Note to the Reader
This version of Robert Hirsch's testimony of November 1, 2001 contains two notes that have been added (at the end of the testimony) to (1) explain the scope of USGS's sampling in the National Water-Quality Assessment Program and (2) to correct the wording of a sentence concerning MTBE detections in urban studies. Please contact John Zogorski, USGS, if you have questions concerning the two footnotes or any other aspect of the testimony.
Chief, VOC National Synthesis
National Water-Quality Assessment Program
U.S. Geological Survey
1608 Mt. View Road
Rapid City, SD 57702
Ph: 605-355-4560 x214
Chairman Greenwood and other committee members, I appreciate the opportunity to appear before the Subcommittee on Oversight and Investigations to testify on the findings of U.S. Geological Survey (USGS) studies on water-quality issues related to methyl tertiary-butyl ether, commonly referred to as MTBE.
As you may know, the mission of the USGS is to assess the quantity and the quality of the earth's resources and to provide information that will assist resource managers and policy makers at the Federal, State, and local levels in making sound decisions. Assessment of water-quality conditions and research on the fate and transport of pollutants in water are important parts of the overall mission of the USGS.
USGS studies over the past 8 years have shown that MTBE typically is present at very low concentrations in shallow ground water within areas where MTBE is used. Our studies also suggest that MTBE levels do not appear to be increasing over time and are almost always below levels of concern from aesthetic and public health standpoints. The few locations in our database with high concentrations of MTBE may be associated with leaking underground storage tanks.
Based on comparisons with the U.S. Environmental Protection Agency's (USEPA) drinking water advisory, the health threat to water supplies is small compared to other water-related issues. MTBE is primarily an aesthetic (taste and odor) problem. However, we believe it may be prudent to continue our monitoring and research within available resources so that we can verify that the threat remains low and to further the understanding of this chemical to contribute to effective strategies to protect our Nation's water supplies and to efficiently remediate those ground waters that have become contaminated.
The results I will present today come from about a decade of sampling and study of MTBE and other Volatile Organic Compounds (VOCs). MTBE is one of about 60 VOCs that we measure on a routine basis in our water-quality studies.
The single largest study we have made of MTBE is part of our National Water Quality Assessment (NAWQA) Program.1 Based on initial monitoring data for wells sampled in 1993-94 in the NAWQA Program, we published a report on the occurrence of MTBE in shallow ground water in urban and agricultural areas. At that time our data set was fairly small-about 200 randomly selected wells in urban areas and 500 randomly selected wells in agricultural areas. We reported finding MTBE in about 25 percent of urban wells and 1 percent of agricultural wells. Many of the MTBE detections were low concentrations. In fact, only 3 percent of the urban detections exceeded 20 micrograms per liter, the lower limit of USEPA's consumer advisory for taste and odor.2 Also, many of the urban wells that contained MTBE were located in Denver, Colorado, and in New England, both areas with extensive use of MTBE prior to our sampling. At the time, MTBE was a chemical for which usage had increased dramatically in recent years and we knew it moved in the subsurface differently from other gasoline components. Thus, even though it was detected in few wells and at very low levels, we believed it would be prudent to continue studying it at many locations and over a period of several years to learn more about its national distribution and fate.
Since our first report in 1995, we have sampled additional wells in the NAWQA Program. This now gives us much better coverage of aquifers across the Nation. For the period 1993-2000, we sampled 4,260 wells (or springs) for MTBE and a wide range of other compounds. Of this total, 396 are public water-supply wells; 1,847 are domestic wells; and 2,017 are monitoring wells (or other wells not used for drinking water). At a reporting level of 0.2 micrograms per liter (a level that is one one-hundredth of the USEPA advisory level), we detected MTBE in 5.2 percent of the wells sampled. Most of the MTBE detections are low concentrations. None of the public water-supply wells and only one domestic well had MTBE at a concentration above the lower limit of USEPA's advisory. Through our interpretations of this large data set we have also determined that low-levels of MTBE are detected in about 1 out of 5 wells in MTBE high-use areas. Although we do not expect to see a great change in these results over time, we recognize that there may be a delay in the detection of MTBE in some wells-particularly those that are deeper and may be farther from the source of contamination. MTBE is the second most frequently detected volatile organic compound (VOC). Chloroform, a drinking-water disinfection by-product and a commercial solvent, is the most frequently detected VOC.
Based on our NAWQA findings and interests of other agencies, we have undertaken two allied, large-scale studies to further our understanding of the occurrence of MTBE and other VOCs. We have completed a study in cooperation with the USEPA's Office of Ground Water and Drinking Water. For the period 1993-98, we have compiled information on the occurrence of MTBE and other VOCs in drinking water supplied by Community Water Systems in 12 States in the Northeast and Mid-Atlantic Regions of the United States. Parts of these Regions are designated Reformulated Gasoline (RFG) Areas and, in general, these RFG Areas have used MTBE in gasoline in large amounts for many years. USGS obtained the MTBE/VOC data from each State's drinking-water program. We then randomly selected about 20 percent of the almost 11,000 Community Water Systems in the study area for our analysis. States with MTBE data included Connecticut, Maine, Maryland, Massachusetts, New Hampshire, New Jersey, New York, Rhode Island, Vermont, and Virginia. Data for MTBE were not available for Delaware and Pennsylvania, at the time the study was completed.
At a reporting level of one microgram per liter, about 9 percent of the Community Water Systems had detectable MTBE in their drinking water; however, most of the detections were low concentrations. Ten Community Water Systems had MTBE concentrations that equaled or exceeded the lower limit of the USEPA advisory, or about 1 percent of all Community Water Systems with MTBE data. We also confirmed that MTBE was detected more frequently in RFG Areas than elsewhere in the two Regions. Furthermore, larger Community Water Systems located in urban centers had a larger incidence of MTBE detections.
We are also working with the Metropolitan Water District of Southern California, and the Oregon Graduate Institute of Science and Technology, to complete a study of MTBE, other ether gasoline oxygenates, and other VOCs in select reservoirs, rivers, and wells that supply Community Water Systems. This study was partly funded through the American Water Works Association Research Foundation (AWWARF). We are in the final year of this 4-year project.
For this study, we tested the source water of 954 randomly selected Community Water Systems, including 579 wells, 171 rivers, and 204 reservoirs. Samples were collected in all 50 States and Puerto Rico, and varied sizes of systems were included. All sampling for this project is completed; however, some of our intended interpretations and report writing are not yet completed and peer reviewed. Initial findings, which were reported on June 20, 2001, at the Annual Conference of the American Water Works Association, were similar to our findings noted earlier in this statement. Specifically, when detected in source waters, the concentrations of MTBE were almost always below the USEPA advisory. However, MTBE was found in about 9 percent of all sources sampled (at a reporting level of 0.2 micrograms per liter), and it was the second most frequently detected VOC. A larger detection frequency of MTBE was found in surface-water sources (14 percent), than ground-water sources (5 percent). In general, the detection of MTBE increased with increasing size of the Community Water Systems. MTBE was detected in about 4 percent of Community Water Systems serving less than 10,000 people, and in nearly 15 percent of systems serving greater than 50,000 people. Many of the surface-water sources sampled in the AWWARF study were large rivers and reservoirs that had recreational watercraft usage. Older models of watercraft motors are known to release a fraction of non-combusted gasoline to water and this, in part, may explain the larger occurrence of MTBE in surface-water sources.
We also conduct research on the fate and transport of MTBE in ground water and surface water through the USGS Toxic Substances Hydrology Program. In this program, we explore the range of geochemical and microbiological processes that determine how MTBE will behave when it enters soil, ground water or surface water. This research is demonstrating that MTBE does biodegrade under a wide range of environmental settings although at slower rates than many of the components of traditionally formulated gasoline. These ongoing studies have important implications for predicting the future concentrations of MTBE in water, where contamination has already occurred. These results are also important for the design and selection of remediation plans.
As part of the Toxic Substances Hydrology Program research, USGS scientists have demonstrated that naturally occurring microorganisms can biodegrade MTBE in many hydrologic environments, and in some cases, to harmless by-products. In some situations, however, biodegradation may be incomplete and tert-butyl alcohol (TBA) can be formed. Especially noteworthy are the observations that MTBE biodegrades in ground water and soil where sufficient oxygen is present and in bed sediments of streams, lakes, wetlands, and estuaries where MTBE-contaminated ground water can ultimately discharge. Essentially, these environments can be considered to be natural sinks for MTBE removal. As noted earlier, MTBE is expected to degrade slower in ground water than gasoline hydrocarbons of traditional gasoline formations. The length of time required to complete this removal is currently a topic of ongoing investigation.
The USGS has actively participated in two previous Federal reviews of MTBE and other oxygenates in gasoline. A Blue Ribbon Panel was appointed by the Administrator of the USEPA to investigate the air-quality benefits and water-quality concerns associated with oxygenates in gasoline, and to provide independent advice and recommendations on ways to maintain air quality while protecting water quality. In 1998-1999, Dr. John Zogorski of the USGS served as a water-quality consultant to the Blue Ribbon Panel and three USGS scientists testified before the Panel. An important finding of the Blue Ribbon Panel is that the major source of MTBE ground-water contamination appears to be releases from underground gasoline storage systems. Many of these tanks have been removed permanently or upgraded in the 1990s, and thus this source is likely to diminish in the coming years. Other major sources of water contamination were stated to be from small and large gasoline spills and from recreational watercraft, especially those with older model 2-cycle motors. USGS has documented low levels of MTBE in urban air, urban precipitation, and urban stormwater, and these sources may cause low concentrations of MTBE in surface water and ground water. MTBE has also been found in spills of home fuel oil in Northeastern States.
During 1995-96, at the request of the USEPA and the Office of Science and Technology Policy (OSTP), the USGS co-chaired an interagency panel to summarize what was known and unknown about the water-quality implications of the production, distribution, storage, and use of fuel. Our efforts were published in 1997 as a chapter in a report entitled "Interagency Assessment of Oxygenated Fuels" prepared by the National Science and Technology Council, Committee on Environment and Natural Resources. The chapter summarizes the scientific literature and data on the sources, occurrences, concentrations, behavior, and the fate of fuel oxygenates in ground water and surface water. We also discussed the implications for drinking water and aquatic life, and made recommendations of information needed to better characterize the occurrence of MTBE and other oxygenates in the Nation's drinking-water supplies.
Furthermore, last year, USGS and Oregon Graduate Institute scientists co-authored a feature article in the journal Environmental Science and Technology, a publication of the American Chemical Society. A salient part of the article summarized important information about MTBE including: growth in production; solubility, transport and degradation in ground water; releases from leaking underground fuel tanks; and the effect of select factors, such as aquifer recharge, the presence of low permeability stratum, and water utility pumping rates. This information helped to determine the likelihood of MTBE reaching community water-supply wells. Based on available but admittedly incomplete data for 31 States, the authors determined that about 9,000 community wells may have one or more leaking underground storage tanks nearby (i.e., within 1-km radius of the well). Because detailed information on the community wells, storage tanks, and hydrogeology were not available, the authors could not determine the number of wells at risk.
Unfortunately, some of the press coverage of this article inaccurately stated that 9,000 drinking-water wells were contaminated with MTBE. As stated in the journal publication, not all community wells with gasoline releases nearby are at risk because not all gasoline releases contain MTBE, and not all MTBE-gasoline releases are sufficiently large to pollute a nearby well. Also, many wells draw water from the deeper zones of aquifers and many wells are largely isolated from land-surface contamination by low permeability stratum, technically called aquitards. Based on these factors, data from the studies mentioned previously, and a recent survey by others, we would estimate that the number of community wells contaminated is far lower than 9,000 for 31 States.
In summary, the USGS has not found widespread, high-level MTBE contamination in rivers, reservoirs, and ground water that are actively used as the sources for Community Water Systems. Furthermore, we have not found such contamination in public wells and domestic wells sampled in our NAWQA Program, or in the drinking water of Community Water Systems in 10 Northeastern and Mid-Atlantic States. We have, however, identified MTBE (and some other VOCs) fairly frequently in ground water, source water, and drinking water at concentrations below USEPA's advisory. We also conclude that the frequency of detection of MTBE is larger in RFG Areas, in comparison to other areas of the Nation. Approximately 85 million people reside in RFG areas that use MTBE extensively, and drinking water in these areas is provided almost equally from surface water and ground water.
There are multiple strategies for dealing with situations where MTBE contamination of ground water has taken place and these should include strategies that take maximum advantage of the natural attenuation that we observe in our research. Within available resources, more research would be helpful to provide guidance on the most cost-effective strategies for protecting drinking water sources in those areas that have become contaminated.
I appreciate the opportunity to testify on the results of USGS assessments and research on MTBE. I am happy to try to respond to any questions of the Subcommittee.
1Monitoring for MTBE and other VOCs in ground water, in the NAWQA Program, may be conducted in three different ground-water studies including Major Aquifer Studies, Urban Land-Use Studies, and Agricultural Land-Use Studies. The specifics about the purpose, scope, and design of these studies are given in Gilliom and others, 1995, Design of the National Water-Quality Assessment Program: Occurrence and distribution of water-quality conditions, U.S. Geological Survey Circular 1112, 33 p. It is important to note that the location of wells sampled in the NAWQA Program are randomly selected and areally distributed across large aquifer study areas. The intent of these design criteria is to provide a general description of water-quality conditions for the aquifer at large. The NAWQA Program does not attempt to sample ground water at known/regulated release sites with extensive ground-water contamination, such as Superfund sites, RCRA sites, and locales with leaking gasoline storage tanks. Other agencies have responsibility for monitoring at these extensively contaminated sites.
2This statement should read "In fact, only 3 percent of urban wells exceeded..."