Limited temporal variability of arsenic concentrations in 20 wells monitored for 3 years in Araihazar, Bangladesh.

Millions of people in Bangladesh have probably switched their water consumption to wells that meet the local standard for As in drinking water of 50 microg/L as a result of blanket field testing throughout the country. It is therefore important to know if As concentrations in those wells could change over time. To address this issue, we report here precise groundwater As analyses for time-series samples collected from a suite of 20 tube wells containing < or =50 microg/L As and ranging from 8 to 142 m in depth. For 17 out of 20 wells, the standard deviation of groundwater As concentrations was <10 microg/L over the 3-year monitoring period (n = 24-44 per well). Six of the 17 wells are community wells, each of which serves the needs of several hundred people in particularly affected villages. Of the three wells showing larger fluctuations in chemical composition including As, two are very shallow (8 and 10 m). Variations in As concentrations for one of these wells (50 +/- 32 microg/L, n = 36), as well as another shallow well showing smaller variations (48 +/- 5 microg/L, n = 36), appear to be coupled to seasonal precipitation and recharge linked to the monsoon. The other shallow well showing larger variations in composition indicates a worrisome and steady increase in As concentrations from 50 to 70 microg/L (n = 36) over 3 years. The time series of As (30 +/- 11 microg/L, n = 24) and other constituents in one deep community well (59 m) show large fluctuations that suggest entrainment of shallow groundwater through a broken PVC pipe. Even though the majority of wells that were initially safe remained so for 3 years, our results indicate that tube wells should be tested periodically.