Executive Summary

Health Canada Ethanol Expert Panel Workshop Report - Executive Summary

Table of Contents

• Executive Summary
• Expert Panellist Presentations
  • Review of Emission Studies of Ethanol Fuel
  • Ethanol-Blended Fuel Photochemistry / Air Quality Modelling
  • Exposure Assessment
  • Toxicology
• Ethanol Expert Panel Evaluation
  • Summary of Responses to Specific Topics
    • Vehicule Emmissions
    • Ethanol-Blended Fuel Photochemistry / Air Quality Modelling
    • Exposure
    • Toxicology
  • Summary of Overall Evaluation
• The Ethanol Expert Panelists

Executive Summary

Ethanol has been used as a transportation fuel for more than a century. Recently there has been renewed interest in ethanol-blend gasoline as a potentially cleaner burning alternative fuel as a result of concerns over emissions (i.e., air pollutants) from vehicles which use gasoline only. The most common form of ethanol-blend fuel is known as E10, which is a mixture of 90% gasoline and 10% ethanol. The ethanol acts as an oxygenate (i.e., a source of oxygen) and octane-enhancer, increasing combustion efficiency. Ethanol has also been used as an additive to replace other gasoline additives (e.g., methyl tertiary butyl ether, or MTBE) which were seen as worse in terms of environmental and health impacts. Most recently, increased use of ethanol-blend fuel has been suggested as a measure to reduce greenhouse gas emissions on a fuel-cycle basis from the transportation sector.

Some of the observed air quality benefits of ethanol-blend fuel include reduced emissions of carbon monoxide (CO) and exhaust hydrocarbons, and the displacement of some air toxics such as benzene. However, advances in emission control technology over the years have reduced the relative advantage of ethanol as a cleaner fuel. In addition, there are some concerns over potential human exposure to certain other emissions related to the use of ethanol-blend fuel (e.g., ethanol, acetaldehyde, formaldehyde, peroxyacetyl nitrate).

Evaluation of the influence on human health and the environment of various positive and negative changes in emissions associated with ethanol-blend gasoline is difficult to carry out. As well, evaluation of both benefits and concerns are confounded by the ongoing improvements in vehicle emissions control technology and gasoline formulation (e.g., reduced sulphur and benzene levels) which have substantially reduced automobile exhaust emissions over the last couple of decades and have produced improvements in air quality in many jurisdictions.

Overall, motor vehicle emissions continue to be of concern from a human health perspective. Due to the major uncertainties in emissions inventories and emission modelling, Health Canada recognizes the need to work with its partners to understand what additional information is needed to effectively evaluate ethanol-blend fuel emissions and the potential health effects associated with them. Consequently, in May 2003 Health Canada held an Ethanol Expert Panel Workshop sponsored by Environment Canada with the aim of engaging experts to consider the potential human health impact of possible future widespread use of ethanol-blend gasoline in Canada. The key points covered in the Expert Panel presentations (section A) and in the structured discussion sessions to develop responses to Health Canada's specific questions (section B) are presented below.

A) Expert Panellist Presentations

Review of Emission Studies of Ethanol

Fuel Blends
Tom Durbin
College of Engineering-Center for Environmental Research and Technology
University of California

Deniz Karman
Department of Civil and Environmental Engineering
Carleton University

This presentation focussed on the results of various studies of ethanol fuel blends to provide a review of the expected impacts of increased ethanol use in gasoline on emissions inventories. A 1983 study by the Colorado Department of Public Health was one of the first to be done (CDH, 1983). In 1988 the U.S. Environmental Protection Agency summarized the work that had been done to date and determined estimates of the reductions in carbon monoxide (CO) and total hydrocarbon (THC) emissions of 21-35% and 5-15%, respectively (U.S. EPA, 1988). These estimate ranges indicate differences in reductions attributed to different technologies.

Based on an evaluation of studies addressing emissions benefits for oxygenates, and emissions results and reactivity for toxic air contaminants (e.g., acetaldehyde, formaldehyde, benzene), it can be concluded that the use of ethanol in gasoline can provide reductions in CO and HC vehicle emissions, and some reduction in benzene emissions. On the other hand, ethanol may also cause some increases in emissions of nitrogen oxides (NOX), and also tends to increase acetaldehyde emissions. While a number of studies have shown mixed results, these trends are more consistent in larger research programs.

In terms of future research needs, additional data on more modern vehicles would be useful (e.g., E-67 Program), as well as data comparing ethanol blends with non-oxygenated reformulated fuels. Finally, controls on vapour pressure (RVP) should be adopted to minimize increases in in-use evaporative emissions.

Ethanol-Blended Fuel Photochemistry / Air Quality Modelling

Mike Lepage
Consulting Engineers
RWDI West Inc

Little research had been done on photochemistry and air quality modelling with regard to ethanol-bend gasoline until the late 1990s, when a number of studies were conducted on ethanol and other oxygenates (e.g., MTBE) in gasoline. This review focuses on such issues as the formation of tropospheric ozone; air quality impacts of the use of ethanol in California Phase 2 reformulated gasoline; fuel composition in various studies; a summary of results from modelling studies; an analysis of field monitoring data; and emission inventory uncertainties.

Based on the current state of knowledge in this field, it can be safely concluded that the use of E10 would result in a 5-15% reduction of CO; a near-neutral effect for NO2 emissions; a fairly neutral effect for ozone in smog events; small increases in aldehydes during smog events; possibly large increases in longer-term average aldehyde (e.g., acetaldehyde) levels; small increases in longer-term average levels of peroxyacetyl nitrate; and a small effect on benzene emission levels, dependent on fuel formulation.

The major areas of uncertainty regarding the use of E10 vs. gasoline are associated with the following factors: E10 versus gasoline emissions for realistic on-road conditions; temporal resolution of modelling; spatial resolution of monitoring data (especially for aldehydes); spatial resolution of modelling; variations in fuel composition; and the lack of data on the use of E10 in non-road gasoline vehicles (e.g., leaf blowers, lawn mowers, and marine vehicles, all of which are becoming evermore common in Canada).

There are currently a number of gridded photochemical modelling opportunities for Canada (e.g., Ontario, Quebec, the Maritimes, Alberta, the Lower Fraser Valley and southern BC). Such models are designed to examine transboundary air pollution effects but could also be used for other purposes.

Exposure Assessment

Jeanette Southwood
Golder Associates Limited

Lynn McCarty
L.S. McCarty Scientific Research & Consulting

There is a current lack of adequate data on human exposure to emissions resulting from the use of ethanol-blended fuel. As a result, "we are trying to hit a moving target, and to do anything of utility would require more data on which to build." One of the problems is that detailed specifications of what "Canadian" E10 gasoline would be are not available. There are existing regulations, including some new ones for sulphur and benzene, but exact characteristics are not clear and thus it is difficult to conduct an exposure assessment. As well, there are related issues such as greenhouse gases and global climate change to consider, but there is much uncertainty in these areas.

An assessment of the available information on exposure to ethanol-blend gasoline suggests that the widespread introduction of ethanol-blend gasoline in Canada will produce modest positive and negative changes in vehicle emissions, air quality, and subsequent human/environmental exposures to various substances of concern. However, to conduct a proper exposure assessment, sufficient and reliable data on exposure are needed.

With regard to health and exposure surveillance, it is premature to comment on health and exposure surveillance issues, as much depends on policy objectives. The important caution is that the available information suggests very modest differences between various types of gasoline (and none is all positive). Large samples will be needed for statistical analysis in order to address significant differences in health or exposure. A major effort is required, and even then the results may be equivocal. Differences might be observed, but may not be statistically significant. A key consideration is that if the outcome is uncertain, the question arises as to whether the effort would be worthwhile.

Qualitative judgments will therefore heavily influence the analysis -- e.g., how the weightings of different compounds are arrived at in the analysis; where exposures occur; etc. These are judgments which will affect the outcome. Hence, there is a need to be transparent about how the data are generated.

This discussion of exposure assessment focussed specifically on the literature addressing exposure concerns with regard to ethanol and aldehydes (acetaldehyde and formaldehyde) in ambient air resulting from the use of ethanol-blend gasoline. Based on the available information, two key conclusions can be drawn. First, prior to the widespread introduction of ethanol-blend gasoline, additional analyses are needed to estimate ambient exposure to ethanol and its atmospheric breakdown products, including constituents such as acetaldehyde, to assess the potential public health impacts of increased ethanol use. In addition, evaluations should consider sensitive subpopulations. Secondly, surveillance programs of various types have been recommended in some studies relating to exposure assessment. Some programs are widespread and others fairly focussed, for various reasons. A very preliminary recommendation based on these studies would be to implement focussed surveillance programs (e.g., measuring ambient concentrations in probable worst-case locations) if the use of ethanol-blend gasoline is widespread.

Toxicology

Michel Charbonneau
INRS - Institut Armand-Frappier
Université du Québec

Robert Tardif
Université de Montréal

Previous modelling studies have suggested that the addition of ethanol to gasoline could modify human exposure to certain emissions constituents, such as ethanol, acetaldehyde, formaldehyde and peroxyacetyl nitrate (PAN). These emissions are anticipated to increase as a result of the potential widespread use of ethanol-blend gasoline in Canada.

The presentation on the toxicological profiles of these major emissions focussed on dose-response assessment, with an emphasis on low inhalation exposure levels in both human and animal species and on quantitative data on dose-response relationships in order to identify threshold levels for derivation of safe exposure concentrations. With respect to potential health effects, the emphasis was on sub-chronic/chronic and acute effects of the four pollutants cited above.

Analysis of the data collected for inhalation exposure yielded the following derived tolerable continuous exposure concentration (i.e., TCEC) values for the three chemicals assessed: acetaldehyde: 270 to 540 ug/m3 ; formaldehyde: 7 ug/m3; peroxyacetyl nitrate: 4.95 to 9.9 ug/m3. Data on the toxicity of inhaled ethanol are sparse and do not permit the identification of a safe level of exposure in humans. There is some evidence that ethanol vapours can cause bronchial hyper-responsiveness in humans at high concentrations.

The rule of thumb in toxicology is that whenever chemicals are causing toxicity in the same target organism by way of the same mechanism, the responses should be summed (i.e., they should be considered additive). So, looking at effects such as irritation and cytotoxicity, that rule applies. The chemicals discussed here should be considered as a group, and their levels in the environment should be summed. Other pollutants in motor vehicle emissions and their possible interactions have not been studied.

B) Ethanol Expert Panel Evaluation

1) Summary Responses to Health Canada's Topic-Specific Questions

Vehicle Emissions

• The case for ethanol as a reducer of regulated and non-regulated exhaust emissions is becoming weaker as the vehicle stock changes and new vehicles take a larger share of the market. By 2010 it is expected that any benefits will probably be minor. With respect to evaporative emissions, emissions of some compounds are higher with ethanol use. There is a need to recognize that these emissions will be even higher with non-tailored ethanol blends, and our technologies to handle these emissions may not be as effective as currently assumed.
• Caveats: There is an agreement that the case for ethanol as a producer of emission reduction benefits is declining. The assumptions made about vehicle stock turnover will affect how quickly this decline will occur. It is important to also consider non-road vehicles, however, as larger benefits might be possible in this area.
• A second caveat is that the results depend on such factors as temperature, vehicle age, etc. Trends can be shown, but we must also consider modelling some aspects.
• A waiver to allow increases in RVP (i.e., vapour pressure) should not be granted for splash-blended ethanol.
• Mixing E10 and conventional transportation fuels (commingling) will have real impacts for the consumer in terms of fuel efficiency and vehicle drivability. These impacts need to be better understood and publicized.

Ethanol-Blended Fuel Photochemistry / Air Quality Modelling

• E10 will have a small impact on atmospheric levels of most relevant emissions (e.g., CO, NOx, benzene, 1,3-butadiene, formaldehyde, peroxyacetyl nitrate) and in many cases transportation fuels are only one of several sources of these emissions. Some emissions, however, do increase with the use of E10 (e.g., acetaldehyde, formaldehyde, ethanol) and the contribution made by these increases might be very important.
• The potential emissions reductions (i.e., air quality benefits) associated with the use of E10 versus conventional gasoline are mainly of CO. There is mixed evidence on how extensive these reductions would be, and some debate as to whether or not these reductions (even if they are real) would provide any public health benefits. We need to better understand this issue in order to fully answer the question.
• There are potential disbenefits associated with ethanol use -- specifically, increased emissions of acetaldehyde, formaldehyde, and ethanol. Potential benzene emission increases associated with splash blending may also be an issue.
• Modelling tools (i.e., current photochemical models) have the ability to address the potential impacts of E10, but there is uncertainty about the data entered into the models (i.e., 'garbage in, garbage out'). Even with good data, however, it is hard to identify the small changes in emissions we are talking about. As a result, it is best to think of these models as providing "directional" conclusions, as opposed to detailed and quantified conclusions.
• The approach previously taken by Health Canada to model the effects of E10 penetration within Canada was a good one, but there is a need to broaden the scope to cover more geographic regions and more time periods. Doing this will necessitate the consideration of the availability of modelling data, but an ideal scenario would consider at least the Prairies, the lower Fraser Valley, and the Windsor-Quebec corridor.

Exposure

• Regarding potential changes in exposure of the general population to certain vehicle emissions as a result of increased use of E10, all of the emissions that we are concerned with will decline in absolute terms over time, but this is not necessarily due to the increased use of E10. These declines will offset increased activity levels, producing absolute declines in all emissions. Within the context of lower absolute emissions, using E10 leads to an increase in some emissions (acetaldehyde, formaldehyde, and ethanol) and a potential decrease in others (e.g., carbon monoxide). The significance of these changes with respect to exposure can only be assessed with an understanding of where we are starting from in terms of chemical compositions in the atmosphere. For example, some chemicals (acetaldehyde, formaldehyde) are presumably already at levels in the atmosphere that warrant concern because they are recognized as toxic substances under the Canadian Environmental Protection Act.
• Regarding possible concerns about potential groundwater contamination by ethanol and subsequent exposure, and the potential risk that ethanol would increase the movement of BTEX in contaminated soil, these topics were beyond the scope of the workshop presentations. Health Canada will independently look at some of the sources identified by workshop presenters, and these issues will not be discussed at this time.
• It is important to consider microenvironments in assessing exposure to ethanol-related emissions. In cases where ambient levels are low and not considered problematic (e.g., ethanol), microenvironments are critical. In other cases (e.g., formaldehyde, acetaldehyde) ambient levels are also important because of the toxicity of the emissions. Different microenvironments will be important for different emissions. Some to consider include: refuelling, pedestrian, parking garages, indoor air, and shower/bath.

Toxicology

• There is potential for concern about the effect of ethanol on the general population and on specific sensitive populations such as pregnant women, ALDH deficient individuals, and people taking medication that block ethanol oxidation at the acetaldehyde stage (e.g., disulfiram, used to treat alcoholics). Lack of data on the impacts of ethanol inhalation makes it impossible to provide a definitive answer, but there is a possibility that we need to consider sensitive subpopulations. When addressing some of the broader questions the extent to which this is a high-priority area in filling knowledge gaps is a key consideration.
• With respect to possible concern with the hazard of E10 vehicle-related exhaust, evaporative emissions, and secondarily formed compounds, the four most potent compounds to be concerned about are: formaldehyde, peroxyacetyl nitrate, acetaldehyde, and ethanol. Importantly, the effects of these emissions can be additive and need to be considered in this way (they are not synergistic).
• With regard to other compounds, reductions in emissions of benzene and of 1,3-butadiene would both be considered important (as both compounds are recognized as toxic substances under the Canadian Environmental Protection Act), but it is uncertain if E10 use provides such a benefit.

2) Summary of Overall Evaluation

Listed below are the summary responses to the specific questions prepared by Health Canada for an overall evaluation of the workshop presentations and ensuing discussion. These responses may be considered conclusions and/or recommendations.

• Regarding the potential increase in level and breadth of the population exposure to E10 vehicle emissions and the potential risk, the government should be concerned about emissions of acetaldehyde, formaldehyde, peroxyacetyl nitrates (PANs), and ethanol from E10 use. This concern relates only to microenvironments in some cases, and to microenvironments and the ambient environment in others.
• In order to specify potential risks and/or benefits and significant areas of uncertainty, Health Canada should conduct a risk assessment (due diligence) of E10 with particular attention to acetaldehyde, formaldehyde, PANs, and ethanol. This effort should be staged -- moving from inventories to modelling to exposure to toxicology. It is likely, however, that all stages will need to be completed. The only area where fundamentally new research is needed is on the toxicology of ethanol inhalation, and particular attention should be paid to factors (e.g., vapour pressure, or RVP) that can influence emissions of these four chemicals.
• Decisions on whether the federal government should implement a complete/comprehensive surveillance program to follow exposure and potential health effects as ethanol use is increased should not be made until some risk assessment work is done. Nonetheless, there may be some value in doing work in conjunction with or prior to risk assessment efforts in relation to aldehydes and specific microenvironments. That might even help with the risk assessment work.