Does Ethanol Use Result in More Air Pollution?
David Morris and Jack Brondum*
Institute for Local Self-Reliance
The MTBE crisis has taught us the need to do a comprehensive evaluation of the costs and benefits of fuel additives. It has also reminded us that tradeoffs will be involved no matter what fuel or fuel strategy we embrace. To phase out MTBE while maintaining the requirement for oxygenates in transportation fuels will require a vast increase in the use of ethanol. Ethanol is an attractive fuel with many undisputed environmental and economic development benefits. Yet it suffers from one disadvantage. It raises the volatility of gasoline.
To make the transition from MTBE to ethanol in the most rapid and least disruptive manner possible, ethanol blends could be allowed a volatility waiver. Many in the environmental community vigorously oppose this. They fear that increased hydrocarbon emissions lead to increased ozone levels and adverse public health impacts. Their position seems to be that the adverse impact of these increased emissions is so damaging that it outweighs the many undisputed environmental benefits of ethanol.
This report challenges that position. We argue that volatility should be one of the factors evaluated by transportation fuel policy makers but should not be the single most determining factor. We further argue that although ethanol blends do indeed increase mass emissions, this has little, if any impact on ozone formation. Finally, we argue that ozone concentration itself is a minor public health problem compared to the impact of particulate matter or other more toxic emissions where ethanol has a very beneficial impact.
The methyl-tertiary-butyl-ether (MTBE) crisis resulted in part from environmentalists and policy makers focusing on a single environmental impact while ignoring the larger picture. MTBE, a fossil fueled-derived gasoline additive, was embraced because of its benign impact on local air quality; its adverse impact on water quality was never addressed.
The MTBE crisis has taught policy makers two lessons. First, we should do a full cost, not a partial cost, analysis when designing public policies. Second, there are tradeoffs in any policy we adopt.
As a result of grassroots activism, several states are now phasing out MTBE. The debate now focuses on what clean air additives should substitute for MTBE. Currently federal regulations require that gasoline sold in urban areas that suffer from high concentrations of ground level ozone must contain oxygen. If MTBE is phased out, the only oxygenate available in large quantities is plant matter-derived ethanol. Ethanol has many attractive qualities; however adding ethanol to gasoline in small amounts raises the overall level of hydrocarbon emissions (VOCs) from the gasoline. To reduce ground level ozone formation (smog) the federal government has mandated that gasoline sold in highly polluted areas have lower volatility levels (i.e. reduce VOC emissions). This gasoline is called RFG, or Reformulated Gasoline. Often these urban ozone exceedance areas are called RFG areas.
The measure of volatility in gasoline is the Reid Vapor Pressure (RVP), in this case "pressure" is the tendency of a compound to "offgas" or volatize. A 10 percent ethanol blend raises the RVP of gasoline by about one pound. Therefore, to add ethanol to gasoline sold in polluted areas and still remain within the necessary RVP levels, the base or starting volatility level of the gasoline must be lowered by one pound. Such gasoline is sometimes called sub-RVP gasoline. In Chicago and Milwaukee, where gas stations switched from MTBE to ethanol (as a result of customer complaints), sub-RVP gasoline has been supplied by the oil companies.
But oil companies are reluctant to produce low volatility gasoline, especially if they only need to produce it for a few areas within their market region. Thus there is a problem in obtaining a reliable supply of low volatility gasoline at a reasonable price. The EPA, for example, estimates that the increase in price resulting from lower volatility gasoline should be about 2 cents a gallon but companies may charge 5-10 cents a gallon more. The problem of supplying low RVP gasoline to only a few communities in a region is one reason for the huge spike in gasoline prices in the Midwest this past spring.
Ethanol currently has a 1 pound RVP waiver in areas not suffering from high ozone concentration levels. In those communities, ethanol can be mixed with normal volatility gasoline. No such waiver exists for ethanol blends sold in high ozone concentration communities. Extending that 1 lb waiver to RFG areas would allow ethanol to substitute for MTBE with relatively few logistical problems.
The environmental community vigorously opposes such a waiver. Indeed, some leading environmental organizations want Congress to give states the authority to eliminate the existing one pound RVP waiver currently granted to ethanol blends in non-polluted areas. These groups argue that when mass hydrocarbon emissions increase, ozone levels increase and public health is adversely affected. It is that proposition that this paper challenges.
As noted above, the MTBE crisis reminds us that any transportation fuel strategy will have its costs and its benefits. When it comes to MTBE, the environmental community appears to accept that approach. For example, in mid-1999 the Northeast States for Coordinated Air Use Management (NESCAUM), an interstate association of air quality control divisions of six New England states as well as New York and New Jersey, has supported MTBE's continued use even after it has been shown to contaminate water because "the public health benefits RFG provides by reducing air pollution substantially outweigh adverse health impacts from exposure (of the population) to the oxygenate methyl tertiary butyl ether (MTBE) in the air and water."1
Regrettably, the environmental community rarely adopts such a cost-benefit approach with respect to ethanol. Ethanol has many undisputed benefits compared to either gasoline or MTBE.
Ethanol does not pose a water quality threat.
Ethanol has a more benign impact on greenhouse gas formation than gasoline.
Ethanol's use reduces gasoline's most toxic and harmful emissions, particulate matter (PM) and benzene
Ethanol is made from a renewable fuel.
Using plant matter to substitute for petrochemicals substantially reduces both upstream and downstream pollution.
America's farmers are currently facing a profound crisis and substituting ethanol for MTBE would provide a modest or significant benefit to them, depending on whether the farmers themselves own the ethanol plant.
Despite its health, environmental and rural economic development benefits,
many in the environmental community believe that a single factor---the
increased volatility of gasoline when ethanol is added --- should outweigh
all of ethanol's other advantages. We disagree that this single factor
should be accorded such determinative status. We do so for three reasons:
* Ground level ozone has modest public health impacts
* The entire RFG program has a very modest impact on ozone formation
* Ethanol's contribution to ground level ozone formation is trivial or nonexistent
A comprehensive cost-benefit analysis should lead the environmental community not only to support the substitution of ethanol for MTBE, but to support a volatility waiver for ethanol so that it can be substituted more rapidly and easily.
1. High concentrations of ground level ozone have undeniable but modest adverse health impacts.2
Many toxic pollutants are generated when gasoline is burned. Some, like particulate matter, pose significant threats to public health. The empirical evidence suggests that others, like VOCs, which are one of the chemicals that contribute to the formation of ozone, do not.
Clean air regulations target hydrocarbon emissions (VOCs). But the real target is ground level ozone. We will explore in some detail below the relationship of ethanol blends and VOC emissions. But ultimately, the public health issue is not about VOCs but about ozone concentrations. One medical researcher summarized the clinical evidence regarding ozone and public health in this way, "Although ozone has been demonstrated in clinical studies to cause undesirable physical reactions at levels that occasionally occur in ambient air, these effects are minor, temporary, and for the most part, unnoticeable unless an individual is engaged in moderate to heavy exercise."3
The study upon which the current EPA standard is based showed reversible physiological effects in adults engaging in heavy exercise at ozone concentrations above 150 ppb (parts per billion).4 The present standard of 120 ppb was set to allow for an ample margin of safety.
Recently the EPA proposed to reduce the National Ambient Air Quality Standards (NAAQS) for ozone from 120 ppb to 80 ppb. Its staff analysis concluded that such a reduction would prevent about 30 asthmatic admissions in New York City during the summer when ozone levels are high. This represents one tenth of 1 percent of the 28,000 total New York City asthmatic admissions each year.5
In its Regulatory Impact Assessment (RIA), the EPA concluded that the national benefits of such a standard would be modest and there was a possibility that zero health benefits would result from such a reduction. On the other hand, the EPA concluded that there were significant health benefits gained from reducing particulate matter emissions.
2. The Reformulated Gasoline (RFG) program has a modest impact on ozone formation.
The RFG program, in effect since 1996, has significantly reduced allowable volatility levels of gasoline, including ethanol blended gasoline. Its impact on ozone concentrations has been very small.
The national goal is to reduce ozone concentrations to a maximum of 120 ppb. The entire RFG program might reduce ozone levels by 1-3 ppb. Indeed, the National Research Council concluded that "the net impact of RFG on ambient ozone concentrations...is a few percent. For this reason, it is difficult to quantify the specific contribution of the RFG program to the apparent downward trend in ozone."6
Ethanol increases mass hydrocarbon emissions by 15-20 percent. If the entire RFG program will reduce ambient ozone concentrations by 1-3 ppb and if we assume a one-to-one linear correlation between increases in VOCs and ozone formation, then a 100 percent use of ethanol blends might diminish the expected ozone reductions by .15-.60 ppb, a remarkably small amount.
3. Increased volatility from ethanol blends has a very small impact on ozone formation. This is true even if ethanol blends were given a 1 pound volatility waiver in cities that now exceed national safe ozone concentrations.
The last paragraph indicated what the impact would be on ground level ozone concentrations if volatility increases from ethanol blends were directly related to ozone formation. But scientists now agree that increases in mass VOCs are in fact not linearly related to increases in ozone concentrations. The reactivity of the different hydrocarbon gases emitted from the car is important. Thus, if ethanol blends result in a 15 percent increase in emissions but because the composition of those emissions is different their reactivity with ozone-producing chemicals is reduced by 15 percent, then no more ozone will be formed.
This approximates what happens. It is widely accepted that ethanol blends will increase the volume of evaporative VOCs, which are less reactive (e.g. butane), and decrease the volume of exhaust pipe VOCs, which are more reactive and more toxic (e.g. benzene).
As noted above, ozone is not emitted from a car but is formed by the interaction of various chemicals in the presence of sunlight. These chemicals include nitrogen oxides (NOx) and carbon monoxide (CO). Ethanol blends increase NOx emissions but significantly decrease carbon monoxide emissions. A 10 percent ethanol blend contains 3.5 percent oxygen, which reduces carbon monoxide emissions by 15 percent or more.7 The contribution of carbon monoxide to ozone formation is now widely recognized by among others, the California Air Resources Board (CARB), the EPA and the National Research Council.8
CARB has presented the following table showing the difference in emissions between a hypothetical zero oxygen gasoline that meets RFG standards versus two that meet the same standard but contain increasing proportions of oxygen from ethanol. Two percent oxygen means a 5.7 percent ethanol blend. Three and a half percent oxygen translates into a 10 percent ethanol blend.
Expected Percent Change in Reactivity-
Adjusted Ozone Forming Emissions Pollutant Zero oxygen 2.0% Oxygen 3.5% Oxygen
NOx -5.4% -2.4% -0.8%
Exhaust Hydrocarbons -1.5% -3.3% -5.7%
Evaporative Hydrocarbons -7.2% -5.0% +2.0%
CO 9 0% 0% -0.2%
Total emissions 10 -3.1% -3.5% -4.5%
As we can see, the reduction in expected ozone forming emissions is modest for all three gasolines, although it is highest for a 10 percent ethanol.
CARB focuses on reducing NOx. But atmospheric scientist Gary Whitten of ICF notes that if the tradeoff of reducing NOx is to increase hydrocarbon and carbon monoxide emissions, the environment would be poorly served. The reason, according to Whitten, is that a reduction in hydrocarbon and carbon monoxide emissions has a much greater beneficial impact on ozone formation than an equivalent reduction in NOx. Whitten concludes, "The effectiveness of THC for reducing ozone in these simulations must be as much as 8 times better than NOx reductions on an equal percentage of the mobile emissions basis."11
If MTBE is phased out and ethanol is not used, the alternative is for oil companies to reformulate their gasoline. A reformulated gasoline that contains no oxygenate will often contain higher proportions of aromatic chemicals to achieve sufficient octane ratings. Dr. Michael Graboski of the Colorado School of Mines notes that California's new fuel standards allow for an increase in aromatics and concludes, "Considering CO effects only, the ozone reduction benefit of using a 3.5% oxygen provided by an ethanol blend with 26% aromatics compared to a non-oxygenated fuel with 34% aromatics is nearly 5%."
A vigorous debate is going on among scientists as to the precise impact of ethanol blends on ozone formation. For us, the most important aspect of this debate is that the difference of opinion revolves around an astonishingly small impact. Whitten, concluded that a 10 percent ethanol blend resulted in a 0.1 ppb increase in ozone formation compared with an MTBE blend, but that was still some 0.2 ppb less than the peak ozone emissions that would result from using a 100 percent gasoline, non-oxygenated fuel.12 In another analysis, ICF took into account CARB's projected 5 percent NOx increase and 5 percent VOC decrease from a 10 percent ethanol scenario and concluded that it decreased ozone formation compared to the non oxygenated fuel by 1.0 ppb.
The bottom line is that whichever side one supports, we are talking about an impact that is trivial. Whether ozone concentration levels increase or decrease by one part per billion should not be the determining factor in deciding whether the environmental community and policy makers support ethanol.
The CARB and Whitten studies assumed current standards. That means that ethanol blends used in these ozone exceedance areas had to meet the same volatility standards as MTBE blends. However, a series of studies done in the early 1990s concluded that even if ethanol blends were allowed a higher volatility in RFG areas (that is, urban areas that exceed ozone standards), the reduction in VOC emissions' reactivity and in carbon monoxide emissions resulting from a 10 percent ethanol blend offsets the increase in mass VOC emissions and in NOx.
The most comprehensive study was done in Chicago. "(T)he objective of this study is to examine the net balance between the compensating effects of the higher vapor pressure," the authors explained. "(T)he results of this study essentially show no difference in ozone formation between the two candidate oxygenates...As used in this study, the model generated a maximum ozone concentration of 116.7 ppb if all gasoline contained ll percent MTBE while the simulated maximum ozone concentration from the l0 percent ethanol blend was 116.2 ppb."13
Several years later the National Research Council came to the same conclusion. As the Chairman of its Committee on Ozone-Forming Potential of Reformulated Gasoline reported, "the overall impact on ozone of ethanol-containing fuel with 1 psi higher RVP would likely be quite small".14
*Dr. David Morris is Vice President of the Institute for Local Self-Reliance and the author of many reports on ethanol and rural economic development. Dr. Jack Brondum is an epidemiologist and former public health official.
1 The NESCAUM document continues, "This conclusion stems largely from the fact that many of the toxins reduced by using RFG are far more potent than MTBE and that tens of millions of Northeasterners benefit from ozone reductions and reduced vehicle air toxic emissions, while contaminated drinking water affects a relatively small percentage of Northeasterners." The Health Effects of Gasoline Constituents. August 1999. NESCAUM. p. 4.
2 Background air has concentrations of ozone of 30-50 ppb or higher. In some western areas of the country with heavy vegetation, ozone concentrations of 50-75 ppb are not uncommon. EPA Ozone Staff Paper. EPA/R June 1996. pp. 20-21.
3 Kenneth Chilton, EPA's Case for New Ozone and Particulate Standards: Would Americans Get Their Money's Worth? June 1997. In 1986 the EPA concluded, "reported effects on the incidence of acute respiratory illness and on physician, emergency room, and hospital visits are not clearly related with acute exposure to ambient ozone or oxidants and, therefore, are not useful for deriving health effects criteria for standard-setting purposes. Likewise, no convincing association has been demonstrated between daily mortality and daily oxidant concentrations; rather, the effect correlates most closely with elevated temperatures." EPA, Air Quality Criteria for Ozone and Other Photochemical Oxidants. EPA August 1986. More recently, the EPA has concluded that ozone can cause a wide range of respiratory symptoms including coughing, throat irritation, chest pain, shortage of breath and increased susceptibility to respiratory infection. The effects on people who are asthmatics can be severe and result in hospital admissions. Most if not all of the acute effects are temporary and reversible. The EPA cites a "possibility" that repeated inflammation associated with ozone could do sufficient damage to lungs to affect people's quality of life as they grow older, but concedes that these relationships are "highly uncertain". The EPA does not assert that ozone kills people. J.W. Anderson, Revising the Air Quality Standards. Resources for the Future. 1999. "Typical subjects experience less than a 5 percent loss in lung function even at the highest ozone levels recorded in the United States in 1996(about twice the current standard)". While noting that the decrement in lung function is transitory and reversible, the staff paper conjectures that people with preexisting illnesses that limit pulmonary function may suffer more significant effects but notes, "Unfortunately, not enough is known about the responses of these individuals to make definitive conclusions regarding their relative sensitivity to O3". Ozone Staff Paper. EPA/R June l996. pp. 20-21 and p. 55.
4 44 Federal Register 8220, February 8, 1979. Also see Kenneth Chilton and Anne Sholtz, Battling Smog: A Plan for Action. Formal Publication No. 93. St. Louis. Center for the Study of American Business. Washington University. September 1989 pp 7-14 for review of clinical studies.
5 R.G. Whitfield, A Probabilistic Assessment of Health Risks Associated with Short Exposure to Tropospheric Ozone: A Supplement (produced by Argonne National Laboratory, Argonne IL for U.S.EPA. January 1997. Table 6, as cited in Chilton, EPA's Case for New Ozone and Particulate Standards, Op. Cit.
6 Ozone Forming Potential of Reformulated Gasoline. NRC. Washington, D.C. September 1999. Statement by William L. Chameides, Chair of the National Research Council Committee on Ozone Forming Potential of Reformulated Gasoline before the Subcommittee on Energy and Environment, Committee on Science, U.S. House of Representatives. September 14, 1999.
7 Statement of Dr. Michael S. Graboski, Director, Colorado Institute for Fuel and High Altitude Engine Research, Colorado School of Mines, on behalf of NCGA. Committee on Commerce. Subcommittee on Health and Environment, US House of Representatives. March 2, 2000.
8 The NRC's report concludes, "The contribution of CO to ozone formation should be recognized in assessments of the effects of RFG".p. 6. CARB now gives a credit for carbon monoxide reductions in its predictive atmospheric air quality model. The EPA may soon issue rules that would give gasoline blends that reduce carbon monoxide emissions a volatility credit of .2-.5 pounds per square inch.
9 Changes in carbon monoxide presented as reactivity adjusted to evaporative hydrocarbons.
10 Total HC and CO emissions. This row of figures contains weighted sums and is not a simple arithmetic sum of the four rows above.
11 Gary Whitten, Letter to the EPA regarding CARB request for waiver. February 7, 2000
13 Gary Z. Whitten, et. al. Comparison of the Air Quality Effects of Ethanol and MTBE in Reformulated Gasoline in the Lake Michigan Region. July 26, 1993. SYSAPP-93/083. Prepared for Council of Great Lakes Governors, Chicago, IL.
14 Statement by William L. Chameides, Chair of the National Research Council Committee on Ozone Forming Potential of Reformulated Gasoline. Before the Subcommittee on Energy and Environment, Committee on Science, U.S. House of Representatives. September 14, 1999.