Statistical Evaluation of Factors Affecting Occurrence of Organic Constituents from Fuels in Public-Supply Wells in Southern California

Science Center Objects

Biofuels, including ethanol, are expected to provide a greater contribution to the fuel supply in the United States. Research in the past decade has shown that ethanol in gasoline can cause plumes of benzene in groundwater to be larger than they would otherwise be.

Map of Hydrogeologic provinces of California and distribution of wells

Hydrogeologic provinces of California and distribution of wells in the CDPH database and grid and understanding wells sampled by GAMA during 2004-2009.

(Public domain.)

A key step in determining the vulnerability of public-supply wells to the increased use of biofuels is to identify public-supply wells in key settings of the United States that have had historical detections of fuel components and to evaluate what factors may explain those detections. California provides an ideal location to evaluate the relation of environment factors to the occurrence of fuel components in public-supply wells for several reasons: (1) there are thousands of public-supply wells, providing the statistical mass necessary to do a comprehensive statistical analysis, (2) there is an abundance of ancillary data sets, mostly assembled by the GAMA program, available to facilitate analysis of statistical relations of fuel components in public-supply wells to ancillary factors, and (3) Southern California is an important setting in the national landscape, superimposing very large urban populations dependent on groundwater supplies, with a high density of fuel sources, and large groundwater withdrawals.

The objectives of this study are to identify relations of land use, hydrogeology, geochemistry, and well construction and well-operation factors to the occurrence of gasoline hydrocarbons and other constituents derived from fuels in public-supply wells in Southern California.

The study will consist of four phases: (1) identify wells Southern California hydrogeologic provinces in the California Department of Public Health, USGS Groundwater Ambient Monitoring and Assessment (GAMA) and National Water Information System (NWIS) databases with detections of any of 25 gasoline hydrocarbons or gasoline oxygenates and attempt to fill gaps in ancillary factor data for those wells, (2) conduct nonparametric statistical analysis of relations of fuel constituent detections in public-supply wells with potential explanatory factors to identify significant relations, (3) build logistic regression model(s) based on significant explanatory relations to estimate the probability of detection of fuel constituents in public-supply wells, and (4) present the results in a report.