How To, Education & Training

Groundwater arsenic contamination on the Ganges Delta: biogeochemistry, hydrology, human perturbations, and human suffering on a large scale

Groundwater arsenic contamination on the Ganges Delta: biogeochemistry, hydrology, human perturbations, and human suffering on a large scale
of 12
All materials on our website are shared by users. If you have any questions about copyright issues, please report us to resolve them. We are always happy to assist you.
  C. R. Geoscience 337 (2005) 285–296  External geophysics, Climate and Environment Groundwater arsenic contamination on the Ganges Delta:biogeochemistry, hydrology, human perturbations, and humansuffering on a large scale Charles F. Harvey a , ∗ , Christopher H. Swartz a , e , Abu Bohran M. Badruzzaman b ,Nicole Keon-Blute a , Winston Yu a , M. Ashraf Ali b , Jenny Jay a , Roger Beckie f  ,Volker Niedan a , Daniel Brabander a , c , Peter M. Oates a , Khandaker N. Ashfaque a ,Shafiqul Islam d , Harold F. Hemond a , M. Feroze Ahmed b a Parsons Laboratory, CEE, 48-321 MIT, Cambridge, MA 02139, USA b  Bangladesh University of Engineering and Technology, Dhaka, Bangladesh c Wellesley College, 106 Central Street, Wellesley, MA 02481, USA d Tufts University, Medford, MA 02155, USA e Tellus Institute, 11 Arlington Street, Boston, MA 02116-3411, USA f  University of British Columbia, 2329 West Mall Vancouver, BC, Canada V6T 1Z4 Received 3 March 2004; accepted after revision 11 October 2004Available online 8 December 2004Written on invitation of the Editorial Board Abstract Over the last several decades, much of population of Bangladesh and West Bengal switched their water supply from surfacewater togroundwater. Tragically, much of theregion’s groundwater isdangerously contaminated byarsenic, and consumption of this water has already created severe health effects. Herewe consider how groundwater flow may affect arsenic biogeochemistryand we compare the vertical patterns of groundwater chemistry at our intensive study site with the average values across thecountry. Detailed hydraulic data are presented from our field site that begins to characterize the groundwater flow system.  To cite this article: C.F. Harvey et al., C. R. Geoscience 337 (2005). © 2004 Published by Elsevier SAS on behalf of Académie des sciences. RésuméContamination par l’arsenic des eaux souterraines du delta du Gange : biogéochimie, hydrologie, perturbationshumaines et conséquences pour l’homme à grande échelle.  Pendant les dernières décennies, la plus grande partie de lapopulation du Bangladesh et de l’Ouest du Bengale sont passées d’une alimentation en eau potable par les eaux de surface à * Corresponding author.  E-mail address: (C.F. Harvey).1631-0713/$ – see front matter  © 2004 Published by Elsevier SAS on behalf of Académie des sciences.doi:10.1016/j.crte.2004.10.015  286  C.F. Harvey et al. / C. R. Geoscience 337 (2005) 285–296  l’utilisation des eaux souterraines. De façon dramatique, la majorité de ces eaux souterraines est dangereusement contaminéepar de l’arsenic, et la consommation de cette eau a déjà eu des conséquences sanitaires très sérieuses. La comparaison desprofils verticaux de la composition géochimique des eaux sur notre site de recherche intensive avec les valeurs moyennes pourle pays suggère que les propriétés générales du système souterrain seraient déterminantes pour expliquer certains aspects de lamobilisation de l’arsenic. Des données hydrauliques détaillées obtenues sur le site expérimental sont présentées, qui permettentd’aborder la caractérisation du système d’écoulement souterrain.  Pour citer cet article:C.F. Harvey et al., C. R. Geoscience 337 (2005). © 2004 Published by Elsevier SAS on behalf of Académie des sciences. Keywords:  Arsenic; Bangladesh; Bengal; Aquifers; Biogeochemistry  Mots-clés :  Arsenic; Bangladesh ; Bengale ; Aquifères; Biogéochimie 1. Introduction Over the last several decades, much of populationof Bangladesh and West Bengal switched their watersupply fromsurface water to groundwater.As many as10 million new domestic wells were installed, provid-ing drinking water for over 100 million people. Thislarge-scale transition to groundwater as the source fordomestic water was motivated by the necessity of pro-viding water free of pathogens – diarrheal diseasessuch as cholerawereinfectingmillions ofpeople.Thistransition to well water was readily adopted becauseof the convenience of having a water supply in closeproximity to homes and the ease of drilling in the re-gion’s high-yielding aquifers. At the same time thatdrinking wells were drilled, irrigation wells were alsoinstalled across the country.Groundwaterpumpingforirrigation greatly increased food production enablingBangladesh to become self-sufficient in food, eventhough the population nearly tripled over the last fourdecades. Irrigation is necessary for the newly intro-duced dry-season rice called ‘Boro’ that now providesmore yield than the traditional rice grown during thewet season, and Boro rice cultivation and irrigationincreased together from 1970 to 2002 (Fig. 1). Thus, issues of groundwater quality and quantity have be-come vital for both the supply of drinking water andthe production of food in Bangladesh.Tragically, much of the region’s groundwater isdangerously contaminated by naturally occurring ar-senic, and consumption of this water has already cre-ated severe health effects. About half of the wells inthe countryhaveconcentrationsgreaterthan10µgl − 1 ,now a common standard. Much of our understand-ing of the distribution of arsenic across Bangladesh’s Fig. 1. Cultivation of high yielding Boro rice has greatly ex-panded over the last several decades to cover approximately 20%of Bangladesh, or approximately 45% of the cultivatable area. MostBoro is irrigated by groundwater so extraction has also risen. Datataken from [10].Fig. 1. La culture de riz Boro à haut rendement s’est fortement éten-due sur les dernières décennies, pour couvrir 20% de la surface duBangladesh ou 45% de la surface cultivable. La majorité des terresà riz Boro est irriguée par de l’eau souterraine, dont l’extraction adonc aussi augmenté. D’après [10]. groundwater comes from the comprehensive work of the British Geological Survey (BGS) [6,12]. Theirwork shows that high arsenic concentrations may befound throughout the flood plains and delta of theGanges Brahmaputra and Megnha Rivers, but theDelta region of the southern half of the country isthe most contaminated. Yu et al. [20] combine theBGS’s database with dose-response models to esti-mate that, if consumption of contaminated water con-tinues, the prevalence of arsenicosis and skin cancerin Bangladesh will be approximately 2000000 and100000 cases per year, respectively, and the incidenceof death from cancer induced by arsenic will be ap-proximately 3000 cases per year. Because detailedhealth records are not kept in Bangladesh, these esti-mates were made using dose-responsecurves from the  C.F. Harvey et al. / C. R. Geoscience 337 (2005) 285–296   287 literature. How accurately these dose-response rela-tionships apply to the broad population of Bangladeshremains an open question and existing epidemiologi-cal surveys show a wide spread of arsenicosis preva-lence estimates [20]. 2. Biogeochemistry and arsenic mobility Researchers largely agree that dissolved arsenic inthe groundwater of Bangladesh srcinates from thesediments. However, there is no evidence of wide-spread, unusually high, levels of solid phase arsenicin the aquifer material – concentrations are typicallyless than 10 ppm in sandy sediment and less than100 ppm in clays and peats [14–17]. High solid-phase concentrations have been reported in the soils of ir-rigated fields, but these could be the result of ar-senic input from groundwater irrigation and sorptionto the soils [1,13].Thus,it appearsthat highdissolved- arsenic concentrations are the result of particular hy-drologic and biogeochemical conditions that partitionarsenic from the solid to aqueous phase, but have notyet flushed dissolved arsenic from the subsurface.The reducing conditions of almost all groundwaterin Bangladesh (demonstrated by high levels of dis-solved ferrous iron and methane, and low measure-ments of Eh), and the weak but statistically-significantpositive correlationof dissolved arsenic to iron and bi-carbonate, suggest that most arsenic is liberated bydissolution of iron (oxi)hydroxides, or perhaps des-orption of arsenic after reduction from arsenate to ar-senite [6,8,16]. The low concentrations of sulfate (and in some areas the negative correlation between ar-senic and sulfate) as well as the generally reducingconditions indicate that arsenic has not been directlymobilized from sulfide minerals (e.g., [9]). However,the arsenic in the Ganges Delta sediments likely src-inated from sulfide minerals that weathered out of the granitic and metamorphic source rock of the Hi-malayas, and it remains a possibility that at the landsurface, where oxygen is introduced as the water tablerises and falls, sulfide minerals could be oxidized anddissolved thereby liberating arsenic. The long-termimplications of such cyclical near-surface processes ina rapidly accreting  ( ∼ 1 cm/yr )  aquifer remain to bestudied.Several research teams [6,8,15,19] describe two distinct types of aquifer sediment: brown (or or-ange to yellow) sediment presumably containing iron(oxi)hydroxides where dissolved arsenic concentra-tions are low, and grey sediments where dissolvedarsenic concentrations may be high. The brown sed-iments are found at depth in the older Pleistoceneaquifers such as the Dupi Tilla formation, where low-arsenic water is obtained, as well as near the surface.Dissolved arsenic is presumably low in these sedi-ments because of the capacity of iron (oxi)hydroxidesto adsorb arsenic. Islam et al. [11] showed that ar-senic is liberated from sediments collected in WestBengal by the addition of organic carbon. They do notreport the in-situ arsenic concentration in the pore-water, but the sample contains iron (oxi)hydroxidesand is described as coming from a transition zone be-tween a region of oxidizing conditions and a regionwith reducing conditions. The role played by iron(oxi)hydroxides within the contaminated grey sedi-ments ofthe Holoceneaquifer,where mostwells with-draw water, is much more enigmatic. Iron (oxi)hydro-xides must exist, or have existed veryrecently,accord-ing to the theory that arsenic is released from iron(oxi)hydroxides in local sediments by organic car-bon oxidation. However these iron (oxi)hydroxideshave not been definitively demonstrated in the greysediment and high concentrations of methane and hy-drogen [8] in strongly reducing water indicate thatgeochemical conditions are not conducive to stabil-ity of iron (oxi)hydroxides.On the other hand, Swartzet al. [17] show that only small quantities of iron(oxi)hydroxides would be required to explain currentgeochemical conditions, and McArthur et al. [15] pro-vides a geologicexplanationforwhythe GangesDeltasediment would have been deposited with relativelylittle iron (oxi)hydroxides. Thus, it is conceivablethat slow reductive dissolution within aquifer sedi-ments could be responsible for high dissolved arsenicconcentrations, but only if the geochemical systemhappens to be in a state where iron (oxi)hydroxideshave released almost all of their sorbed arsenic. Inother words, the aquifer sediments must be poisedin a geochemical state where the inventory of iron(oxi)hydroxides is nearly (or recently) exhausted, yetarsenic has not been flushed away by flowing ground-water. Other explanations, which we explore below,are that both the physical flow system and the bio-  288  C.F. Harvey et al. / C. R. Geoscience 337 (2005) 285–296  geochemical system have recently been perturbed,and that dissolvedarsenicsrcinatesfromnear-surfacesediments above the aquifer.Dissolved arsenic concentrations are maintained ingrey sediment because several geochemical factorsconspire to prevent arsenic that has been dissolvedfromsorbingbackontothisaquifersediment.First,thepaucity of ferric (oxi)hydroxides means there are fewadsorption sites. Second, high concentrations of otheranions, such as silicate and phosphate, which competewith arsenic for surface sorption sites, are prevalent ingroundwater throughout most of the arsenic-affectedareas. (However,thereis noconvincingcorrelationbe-tween these anions and arsenic to indicate that theseanionsexplainthe spatial patternof dissolvedarsenic.)Appelo et al. [5] have also suggested that competitionby bicarbonate, which correlates better with arsenicover the country, might explain the distribution of dis-solved arsenic. By this scenario, oxidation of organiccarbon liberates arsenic indirectly through desorptioncaused by its byproduct, bicarbonate, rather than di-rectly by reduction of iron oxides or arsenate. How-ever, equilibrium chemical modeling using the para-meters measured at our site indicates that the effectof bicarbonate on arsenic sorption is less than that of silicate and no more than phosphate [17]. These con-ceptual geochemical models are further complicatedby the fact that arsenic likely adsorbs to surfaces of many solid phases other than oxihydroxides, such asmagnetite, green rust, and potentially siderite and ap-atite. Arsenic is known to sorb readily to magnetite[7] and the results of our density and magnetic separa-tions show that the magnetite fraction has the highestarsenic concentration by weight [17].Several research groups postulate that irrigationpumping may flush arsenic from aquifers [8,15]. Har- vey et al. [8,9] support this contention by comparing concentrations sampled from irrigation wells to con-centrations from drinking water wells to show thatirrigation wells, which flush much greater quantitiesof water, have significantly lower arsenic concentra-tions. At a national scale, Ali [4] estimated, using1996 irrigation data, that each year, groundwater ir-rigation removes from aquifers, and then applies tofields, 1 kg of arsenic per hectare of irrigated area.Thus, in 2001 (Fig. 1), irrigation pumping would haveextracted much more than a million kilograms of ar-senic per year from the aquifers and moved it into ricefields. On the other hand, some evidence suggests thatarsenic concentrations may rise after pumping com-mences. Kinniburgh et al. [12], van Geen et al. [19], and McArthur et al. [15] all provide strong statisti-cal evidence that arsenic concentrations in domesticwell water correlate to the age of the well, suggestingthat arsenic concentrations may rise after a well is in-stalled, perhaps because irrigation wells, which havemuch greater effects on the local groundwater system,are installed in the region at the same time as the do-mestic wells where arsenic is measured. Can these ap-parently contradictory suggestions of both falling andrising arsenic concentrations be reconciled? Clearlypumping removes some arsenic from the aquifer. (Infact, irrigation pumping can be viewed as analogousto‘pump-and-treat’groundwaterremediationmethodsemployed in North America and Europe, but withoutthe ‘treat’ and with extraction rates that are actuallyhigher than at many sites!) However, increased flush-ing may be concurrent with increased arsenic inputeitherby simple transportor releasecaused byinputof organiccarbonfromthe surface sources,or potentiallydrawn from peat layers. This concurrent enhancementof both sinks and sources of arsenic to the ground-water system by human perturbation could potentiallycreate very complex temporal and spatial behavior of dissolved arsenic.At our site in Munshiganj, processes appear to becompeting to both increase and decrease arsenic con-centrations: arsenic is being extracted from the sys-tem and radiocarbon dating of dissolved carbon in-dicates that arsenic has been mobilized recently [8].The radiocarbon data show that detrital organic car-bon has not driven recent biogeochemical reactions.The byproducts of microbial activity, both inorganiccarbon and methane, have much younger dates thanthe dissolved organic carbon or the sediment, and theconcentration of this inorganic carbon is much largerthan that of the older organic carbon. In fact, at 20-mdepth, the inorganic carbon has levels of carbon-14higher than 100% modern. This is carbon from bombtesting, so it entered the aquifer in the last 50 years.At three 30-mwells, dissolvedinorganiccarbon(DIC)radiocarbon ages are 462, 770, and 823 years, muchyounger than the local sediment and the radiocar-bon age of dissolved organic carbon (DOC) (3636,1538, and 1890 years). Tree roots and burrowing an-imals are unlikely to penetrate below 10 m because  C.F. Harvey et al. / C. R. Geoscience 337 (2005) 285–296   289 the aquifer remains saturated all year. (Later, for thehydrologic model, we will consider whether roots invillages maypenetratethrough6 mof claytoreachtheaquifer.) Thus, dissolved carbon with a radiocarbonage younger than the sediment age was transporteddownward, and laterally, by flowing groundwater. Thepresence of young DIC and old DOC in the same wa-ter does not appearto result from the mixingof young,DIC-containingwater andold, DOC-containingwater.The concentrations correlate strongly (i.e. water highinyoungDICis alsohighinoldDOC);theydonotfol-low a mixing-line that would have a negative correla-tion.ThusitappearsthattheolderDOCwas mobilizedfrom the sediment concurrentlywith the productionorinflow of young DIC. McArthur et al. [15] argue thatburied peat deposits have provided the organic carbonthat drives reduction at our field site, but they do notattempt to reconcile the different radiocarbon ages of dissolved organic carbon and inorganic carbon. 3. Geochemical profiles with depth In this section, we consider how geochemical char-acteristics vary with depth in aquifers, and hence howchemical conditions relate to flow paths and ground-water age. Fig. 2 compares depth profiles of soluteconcentrations measured at our field site in Munshi-ganj with averaged values from the BGS and DPHE[6] dataset. Our site in the Munshiganj district (Fig. 4) is located 30 km south of Dhaka and 7 km north of the Ganges. It contains a small intensive-study area(100 m 2 ) with 25 sampling wells that extract waterfrom depths ranging between 5 and 165 m below theland surface. We also monitor water levels at 87 otherlocations in the surrounding 16-km 2 region. We de-scribe some similarities between results at our sin-gle site and the averaged national dataset that suggestsome general characteristics of geochemical evolutionand transport across the region. 3.1. Arsenic as a function of depth At the site in Munshiganj, dissolved arsenic hasa distinct peak at approximately 30 m depth, but wefind no chemical characteristic of the solid sedimentto explainthispattern[8,17].Dissolvedcomponentsin the groundwater indicate an arsenic source that is hy-drologically upgradient. Furthermore, several types of data, when taken together, suggest a relation betweenthe arsenic peak and groundwater flow patterns. Thehydraulic conductivity data (Fig. 3A) indicate varia- tions less than an order of magnitude, but suggest thata lower conductivity layer at 22 m may work to sep-arate horizontal flow paths. The head data (Fig. 3B)show that, at least during some times of year, there isconvergent vertical flow that mixes water from aboveand below 30 m, thus horizontal flow must accelerateto conserve mass at this depth. Evidence for ground-water mixing at 30 m is supported by the  18 O profile(Fig. 3C); the range of isotope ratios at 30 m is con-sistent with mixing of lighter water from above andheavier water from below. The heavier water below30 m could represent infiltrated pond, river or rice-field water that has been subject to relatively moreevaporation. Measurable tritium values are found toa 60-m depth [9], indicating the presence of at leasta component of water that is less than 40 years oldthroughout this depth interval. Furthermore, the tri-tium values show a sharp decrease to less than 1 T.U.below 24 m, so that the peak of high dissolved arseniccorresponds with the depth where older water mixeswith younger recharge.Kinniburgh et al. [12] and McAthur et al. [15] both describe typical depth profiles of arsenic concentra-tions as ‘bell shaped’. Although this pattern is not ob-viously evident from the national dataset (Fig. 2), it isfound at a variety of study sites [8,15]. Depth trends can also be considered within different geologic re-gions, and Yu et al. [20] tabulate the geologic regionsof Bangladesh where there is a statistically signifi-cant trend of decreasing arsenic with depth. (they didnot consider non-monotonictrends). Their geostatisti-cal analysis shows that the trend of decreasing arsenicconcentrations with depth explains much of the differ-ences in arsenic concentrations between nearby wells– neighboring wells often have different arsenic lev-els because one withdraws water from deeper in theaquifer where arsenic concentrations are lower. 3.2. Sulfate, calcium and ammonium as a function of depth At the Munshiganj site, the inverse relation of dis-solved sulfate with As (Fig. 2), and the presence of 
Similar documents
View more...
Related Search
We Need Your Support
Thank you for visiting our website and your interest in our free products and services. We are nonprofit website to share and download documents. To the running of this website, we need your help to support us.

Thanks to everyone for your continued support.

No, Thanks

We need your sign to support Project to invent "SMART AND CONTROLLABLE REFLECTIVE BALLOONS" to cover the Sun and Save Our Earth.

More details...

Sign Now!

We are very appreciated for your Prompt Action!