Karen Broyles
OGJ Online
The contamination of methyl tertiary butyl ether (MTBE) in California's drinking water supplies might not be as widespread as originally thought, said one scientist speaking at the DeWitt & Co. MTBE, Oxygenates & Methanol conference Wednesday in Houston. The contamination is unlikely to pose a significant health risk in California, she added.
The methods of data interpretation used to analyze the issue could be at fault, said Pamela Williams, senior scientist with the California-based research firm Exponent Inc. Williams told audience members Wednesday that monitoring data doesn't support claims that MTBE contamination is affecting the general population as widely as believed. She also questioned the decisions of California and other states to phase out MTBE or closely regulate its use.
"Policy decisions to ban MTBE should be based on an appropriate characterization of the available data," Williams noted. A more accurate picture of MTBE's affect on the general population's health could be drawn by evaluating MTBE's toxicity with more accurate modeling; identifying areas at high risk for MTBE water contamination; and comparing the cost-benefits of MTBE alternatives, she said.
Williams discussed the findings of a 1998 report by the University of California at Davis, titled Health & Environmental Assessment of MTBE, which said that "significant risks and costs" accompany the MTBE water contamination. It recommended that MTBE use be phased out over the next several years.
California Gov. Gray Davis in 1999 issued an executive order ordering MTBE be removed from California gasoline supplies by the end of 2002. Other states, including Iowa and New York, are working on legislation to ban MTBE or allow states to opt out of the oxygenate mandate. They also are proposing more stringent secondary water quality standards for MTBE and regulation of MTBE under the Toxic Substances Control Act.
However, Williams said that she believes these decisions are based on "aesthetic" reasons, or secondary health risks, rather than evidence concerning primary health risks. Williams researched in her own study the health risks posed by MTBE, and concluded that the assumption that MTBE posed a significant health risk and other assumptions about MTBE were untrue.
Williams also challenged the idea that MTBE drinking water contamination is widespread in California.
UC-Davis report
She noted one figure from the UC-Davis report said that 78% of leaking underground storage tanks in California have detectable levels of MTBE. While the number may imply that 78% of drinking water supplies have been affected, Williams said that MTBE has been detected in about 2% of public drinking water systems sampled in California as of September 2000.
The majority of California residents�about 29.9 million people, or 88% of the state's population�are served by public water supply systems.
She added decisions to close drinking water wells in Santa Monica and South Tahoe, Calif., due to contamination don't represent what's happening in other communities in California. "Reports of the incidences have created misperceptions about the magnitude of MTBE drinking water contamination," Williams noted.
Williams said the UC-Davis report recorded 4 detections of MTBE, or 3.4% of the 117 sources sampled, in 1995. The number of sources sampled and detections rose through 1998, when 44 cases, or 1.8%, of MTBE contamination were found in 2,435 sources sampled. In 1999, 35 MTBE detections, or 1.8%, were recorded out of a total sampling pool of 1,951.
The percentage of affected drinking water sources, however, hasn't increased from 1995 to 1999, Williams said, adding that the detection frequency for MTBE has remained relatively stable over the last few years.
Williams said the findings illustrated the problem of using "snapshot" data. MTBE contamination of drinking water sources isn't necessarily consistent over time. For example, MTBE was generally detected in only 1 of 2 consecutive years for drinking water sources sampled for 3-5 consecutive years. "Snapshot data may therefore not be adequate for evaluating the extent of MTBE contamination over time."
She estimated the average concentration of MTBE in drinking water sources for the general population between 1 and 5 particles/billion (PPB).
Williams said she found that MTBE levels in drinking water are fairly low on average and have remained relatively stable the past few years. "Even for sources with at least one MTBE detection, average concentrations are less than 10 particles/billion," Williams said.
Average MTBE levels are below California's primary, or health-based, standard of 13 PPB, but exceed the secondary or aesthetic standard of 5 PPB. Available monitoring data, however, doesn't provide information on the distribution of MTBE exposures or risk in the population. Williams did note that some people might be more sensitive to MTBE, but that the ability to detect or "sense" MTBE in drinking water doesn't imply a human health threat.
Using probabilistic methods to characterize MTBE exposures would allow more accurate conclusions to be drawn about the MTBE contamination threat in California, Williams said. She added several routes of exposures should be considered when evaluating health risks, such as ingestion of contaminated water, contact with the skin, and inhalation of MTBE vapors during showers or other uses of water.
Williams found that the daily consumption of water made the largest contribution to total daily MTBE dosage. Inhalation of MTBE vapors does play a role, but exposure through the skin from water-related activities had a negligible impact.
In her report, Williams found that 95% of the general population in California are exposed to an average dose of less than 1 mg of MTBE per kg of water per day (mg/kg/day).
So far, the US Environmental Protection Agency hasn't formally classified MTBE as a human carcinogen. And the National Institute for Environmental Health Sciences has omitted MTBE from its list of known carcinogens (OGJ Online, May 25, 2000).
While tumors have been observed in animals exposed to high doses of MTBE, some controversy has arisen in how these data sets have been interpreted, Williams noted. This piece of information has been used by California to derive an upper-bound cancer slope factor for MTBE in drinking water of 0.0018 mg/kg/day.
But Williams noted that only limited data is available on potential chronic or non-cancer health effects from MTBE air and drinking water exposures. Williams estimates the cancer risk from MTBE exposures through drinking water at less than 1 in a million for all exposure scenarios. The estimated hazard index, which is based on all routes of exposure, is also less than 1.
