in Aquifers and Wells Larry McKay Dept. of Earth & Planetary Sciences University of Tennessee Co authors & Collaborators

Transcription

1 of TENNESSEE THE UNIVERSITY T The Value of Field Occurrence Studies of Fecal Contamination in Aquifers and Wells Larry McKay Dept. of Earth & Planetary Sciences University of Tennessee Co authors & Collaborators University of Tennessee Alice Layton & Peter Knappett Columbia University & Barnard College Lex van Geen, Patricia Culligan, Brian Mailloux, Andrew Ferguson & JohnFeighery University of North Carolina Mike Emch, Marc Serre, Veronica Escamilla, Yasu Akita & Jianyong Wu Bangladesh Kazi Matin Ahmed, Mohammad Yunus & Peter Streatfield

2 Outline The need for field scale occurrence studies and case histories of pathogens & fecal indicators Examples from a field study in Bangladesh Testing the main hypothesis Serendipitous findings Summary Pathogens in Groundwater Wide variety of pathogen types & inputs Variable correlation with other fecal microbes Strong influence of environmental factors Hydrogeologic setting, precipitation, well type, sewage treatment, population density, etc. Poor ability to predict waterborne pathogen occurrence or disease

3 Pathogen/surrogate Papers in Hydrogeology Lab-based or modeling transport studies Field transport experiments Pathogen occurrence or case histories WHY? Case against occurrence/case histories Expensive, time consuming, messy, unconstrained, seasonal, irreproducible, local relevance, etc. Viewed by some as a public health topic, not a hydrogeological topic Not attractive to hydrogeological funding agencies or journal editors

4 Case for occurrence/case histories Distinguish i what is from what might be Insights into transport processes Aid in experimental & modeling design Water management & public health Can lead to serendipitous discoveries Medical Case Histories are highly respected, so why are hydrogeological Case Histories discouraged? Examples from Bangladesh Study Funded by NIH Investigate relationship between fecal lbacteria & arsenic in rural wells Collaboration with Columbia Univ. Barnard College Dhaka Univ. Univ. of North hcarolina

5 Bangladesh Arsenic & Pathogens Children drinking from shallow wells that are high in arsenic are less likely to have diarrhea than children drinking from wells low in arsenic Can this be explained by a geologic g control? Bangladesh Hypothesis Sandy sediments near ground surface Short resident time & rapid transport of pathogens Oxic conditions favor sorbed arsenic Clayey sediments near ground surface Longer residence time & slow pathogen transport Rd Reducing conditions favor dissolved d arsenic

6 Revised Conceptual Model Latrines and wash water typically discharge into excavated pits These ponds act as sources of recharge to aquifer Many transport pathways Latrine infiltration Leaks in or around well casing Seepage from ponds Priming water Latrine Pond Water Table Influence of sediment type? Dry season vs Monsoon?

7 Primary Results Geology is a major control on both arsenic and E. coli (GW 2010) and there is an inverse correlation between As & E. coli (ES&T 2010) Consumers switching to low Arsenic wells in villages can increase risk of diarrhea (In-review) Freque ency of E. coli de etection Well As (ug/l) Secondary (often serendipitous) Results Role of ponds as fecal sources (ES&T in-review) Cast iron hand pumps as reservoirs of microbial contamination (J. Water & Health in-press) Improved sampling methods and influence of pumping on E. coli concentration (GW 2010) Influence of human population density on E. coli Comparison of field-scale and column-scale transport Molecular E. coli correlation with pathogens Comparison of well pathogens with those detected in clinical disease studies

8 Transport from pond experiments (Knappet et al.) Fine-medium grained sand Monitoring well transect Pond Flooding Experiment Simulate influence of major rainfall event on E. coli transport a b c d 10 m 4 m

9 Results 1e+6 E. coli concentrations increase in transect near sandy pond, but not at silty pond Fitted β values s s indicated a 6-log 10 ( %) 9999%) removal distance of < 5-12 m MPN/100 ml) E. coli ( E. coli (MPN N/100 ml) T1.1 Natural Medium β = 2.9 ( ) 1e+5 sand T1.1 1 Post-filling β = 1.6 ( ) 1e+4 β = 0.7 gw 1e+3 1e+2 1e+1 1e+0 1e-1 (a) 1e+6 T3.1 Post-filling Silty fine T4.1 Post-filling 1e+5 1e+4 1e+3 1e+2 1e+1 1e+0 1e-1 sand Distrance from Pond (m) (g) Hand pump experiment (Ferguson et al.) Test whether pumps act as secondary reservoirs for contamination

10 Results Added clean water to reservoir daily E. coli detected in discharge for 30 days Results Added clean water to reservoir daily E. coli detected in discharge for 30 days Repeated experiment with higher C/Co Shock chlorination ineffective at removing E. Coli or coliform

11 Summary Field studies are needed to better understand causes & impacts of fecal contamination These studies often yield valuable unexpected insights