Improving safe sanitation practices using groundwater transport modelling and water quality monitoring data

Abstract

In many developing countries, poorly constructed pit latrines are the primary source of groundwater contamination. Approaches are needed to identify site-specific separation distances between domestic wells and pit latrines. In this study, tracer transport simulations are combined with water quality monitoring data to identify separation distances in peri-urban Dar es Salaam. Bivariate correlation and linear regression models were used to find the relationships between (1) simulated tracer and distances from the wells to the nearest pit latrines (2) simulated tracer and observed contaminants (nitrate, E. coli, and total dissolved solids). The results showed a strong correlation between tracer with nitrate and E. coli, with Pearson coefficient (r) values of 0.80 and 0.79, but a weak correlation with total dissolved solids (TDS) (r = 0.23). A strong correlation between tracer and distance for shallow and deep wells (r = −0.96, −0.76) was found. Based on the soil type and aquifer properties in the area, wells must be placed at least 34 m from a pit latrine to minimize contamination. With recent advances in sensor technologies and the availability of low-cost sensors, linking simulated tracer with observed contaminant levels may provide an alternative first approach to quickly assess human health risks associated with groundwater contamination. HIGHLIGHTS A method to determine site-specific separation distance between wells and pit latrines was developed.; Tracer transport simulations and statistical analyses were used to find relationships between tracer, distance and contaminants.; A strong correlation was found between simulated tracer, distance to the nearest pit latrine and contaminants.; In this area, wells must be placed >34 m from pit latrines to minimize contamination.;

Keywords

• contaminant transport
• developing country
• domestic wells
• groundwater flow
• pit latrine
• separation distance