggest a vertical succession from brackish water to freshwater species, with significant accumulation of organic matter occurring after the delta surface was high enough to avoid regular flooding (Clague et al. 1983). Since the surficial geology was strongly influenced by episodic flooding events, the Fraser River delta plain is markedly level, and mainly below an elevation of 2m above sea-level (Clague 1998). Figure 3.3 illustrates the present-day surficial geology of the Fraser River delta plain. 3.3 Mineralogy and aqueous geochemistry The following section provides a detailed discussion of the geochemistry and groundwater hydrology of the topset units of the Fraser River delta. Similar to those of the Bengal Bengal basin, the sediments of the Fraser River delta were coated with varying degrees of ferrihydrite and other H F O mineral precipitates. Once the organic-rich silt was deposited as a cap over the distributary-channel sands and interbedded sands and silts, reducing conditions developed. 38 Figure 3.3: Surficial geology of the present-day Fraser River delta, source: Clague (1998) Mackintosh and Gardner (1966) conducted an investigation of the mineralogical and chemical properties of the shallow overbank sediments of the lower Fraser Valley. The results of their investigation suggest that the sediments in the top 24 inches (0.6m) of the profile are detrital in nature, reflecting their source area, with some influence from marine diagenesis processes. Samples that were collected at different depths and locations throughout the lower Fraser River delta all contained a similar suite of minerals: the silt fractions (2-50um) contained primarily quartz, feldspar, mica and chlorite, with minor amounts of amphiboles and pyroxenes, and the clay fraction (<2um) contained mainly montmorillonite and chlorite, with some minor kaolin and micaceous minerals present (Mackintosh et al. 1966). 39 Mathews and Bustin (1994) investigated the effectiveness of the surficial peat deposits of the Fraser River delta in sequestering heavy metals from leachate beneath a sanitary landfill. The results of the investigation suggest that heavy metals, notably As , were highest in the lower part of the peat succession where the clean bog sediments are interbedded with significant inorganic sediments (Mathews et al. 1994). The authors note that consistently high concentrations of Fe and M n were associated with accumulations of other metals, suggesting that Fe and M n may have been reduced into solution and transported down into favourable Eh conditions, where they precipitated as minerals. These Fe and M n precipitates then adsorb dissolved metals that are transported from the surface, sequestering them from solution. Mathews and Bustin (1994) concluded that metals such as As could be retained in these interbedded peat deposits for long periods of time. Mackintosh and Gardner (1966) and Mathews and Bustin (1994) are limited to the shallow overbank deposits, and do not discuss the mineralogy and geochemistry of the distributary-channel sands. Simpson and Hutcheon (1995) provide the results of a detailed investigation of both sediment and pore-water geochemistry at various locations on the Fraser River delta and in the adjacent off-shore environments, at depths ranging from ground surface to up to 40m. Simpson and Hutcheon (1995) classify the sediments of the Fraser River delta into two general categories: shallow clay-rich sediments and deeper sand-rich sediments. The sand-rich sediments are generally immature lithic arenites, containing a mixture of igneous and metamorphic rock fragments, reflecting the diversity of the Fraser River drainage basin (Simpson et al. 1995). Similar to the results published by Wride, et al. (2000), Simpson and Hutcheon (1995) estimate the approximate mineral composition (by weight %) of the channel sands to be mainly quartz (50-60 %) and feldspar (30-40 %). Mica is reported as ubiquitous at concentrations up to 15 %, calcite as detectable at concentrations up to 11 %, with chlorite, amphibole and pyrite present at concentrations generally <2 %. The clay-rich sediments (0.2-2 urn) are predominantly illite at concentrations >45 %, with concentrations of quartz and feldpar each at approximately 10 %, smectite and chlorite at 5-10 %, and kaolinite at up to 10 %. 40 Siliceous diatoms (single-celled organisms that secrete siliceous frustules/cell walls) and pyrite framboids were noted in both the sand and clay units at certain locations. Results from Simpson and Hutcheon (1995) also suggest that the pore-waters in some of the deep zones of the Fraser River delta appear to be connate, with little evidence of extensive meteoric flushing since the time of deposition. Although the pore-waters in the deltaic sediments appear to be a function of the mixing of the meteoric and marine water at the time of deposition, some deviations from the mixing line are observed, suggesting that there has been some diagenesis since the time of deposition. Due to the presence of the low hydraulic conductivity overbank and organic peat deposits, the groundwaters in the Fraser River delta tend to be very reducing. In general, Na + and M g 2 + are linearly correlated with Cl", following a mixing trend between the meteoric and marine water at the time of sediment deposition. Ferrous iron (Fe 2 +), M n 2 + , and HCO3" exhibit a very poor linear trend with Cl", suggesting that the distribution of 2 2+ these species is strongly controlled by diagenetic reactions, while SO4 " and Ca profiles show control from both mixing and diagenetic processes. The concentrations of Fe 2 + in the pore-water are higher than in marine water, suggesting that there has been some addition of Fe 2 + to the pore-water since deposition. Simpson and Hutcheon (1995) suggest that the likely source of the increased Fe 2 + is the bacterially mediated reduction of HFO minerals. The presence of pyrite framboids in certain sediments suggests that precipitation of iron monosulphides is occurring as the Fe 2 + reacts with the HS" that is produced by the bacterially mediated reduction of SO42". The pattern of M n 2 + is similar to the pattern observed for Fe 2 + , and the authors suggest that the source of the M n 2 + is the bacterially mediated reductive dissolution of Mn oxide minerals. As discussed in Section 2.2.2.2, HCO3" is influenced by mineral precipitation and dissolution, natural pore-water buffering processes, and microbial respiration during the oxidation of organic matter and reduction of HFO minerals, Mn oxides, and SO42", and during methanogenesis. However, Simpson and Hutcheon suggest that Fe and Mn reduction occur in smaller quantities than the extensive SO4 " reduction that is associated with marine depositional environments. The authors also note that the concentrations of 41 HCO3" and SO42" in Fraser River delta pore-water are inversely proportional. This 2 2 relationship is pronounced in marine SO4 "-rich waters, and much weaker in less SO4 "-rich meteoric waters. At depth, the pore-waters in the Fraser River delta sediments are saturated with respect to calcite. In particular, the pore-waters are super-saturated with respect to calcite at a depth of 5-10m. The authors speculate that at this depth in the profile, the rate of HCO3" production is greater than the rate of calcite precipitation. The patterns of SO42" concentrations suggest that there has been a deviation from the original mixing trend between marine and meteoric waters, as some SO4 " has been removed from solution since deposition. Sulphate is reduced by anaerobic bacterial activity, producing HS" (equation 2.12). This is supported by the fact that the remaining S0 4 2" is enriched in 3 4 S, as the 3 2 S in S0 4 2" is preferentially reduced. However, the proportion of 3 4 S in the SO42" is lower in the zones in which framboidal pyrite has been deposited. This is because 3 4 S is preferentially incorporated into the mineral aggregate during precipitation, as indicated by positive 534S-values in sedimentary sulphide minerals (Cameron et al. 1995). Simpson and Hutcheon (1995) note that in aquifer units that are hydraulically connected to distributary channels, the observed geochemical relationships are influenced by the episodic tidal influx of marine water. Neilson-Welch and Smith (2001) investigated saline water intrusion processes at the Kidd2 site in Richmond. For approximately 10 months of the year, at high tide, seawater from the Georgia Strait extends into the Fraser River along the riverbed. The results of their investigation suggest that this relatively dense water infiltrates into the adjacent permeable aquifer units that are hydraulically connected to the Fraser River (Neilson-Welch et al. 2001). Thus, the geochemical relationships between the pore-water and sediments would be significantly influenced by the presence of saline water in such aquifer units. Cameron (1995) also notes that the concentrations of cations and anions in groundwater in the vicinity of the hyporheic zone tend to get diluted to some degree during the freshet, at a time when water levels in the Fraser River are high and groundwater tends to be recharged by the river (Cameron 1995). Such seasonal influences would also affect the geochemical relationships between the pore-water and sediments near the hyporheic zone. 42 The British Columbia Ministry of Water, Land and Air Protection ( B C M W L A P ) - Lower Mainland Region conducted an investigation of concentrations of dissolved As in domestic water wells in Mission, B C (Zubel 2002). The author notes that elevated concentrations of dissolved As (>25 ug/L) were found in 9 out of 129 water well sites that were sampled. These results suggest that the occurrence of dissolved As in the area is not widespread, as 7 out of the 9 contaminated wells were located within a small, localised area. After consideration of the local geology, Zubel concludes that the elevated concentrations of dissolved As are a result of local geologic conditions, suggesting that the dissolution of As mineralisation along fractures in the granitic bedrock is the source of the dissolved As. The author also speculates that the drawdown associated with operating the domestic wells may introduce oxic waters to the screened depths, possibly causing desorption of As from oxide surfaces. Mission is situated in the uplands area of the Lower Mainland, and is located to the east of the Fraser River delta (see Figure 3.4). Thus, the aquifer units in the Mission area are not hydrogeologically connected to the Fraser River delta sand aquifers. t W e « Vancouver North Vancouver Downtown Vancouver Vancouver it**-*. Port Moody t>ort Maple Radge New Wesmirciiste* ftkhmond Ladner South Delta Figure 3.4: Location map of the Municipality of Mission, note: Mission is east of the Fraser River delta (source: www2.jurock.com/areainfo/ mission.asp) Therefore, the relatively oxic conditions and local geology in the Mission area are unlike the reduced sand aquifer units that are located in the Fraser River delta. The results of a literature search indicated that the Zubel (2002) paper is the only documented investigation into concentrations of dissolved As in the Lower Mainland. 43 3.4 Groundwater hydrology The public water supply for the City of Richmond and the Municipality of Delta is currently serviced by the Greater Vancouver Regional District ( G V R D ) distribution network, which is sourced from the Coquitlam, Capilano, and Seymour watersheds. A search of the B C M W L A P "Aquifers and Water Wells of B C " webpage indicated that there are no records of water wells in the Fraser River delta area, suggesting that there has been little development of the groundwater resources of the delta (see Figure 3.5). Therefore, little information has been published on hydrogeological parameters such as hydraulic conductivity, hydraulic gradients, etc. of the Fraser River delta sediments. Figure 3 . 5 : Resu l ts of well water search for F raser River del ta, note: sea rch conducted on B C M W L A P "Aquifers and water wel ls of B C " webpage : ht tp: / /maps.gov.bc.ca/apps/wlap_aqui fer / Ricketts (1998) presents a three-dimensional numerical simulation of groundwater flow in the Fraser River delta (Ricketts 1998). The results of the model suggest that shallow groundwater beneath the Fraser River delta is recharged both by direct precipitation and topography-driven flow from the adjacent uplands. The results of the model also suggest a simulated average horizontal flow velocity that is generally less than 1 m/day. However, this flow velocity at any given location within the Fraser River delta could vary 44 due to aquifer anisotropy and heterogeneity. Neilson-Welch and Smith (2001) estimate hydraulic conductivity (K) of the distributary-channel sands at the Kidd2 site to be in the order of 4 x 10"4 m/s, with hydraulic gradient (i) estimates ranging from 0.005 in the deep piezometers (16-18m depth) to 0.0004 for the intermediate depth piezometers (10-12m depth) (Neilson-Welch et al. 2001). From these estimates, a Darcy flux (q = K x i) in the order of 2.0 x 10"6 m/s and 1.6 x 10"7 m/s, is estimated for the deep and intermediate zones at the site, respectively. Furthermore, assuming a porosity (r|) of 0.3 for the site, the groundwater velocity (v = q / TJ) for the deep and intermediate zones is estimated to be in the range of 0.58 m/day and 0.05 m/day, respectively. These estimates are consistent with Ricketts' suggestion that groundwater flow in the region is relatively slow. 45 4.0 F I E L D S I T E D E S C R I P T I O N Two field sites in Richmond, B C were chosen for this investigation of the Fraser River delta: the Department of National Defence (DND) site is located in the central, uplands area of the delta and the Kidd2 site is located adjacent to, and to the south of, the north arm of the Fraser River (see Figure 4.1). Figure 4.1: Locat ions of the D N D and Kidd2 field s i tes, note: the D N D site is located in a relatively central up lands a rea of the Fraser River del ta, and the Kidd2 site is located adjacent to the north arm of the Fraser River, (source: http:/ /map.ci ty.r ichmond.bc.ca/website/gis/viewer.htm) 4.1 Department of National Defence (DND) site The Department of National Defence (DND) site is located north of Westminster Highway, between No. 4 Road and Shell Road. Figure 4.2 presents an aerial photograph of the D N D site. A n army battalion is located in the northern corner of the property. Discussions with personnel at the D N D battalion indicated that a radar site that was equipped with buildings and towers was located in the central portion of the property in the 1950's. However, other than an overgrown road that extends roughly north from Westminster Highway through the central portion of the property, there are no visible 46 remains of the radar station on the site. Otherwise, the rest of the site is undeveloped bog that is covered in brush and small trees. Figure 4.2: Aerial photograph of the DND site. The approximate profiling and coring location (P1 & C1) and profiling location P2 are indicated as yellow circles (source: http://map.city.richmond.bc.ca/website/gis/viewer.htm) Inspection of the surficial geology map presented in Chapter 3 (Figure 3.3) suggests that the stratigraphy in the vicinity of the D N D site is capped with an organic peat bog deposit that is up to 4 meters thick. Thus, it is expected that infiltrating water at the D N D site contains a high load of dissolved organic carbon (DOC) and other associated species. However, the low permeability of the peat bog deposits and the underlying organic-rich silt unit would limit infiltration, as indicated by the standing water observed throughout the majority of the D N D site. The topography of the site is relatively level, and the elevation of the property is less than 2 meters above geodetic datum (mean sea level). 4 .2 Kidd2 Field Site The K i d d 2 site is located near the northern arm of the Fraser River, west of No. 4 Road, and south of River Road (see Figure 4.3). The site is owned by B C Hydro, and the majority of the site includes the K i d d 2 substation, with a Terasen Gas works yard located 47 to the south of the property. The areas adjacent to the site are developed: industrial properties are located to the north and west, and residential properties are located to the south and east of the site. The eastern portion of the Kidd2 site, immediately adjacent to No. 4 Road, is a cleared park area. This portion of the property is operated by U B C as a research site. As a result, a number of field investigation programs have been conducted at the Kidd2 site, including cone penetrometer tests, ground freezing and sampling, and the completion of a number of groundwater monitoring wells [(Neilson-Welch 1999), (Wride et al. 2000)]. Therefore, the local stratigraphy is well characterised. Appendix A presents borehole logs from previous investigations at the Kidd2 site. Figure 4.3: Aerial photograph of the Kidd2 site. The approximate profiling and coring location (C3 & P3) is indicated by the yellow circle (source: http://map.city.richmond.bc.ca/ website/gis/viewer.htm) 48 5.0 M E T H O D O L O G Y 5.1 Sample collection 5.1.1 Water samples The D N D site is representative of natural, undisturbed conditions of the delta in the vicinity of overlying peat deposits. The Kidd2 site is adjacent to the north arm of the Fraser River, and therefore permits an investigation of the influence of the saline wedge that is present along the western reaches of the river. The locations of profiles PI and P2 at the D N D site, and P3 at the Kidd2 site, are provided in Figures 4.2 and 4.3, respectively. At each profiling location, a Waterloo Drive Point Profiler (WDPP) system was used to collect water samples. The W D P P was advanced through the surficial peat and silt units into the underlying sand and silt unit. Once sediments with a hydraulic conductivity sufficient to permit the extraction of groundwater were encountered, the collection of samples commenced. Samples were generally collected every 5 feet (1.52 m) of depth, although the sampling frequency was modified depending upon the hydraulic conductivity of the sediments that were encountered. For example, at depths where low hydraulic conductivity sediments were encountered, the W D P P was advanced in 2.5-foot (0.76 m) increments until a sample could be collected. Each profile was terminated when low hydraulic conductivity materials were encountered at depths of approximately 79 feet (24 m), 76.5 feet (23.3 m) and 74 feet (22.5 m) in profiling holes PI, P2 and P3, respectively. These low hydraulic conductivity units are interpreted to be the clay and silt unit that comprises the base of the topset unit. The W D P P system, which was developed at the University of Waterloo, enables the collection of discrete samples from specific depths, thereby permitting the construction of a detailed vertical characterisation at a given location. The W D P P consists of a profiler tip that is connected to 5-foot (1.52m) lengths of 1% inch (3.5cm) internal diameter (ID) PA inch (4.4cm) outer diameter (OD) A W rods. The profiler tip is fashioned with six 6mm diameter screened ports. Each port is connected to a sealed central reservoir, which is connected to lA inch O D low density polyethylene (LDPE) tubing that runs internally through the entire length of the A W rod assembly. At ground surface, the H D P E tubing 49 was connected to a Geotech Geopump 2 (300 rpm) bi-directional, variable speed peristaltic pump, and then through a flow through cell which discharged into a purge container. Appendix B presents photographs that show the profiling equipment. The profiler tip and A W rod assembly was advanced with a pneumatic hammer that was suspended from scaffolding. As the assembly was advanced into the ground, additional A W rods were connected as necessary. As the assembly was advanced, distilled de-ionised water (DIW) was pumped down through the internal tubing system in order to prevent the ports from clogging. Once the profiler tip was advanced to the desired depth, the direction of the peristaltic pump was reversed and groundwater was drawn up through the tubing, the flow through cell, and into a purge container. The flow through cell was equipped with a Denver Instrument™ UP-25 p H meter, and an O r i o n ™ model 115 electrical conductivity meter. The pH, temperature and electrical conductivity of the groundwater were monitored, and once stability was achieved, the corresponding values were recorded and the flow rate, with the pump running at full speed, was measured in order to provide a relative estimate of the hydraulic conductivity of the sediments at each depth. . .* * '' Select parameters such as dissolved oxygen (DO), ferrous iron (Fe 2 +), ammonium (NFl4+), phosphate (PO43), arsenic (As) and alkalinity were analysed in the field. Samples were also collected for laboratory analyses such as concentrations of dissolved cations, dissolved anions, dissolved organic carbon (DOC) and concentrations of arsenite [As(UI)]. A l l samples were extracted from the H D P E tubing through a 3-way stopcock with a 60mL syringe, thereby minimising oxidation of the sample prior to filtration. The samples were filtered with dedicated 30mm 0.45um cellulose acetate syringe filters directly into high-density polyethylene (HDPE) sample bottles. The samples that were collected for concentrations of As(III) included filtering the sample through both a syringe filter and a Supelco Supelc lean™ L C - S A X SPE tube, which contains ion-exchange packings that efficiently remove As(V) anions from solutions, but permit the passage of As(III) anions. 50 The samples for dissolved cations were preserved with nitric acid (HNO3) to a p H of approximately 2 and the samples for dissolved anions were not preserved in the field. The samples for D O C that were collected at the D N D site were preserved with hydrochloric acid (HC1) to a p H of 2, while the D O C samples that were collected at the Kidd2 site were not preserved with acid. Upon collection, all samples were placed in a cooler with ice packs for transport to the laboratory, where all samples were frozen within 12 hours of collection. 5.1.2 Sediment samples Based on the analytical results of the water samples from profiles PI , P2 and P3, sediment cores were retrieved from specific depths at locations C l and C3, located adjacent to profiles PI and P3, respectively. 5.1.2.1 D N D site Two sediment cores were collected at location C l , at depths of 25-29.5 feet (7.6-9.0 m) and 40-43.75 feet (12.2-13.3 m). The location of C l , which is approximately 1 meter east of profile PI , is presented in Figure 4.2. The sediment cores were collected using a coring system that was developed based on a sample-freezing drive shoe core sampler system developed by the U.S. Geological Survey (USGS) (Murphy et al. 1996). The corer that was used for this project consists of a 5 foot (1.52cm) 2 inch (5.1cm) ID 2xh inch (6.4cm) O D core barrel that is equipped with a drive point shoe. The interior of the core barrel is equipped with a 2 inch (5.1cm) O D V/% inch (4.8cm) ID p o l y v i n y l chloride) (PVC) core liner. The core barrel is attached to 5-foot (1.52 m) PA inch (4.4cm) ID 2lA inch (5.7cm) O D rods of casing. The drive point piston, which fits inside the drive.point shoe, is threaded onto 5-foot (1.52m) P/% inch (3.5cm) O D internal rods that run the length of the core barrel and casing. The piston is equipped with rubber o-rings to ensure a tight seal between the inside of the drive shoe and the piston. The drive shoe is also equipped with an external gas line that runs the length of the assembly, and is attached to a tank of compressed carbon dioxide (CO2) gas at the ground surface. As with the W D P P system, the coring assembly was advanced with a pneumatic hammer that was suspended from scaffolding. As the coring assembly was advanced, additional 51 casing and internal rods were added as necessary. Once the drive point shoe was located at the top of the desired coring interval, the internal piston and drill rod assembly was secured to the scaffolding, so as to remain stationary. The external drive point shoe and core barrel assembly was then advanced into the sediment, 5 feet (1.52m) past the internal piston. The sediment in the drive shoe was then frozen with C O 2 gas. This ensured complete recovery of the core, as the plug of frozen sediment in the drive shoe prevented loss of sediment from the core liner as the entire assembly was retrieved. Appendix B provides photographs of the coring system. Once the coring assembly was retrieved, the internal core liner was removed from the core barrel. The ends of the core liner were cut to length and capped with P V C caps and sealed with duct tape in order to maintain anoxic conditions in the core sample. The core samples were transported back the laboratory and frozen within 6 hours of collection. The core samples remained frozen until sequential extractions were completed. 5.1.2.2 Kidd2si te Three sediment cores were collected at location C 3 , at depths of 25-26.5 feet (7.6-8.1 m), 40-43 feet (12.2-13.1 m) and 50-53 feet (15.2-16.2 m). The location of C 3 , which is approximately 1 meter east of profile P3, is presented in Figure 4.3. The sediment cores at the Kidd2 site were collected with a B-80 mud rotary drilling rig that was operated by M u d Bay Drilling Co. Ltd. The borehole was advanced to the top of the desired core interval at 25 feet (7.6 m) below ground surface. A geoprobe tool, which is equipped with an internal 4-foot (1.2 m) P V C liner that is fashioned with a sediment trap and connected to internal drill rods, was advanced with a pneumatic hammer 3 to 4 feet (0.9-1.2 m) past the bottom of the borehole. The assembly was then retrieved and the core liner was removed from the geoprobe. The core liner was cut to length, capped with P V C caps and sealed with duct tape. The borehole was then advanced to the next core sampling depth and the core sampling process was repeated at the 40 foot (12.2 m) and 50 foot (15.2 m) sampling depths. The core samples were transported back to the laboratory and frozen within 4 hours of collection. 52 5.2 Sample analyses 5.2.1 Water samples 5.2.1.1 Field analyses As discussed in Section 5.1.1, during profiling, at each depth the p H and the electrical conductivity (EC) of the groundwater was monitored. Once stabilisation occurred, the p H and E C were recorded, an estimate of the relative hydraulic conductivity was recorded, and a number of field analyses were conducted. A l l field analyses were conducted on groundwater samples that were extracted from the 3-way stopcock with a 60mL syringe and filtered with a 30mm 0.45um cellulose acetate syringe filter. Samples that were extracted from the top 4 profiling depths were analysed for concentrations of dissolved oxygen (DO) with C F f E M e t s ® K-7501 kits. A l l samples were analysed with a H A C H ™ DR/2010 spectrophotometer to determine concentrations of ferrous iron (Fe 2 + ) and ammonium (NFLt+), utilising the 1,10 Phenanthroline and Nessler methods, respectively. At the Kidd2 site, samples were also analysed with the spectrophotometer for concentrations of phosphate (PO4 "), utilising the PhosVer3 method. A H A C H ™ 28000-88 arsenic (As) test kit was used on selected samples to determine the absence or presence of dissolved As and, if present, provide an estimate of the concentration. Alkalinity titrations were performed to determine the concentrations of bicarbonate (HCO3") at each sampling depth. Hydrochloric acid (0.1N HC1) was gradually added to 25, 30 or 35 m L of sample over a number of small increments with a micropipet. The p H of the solution was monitored and recorded until the p H was past the inflection point, in the range of 2.90 or lower. The results of each titration were analysed with the Gran titration mathematical technique. 5.2.1.2 Laboratory analyses 5.2.1.2.1 DND site 53 Groundwater samples from the D N D site were analysed in the laboratory for concentrations of dissolved cations, dissolved anions, and dissolved organic carbon (DOC). The samples were thawed and shaken prior to all analyses. Concentrations of dissolved cations and concentrations of As(III) were analysed with an inductively coupled plasma - mass spectrometry (ICP-MS) instrument using the semi-quant method at the Department of Earth and Ocean Sciences (EOS) at U B C . Concentrations of dissolved anions were anaysed with a Dionex DX100 ion chromatograph (IC) at E O S , U B C . Concentrations of D O C were anaysed with a Shimadzu T O C (Infra-red) analyser, using the High Temperature Combustion method. The D O C samples were analysed at the Environmental Engineering Laboratory, Department of Civi l Engineering at U B C . 5.2.1.2.2 Kidd2site Groundwater samples from the Kidd2 site were also analysed for concentrations of dissolved cations, dissolved anions, and D O C . Due to the presence of the saline wedge at the Kidd2 site, concentrations of dissolved chloride in the groundwater at the Kidd2 site are much greater than the groundwater, at the D N D site. As a result of the associated chloride interference effects, samples from the Kidd2 site were analysed for concentrations of D O C with an Elementar High T O C analyser, and concentrations of dissolved As were analysed with a Perkin-Elmer 3300 Hydride Vapour Atomic Absorption Spectrophotometer. Both of these analyses were performed at A L S Environmental analytical laboratory in Vancouver, B C . 5.2.2 Sediment samples 5.2.2.1 Polished thin sections Two sediment samples were collected from coring locations C I and C3, both at a depth of 41 feet (12.5 m). Polished thin sections of these samples were made at Vancouver Petrographies Ltd. in Langley, B C . 54 5.2.2.2 Sequential extractions Sequential extraction procedures (SEPs) permit the quantification of various pools of solid phase As in a given sediment sample. Appendix C presents a background discussion on sequential extraction procedures (SEPs) and the exchange mechanisms by which the various forms of As are released from sediments. Although SEPs have been used by a number of different researchers, there is no standard procedure associated with implementing SEPs [(Li et al. 1995), (Gleyzes et al. 2001), (Keon et al. 2001), (Shiowatana et al. 2001), (Wenzel et al. 2001), (Van Herreweghe et al. 2003)]. Therefore, when using SEPs, the researcher must develop a protocol that is tailored for the specific application. Table 5.1 provides a summary of publications that present SEPs for As-bearing sediments. It is also important for a chosen SEP to be consistent with published methods to provide results that are directly comparable to other relevant investigations (see discussion in Appendix C). A number of publications have employed SEPs to assess the nature and concentrations of As in Bengal Basin sediments [(Anawar et al. 2003), (Ahmed et al. 2004), (Akai et al. 2004), (Harvey et al. 2002), (Tareq et al. 2003)]. The details of the various SEPs that were utilized in these papers are summarized in Table 5.2. The objective of this current investigation is to compare the concentrations of aqueous and sedimentary As in the Fraser River delta sediments to the results that are published from similar, more documented locations such as the Bengal Basin. A SEP that is consistent with the publications presented in Table 5.2 was developed for this current investigation. The SEP that was developed for this investigation incorporates the first 4 steps of the SEP presented by Keon, et al. (2001), and includes a fifth extraction for crystalline Fe oxyhydroxides that is similar to the extraction procedure presented by Wenzel, et al. (2001). Concentrations of total solid phase As on split samples were determined neutron activation analysis (NAA) . Details of the SEP that was developed for this investigation are provided in Table 5.3. 5.2.2.2.1 Sequential extraction procedure 55 The core samples and all the necessary equipment were transferred to an anaerobic chamber which was purged with helium until less than 2 % atmospheric oxygen was present in the chamber. Appendix B presents a photograph of the anaerobic chamber. The liner from each core sample was cut open and details of the sediments such as grain size, sorting, etc. were recorded. Based on the observations, specific sampling depths within each core sample were identified for analyses. At each sampling depth, a total of four sediment samples were collected. One sample was placed in a plastic cup and the mass of the sediment was recorded. This sample was transferred from the anaerobic chamber to an oven where it was dried for 5 days at a temperature of approximately 7 0 ° C . Upon drying, the dry mass of the sediment was recorded and an accurate equivalent dry mass value for each sediment sample was calculated. Two sediment samples, were also collected from each sampling depth^ for SEP analysis. For each sample, approximately 0.4 grams equivalent dry mass of sediment was transferred into a H D P E 50mL centrifuge tube and the mass of the sediment was recorded. A fourth sample was collected in a sealed, plastic bag for thin sections and neutron activation analysis. Each sample was collected from the center of the core, away from any possible contamination effects associated with the core liner. Sequential extraction steps 1 through 5 outlined in Table 5.3 were then completed in the anaerobic chamber. A l l of the extractant solutions were prepared with distilled, de-ionised water (DIW) and the p H of each solution was adjusted with environmental grade hydrochloric acid (HC1) or sodium hydroxide (NaOH). Once prepared, the extractant solutions and DIW were placed in a nitrogen-filled glove bag and de-aired by bubbling nitrogen gas through each solution. The extractant solutions and DIW were then transferred directly into the anaerobic chamber. During each extraction step, the mass of each centrifuge tube was recorded prior to, and after, the addition of each extractant. Once a reagent was added to the centrifuge tubes, the tubes were sealed, removed from the anaerobic chamber, and transferred to a shaker table. Upon completion of the each extraction step, the samples were centrifuged at 3000 rpm for 15 minutes. The samples were then transferred back to the anaerobic chamber and the extractant solutions were decanted into dedicated 20mL syringes and filtered with 30mm 0.20um cellulose acetate 56 filters into H D P E sample bottles. The mass of each sample bottle was also recorded prior to, and after, the addition of the extractant. The extractant from extraction steps 1, 2, 4 and 5 were preserved with nitric acid (HNO3) to a p H of approximately 2. 5.2.2.2.2 Analyses of extractant solutions The SEP extractant solutions were analysed for concentrations of dissolved Fe and As. The extractant samples were analysed for As with a Perkin-Elmer 3300 Hydride Vapour Atomic Absorption Spectrophotometer at A L S Environmental analytical laboratory in Vancouver, B C . The Fe analyses were conducted with a H A C H ™ DR/2010 spectrophotometer, using the T P T Z method. 5.2.2.3 Analysis of total solid phase arsenic in sediments Select sediment samples were analysed for total solid-phase As by neutron activation analysis ( N A A ) at Becquerel Laboratories Inc. in Mississauga, O N . 5.2.2.4 Analysis of total sulphur in sediments Select sediment samples were analysed for total solid-phase S at U B C with a Carlo Erba NA-1500 N C S Elemental Analyser. 57 T3 CD X o CN •a 2 ft X ft (3 § '3 •_ c T3 c 3 O X -a xo c S x 2 S •a § o X T3 § O X S3 X ft e o 1 x o 43 >. x X o c s X >^ X o tu • U T3 c n 3 3 o o X ) X T3 c 3 O X T3 >> X! >, X o CD fc CD § .5 x —& _ MI hi 3 3 3 T3 c 3 o x o T3 (3 •53 3 60 O o T3 c 3 O X e o 13 o IS 3 O X 3 O -o >> X T3 C O X 3 S T3 >> X O <si x NO o" < O o CN SC ft u 60 © X o SC u C3 X X ft < o z T3 o WD s o o CN c U X X q X Z U O ON - < SC Z CN S 3 •5 c ft d CD "2 'x o CD K 60 «"> o + T3 >N XI M © 7 3 in — + I SC i X ^ s in N_J — 60 . CD + T3 I i NO , C? U o z _ SC o 2 SC CN O 60 CD T3 o CN s - ^ ^ IT) 60 CN CD TS ft o CN SC _. + t# Q ; O O W >-i ^ o o ^ z SC u SC o 3 ts o XI >N X ' SC 3 X u 60 T3 NO ON SC ft TD O O S3 C x" O SC U SC SC o o O oo SC u o SC o sc -x o X o CN O O SC o o o o u u cs _ _ . 3 X o s CN d c o CD <D in ^ « ON « X o 60 T3 X o SC z S S </-> m o o d d x NO O ft SC z £3 T3 X X ft 13 x o I SC Z N O O N O CO SC u X o o + 13 X o I SC Z oo in 60 T3 O o z SC o C3N ON O o CN O O (_ ' -a s PH ce ce c o 1 - J 3 « < < -g C Is 3 O O -o c '« 2 < ° a p. ca CT ."S o .£• G 8 > o o o T f T f T f T f T f o d d T3 J3 X O 3 O J 3 ' 5 T3 o o T t -* O T3 2 O O' CD. an < IS S3 b ^ c 0 -a 3 6 1 2 rt c c3 rt fi> 1 * I t T3 •s • c a S ca i-c a. a CJ VI O <U <U CJ CJ O c/i c/3 < < o o T f T f d d a. u o B cj .s < o o o T f T f T f T f d d I 3 o o ro cn ca o — -9 •< a •a 2 PH O g * •3 ^ c3 3 J3 ca o PL, X 33 ca . z z a o ro 3 T 3 o o >o CO ro ir> iri o d d B O X o ca 3 • T 3 a Cu T f CN *0 a. CN oo" X „ ^ O U 60 2 S o o tN c o cu CN co" D, o tN X tN O Tt 33 z o o u O T3 si CN n. c o 1 < H Q UJ £ J3 T f CN 5 <« O CN K o si ^ CN + CO O O z S O — - H ^ . 5 O cd U o rt z z 3 d o o CN ca o 2 t/3 U a. < X I CN g o . o + o u X ca Z s CN + ca Z d <u ^ 5 oo -a o J3 CN O X ro O o CN 4= 60 o ca Z z m CN O tN 00 60 •o 00 o Z -S CN CN O a z CN O O a"^ oo -• ° d 0 0 z s ^ CN C _o CD 60 •3 "H 'o ca •c IU crt i-S T 3 o 3 1 ^ 3 § I C w > § 3 ^ •S S g a 6a a o cfl 00 CJ o xi Cfl <U X) JS B CQ cn e o « cn > B .8 B cu a. 6D 1. a H co •a ixides 2 CO (3 co CO ixides X l U •silic onat o 1 hide X ) ulp O a ulp B CO o - and/i co with alysi _e - and/i ,B with alysi fixed - and/i _>> ed an fixed Fe ain CM M u >. c .o S assoc ole ro mainl fixed anic f fixed co X i CO co oo CO < <c <: o < CM X •a CM >> X <u x cn >. CM X •5 ° tt c co «S « S 7 3 O B X l CM c3 ,o cj tU 0 0 o o o CN CN in iri d d x; ft ft 3 T3 CO CD -d 'x o a s CO u oil l e g u > „ <C x i _ >, o X l u X I I x >> X o , u P-: s & 1 1 JO CO CM x -a is -a x x X I =2 — •a CM 0 0 3 O « T3 m -a CO co CD 3 § 3 "p. O f O o S co B t-ft B a. o ™ o o >, o o £ 3 13 >> ft cp oo (3 •a. S CO 1-1 XI c •3 ft •> o > o xl "J o •< w xi ' i j ft 3 u o < < a a & o O O O O O O O •* Tf Tt Tt Tt Tt Tt Tt Tt Tt Tt Tt Tt Tt o d d d o d o 3 O X & a ,? 3 - W 3 ' co £ :s —. _ g S CM -g c U Q X CO CO <* CQ X d o xj < £ B TS cu (A 6JQ B . "C cs cu Xi u CA 3 X) CU u o i . C H B O • P N * J U 2 W is B cu S O" <u t / ) CN v i 9) Xi CQ H a cu C ts 2 O Tt o o CN CM «S xi Tt o o CN CM to •a 3 r<1 O O CN a 1 oo % N ft u >_ — M la u S z x s s s CN o o CN CD CM B O o IB • < @ E 3 H3 o CO CM a XI # B X" CN en" ft CO 13 S 3 O 2 o K t £ o CO + CO o s o > N CD-. > a u 0 0 u xi o 1/-1 o CO U X in CN O + U X X o m CN d CO o o CN CM 13 cr CD a U oo CD XI O 0 0 3 X X I CM DC O CM z T3 XI E O CH 0 0 o Xi O c * M l CD -a • t_ o 5. ° CT co ft Pi to CO cs o P . O < - a O < o c o P-. o a. J3 a. 3 X <U T3 s • a c 3 - a CO CO o T t • i t d o x t • i t d o T t T t d o T t T t O o T t T t O • a c SC P . C •c 3 00 o -a c c o E « O " S a m" -2 l-i X . a. O a. tN SC ca Z oj • a. oj .c T t CM c P . E E o o «J o ca U .5 ~ Q s OJ Xj - a d c OJ + • a aci 3 nse aci ta nse pei •c ra X E o 5 OJ 1 ala o o c X i-c o o ;s E CN o. 3 .cn" p . '2 SC a o p . X i C-M o i a. SC CN ca d Z T 3 c • o o p . T3 C ca § o es 00 e o cj ca £ S -a • « 1 ^ ca o 2 •§ * -S c3 3 Pc O o 2 £ & > S 3 O X i & i "<3 S3 .& 2 CJ T3 U o X ! -O OJ X o 1 -3 .§< 2 cj •u " X o T3 o U • a i4 E 1 8 a 0-o ' 3 "3 g a 6.0 RESULTS AND DISCUSSION 6.1 Groundwater analyses 6.1.1 DND site As discussed in Chapter 4, groundwater samples were collected from two profiling locations, PI and P2, at the D N D site. Figure 4.1 shows the location of the D N D site, and Figure 4.2 shows the locations of profiles PI and P2. Table 6.1 presents the measured values for field parameters such as pH, electrical conductivity and flow rate, results of field analyses that include concentrations of ammonium (NFL"1") and dissolved inorganic carbon HCO3", and laboratory analytical results that include concentrations of anions and dissolved organic carbon (DOC). Table 6.2 presents the concentrations of dissolved cations that were analysed in the groundwater samples that were collected in profiles PI and P2. Alkalinity titrations were conducted in the field to determine the concentrations of dissolved inorganic carbon (HCO3) in the groundwater samples from profiles PI and P2. The concentrations of HCO3* that are presented in Table 6.2 were estimated from Gran titration plots. Example copies of the titration plots and the associated data are presented in Appendix D . 62 ,1 CJ r-Q. ft "~ 7. § 5 rt CO *~ oo 6 *? £ gj CM Q_ CO *" i » s TT e -2 I 5 9 5 CO 1- fvj 00 1^ o s § Q ^ co T - in 0_ co s I 5 3 9 - '-. r S « 0. co » Q Q. E CO CO co k_ a> CD E cs k_ ID Q. CU CO « ^ e § S5 n o to co m CD I s T _ m CM CD co O in LO LO (D N T-? !M O r- CM O to N ro o o "* 1- CM CD CD V i i W - J ;> o ~J T3 C o 5 j » o Q. LU LL o CD m CO Tt CO xf o i n 6 CO LO CD LO Tt o i n o co o CO LO cn LO o 1 «5 E o o JZ Q. o k_ +*• o u a. co u 'E cu ca < X < co X < o CM CM CO CM cn co CD D) E O o X 1 1 2> o. 1 g ro g> 2 o o to >. cs c < o o n cs cn ™ n 6 ^ v o V Tt CM LO O O V o V o V CM LO t-_ cri 0 V o V o V o V r~ i n co 0 D> O l O) O) E E E E O O O ™„ EL O 3 52. M . CD x: D a. CO c o XI i-w. o o 'E co O) O O E I LU CL a I a O = = - E r i E a> S S =1 z g 2 9- a, a| a. 0 s-x 2 E o O S> S I 3 a I ra E J= •? 5 CD £ 01 8 I CO o E 0 g- s- < i m E Q co * T- 3 O CO T> = o 5 6 S co 5 oj o <5 O 5 Si r 5 H E Q O C D X I O O < < X I sz sz I .1 1 1 ra Q c o ~ g g g g CD CD CD CD CD 13 TJ 0) CD CD E E CD CD E E !5 " CO 1 2 a> a CM a. •a c CO (0 J D o a. c U O a •o c co w c o c co w c o "«p (0 k_ c a> u c o u </f k_ a> • E CO CO a 0 c o o CO J D n co O ^ CV, CM C! <o, n co 9 * •* CM g m O. co CN § j n co II 8 S ^ CM 1— > «•> s- 1 -T- 5! Q. CO v § s ,1 N M Q. ?? "~ gj 2 Q_ CO e e 11 0) Q Q. E CO C7> (0 t_ (V E CO CO a. CD ii ^ Ol O m o o i n CD 00 i— CD N C\J co o O) o tO N ^ 3 I S to CO o> co _ 5 8 CD S i i > o "D C o O o ro 2 O CO n> o 5 X -5? .2 Q. LU LL O CD i n o CO a> CD CM CO CO c o c \ i CO o o o e g i n CM i n c o E E x 2 E a n P c o a (0 c a> co < T < CO X < O O X c o . Q CO o o 'c ca U) cu V) >. ro c < o o .n (0 CM N-v CM o o i -•<t d v O i r i d CM i-; co d V CM o V o V o V o V o V o V o V o V O) O) O) O) E E E E 1 CO o o 0 <% 1 O S ™ ro CD 5 CD Q . " O Z CL CO c o k_ rs O u 'E rs 6 O E x CL Q X Q O = = - E 3 S => z ® *= a> m o> cu Q_ Q> E 2-E rs o t3 Q O O £ .5 cn t3 o f X S g 5 £ o E « co -o £ •g a. S § c • i o x x o o < < X X CD X " D 5 .5 _ _ g g g "o TD "o "a "a CD CU CO CO CO 3 3 3 3 3 co co y) co co co ni oj co co CU CD CO Q CO E E E E E S '5 '5 _1 T3 "O E CD CO s , , o <£> 3 T S ° ! 2 S ~ T ? • 1 0 CM 0-oo 9 CM 0. CO o I CN 0. co CJ o CO S CN a CN S CM a. 55 " r j cn rt CN a. 2 S o • CM a ?5 ,~~ rt § „ CXI Sj ° n rt o s T? co a, Q 1 Q. E CO CO (A h. V V E to co OL OJ C O •<t L O r*~ c n o ° o ° ° ^ OJ 1^ L O <o Ol s t- 1-co £2 c o o 0 3 i - 0 0 to S O) s o Is- o c o t o N t-» ?3 <g C O C O r- DO O c o c o c n CO tO r-1^ i - o tO Tf Tf c o o o i -C M r- _ <o n 9 co cn c 'E CO 1-1 =1- E E o > CJ r j -a c o O o • ra 2 O CO ' l _ , » o J X -5? -2 Q. LU LL L O Tf L O cn L O C O C M C M Tf C M CO CO O C O L O C O C M OJ E OJ E E o O -C Q. O o CD Q. CO O 'E u o -< CO X < C O C O c n CM C O C M C M C M O cn L O to c o C M 'E, C \ J 0 0 C M OJ E O o i c o n to o g 'c ra OJ f £ o to CD CO >. CO c < o 6 co Tf OJ ^ V Tf O V C M C M t-Tf d C M o o o V C M C O r-c o d V C M O i -Tt' d V C M c o d V CM Tf T-Tf d V o V o V o V o V o V o V o V o V o V OJ OJ OJ OJ E E E E o o T-O Eb O CD co <D =- ™ 9i ~ CD Q . O t5 <B O J-f c o n k_ (0 o c CO D) < O Z l ffl o o E X L U Q_ D I Q o 2 3 5 z & 2 S" {? « •= « c c 3 CD CD Q. CD E E i & u o X ? - B 5 ci E O « » » r S n i-i Q. .-r §• b 6 2 o OJ CO OJ o 2 o J= o o gj o i l E £ 1.1 CD B g CD W © 3 "O >—' m CD . c ' c , CO CO to £ OJ X w ro — i : c o c c CO Q O O Q CM < < X X S o y = r\i i_ ^ I- "5= o ! u | Q 5 5 5 5 3 "O T3 -D T3 "O CD CD CD 0J CD I I I ! CD • CO ~ -I T3 U » l CD CD E E CD CD E E CO CO ^ CO 1 £ T - gj 2 § 3 CM CM r» S 9 CM °? Q. | 9 Si ^ Q. CO *~ 10 o g; to CO P ^ CO r- Q S 9 l T -Q_ CO 9 ^ co T- g; •* CL g| CM CO Q i I* 0) Q j . I I CS CO c O a 1 1 tv — T3 cn c g IS O CM TT V CM S ^ C O C M ^ ^ l o g o - ^ n C M C M i n W C N j W c ^ V V v CM V P £ S £ CM Jj * £ o S v ^ Tf „, m co m 41 CD q T-00 V V T- TJ T CO CM 05 V V CM IO CM CM CM CM V V V V £ j V V CM CM V V O m in co Kj --V v V CM 00 V in CD co "2 o ^ V in L n ^ O T - C T > C O C M C M L O C M C M C ^ C M C M C M C M S c M i 2 v l R - ^ r - - v v v v v v r : v v v v v v \ ; v ^ v CM CD ^ 00 CO « b O 00 CM 0> . CO V LO "* ' r- y r- " o n in co m S S ' - r W t M C M i n C M C M ^ R - T - O O V V V V V V CM CM CM CM S CM 2^ v v v v \i v TJ CM w v m CM CM v i*» 0 IJ 5 N CO v O CO 10 m CO in . CM S S ' - ^ C O C M C J i n i M C M ^ ^ ^ j t - T l - V V V V V V O TJ V 10 S o ^ n n w N i o w c \ | C O N C N w o i S i \ i S 2 ^ ^ J I - T ^ V V V V V V O S V V V V ^ V T J CM CM CM CM S CM J2 V V V V V TJ in ^> ,-m. to r in o co v ^ LO S o ^ - r - C O C M C M i n C M C M L O C M C M C M C M S ^ ^ j i - V V V V V V V V V V V V ^ 'v v CM £ v T: T f B i Q i Q ^ ^ L o c o i n o ^ i P C D C D c M O ' r ^ i n g o ' ^ ' ^ C O C M C M L O C M C M l O C M C M C M C M c ^ • j O I O I O O I r n D J D ) 5 E E E E 1 E E I E i i i i i i i i i i i i i i i l l — Z E O — .2 E E 3 E E o .2 ro < O o CD ~—' L L ro c -s 2 6 ~ o — CD .9 O 2 O co N o c o cu c <2 N < O E § s -5 r B E o c E ^ ? •& I - i -I o = co .E c CD ro O H < O m E O 1| ro co _J O P = ~a CD ? « m S £ E 3 .2 E c= in z> <B in co o l s = S 3 9 ^ CO M SS N a g 9 T? CM CN g co S 3 9 * CN g Q. g a> a E CO CO Lfj g 3 9 T? . CM 5j f CO 5 3 9 * ^ N 5 CM o. g 1- ^ s Q_ CO 8 si ' S o CM CM CJ> co" "~ CM CO ?5 *~ 10 c O ra o T- W CM «> • <0 £ CM 5 CM v CM V O m CD •n S S ' ^ ' - C O C M C M i n C M C M L r C M C M C M C M C D O V ^ V i - V V V V V V V V O CM V V V V CM CM V 00 °> c8 Tf L O C O o OS ^ « L O u C M O C O l t / • l - ^ V V V V V V O v V V V ^ " , ^ V 1— V V Tf m CM CD g CD V CO K| o ^ V CJ) C M LO . 00 v/ O C O LO Kc§'-'^ COCMCMLOCMCMLOCMCMCMCMC5cMi2 . — \/ y v V V V V V V V V V - v V C M 5 S J C M g 2 cn S § « V " > T ? ° 8 v d ^ v v 9 -• C O C M C M i n C M C M ' f c M C M C M C M S V V V V V V ^ j V V V V 1 ^ CM CM CD . CD V cri 5 cS t Tf CJ) Tf CM CO Q 3- v CM $ O i -tn , „ in co cn C D O ^ T - C O C M C M c n C M C M v ^ v •r- v v V V V V V V CO CM CM CM CM S CM !£? v v v v co v ^ T t < P S ^ 8 ° o c o ^ . C O ^ O i -L O c n O ^ ^ n w w i n w w ^ w w w c N S c j i ! ? < ^ ^ : • l - v v v v v v v J 5 V V V V ^ 5 v , ^ "v v V CO to . in CO L O r** f—i C M ^ ^ c n f f i o ^ ^ C T l C M C M c n C M C M C M C M C M C M S V ^ ; ^ - V < R - T - V V V V V V V ' - V V V V C O CM « v \ ; T- CM Tf co o ^ ^ V ^ C O ^ C M S ^ I O S O ^ ^ C O C M C M L O C M C M ^ C M C M C M C M S V ^ V ^ i n i n i n f e o fiio C M ^ i P C D C O c M o T ^ c n g o ^ ^ C O C M C M i n C M C M L n C M C M C M C M S c M ^ w CO O CO X *- _ ' CD m £ • I1 CD m 5 E t CD 3 c ^ £? o ™ CO 2 i l l • I . ! E E o o ^ CO JZ < O O 2 CD CO s CO LL ~ O) CO CD "O o CO l o — O C , = s , l r H i P l I c o g o . 2 o c i 2 o ^ c o g c c D ~ ' O Z O N < 5 C 0 O h < O CO a? • aT m E o E « E co ca cp m J o ^ xi CO CO _ CD CD S E F c § £ in z> >. i_ CO T3 I CM Q. W § „ V 'in • SI « Q. g £ 1 s D. ft » 3 s T <" ~< CM 2 Q. g CM S CM §J t n ft o _^ Tt Q. CO *~ o s • CM 0-CM £j ™ a ft (-1 g I 9 S CM S! Q. g l» S O f t CM 5 ° a. g co S 9 TT <o CM C| oo Q. g _ cd 4) Q Q. E CS c O j) 2 1 cu — •D (A C o ro O CM V T t T f CM V 10 CM TJ- CM LO LO co Tl- CM LO LO iri CO o T d V CO V 0> V CO O V V T - ^ V V V V V V V V V V V 8* io io 10 i - V ' S - ' ^ V V V V V V V \ , V -i- y CM ^ v s v " 5 w . CO v m 5S 2 T-V V V CO CM CM LO CM CM V V V V V V CM CM CM CM S CM V V V V V TJ CM CM CM CM S CM i2 V V V v V V TJ Tt CO CM • • V CM CO V Is- ~. ^ °> « CM CO I'J y-, CO V • - f _ T- y C\J O CM .n 10 S LO io o »-v <B • i - v v i - v v CO CM CM 10 CM CM ZJ-V V V V V V CO CM CM CM CM S CM i2 V V V V ^ V TJ 5 " s v LO ~ IO co LO S o ^ ^ n c M N i n N c g V ^ V T~ CO V V V V V XiCMCMCMCMScM^ v ^ v v v v ^ v T J W . Q O V 10 w 10 00 10 v LO CO CO Tf iri q d CM V  $ -^ W COCMCMLOCMCM2CUCMCUCM !2CMLO ^ I J I - ^ V V V V V V ^ V V V V ' ^ V T J CM]?. k q P c M c S ^ L o S g - T - c o w ^ v ^ ^ d ^ v ^ c g v ' R - T - v v v v v v v c O v v v v v v T J v cn "rf i - V V CM V ^ o 1 J 00 O O) • ^ LO O f IO C ^ O - ^ ^ C O C M C M L O C M C M ^ C M C M C M C M S C M ^ " y ' ^ j - ^ V V V V V V V ^ V V V V ' y ' v T J LO ro T— . CO . CM V Ol v CM Tt T-r~- ^ LO 1 . 00 V o t C o O ^ T T W C M C M l f l C M C N I C M C M ^ N ^ ' N V V t- V V V V V V V V T - V V V V ^ C M c M ^ L o f f i o ^ - ^ C O C M C M l O C M C M i n C M C M C M C M S c M L O o v ^ j T . v t R - ^ v v v v v v v v v v v v c 5 v T J ^ ° V TJ V Tt O) CM o l T-V CM . LO _ ^ <M O T" ^ od O CO lO K O ^ ^ W C M C M I O C M C M S C M C M C M C M S C M L O V A J . — y y y y y y V 0 " " " v/ V \/~ CD V V V V V i E 10 IO lO LO 1— LO o CO CO CM o 10 CO d CO d 1 _i _1 _ l 1^ _ l O) E E E =- E E 3. LO ^ l o o ^ ^ n w N i o c \ i w i n c \ i N c \ i ( A i g ( A i J 2 T - u ' co q r_ 2 S 2 S 2 S cu m " z E o_ !> cu m - 5 - - < - r o ° . ^ „ i l l I 3 E o 2 ro LL - _ ro 9> -a O at J 3 ^ J2. " ro E ^ > - " & .•15 § | E o H 5 « r » l o 0 . y o c : i » o ^ a j c C ( i ) 5 | w S < o o 2 i _ o z O N < 2 c o o i - < o m cu -— " ° — m w E 3 I E 3 CO CD _ J O P = -D ni rt ro . £ E B .3 E c _ 2 m i5 ni i °- E ts ... ai co co o I c to co .9 g E to I 2 HI if = ? ^ 2 .E 8 ? 1 . ° K — £ Detection limits were determined for the concentrations of anions, D O C , and cations that are presented in Tables 6.1 and 6.2. Sample P2-05 was independently prepared and analysed 5 times to estimate the variance between the measured concentrations. The variance term was then multiplied by a factor of 3 to provide a detection limit below which a reported concentration cannot be distinguished from the background noise associated with the instrument (Skoog 1985). However, this detection limit only includes the variability associated with the analytical instrument and sample preparation in the lab. Therefore, a duplicate sample for P2-05 was also collected in the field with a separate, dedicated syringe, filter and bottle. The duplicate sample was also independently prepared and anlysed 5 times. The variance within each duplicate, and between the two duplicate samples, was assessed to determine the highest, and thus most conservative, detection limit for a given species. Therefore, the detection limits reported in Tables 6.1 and 6.2, incorporate the variability that is associated with both sample collection and laboratory analyses. For example, the detection limit for concentrations of dissolved iron (Fe) in Table 6.2 is reported at 11.5 mg/L. This detection limit, which is high relative to the detection limits reported for other cations, includes a certain degree of variability detected between the two P2-05 duplicate samples. Although the resolution of the ICP-M S instrument is much higher than what is indicated by a detection limit of 11.5 mg/L, the more conservative estimate of variability between the two duplicate samples is provided in Table 6.2. Concentrations of cations were measured with an inductively coupled plasma - mass spectrometer (ICP-MS) using the semi-quant method. The semi-quant method involves analysing both a certified reference material that has been prepared to a known dilution, and a reference blank. The composition of the certified reference material is entered into the ICP-MS and the instrument correlates the specified concentrations of cations to the response of the element detector during analysis (Meuller 2004). A straight-line response curve between the blank and the certified reference material is then established for each of the cations specified in the reference material. Response curves for cations that are not specified in the reference material are estimated from the instrument response to the cations specified in the reference material. The reference material that was used to 69 calibrate the ICP-MS instrument for this investigation is T M D A - 5 4 . 3 . A copy of the Certificate for Certified Reference Materials is provided in Appendix E . Application of the semi-quant method assumes a linear relationship between the response of the detector in the ICP-MS and the concentration of cations in solution. A more accurate calibration method is to prepare a number of cation-specific standards at different dilutions. This detailed method would develop a non-linear relationship that more accurately characterises the relationship between instrument response and cation concentration. However, the accuracy of the ICP-MS when using the semi-quant method is still relatively high and appropriate for estimating overall patterns in groundwater profiles. When using an ICP-MS, the resolution of the instrument must also be specified as low, medium or high for each cation to be analysed. The ICP-MS detects cations as isotopes with a specific charge and mass. Depending upon the matrix of a sample, at low resolution it may not be possible for the instrument to distinguish between species that are similar in charge and mass. Specifying a higher resolution for certain isotopes provides greater separation between interfering species. However, when the ICP-MS is operated at higher resolution, the sensitivity of the instrument is significantly lower. Thus, although operating the ICP-MS at a higher resolution may minimise interference effects associated with certain cations, if the cations are present in the sample matrix at low concentrations, then the instrument may not be able to detect the signal. Therefore, there is a compromise between the sensitivity of the ICP-MS and the suite of cations that the instrument can accurately detect within a given sample matrix. Due to interference effects, the composition of reference material T M D A - 5 4 . 3 and the limitations associated with the semi-quant method, certain cations were not analysed with acceptable accuracy. In particular, concentrations of potassium (K) are not reported in Table 6.2. A l l other majors are reported in Tables 6.1 and 6.2. Charge balances of the analytical results were calculated for depths in profiles PI and P2 at which complete suites of analyses are reported. The average charge balance value is 4.2 %. However, there is some variability noted in the charge balance values, as a number of the values are 70 greater than 10 %. The results of the charge balances indicate that there are consistently more cations than anions in the groundwater samples, suggesting the presence of significant concentrations of organic acids. Although a charge balance value of less than 5 % is ideal, the results presented in Tables 6.1 and 6.2 are generally acceptable for the purpose of illustrating the vertical distribution of dissolved species in profiles PI and P2. Error analyses were also conducted for the analytical results presented in Tables 6.1 and 6.2. As discussed above, sample P2-05 and a duplicate sample that was collected with a separate, dedicated syringe, bottle and filter, were both independently prepared and analysed 5 times for concentrations of anions, D O C and cations. The standard deviation within, and between, each of the P2-05 samples was assessed. The most conservative standard deviation value associated with each parameter was then selected to represent the error associated with both the collection and analysis of that particular parameter. The error associated with field analyses, such as concentrations of ammonium (NFLi+) and bicarbonate (HCO3) , were estimated to be. a conservative 10 % (i.e. plus or minus 5 %). Estimations of the error that is associated with the analytical results are shown on Figures 6.1 through 6.3 and 6.6 through 6.8, which plot concentrations of select parameters with depth in profiles PI and P2, respectively. 6.1.1.1 Interpretation Groundwater samples from the top 4 sampling depths in both PI and P2 were analysed in the field for concentrations of dissolved oxygen (DO) with C H E M e t s ® K-7501 kits and concentrations of ferrous iron (Fe 2 +) with a H A C H ™ DR/2010 spectrophotometer, using the 1,10 Phenanthroline method. The results indicated that concentrations of D O in the groundwater were below detection limit, and the majority of the dissolved iron was present as ferrous iron (Fe 2 +). The results presented in Table 6.1 also indicate that the concentrations of nitrate (NO3) in samples from PI and P2 are generally low. The presence of iron in the reduced form as F e 2 + and the absence of D O and NO3" suggest that, in general, the groundwater at the D N D site is anoxic and relatively reduced. Concentrations of select cations with depth below ground surface, concentrations of dissolved organic carbon (DOC) and dissolved inorganic carbon ( H C O 3 ) with depth, and 71 concentrations of ammonium (NFL*4-) and As with depth, in profile PI , are presented in Figures 6.1, 6.2 and 6.3, respectively. The concentrations of dissolved As are less than 2 pg/L at depths shallower than 7.3 m and greater than 18 m, with a distinct zone of elevated As concentrations ranging from 7.9 to 9.1 ug/L present from a depth of approximately 10 m to 14 m. Concentrations of dissolved magnesium (Mg), calcium (Ca) and iron (Fe) exhibit a vertical pattern similar to that of As. Elevated concentrations of M g , C a and Fe in the range of 15-17 mg/L, 20-25 mg/L and 70-80 mg/L, respectively, are observed at the depths that exhibit relatively high concentrations of As. However, unlike the pattern observed for As, after decreasing in concentration between depths of 16 to 20 m, the concentrations of M g , Ca and Fe increase at depths greater than 20 m. Inspection of Figure 6.2 suggests that concentrations of D O C in profile PI also follow a pattern with depth similar to the one observed for concentrations of As. The concentrations of D O C range from generally less than 25 mg/L at depths less than 8 m and greater than 15 m, to as high as 40 mg/L to 50 mg/L in the zone between 8 m and 15 m. The concentrations of HCO3" generally increase with depth, ranging from 164 mg/L at a depth of 5.8 m to 287 mg/L at a depth of 22.5 m. A slight deviation from this pattern of increasing concentration with depth is observed in the zone from 8 m to 15 m, where slightly elevated concentrations of HCO3" coincide with the elevated concentrations of As. Figure 6.3 illustrates the relationships of N H / and As with depth. Concentrations of N H i + drop from 5.95 mg/L at a depth of approximately 5.8 m to 4.15 mg/L at 8.8 m below ground surface. Concentrations of N H i + are similar to those of As in that a zone of slightly elevated concentrations of N H / . is observed between 10 and 14 m, with concentrations of N H / decreasing to less than 3 mg/L below a depth of 17 m. 72 Figure 6.1: Concen t ra t i ons of ca t ions with depth in prof i le P1, DND si te concentration Na, Mg, Ca, Fe, Mn (mg/L) 0 10 20 30 40 50 60 70 80 90 0 i ' ' ' ' ' ' 1 1 1 73 Figure 6.2: Concentrations of DOC and HC0 3 " with depth in profile P1, DND site 10 concentration DOC (mg/L) 20 30 40 H h 50 60 50 100 150 200 concentration HC03" (mg/L) 250 300 350 • HC03- • DOC 74 Figure 6.3: Concentrations of As and N H 4 + with depth in profile P1, DND site 75 50 cone Fe (mg/L) 100 20 40 cone D O C (mg/L) 60 • • • 100 200 300 cone H C 0 3 ' (mg/L) 400 10 20 30 40 cone C a (mg/L) 50 8 -13 6 -CO < 0 4 -c 0 0 2 5 10 15 cone C l ' (mg/L) 20 2 4 6 cone N H 4 + (mg/L) Figure 6.4: The relationship of A s to other species in profile P 1 , D N D site; (a) A s to Fe; (b) A s to D O C ; (c) A s to H C 0 3 ' ; (d) A s to C a 2 + ; (e) A s to Cl" ; and (f) A s to N H 4 + . *note: the R 2 value presented in (d) excludes the three data points at depths below 21 m ( C a 2 + concentrations > 35 mg/L) 76 100 200 300 cone HC03" (mg/L) 400 60 i 100 200 300 cone HC03" (mg/L) 400 100 200 300 cone HCO3" (mg/L) 400 100 200 300 cone HCO3" (mg/L) 400 Figure 6.5: The relationship of H C 0 3 " to other species in profile P 1 , D N D site; (a) Fe to HCO3"; (b) C a 2 + to H C O 3 ; (c) M g 2 + to H C 0 3 ; and (d) D O C to H C 0 3 " . Figures 6.4 and 6.5 present the relationship of As to various other parameters, and the relationship of HCO3" to various other parameters, respectively, in profile PI. Arsenic exhibits a strong positive correlation with Fe (R 2 = 0.82) and, to a lessor degree, D O C (R 2 = 0.57). This is consistent with the hypothesis that the elevated concentrations of As are associated with the reductive dissolution of arsenical hydrous ferric oxide (HFO) minerals. At depths where D O C is present in high concentrations, H F O reduction is hypothesised to proceed, producing more Fe, HCO3" and the associated As.. However, recently published results from column experiments suggest that the reduction of H F O minerals to Fe(II) or mixed Fe(II/III) solid phases can release significant concentrations of As into solution (van Geen et al. 2004). Therefore, the dissolution of H F O minerals, and the associated release of Fe(II) into solution, does not appear to be necessary for the release of solid-phase As into solution. 77 Bicarbonate, which is also a by product of HFO mineral dissolution, appears to be poorly correlated to As, Fe and DOC. Bicarbonate is not a conservative species in solution, as it is affected by a number of geochemical processes, including mineral dissolution and precipitation. Saturation indices (SI) for select samples in PI were calculated with the geochemical code PHREEQC. The SI values that were calculated for relevant mineral phases for samples Pl-06, Pl-09, Pl-12, and Pl-15 are presented in Table 6.3. The negative SI values suggest that throughout the vertical profile PI , the groundwater is undersaturated with respect to calcite (CaCOa), dolomite [MgCa(C03)2] and magnesite (MgCOs), and these minerals are likely to dissolve into solution. The dissolution of calcite is supported by the significant positive correlation between C a 2 + and HCO3" observed in Figure 6.5b (R 2 = 0.69). The relatively poor correlation between M g 2 + and HCO3" suggests that if dolomite and magnesite are dissolving in solution in PI , this process is less dominant. The results in Table 6.3 also indicate that the groundwater in PI is saturated with respect to siderite (FeC03), suggesting that siderite will tend to precipitate out of solution throughout the profile. Table 6.3: Calculated SI values for relevant minerals in profile P1, DND site Sample ID d Saturation Index (SI) va lues a , b , c Calcite (CaC0 3) Dolomite [CaMg(C03)2] Magnesite (MgC03) Siderite (FeC03) P1-06 -1.36 -2.16 -1.75 1.10 P1-09 -1.17 -2.29 . -1.62 1.23 P1-12 -1.32 -2.50 -1.68 0.72 P1-15 -0.70 -1.36 -1.15 0.84 n o t e s : a saturation indices for selected minerals, calculated using P H R E E Q C geochemical code b negative SI values indicate that a solution is undersaturated with respect to a given mineral, suggesting that the mineral will tend to dissolve into solution - positive SI values indicate that the solution is saturated with respect to a given mineral and that mineral will tend to precipitate out of solution 0 reported saturation indices represent thermodynamic equilibrium values, and may be influenced by kinetic limitations d selected sampling depths in profile P1, DND site; samples P1-06 and P1-09 are located within distinct zone of elevated dissolved As concentrations In a reduced environment, it is expected that some sulphate (SO4 ") may be reduced to hydrogen sulphide (HS"). This is supported by the low SO4 2 " concentrations reported in 78 Table 6.1. In a solution containing Fe + and HS", sulphides such as mackinawite (FeS) would precipitate out of solution and, to some degree, sequester As out of solution. Therefore, the dissolution and precipitation of both CO3 2"- and Fe- bearing minerals may further explain the poor correlation between HCO3" and other species such as As, Fe and D O C . However, it must be noted that the SI values presented in Table 6.3 are equilibrium based, and the dissolution and precipitation rates of minerals are kinetically limited in most natural groundwater settings. When all the samples from profile PI are plotted, there is no significant trend observed between As and C a 2 + (Fig. 6.5d). However, when the three samples from depths of 21 m 2+ 2 and greater are omitted, the positive correlation between As and C a is significant (R = 0.57). This suggests that the dissolution of calcite, and hence the generation of C a 2 + and HCO3" in solution, is occurring in the vicinity of the elevated concentrations of As. As with profile PI , the groundwater in P2 is relatively reduced. Figure 6.6 presents the relationship of selected cations with depth in profile P2. Figures 6.7 and 6.8 present the relationships of D O C and HCO3" with depth, and N H / and As with depth, respectively, observed in profile P2. In comparison to profile PI , the concentrations of As in P2 are relatively lower, ranging from less than 1 ug/L throughout much of the profile to as high as 3.2 ug/L at a depth of approximately 17.2 m. Concentrations of M g 2 + , C a 2 + and Fe all follow a similar random pattern with depth, ranging in concentrations from approximately 4-29 mg/L, 8-51 mg/L and <12-76 mg/L, respectively. Although the patterns of M g 2 + , C a 2 + and Fe with depth observed in profile P2 are different from those observed in PI , the cations are present in the profile at similar concentrations. 79 Figure 6.6: Concentrations of cations with depth in profile P2, DND site concentration Na, Mg, C a , Fe, Mn (mg/L) 0 10 20 30 40 50 60 70 80 90 0 . , , , , , , , , j 8 0 Figure 6.7: Concentrations of DOC and HC0 3 " with depth in profile P2, DND site concentration DOC (mg/L) 0 10 20 30 40 50 60 0 -j 1 1 1 1 1 i 25 -I 1 1 1 1 1 1 1 0 50 100 150 200 250 300 350 concentration HC0 3 ' (mg/L) I • HC03- a DOC I 81 Figure 6.8: Concentrations of As and N H 4 + with depth in profile P2, DND site 82 10 10 • 50 cone Fe (mg/L) 100 rj co < o c o o • • 20 40 cone DOC (mg/L) 60 • ' • 100 200 300 cone HC0 3' (mg/L). 400 20 40 cone Ca 2 + (mg/L) 60 10 - i 8 - e IT 6 -CO < o 4 -c • o o 2 • 0 -4 - ^ . r »-50 cone CI" (mg/L) 100 1 2 3 cone NH 4 + (mg/L) Figure 6.9: The relationship of As to other species in profile P2, D N D site; (a) As to Fe; (b) As to D O C ; (c) As to H C 0 3 ; (d) As to C a 2 + ; (e) A s to CI"; and (f) A s to N H 4 + . 83 100 100 200 300 400 cone H C 0 3 ' (mg/L) 60 i 100 200 300 400 cone HCO3" (mg/L) 100 200 300 cone HCO3" (mg/L) 400 100 200 300 cone H C 0 3 " (mg/L) 400 Figure 6.10: The relationship of H C 0 3 ' to other species in profile P 2 , DND site; (a) Fe to HCO3'; (b) C a 2 + to HCCV; (c) Mg 2 + to H C 0 3 " ; and (d) DOC to H C 0 3 " . Concentrations of D O C in P2 range from 23-37 mg/L at depths of 5.4 to 15.7 m and 21.8 m, with concentrations in the range of 11-19 mg/L between depths of 17 and 20 m (Fig. 6.7). The distributions of D O C with depth in profiles PI and P2 are similar in both pattern and magnitude. However, in contrast to PI , the highest concentrations of As in P2 are in the vicinity of relatively low D O C concentrations. Profiles PI and P2 exhibit similar patterns of HCO3" with depth, at a similar range of concentrations. The concentrations of NFL"1" in profile P2 remains relatively constant with depth, ranging from 1.1 mg/L to 3.15 mg/L, which is lower than the range of approximately 3-6 mg/L observed in PI. Figures 6.9 and 6.10 present the relationship of As to various other parameters, and HCO3" to various other parameters, respectively, in profile P2. As discussed previously, 84 the concentrations of dissolved As remain quite low throughout profile P2. As such, it is difficult to interpret the correlation of As to other parameters. The results plotted in Figure 6.9 suggest that the correlation of As to Fe, and As to D O C in profile P2 are much weaker than the relationships observed in PI. However, Fe and HCO3" are present in profiles PI and P2 at similar concentrations, suggesting that the reductive dissolution of H F O minerals may be occurring at similar rates in both profiles. Nonetheless, the process is either not generating the same concentrations of dissolved As in P2 as in PI , or sequestration mechanisms are removing As from solution at a rate greater in PI than P2. Similar to the patterns observed in PI, the correlation of As to HCO3", Cl" and N H / in P2 is poor. Saturation indices (SI) for samples P2-08, P2-09, P2-13 and P2-14 were calculated with the P H R E E Q C code, using the minteq database. The results are presented in Table 6.4. As with profile PI , the results suggest that calcite, dolomite and magnesite are likely to dissolve into solution, and siderite will tend to precipitate out of solution in P2. However, the correlation between C a 2 + and HCO3" in P2 is much weaker than in PI , suggesting that the dissolution of calcite exerts less influence on concentrations of HCO3" in P2 than PI. The correlation between As and C a 2 + in P2 is also poor. 85 Table 6.4: Calculated SI values for relevant minerals in profile P2, DND site Sample ID d Saturation Index (SI) values a , b ' c Calcite (CaC03) Dolomite [CaMg(C03)2] Magnesite (MgC03) Siderite (FeC03) P2-08 -1.02 -2.23 -1.71 1.08 P2-09 -1.51 -3.05 -2.03 0.86 P2-13 -0.81 -1.76 -1.45 0.96 P2-14 -0.81 -1.76 -1.45 0.96 notes: • saturation indices for selected minerals, calculated using PHREEQC geochemical code b negative SI values indicate that a solution is undersaturated with respect to a given mineral, suggesting that the mineral will tend to dissolve into solution - positive SI values indicate that the solution is saturated with respect to a given mineral and that mineral will tend to precipitate out of solution c reported saturation indices represent thermodynamic equilibrium values, and may be influenced by kinetic limitations d selected sampling depths in profile P2, DND site 6.1.1.2 Arsenic field kit results Select water samples in profiles PI and P2 were analysed in the field for concentrations of dissolved As with a HACH™ 28000-88 test kit. The results of the field screenings are presented in Table 6.1, and the results of ICP-MS analyses are presented in Table 6.2. Analysis with the field kit method involves the oxidation of hydrogen sulphide (HS") within the sample to sulphate (SO4 2"), to prevent HS" interferences. Sulfamic acid and powdered zinc are then added to create the strong reducing conditions that reduce the As from any As compounds in the solution into arsine gas ( A S H 3 ) . ' The AsH 3 reacts with the mercury bromide on the test strip to form mercury halogenides, which causes a discolouration of the test strip that is proportional to the concentration of As. A colourimetric comparison of the test strip to a concentration chart is performed to estimate the concentration of As in the sample. When compared with results of the ICP-MS analyses, the concentrations of As reported with the field kit are consistently and significantly overestimated. The 28000-88 test kit lists a number substances, at specified concentrations, which are likely to interfere with the method. In particular, the presence of HS" at concentrations > 5 mg/L is identified as a problem when using the field kit. The groundwater in profiles PI and P2 is reduced, 86 suggesting that any aqueous S will be present as HS". Antimony (Sb), when present in solution at concentrations > 250 u,g/L, is also identified as a substance that will interfere with the operation of the field kit. However, as indicated in Table 6.2, the concentrations of Sb in the groundwater samples from profiles PI and P2 are below the detection limit of 2 Ug/L. Therefore, although test kits can be useful tools for screening samples for As concentrations in the field, researchers must be aware of the limitations associated with the test kit results. This is especially true for areas like Bangladesh, where test kits have been used to assess well water potability and guide important public policy decisions. 6.1.1.3 Arsenic speciation results As discussed in Chapter 5, four samples were collected from profiles PI and P2 for As speciation analyses. A comparison of the ICP-MS analytical results for the samples that were passed through ion-exchange tubes to remove As(V), and the split samples that were not speciated, is presented in Table 6.5. Results of the analyses show that there is no significant difference between the speciated and non-speciated samples in Pl-05. However, the total As reported for the speciated samples in both Pl-06 and Pl-09 are significantly lower than the corresponding non-speciated samples. This suggests that both As(III) and As(V) are present in the groundwater at certain depths within profile PI. Low concentrations of As were detected in sample P2-08. Thus, an assessment of possible As speciation in profile P2 is not possible. The analytical results in Table 6.5 also indicate that the concentrations of certain cations analysed in samples that were passed through an ion-exchange tube are higher than the concentrations in split samples that were not passed through an ion-exchange tube. In particular, concentrations of Na, M g , Ca, M n and Fe were generally higher in speciated samples, by up to 25 %. The concentrations of Cr were up to 40 % lower in speciated samples and concentrations of B a were up to 968 % higher in speciated samples. Therefore, this influence must be considered when interpreting the concentrations of other cations reported for a sample that has undergone speciation. 87 c o (0 '5 <D Q. (A (0 < O T3 C (0 (0 o. E to c o (0 o (A C o CO L . c 0) o c o u o c o .52 CO a E o o in cd 0) Si CO < co o • CM a. CO o I CM CL < C O o p co o s Si < CO o - SI 1 < LO o LO o a. E <° CO cu CD E CO h-, <° o. x- CO , CM i - i - 1 CM LO OJ CM T - O LO LO CO CO CM LO % V CO $2 V CO ^ LO CO d v w ° x V * CO LO LO CO d V v ^  ,j v v v v v CM CM V V O LO CO i -V V CO LO § ° - - V CO CM CM LO CM V V V V V CM CM Tf CO ••cl-l O O CO CM CM LO CM 2 CO V V V V V ™ CM O V CO c N S ^ c M C M ^ L 0 S g - - C 0 C M C M L 0 C M ^ C M C M O | ^ LO CD CD CO <M CM Tf CO CD CO CM V o <* 8 ® « ? ° CM C O C M C M L O C M S C M C M S!^ CM CD V LO CD r- t: o ^ V LO L O t S o ^ - T C O C M C M L O C M S c M C M S L O V u - x - c o v v v v vco y v v ^ / V T -V O O) CM T - CM CO CO CO CM . CM *-CM Tf ^ co L° LA CD . CM CO d 2 v d V ^ LO Tf .„ LO CD LO ti v co 9 £ CO V T— y CD CO CM CM IT) Tf V V V V CM CO Tf CM ^ CM CM 2 v g v v co -CM CJ) V LO CO d V CM CD CM O 8 LO o * -"*COCMC\ILOCM^CMCM o uo CO T -v v ~c5)"3)^ )^ )^ )"c3) D ) ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ ^ " C T " O ) " D ) " C 3 ) =L E E E E :t E E =s. s. E i i i i i i i i i i i n . O ) CD _ m to 5 -•z. E CO O I E CD c T T 2 <= 'E 0) O ^ — m co ^ E i f o 2 J3 co .c CO "o co X CD - — . w C ® CO LL D ) , „ -3 O ~ -a lo E Si w ~ E fS_5 < O O 2 ^ O Z O N < CO O f— > , S L I I C CO < m £1 = T 3 t CO CO CO f 3 5 < LU 0_ CO X < CO 6 _p t— .ra -- -5 o O 4) o U), CD — CD -a I % % & col CO § • ra s S H I 3 ' o Q) ca I ^ 3 5 CD < ~ CO i _ r -| 1 1 * ' ra ~ .2 ^ • •2- f o) 5 c= S- c » 4 §j ¥ - < £ £ o CD c i= X) a> _CD CL Q. <D E E> a CO CO '-G0 oo 0 0 6.1.2 K i d d 2 site One profile, P3, was advanced at the Kidd2 site. The location of profile P3 is shown in Figure 4.3. Tables 6.6 and 6.7 present the field and analytical results for samples collected from P3 and one sample that was collected from the riverbed of the Fraser River, immediately adjacent to the Kidd2 site. Results from previous investigations at the Kidd2 site confirm elevated concentrations of Cl" with depth (Neilson-Welch et al. 2001), suggesting that the site geochemistry is influenced by the presence of the underlying saline water wedge. Therefore, certain analytical procedures that were utilised for the D N D site samples were slightly modified for the Kidd2 site samples. In particular, due to anticipated interferences, and the limited ability of IC analyses in the lab to detect any anions in the presence of overwhelming concentrations of Cl", PO43" concentrations were estimated with a spectrophotometer, using the PhosVer3 method. Due to dilution effects, and the high concentrations of Cl", concentrations of SO42" were not detected with the IC. Concentrations of As were originally analysed with an ICP-MS, which uses argon (Ar) as a carrier gas. Sample matrixes that have high concentrations of Cl" are subject to the formation of A r C l ions, which have a similar charge and mass to As. Therefore, the ICP-M S results for the P3 samples overestimated the concentrations of As. To address this issue, the P3 samples were reanalysed for concentrations of As with a hydride vapour atomic absorption spectrophotometer ( H V A A S ) . The difference between the two analytical instruments was significant, as the values reported for concentrations of As reported with the H V A A S were approximately 30-40 % the values that were originally reported with ICP-MS analysis. The concentrations of As that are presented in Table 6.7 are from the H V A A S analyses. The methodologies for estimating laboratory detection limits, and sampling and analytical errors associated with the samples from profile P3 are the same as those outlined for profiles PI and P2. 89 0) w OJ TJ TJ CO Q. OJ o Q . o o Q TJ c CO CO c o ~E co o CO c o co •*-» c 0) u c o o w k_ V •«-» o E CO fc-(0 Q. TJ 0> CO CO 0 n CO H co o o CO o c6 CL s o O a> co p y~ a. o CO s o CO o d CL <<t o CO o 9 oo co o od Cu o TT o O CO CO o 0. o co o o CO CO o ui a. o CM s o CO CO p TT" 0. o o © h-CO p cvi a. o a o ro its a> a its a c E De 3 ca w i_ a •*•* i ro CO a 14 a> cu E CO CO a, cu CD • LO ,x ZZ r-o cn ^ CO CO CO CM CM S o CO CM -5-CD CO CM 0> o CO t - o CD CO CM co cn o co LO CD CM CM E E '> o zi •a c o « £ O CO « u 3 X -52 2 a tu i t LO h- LO LO L". 5 2 CM LO <N O CO CM LO LO f-~ 00 O h~ CO CO CO LO CM O CO cn oo o LO Tt LO E cu 5 E ~ Q. 3 J= o CU g CO < cu ro sz Q . CO O sz a. o c a V) O . < O O .2 C I I •S « *.S I g co E? J J - O V) cu (0 >. ca c < o c o Tf "O T3 CO C C CD CO s -g -g LO c n O) o D) D) D) E E E o o S cu 5- 3 2 cu co v - - ~ SZ o ro _ Q . c o .a i_ ca O u "E ro ro. < O Z W O O LU 0. a i Q o E G 3 a; T T E co 5 5 Z 0-cn cn E E E X I i $ $ a a ! K c S : O o i co i O CD " D T D T3 T3 " O U CO CO CD CO CO co CO CO E E CO CO E E o CO 8 £ s r i= ^ O a) T T ™ SI 2. ° u X p c o *y o | l_L U . J> o o co -o T- i - 03 ]) S - ° S — o I I I u u o < < < X X J 3 5 2 2 5 S o E 8 2 i I X Li. O ° ° s » i s X C S ro ^3 * 1 I E m _ - TJ (0 LU <S> Z JZ V> J5 E OJ 03 0) CO CM TJ TJ CO OL Q) O mm a. c O O Q TJ C 10 (0 co CO c o co L a •4-* c CD o c o o (0 k. V 0) E CO fe-re a c o u CD CD X i CO H °- 3 2 S Q 2 p » Q 1 Q. E CO co o o co CD ^ C D r- oo co CD _ in .• O s in C D L O Tf C M L O " ? C N -C O L O x-L O C D o ^ C O L O -C O Tf ^ CO . o C P t j W CO 1 co __i co mm CJ o E cs 1_ cn a. a LL T J C o CO o C O Tf Q. LU C O L O C O co c? co Tf . J O) CD E E 0 u <D E x o Z l i i : o cu E a. E co < o L O O C M o X CM r; C O _ ; C O 05 O i -co C M O CO CO C O co -a "o co c c O) T3 TJ T - C C CJ) CD E o C O CO co eg CD E L O S -o "a CM Tf T J T J CJ) C C C O O) CD CD E E E o "E a> ( A 0 _ < CO 1 < o CO o 'c CO CD < £ ( A CD (0 >. co c < o *m (0 o CO 8 o o 52-CD 5 CD O JO ( A C g < o z c o JO mm CO _ O « o c o "c _2, co O O LU CL a I o O 2 2 S £ " D O ™ m 5 3 C C 9 ° 5 '5 'S « S fl) S E ^ 2 E E o o c5 o E — o Q O O g 5 3 S o m Q D-E x !2 t l o s 2 o i o o go -o .— .— CO Q) 8 £ E TZ CM ra — •fc o o x Q E 8 .1 I I I o o o < < < I 1= 2 £ $ 2 £ £ ^ I I *Lt co 5 3 CD CD E E CO CO « ra ca CD E E 1 9 f 1 w ra | g CD •*-« » CM TJ TJ CO CL CD O L a Q. C (0 C o co U H— o CO c o C O c 0) o c o o CD 9) n CO 00 o I CO Q. o • co a o CO S ° £ co S 0- •* o co S ° 5 co S 00 o • CO CL 8 I ° 5 co S m CM O i CO a. 00 CO a OJ a CS CO CD E * - • — cu — T 3 CM V CM V CM V CM V CM V O L". CJ) ^ IT) CD v CO V CO ^ ^ C O ^ p ^ ^ C O C N C M L O C M C M O CO V , -V o CO V V V V V . _ * CM CM CM CM V ™ V V V V LO , t 0 0 CD O CD v LO LO ^ S S C O S O ^ ^ C O C ^ C V I L O C N J C M ^ C M C M C M C M S C M ^ y O C O y C D V T - O v V V V V V ^ V V V v ^ v ^ P CM CD CD 00 v t V CO 1 . u ; x CM CM CO LO M V CO " O) i - in CO ^ LO CO c - j i n c n ^ c o S o ^ ^ ' C O C M C M L O C M C M ^ C v l C v J C N C v l H c M i f i ^ C O v ^ ^ j - ^ ^ V V V V V V g V V V V ^ V ^ CM °°. LO O LO T - in CO ^ CM CD r~i i n ^ L o K o ^ ^ C O C M W L O C M C M ^ ' C M C M C M C M S c M b i i LO CO CO o CO V CO CD ^ O ^ C M C 0 C M C M L n C M C M 2 0 U C U C J 0 J C 0 C > J ^ ° I S" , N l ^ . • LO tr\ O *~ ^ ^ ^ ^ u > ^ ^ ~C ^ ^ ^ ^ m C M _ J ^ ! V 1 ^ ^ T - V V V V V V V 0 O V V V V \ ; V V v CO q c q C O c M o ^ c o S o ^ C > J C O C V I C g L O C \ l C M L O < N C M C M C M S CO O CO r-i i - ' *~ "oi CT D> gi g>"a) o) ra'a)^) 2>"O) CJO ^  ^ ~3> ^ > l5> ^ > a> oi o> a> u> rad) o> I E E E E 3 - E E =5-=S-E i i i i i i i i i i i i i i i i e s O CD „ CQ CO 3 E Q) O CO CO I E CO u CD C S I CO 3 2 < ^ E I I !y 2 CO sz O O CD CO £ aT co u-oo o >^ CO I I o .a g o O . £ O N « E a < 2 C 0 O h < y . c Q .11 co ~ CD co o m E O § E 5 5 E i Z} CO CO £ E =j .2 £ c « * in z> | B E F +- co CO < < > I 1: > ^ o « -° B c o o "S co S " o co co oj ro .o a <2 E > I S 5 ! ? CD W CM T J T J CO Q. JD «^  o Q . C CO c o CO o CO c g to &_ *-* c CD O c o o c o u r» CD 0) J3 C O V r3 10 o co V ft co g a- S CO a. CO a T f ^ o V ft co S CL S CO CL CO o co a. g i 1 <0 a sz Or S E Q CO CO c 0 u 1 I cu — •a i n 9 o i ^ N O u ) j ; o 9 o i o 9 9 o o 9 S o . § 8 - -<M V CM V co LQ o _. co co C M . LO ,„ LO co / x ^ C M C O ^ L O j y o ' - COCMCMLOCMCMiiCM CM CM CM S, CM S2 V V V C J V V CMCM2CMCMCMCM9CM!£? v v ^ v v v v c o v 1 ^ CM V CM V Cn m Cn • Cn . . CD ^ — OQ y y y y y y CJJ x / v \ / \ / V J v T T cn cd oi v CD y LO . • O " i i O X ^ T - V V co V CM V V V V S ^ ^ C M C I ) ^ L O S § ^ ^ C O C M C M L O C M C M ^ C M C M C M C M 2 C M ^ V O C D C O ^ ^ ^ i n g g ^ c ^ j c v j C M C M L O C M C M L O C M C M C M C M S i C M ^ C O O C O Q I - ^ ' -~D) D) O) D) D>~o> D) D) CDCJ) ™ C B C » C B " c b ^ ) " c i ) ^ ) ^ > CJ) C» CJ) CJ) CJ) CJ) CJ) 0)| CD „ m co CO c o (0 O .2 E ! • ! CD O m co - '< I E SS ? E o P CO E CB O CD S ! CO 3 -> < l3 2 CO -C O O CD  DCO LL O) 5 C CO n O o-P . Z 3 3 o .9 O Z co „ < c ^ ~ IM o c O CD C W N < - a - t-O = CO E c CD CO 2 co o i- < o m I E i.I o = E CO TJ ==- CO '{/) Q. CM O E co > X E -S CD 1 0 CD ra E 5 E ~° p 5 °> * 2 c .E •o o o „ f a i l 8 s 6.1.2.1 Interpretation Figure 6.11 presents the concentrations of select cations with depth in profile P3. Concentrations of dissolved organic carbon (DOC) and bicarbonate ( H C O 3 ) with depth, concentrations of ammonium (NFI/) , phosphate (PO3) and As with depth, and concentrations of chloride (CI") and As with depth, are presented in Figures 6.12 through 6.14, respectively. Similar to the D N D site, the groundwater at the Kidd2 site is anoxic and relatively reduced. Concentrations of As are in the range of 3 pg/L to a depth of approximately 9 m, increasing to concentrations that range from 27 to 32 pg /L in the zone between 12 to 21 m. The pattern of As concentrations with depth in P3 is distinct from those observed in profiles PI and P2 at the D N D site. In PI, a distinct zone of elevated As concentrations is observed between depths of approximately 8 to 16 m, whereas in P3, concentrations of As are relatively high from a depth of approximately 12 m to the bottom of the profile. The concentrations of As in P3 are also up to 3 times greater than those observed in PI or P2. Concentrations of C a 2 + and Fe in P3 also follow a pattern with depth similar to that observed with As, as the two cations are present at concentrations of approximately 500-800 mg/L and 120-230 mg/L, respectively, in the zone from 12 to 21 m. Concentrations of M n 2 + and M g 2 + also follow a similar pattern with depth, .but at significantly lower concentrations, in the range of 10 mg/L and less than 1 mg/L, respectively. 94 Figure 6.11: Concentrations of cations with depth in profile P3, Kidd2 site concentration Mg, Ca, Mn, Fe (mg/L) 0 100 200 300 400 500 600 700 800 900 0 H 1 1 1 1 1 1 1 1 1 25 1 1 1 1 1 1 1 1 0 5 10 15 20 25 30 35 40 concentration As (ug/L) Mg ca — • — Fe Mn - « - As 9 5 Figure 6.12: Concentrations of DOC and HC0 3 " with depth in profile P3, Kidd2 site 4 concentration DOC (mg/L) 8 10 12 H h 100 200 300 400 concentration HC03' (mg/L) 500 600 HC03- • DOC 96 Figure 6.13: Concentrations of As, N H 4 \ and P0 4 3 " with depth in profile P3, Kidd2 site 97 Figure 6.14: Concentrations of As and Cl ' with depth in profile P3, Kidd2 site concentration As (ug/L) 15 20 6000 8000 concentration Cl" (mg/L) 10000 12000 14000 Cl- '- As 98 cone Fe (mg/L) cone DOC (mg/L) 200 400 cone HC03" (mg/L) 600 500 1000 cone Ca + (mg/L) 0 5000 10000 15000 u 5 l u cone Cl- (mg/L) cone NH 4 + (mg/L) Figure 6.15: The relationship of A s to other species in profile P 3 , Kidd2 site; (a) A s to Fe; (b) A s to D O C ; (c) A s to H C 0 3 " ; (d) A s to C a 2 + ; (e) A s to Cl " ; and (f) A s to N H 4 + . 99 250 2T 200 £ 150 t 100-I 8 50 H = 0.55 200 400 cone HC03" (mg/L) 600 1000 •"800 — 600 + CM CO O400 o c • b R2 = 0.78 \ • X t 200 400 cone HCO3' (mg/L) 600 0 200 400 600 0 200 400 600 cone HCO3" (mg/L) cone HC0 3' (mg/L) Figure 6.16: The relationship of H C 0 3 ' to other species in profile P 3 , Kidd2 site; (a) Fe to HCO3"; (b) C a 2 + to HCO3"; (c) M g 2 + to H C 0 3 ' ; and (d) D O C to H C 0 3 . Concentrations of D O C decrease with depth, from a concentration of 17.5 mg/L at 7.3 m below ground surface to less than 0.5 mg/L at a final depth of 21.0 m. Concentrations of HCO3" also generally decrease with depth from 541 mg/L to 213 mg/L, with a deviation from this pattern occurring in the zone from 13.4 to 19.5 m, where concentrations of HCO3" are slightly elevated. The D O C and HCO3" patterns observed in P3 are different from those observed in profiles PI and P2 at the D N D site. In particular, the concentrations of D O C on P3 are lower than the concentrations observed in PI and P2, while the concentrations of HCO3" tend to be higher in P3 than PI and P2. Concentrations of D O C and HCO3" with depth are generally inverse to concentrations of As in P3, which is generally opposite to the patterns observed at the D N D site. 100 Concentrations of ML,"*" generally follow the same pattern with depth as As in profile P3, ranging from 2.6 mg/L at a depth of 7.3 m to as high as 17.0 mg/L at a depth of 19.5 m. The concentrations of N l V " in P3 are significantly higher than those observed in PI and P2, suggesting that relatively more organic matter is undergoing decomposition in the subsurface at the Kidd2 site. Concentrations of PO43" in P3 are generally in the range of 2.5 mg/L or lower, showing no particular pattern with depth. Figure 6.14 presents a plot of Cl" and As concentrations with depth in P3. Because Cl" is a conservative species, it is a good indicator of saline water intrusion. The distribution of Cl" with depth follows a similar pattern to As, but at much higher concentrations that range from less than 120 mg/L at a depth of 10.4 m, to as high as 11695 mg/L between 11.9 m and 21 m.. This suggests that groundwater in P3 to a depth of 10.4 m is generally fresh, meteoric water, and the groundwater below 10.4 m is influenced by porewater mixing processes between meteoric water and invading saline water. The maximum concentrations of Cl" at the Kidd2 site are over two orders of magnitude higher than those observed at the D N D site. Figures 16.15 and 16.16 present the relationships of As to various other species and HCO3" to various other species, respectively, in profile P3. Arsenic exhibits a relatively strong positive correlation with Fe (R 2 = 0.73) and a strong negative correlation to both D O C (R 2 = 0.91) and H C 0 3 " (R 2 = 0.86), indicating that, unlike PI and P2, elevated concentrations of As and Fe in P3 are located in zones that are relatively low in D O C and HCO3". Concentrations of Fe are present in profiles P3 and PI at maximum concentrations of 229 mg/L and 81 mg/L, respectively. This suggests that H F O minerals are not just undergoing the reduction of Fe(III) to Fe(II) or mixed Fe(II/III) solid phases, but rather complete reductive dissolution. Although HCO3" decreases with depth in P3, the concentrations in the vicinity of the elevated As are similar to those observed in PI. This suggests that, although concentrations of D O C are significantly lower in P3 than those observed in PI and P2, the rate at which the reduction of H F O minerals is proceeding may be greater in the subsurface at the Kidd2 site than the D N D site. This is supported by the positive 101 correlation between both As and Fe with NH/, which is present in deep P3 porewaters at concentrations approximately 2.5 times greater than those observed in PI . Alternatively, the reduction of FfFO minerals may be occurring at an area that is hydraulically upgradient, and the dissolved species may have been transported to P3. As previously discussed, the relationship between As and C T is strong and positive, suggesting that significant concentrations of As are present in P3 at a depth where intruding saline water is present. Figure 6.17 presents a cross-section that illustrates the general pattern of groundwater flow at the Kidd2 site. At high tide, dense saline water infiltrates down from the bed of the Fraser River into the sand units that underlie the Kidd2 site. The dense saline water migrates inland along the top of the basal silty clay unit, circulates as it mixes with fresh meteoric water and discharges back to the Fraser River along the transition zone. approximate location 0 50 TOCim v Horizontal Scald <m) -' • . Figure 6.17: Cross-sect ion illustrating conceptual model of groundwater flow at the Kidd2 site, source: Nei lson-Welch (1999) The results of an investigation of the hyporheic zone along the Fraser River indicate that during the winter months, the saline water along the bed of the river and adjacent to the Kidd2 site, contains dissolved oxygen (DO) at concentrations estimated at 5-8 mg/L (Bianchin 2004). The results presented in Tables 6.6 and 6.7 indicate that the deep Fraser River water is also characterised as containing relatively high concentrations of CT, NO3" 102 0 9-4-and SO4 ", and relatively low concentrations of HCO3", C a and Fe. As intruding saline water flows into the deep, reduced zone of the aquifer unit at the Kidd2 site, in the presence of D O C oxidation, SO42" would tend to reduce to HS". In the presence of the high Fe concentrations in the porewater at the Kidd2 site, the precipitation of monosulphides, a process that sequesters dissolved As to some degree, is thermodynamically favourable. However, Neilson-Welch (1999) reports S0 4 2 " within the deep aquifer unit at the Kidd2 site at concentrations of approximately 1000 mg/L, suggesting kinetic limitations to the reduction of SO42" at the site. Arsenic was not an analyte of concern for the investigation from which the Fraser River water sample analyses were obtained. As such, the reported detection limit of 200 ug/L does not provide the precision necessary to compare concentrations of As in the Fraser River sample to those obtained from P3. However, Cameron et al. (1995) report concentrations of less than 1 ug/L dissolved As in Fraser River water from 17 sampling locations. The natural concentration of As in seawater is also low, averaging less than 2 L i g / L (Ahmed 2003). In the oxygenated and relatively saline waters along the base of the Fraser River, H F O minerals would tend to flocculate out of solution and, in the process, sequester As. Therefore, significant concentrations of As are not expected in the saline water along the bed of the Fraser River. The concentrations of D O C are relatively low at depths greater than 11 m in profile P3, whereas concentrations of Fe and As generally increase with depth. This suggests that either the D O C is more efficiently oxidised at depth, or perhaps the various dissolved species have undergone chromatographic separation during transport processes. Mixing processes between relatively DOC-rich meteoric water and DOC-deficient saline water could influence the distribution of D O C in profile P3. Therefore, the source of the dissolved As could be located hydraulically upgradient, and the distribution of As with depth may correspond to the pattern of groundwater flow. This is possible, as the retardation factor for arsenic can be as low as two (Kinniburgh et al. 2003). Neilson-Welch (1999) estimates the horizontal groundwater flow velocity within the saline wedge to range between 0.5 to 6.3 m/y in the central portion of the wedge to 103 approximately 0.4 to 1.4 m/y at the toe of the wedge. Inspection of the flow paths illustrated in Figure 6.17 suggests that saline water at the base of the Fraser infiltrates into the adjacent aquifer unit and travels as far as approximately 600 meters inland, through the deep sediments. The saline water mixes with meteoric water and this relatively less dense saline water flows back along a transition zone, discharging back into the Fraser River. Assuming groundwater velocities of 3 m/y in the central portion of the wedge and 1 m/y at the toe of the wedge, the residence time for groundwater within the saline wedge is approximately 530 years. However, the groundwater velocities published by Neilson-Welch (1999) are average values that do not account for the porewater mixing that occurs due to tidal processes. Appendix F contains sample equations for the calculation of porewater residence times, and the solid-phase source of the dissolved As is discussed in more detail in Section 6.2. Table 6.8 presents saturation indices that were calculated with the P H R E E Q C geochemical code, using the minteq database, for relevant mineral phases in select samples from profile P3 and the sample that was collected along the riverbed of the Fraser River. The SI values suggest that calcite, magnesite and dolomite will tend to dissolve into solution in the shallow and intermediate depths of P3. However, SI values for these minerals are closer to zero in P3 than those observed in PI and P2, suggesting that the relatively deep porewaters at the Kidd2 site are close to equilibrium with respect to these mineral phases. The positive Slvalues for siderite are greater in P3 than PI , suggesting that the system is further from equilibrium with respect to siderite in P3 and the mineral will tend to precipitate out of solution at depth, possibly at a rate greater than at the D N D site. The relationship of D O C to HCO3" is positive, suggesting that the oxidation of D O C is a significant source of HCO3" to the porewater in P3, and the negative correlation between C a 2 + and HCO3" suggests that calcite is a less significant source. The correlation between Fe and HCO3" is relatively weak, supporting the hypothesis that each species is influenced by geochemical processes that are independent of H F O mineral reduction. 104 The SI values that were calculated for the Fraser River sample and presented in Table 6.8 indicate that the saline water is undersaturated with respect to calcite, magnesite and siderite, and oversaturated with respect to dolomite. Table 6.8: Calculated SI values for relevant minerals in profile P3, Kidd2 site Sample ID d Saturation Index (SI) values a b ' c Calcite (CaC03) Dolomite [CaMg(C03)2] Magnesite (MgC03) Siderite (FeC03) P3-05 -0.97 -1.66 -1.18 1.12 P3-08 -0.22 -0.16 -0.43 1.02 P3-10 -0.21 -0.13 -0.42 1.13 P3-13 -0.06 0.06 -0.38 1.24 Fraser River -0.23 0.24 -0.03 -1.14 notes:" saturation indices for selected minerals, calculated using PHREEQC geochemical code b negative SI values indicate that a solution is undersaturated with respect to a given mineral, suggesting that the mineral will tend to dissolve into solution - positive SI values indicate that the solution is saturated with respect to a given mineral and that mineral will tend to precipitate out of solution 0 reported saturation indices represent thermodynamic equilibrium values, and may be influenced by kinetic limitations d selected sampling depths in profile P3, Kidd2 site 6.2 Sediment analyses Sediment cores were collected from specific depths at locations C I and C3, which are approximately 2 m adjacent to profiles PI at the D N D site and P3 at the Kidd2 site, respectively. Two sediment cores were collected at depths of 25-29.5 feet below ground surface (7.6-9.0 m) and 40-43.8 feet (12.2-13.3 m) at coring location C I , and three sediment cores were collected at depths of 25-26.5 feet (7.6-8.1 m), 40-43 feet (12.2-13.1 m) and 50-53 feet (15.2-16.2 m) at coring location C3. Coring locations C I and C3 are presented in Figures 4.2 and 4.3. Two polished thin sections from the sediment cores were prepared for mineralogical interpretation. The sediments for thin section CI-41 were collected from the core sample that was retrieved from coring location C I , at a depth of approximately 41 feet (12.5 m), and the sediments for thin section C3-41 were collected from the core sample from C3, 105 also at a depth of approximately 41 feet. A sequential extraction procedure (SEP) was performed on sediments collected from each of the core samples that were collected at coring locations C l and C3, to permit the quantification of the various pools of solid-phase As in the sediments. 6.2.1 Mineralogical analyses A mineralogical inspection of thin sections Cl-41 and C3-41 was performed with a petrographic microscope. Figures 6.17 and 6.18 present photomicrographs of thin sections Cl-41 and C3-41 at different magnifications and under both polarised and cross-polarised light. Inspection of thin section Cl-41 indicates that the sediment is texturally immature, as the majority of the grains are sub-angular. Thin section Cl -41 is composed predominantly of lithic fragments and quartz grains, with minor amounts of feldspars and muscovite mica and a trace of opaque minerals, suggesting that the sediment is also chemically immature. The sediment grains in thin section C3-41 are slightly coarser than those observed in thin section Cl -41 . Otherwise, thin section C3-41 is similar to Cl -41 in both textural and chemical maturity. 6.2.2 Solid phase analyses Table 6.9 presents the results of the solid phase analyses for As, Fe and S in the sediments from the core samples that were collected at locations C l and C3. The sequential extraction procedure (SEP) that was performed on the sediment samples was specifically designed for assessing the different pools of solid-phase As. Extractant solutions were analysed with a hydride vapour atomic absorption spectrophotometer ( H V A A S ) , which is an analytical technique that is particularly effective at measuring As concentrations in solution. Each sediment sample was split into duplicate samples for SEP analysis. Both splits from samples Cl-41 and C3-52 were submitted for H V A A S analysis to assess the variability associated with the method and analyses. Review of the data presented in Table 6.9 indicates that target phases of As present at concentrations greater than 0.3 U-g/g show variability in the range of 20 % or less. Although this 106 variability is significant, it represents not only the error associated with the SEP methodology and analyses, but also the heterogeneity within the sediments. 107 0 i ft © • i t 1 >^E tc « cd <o E >-U i >- E E OJ o •D T3 £ § 1 " 2 0 co" to .c c 5 5 CO *-— CO o — TJ co c .c — CL CO to co cn m 2 § o *-o ->»' 0) ;5 CT N t t CO cr "D C ro c CD E CD ro o ro c Q3 C CO c to CO "co 0 3 r-> co Z I "k CD r " 5 § O OJ o o o = ro C L  o CD o ro OJ 1=h.-= CO J ? £ -a SZ OJ — CO c CO 4— c CD E to CO S -a CO c o to to co co TJ Q) CJ CO i_ OJ 0 TJ to CO I s .2 TJ 1 « 03 T -CO OJ i_ 3 CO CD T3 C ro CO > o o C O E •a -° to c ro co CD M c 3 Q. CT CO C TJ _LJ H — to o fc4 • ^ 1 -d;*. tMTn;, r-fc fcV - -* *• ? ^ CD OJ TJ TJ CO O c o -i—" o CD CO c CD E TJ CD CO sz CL ro CT TJ CD .co o -° C L T _ " cb co 1 § o !_ C OJ CO TJ •£ c *^ d T -CO t "~ CO g o i l CO O *- OJ x: co •M £ TJ £ CO .E _co tu O co C L cO fc o OJ O TJ ^ CO CT -=> W 3, OJ ro to Si CD o o E o o CT CD j . i E "§ o to O 0) sz sz C L h-_ 2 ro E ro o E CD TJ C ro ro 3 >< CD O CO E CD CD O cn T— CD OJ 1-3 D) L oo o The SEP step 2 extractant samples consist of a 1 M N a H 2 P 0 4 3 " solution matrix. When these samples were analysed by H V A A S , interferences were encountered. In particular, standards were prepared in the laboratory by adding specific concentrations of As to volumes of blank NaH 2 P04 3 " solution. When these standards were analysed, the results indicated significantly higher concentrations of dissolved As. Inspection of the results indicated that such low concentrations of As in the SEP step 2 extractant samples, when compared to a relatively high background signal strength that is not stable, could not be reliably calculated. Therefore, the results of the SEP step 2 extractant samples are not reported in Table 6.9. Rather, the reported residual solid-phase As includes both the concentrations of recalcitrant mineral phases and the concentrations of strongly adsorbed As. Split sediment samples that had not undergone SEP analysis were analysed for concentrations of total solid-phase As by neutron activation analysis (NAA) . The N A A method involves the irradiation of the sample in a reactor. Each element within the sample becomes radioactive when the bombarding neutrons interact with the nuclei of the element, forming a radioactive nucleus (radioisotope). The radioisotope will de-excite into a more stable configuration through the emission of gamma rays. The energy signature (i.e. half-life) of the emitted gamma rays is unique for each element, and the intensity of each energy signature is proportional to the concentration of that element within the sample. The major advantages of N A A include the ability to achieve low detection limits and the absence of matrix effects. The concentrations of total solid-phase As that were determined with N A A are reported in Table 6.9. The N A A results indicate total As concentrations in the C l and C3 samples that range from 2.2 to 4.1 ug/g. The total solid-phase As concentrations that are presented in Table 6.9 are comparable to those reported by Harvey et al. (2002) for the sediments at their field site in Munshiganj, Bangladesh. The reported N A A results for the C l and C3 samples also include low relative standard deviation (RSD) values that range from roughly 3 to 5 %, confirming the reliability of the results. 109 The extractant solutions from the SEP were also analysed for concentrations of total Fe with a spectrophotometer, to estimate the concentrations of different pools of solid-phase Fe. Inspection of the data in Table 6.9 indicates variability as high as 50 % in the concentrations of solid-phase Fe between split samples from Cl-41 and C3-52. Although the SEP that was utilised for this investigation was specifically designed for solid-phase As analysis, the results of the Fe analyses provide a general estimation of the concentrations of different solid-phase pools of Fe within the sediment samples. Concentrations of total solid-phase sulphur were analysed with an elemental analyser. Sample C3-52 was analysed for concentrations of total S twice to assess the reliability of the analytical results. The concentration of total S in sample C3-52 was reported as 180 u,g/g each time it was analysed, confirming the reliability of the method. 110 CN oo I CO O CN 00 CO, O CO O 1 co O JD • * (0 T— 5 cn CM CO CM IS M 9 II a. E Q| CO CO LO _ CO O ^ CM d d o CD CO 3 CD o Q. "D c CO "D c 3 O CO CD O ^ CM O O CM o d O N , . • * 0 0 CM d d CO CO i n CM i— . r-i r-i CM O CM O CD _ t o p CO d o ^ _ LO O ^ CN d o O ^ CM d o _0> _C0 D) D> D) =L, A =5. _ Q . ZD CO « 3 O CO ° CO 1^ CM CM CO CO CO o o CM f-d d CM d CM LO CM O co ^ o ? ^ CO CO LO CM LO Cvi CO CM CM CM CO CD ra c o J D t_ CO o CO ^ CD CO 2 > x < 2 — "D CO >• CD £ s ]? o ' CD L L Cfl _ Q . O 5 E y CO ® co co -. t ; CO CD. CD o S CD X * J O -> CD •g x o 3 0 c 9- .y o CO CD O c i 2 8 o 3 3 "D 0 ) 2 'g . o CD 1~ Q . O CO CO T3 E CD CD JD *-O ? -o « CO o, _>. CD CD 5 C J Z O Q . o o CM O O LO LO T - CD d o CO i - CO CO d T— co T— CO CD d ^ d LO CM T - o CO CO CM CO co cvi o r-CM CO CM CO CD f-O CM CM T t CM T - 0 0 CM d T— CM CO O CD O CM CM CD CO CM CM CM CO Cvi Tt LO CO CM CD O T T CO CM _cp _o> c^o CO CO O) E E E CO E E CD O CL 2 CO C o JD i— CO o CO ^ CD CO s > x 5 2 ~- -g CO >> CD J= ? X CD £  u-2 3 "•a £ C L CO "D C D O JD "CO g 'c o § •D CJ CO sz I o E co £> CD > -o ? » ro ro -~ CO CL CD o 2 CD X ° £ CO CD T3 CD "D X o t J : >> x o CD c ro I o 3 3 •a 0 ) ro CD. a. o T3 £ ro 5. O O IB .o .o n .o <u a a) CO <D (As) CM CO Tt LO i— CM CO •<t LO (As) Q. CD. CD- Q . a. 1_ CD- CD- a. C L C L V (As) CD CD CD CD CD ro CD CD CD CD CD mel nic Psl Psl Psl r S1 Psl side • 5 (Fe Psl Psl Psl r SI Psl CO o UJ LU L U LU LU CD o LU LU LU LU LU (9 m \— co co CO CO CO CC H o co CO CO CO CO a < O 0 0 o 0 0 o CO o CO CO o o o o o o CO CO sz « Q . o 3 H CO I ? E e E z s . o S ~ P co K t » t S H o " N >• S ' > - h - CO D . S S H < o 4^1 Si 1 3. O < CD co ca I ? CJ O. o a) I I CO "O a) a) o. E ro s S " e P o a S " " O JO o o 5 3 -2-co -a I ro _ co .ts E o l ro o I ro = <u ro « c ro "D ^ ro ts ro CO CD C W 0) 6.2.2.1 Interpretation Review of Table 6.9 indicates that the concentrations of solid-phase As in the sediment samples from coring locations C I and C3 range from 2.2 to 4.1 M-g/g. Figures 6.20 and 6.21 present plots of the concentrations of different solid-phase pools of As in sediments with depth for coring locations C I and C3, respectively. Inspection of the plots indicates that total concentrations of solid-phase As in the sediments at the Kidd2 site appear to decrease with depth, and are generally higher than those observed in the sediments at the D N D site. The concentration of solid-phase As that is coprecipitated with crystalline H F O minerals appears to be significantly higher in core sample C l - 2 9 than the other core samples at locations C I and C3, at approximately 2.12 jxg/g. However, this may be due analytical or sampling errors during the extraction process. Otherwise, the distribution of As among the pools targeted by SEP steps 1 and 3 through 5 appears to be relatively consistent throughout the subsurface at coring locations C I and C3. The majority of the solid-phase As in the sediments at both the D N D and Kidd2 sites appears to be either strongly adsorbed, or present in more recalcitrant mineral phases. However, there are also significant concentrations of solid-phase As that are associated with H F O minerals, and possibly the A V S fraction. 112 Figure 6.20: Concen t ra t i ons of so l i d -phase A s in sed iment in co re C 1 , DND si te o 5 o CD n £ 15 Q . CD T3 20 0 1 2 3 4 concentration As (ug/g) • Step 1 • Step 3 • Step 4 • Step 5 W Residual 113 Figure 6.21: Concentrations of solid-phase As in sediment in core C3, Kidd2 site o 5 20 2 5 ] , , , , 1 0 1 2 3 4 5 concentration As (ug/g) • Step 1 El Step 3 El Step 4 • Step 5 % Residual 114 The results in Table 6.9 suggest that significant concentrations of Fe oxyhydroxide (FIFO) minerals are present in the sediments from the C l and C3 samples. The sum of the concentrations of the Fe that is associated with both amorphous and crystalline H F O mineral phases in the C l and C3 sediment samples range from approximately 3.6 to 7.3 mg Fe per gram of sediment. The corresponding molar concentrations of H F O minerals are approximately 65 to 131 u M per gram of sediment. However, the concentrations of solid-phase As that are associated with H F O minerals are estimated to be in the range of 0.2 to 2.3 ug As per gram of H F O mineral (2.7 to 31 n M As/g sediment), at an A s : H F O molar ratio of approximately 4.1 x 10"5 to 2.3 x 10"4. Thus, the release of all the solid-phase As that is associated with the amorphous and crystalline H F O mineral phases in the sediments from cores other than Cl -29 , has the potential to generate concentrations of dissolved As of approximately 1400 to 3700 ug /L. Therefore,- even though solid-phase As that is associated with H F O minerals is present in the sediment at trace concentrations, it has the potential to generate significant concentrations of dissolved As. Sample calculations for the determination of maximum porewater concentrations are contained in Appendix F. As discussed in Section 6.1.2.1, the elevated concentrations of dissolved As at the Kidd2 site are in the vicinity of the relatively deep, density-driven flow cell (see Figure 6.19). If the dissolved As were to act as a conservative species, once released from the solid phase, it would be flushed out of the aquifer. Assuming the potential source of dissolved As from H F O minerals at 1400 Ug /L, the concentration of As in the porewater remains relatively constant at 30 Ug /L, and a groundwater residence time of 530 years, the amount of time required to flush the HFO-associated solid-phase As out of the aquifer sediments within the flow cell is estimated at 24,750 years (see Appendix F for sample calculations). The hydraulic conductivity of the distributary channel sands throughout the Fraser River is approximately 2 x 10"4 m/s to 4 x 10"4 m/s, with estimated porosity values of 0.3 (Neilson-Welch 1999). In upland areas of the Fraser River delta, where hydraulic gradients within the distributary channel sand unit are approximately 2 x 10"4 (Smith 115 2004), the. groundwater velocity is approximately 6.3 m/y (see Appendix F for calculation). The D N D site is located approximately 1.5 km downgradient from the groundwater divide that intersects the north delta area (Ricketts 1998). The corresponding residence time for groundwater that flows horizontally from the groundwater divide to the D N D site is approximately 238 years (see Appendix F). However, this estimate of groundwater residence time only includes the horizontal component, and does not account for the vertical component of flow from either the overlying clayey silt unit or the underlying silty clay unit. Although the estimated residence time for groundwater that flows from the groundwater divide to the D N D site is less than the residence time estimated for groundwater to cycle through the saline wedge at the Kidd2 site, the concentrations of As are lower at the D N D site. This supports the assertion that the relatively elevated concentrations of dissolved As that are present at the Kidd2 site are a result of increased mass loading at the site. A number of attenuation mechanisms retard the migration of As through the subsurface. For example, once released into solution, dissolved As can be sequestered from reduced groundwater during the precipitation of sulphide minerals such as mackinawite. Dissolved Fe is present in the deep porewater at the Kidd2 site at concentrations in the range of 150-230 mg/L, and SO42". has been measured in the deep porewaters at concentrations ranging from approximately 800 to 1400 mg/L (Neilson-Welch 1999). This suggests that there are kinetic limitations to the reduction of SO4 " at the site. However, it is expected that some SO4 " will reduce to HS" in the presence of D O C oxidation and sulphides will precipitate out of solution. The analytical results presented in Table 6.9 indicate that concentrations of total sulphur in the C3 samples ranges from 180 ug/g at a depth of 52 feet (15.8 m) below ground surface to 370 ug/g at 41 feet (12.5 m), confirming the presence of sulphur-containing minerals in the sediments at the Kidd2 site. In contrast to the Kidd2 site, concentrations of total sulphur in the C l sediments are relatively low, at concentrations of 110 ug/g or lower. As discussed in Section 6.1.2.1, the results of thin section analyses indicate that the sediments at the D N D and Kidd2 sites are very similar in mineralogical composition. Therefore, it is presumed that the concentration of S042"-containing lithic fragments would be similar in the sediments from 116 both sites, supporting the assertion that the elevated concentrations of total sulphur in the Kidd2 sediments are associated with the presence of authigenic sulphide minerals. The data presented in Table 6.9 indicates that the concentrations of solid-phase As that is coprecipitated with acid volatile sulphides (AVS) , carbonates, M n oxides, and very amorphous Fe oxyhydroxides (HFO), ranges from 0.25 to 0.34 ug/g. The groundwater within the deep aquifer sands contains SO42" and dissolved Fe and As at concentrations as high as approximately 1400 mg/L, 229 mg/L and 32 ug/L, respectively. However, the solid-phase As that is associated with A V S minerals appears to be similar at both sites and exhibits no pattern with depth. In contrast, the concentrations of solid-phase As that is strongly adsorbed and/or associated with more recalcitrant minerals such as As oxides and sulphides are higher in the sediments at the Kidd2 site than the D N D site, at concentrations ranging from 1.7 to 1.9 ug/g in the C l samples and 1.9 to 3.7 ug/g in the C3 samples. Although the concentrations of solid-phase As that is adsorbed and associated with more recalcitrant minerals tend to decrease with depth, the pattern is not strongly correlated to concentrations of total sulphur. This suggests that solid-phase As may be associated with recalcitrant phases other than just sulphides. Review of the solid-phase analyses indicates that the amount of solid-phase As that is associated with A V S may be similar in the sediments at both the D N D and Kidd2 sites. This suggests that similar amounts of dissolved As may have been sequestered from the groundwaters at each site during the precipitation of authigenic sulphides. Alternatively, the authigenic minerals in the sediments at the Kidd2 site may have undergone maturation processes. This is consistent with the presence of higher concentrations of solid-phase As that is associated with more recalcitrant mineral phases such as more crystalline sulphide minerals in the sediments at the Kidd2 site. In addition to the precipitation of sulphides, adsorption processes can also sequester As from solution. As discussed in Section 6.2.2, difficulties were encountered in analysing the SEP extractant solutions that targeted strongly sorbed As. Therefore, it is not possible, with the current data set, to estimate the concentrations of strongly sorbed As in the sediments at the Kidd2 site. However, the presence of elevated concentrations of 117 dissolved As at the Kidd2 site suggests that more dissolved As has been generated than has been attenuated. As discussed in Section 6.1.2.1, the presence of elevated concentrations of dissolved Fe, As and HCO3" in profile P3 suggests that H F O minerals are undergoing reductive dissolution in the subsurface at the Kidd2 site. However, concentrations of D O C exhibit a negative correlation with dissolved As in P3 and the concentrations of solid-phase As that is associated with H F O minerals are generally similar throughout the vertical profile. This further suggests that the reductive dissolution of H F O minerals may not be occurring at a significant rate in the vicinity of P3, but rather the dissolved Fe and As are transported to P3 from an area that is hydraulically upgradient. The absence of significant concentrations of dissolved Fe and As at depths not within the saline wedge also supports the hypothesis that the elevated concentrations of dissolved As are related to increased mass loading within the saline wedge. The presence of relatively high concentrations of As within the saline wedge suggests that the sorptive capacity of the sediments is approaching saturation, and the precipitation of arsenical sulphides is not sufficient to mitigate the dissolved As that is generated in the porewater. As discussed in Chapter 2, As(III) is more mobile than As(V) in solution (Watt et al. 2003). Therefore, the concentration of approximately 30 Ug/L dissolved As within the saline wedge may be a function of the rate at which solid-phase As is released to solution in upgradient areas, and less so the rate at which As is transported along the flowpath. 118 7.0 S U M M A R Y A N D C O N C L U S I O N S The results of this investigation indicate that significant concentrations of naturally-occurring dissolved arsenic (As) are present in the reduced, circumneutral groundwaters in the deltaic sediments of the Fraser River delta at certain locations. The D N D site is located approximately 3 km upgradient from the Fraser River. The groundwater throughout the vertical profile at this site is characterised as fresh, meteoric water. Concentrations of dissolved As are relatively low in the two vertical profiles at the D N D site, with a slight peak in profile PI of approximately 8 to 9 ug/L between depths of 10.4 and 14.2 meters below ground surface. Residence times for the groundwater at the D N D site are estimated to be approximately 240 years. This suggests that porewaters in the Fraser River delta have experienced more meteoric flushing since deposition than originally, hypothesised by.Simpson and Hutcheon,(1995). Adjacent to the north arm of the Fraser River, at the Kidd2 site, a distinct zone of dissolved As and iron (Fe) is observed in .the vertical profile from a depth of 11.9 m below ground surface to the bottom of the aquifer unit at approximately 22m. The dissolved As and Fe is present at concentrations of approximately 30 Ug/L and 120 to 230 mg/L, respectively. This zone of elevated dissolved As appears to be correlated to groundwater that is circulating within a relatively deep saline water wedge. Residence times for the groundwater within the saline wedge are estimated at approximately 530 years, suggesting that the elevated concentrations of dissolved As are a result of increased mass loading of reactants to the aquifer at that location. Interpretation of the analytical results suggests that the reduction of hydrous ferric oxide (HFO) minerals is the mechanism by which solid-phase As that is coprecipitated with, and strongly adsorbed to, H F O minerals is released into solution. Although the solid-phase As that is associated with H F O minerals is only present at trace concentrations of generally less than 1 ug As per gram of aquifer sediment, it represents a significant potential source of dissolved As. The analytical results also suggest that the precipitation of sulphide minerals, a process that sequesters As from solution, is occurring within the sediments of the Fraser River delta. However, the presence of significant concentrations 119 of dissolved As at the Kidd2 site indicates that more As has been released into solution than can be sequestered. Concentrations of sulphur are relatively lower in the sediments at the D N D site, suggesting that less sulphide minerals are present at that location. However, the absence of significant concentrations of dissolved As suggests that the release and sequestration of As is closer to equilbrium at the D N D site than the Kidd2 site. The absence of significant concentrations of dissolved organic carbon (DOC), the species which drives the reduction of H F O minerals, within profile P3 suggests that more complete oxidation of the D O C load is occurring and/or the elevated concentrations of dissolved As and Fe have been transported to that location from an area that is hydraulically upgradient. In the absence of sufficient sorption processes and precipitation of sulphides, it appears that the dissolved As may be slowly flushing from the aquifer. However, with rough estimates of 530 years for groundwater residence times within the saline water wedge and a sustained concentration of approximately 30 ug/L dissolved As in the groundwater, it would take approximately 24,750 years for the aquifer at the Kidd2 site to be flushed of the solid-phase As that is coprecipitated with, and adsorbed to, H F O minerals. 7.1 Conceptual model Based on the analytical results from this investigation, a conceptual model of As geochemistry at the Kidd2 site is presented and discussed. Figure 7.1 provides an illustration of the conceptual model. 120 CD ^ CO £ CM < 0 e ® / o 0 ° 0 o cn 0} "O X !c Q. C\l o SU o O <D CO <D CO D UL < U-0 • 0 • ui CO X co < T3 C ca <D CD •7) U_ S " ? CO CD 2^ CD zs CD 'S < s >- O CD CO 0 ^ cn c o i f i CO O < CD X *? — CD C t S C O E g ^ CD CO CD 1- - r , S C CO XI (D CD p ±2 3 CD o -O) M - > co C (- ® -CD CD W = -Q co IE CD CD — > C C CO 7= CO CO CM o C O CD E o CO 2 'Lo ii eg T , 1 c CD CO E co = CD CJ rn CD ^ co S TS o CO W CD CO CD £ O O £ T5 W CO C _ Q. CO •§ o x: ® ~ &• w c= ™ CD £ g> o-CD CO O CO CO c CM CO O - o CO . 5 SU ~o CO 11 o D ) E 03 CD T S TS CD CO >> .c 0 co 8 1$ 0 .2 c cn CD D xi o T> g CD CD CD CO G i _ * 8 ® o o I 0 CO c TS ' ^ O C O E .2 CT) TS S ® CO CO r 3 0 * -O 8 | 5 § o 5 O 2 c . 0 IS" — CO 0 Sr. O) E © c . I I I u. co Q CO i ' CO — JO CS) CO CO 'F C >- c o J> >_ £ g 0 co o l 8 - e g o Q. O O ZJ cn to W 0 < o £ ~° a *- 0 C CO > ~ < o 0 CO | £ | ; Q . co c 0 3 0 . n § OT co-c X 05 Z, C T3 C C0 0 o _> c CD o *-L L . CO => CO XI 1 B | p e c CO 0 ° O o c If .£ o X r i Cameron et al. (1995) report concentrations of D O C that range from 1 to 10 mg/L in the Fraser River, and analytical results for a sample of the saline water at the bed of the Fraser River and adjacent to the Kidd2 site indicate that the concentration of sulphate (SO42") was 1310 mg/L in early February, 2004. This dense saline water infiltrates through the hyporheic zone and down into the reduced aquifer that underlies the Kidd2 site. This relatively dense water migrates down into the aquifer and, in the presence of more crystalline, lower energy yielding H F O minerals, the oxidation of D O C is accompanied by the reduction of SO42" to sulphide (HS). As the dense groundwater migrates inland along the top of the basal silty clay unit, it encounters sediments with more amorphous H F O minerals that contain adsorbed and coprecipitated solid-phase As. In the absence of higher energy yielding terminal electron accepting processes (TEAPs), the microbial decomposition of D O C is accompanied by the simultaneous reduction of both SO42" and the amorphous H F O minerals. In the process, the solid-phase As that is assoicated with the amorphous H F O minerals is released into solution. Once in solution, As will sorb onto available sorption sites on the adjacent sediments. Dissolved Fe and HS" will react and precipitate into sulphides, a process that sequesters trace concentrations of dissolved As. However, sequestration processes such as sorption and sulphide precipitation only remove a certain amount of As from solution, with the residual load of dissolved As remaining in solution as it is transported downgradient. Under the reducing conditions, it is proposed that the dissolved As reduces to As(IU), a species that is relatively mobile in the subsurface. In the presence of H F O minerals with sorption sites that are saturated, no net change in the concentrations of dissolved As will be observed along the flowpath. The data that are presented for this investigation are consistent with the proposed conceptual model. However, there are a number of issues that need to be clarified in order to support or refute the model. In particular, the key topics that must be investigated are: • Is a front of depleted H F O minerals and the associated solid-phase As present in the subsurface at the Kidd2 site? If so, what is the rate at which the front is advancing 122 along the flowpath? Is the delta evolving into a more mature delta environment in which the sediments are flushed of the mobile forms of solid-phase As? • What is the source of the D O C that is driving the reduction of H F O minerals? Is the D O C sourced from river water and the organic material that is deposited along the banks of the Fraser River? Is the D O C that drives the redox reactions infiltrating down into the aquifer from the clayey silt deposits that cap the Kidd2 site? • If the D O C is associated with the water that is infiltrating into the aquifer through the hyporheic zone, are there any seasonal effects that influence the introduction of D O C to the subsurface? • Could other processes such as competitive anion exchange also be responsible for the release of solid-phase As into the groundwater? " '. • Although rough estimates have been provided in this investigation, what are the potential concentrations of dissolved As that could be generated in the Kidd2 sediments? What are the potential concentrations associated with the D N D site? What is the sensitivity of the system to changes? • To what degree is the dissolved As sequestered in the sediments? What is the sorption potential of the sediments? Are the sorption sites on the sediments saturated? • Is the precipitation of acid-volatile sulphides (AVS) a process by which significant concentrations of dissolved As are sequestered from solution? Is there a difference between the concentrations of A V S in the sediments at the Kidd2 site and the D N D site? Is there a difference between the ratio of solid-phase As to A V S in the sediments at the two sites? 7.2 Recommendations Based on the outstanding issues discussed above, a number of recommendations are proposed for future research. A number of profiles could be advanced along a north-south cross-section at the Kidd2 site. Analysis of the groundwater and sediments along this cross-section would confirm the presence of a front of depleted H F O minerals and 123 the associated solid-phase As in the subsurface. If identified, the location of the front would provide insight into the rate at which the front is advancing along the flowpath. A n investigation of the groundwater and sediments along the Fraser River, at a location on the delta that is farther from the ocean would provide information on the geochemistry of more mature areas of the delta. This would provide information on what the Kidd2 site might be evolving towards as mobile solid-phase As is flushed from the sediments. Isotopic analyses of the carbon would provide information on the source of the organic carbon that is driving the reduction of H F O minerals. Radiocarbon ages for the D O C in deep zone of the aquifer could be compared to those in the shallow groundwater at the site and those in the relatively dense water along the bed of the Fraser River. Analytical data for water along the bed of the Fraser River at different times in the year could provide insight into possible seasonal effects and influences on the introduction of D O C and SO42" to the hyporheic zone. Field redox manipulation experiments, or "push-pull" experiments could be conducted at the Kidd2 site to determine the concentrations of dissolved As that could be generated by changes in the geochemistry of the subsurface. For example, the addition of a labile organic carbon to the groundwater would drive the reductive dissolution of H F O minerals and the release of associated solid-phase As. The concentrations^ of dissolved As following the introduction of the organic carbon would provide insight into the rate at which the dissolved As is sequestered. A sensitivity analysis could be performed with the completion of a number of push-pull tests, at different concentrations of injectate. The potential for competitive anion exchange in the sediments at the Kidd2 site could also be determined by the addition of a competing anion such as phosphate (PO4 3 ) . Similar manipulations could also be performed at the D N D site to compare dissolved As potential at the Kidd2 site to an upland, relatively undisturbed location. Analysis of the sediment samples from the Kidd2 site and the D N D site should be expanded to include an estimate of the strongly sorbed concentration of solid-phase As. This information is necessary to develop an isotherm for the sediments of the Fraser River delta, and characterise the mobility of dissolved As in the subsurface. 124 Concentrations of solid-phase As appear to be similar in the sediments at the Kidd2 site and the D N D site. However, the concentrations of total S are higher in the Kidd2 sediments, suggesting that more mature sulphides might be present in the subsurface at the Kidd2 site. A n estimation of the concentrations of A V S in the sediments at the Kidd2 site and the D N D site will determine the influence of sulphide precipitation processes in the sequestration of dissolved As. The ratio of solid-phase As to A V S in the sediments may be similar in the sediments at the two sites, or may be lower at the Kidd2 site, suggesting kinetic limitations to the sequestration of As in A V S . 7.3 Implications The maximum concentration of dissolved As that was identified at one of three sites in the Fraser River delta was approximately 32 u.g/L, which is over an order of magnitude lower than the highest concentrations documented in Bangladesh. However, concentrations of solid-phase As in the sediment are similar to those reported by Harvey et al. (2002) at their field site in Munshiganj, Bangladesh. This suggests that the sediments in the Fraser River delta have the potential to generate significant concentrations of dissolved As in the groundwater. Unlike areas such as Bangladesh and West Bengal, the Fraser River delta has not experienced significant development of the groundwater resources. Therefore, the results from this investigation of the Fraser River delta are useful for comparison to results of field studies in Bangladesh. Disturbances to the groundwater regime in the Fraser River may trigger the release of significant quantities of As from the solid-phase into the groundwater. Increased mass loading of reactants to the groundwater has the potential to drive the reduction of H F O minerals, causing the associated solid-phase As to be released into solution. In the absence of adequate sequestration processes, concentrations of dissolved As may increase and migrate downgradient from the source. Therefore, pumping of the groundwater resources in the Fraser River delta may generate relatively elevated concentrations of dissolved As. Alternatively, the introduction to the subsurface of a labile organic carbon, such as hydrocarbon fuel, also has the potential generate elevated concentrations of 125 dissolved As. The application of phosphate-containing fertiliser could induce anion exchange processes in the subsurface, releasing solid-phase As into solution. 126 8.0 R E F E R E N C E S Acharyya, S. K . , Lahiri, S., Raymahashay, B . C . and Bhowmilk, A . (2000). "Arsenic toxicity of groundwater in parts of the Bengal Basin in India and Bangladesh: the role of Quaternary stratigraphy and Holocene sea-level fluctuation." Environmental Geology 39: 1127-1137. Adamsen, K . R. and Pokhrel, A . (2002). The arsenic contamination of drinking water in Nepal: draft 1. Kathmandhu, Nepal, Nepal Water for Health: 72. Aggett, J. and Kriegman, M . R. (1988). "The extent of formation of arsenic(HI) in sediment interstitial waters and its release to hypolimnetic waters in Lake Ohakuri." Water Resources 22: 407-411. Aggett, J. and Roberts, L . S. (1986). "Insight into the mechanism of accumulation of arsenate and phosphate in hydro lake sediments by measuring the rate of dissolution with ethylenediaminetetraacetic acid." Environmental Science and Technology 20: 183-186. Ahmed, K . M . , Ed . (2003). Arsenic contamination of groundwater and a review of the situation in Bangladesh. Groundwater resources and development in Bangladesh: background to the arsenic crisis, agricultural potential and the environment. Dhaka, University Hi l l Press. Ahmed, K . M . , Bhattacharya, P., Hasan, M . A . , Akhter, S. H . , Alam, S. M . M . , Bhuyian, M . A . H . , Imam, M . B. , Khan, A . A . and Sracek, O. (2004). "Arsenic enrichment in groundwater of alluvial aquifers in Bangladesh: an overview." Applied Geochemistry 19: 181-200. Akai, J . , Izumi, K . , Fukuhara, H . , Masuda, H . , Nakano, S., Yoshimura, T . , Ohfuji, H . , Anawar, H . M . and Akai, K . (2004). "Mineralogical and geomicrobiological investigations on groundwater arsenic enrichment in Bangladesh." Applied Geochemistry 19: 215-230. Alpers, C . N . , Blowes, D . W. , Nordstrom, D . K . and Jambor, J. L . (1994). Secondary minerals and acid mine-water chemistry. Short course handbook on environmental geochemistry of sulfide mine-wastes. D . W . Blowes. Waterloo, Ont., Mineralogical Association of Canada: 247-270. Anawar, H . M . , Akai, J . , Komaki, K . , Terao, H . , Yoshioka, T . , Ishizuka, T. , Safiullah, S. and Kato, K . (2003). "Geochemical occurrence of arsenic in groundwater of Bangladesh: sources and mobilization processes." Journal of Geochemical Exploration 77: 109-131. Appelo, C . A . J. and Postma, D . (1993). Geochemistry, Groundwater and Pollution. Rotterdam, A . A . Balkema Publishers. Appelo, C . A . J. , Van der Weiden, M . J. J. , Tournassat, C . and Charlet, L . (2002). "Surface complexation of ferrous iron and carbonate on ferrihydrite and the mobilization of arsenic." Environmental Science and Technology 36: 3096-3103. 127 Aurillo, A . C , Mason, R. P. and Hemond, H . F. (1994). "Speciation and fate of arsenic in three lakes of the Aberjona Watershed." Environmental Science and Technology 28: 577-585. Bearak, B . (2002). Bangladeshis sipping arsenic as plan for safe water stalls. New York Times. New York. Beckie, R. (2002); Lecture: Arsenic contamination of groundwater in Bangladesh. University of British Columbia. Belzile, N . and Lebel, J. (1986). "Capture of arsenic by pyrite in near-shore marine sediments." Chemical Geology 54: 279-281. Belzile, N . and Tessier, A . (1990). "Interactions between arsenic and iron oxyhydroxides in lacustrine sediments." Geochemica et Cosmochimica Acta 54: 103-109. Berner, R. A . (1970). "Sedimentary pyrite formation." American Journal of Science 268: 1-23. Berner, R. A . (1971). Principles of chemical sedimentology. New York, McGraw-Hil l Inc. Berner, R. A . (1980). Early diagenesis: A theoretical approach. Princeton, NJ , Princeton University Press. Bianchin, M . (2004). PhD thesis in progress. Department of Earth and Ocean Sciences, U B C . Bostick, B . C . and Fendorf, S. (2002). "Arsenite sorption on troilite (FeS) and pyrite(FeS2)." Geochemica et Cosmochimica Acta 67: 909-921. Cameron, E . M . (1995). "Hydrogeochemistry of the Fraser River, British Columbia: seasonal variation in major and minor components." Journal of Hydrology 182: 209-225. Cameron, E . M . , Hall , G . E . M . , Veizer, J. and Krouse, H . R. (1995). "Isotopic and elemental hydrogeochemistry of a major river system: Fraser River, British Columbia, Canada." Chemical Geology 122: 149-169. Chen, C . J. and Lin , L . J. (1994). Human carcinogenicity and atherogenicity induced by chronic exposure to inorganic arsenic. Arsenic in the environment, Part II: Human health and ecosystem effects. J. O. Nriagu. New York, John Wiley & Sons. 27: 109-131. Christensen, T. H . , Bjerg, P. L . , Banwart, S. A . , Jakobsen, R., Heron, G . and Albrechtsen, H . J. (2000). "Characterization of redox conditions in groundwater contaminant plumes." Journal of Contaminant Hydrology 45: 165-241. Clague, J. J. (1998). Geological setting of the Fraser River delta. Bulletin 525: Geology and natural hazards of the Fraser River delta, British Columbia. J. J . Clague, J. L . Luternaurer and D . C . Mosher. Vancouver, B C , Canada, Geological Survey of Canada: 7-16. 128 Clague, J. J . , Luternauer, J. and Hebda, R. J. (1983). "Sedimentary environments and postglacial history of the Fraser Delta and lower Fraser Valley, British Columbia." Canadian Journal of Earth Sciences 20: 1314-1326. Clague, J. J . , Luternauer, J . , Monahan, P. A . , Edwardson, K . A . , Dallimore, S. R. and Hunter, J. A . (1998). Quaternary stratigraphy and evolution of the Fraser delta. Bulletin 525: Geology and natural hazards of the Fraser River delta, British Columbia. J. J. Clague, J. L . Luternaurer and D . C . Mosher. Vancouver, B C , Canada, Geological Survey of Canada: 57-90. Cullen, W . R. and Reimer, K . J. (1989). "Arsenic speciation in the environment." Chemical Reviews 89: 713-764. Cummings, D . E . , Caccavo, F . J . , Fendorf, S. and Rosenzweig, R. F . (1999). "Arsenic mobilization by the dissimilatory Fe(III)-reducing bacterium Shewanella alga BrY." Environmental Science and Technology 33: 723-729. Dean, W . E . and Arthur, M . A . (1989). "Iron-sulfur-carbon relationships in organic-carbon-rich sequences I: Cretaceous Western Interior Seaway." American Journal of Science 289: 708-743. Edenborn, H . M . , Belzile, N . , Mucci, A . , Lebel, J. and Silverberg, N . (1986). "Observations on the diagenetic behavior of arsenic in a deep coastal sediment." Biogeochemistry 2: 359-376. Farquhar, M . L . , Charnock, J. M . , Livens, F. R. and Vaughan, D . J. (2002). "Mechanisms of arsenic uptake from aqueous solution by interaction with goethite, lepidocrocite, mackinawite, and pyrite: an X-ray absorption spectroscopy study." Environmental Science and Technology 36: 1757-1762. Freeze, R. A . and Cherry, J. A . (1979). Groundwater. Upper Saddle River, NJ , Prentice Hall, Inc. Fuller, C . C , Davis, J. A . and Waychunas, G . A . (1993). "Surface chemistry of ferrihydrite: Part 2. Kinetics of arsenate adsorption and coprecipitation." Geochemica et Cosmochimica Acta 57: 227'1-2282. Gleyzes, C , Tellier, S., Sabrier, R. and Astruc, M . (2001). "Arsenic characterisation in industrial soils by chemical extractions." Environmental Technology 22: 27-38. Goldhaber, M . B . and Kaplan, I. R. (1974). The sulfur cycle. The sea. E . D . Goldberg. New York, Wiley & Sons. 5: 569-656. Gulens, J . , Champ, D . R. and Jackson, R. E . (1973). Influence on redox environments on the mobility of arsenic in groundwater. Chemistry of water supply treatment and distribution. A. J. Rubia. Ann Arbor, Ann Arbor Science Publishers. Harvey, C . F . , Swartz, C . H . , Badruzzaman, A . B . M . , Keon-Blute, N . , Y u , W. , A l i , M . A . , Jay, J . , Beckie, R., Neidan, V . , Brabander, D . , Oates, P. M . , Khandaker, N . A . , et al. (2002). "Arsenic mobility and groundwater extraction in Bangladesh." Science 298: 1602-1606. - . 129 Howard, A . G . , Apte, S. C , Comber, S. D . W . and Morris, R. J. (1988). "Biogeochemical control of the summer distribution and speciation of arsenic in the Tamar Estuary." Estuarine, Coastal and Shelf Science 27: 427-443. Huang, H . and Dasgupta, P. K . (1999). "A field-deployable instrument for the measurement and speciation of arsenic in potable water." Analytica ChimicaActa 380: 27-37. Keon, N . E . , Swartz, C . H . , Brabander, D . J. , Harvey, C . F . and Hemond, H . F . (2001). "Validation of an arsenic sequential extraction method for evaluating mobility in sediments." Environmental Science and Technology 35: 2778-2784. K i m , M . J. , Nriagu, J. and Haack, S. (2000). "Carbonate ions and arsenic dissolution by groundwater." Environmental Science and Technology 34: 3094-3100. Kinniburgh, D . G . and Smedley, P. L . (2001). Technical report: Arsenic contamination of groundwater in Bangladesh. Keyworth, British Geological Survey. Kinniburgh, D . G . , Smedley, P. L . , Davies, J . , Milne, C . J . , Gaus, I., Trafford, J. M . , Burden, S., Huq, S. M . I., Ahmad, N . and Ahmad, M . K . (2003). The scale and causes of the groundwater arsenic problem in Bangladesh. Arsenic in groundwater: Geochemistry and occurrence. A . H . Welch and K . G . Stollenwerk. Boston, Kluwer Academic Publishers: 211-257. Korte, N . E . and Fernando, Q. (1991). "A review of arsenic (III) in groundwater." Critical Reviews in Environmental Control 21: 1-39. Langmuir, D . (1997). Aqueous Environmental Geochemistry. Upper Saddle River, NJ , Prentice-Hall Inc. Langner, H . W . and Inskeep, W . P. (2000). "Microbial reduction of arsenate in the presence of ferrihydrite." Environmental Science and Technology 34: 3131-3136. Le, X . C , Yalcin, S. and M a , M . (2000). "Speciation of submicrogram per liter levels of arsenic in water: on-site species separation integrated with sample collection." Environmental Science and Technology 34: 2342-2347. Leventhal, J. S. (1982). "An interpretation of carbon and sulfur relationships in Black Sea sediments as indicators of environments of deposition." Geochemica et Cosmochimica Acta 47: 133-137. L i , X . , Coles, B. J . , Ramsey, M . H . and Thornton, I. (1995). "Sequential extraction of soils for multielement analysis by ICP-AES. "Chemical Geology 124: 109-123. Lin , Z . and Puis, R. W . (2000). "Adsorption, desorption and oxidation of arsenic affected by clay minerals and aging process." Environmental Geology 39: 753-759. Lovley, D . R. and Chappell, F . H . (1995). "Deep subsurface microbial processes." Reviews of Geophysics 33: 365-381. Mackintosh, E . E . and Gardner, E . H . (1966). "A mineralogical and chemical study of Lower Fraser River alluvial sediments." Canadian Journal of Earth Sciences 46: 37-46. 130 Mandal, B . K . , Chowdhury, T. R., Samanta, G . , Mukherjee, D . , Chanda, C . R., Saga, K . C . and Chakraborti, D . (1998). "Impact of safe water for drinking on five families for 2 years in West Bengal, India." The Science of the Total Environment 218: 185-201. Manning, B . A . and Goldberg, S. (1997). "Adsorption and stability of arsenic(III) at the clay mineral-water interface." Environmental Science and Technology 31: 2005-2011. Manning, B. A . and Martens, D . A . (1997). "Speciation of arsenic(UI) and arsenic(V) in sediment extracts by high-performance liquid chromatography-hydride generation atomic absorption spectrophotometry." Environmental Science and Technology 37: 171-177. Mass, M . J. , Tennant, A . , Roop, B . C , Cullen, W . R., Styblo, M . , Thomas, D . J. and Kligerman, A . D . (2001). "Methylated bivalent arsenic species are geotoxic." Chemical Research in Toxicology 14: 355-361. Masscheleyn, P. H . , Delaune, R. D . and Patrick, W . H . J. (1991). "Effect of redox potential and p H on arsenic speciation and solubility in a contaminated soil." Environmental Science and Technology 25: 1414-1419. Masud, K . (2000). "Arsenic in groundwater and health problems in Bangladesh." Water Resources 34: 304-310. Mathews, W . H . and Bustin, R. M . (1994). "Trace metal geochemistry of peat under a sanitary landfill - a reconnaissance." Environmental Geology 23: 14-22. McArthur, J. M . , Ravenscroft, P., Safiullah, S. and Thirlwall, M . F. (2001). "Arsenic in groundwater: testing pollution mechanisms for sedimentary aquifers in Bangladesh." Water Resources Research 37: 109-117. Meuller, B . (2004). Analytical Laboratory Technician, p. communication. Oceanography, U B C . Monger, J. W . H . and Journeay, J. M . (1994). Basement geology and tectonic evolution of the Vancouver region. Bulletin 481: Geology and geological hazards of the Vancouver region, Southwestern British Columbia. J. W . H . Monger. Ottawa, O N , Canada, Geological Survey of Canada: 3-25. Morse, J. W. , Millero, F . J . , Cornwell, J. C . and Rickard, D . (1987). "The chemistry of hydrogen sulfide and iron sulfide systems in natural waters." Earth Science Reviews 24: 1-42. Morton, W . E . and Dunnette, D . A . (1994). Health effects of environmental arsenic. Arsenic in the environment, Part II: Human health and ecosystem effects. J. O. Nriagu. New York, John Wiley & Sons. 27: 17-34. Murphy, F . and Herkelrath, W . N . (1996). "A sample-freezing drive shoe for a wireline-piston core sampler." Ground Water Monitoring and Remediation 16(3): 86-90. Naqvi, S. M . , Vaishnavi, C . and Singh, H . (1994). Toxicity and metabolism of arsenic in vertebrates. Arsenic in the environment, Part II: Human health and ecosystem effects. J. O. Nriagu. New York, John Wiley & Sons. 27: 55-91. 131 Neilson-Welch, L . (1999). Saline water intrusion from the Fraser River estuary: a hydrogeological investigation using field chemical data and a density-dependent groundwater flow model. Department of Earth and Ocean Sciences. Vancouver, University of British Columbia: 209. Neilson-Welch, L . and Smith, L . (2001). "Saline water intrusion adjacent to the Fraser River, Richmond, British Columbia." Canadian Geotechnical Journal 38: 67-82. Nickson, R. T . , McArthur, J. M . , Ravenscroft, P., Burgess, W . G . and Ahmed, K . M . (2000). "Mechanism of arsenic release to groundwater, Bangladesh and West Bengal." Applied Geochemistry 15: 403-413. Oremland, R. and Stolz, J. F . (2003). "The ecology of arsenic." Science 300: 939-944. Pierce, M . L . and Moore, C . B. (1982). "Adsorption of arsenite and arsenate on amorphous iron hydroxide." Water Resources 16: 1247-1253. Postma, D . and Jakobsen, R. (1996). "Redox zonation: equilibrium constraints on the Fe(III)/S04- reduction interface." Geochemica et Cosmochimica Acta 60: 3169-3175. Pyzik, A . J. and Sommer, J. E . (1981). "Sedimentary iron monosulfides: kinetics and mechanisms of formation." Geochemica et Cosmochimica Acta 45: 687-698. Rahman, M . , Mukherjee, D . , Sengupta, M . K . , Chowdhury, U . K . , Lodh, D . , Chanda, C . R., Roy, S., Selim, M . , Quamruzzaman, Q. , Milton, A . H . , Shahidullah, S. M . , Rahman, M . T., et al. (2002). "Effectiveness and reliability of arsenic field testing kits: are the million dollar screening projects effective or not?" Environmental Science and Technology 36: 5385-5394. Raven, K . P., Jain, A . and Loeppert, R. H . (1998). "Arsenite and arsenate adsorption on ferrihydrite: kinetics, equilibrium, and adsorption envelopes." Environmental Science and Technology 32: 344-349. Ravenscroft, P., McArthur, J. M . and Hoque, B . A . (2002). Geochemical and palaeohydrological controls on pollution of groundwater by arsenic. Fourth international conference on arsenic exposure and health effects, Dhaka, Bangladesh. Ricketts, B . D . (1998). Groundwater flow beneath the Fraser River delta, British Columbia; a preliminary model. Bulletin 525: Geology and natural hazards of the Fraser River delta, British Columbia. J. J. Clague, J. L . Luternaurer and D . C . Mosher. Vancouver, B C , Canada, Geological Survey of Canada: 241-255. Rochette, E . A . , Bostick, B . C , L i , G . and Fendorf, S. (2000). "Kinetics of arsenate reduction by dissolved sulfide." Environmental Science and Technology 34: 4714-4720. Seyler, P. and Martin, J. M . (1989). "Biogeochemical processes affecting arsenic species distribution in a permanently stratified lake." Environmental Science and Technology 23: 1258-1263. Shiowatana, J . , McLaren, R. G . , Chanmekha, N . and Samphao, A . (2001). "Heavy metals in the environment: fractionation of arsenic in soil by a continuous-flow sequential extraction method." Journal of Environmental Quality 30: 1940-1949. 132 Simpson, G . and Hutcheon, I. (1995). "Pore-water chemistry and diagenesis of the modern Fraser River delta." Journal of Sedimentary Research A65: 648-655. Skoog, D . A . (1985). Principles of Instrumental Analysis. Orlando, Saunders College Publishing. Smedley, P. L . and Kinniburgh, D . G . (2002). "A review of the source, behavior and distribution of arsenic in natural waters." Applied Geochemistry 17: 517-568. Smith, A . H . , Lingas, E . O. and Rahman, M . (2000). "Contamination of drinking-water by arsenic in Bangladesh: a public health emergency." Bulletin World Health Organization 78(9): 1093-1103. Smith, L . (2004). Professor of Groundwater Hydrology, p. communication. Department of Earth and Ocean Sciences, U B C . Spliethoff, H . M . , Mason, R. P. and Hemond, H . F. (1995). "Interannual variability in the speciation and mobility of arsenic in a dimictic lake." Environmental Science and Technology 29: 2157-2161. Styblo, M . , Del Razo, L . M . , Vega, L . , Germolec, D . R., E . L . , L . , Hamilton, G . A . , Reed, W. , Wang, C , Cullen, W . R. and Thomas, D . J. (2000). "Comparative toxicity of trivalent and pentavalent inorganic and methylated arsenicals in rat and human cells." Archives of Toxicology 74: 289-299. Sullivan, K . A . and Aller, R. C . (1996). "Diagenetic cycling of arsenic in Amazon shelf sediments." Geochemica et Cosmochimica Acta 60: 1465-1477. Swartz, C . H . , Keon Blute, N . , Badruzzaman, A . B . M . , A l i , M . A . , Brabander, D . , Jay, J. , Besancon, J. , Islam, S., Hemond, H . F. and Harvey, C . F . (2003). "Mobility of arsenic in a Bangladesh aquifer: inferences from geochemical profiles, leaching data, and mineralogical characterisation." Geochemica et Cosmochimica Acta in press. Takamatsu, T. , Kawashima, M . and Koyama, M . (1985). "The role of Mn 2 + - r i ch hydrous manganese oxide in the accumulation of arsenic in lake sediments." Water Resources 19: 1029-1032. Tareq, S. M . , Safiullah, S., Anawar, H . M . , Rahman, M . M . and Ishizuka, T . (2003). "Arsenic pollution in groundwater: a self-organizing complex geochemical process in the deltaic sedimentary environment, Bangladesh." The Science of the Total Environment 313: 213-226. van Geen, A . , Robertson, A . P. and Leckie, J. O. (1994). "Complexation of carbonate species at the goethite surface: implications for adsorption of metal ions in natural waters." Geochemica et Cosmochimica Acta 58: 2073-2086. van Geen, A . , Rose, J . , Thoral, S., Gamier, J. M . , Zheng, Y . and Bottero, J. Y . (2004). "Decoupling of As and Fe release to Bangladesh groundwater under reducing conditions. Part II: Evidence from sediment incubations." Geochemica et Cosmochimica Acta 68: 3475-3486. Van Herreweghe, S., Swennen, R., Vandecasteele, C . and Cappuyns, V . (2003). "Solid phase speciation of arsenic by sequential extraction in standard reference materials and industrially contaminated soil samples." Environmental Pollution 122: 323-342. 133 Watt, C . and Le, X . C . (2003). Arsenic speciation in natural waters. Biogeochemistry of environmentally important trace elements. O. C . Braids. Washington, D . C . , American Chemical Society. 835: 11-32. Wenzel, W . W . , Kirchbaumer, N . , Prohaska, T. , Stingeder, G . , Lombi, E . and Adriano, D . C . (2001). "Arsenic fractionation in soils using an improved sequential extraction procedure." Analytica ChimicaActa 436: 309-323. Wilkin, R. T . and Ford, R. G . (2002). "Use of hydrochloric acid for determining solid-phase arsenic partitioning in sulfidic sediments." Environmental Science and Technology 36: 4921-4927. Williams, H . F. L . and Roberts, M . C . (1988). "Holocene sea-level change and delta growth: Fraser River delta, British Columbia." Canadian Journal of Earth Sciences 26: 1657-1666. Wride, C . E . , Hofmann, B . A . , Sego, D . C , Plewes, H . D . , Konrad, J . - M . , Biggar, K . W. , Robertson, P. K . and Monahan, P. A . (2000). "Ground sampling program at the C A N L E X test sites." Canadian Geotechnical Journal 37: 530-542. Yamauchi, H . and Fowler, B . A . (1994). Toxicity and metabolism of inorganic and methylated arsenicals. Arsenic in the environment, Part II: Human health and ecosystem effects. J. O. Nriagu. New York, John Wiley & Sons. 27: 35-54. Zachara, J. M . , Girvin, D . C , Schmidt, R. L . and Resch, C . T. (1987). "Chromate adsorption on amorphous iron oxyhydroxide in the presence of major groundwater ions." Environmental Science and Technology 21: 589-594. Zubel, M . (2002). Hydrogeologic investigation of arsenic in domestic water wells, Mission, B C . Surrey, B C , B C Ministry of Water, Land & Air Protection - Lower Mainland Region: 1-57. 134 Appendix A : Borehole logs for the Kidd2 site 135 Fraser River H e N i *m><n—~—r~ River Drive BC Hydro Kidd 2 Substation | *W3 Mill »1 ()^PW|K«i)l JIJJIJLO i i*J5pfOK*rpate Ar<M -,: or RC C5«ll lfe4G>VlO 8 act 6 Condominium Development Residential Houses . W o s T B a y W e * W B » W « * K - C J ^ 9 PorffitrornatorTost 1 0 1 - 1 1 4 = Z o r t « ^ [ » c l r f c W a j * W 1 - W 3 - M u i H l o v e ! S a m p O n f l 0 V t t § ) W W l s Figure A . 1 : Kidd2 site plan (Neilson-Welch, 1999) 136 Site Location: UBC Field Site, Richmond, BC Drilling Date; 1996 03 26 Drilling Method: Sonic Drill Borehole Diameter 0.152 m Borehole Depth: 22,07 m Well Material: PVC Glued Pipe with Polyethylene sample tubes Well Screen Length: N/A Screen Slot Size: N/A |Well Diameter: PVC pipe = 0.0508 m Sample Tubes 0.478 cm ID, 0.787 cm OD Well Depth (below grade): 22.07 m Top of Well Elevation (from survey); MONITORING W E L L L O G INotes: Multi-level sampling well constructed with PVC pipe. Fifteen holes were aYUledin the PVC pipe for the sampling ports. Fifteen polyethy lene tubes were threaded inside the PVC pipe and through each hole. A nylon mesh was fastened to the end of each sampling tube. Depth; (tn) •o-r .2-13-4-5 6-'7-8-9-10-I I -12-13-14-15-16-17-18 '•}*~ 20-21 22—1 Ground Surface Native Sand Collapse W E L L N O : WI I'm Bcntonite Grout — 6:H» Sample Port Depth Number <M 1 2 % 4 •5: 6 7 8 9 10 11 12 13 14 IS 8,08 907 10.08 11.08 12.08 13.07 14.08 15,08 16.17 17.08 18.08 19.08 20.06 21.07 22.03 Figure A.2: Borehole log for multi-port well M1 (Neilson-Welch, 1999) 137 DEPTH INTERPRETED STRATIGRAPHY (m) - : l - 2 C L A Y E Y SILT Clayey sik,irace fine sand, trace to some organic matter (leaves.wood, etc.), light grey. 3 - 4 SILTY SAND Laminated silty fine sand - 5 Laminated medium and fine sand, some silt - 6 - 7 " ~ tilhtom<a^ IWTiHo"Sah<t r - - - - - - - -- 8 Z:Z Z Z-Z.Z-Z-. ,ZZ WbS^3tflI<&Z&i&^&iIiZ -Laminated silty tine sand 9 Medium sand ~ 10 FINE AND MEDIUM SAND Medium sand, some fine sand, grey, occasional silty sand lenses _ 11 MEDIUM SAND - 12 Medium Sand, uniform - 13 - 14 - 15 - 16 - 17 - 18 - 19 - 20 - 21 - 22 SILTY C L A Y - 23 Qayey silt to silty clay, thinly laminated, light grey SOIL L O G : BOREHOLE DRUJ.I-D; March 1995 LOCATION: Approx. 320 m South of River METHOD: Mud rotary drill and core barret Figure A.3: Interpreted stratigraphy - Kidd2 site (Neilson-Welch, 1999) 138 DEPTH (m) INTERPRETED STRATIGRAPHY RESISTIVrTY DATA - 1 - 2 - 3 FILL 1 1 1 1 1 1 1 10 20 30 40 50 60 70 CLAYEY SILT Sandy or clayey silt to silty clay and clay, trace fine sand. Resistivity (ohm-m) DATA NOT AVAILABLE - 4 - 5 - 6 - 7 - 8 SILTY SAND Silty sand with some layers of fine sand, some clayey and sandy silt lenses 6.0m PINE AND MEDIUM SAND Medium arid fine sand, uniform, some fine sand lenses, unit thickness unknown 9 - 10 MEDIUM SAND Medium Sand, uniform - 11 - 12 - 13 - 14 - 15 '—" Frcsh-Sallnc Transition Zone or "Interface" - 16 - 17 - 18 - 19 - 20 - 21 - 22 - 23 22.0m Saline-Fresh Transition Zone SDLTY CLAY Clayey silt to slliy clay SOIL LOG; CPTK9301 LOGGED B Y U B C Civ. Eng. DRHJ.F.D; 1993 LOCATION: Appro*. 375m south of River METHOD: CPT Figure A.4: Borehole log for K9301 (Nei lson-Welch, 1999) 139 DEPTH INTERPRETED STRATIGRAPHY RESISTIVITY D A T A 1 F H X 1 1 I I i 1 I 10 20 30 40 50 60 70 C S A Y E Y S I L T Resistivity (ohm-m) -2 3 Sandy or clayey silt to silty clay and clay, trace fine sand. 4.0m D A T A NOT A V A I L A B L E -4 5 6 SILTY SAND . . • Silty sand with some layers of fine sand, some clayey and sandy silt l e n s e s 7 8 9 - 10 11 12 11.0m —— Fresh-Saline Transition Zone or "Interface" - FINE A N D M E D I U M SAND — 13 14 Medium and fmc sand, uniform, some fine sand lenses, unit tliickness unknown _ 15 M E D I U M SAND Medium Sand, uniform _ 16 _ 1 7 _ 1 8 - 19 - 20 - 21 — 22 23 23.0m SILTY C l A Y Clayey silt to silty clay SOIL LOG: C P T K9612 DRILLED: 1996 LOCATION: Approx. 341m south of River METHOD: CPT Figure A.5: Borehole log for K9612 (Neilson-Welch, 1999) 140 Appendix B: Selected photographs 141 Photo 1: The Waterloo Drive Point Profiling system. The pneumatic hammer, which is suspended from the scaffolding, is used to advance the profiler tip and rod assembly. 142 Photo 3: Waterloo Drive Point Profiler (WDPP) groundwater sampling system. During advancement of the profiler assembly, de-ionised, distilled water (DIW) is pumped down through the internal tubing to prevent the ports (see Photo 2) from clogging. When the desired depth is achieved, the 3-way stopcock is adjusted to close the DIW reservoir, the peristaltic pump is reversed and groundwater is drawn up through the flow through cell and into the purge container. Note: groundwater samples are extracted from the 3-way stopcock with a syringe, thereby minimising sample oxidation. Photo 4: Groundwater sampling equipment Groundwater samples are extracted directly from the 3-way stopcock (see Photo 3) with the 60mL syringe to prevent sample oxidation. A 30mm 0.45um cellulose acetate syringe filter is then attached to the syringe, and the groundwater sample is filtered directly into sample bottles that, if necessary, contain preservative (eg. sample bottles for dissolved cations analyses are preserved with nitric acid (HN0 3) preservative). 1 4 3 Photo 5: Groundwater sampling equipment for arsenic speciation. The groundwater samples are extracted and first prefiltred with a 30mm 0.45um cellulose acetate syringe filter into a clean sample bottle. The fitred sample is decanted into a separate, clean 60mL syringe, and the sample is then syinged through a Supelco LC-SAX S P E ion exchange cartridge which retains As(V) ions from solution but permits the passage of As(lll) ions. Photo 6: Equipment for ammonium analyses in the field. Nessler reagents are mixed with 25mL of filtered groundwater sample in a cuvette and left for a specified reaction time. The HACH DR/2010 spectrophotometer is zeroed with the blank sample, and the groundwater sample is then analysed for concentrations of ammonium. Note: the spectrophotometer was used to measure concentrations of ammonium and ferrous iron in profiles P1-P3, and concentrations of phosphate in profile P3 (Kidd2 site). 144 Photo 7: H A C H 28000-88 Arsenic Test Kit. The reaction vessel is filled with 50mL of filtered sample, a test strip is placed face-down in the lid, and 5 reagents are sequentially added at specified time intervals. After a 30 minute reaction, the test strip is removed and the developed colour is compared to the chart on the test strip bottle (see Photo 8). Photo 8: H A C H 28000-88 Arsenic Test Kit. Upon completion of the reaction procedure, the developed colour on the test strip is compared to the chart on the test strip bottle. 145 Photo 9: Alkalinity titration equipment. The filtered groundwater sample is transfered into the beaker, and placed on the magnetic stir plate. Hydrochloric acid (0.1 N HCI) is added in small increments with the micropipet. The pH of the sample is monitored and recorded, and the titration is complete once the pH is in the range of approximately 2.90. Photo 10: Coring system drive shoe. A core barrel that is equipped with an internal PVC liner is attached to the top of the drive shoe (see Photo 11). Note the gas line that transports C 0 2 gas to the drive shoe, enabling the bottom 6 inches (15cm) to be frozen, thereby increasing sample recovery. 146 148 Appendix C: Sequential Extractions 149 C.I Background For the successful assessment of As in the solid form, in addition to its concentration, the solubility, speciation and mode of retention must be quantified (Loeppert, 2003 #191). This information is necessary in order to determine the release and mobilization potential of As from a given sediment. Chemical speciation can be subdivided into (i) functionally defined species that are defined by their role (e.g. bioavailable), (ii) operationally defined species that are defined by the extractant that removed them from the sediment (e.g. HC1 extractable fraction), or (iii) specific chemical or oxidation state species (e.g. A S H 3 ) (Van Herreweghe, 2003 #202). In the literature, functionally and operationally defined species are commonly referred to as the fraction of an element in the sediment, while the specific chemical or oxidation state of a species is referred to as the phase. Arsenite [As(III)] is in aqueous solution predominantly as H 3 A s O 3 0 in most p H 4 to 8.5 conditions, and arsenate [As(V)] is present as H 3 A S O 4 " or HA .SO4 2 " under most conditions (Loeppert, 2003 #191). Arsenic exists in natural sediments both as soluble species and bound species. The predominant mode of inorganic As bonding under oxidizing conditions involves complexation at Fe, A l and M n oxide surfaces (although strongest with Fe oxides). With Fe oxides (also known as H F O minerals), the complexation of both As(III) and As(V) is predominantly inner sphere complexation (chemisorption), as As forms bidentate, binuclear bridging surface complexes. The results of numerous studies show that both As(III) and As(V) are adsorbed to Fe oxide surfaces through inner-sphere mechanisms (Pierce, 1982 #67); (Jain, 1999 #195); (Fendorf, 1997 #196); and (Sun, 1996 #197). Therefore, the As is considered to be chemically bonded to the surface and therefore is a chemical component of the mineral grain. Lombi et al. (2000) found that the distribution of total As among the various particle size fractions in contaminated soil samples was clay » silt > sand, as a result of the greater surface area and greater Fe oxide content (Lombi, 2000 #200). In general, Lombi et al. also found As to be mainly bound to amorphous and crystalline Fe oxides, with the largest proportion of the exchangeable As released from the sand and silt fractions. The retention of As in sediments is a function of a series of redox sensitive complexation and mineral dissolution and precipitation processes. For example, in an oxic, fluvial 150 environment (i.e. the environment in which the H F O minerals in the Fraser River delta were originally deposited), the following complexation processes occur during deposition: Fe-oxide + H A s 0 4 2 " => Fe-oxide=HAs0 4 + 2 H 2 0 (C. 1) Following deposition and after the onset of reducing conditions, Keon et al. (2001) note that iron-reduction often coincides with As reduction. Under such conditions, the liberated As(V) would reduce to As(III), and readsorb to remaining Fe oxides. In sulphate-reducing environments, As(III) can also be strongly retained by the precipitation of amorphous and crystalline sulphide minerals such as arsenian pyrite and arsenic sulphides such as orpiment (As 2 S 3 ) (Wilkin, 2002 #194). In low p H solutions, Fe monosulphides dissolve to yield metal cations and hydrogen sulphide gas: MeS(s) + 2 H + <=» M e 2 + + H 2S(aq) (C.2) These components of the sulphides in a sediment suite are referred to as the acid-volatile sulphides (AVS) . Under low p H solutions, arsenic sulphides such as orpiment also dissolve to produce As(IU) species via: As 2 S 3 (s) + 6 H 2 0 2H 3 As0 3 (aq) + 3H 2S(aq) (C.3) As 2 S 3 (s) + H 2S(aq) 2AsS 2" + 2 H + (C.4) Orpiment decreases in solubility with decreasing p H . Therefore, as the p H of a solution decreases, A V S such as Fe monosulphides will tend to dissolve and arsenic sulphides such as orpiment will tend to precipitate. In addition to Fe oxide and sulphide minerals, both As(ffl) and As(V) can complex with humic acid amine groups in organic-rich environments. Thus, As cycling in reducing environments can be complicated by a number of complexation and mineral dissolution and precipitation processes. Methods to determine the various solid phases of As in sediment include X-ray adsorption spectroscopy (XAS) , X-ray diffraction (XRD) and sequential extraction procedures (SEPs) (Keon,' 2001 #192). Due to difficulties in assessing the phase of 151 elements associated with amorphous (as opposed to crystalline) minerals within natural sediment samples, X A S and X R D methods are not as practical as the use of SEPs. C.2 Sequential extraction methodology for arsenic The original scheme upon which most SEP methodologies are based is the scheme originally proposed by Tessier (Tessier, 1979 #203). Since its introduction, this scheme has been adapted by a number of different researchers. Many of the SEPs that were developed for As are based upon the extraction schemes for phosphorus, as the two anions share very similar properties. Loeppert et al. (2003) present a discussion of the mechanisms of extraction for soluble As and the various forms of sediment-bound As. A summary of this discussion is presented below. C.2.1 Extraction of soluble arsenic The two most effective extractants for the removal of soluble As are deionized water and 0. 0 1 M calcium chloride CaCl2. During this step, it is critical to ensure that the extractions are conducted under the same redox conditions from which the sample was collected. This will ensure that neither the quantity of dissolved As nor the concentration of the available surface adsorption sites is altered as a result of the process. C.2.2 Extraction of bound arsenic 1. Ligand exchange: This step involves the desorption of As by a competing ligand (molecule that surrounds the metal in a complex ion). Since As(V) and phosphate (PO4 ") are similar in ionicradius and charge, PO4 Vis. a common extractant for this procedure where: Fe-oxide=HAs0 4 + H P 0 4 2 ~ => Fe-oxide=HP0 4 + H A s 0 4 2 " (C.5) However, P0 4 3 " is not as effective for causing the desorption of As(III). Also, Loeppert et al. (2003) note that due to slow kinetic rates, the desorption of As(V) by phosphate is a slow process that can take over 100 hours to reach completion. However, this rate can be both acid and base catalyzed. 152 2. Ligand-enhanced dissolution: This step involves the complexation of the ligand with the surface structural cation [i.e. the Fe(III) ion in an H F O mineral grain] and the dissolution of the metal oxide ligand-binding site at the mineral surface, resulting in the release of the As: Fe-oxide=As0 4 + L" => Fe 3 +=L" + As(aq) (C.6) This reaction occurs in two steps: the rapid adsorption of the ligand at the mineral surface is followed by the slow dissolution of Fe(IU) from poorly crystalline (amorphous) Fe oxides. When used in the dark and at a p H of 3, ammonium oxalate [(COONFLO2H2O)] is an effective extractant for ligand-enhanced dissolution. Because the more crystalline forms of Fe oxides are much more kinetically limited than the amorphous forms, this step permits the selective dissolution of the amorphous minerals. It is critical to perform this step in the dark because, when exposed to light, the more crystalline Fe oxides dissolve due to the photoreduction of Fe(III). This reaction occurs rapidly, within 30 minutes. 3. H +-enhanced dissolution: This step involves the H +-enhanced dissolution of the more amorphous Fe oxides and reactive sulphide minerals [i.e. the acid volatile sulphides (AVS) as opposed to crystalline sulphides] via the following reactions: Fe-oxide=As0 4 + H + => F e 3 + + H 2 0 + As(aq) (C.7) F e ( 1 . x ) A s x S + 2 H + => ( l -x)Fe 2 + + H 2 S + x-As(aq) (C.8) respectively. This reaction rate is initially relatively quite rapid, as the amorphous Fe oxides rapidly dissolve, and the associated As desorbs. This is followed by a gradual readsorption of As, as the remaining Fe oxide approaches equilibrium with respect to the dissolved Fe. As a result, this extraction is highly dependent upon methodological variables such as acid concentration, reaction time, and the soil-to-solution ratio. Therefore, in order to compare data sets between different locations, it is imperative to follow similar analytical methodologies and protocols. 4. OH-enhanced dissolution: Under high p H conditions, the reduction of Fe oxides is enhanced and competitive adsorption of OH" at the oxide surface occurs 153 because the surface potential of the mineral becomes increasingly more negative, causing the competitive anion exchange between As(V) and OH": Fe-oxide=As0 4 + OH" => Fe(OH) 4~ + As(aq) (C.9) Loeppert et al. (2003) note that N a O H (0.1 M ) is an effective extractant of As(V) from Fe oxides. Jackson et al. (2000) included 0.1 M P 0 4 ." as a competing ligand in order to prevent the readsorption of As on crystalline Fe oxides after it has been displaced due to the dissolution of the amorphous Fe oxides. The results suggest that when 0.1 M P0 4 3 " is included in ligand-enhanced dissolution extractant, up to 100% of the As that is liberated from amorphous oxides remains in solution and does not readsorb onto the crystalline Fe oxides. However, the authors note that complications can arise because P0 4 3 " is nonspecific and could displace As that is originally sorbed to the crystalline Fe oxide surfaces. Therefore, all of the As in the extractant from this step cannot necessarily be attributed to the dissolution of amorphous Fe oxides with certainty. Wenzel, Kirchbaumer et al. (2001) also report that N H 4 O H is generally less effective in extracting strongly adsorbed As than N H 4 - H 2 P 0 4 . 5. Reductive dissolution: Reducing agents extract free M n and Fe oxides and the associated As via the following reductive dissolution process: Fe-oxide=As0 4 + e" + L" => F e 2 + = L + As(aq) (C. 10) Effective agents include 0.1 M hydroxylamine hydrochloride (NH20H-HC1) or ammonium oxalate/oxalic acid [(C00NH4)2-H20)/C2H20 4] mixtures at a p H of 2.0. 6. Total digestion: This step uses concentrated mineral acids to determine the total As present in a sample (i.e. the total remaining As after the previous sequential extractions are completed). C.3 Literature review - arsenic SEPs Numerous authors have published results for SEPs for As in sediment (Gleyzes, 2001 #210); (Keon, 2001 #192); (Li, 1995 #204); (Shiowatana, 2001 #198); (Van Herreweghe, 154 2003 #202); and (Wenzel, 2001 #199), with the SEP published by L i et al. recognized as one of the most popular and accepted methods by authors such as van Herreweghe et al. Table 5.1 in Chapter 5 presents a detailed summary of the SEPs presented by the aforementioned authors. Many of these authors note a lack of standard precision and procedure associated with the various SEPs that are employed by different researchers. For example, Keon et al. (2001) report that some of the weaknesses associated with extraction techniques include potential for alteration of the sediment during extraction, and the lack of well-tested extraction techniques. Due to this lack of consistency between the various methods, the Community Bureau of Reference (BCR) formulated a standardized sequential extraction protocol (Van Herreweghe, 2003 #202). However, the results presented by Van Herreweghe et al. (2003) suggest that their more comprehensive SEP was more accurate than the three-step protocol proposed by the B C R . In their paper, Wenzel, Kirchbaumer et al. (2001) discuss a number of different SEP techniques and present a modified SEP that is both adaptable and accurate. Although the authors first began their verification experiments with a thorough 10-step procedure, they condensed their SEP into a 5-step procedure that is appropriate for sediments in which As is associated with hydrous oxides. Of critical importance to any proposed SEP is the consideration of which extractants to use, as they must be appropriate for the project objectives. As discussed earlier, Jackson et al. (2000) report that the inclusion of P0 4 3 " to the extractant during the reductive dissolution of amorphous Fe oxides prevents the readsorption of As on crystalline Fe oxides. Also, in addition to Fe oxides, As(III) is potentially associated with iron sulphides such as arsenian pyrite and arsenic sulphides under low pH, reducing conditions. This process complicates the interpretation of SEP results that incorporate low p H extractants. In contrast, the results of Wilkin and Ford (2002) show alkaline extractant solutions of sodium sulphide and sodium carbonate to be more efficient in removing As from iron sulphides and arsenic sulphides. This is due to the high solubility of As in high p H solutions. Therefore, the authors suggest that alkaline extractants are more efficient than acidic extractants when considering the solid-phase As in sulphidic sediments. 155 Keon et al. (2001) also suggest practical modifications to their published SEP, including the use of more dilute orthophosphate concentrations such as (i.e. 0.1 M ) if the concentration of As in the sediments is expected to be less than 100 mg As kg"1. The authors also note that the use of H F , which is hazardous and dangerous, can be omitted if As oxides or As incorporated into silicates are not likely to be significant sources or sinks. In addition to the determination of appropriate chemical extractants, a number of other practical issues must be considered. For example, in order to minimize any changes to the ambient redox conditions, SEPs must be completed in an anaerobic chamber. Furthermore, a researcher can monitor certain parameters such as solution p H to ensure that a certain extraction step is proceeding under optimal conditions (Van Herreweghe, 2003 #202). Keon et al. (2001) also note that low sediment-to-extractant ratios ensure that the reactivity of extractants is not exhausted during the extraction process. However, Shiowatana, McLaren et al. (2001) note that if the sample weight is too low, then sample non-homogeneity can affect the precision of the extractions. Shiowatana, McLaren et al. (2001) incorporate a continuous-flow SEP as opposed to batch methods. The authors report that in addition to benefits such as simplicity, rapidity, less risk of contamination, and less vulnerability to changes in extraction conditions, of particular importance is the fact that continuous-flow SEPs remove As that is liberated from the system continuously, thereby reducing the opportunity for resorption to occur. 156 Appendix D : Sample calculations for alkalinity titration analyses 157 Sample ID: Date: Time: Normality: sample volume (mL): P1-03 21-Mar-04 2:48 0.0993 25 equation: mL acid: moles acid: moles HC0 3 ' : cone. HCO3" (mg/L): y = 0.0359x - 0.0243 0.677 6.72E-05 6.72E-05 164 HCI volume (mL) PH Gran function 0 6.72 4.76365E-06 0.2 6.41 9.80394E-06 0.3 6.24 1.45586E-05 0.4 6.06 2.21225E-05 0.5 5.83 3.77173E-05 0.55 5.67 5.46249E-05 0.6 5.50 8.09543E-05 0.65 5.20 0.000161841 0.7 4.64 0.000588753 0.75 4.05 0.002294971 0.8 3.76 0.004483526 0.85 3.59 0.006644473 0.9 3.49 0.008381076 1 3.35 0.011613773 1.2 3.15 0.01854818 0.02 s 0.01 \ I 0.005 0 0 Gran Plot P1 -03 = 0.0359X - 0.0243 R2 = 0.9976 > * « >>•< 0.5 1 mL of 0.0993 acid 1.5 Sample ID: Date: Time: Normality: sample volume (mL): P1-05 21-Mar-04 4:45 0.0993 35 equation: mL acid: molarity acid: molarity HC0 3 ' : cone. HCO3" (mg/L): y = 0.0447X • 0.966 9.60E-05 9.60E-05 167 0.0432 HCI volume (mL) pH Gran function 0 6.82 5.29746E-06 0.2 6.59 9.04779E-06 0.4 6.31 1.73382E-05 0.5 6.24 2.04281 E-05 0.6 6.14 2.57899E-05 0.7 5.98 3.73825E-05 0.8 5.78 5.94132E-05 0.9 5.47 0.000121645 1.0 4.70 0.000718294 1.05 4.07 0.003068353 1.1 3.78 0.005991109 1.2 3.50 0.011447445 1.3 3.36 0.015845525 1.4 3.26 0.020003288 1.6 3.13 0.027131955 1.8 3.03 0.034343758 2.0 2.96 0.040569693 0.05 -1-0.04 -§ 0.03 -<S 0.02 -0.01 -0 4-Gran Plot P1-05 y = 0.0447X - 0.0432 R2 = 0.9914 0.5 1 1.5 mL ot 0.0993 acid 2 5 158 Sample ID: Date: Time: Normality: sample volume (mL): P1-16 23-Mar-04 11:10 0.0993 25 equation: mL acid: molarity acid: molarity HC0 3 " : cone. HCQ 3" (mg/L): f = 0.0347X • 1.184 1.18E-04 1.18E-04 287 0.0411 HCI volume (mL) PH Gran function 0.0 6.93 2.93724E-06 0.2 6.86 3.47857E-06 0.4 6.65 5.68635E-06 0.6 6.46 8.87646E-06 0.7 6.34 1.17472E-05 0.8 6.20 1.62787E-05 0.9 6.05 2.30834E-05 1.0 5.86 3.589E-05 1.1 5.62 6.26095E-05 1.2 5.00 0.000262 1.25 4.27 0.001409708 1.3 3.82 0.003980666 1.4 3.50 0.008348413 1.5 3.35 0.011837115 1.6 3.25 0.014958279 1.7 3.17 0.018051415 1.8 3.12 0.020329879 2.0 3.02 0.0257848 2.2 2.95 0.030518902 Gran Plot P1-16 0.04 0.03 0.02 0.01 0 y = 0.0347X - 0.0411 R2 = 0.9908 0.0 0.5 1.0 1.5 2.0 2.5 mL of 0.0993 N acid Sample ID: Date: Time: Normality: sample volume (mL): P2-03 24-Mar-04 1:50 0.0993 30 equation: mL acid: moles acid: moles HCO3": cone. HCO3" (mg/L): y = 0.0359X - 0.0243 0.677 6.72E-05 6.72E-05 137 HCI volume (mL) PH Gran function 0 6.36 1.30955E-05 0.2 6.26 1.65961E-05 0.5 6.14 2.20953E-05 0.75 5.88 4.05364E-05 0.9 5.61 7.58505E-05 1.0 5.35 0.000138472 1.1 4.81 0.000481682 1.15 4.31 0.001525661 1.2 3.92 0.003751065 1.25 3.75 0.005557123 1.3 3.62 0.007508347 1.4 3.45 0.01114114 1.5 3.34 0.014398278 1.6 3.26 0.017365492 1.8 3.14 0.023037064 2.0 3.06 0.027870835 2.2 2.98 0.033717539 Gran Plot P2-03 0.04 I 0.03 § 0.02 I 0.01 ° 0 y = 0.0359X - 0.0243 R2 = 0.9976 • 0.5 1 1.5 mL of 0.0993 acid 2.5 1 5 9 Sample ID: P3-05 Date: 7-Apr-04 Time: 1:33 Normality: 0.0993 sample volume (mL): 25 equation: mL acid: molarity acid: molarity HC0 3": cone. H C 0 3 ' (mg/L): y = 0.0465x 2.095 2.08E-04 2.08E-04 508 0.0974 HCI volume (mL) 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 2.8 PH 6.61 6.55 6.47 6.38 6.30 6.19 6.07 5.97 5.89 5.84 5.73 5.61 5.45 5.16 4.64 3.81 3.46 3.28 3.16 3.07 2.93 Gran function 6.13677E-06 7.10232E-06 8.60664E-06 1.06719E-05 1.29306E-05 1.6787E-05 2.22998E-05 2.82881 E-05 3.41386E-05 3.84487E-05 4.97177E-05 6.57862E-05 9.54448E-05 0.000186794 0.000620825 0.004212781 0.009465916 0.014379724 0.019025352 0.02349141 0.032662152 P3-05 8 -6 • X 4 . 2 n 0.0 0.5 1.0 1.5 2.0 2.5 3.0 mL of 0.0993 acid Gran Plot P3-05 0.04 0.03 0.02 0.01 0 y = 0.0465X - 0.0974 P R 2 = 0.9991 X f 0.0 0.5 1.0 1.5 2.0 2.5 3.0 mL of 0.0993 acid Sample ID: P3-06 Date: 7-Apr-04 Time: 2:05 Normality: 0.0993 sample volume (mL): 25 equation: mL acid: molarity acid: molarity HC0 3 " : cone. HC0 3 " (mg/L): y = 0.0511x- 0.1009 1.975 1.96E-04 1.96E-04 479 HCI volume (mL) 0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.5 1.6 1.7 1.8 1.9 2.0 2.1 2.2 2.3 2.4 2.5 2.6 pH 6.77 6.69 6.61 6.51 6.40 6.29 6.15 6.01 5.93 5.83 5.67 5.52 5.18 4.30 3.66 3.37 3.21 3.10 3.01 2.94 Gran function 4.24561 E-06 5.14518E-06 6.23496E-06 7.91116E-06 1.02712E-05 1.33344E-05 1.85482E-05 2.57991 E-05 3.11348E-05 3.93443E-05 5.70836E-05 8.09347E-05 0.000177727 0.001353206 0.005928834 0.011602963 0.016833044 0.021764594 0.026874024 0.03168904 Gran Plot P3-06 0.04 0.03 2 0.02 ID 0.01 y = 0.0511x-0.1009 R 2 = 0.9995 0.5 1 1.5 2 2.5 3 mL of 0.0993 N acid 160 Appendix E : Certificate for certified reference material T M D A - 5 4 . 3 161 m Canada Cnnud» C E R T I F I E D R E F E R E N C E M A T E R I A L TMDA-54.3 A fortified calibration standard for mce elements N W K t certified faCindards tor K a t e tier mint:** un: m.itii: up hi iihwm't *tui itiuiiwi i :tki* Onninrs w;tt^r n n d a f C i c i V c i witti 0.2% tutu*., .und. The. iwkutinmti tu thrs uinmed iuuk samptt* Ot JOOL rvvmic-- a s<~*rr v-asci with cation concentrations as listed b e i o w . i^us C R M tus i:t.m<:<*imaiion- i n tisw s(st>-sOO n^/l rAnf4> :%nrl s--. Cie<>i£ned k)-: t .»ihb(j iJujn t;nt^>^; Ttar.c f f ln inwHI ^ n t t H r d s *<m n o ! « d ror ih^sr jnlrf^'/tty And C0rK-ir4?>ncv i y l d fii'O mi nitvjinl in iiiilitiix if di ' i il i •> ^t .Mi tP '"* (h*- value* a m i Ma-i tirs *OM thi* C R M ,H< d n.t J n i <#u « i ' « n i * r round rr hm Mxrficis- w h i c h tor this samole v>cie s tud io , &0 ~2 £ ""* il i lw' Sep ' / o v « l-( <nH't>p O'S c. pect vi.lv \ U t » ot the t a b u i a t u i i o (34 w h u h iwr lu nwtwl in I I I I M I t^-\ s u t i l e - ,rr ut «<)r«<t urW i a * . c b i. tu T N I S I im! utht f C R M - . P I H I%*» itui** ih<*t * » \ p s r y ttatMs of 1 y e t M r o m d ot h i p f - i n " 1 asc ut i m ' i i i i u i i t s,ini,ile itrthiliiy, hut M i k c r thai nf s-imple tran«;t>art, band! Int; and stuiaRt. \ U vtoi ivl ' i<* w n m w u l th.r! the C K M he t i f+ i t l y c a p p e d and rerrif,i~rated Imined t H y alu < u,t Andlyli. studies' Results 1, ^ / . ^ ^ Arttcfnortv 2*> t t "•• Ar*.t nt <SS t 7.41. 1, 7 J B>uimrr 1,17 i J . 5 I 0 * HpryHfiam .i. ; 2 n ) r . 1 J / ^ J i i u n t i u M i < j» lw. l l - » " J t i . p j w .. : • > ) 5 2 . : , ± 5 I . S . ^ <>S iron • • ' M . '. 4.5; f> i Kit . S J 3 . * SIM i i itrmins 26.7. ) r 24.a • i.i'i.i • • 3 CM. * /<» u t (. + . <i J 0^ 30.3 i: 6.95 «s Strontium C14. .1: :5 / 5*) Thi'iliiuni .;.>?. o 4,.,i't 3 / 4 2 U F J I H I U M : 1 1 . » :i / 47 Vaividsum 360. ±. 2".a . I 74 Zirsc 5 5 5 . 6 I . 0 • S f 92 1 Ti i« ys% «>nfi<j«nca (n!«rva!'is t'i standard dcviiiiioms, oui!U;rs o l > % Mi l . «w;lticlwl. ' Th« lwckt!<t.H!i>d <-«n(.«rttrations for Ca , M & Na ai)c) K me 6 , 1 , , I 2 .S , w « l 0.S m s t I K S J I W tivniy. liifnmh . . ' (9 . 4 20 i « *u l ls Sii.vcf - 14. . 4;> results T in 22. . ta tvsuilo. January 2 0 0 1 Canadl E . 1 : Certified Reference Material TMDA-54.3 trace element composition M * 'l.f o t I t **i!H 'I It . J. I." It i ^ 162 •Certificate Qualification for Cer i i i ied Reference Mater ia ls Certified Referent Materials (.CRMs) are valuable and neeMsacy roofs for validating •analytic results in'r-psca'rch and routine analysis. In termsof naf tonaland ' • ' ; internal tonal environmental programs for monitoring and.research, CRMs arc essential in the management and trace-ability of the. quality of result's. Tlie.fnstittite CRMs are developed by its' research programs', certified values lor these CRMs arc based on performance based • methods used by laboratories in Environment Cauudii'sintcrfaboraiory i*E prxsgratri,-'NWR.l. warrants that the materials cbrtforot fc> the Certificate values. In the event of a breach ot this warranty. thtrinstituie w i l l only be .liable./ for-a replacement sample, an equivalent substitute., or the. invoice pri«:e of the, O K M . /Disc ia i i t ie r & - y a b i l l i y ' • *., • " ' C R M s are used to verify the accuracy and iraeeabihty of analytic data. This is their primary scienhTte purpose. The. Institute, can only recommend their use when it conforms togood laboratory practicesas-outlined in accreditation liucumctilatiori such as thoseof ISO. A secondary pt'tyose'bf CRMs-is analytic audhmsv The Institute cannot !>f. held responsible in any man iter for airy such use or its outcome. , •. ' The eert i te i t ion and stability of C R M s are • subject to uncertainties. The inherently complex nature of cn vhwrtincHlaJ analysts '-should toe observed'as they may not be. tiompkaely fmr of entire. 'In addition, C R M s arc weessttriiy .shipped by commercial couriers under non>cC!titrolled conditions and storage may be subject to lorcixn and *' unknown customs regulations. Also,.the CRMs should te handled with good, laboratory practices ie ensure HKeir Inte^ rriy. Hence, in no event: will NWRi be, liable for ' dftget, Indlrecfj'speetal incidental or • consequential damages arksinx out of the list* at*, or inability to use tlte material or. dc*'a:mehtanon, nr for loss of nevenue.or profit, even if advised of-the possibility of such damages. - -Storsxc -.All wafer CRMs should refrigerated and stored if) the-dark.. This storage ts required as natural. Wsiter standards are not chemically • preset veil, and adsorption tutd.evaporatkai losses would be- min imized , 1 Sale 1 larfdltitg Water CRMs-are preserved hi the ..usual" manner of natural samples and standards. Normal, good, and. safe -laboratory practices by qualified }'s;rsotinel should be used. Farther information • ' • The Institute promotes its research and findings and.makes.them available te^  the scientific- community. Additional information is available on re-t)tte.*it'. Analytical result*;, any ' continents or suggestions w i l l Ix*. most weicofne.'Difficulties ordiscrepancies «rbut& with lite certified standards should be.. comiiMmica.U'd immediately. • Costtsnettts, concerns, or information ir«{aii-fcs may be addressee! to i l l A lire ma, QA Specialist at-E-mail address: U;ttfy.Alkema(<JK<y^c,ca. E.2: Certified Reference Material TMDA-54.3 trace element composition 163 Appendix F : Sample calculations 164 1. Sample calculation for groundwater velocity (DND site): The groundwater velocity (v) is defined as: v = —, where q = K x i n where q is the Darcy flux, n is the porosity, K is the hydraulic conductivity and i is the hydraulic gradient. Therefore, the velocity of groundwater in the upland areas of the Fraser River delta (i.e. the D N D site) is calculated as: q [(3xl0" 4 m/s) x (0.0002)] _ 7 v = — = 1 1 — — = 2.0 x 10 m/s = 6.3 m/y n 0.3 2. Sample calculation for porewater residence times (DND site): The residence time (T r) can then be calculated over a given flowpath of length L as: TV=i The horizontal groundwater flow velocity within the distributary-channel sand unit of the Fraser River delta is approximately 6.3 m/y. The D N D site is approximately 1.5 km downgradient from the groundwater divide that is proposed by Ricketts (1998). The residence time for groundwater that flows horizontally from the groundwater divide to the D N D site is: V 1500 m 6.3 m/y = 238 years note: this estimate of the residence time does not include the time required for groundwater to migrate to the distributary-channel sand unit from either surface recharge, or recharge from the underlying silty clay unit. 165 3. Sample calculation for porewater residence times (Kidd2 site): The residence time (T r) can then be calculated over a given flowpath of length L as: The horizontal groundwater flow velocity within the saline wedge at the Kidd2 site ranges from approximately 0.5 to 6.3 m/y in the central portion of the wedge to approximately 0.4 to 1.4 m/y at the toe of the wedge (Neilson-Welch 1999). The extent of the saline wedge at the Kidd2 site is approximately 600 m inland from the Fraser River. With groundwater velocity estimates of 3 m/y along 500 m of the flowpath from the Fraser River inland, and 1 m/y along the 100 m within the vicinity of the toe of the saline wedge, the residence time for groundwater within the saline wedge at the Kidd2 site is calculated as: T r = - x 2 = v Where the residence time is multiplied by 2 because the groundwater flows 600 m inland from the river, and then discharges back into the Fraser River. Therefore, a total residence time of approximately 530 years is estimated for groundwater that flows down from the bed of the Fraser River, inland along the basal silty clay unit, and then back up along the transition zone and into the Fraser River. 500 m 100 m + 3 m/y 1 m/y x 2 = 530 years 4. - Sample calculation for the potential concentrations of dissolved As generated by the release of solid-phase As that is associated with HFO minerals The calculation of the potential porewater concentrations generated from 0.16 |ig solid-phase As/g aquifer sedi and a porosity of 0.3, is ment, assuming a sediment density equal to quartz (2.65 g/cm3) 0.16ugAs 2.65gsed. (100) 3cm 3 l m 3 aquifer lLaquifer _ 1 A 1 ~ , n - X X X X — jl^ rJ_0 LI2/1-/ lgsediment l c m 3 sed. l m 3 1000Laquifer 0.3Lporewater 166 5. Sample calculation for the time required to flush the solid-phase As (associated with HFO minerals) from the sediment at the Kidd2 site Assuming a constant porewater concentration ( C o ) , the number of cycles (c) required to flush the aquifer is calculated as C c = C 0 where C m a x is the maximum concentration that could be generated from the solid-phase As. Therefore, the number of cycles necessary to flush the solid-phase As from sediments containing 0.16 |ig solid-phase As/g aquifer sediment, assuming the maintenance of 30 Ug/L dissolved As in the porewater is 1400 ug/L A £ n , c = = 46.7 cycles 30 ug/L Given the estimated residence time (T r) of 530 years for groundwater that is cycling in the saline wedge at the Kidd2 site, the time (t) necessary to flush the H F O mineral-associated solid-phase As from the sediments is calculated as t = 46.7 cycles x 530 years/cycle = 24,750 years 167 

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0052461/manifest

Comment

Related Items