EDP Home Page

MTBE Water Contamination  

slvrline.jpg (1988 bytes)

MTBE Water Contamination: Key Considerations for Remediation,Risk Assessment, and Risk Management

Paper presented at:
Pacific Conference on Chemistry and Spectroscopy
American Chemical Society
Society for Applied Spectroscopy
October 21 - 25, 1997
Irvine, California

Robert P. Ghirelli**, Hassan Amini, Brent D. Kerger, Alexis Hillman, and Richard O. Richter
McLaren/Hart Environmental Engineering Corporation
515 S. Figueroa Street
Los Angeles, CA 90071
[**Dr. Ghirelli is currently the Technical Director of The [California] Orange County Sanitation District]

The discovery of the fuel additive methyl-tertiary butyl ether (MTBE) in ground water and drinking water reservoirs in California and elsewhere has fueled public debate about the health risks associated with this oxygenate. MTBE has been used since the mid-1970's as a fuel additive to reduce carbon monoxide emissions. The City of Santa Monica in 1996 shut down the majority of the City’s drinking water wells due to MTBE contamination. That event has focused much attention on the chemical and prompted public health officials, water quality regulators, the state legislature, and the regulated community to seek more information about the health effects and risk management options associated with the continued use of MTBE in California. What does the future hold for this chemical and how should it be dealt with in the regulatory arena? This paper examines some of the key parameters concerning health risks from exposure to MTBE, the management of risks associated with MTBE leaks from underground fuel tanks, and remediation options for different spill scenarios.

At what levels is MTBE being detected in California waters?

As a result of its widespread use in reformulated gasoline over the last several years, MTBE has been detected in ground water and surface water throughout the United States. In order to measure the presence of MTBE in California drinking water supplies, the California Department of Health Services in February 1997 required public water utilities to begin testing their sources of drinking water for MTBE.

As of August 1997, approximately 2,200 water sources--20 percent of the total number of sources in the state--had been sampled. Twenty seven of the sources (1.2 percent of the total) reported MTBE detections: 15 ground water and 12 surface water sources (Figure 1). In ground water, MTBE concentrations generally ranged from about 1 g/l at the low end to less than 35 g/l at the high end, with 7 detections exceeding the interim drinking water action level of 35 g/l. Surface water concentrations were all below 5 g/l except for two raw lake water detections at 9.5 and 14.0 g/l (1).

In February 1997, the Los Angeles Regional Water Quality Control Board (RWQCB) reported that MTBE had been discovered in 10 of the approximately 436 drinking water supply wells in the region. The City of Santa Monica closed down 7 wells in two well fields due to elevated MTBE levels, the City of Los Angeles Department of Water and Power detected MTBE in two wells in the Burbank area, and California Polytechnic University, in Pomona, detected MTBE in one well (2).

Based on the findings so far, ground water appears to be more vulnerable to MTBE contamination than surface water (Figure 2). At the present rate, less than 10% of the state’s drinking water sources will ultimately show detections of MTBE and far fewer will exceed the interim drinking water action level. Are we just now seeing the tip of the iceberg? Are more water sources at risk? Or, will the regulatory controls now in force and the requirements to test and upgrade underground storage tanks prevent MTBE from becoming the "benzene" of the future? These are questions for which there are no quick and easy answers. Improvements in underground storage tank (UST) technology and research into effective treatment technologies, coupled with efforts to identify vulnerable sites and monitor drinking water supplies, are essential elements of an aggressive strategy to tackle the MTBE threat head-on.

Is MTBE a problem at these levels? What are the health risks?

MTBE exhibits a relatively low acute and chronic toxicity when considering its potential caner and noncancer effects. The California State Action Level of 35 g/l and the U.S. EPA Health Advisory of 70 g/l for MTBE are based on currently available noncancer toxicity criteria developed by U.S. EPA (3). The U.S. EPA health advisory is based on an oral reference dose of 0.005 mg/kg-day, which corresponds to safe lifetime ingestion exposures of 180 g/l for an average adult and 70 g/l for a small child. Although U.S. EPA does not currently list any cancer potency values for MTBE, a screening evaluation by the Office of Science and Technology Policy has proposed oral and inhalation cancer potency values based on available data (4).

At the California State Action Level of 35 g/l, the preliminary potency estimates lead to calculation of lifetime incremental cancer risks for 30 years of continuous exposure of less than 2 per million for MTBE ingestion exposures and less than 5 per million for inhalation exposures related to indoor air pollution from tap water use. Exposure modeling has shown that 76% of MTBE intake is attributable to inhalation of house air. However, the upper bound estimates of exposures and cancer risks from inhalation exposure are not plausible because predicted indoor air concentrations of MTBE are well above the reported odor threshold using the McKone indoor air model (5). The odor nuisance at these predicted indoor air concentrations would likely preclude long-term exposures to the contaminated water source. The distribution of upper bound daily MTBE exposures contributed by ingestion and three indoor inhalation scenarios is illustrated in Figure 3.

How are the regulatory agencies responding?

Several bills dealing with MTBE made their way through the California Legislature during the 1997 session. A bill by State Senator Richard Mountjoy (SB 521) was scaled back from a mandated phaseout of MTBE to instead require a health effects study of the impacts of this and other oxygenates. The bill also prohibits closure letters at underground tank sites, unless soil and ground water have been tested for the chemical. Two other bills require a new drinking water standard for MTBE, new controls on pipelines carrying products containing MTBE, and a determination whether MTBE should be listed under Proposition 65.

In a settlement with the City of Santa Monica, two oil companies have agreed to clean up a MTBE-contaminated site to a standard that is the lesser of 20 g/l MTBE or any state or federal regulatory standard. The current state action level is 35 g/l while the U.S. EPA advisory standard is 70 g/l. The companies also agree that once they meet the 20 g/l limit, it will be maintained for a year before they are given a release (6).

For two years the California State Water Resources Control Board (SWRCB) has been considering adopting a policy that would establish statewide standards for the cleanup of petroleum discharges from leaking underground storage tanks (UST). In October 1996 State Water Board UST staff released an internal draft of the cleanup policy that would have allowed the use of passive bioremediation of petroleum hydrocarbons for "low risk" sites. Following criticism from staff at the California Regional Water Quality Control Boards the State Board staff pulled back its proposal.

The recent findings of MTBE in soil and ground water prompted the State Water Board to fund a study of MTBE by the Lawrence Livermore National Laboratory (LLNL) and results are expected in 1998. New legislation now requires all major agency proposals that rely on scientific concepts to be subjected to peer review. These events will likely delay the release of the draft statewide policy even further.

In the meantime, the Los Angeles RWQCB staff follows an interim review procedure for dealing with UST sites with MTBE contamination. Pending release of the LLNL MTBE report and issuance of further guidance from the SWRCB, the RWQCB uses a 35 g/l trigger level for MTBE. UST sites are evaluated for closure using a tiered risk-based approach (Figure 4).

Sites where the maximum concentration of MTBE in ground water affected by the discharge is less than 35 g/l may be eligible for closure if other specified criteria are met (2). Those criteria include extent of contamination by other gasoline constituents, specifically benzene, distance to the nearest drinking water well or surface water body, and whether petroleum-saturated soil is in contact with ground water at the site.

Sites within the Los Angeles region where MTBE in the ground water is greater than 35 g/l are further reviewed and must meet additional criteria prior to being considered for closure. In all cases, the assessment assumes the site has been fully characterized and the contamination delineated and sources of pollution (e.g., saturated soil) have been removed. Cases that were formerly closed may be reopened in areas where drinking water supply wells contain MTBE concentrations exceeding the 35 g/l trigger. Reviews are typically done on sites within a one-half mile to one mile radius of the affected water supply well but in certain cases sites may be investigated that are located more than one mile away from the affected well.

Remediation options for MTBE soil and ground water contamination

Generally, the techniques used to remediate the BTEX (benzene-toluene-ethyl benzene-xylene) components in gasoline are also applicable for MTBE. Remediation costs for MTBE are highly dependent on the hydrogeological characteristics of the site, the history of the spill, and the residence time of the contaminants. Almost any technique that removes BTEX from water can also remove MTBE (7). However, due to its high water solubility, lower volatility, lower adsorption to granulated activated carbon, and poor biodegradability, removing MTBE from water is typically more expensive than removing BTEX components. Figure 5 highlights some of the key characteristics of MTBE and their impact on remediation.

A MTBE plume is often larger than the BTEX plume and this greater areal extent translates to more wells, pumps, and pipelines, and treatment systems with greater capacities and efficiencies. While the cost of remediation is dependent on site-specific factors it is anticipated that the cost multiplier for remediating MTBE spills will be 1.1 to 2 times in many cases and possibly much higher for municipal water supplies or other special cases (8).

Selection of the correct remediation technology for a MTBE-impacted site, as for most sites, is dependent on a thorough understanding of the site characteristics, history of the spill, and the limitations of the technology. Compared to the main components of gasoline (i.e., BTEX), relatively little data exists for the successful remediation of MTBE-impacted ground water. Most MTBE remediation projects seem to be in the research and development state. A number of investigators have reported the successful application of soil vapor extraction (SVE) to remove MTBE from soil (9). Figure 6 presents some key considerations in the application of typical remediation technologies used for MTBE.

Removal of MTBE from aqueous solution is possible by the application of air-flow enhancement techniques such as air sparging, high vacuum suction, and dual-phase extraction techniques. Natural attenuation of MTBE is limited due to its high solubility, low biodegradability, and poor adsorption to soil particles and organic matter. Experiments conducted on oxidation of a mixture of MTBE and ethyl tert-butyl ether (ETBE) in aqueous solution demonstrated that ozone and ozone/hydrogen peroxide treatment is more effective in eliminating ETBE than MTBE (10).

Pump and treat and well-head treatment systems seem to be successful approaches for containing and removing MTBE plumes in transmissive aquifer conditions. Removal of MTBE from solution seems to be possible in an environment of high air to water volume ratios. These conditions are created by the use of large air blowers and air strippers, or multiple strippers in series. The introduction of heat appears to enhance removal of MTBE from solution.

Risk management considerations

Compared to benzene, MTBE is highly soluble and mobile in both air and water. Unlike benzene, MTBE is fairly persistent in ground water and sediments since it is resistant to most biodegradation processes. Also unlike benzene, MTBE is a more volatile ether compound with an odor that can be readily detected in water at concentrations of 40 to 134 g/l. Peak shower air concentrations of 288 to 451 g/m3, corresponding to the range of reported odor recognition thresholds for MTBE, were generated by MTBE water concentrations in the range of 34 to 53 g/l (11). The apparently low odor detection thresholds for MTBE would therefore provide a very conservative basis for an aesthetic water quality-based standard to prevent odor nuisance complaints.

The physical and chemical properties of MTBE result in ‘early warning’ of potential health risks associated with groundwater transport of MTBE-containing gasoline from leaking underground fuel tanks (12). In differing soil environments with MTBE at 5% to 11% by weight in spilled gasoline it was shown that MTBE arrives at a residential water well 100 meters down gradient anywhere from 54 to 59 days before benzene and the MTBE concentration reaches 100 g/l when the benzene concentration is about 0.1 to 0.3 g/l at the well. In each case, the upper bound cancer risk estimates associated with MTBE and benzene are below 1 per billion incremental risk for exposures prior to water concentrations reaching 100 g/l.

Figure 7 provides estimates of peak indoor air concentrations of MTBE versus water concentration using a conservative application of the McKone model for predicting exposure to volatiles in tap water. Reported odor detection limits for MTBE vary in the literature from air concentrations of 190 g/m3 to 690 g/m3. Figure 7 illustrates that MTBE water concentrations in the range of 30 g/l to 60 g/l are necessary to produce a detectable odor during showering. If MTBE remains below the odor detection threshold during showering, it is unlikely that the resident will detect any adverse aesthetic qualities (odor and taste) of MTBE during tap water use. Because water agencies will not serve water with an objectionable taste or odor to their customers, the odor threshold appears to be more limiting than the U.S. EPA health advisory level.

Since the primary impact of MTBE on water quality is aesthetic rather than health-based at concentrations approaching current advisory levels, risk management decisions regarding cleanup of MTBE in soil and groundwater should be keyed to potability and use of the water supply and fate and transport considerations at potential points of impact on potable aquifers. For example, if an initial spill cleanup has resulted in limited MTBE contamination such that ground water reaching a downgradient well will be below action levels, no further remediation should be required. Similarly, if a large production well has MTBE contamination near or below advisory levels, blending with less contaminated water is a viable option to maintaining both acceptable aesthetic water quality and preventing any appreciable health risks from MTBE.

Risk Management Considerations for Two Hypothetical Case Studies

The purpose of presenting these hypothetical case studies is to illustrate how the approach to managing the problem and selecting a remedial option may vary with site specific factors. Two scenarios are presented: Scenario 1 - high levels of MTBE present in a domestic drinking water aquifer, and Scenario 2 - high MTBE levels in shallow ground water at a corner gas station UST leak site (Figure 8). While there are no guarantees, it may be possible to predict with some certainty the response from the regulatory agencies to these situations. In both cases, MTBE is detected at high levels, well above regulatory thresholds, yet the regulatory response and remediation approach for each case is markedly different.

In the first scenario, a domestic water supply well is contaminated with high levels of MTBE exceeding state and federal regulatory guidelines. All UST sites and pipelines within about one mile of the well will come under close scrutiny by the lead regulatory agency. Under this scenario a full investigation and cleanup of UST sites will be required including an evaluation of the hydrogeological characteristics of the basin and migration pathways from sources in the basin to the production well. Enforcement orders may be issued to ensure the assessment and cleanup work is completed, and periodic monitoring to determine the progress of cleanup. As responsible parties are identified and their contributions to the problem confirmed they may be required to contribute a share of the cost for well-head treatment, or blending, or an alternative water supply until the contaminated well is brought back on line.

The second scenario is at the other end of the spectrum: a single, isolated UST site with MTBE contamination limited to an unuseable, shallow aquifer. This site would appear to be a candidate for closure or at least a monitor-only containment approach under the Los Angeles RWQCB’s interim procedure if:

  • there is no free product or free product is removed;
  • there is no MTBE in any production well within a mile of the site;
  • there is no saturated soil or saturated soil is removed; and,
  • depth to useable ground water is greater than 20 feet.

For sites that fall somewhere in between these two cases the regulatory agencies seem to be open to considering closure only after it can be shown either through investigation or fate and transport modeling that the site has been stabilized and is no longer a source of MTBE to underlying useable ground water or nearby production wells. Each regulatory agency takes a different approach and until there is uniform guidance from the U. S. EPA or State Water Resources Control Board it is best to consult with the RWQCB or local agency to determine its policy with respect to MTBE and UST cleanup.

Concluding Remarks

The detections of MTBE in California waters appear to be random, low-level occurrences associated with leaking USTs in ground water and boating on surface waters. Remediation of leaking fuel USTs will continue to be driven by the toxicity of BTEX, not MTBE, although the presence of MTBE will typically increase the cost of cleanup by 20% or more. For older spills at sites where contamination is limited to shallow ground water and is not threatening a drinking water well a containment approach or monitoring-only scenario may be appropriate

The taste and odor threshold values for MTBE will drive the regulatory cleanup requirements for MTBE as they are roughly one/half the health-based levels. If agencies will accept that MTBE at low levels (<100 g/l) is an aesthetic rather than health-based water quality problem, blending of water to maintain MTBE concentrations below the California action level should be sufficient to achieve safe and nuisance free water quality. MTBE will remain the focus of intense political and regulatory scrutiny as the impact of recent legislation and continuing research expand our knowledge of the impact of this chemical on public health and the environment.

References [.....Figures appear after References]

1. California Department of Health Services. MTBE in Drinking Water. Summary of Sampling of Public Drinking Water Systems for methyl tertiary butyl ether (MTBE). Internet Site. August 20, 1997.

2. California Regional Water Quality Control Board, Los Angeles Region. "Review Procedure for UST Sites with MTBE." April 29, 1997.

3. IRIS. 1996. MTBE. Washington, D.C.: Integrated Risk Information Service, Office of Health and Environmental Assessment, U. S. Environmental Protection Agency. Down-loaded from National Library of Medicine on-line service.

4. National Science and Technology Council. 1996. Interagency assessment of potential health risks associated with oxygenated gasoline: National Science and Technology Committee on Environment and Natural Resources, Office of Science and Technology Policy (OSTP).

5. McKone, T. E. 1987. Human exposure to volatile organic compounds in household tap water: The indoor inhalation pathway. Environmental Science and Technology 21: 1194-1201.

6. California Environmental Insider. "Santa Monica MTBE Settlement May Set Precedent." Volume 11, Number 3. July 15, 1997.

7. Davidson, James M. "MTBE in ground water and drinking water: a technical data summary." Alpine Environmental, Inc., Fort Collins, CO. February 1997.

8. Davidson, James M. and R Parsons. "Remediating MTBE with current and emerging technologies." Alpine Environmental, Inc., Fort Collins, CO. November 1996.

9. Davidson, James M. Existing and emerging MTBE remediation technologies. Alpine Environmental, Inc., Fort Collins, CO. In Workshop on MTBE Ground Water Issues, sponsored by Western States Petroleum Association, Oxygenated Fuels Association, American Petroleum Institute, Los Angeles, CA, November 4, 1996.

10. Karpel Vel Leitner, N. A. L. Papilon, J. P. Crone, J. Peyrot, and M. Dore. "Oxidation of Methyl tert-Butyl Ether (MTBE) and Ethyl tert-Butyl Ether (ETBE) by Ozone and Combined Ozone/Hydrogen Peroxide. Ozone Science & Engineering. Vol 16, pp 41-54. July 21, 1993.

11. Abstract for Society of Toxicology Meeting, March, 1998. Use of odor threshold data and shower air modeling to determine allowable limits of methyl-tertiary butyl ether (MTBE) in drinking water. R O Richter1, D Suder2, and B D Kerger1, 1McLaren/Hart-Chemrisk, Irvine, CA, and 2Precise Environmental Consultants, Davis, CA.

12. Kerger, B.D., D.G. Dodge, R. O. Richter. "Key considerations for risk assessment and risk management of methyl-tertiary butyl ether (MTBE) regarding leaking underground fuel tanks." The Toxicologist. Abstract Issue of Fundamental & Applied Toxicology 36 (1, part 2): 283 [abstract 1438].

13. TRC Environmental Corporation. 1993. Odor threshold study performed with gasoline and gasoline combined with MTBE, ETBE and TAME. API Publication No. 4592. American Petroleum Institute, Washington, D.C.

14. API. 1994. Odor threshold studies performed with gasoline and gasoline combined with MTBE, ETBE and TAME. API Publication No. 4592. American Petroleum Institute, Washington, D.C.

15. Prah, J.D., G.M. Goldstein, R. Devlin, D. Otto, D. Ashley, D. House, K.L. Cohen, and T. Gerrity. Sensory, symptomatic, inflammatory, and ocular responses to and the metabolism of methyl tertiary butyl ether in a controlled human exposure experiment. Inhalation Toxicol. 6:521-538.


Figure 1: Frequency of MTBE Detections
Figure 2: Highest Reported Detections
Figure 3: Dose Distribution for Ingestion and Inhalation of MTBE
Figure 4: LARWQCB Review of UST Sites with MTBE
Figure 5: MTBE Impact on Remediation
Figure 6: MTBE Remediation Technology
Figure 7: Maximum MTBE Concentration in Ambient Air
Figure 8: Two Hypothetical MTBE Case Studies

mbte-fig1.jpg (83419 bytes)

mbte-fig2.jpg (114371 bytes)

fig3.GIF (15010 bytes)

mbte-fig4.jpg (122886 bytes)

Figure 5: MTBE Impact on Remediation

MTBE Characteristic

Impact on Remediation

Low affinity for organic carbon (log Koc = 1.01), high affinity for water (log Kow = 1.1) Minimal retardation in soil; SVE effective only when spill is recent and MTBE is still in the soil
High solubility (43,000,000 g/l) compared to benzene (1,780 g/l) Significantly elevated concentrations in water compared to benzene makes ground water pumping an effective remediation option
Biodegradability substantially slower than aromatic fuel components MTBE has potential to move much further in the aquifer than the aromatics
Low Henry’s Law constant Air:water ratios of 20:1 up to 100:1 needed for 99% removal compared to 10:1 to 20:1 for BTEX

Figure 6: MTBE Remediation Technology

Remediation Approach

Key Considerations Affecting Application


Pump and Treat May be quite effective for MTBE compared to more slowly desorbing BTEX compounds. But treatment is difficult and expensive.


Soil Vapor Extraction Most effective when applied soon after release to soil while most MTBE is still in the soil (typically within several years).
Dual Phase Extraction Most effective in removing MTBE from vadose zone and capillary fringe.
Bioventing Limited application due to low biodegradability.
Excavation Effective in soil only; not a final solution.
Air Sparging More effective in conjunction with SVE when applied long after spill has occurred and MTBE has moved into the ground water.


Biodegradation 80% - 90% destruction reported in aboveground bioreactor, however, MTBE not expected to readily degrade in the subsurface.
Natural Attenuation MTBE does not readily degrade in subsurface due to low affinity for and adsorption to soil particles.
Biological Oxidation Addition of oxygen has not been shown to help biodegradability of MTBE.
Thermal Oxidation Effective in controlling off-gas emissions from air stripper.
Air Stripping 50% - 90% removal rates are typical. Heating water before air stripping improves performance and increases costs. Off-gas treatment usually required. Multiple units may be necessary depending on water chemistry or presence of high concentrations of MTBE.

img001.GIF (7422 bytes)

Figure 8: Two Hypothetical MTBE Case Studies

MTBE Scenario Municipal well field: high levels in drinking water supply.
Former gas station site:
high levels in shallow ground water.
Site History
Large municipal well field
Serving as public water supply from multiple aquifers.
Land use change: former gas station site redeveloping as a fast food restaurant.
Sources Multiple former and current UST sites and several product pipelines within mile radius of well field. 2 gasoline USTs removed 10 years ago; no other sources within mile of site.
Spill History
Relatively recent (<10 years)
with possible continuing
Historical leaks occurred 10 - 20 years ago.
Subsurface hydrogeology Sandy soil aquifer with high transmissivity and limited clay lenses and retardation zones. Shallow, high TDS ground water at 25 feet. Silty sands with clay horizons and lenses.
Ground water regime Ground water at 100 feet; pumping occurs in 120 - 180 foot depth interval. Useable ground water at 120 foot depth; no production wells within 1.5 miles of site.

Factors affecting the selection of a remedial approach

Concentration of MTBE and benzene in water
MTBE = 400 g/l;
avg = 100 g/l
Benzene = 50 g/l;
avg = 2 g/l
MTBE = 500 g/l
avg = 75 g/l
Benzene = 200 g/l
avg = 50 g/l
Regulatory cleanup levels in ground water
MTBE = 35 g/l
Benzene = 1 g/l
Level of Agency Review (High = 5, Low = 1) Highest level of review (Level 5). Lower level of review (Level 1 or 2).
Remediation options Air stripping or blending option. Containment zone concept or close site.

News PageEDP Home PageInformation Request