Comparing the Health Benefits and Risks of Methyl-Tertiary-Butyl Ether (MTBE) in Reformulated Gasoline (RFG) and Winter Oxyfuels in the U.S. R.G. Tardiff, S.R. Baker, and J.S. LaKind, EA Engineering, Science, and Technology, Inc., Silver Spring, MD, USA; and E.J. Burger, Institute for Health Policy Analysis, Silver Spring, MD, USA
INTRODUCTION
In the U.S., air pollution from automobile emissions has been responsible for numerous adverse consequences on health, including increased incidences of cardiovascular incidents, upper-respiratory distress, and an increase in the risk of cancer. Other industrialized nations have realized the same consequences to varying degrees.
For over 20 years, industry and government have been working to find cost-effective means of reducing emissions to improve public health, the result of which has been an evolving set of technologies whose cumulative impact has led to substantially reduced emissions. Since 1990, oxygenates, predominantly MTBE, have been added to gasoline to reduce tailpipe emissions in order to bring about a reduction of the risks of myocardial infarction, respiratory distress, cancer, and other chronic diseases. Oxygenates have been added to fuel at either of two concentrations: either 2.0% (which is equivalent to 11% by volume of MTBE; called "reformulated gasoline" or RFG which is used year round) or 2.7% (which is equivalent to 15% by volume of MTBE; called "oxyfuel" which is used during the winter months only). RFG is used to attenuate the concentrations of volatile organic compounds and ozone, whereas the winter oxyfuel reduces concentrations of carbon monoxide (CO).
Because questions have arisen regarding the safety of MTBE and the effectiveness of MTBE in moving communities closer to compliance with standards, extensive scientific analyses were undertaken in order to provide the scientific community, legislators, authorities, and the public with the most recently available information regarding: (1) human health benefits from the reduction of air pollutants associated with MTBE use; (2) human exposure to MTBE and related compounds; (3) MTBE's potential for causing adverse health effects; and (4) a comparison of MTBE health effects and exposure concentrations causing them, for baseline fuel, MTBE alone, and MTBE in fuel.
HEALTH BENEFITS ANALYSIS FOR RFG AND OXYFUEL
Fuels containing MTBE were compared to "baseline" fuel to assess changes in emissions characteristics and ambient air concentrations of CO, ozone and ozone precursors, volatile organic compounds (VOCs), and lead. Health benefits associated with changes in exposure to concentrations of VOCS, CO, ozone, and lead in air resulting from the use of oxygenating substances in gasoline for light duty vehicles were estimated.
Since 1990, oxygenates, particularly MTBE, have been added seasonally to gasoline in several metropolitan areas in order to reduce vehicle emissions and ambient air concentrations of CO. Modeling and direct measurements have confirmed a consistent record of decreases of 6 to 15% in atmospheric CO as a result of this program and as much as 30% in one community. Benefit to health of this reduction focuses on the cardiovascular system and, in particular, on those persons who have underlying coronary heart disease and are prone to heart attack and death from heart attack. An existing body of evidence from both controlled clinical studies and epidemiological investigations indicates that approximately 1400 excess heart attacks and nearly 500 fatal heart attacks might be prevented or postponed as a result of this program and that the number of persons over 65 admitted to the hospital for cardiovascular disease might be reduced by 400. These estimates have been determined conservatively, and the true numbers could be larger. The calculated benefits should be considered not an absolute figure, but rather as an indication of order of magnitude.
Analyses of the effects of Phase I of the RFG program indicate a consistent reduction in emission of ozone precursors from light duty vehicles, and a small (1-2%) but consistent trend in reduction in ozone formation. For a severe ozone pollution day in metropolitan Los Angeles, analysis estimated that a reduction of ozone by 5 ppb, because of MTBE use, should result in about 90 people avoiding FEV1, losses greater or less than 10%, about 90 people avoiding moderate chest discomfort, and reduction in the death rate by 0.095%, or about 1 in 1,050. For people in the Northeast experiencing 150 ppb exposures, this analysis estimated that a reduction of ozone by 1 ppb due to MTBE use should result in about 100 people avoiding FEV1, losses greater or less than 10%, about 100 people avoiding moderate chest discomfort, and reduction of the death rate by about 0.019%, or about 1 in 5,260.
Use of MTBE in RFG and oxyfuel alters the emission pattern of tailpipe VOCs. The health benefits analysis focused on selected VOCs: benzene, 1,2-butadiene, polycyclic organic matter, acetaldehyde, formaldehyde, ethylbenzene, toluene, and xylene. Using a 30% marketplace penetration, this analysis demonstrates a decline in cancer incidence: use of RFG in winter is estimated to reduce cancer incidence by 33 cases (20%, not seasonally adjusted), while the use of RFG in summer yields a reduction of 12 cases (12%, not seasonally adjusted). Using oxyfuel in winter results in a reduction in cancer incidence of 39 cases (23%). Data indicate that, under certain circumstances, exposure to acetaldehyde, benzene, ethylbenzene, and toluene may exceed established effect levels for effects other than cancer, and that transient formaldehyde exposure at the upper limits of the measured ranges may approach levels associated with discomfort.
HEALTH AND SAFETY ANALYSIS FOR MTBE IN RFG AND OXYFUEL
The focus of this evaluation is on MTBE because it is the most
widely used oxygenate in fuel and because it is used in relatively
large quantities. The evaluation of MTBE's health risks has been
conducted within the context of its presence in gasoline.
Exposure to MTBE
MTBE exposure via inhalation occurs by (1) living in the vicinity
of MTBE manufacture, storage, or commercial or consumer use, (2)
working in occupations related to MTBE where the potential exists
for higher than average exposure, (3) driving an automobile fueled
with MTBE-containing gasoline, or (4) refueling. About 30% of
the U.S. population (73 million individuals) is estimated to reside
in heavy MTBE-use areas of the country. Considering combined
MTBE exposures via inhalation and ingestion, the number of individuals
exposed to average daily doses from 0.00001 mg/kg-day to 0.1 mg/kg-day
is estimated to range from 540,000 to 2,300. Considerable uncertainties
are associated with these estimates.
Acute Toxicity
When oxygenated fuels were introduced into Alaska, after four years of use in other states, complaints of headache, cough, eye irritation, burning of the nose and throat, nausea, dizziness, spaciness and/or disorientation were reported by individuals who tended to associate these symptoms with exposure to oxygenated fuel. Clusters of similar complaints were also reported in Milwaukee, Wisconsin, and New Jersey. Since 1992, several well-designed, methodologically rigorous, epidemiologic and clinical chamber studies have assessed whether exposure to RFG and oxyfuel causes acute health symptoms in humans. None of these studies showed an association between exposure to RFG or oxyfuel and the development of acute health symptoms. An examination of animal data demonstrates that MTBE is of low acute toxicity, unlikely to affect humans at the low concentrations at which it is present in oxygenated fuels. The doses that produce acute health effects in animals are more than three orders of magnitude greater than the maximum short-term human exposures during refueling and commuting. A review of the two chemicals associated with the use of MTBE in gasoline (i.e., tertiary butyl formate and formaldehyde) has led to the conclusion that these chemicals are not likely to be causally associated with the reported health symptoms.
Chronic Toxicity - Carcinogenicity
Scientific evidence demonstrating MTBE's carcinogenicity to rodents at high doses may lead to the conclusion that MTBE is a possible human carcinogen at doses much higher than could be experienced by workers and consumers. MTBE may well cause high-dose tumorigenic effects via a nongenotoxic mode of action. MTBE has produced benign kidney tumors in male rats and benign liver tumors in female mice at doses that appear to have exceeded the maximum tolerated dose. Male rat kidney tumorigenesis was accompanied by significant target organ cellular toxicity and cell proliferation. The potential significance for human carcinogenicity of the male rat kidney tumors is unknown because they may have been caused by a mode of action that is unique to the male rat and thus not relevant to humans. The validity of this hypothesis is currently being tested. Similarly, it has been suggested that the mode of action of MTBE liver tumorigenesis in female mice is similar to that of unleaded gasoline and involves estrogen modulation (i.e., suppression) resulting in an increase in liver tumor promotion. Since estrogenic activity in the mouse is not similar to estrogenic activity in either the rat or the human, this tumor may have no relevance to evaluation of potential human carcinogenicity. The potential for inducing cancer at low doses in humans is low and is likely to be zero
Chronic Toxicity - Health Effects Other Than Cancer
No data show that MTBE produces chronic toxicity in humans. The
results of animal studies demonstrate that MTBE is of low systemic
toxicity and induces adverse effects only at high doses (i.e.,
thousands to hundreds of thousands times higher than those to
which humans are exposed). No reproductive or developmental effects
were observed in animal studies. Significant immunological, cardiovascular,
hematological or pulmonary effects in animals have not been observed.
Following assessment for technical feasibility and appropriateness,
extrapolations of equivalent dose information for the oral route
(for which data are minimal) were undertaken using inhalation
data. The approach was suggested for use in generating oral dose
estimates of NOAELs and LOAELs for use in risk assessment for
oral exposures.
Risk Characterization for Inhalation of MTBE in Air
All available toxicity data derived from animals exposed via inhalation to MTBE, gasoline, and gasoline mixtures containing MTBE were aggregated in a series of comparative plots, consisting of exposure concentrations at which health effects have been observed, normalized to fraction of lifespan (as an expression of dose duration) at which individual effects were observed. The toxicity data from a total of 119 individual points clustered within a range of two orders of magnitude at concentrations that are significantly greater than those likely to be experienced in occupational (transportation worker) or public (consumer) exposure settings.
The data indicated no sensitivity to exposure duration, as evidence
that time-weighted averaging of MTBE exposures is inappropriate
for risk assessment. The toxicity data for pure MTBE, baseline
gasoline, and gasoline mixtures containing MTBE were in the same
range, but the dose duration data for baseline gasoline and mixtures
containing MTBE (not pure MTBE) tended to occur at lower concentrations
than those for MTBE alone. In each case, the margin of safety
was in excess of 100,000.
Risk Characterization for Ingestion of MTBE in Tap Water
This analysis suggested that MTBE in tap water is not anticipated to be a public health concern under the specified conditions of exposure. A range of estimates of human NOAELs were combined with estimates of average daily doses in tap water to assess the potential health risks, and led to estimated margins of safety between 100,000 and 200,000. Since many of the assumptions in the analysis are highly conservative, potential health risks are unlikely to be underestimated. The evidence indicates that lifetime tolerable exposure levels range between 700 ppb and 14,000 ppb.
Some individuals may be exposed to extremely high concentrations
of MTBE in tap water immediately following a localized spill,
and these concentrations will exceed levels that might be excessive
if exposure continued for an indefinite duration. However, high
human exposures are likely to be brief in duration because a major
spill will typically lead to emergency response and remediation
that includes the institution of public health measures to minimize
human exposure, such as closing impacted wells or public water
supplies and providing alternate sources of drinking water. Furthermore,
the low odor and taste thresholds for MTBE ensure that its presence
is readily detected at concentrations that are well below the
human NOAELS.