Ground-water samples were collected from 90 active public supply wells in the Fresno and Modesto metropolitan areas as part of the California Aquifer Susceptibility (CAS) program. The CAS program was formed to examine the susceptibility to contamination of aquifers that are tapped by public supply wells to serve the citizens of California. The objectives of the program are twofold: (1) to evaluate the quality of ground water used for public supply using volatile organic compound (VOC) concentrations in ground-water samples and (2) to determine if the occurrence and distribution of low level VOCs in ground water and characteristics, such as land use, can be used to predict aquifer susceptibility to contamination from anthropogenic activities occurring at, or near, land surface. An evaluation was made of the relation between VOC occurrence and the explanatory variables: depth to the top of the uppermost well perforation, land use, relative ground-water age, high nitrate concentrations, density of leaking underground fuel tanks (LUFT), and source of recharge water.

VOCs were detected in 92 percent of the wells sampled in Modesto and in 72 percent of the wells sampled in Fresno. Trihalomethanes (THM) and solvents were frequently detected in both study areas. Conversely, the gasoline components, benzene, toluene ethylbenzene, and xylenes (BTEX), were rarely, if at all, detected, even though LUFTs were scattered throughout both study areas. The rare occurrence of BTEX compounds may be the result of their low solubility and labile nature in the subsurface environment.

Samples were analyzed for 85 VOCs; 25 were detected in at least one sample. The concentrations of nearly all VOCs detected were at least an order of magnitude below action levels set by drinking water standards. Concentrations of four VOCs exceeded federal and state maximum contaminant levels (MCL): the solvent trichloroethylene (TCE) and the fumigant 1, 2-dibromo-3-chloropropane (DBCP) in Fresno, and the solvents TCE and tetrachloroethylene (PCE) in Modesto. Chloroform, which is a by product of water disinfection and a constituent used in industrial processes since the 1920s, was the most frequently detected compound, whereas the gasoline oxygenate methyl tert-butyl ether (MTBE), which has been in widespread production and use only since the 1990s, was detected in only 2 percent of the samples.

Downward migration of contaminants appears to be a viable pathway of contamination in the unconfined and semi-confined aquifers underlying the Fresno and Modesto study areas. Within the individual study areas, VOCs were detected more frequently and in greater numbers in shallower wells than in deeper wells. Additionally, VOCs were detected more frequently and in greater numbers in Modesto than in Fresno. Wells sampled in Modesto were significantly shallower than the wells sampled in Fresno; the other explanatory variables examined in this report were not significantly different between the two study areas.

VOCs occurred more frequently in younger ground water (water recharged after 1952) than in older ground water (water recharged prior to 1952). Additionally, wells withdrawing younger ground water had a higher number of VOCs detected per well than did wells withdrawing older ground water. Younger ground water was at or near the land surface during a period when VOCs came into widespread production and use. Therefore, wells from which younger ground water is withdrawn may be more susceptible to contamination.

Of the explanatory variables examined in this study, land use was the best predictor of aquifer susceptibility in the Fresno and Modesto study areas. VOCs were detected more frequently in wells located in heavily urbanized areas. The number of VOCs detected in ground water was positively correlated to the degree of urbanization. VOCs are produced and used primarily in urban land use settings; therefore, aquifers underlying urban areas may be more susceptible to contamination from these compounds.

Other variables had little or no predictability. Overall, the presence of high nitrate concentrations was only marginally useful in predicting aquifer susceptibility to VOC contamination. In Fresno, nitrate concentrations had a moderate correlation to VOC occurrence in ground water, whereas in Modesto, nitrate concentrations did not predict VOC occurrence. The density of LUFTs and the stable isotopic content of ground water were not good predictors of VOC occurrence in Fresno and Modesto ground water. LUFT density was not useful because gasoline components comprised less than 2 percent of the VOCs detected in the study areas. Source of recharge water is indicated by stable isotope ratios. The presence or absence of VOCs in ground-water samples was not correlated with stable isotope values. Therefore, source of recharge water was not important in predicting aquifer susceptibility in the Fresno and Modesto study areas.