The high arsenic groundwater of the Bengal Delta Plain (BDP) in India as well as in Bangladesh, has become a priority health issue. A large population (» 50 million) has been affected by arsenic toxicity over the last two decades. The source of arsenic is geogenic where the sediments are deposited with elevated level of arsenic. An expanding exploitation of groundwater and the change in landuse pattern is believed to have triggered the mobilisation of arsenic into the groundwater. The geochemical characteristics of the aquifer play a crucial role in understanding the mobilisation of arsenic.
Arsenopyrite (FeAsS) and/or pyrite (FeS2) are the primary source of arsenic in natural systems. In sedimentary systems, arsenic is also adsorbed onto the reactive surfaces of secondary oxides and hydroxides of iron, aluminium and manganese (HFO, HAO and HMO). An aerobic (1) and an anaerobic (2) hypothesis have been put forward to explain the mobilisation of the arsenic in the groundwater. 1) When groundwater is withdrawn, oxygen may enter the aquifer and oxidise the primary sulphides and subsequently releases arsenic and sulphate in groundwater. 2) Alternately, arsenic may be derived as a result of desorption and reductive dissolution of the surface reactive mineral phases such as hydrous ferric, aluminium and manganese oxides (HFO, HAO and HMO) present as coatings in the aquifer sediments. Phosphate is adsorbed onto HFO in a similar way as arsenic. The mechanism of arsenic and phosphate is initialised by anoxic water in the aquifers.
Watersamples from wells in Chakdaha, West-Bengal, India and Bangladesh were analysed for a number of chemical parameters. Some porewater/BAT-point samples were also taken from a depth of 3 to 5 meters in Chakdaha and compared with the well samples. The groundwater was found to be neutral to slightly alkaline and essentially Ca-HCO3 type. The groundwater is under reducing conditions. The concentration of arsenic was very high in the area. 79 % respectively 65 % (India and Bangladesh) of the well samples had higher concentrations of arsenic than WHO's permissible limit of 10 mg/l. The ratio of As3+ and As5+ was measured, it varied from 1 to 100 % but could not be linked to any other analysed parameter.
Arsenic correlated with iron. This would strengthen the second hypothesis. Likewise, phosphorous correlated with iron. Further, relationship between the amount of precipitated iron (HFO) and adsorbed arsenic was found and it shows the importance of HFO as a controlling factor for the release of arsenic in groundwater. Porewater from 3 to 5 meters depth had low concentrations of arsenic, iron and phosphorous. In this horizon, both arsenic and phosphorous are immobilised due to adsorption on secondary hydroxides phases.
Saturation indexes (SI) were calculated using the equilibrium-based hydrogeochemical code WATEQ4F. The model was used to understand the role of secondary mineral phases controlling the groundwater chemistry. According to the model calculations, the porewater (from 3 to 5 m. depth) was undersaturated with respect to goethite (FeOOH), siderite (FeCO3) and vivianite (2[Fe3(PO4)2·8H2O]). This would strengthen the idea, that under near surface conditions iron is precipitated as HFO and adsorbs arsenic and phosphorous and thereby reduces the abundance of these elements in the porewater. The well samples from the depths »15 to 130 meters are supersaturated with respect to goethite (FeOOH), siderite (FeCO3) and vivianite (2[Fe3(PO4)2·8H2O]). The supersaturation of these minerals indicates that the groundwater chemistry is controlled by the reductive dissolution of FeOOH thereby mobilising adsorbed As3+/5+ and PO43-. All samples were supersaturated with respect to calcite (CaCO3) and dolomite (CaMg(CO3)2), which shows that the system is well buffered with respect to bicarbonate (HCO3-).
This investigation has produced no support for the first aerobic hypothesis. On the contrary, it strongly forwards the second anaerobic hypothesis.
Keywords: Arsenic, groundwater, the Bengal Delta Plain, geochemistry, redox-conditions, aquifer, hydrogeochemical modelling, BAT-sampler