Arsenic source and fate at a village drinking water supply inMexico and its relationship to sewage contamination J.M.. High concenconcen-trations of Cl, K, Na, NO3and SO4 suggest sewage
Trang 1Arsenic source and fate at a village drinking water supply in
Mexico and its relationship to sewage contamination
J.M Cole & M.C Ryan
Department of Geology and Geophysics, University of Calgary, Calgary, Alberta, Canada
S Smith
Caminamos Juntos para Salud y Desarrollo, A.C Cuernavaca, Morelos, México
D Bethune
Department of Geology and Geophysics, University of Calgary, Calgary, Alberta, Canada
arsenic and other metal exposures from shallow well water Arsenic is present in the dissolved phase
up to a concentration of 37g/L, exceeding the World Health Organization guideline of 10 g/L Stable isotope analyses (2Hwater& 18Owater) indicate precipitation is recharging the wells through the soil and shallow groundwater Soil and rock analyses show that arsenic is present at concen-trations up to 56 mg/kg and 26 mg/kg, respectively High concenconcen-trations of Cl, K, Na, NO3and SO4
suggest sewage or manure contamination in the well water occurs via surface runoff into the unpro-tected wells or through the thin soil and shallow groundwater Open air excretion and free-roaming animals are sources of this contamination Arsenic and sewage contaminant concentrations are strongly correlated and the presence of sewage apparently promotes the release of arsenic from aquifer materials It is likely that arsenic mobilization is the result of desorption associated with arsenate-phosphate competition for sorption sites
Tlamacazapa (Tlama) is located in the Sierra Madre del Sur of the state of Guerrero, roughly 160 km southwest of Mexico D.F and 10 km east of the historical silver-mining town of Taxco de Alarcón
It has a population of approximately 9000, living in high-density housing (mixture of concrete hous-ing and simple cornstalk/cedar branch structures) extendhous-ing up a steep mountain incline Sanitation
is a critical problem due to inadequate sewage management, with most households using open-air excretion methods as opposed to constructed sanitary facilities
Symptoms of low level, chronic metal toxicity were recognized among an undernourished, inad-equately hydrated and poor population by health personnel working in Tlamacazapa as part of a small, international non-governmental organization Caminamos Juntos para Salud y Desarrollo (CJ;
or Walking Together for Health and Development) Hair and nail arsenic concentrations reach 23 and 290 mg/kg by dry weight, respectively, when tested on symptomatic persons admitted to hos-pital Observed arsenic-related symptymology includes skin pigmentation changes (hyperpig-mentation), neurological disorders and abdominal pain Symptoms fluctuate with seasonal changes
in well water levels and apparently worsen following periods of high rainfall A multi-disciplinary research programme was initiated to study the environmental source of arsenic and other metals (especially lead), together with other contamination vectors, including soil, clay cooking pots, and palm dye used in basket weaving
As part of this multi-disciplinary research programme, the objectives of the present study are: (1) to geochemically characterize the local hydrological system, (2) to investigate the source(s) of
Natural Arsenic in Groundwater: Occurrence, Remediation and Management –
Bundschuh, Bhattacharya and Chandrasekharam (eds)
© 2005, Taylor & Francis Group, London, ISBN 04 1536 700 X
Trang 2arsenic in the groundwater and (3) to assess geochemical controls on the mobility of arsenic in the aqueous environment Water samples were collected during both wet and dry seasons for major ion, trace metal, and stable isotope analyses Sample locations include drinking water sources for Tlamacazapa Regional water wells and springs were sampled in order to determine the regional extent of the contamination
2.1 Geological characteristics
The study area is located in the Guerrero Terrane of southern Mexico (Campa & Coney 1983) Basement rock of this terrane is composed of schists of the Tierra Caliente Complex and is overlain
by Mesozoic marine and continental sequences (Fig 1) Massive platform carbonates of the Morelos Formation and thin-bedded calcareous shales, sandstones and conglomerates of the Mexcala Formation dominate these Mesozoic sequences
Post-depositional deformation associated with the late Cretaceous Laramide Orogeny affected the Mesozoic sequence extensively and the basement sequences locally Post-Laramide stratigraphy includes intrusive and extrusive volcanic sequences of the Tertiary Volcanic Province of Southern Mexico These volcanics range in age from late Cretaceous to Pleistocene and are associated with
a magmatic arc extending from north of the study area to the coastal regions (Morán-Zenteno
et al 1998) Thin (1 m), poorly developed soils contain abundant unweathered rock fragments Mining in Taxco de Alarcón began in pre-Hispanic times with the exploitation of silver and gold Currently, mining is focused on the recovery of lead, zinc and to a lesser extent silver Hydrothermal metal deposits occur as replacement veins hosted in metamorphic basement rock, Mesozoic marine
TRANS-MEXICAN VOLCANIC BELT
Hiatus in Tertiary Volcanic Activity
TERTIARY VOLCANIC PROVINCE OF SOUTHERN MEXICO BALSAS FORMATION
Coarse continental clastics interbedded with extrusive volcanics
Regional Angular Unconformity
MEXICALA FORMATION Calcareous shales, continental sub-lithoarenites and conglomerates
CUAUTLA FORMATION Limestones and mudstones with interbedded chert
MORELOS FORMATION Platform carbonates
Regional Angular Unconformity
TIERRA CALIENTE COMPLEX Roca Verde Taxco Viejo Taxco Schist
Figure 1 Simplified stratigraphic succession in the study area.
Trang 3sequences, and to a lesser extent in Tertiary intrusives (Salas 1991) Documented ore mineralogy includes source minerals for arsenic such as pyrite, arsenopyrite, and proustite as well as a host of other metal-bearing sulphides (Salas 1991, de Csersna & Fries 1981) As well, sulphide mineraliza-tion hosted in the Coxcatlán granodiorite about 3 km east of Tlama has been prospected and found
to be economically unfeasible Arsenian pyrite is noted to be present in this mineralization (Campa & Ramírez 1979)
2.2 Soil and bedrock geochemistry
As part of their multi-disciplinary research programme, CJ completed bulk chemical analyses of soil and bedrock from Tlamacazapa In total, 38 composite soil samples (surface to bedrock) and two profiles were collected to provide a good geographical sample of the community as well as sites chosen based on reported metal symptoms Bedrock samples were chosen from prominent geo-logical formations of the Tlama area
Average soil results show consistently high concentrations of arsenic, iron, manganese and aluminum (Table 1) Soil arsenic concentrations are elevated with respect to world background soil concentrations (Boyle & Jonasson 1973) High aluminum and iron concentrations are typical of soil laterites developed in tropical climates Bedrock results show a large variation between rock types, with elevated arsenic, copper, and strontium in the vein deposits suggesting the presence of sulphide mineralization
2.3 Water resources
Precipitation in Tlamacazapa occurs seasonally, with arid conditions prevailing roughly from December to May Drinking water supply is provided by up to eight shallow wells (depending on the season) excavated into existing fractures in the limestone (Fig 2) Low concrete walls surround-ing the wells limit direct surface runoff, with stairs providsurround-ing access for water fetchsurround-ing by hand
An alternate source of water called Los Sabinos is located 5 km northeast of Tlama (Fig 2) Water
is collected and pumped uphill through three pump-and-hold stations and held in a storage tank before gravity distribution through a rudimentary piping system This inadequate and undepend-able system pumps water to each neighbourhood roughly every nine (dry season) to 15 days (wet season) The water source is a shallow aquifer (1 m below ground surface) located in an agricultural valley
A total of eight local, ten regional, and four surface water samples (Fig 2) were collected at the end
of both the wet (December, 2002) and dry (May, 2003) seasons for determination of field param-eters (Temperature, pH, Eh, DO, EC), and analyses of major ions, trace metals and environmental isotopes (2Hwater, 18Owater, 13CDIC, 34SSO4) Trace metal samples were collected for total dissolved and total recoverable analysis, with filtered (0.45 micron) and raw water preserved with spec-pure nitric acid to a pH below 2 Arsenic analyses were done by ICP-MS (Perkin Elmer 6001)
Table 1 Bulk metal analyses (in mg/kg) from different rock units (one sample each) and
average soil composition (53 samples) (Cole 2004).
Trang 44 RESULTS AND DISCUSSION
4.1 Groundwater geochemistry
Ground water is near-neutral pH (6.7 to 7.7) and mildly oxidizing (Table 2) Oxidized species include nitrate, phosphate, and sulphate, and redox potential (Eh) ranges from 170 to 220 mV Mean water temperatures reflect average annual air temperatures of 21°C Well water geochemistry is domin-ated by Ca, Mg, and HCO3, as expected for waters interacting with carbonate rocks and/or cal-cium-rich soils
High concentrations of Cl, K, Na, NO3-N, and SO4, and high electrical conductivity (EC) values (Table 2) indicate sewage contamination in four lowest-elevation Tlama wells (Fig 3) The com-position of the most impacted wells is comparable to septic system effluent and domestic wastewater (Table 2) Nitrate values exceed the World Health Organization (WHO) drinking water guideline (11.3 mg/L NO3-N; WHO 1998) in the four topographically lowest wells in Tlama, while the higher elevation wells have concentrations below the guideline The presence of sewage impact in lower ele-vation wells (Fig 3) suggests that Tlama is the source of its own contamination
A mixing line between uncontaminated (higher elevation) wells and sewage impacted wells is evident by increasing Na K, Cl, and SO4concentrations and relatively high estimated total dis-solved solid concentrations (Fig 4) In some instances, Tlama well water is more concentrated than domestic sewage (Table 2) This is likely due to the lack of dilution by grey water (e.g toilet flush-ing and wash water) Wet season samples are more concentrated than dry season samples in all sewage-related parameters The elevated wet season concentrations are likely related to increased recharge, with active sewage transport to the wells through the thin soil and shallow groundwater Figure 2 Location of study area and water sampling locations Inset shows location and elevation of wells sampled in Tlamacazapa
Trang 5The carbonate system is the dominant control on major ion concentrations in the unimpacted wells Most waters are at or near equilibrium with calcite and dolomite, where equilibrium is defined as a saturation index [SI Log (IAP/K)] range of 0.5 to 0.5 A seasonal variation in saturation indices exists, with wet season samples showing a strong tendency towards undersaturation with respect to calcite while dry season samples are all saturated (Cole 2004) This seasonal variation is likely related
to increased recharge during the wet season, resulting in dilution and less water-rock interaction
Table 2 Dissolved arsenic, field parameters and major ion concentrations for wet season samples from Tlamacazapa drinking water wells, Los Sabinos, and regional wells All values in mg/L unless noted.
( g/L) (°C) pH (mV) DO (S/cm) Ca Mg K Na Cl NO 3 -N PO 4 SO 4 HCO 3
Tlama water wells
Colontsintla – 18 7.5 197 0 740 99 43 2 4 10 22 0.2 78 220 Well 2** 22 21 6.7 178 6 1530 120 54 69 79 110 85 2 135 260 Well 3a** 32 21 7.0 192 1 1080 110 48 49 63 95 65 3 115 340 Well 3b** 37 21 7.7 209 6 1040 110 47 51 49 98 51 0.2 120 290 Well 4** 27 21 6.7 170 1 1270 130 54 63 66 145 68 1 160 460
Regional water wells
Taxco el – 24 6.3 227 1 1037 161 11 1 10 42 27 0.3 205 190 Viejo
Juliantla – 22 6.6 173 1 994 138 17 1 19 64 22 0.2 211 243 Coxcatlán 35 18 6.9 181 1 1062 179 18 14 36 84 22 – 193 437
Regional spring waters
Typical –*** – 7.2 – – 1336 57 17 15 112 100 2.6 11 129 241 sewage
composition
Note: * Results below analytical detection limit; ** Lower elevation Tlama wells; *** Data on As is not available from sources Typical sewage composition is included (Kim et al 2002, Robertson et al 1991, Crites & Tchobanoglous 1998, City of Calgary unpublished data).
0 10 20 30 40
0 40 80 120 160
elevation (masl) Figure 3 Relationship of elevation with As and Cl concentrations.
Trang 64.2 Arsenic geochemistry
Arsenic is present in five of the nine drinking water sources of Tlamacazapa Four of the five wells used for drinking water contain arsenic concentrations exceeding the WHO drinking water guide-line (10g/L; WHO 1998) and the maximum concentration is 37 g/L Although concentrations are not present at levels similar to other affected areas of the world (i.e Bangladesh, Argentina), it is believed that detrimental health effects occur at exposure to concentrations less than 50g/L (WHO 2001) Malnutrition and dehydration are believed to compound the adverse toxic health effects Total and dissolved arsenic results correlate closely (R2 0.997) suggesting that arsenic is pre-sent primarily in the dissolved phase Arsenic concentrations are also correlated to sewage contam-ination (Figs 3 and 4), with positive correlations including Cl (R2 0.83), NO3-N2(R2 0.73),
K (R2 0.82), SO4(R2 0.72) and HCO3 (R2 0.93) (wet season, Tlama water wells) These correlations are stronger during the wet season, which is consistent with increased sewage impact during the period of higher ground water recharge
Among the sewage-related anions introduced into the Tlama wells, PO4, SO4, Cl and NO3 com-pete with arsenic for sorption sites (Manning & Goldberg 1996 a,b) The relative affinity for sorption
to Fe (III)-oxide surfaces is PO4 SO4 Cl NO3(Rau et al 2003) Phosphate is the best known competitor to arsenic, and displays similar geochemical behaviour to arsenic oxyanions Phosphate concentrations decrease arsenate adsorption over a large pH range (2–11) (Manning & Goldberg 1996a, b) Arsenic mobilization by competitive sorption due to phosphate fertilizer application has been observed in agricultural soils (Peryea & Kammereck 1997) Phosphate is also known
to strongly adsorb to carbonate minerals (De Kanel & Morse 1978) Although phosphate is considered the most important competitor, it is likely that all anions introduced through sewage contamination are involved as they are present at concentrations much higher than phosphate The poor correlation of As and PO4 (R2 0.22) in Tlama well waters, combined with the cor-relation of arsenic and other sewage-derived parameters, suggests that arsenic mobilization is
Figure 4 Piper plot of Tlama wells and typical sewage concentrations (Kim et al 2002, Robertson et al 1991, Crites & Tchobanoglous 1998, City of Calgary unpubl data) Circle diameter of central diamond indicates total dissolved solids High arsenic locations indicate concentrations 10 g/L.
Trang 7occurring by PO4-As competitive sorption Phosphate is sorbed and arsenic desorbed during the infiltration of phosphate-rich sewage impacted recharge through soil and shallow groundwater The consequent immobilization reduces its aqueous concentration with respect to the other sewage contaminants
It is also possible that faecal organics are driving the reduction of iron oxyhydroxides and the subsequent release of bound arsenic (McArthur et al 2004) This is not believed to be the domin-ant process liberating arsenic because of the prevalent oxidizing conditions and lack of dissolved iron in As-rich waters
Arsenic is present at five of eight regional water wells and springs at an average concentration
sewage contamination (Cole 2004) At these regional locations, arsenic in drinking water is a con-cern for lower income families, as higher socio-economic levels allow the majority of residents to purchase clean drinking water
4.3 Stable isotopes
Stable isotope results are consistent with groundwater recharge by precipitation Tlama well water
2H and 18O plot closely with the local meteoric water line (LMWL) for Mexico City (Cortés et al 1997) (Fig 5) Dry season waters may be slightly more evaporated than wet season, as would be expected
Arsenic is present in the drinking water supply of the village of Tlamacazapa, Mexico at levels above
bedrock-hosted well waters are 13g/L with a maximum concentration of 37 g/L Although arsenic is not present at levels similar to other affected areas of the world, its concentrations are significant enough
to be contributing to the observed adverse toxicological effects in the undernourished residents of Tlamacazapa
Elevated levels of arsenic in the bedrock and soil of Tlamacazapa suggests that the source of arsenic contamination to the well water is geologic Arsenic is present in well water with geochemical conditions (an oxidizing environment with neutral pH) that suggest significant immobilization by
-75
-70
-65
-60
-55
-50
2 Η ‰
δ 18 Ο ‰
y = 7.97x + 11.03
Figure 5 18 O vs 2 H for Tlama well waters plotted with LMWL-Mexico City.
Trang 8sorption should occur The correlation of arsenic contamination with sewage impact suggests that the arsenic release mechanism is anthropogenic
Sewage contamination is due to a lack of proper sanitation facilities Sewage impact is indicated
by elevated concentrations of Cl, Na, K, NO3, and SO4, and elevated EC with respect to other wells
in the village The most contaminated Tlama wells are comparable to raw sewage (Kim et al 2002, Robertson et al 1991, Crites & Tchobanoglous 1998, City of Calgary unpubl data) The occurrence
of sewage impact in lower elevation wells suggests that inadequate sewage management in Tlama
is the source of the contamination It is thought that the sewage contaminants are transported to wells
by recharge through thin soil and shallow groundwater, particularly in the wet season
Sewage contamination introduces elevated concentrations of anions known to compete with arsenic for sorption sites, including SO4, Cl, NO3, and in particular PO4 Phosphate and arsenic have long been known to behave similarly and it is well documented that phosphate has a stronger affin-ity for sorbents, thus removing arsenic through competitive desorption All sewage related param-eters are correlated to arsenic in Tlama well waters, with the exception of phosphate This suggests that phosphate is being removed by sorption while liberating sorbed arsenic
ACKNOWLEDGEMENTS
This work was supported greatly by the local volunteers working with Caminamos Juntos para Salud y Desarrollo and I Hutcheon of the University of Calgary Funds from the Canadian International Development Agency, the Geological Society of America, and the Central American Water Resource Network (CARA) were appreciated We are thankful to Prosun Bhattacharya and Kazi Matin Ahmed for their comments on an earlier draft of this manuscript
REFERENCES
Boyle, R & Jonasson, I 1973 The geochemistry of As and its use as an indicator element in geochemical
prospecting Journal of Geochemical Exploration 2: 251–296.
Campa, M & Coney, P 1983 Tectono-stratigraphic terranes and mineral resource distributions in Mexico.
Canadian Journal of Earth Sciences 20: 1040–1051
Campa, M & Ramírez, J 1979 La evolución geológica y la metalogénesis del noroccidente de Guerrero.
Universidad Autónoma de Guerrero Serie Técnico-Cientifica 1: 1–84.
Cole, J 2004 Arsenic in a village drinking water supply, Mexico University of Calgary unpublished M.Sc thesis University of Calgary, Calgary.
Cortés, A., Durazo, J & Farvolden, R 1997 Studies of isotopic hydrology of the basin of Mexico and
vicin-ity: annotated bibliography and interpretation Journal of Hydrology 198: 346–376.
Crites, R & Tchobanoglous, G 1998 Small and decentralized wastewater management systems McGraw – Hill Series in Water Resources and Environmental Engineering Boston 1084 p.
De Cserna, Z & Fries, C 1981 Hoja Taxco 14Q-h(7) Resumen de la geología de la Hoja Taxco, Estados de Guerrero, México y Morelos Carta Geológica de México Serie 1:100,000 Instituto de Geología, Universidad Nacional Autónoma México.
de Kanel, J & Morse, J.W 1978 The chemistry of orthophosphate uptake from seawater on to calcite and
arag-onite Geochimica et Cosmochimica Acta 42(9): 1335–1340.
INEGI 1984 Carta geologica E14-5, Hoja Cuernavaca, E 1: 250 000 Instituto Nacional de Estadística Geografía e Informática, Mexico.
Kim, K., Lee, J.S., Oh, C., Hwang, G., Kim, J., Yeo, S., Kim, Y & Park, S 2002 Inorganic chemicals in an
effluent-dominated stream as indicators for chemical reactions and streamflows Journal of Hydrology
264: 147–156.
Manning, B.A & Goldberg, S 1996a Modeling arsenate competitive adsorption on kaolinite,
montmoril-lonite and illite Clay and Clay Minerals 44(5): 609–623.
Manning, B.A & Goldberg, S 1996b Modeling competitive adsorption of arsenate with phosphate and
molybdate on oxide minerals Soil Science Society of America Journal 60: 121–131.
Trang 9Morán Zenteno, D., Alba Aldaves, L., Martínez Serrano, R., Reyes Salas, M., Corona Esquivel, R & Angeles García, S 1998 Stratigraphy, geochemistry and tectonic significance of the Tertiary volcanic sequences of
the Taxco-Quetzalapa region, southern México Revista Mexicana de Ciencias Geológicas 15(2):
167–180.
Peryea, F & Kammereck, R 1997 Phosphate-enhanced movement of arsenic out of lead
arsenate-contaminated topsoil and through unarsenate-contaminated subsoil Water, Air and Soil Pollution 93: 243–254.
Rau, I., Gonzalo, A & Valiente, M 2003 Arsenic (V) adsorption by immobilized iron mediation Modeling
of the adsorption process and influence of interfering anions Reactive and Functional Polymers 54:
85–94.
Robertson, W., Cherry, J & Sudlicky, E 1991 Ground-water contamination from two small septic systems on
sand aquifers Ground Water 29(1): 82–92.
Salas, G 1991 Taxco Mining District, state of Guerrero In G Salas (ed), The Geology of North America, Economic Geology, Mexico Vol P-3: 379–380 Colorado, Geological Society of America.
WHO 1998 Guidelines for Drinking-Water Quality, 2nd edition World Health Organization, Geneva Addendum to Volume 2
WHO 2001 Arsenic in drinking water, Fact Sheet No 2010 World Health Organization, Geneva.