feature MTBE To What Extent Will Past Releases Contaminate Community Water Supply Wells? An improved understanding of the factors that affect the magnitude of the problem is needed. RICHARD JOHNSON, JAMES PANKOW, DAVID BENDER, CURTIS PRICE, AND JOHN ZOGORSKI T he increasing frequency of detection of the widely used gasoline additive methyl tert- butyl ether (MTBE) in both ground- and sur- face waters is receiving much attention from the media, environmental scientists, state environ- mental agencies, and federal agencies. At the na- tional level, the September 15,1999, Report of the Blue Ribbon Panel on Oxygenates in Gasoline (i) )tates that between 5 and 10% of community drinking water sup- plies in high MTBE use areas show at least detectable concentrations of MTBE, and about 1% of those sys- tems are characterized by levels of this compound that are above 20 pg/L. In Maine, a desire to determine the extent of MTBE contamination led to a 1998 study (2) that revealed that this compound is found at levels above 0.1 pg/L in 16% of 951 randomly selected house- hold wells and in 16% of the 793 community water sys- tems tested in that state (37 wells were not tested). The study also suggested that between 1400 and 5200 household wells may have levels above 35 pg/L, al- though no community water supplies were found to be above that concentration. For comparison, Mary- land, New Hampshire, New York, and California have set MTBE remediation "action levels" at or below 20 pg/L, and EPA has set its advisory level for taste and odor at 20-40 pg/L (3). In California, concern regarding MTBE reached statewide levels in 1996 when seven wells supply- ing 50% of the water for the city of Santa Monica were removed from service because of MTBE at concen- trations as high as 600 pg/L. For the city's Char- nock well field, an initial review of known and sus- pected petroleum spill sites identified about 10 potential sources that lay within 1 km of the well field, lay above the hydrologic unit accessed by the well field, and were created after MTBE use began in the state (4). At the time that contamination of die wells was discovered, pumping of the Charnock well field was at 5 million gallons/day (mgd). This aggressive pumping was approximately twice the total natural flow of water moving into the aquifer. Despite the presence of a protective aquitard in the system, the pumping had dewatered a significant portion of the upper aquifer, caused water to flow toward the well field from all directions, and had greatly increased the likelihood that the community water supply (CWS) wells in Santa Monica would in fact become contaminated by one or more persistent organic pol- lutants such as MTBE. Besides leaking underground fuel tanks (LUFTs) and leaking pipelines, other sources of MTBE in groundwater include tank overfilling and faulty con- struction at gas stations, spillage from vehicle acci- dents, and homeowner releases. In Maine, it is pos- 210A• MAY 1, 2000 / ENVIRONMENTAL SCIENCE & TECHNOLOGY / NEWS © 2000 American Chemical Society