Академический Документы
Профессиональный Документы
Культура Документы
PROBLEM INVENTORY
and
EVAPORATIVE MEASUREMENTS
REGISTRERINGSUPPGIFT Rapport nr
MTC 5407
Utgivare: Projektbeteckning Utgivningsr/mn
8050407 2005-05
AVL MTC Motortestcenter AB
Box 223 Uppdragsgivare: ISSN: 1103-0240
SE-136 23 Haninge EMFO
Frfattare: ISRN:
Tel: +46 8 500 656 00 Prof. Karl-Erik Egebck ASB-MTC-R05/2--SE
Fax: +46 8 500 283 28 Mr Magnus Henke
Mr Bjrn Rehnlund
e-post:
Mr Mats Wallin (coordinator)
magnus.henke@avlmtc.com
Associate Prof. Roger Westerholm
Rapportens titel Sprk:
Engelska:
ENG
Blending of Ethanol in Gasoline for Spark Ignition Engines Problem Antal
Inventory and Evaporative Measurements. sidor:
Svenska:
Inblandning av Etanol i Bensin fr Ottomotorer Probleminventering 111+ 20
och Avdunstningsmtningar.
Sammanfattning:
Presently all gasoline sold in Sweden contain 5 % of ethanol. Ethanol mixing above 5% is not possible because of
the EU-fuel directive as well as The European standard EN 228. However, there is in Sweden an interest in further
increasing the bio-ethanol content in gasoline up to at least 10 %. The main reason is that even relatively small
percentage additions will result in a substantial total volume of gasoline substitution, and the present infrastructure
for distributing fuels can be used largely unchanged.
Based on that background the purposes of the study were to collect information on national and international
findings and experience related to the use of blends of ethanol in gasoline as fuels in spark ignition engines. The
project also included a first study on the impact of the evaporative emissions with different grades of base gasoline
and different blending proportions of ethanol.
The main conclusion from using ethanol-gasoline blends in practice is that blends with up to 15 percent ethanol
will not have any significant negative effects on the wear of the engine or vehicle performance.
No significant difference can be seen in regulated emissions when comparing the use of blended fuel (with up to 10-
15% ethanol) to the use of neat gasoline. Concerning unregulated emissions views differ. Regarding the emissions of
benzene, toluene, ethyl benzene and xylene (BTEX) the main conclusion is that there is a slight decrease when using
ethanol blends, while for aldehydes there is a significant increase, especially of acetaldehyde and (to a lesser extent)
formaldehyde emissions. However, there is a serious lack of data describing the effects under Swedish conditions.
There will be a slight increase (~2-3%) in fuel consumption when shifting from neat gasoline to a 10 percent
ethanol-gasoline blend, depending on the design of the vehicle. Cold starts, in particular, will affect fuel
consumption more when using blended gasoline than when using neat gasoline.
There is a need to generate data and experience by running tests and analysing the environmental effects of
blending ethanol with gasoline. The lack of data is more marked for blends with high ethanol contents (~20 %).
Such blends should be avoided before a thorough analysis has been carried out and more data are available.
PROBLEM INVENTORY
and
EVAPORATIVE MEASUREMENTS
Study performed by Stockholm University, ATRAX AB, Autoemission KEE Consultant AB,
AVL MTC AB.
2004-2005
Page 2
Contents
0. Summary ...................................................................................................... 4
1. Introduction.................................................................................................. 6
1.1. Background.....................................................................................................6
1.2. Alcohol Blended Fuels ...................................................................................8
2. The Use of Ethanol in Different Countries.............................................. 10
2.1. Prerequisites for the Literature Study...........................................................10
2.2. Fuel Ethanol in Sweden................................................................................10
2.3. Fuel ethanol in the USA/California ..............................................................11
2.4. Fuel Ethanol in Japan ...................................................................................15
2.5. Fuel ethanol in Brazil ...................................................................................15
2.6. Fuel Ethanol in East Asia .............................................................................17
2.7. Fuel Ethanol in Australia..............................................................................19
3. Data and Experience................................................................................. 23
3.1. Vehicle Performance ....................................................................................23
3.2. Cold Starts and Driving ................................................................................25
3.3. Impact of Fuel...............................................................................................25
3.4. Impact of Lube Oil .......................................................................................28
3.5. Impact on Service and Maintenance.............................................................29
3.6. Compatibility and Wear................................................................................30
3.7. Impact of Vapour Lock.................................................................................33
4. Air Quality and Health Effects................................................................. 35
5. Emissions .................................................................................................... 37
5.1. Regulated exhaust emissions ........................................................................37
5.2. Theoretical Discussion about HC Emissions................................................38
5.3. Unregulated Exhaust Emissions ...................................................................38
5.4. Characterisation of Exhaust Emissions ........................................................42
6. Fuel Energy Content Engine Power...................................................... 45
Page 3
10. Life Cycle Analysis (LCA) of Gasoline and Ethanol Blends ................. 80
10.1. LCA in the Context of This Project..............................................................80
10.2. LCA Studies for Ethanol in Neat Form or Blended with Gasoline ..............81
10.3. Discussion.....................................................................................................83
10.4. Findings ........................................................................................................84
11. Conclusions and Recommendations ........................................................ 85
11.1. Fuel Reid Vapour Pressure ...........................................................................86
11.2. Regulated Emissions.....................................................................................87
11.3. Unregulated Emissions .................................................................................88
11.4. Performance and Wear .................................................................................88
11.5. Life Cycle Analysis ......................................................................................89
12. References................................................................................................... 90
0. SUMMARY
The purpose of the study reported here was to collect information on national and international
findings and experience related to the use of blends of ethanol in gasoline as fuels in spark
ignition engines. The main reason for blending ethanol with gasoline is to reduce fossil carbon
dioxide emissions (and thus the greenhouse effect) from vehicles by using bio-ethanol
originating from renewable sources. Presently all gasoline sold in Sweden contain 5 % of ethanol
(here and throughout the text such percentages refer to the alcohol content of blends by volume)
and the relevant authorities are interested in further increasing the bio-ethanol content in
gasoline. Blending bio fuels with a petroleum-based fuel has the twin advantages that even
relatively small percentage additions will result in a substantial total volume of gasoline
substitution, and the present infrastructure for distributing fuels can be used largely unchanged.
Another aim of the project was to study the impact of using such blends on evaporative
emissions by carrying out measurements with different grades of base gasoline and different
blending proportions of ethanol. A report on these measurements can be found in the appendix to
this report.
The results and experience presented in this report originate from Sweden, the USA, Japan,
Brazil, China, India, Thailand and Australia. The main topics discussed are the effects of using
blends on: vehicle performance, cold starts and drivability, fuel and lubricating oil performance,
service and maintenance, compatibility and wear, vapour lock, emissions (regulated and
unregulated), fuel consumption, fuel energy content, Reid Vapour Pressure (RVP) and Life
Cycle Analyses. The main conclusions of the study are as follows:
There is a need to generate data and experience by running tests and analysing the environmental
effects of blending ethanol with gasoline. The lack of data is even more marked for blends with
high ethanol contents (~20 %). Such blends should be avoided before a thorough analysis has
been carried out and more data are available.
In the light of the situation and conditions in Sweden and the other countries belonging to the
European Union there are certain barriers to overcome in order to succeed with the intention to
increase the content of ethanol in blended gasoline. In addition, given the differences in
conditions and regulations between Sweden, other countries belonging to the European Union,
and regions where there is long experience of running vehicles on blended fuels, a number of
issues have to be addressed before the alcohol content of blends is increased.
A first issue to address is the problem that the RVP increases when ethanol is blended with
gasoline since current gasoline standards impose limits on its RVP. Therefore, either there must
be an exemption for ethanol blended fuels or the base gasoline RVP must be adjusted. Such
adjustments are already made today to the base gasoline used in the 5 % ethanol gasoline blends.
A second issue is concern about the performance and start-ability of vehicles at low
temperatures, which commonly occur in wintertime, especially in the northern parts of Sweden.
A third issue is whether blends with 10 to 15 percent ethanol in gasoline will affect human health
and the environment (both local and regional).
The report includes 181 references and was financed by the Swedish Emission Research
Program (Emissionsforskningsprogrammet, EMFO).
Page 6
1. INTRODUCTION
1.1. Background
A challenge that humanity must take seriously is to limit and decrease the greenhouse effect
caused by various human activities.
A major contributor to the greenhouse effect is the transport sector* due to the heavy, and
increasing, traffic levels. In spite of ongoing activity to promote efficiency, the sector is still
generating significant increases in CO2 emissions. As transport levels are expected to rise
substantially, especially in developing countries, fairly drastic political decisions may have to be
taken to address this problem in the future. Furthermore, the dwindling supply of petroleum fuels
will sooner or later become a limiting factor.
An important step in efforts to solve the problem is to replace fossil source energy with
bioenergy. In the transport sector this means either introducing bio fuels and using adapted
vehicles, or blending bio fuels with petroleum-based fuels for use with present vehicle fleets.
The two alternatives are not, of course, mutually exclusive. However, blending bio fuels with
petroleum-based fuels for use by the present conventional vehicle fleets has the advantages that
even using quite low blending concentrations will result in substantial total volumes of gasoline
being substituted by bio fuels, and that the present infrastructure for distributing fuels can be
used.
Today, the transport sector is a major contributor to net emissions of greenhouse gases, of which
carbon dioxide is particularly important. In Sweden this sector accounts for roughly 20 % of
total energy consumption, and almost 50 % of the total net emissions of carbon dioxide. The
carbon dioxide emissions originate mainly from the use of fossil fuels, mostly gasoline and
diesel oil in road transportation systems, although some originates from other types of fossil
fuels such as natural gas and Liquefied Petroleum Gas (LPG).
If international and national goals (such as those set out in the Kyoto protocol) for reducing net
emissions of carbon dioxide are to be met, the use of fossil fuels in the transport sector has to be
substantially reduced. This can be done, to some extent, by increasing the energy efficiency of
engines and vehicles and thus reducing fuel consumption on a volume per unit distance travelled
basis. However, since the total transportation work load is steadily increasing such measures will
not be sufficient if we really want to reduce the emissions of carbon dioxide. In order to reduce
absolute amounts of these emissions we have to go further and an additional measure that will be
required is to replace fossil vehicle fuels with renewable ones. Primarily, especially in the short
term, this means bio-based fuels.
Probably the best candidate bio fuels to replace gasoline in the short term are alcohols. Alcohols
can be blended with gasoline or used as neat fuel in both optimised spark ignition engines and
compression ignition engines. In the medium term ethanol produced from grain will probably be
*
Approx 30% according to the Annual Report of Greenhouse Gas Emissions from the Swedish EPA, 2002
Page 7
the most important alternative fuel for replacing gasoline, and in the long term ethanol produced
from cellulose might take over from grain ethanol.
Today, ethanol accounts for a substantial part of the alternative fuel market, especially in Brazil,
the USA and Sweden. The advantages of ethanol are that it can:
From an international perspective, most research up to 1990 was focused on blends of methanol
and gasoline, but some studies were carried out on ethanol-gasoline blends. Since these studies
were carried out in the USA, it can be assumed that they mainly included vehicles with efficient
emission control systems, but at the same time technical features of cars in the USA have
historically differed, at least in part, from those in Sweden. It should also be noted that for a long
time 10% ethanol has been added to commercial gasoline in many parts of the USA. In the USA
there is considerable experience of adding higher proportions of ethanol to gasoline than those
allowed by gasoline regulations in Sweden (Europe). The primary advantage of adding a bio-
based alcohol to gasoline is that it reduces net CO2 emissions but it also has other positive
effects, such as increasing the octane value of the fuel and reducing the benzene content of the
exhaust gases.
The use of alcohol blended gasoline and neat fuel alcohols as substitutes for neat gasoline have
become matters of interest in many countries. The International Energy Agency (IEA),
established in 1974, follows the development, and data and other experience from various trials
have been presented and discussed at symposia organised by the International Symposium on
Alcohol Fuels (ISAF). The first ISAF-symposium was organized in Stockholm 1978 and since
then a symposium has been organized every 2 to 4 years.
Today, almost all the alcohol fuel used is ethanol and it has three main uses in Sweden: as neat
ethanol in ca 400 buses; in a gasoline blend (E85) for Flexible Fuelled Vehicles (FFVs), of
which approximately 17 000 were being used in Sweden in February 2005; and as a component
of all of the other gasoline (E5) used throughout the country. This means that the most common
alternative fuel used in Sweden is ethanol. Only approximately 65 000 m3 (50 000 m3 from
wheat and 15 000 from cellulose) of this alcohol is domestically produced and at the time of
writing around 165 000 m3 is imported from Brazil. The goal for the future is to increase the
amount of domestically-produced ethanol from cellulose (ligno-cellulose) and one step toward
this goal is research to be carried out at a pilot plant.
The goals of the project presented in this report are to accumulate the data required to facilitate
increased use of bio fuel by:
Studying available literature, collected knowledge, identified data as well as yet
undocumented experience concerning emissions when using ethanol blended gasoline.
Evaluating the relevance of existing investigations and the data generated in them.
Assessing what (if any) emission studies are needed to estimate reliably the effects of
using ethanol blended gasoline on total emissions, both qualitative and quantitative.
Measuring evaporative emissions from the combustion of different blends of ethanol and
neat gasoline.
Page 8
The use of E85, a mixture of 85 % ethanol and 15 % gasoline, for FFVs has become common.
Blends with other percentages of ethanol in gasoline are commonly used in various countries
around the world, especially Australia (officially 10 %), Brazil (up to 25 %), Canada (10 %),
Sweden (5 %) and the USA (up to 10 %). There is still debate about whether, how and to what
extent ethanol in gasoline may affect the materials in the vehicle and cause excessive wear of
parts in the fuel system and the engine. However, in the USA, car manufacturers have agreed
that use of gasoline with up to 10 % ethanol will not affect the warranties of their vehicles
(Science Fair Projects Encyclopaedia, 2004; Launder, 2001). In his MSc thesis Launder
described the development of the use of fuel alcohol, especially the use of ethanol in the USA.
Amongst other salient facts noted by Launder Minnesota has also passed legislation requiring
the use of 10% ethanol in all gasoline.
Since both methanol and ethanol have considerably lower energy contents (15.7 MJ/l and 21.4
MJ/l, respectively) compared with gasoline (approximately 35 MJ/l) use of an alcohol-
containing blend may affect the power output of the engine to varying degrees, depending on its
design. In section 6 the calculated effects on the energy content of the fuel of blending ethanol
with a specific gasoline are presented. According to these calculations, adding ethanol to a final
volume of 10 % to a gasoline with an energy content of 32.3 MJ/litre will decrease that value by
3.4 %.
Blending alcohol in gasoline will affect, inter alia, the vapour pressure of the fuel and, as shown
in section 8, the increase in vapour pressure is considerably larger when blending methanol than
when blending ethanol. The alcohol is commonly added to gasoline when filling the tank of the
vehicle that will deliver the fuel to the gas stations. More sophisticated blending technologies
such as Ratio Blending, Sidestream Blending and Wildstream Blending are described and
discussed in a paper by Toptech (2004). Of these Ratio Blending is designed for use when up
to six components are to be blended. Sequential Blending of ethanol in gasoline means that the
two components are pumped to the delivery truck in sequence. Even if it is computer controlled
there are some uncertainties about whether the resulting mixtures will fulfil their specifications.
Sidestream Blending is similar to ratio blending and is used when two or more components are
to be mixed together and Wildstream Blending can be used when blending ethanol with
gasolines of many different qualities simultaneously. In Figure 1.1 the configuration of
Sidestream Blending is shown.
Page 9
Since there are concerns related to the fact that ethanol is readily miscible with water it is
important to use water-free systems when blending ethanol in gasoline. It is, of course, also
important to use cost-efficient methods.
At least five major reasons for blending ethanol in gasoline are discussed, including the
following advantages of using ethanol as an alternative fuel:
3. The need (especially in the US) to improve the air quality in non-attainment areas, i.e. areas of
the country where air pollution levels persistently exceed the national ambient air quality
standards.
In Table 1.1 selected fuel properties for a neat gasoline and ethanol blends with the neat gasoline
are shown. Density, RON and MON numbers and the IBP (initial boiling point) of the blended
fuels increase as ethanol contents increase, while 10 %, 90 % and FBP (final boiling point)
decrease. Using the blended fuels reduces brake specific energy consumption.
Table 1.1, Fuel properties of ethanol gasoline fuel blends (He et al., 2003)
Fuel parameters Neat gasoline E10 (10 % ethanol) E30 (30 % ethanol)
Density at 19C 0.736 0.741 0.751
RON 92.4 95.0 99.7
MON 81.2 82.3 86.6
IBP 36.0 37.5 40.0
10% 55.2 49.0 52.7
50% 92.5 73.2 72.5
90% 153.7 149.8 145.7
FBP 184.5 181.0 181.5
Page 10
Vehicle performance/wear.
Fuel.
Vapour pressure.
Emissions.
Air quality and health effects.
Life cycle analyses.
The Swedish producer of ethanol, Svensk Etanolkemi AB (SEKAB) has presented specifications
for fuel ethanol to be used when blending ethanol with gasoline, see Table 2.1 (SEKAB, 2004).
The oil crises of 1973 and 1976 prompted a search in Sweden for alternatives to gasoline for
light duty vehicles and a project was initiated to study the possibility of replacing some gasoline
with an alcohol. Since a governmental investigation had been launched in 1965 in order to study
an introduction of emission regulations it was also seen as the use of an alcohol blended gasoline
could be one measure to reduce especially the emission of CO
The work within the project included both laboratory and field tests. The results of laboratory
investigations and field trials that had been carried out in various locations around the world
were reported to the IEA in 1986 by STU and SDAB. (STU, 1986;1987). The countries that
helped prepare the report were Canada, Japan New Zealand, Sweden and the USA. A great deal
of valuable experience concerning the use of alcohol and (especially) alcohol blended gasoline
was gathered during this co-operative assessment of alternative fuels. However, at this time
Brazil was the only country in the world that was focusing on ethanol as an alternative to
gasoline.
Page 11
Table 2.1. Sales specification of technical ethanol 99.5% SPSE-Ethanol 99.5% 3 4 1 (1)
2004-11-03 SPSE-410.
Parameter Limit Method of analysis
Ethanol % by volume min 99.8 ASME 1112
% by weight min 99.7
Density (D 20/4) g/ml max 0.790 SS-ISO 758
Appearance Clear, without particles ASTM D 2090
Colour Hazel max 5 AMSE 1102
Water % by weight max 0.3 SS-ISO 760
Aldehydes (as acetaldehyde) % by weight max 0.0025 AMSE 1118
Acidity (as acetic acid) % by weight max 0.0025 AMSE 1114
Fusel oil mg/l max 50 AMSE 1107, GC-
method
Methanol mg/l max 20 AMSE 1107, GC-
method
Distillation interval*: ASTM D 1078
- starting point C min 77
- drypoint C max 81
Flashpoint* C +12 SS-EN 22719
Explosion limits* % by volume in air 3.5 - 15 Accepted from
literature
Refractive index* nD20 1.3618 Accepted from
Literature
Evaporation residue* mg/l max 10 AMSE 1124
*The seller guarantees these properties, although they are not tested on each delivery.
During the 1980s the focus of investigations and trials with alcohol fuels in Sweden switched to
ethanol, both as an alternative for diesel oil and as an additive to gasoline. At this time the main
rationale for introducing an alternative was to reduce exhaust emissions. Since the introduction
of emission control systems with catalysts for gasoline-fuelled vehicles had efficiently reduced
exhaust emissions, most of the research on alternative fuels in Sweden during the 1980s and
1990s was related to diesel-fuelled vehicles. Several comparative reports on different fuels,
including alcohol-gasoline blends, were prepared, and the most comprehensive investigations are
summarised and discussed in the following text.
MathPro The mandate volume would increase on an annual schedule, reaching 5 billion
gallons per year (bgy) in 2012 of ethanol. Thereafter, annual mandate volumes would be set to
maintain the percentage share of the U.S. gasoline pool that ethanol held in 2012.
Table 2.2 Estimated US Consumption of Fuel Ethanol, MTBE and Gasoline (Thousand Gasoline
Equivalent Gallons), (Department of Energy,1998).
1994 1996 1998 2000 (projected)
E85 80 694 1,727 3,283
E95 140 2 699 59a 59
Ethanol in Gasohol (E10) 845 900 660 200 916 000 908 700
MTBE in Gasoline 2 108 800 2 749 700 2 915 600 3 111 500
Gasolineb 113 144 000 117 783 000 122 849 000 127 568 000
The US Renewable Fuels Association (RFA) has presented industry guidelines for the use of
ethanol in the American market and according to the Association the guidelines represent a
compilation of the key technical aspects of fuel grade ethanol use based on the collective
experience and expertise of our member companies (Renewable Fuels Association, 2002).
The following industry standard is valid as the industry standard for fuel ethanol to be blended in
gasoline to any rate in the USA, Table 2.3.
Table 2.3. Industry standard for fuel ethanol to be blended in gasoline in the US.
ASTM 4806
Property Specification ASTM Test
Method
Ethanol, vol% 92.1 D5501
Methanol, vol% 0.5
Solvent-washed gum, mg/100 ml max 5.0 D381
Water content, vol%, max 1.0 E203
Denaturant content, vol%, min 1.96
Denaturant content, vol%, max 4.76
Inorganic chloride content, mass ppm (mg/L), max 40 (32) D512
Copper content mg/kg, max 0.1 D1688
Acidity (as acetic acid, mass-% (mg/L), max 0.007 (56) D1613
pH 6.5 9.0 D6423
Appearance Visibly free of suspended or
precipitated contaminants (clear &
bright)
In the cited report the RFA gave certain recommendations and discussed a number of effects
linked to the use of ethanol blended gasoline. The producers of ethanol and member companies
of the RFA were recommended to add corrosion inhibitors to all of their fuel grade ethanol at a
treat rate to provide corrosion protection comparable to the treatments applied to fuels such as
neat gasoline
Page 13
According to the RFA adding ethanol will affect several properties of gasoline, including its
octane number, volatility, water solubility and oxygen content (the oxygen content of ethanol is
approximately 33 % by weight). The octane numbers of ethanol quoted in various reports and
other data sources differ, but in the cited paper from RFA (Renewable Fuels Association, 2002)
the blending research octane number (RON) of ethanol is 129.0 and the blending motor octane
number (MON) 96.0. Chevron has also reported that ethanol has a Blending Research Octane
Number (BRON) of 129 (Chevron, 2002-2005).
An issue that should be addressed is whether (and if so, to what degree) the octane numbers will
increase when blending ethanol in gasoline, and here too opinions differ. In the paper from RFA
it is stated that the octane numbers (R+M)/2 will increase by 2 to 3 units.
Since fuel ethanol has a higher octane number than gasoline it may well be true that adding
ethanol to gasoline increases the octane number, and thus improves the performance of the
vehicle. However, there is considerable uncertainty about the extent to which the engine power
will improve and whether this improvement will occur across the whole range of ethanol
contents in gasoline. During an experimental investigation of ethanol blends in gasoline Abel-
Rahman and Osman found that the maximum engine indicated power improvement occurred
with a 10 % ethanol blend when adapting the compression ratio of the engine to the fuel Abel-
Rahman and Osman, (1997). Performance tests were carried out when using different
percentages up to 40 % of ethanol in gasoline.
Maintaining the safety of the working environment is an important issue when working with
automotive fuels. In its Industry Guidelines, Specifications, and Procedures, the RFA
recommends that ethanol should be handled with the same safety precautions as gasoline and
that sparks and flames should be avoided when handling ethanol. A further recommendation is to
wear safety goggles when handling ethanol Renewable Fuels Association, (2002). In addition,
a guidebook released by the US Department of Energy (DOE) recommends that skin and vapour
contact with E85 should be avoided and that ethanol-resistant gloves should be used (US
Department of Energy, 2001). Further valuable information on many aspects of using alcohols in
gasoline can be found in literature from the USA/California, where experience in their use
(especially ethanol) and test data (both laboratory and field based) have been gathered over a
long time. Unfortunately, however, few reports on measurements of emissions obtained when
using ethanol blended gasoline have been found.
Progress towards the use of alcohol-gasoline blends as fuels has been underway since at least the
1970s, according to a report by Launder (2001). The Arab countries embargo in the 1970s
against the USA was the main factor that initially prompted use of alcohol fuels (ethanol and
methanol) as substitutes to compensate for the resulting drop in gasoline supplies. An Energy
Tax Act was passed, exempting 10 % ethanol-gasoline blends from the 4% gallon tax on motor
fuels applied at the time. In the 1980s three additional Acts were passed that promoted
production of ethanol from corn, inter alia, and these measures were reinforced by a blenders
credit of 40 cents per gallon. It should be noted that the big automobile manufacturers Chrysler,
Ford and GM were in favour of the E10 (gasohol) and they stated that its use would be covered
in their vehicle guarantees, a move that was subsequently followed by almost all manufacturers.
Since the environmental protection authorities also favour the use of ethanol blended gasoline it
has became common in US to blend 10 % ethanol in commercial gasoline. The Clean Air Act of
1990 mandated the use of reformulated gasoline (RFG) in certain areas of the USA to ameliorate
air quality problems, such as high ozone levels, and the use of oxygen blended fuel in areas with
high levels of carbon monoxide during the winter. In order to meet the demand for ethanol as a
Page 14
blending component a 1-psi waiver in the permitted vapour pressure parameter (RVP) has been
allowed for ethanol-gasoline blends, although it has been shown that ethanol will raise the
vapour pressure of the fuel. An increase in vapour pressure has been shown to increase the
emissions of VOCs (volatile organic components), which may in worst case scenarios result in
enhanced levels of ozone (US EPA, 1993). Therefore, the EPA has argued that no waiver should
be applied, as can be seen in the following quotation: EPA believes that ethanol can and will
play an important role in reformulated gasoline without a 1.0 psi waiver, and that granting such a
waiver would therefore be unreasonable (US EPA, 1993).
At the end of 2001 A bill to amend the Clean Air Act to address problems concerning methyl
tertiary butyl ether, and for other purposes was introduced to the US Senate concerning
reformulated fuel. In a summary of the bill the following seven amendments (abbreviated here)
were recommended to the Senate:
Actions should be taken concerning the leakage of MTBE from corrodible underground
tanks, which represents a risk for public health, welfare and the environment and
inspections of underground tanks should be carried out.
The Clean Air Act should authorize any State for which a waiver is in effect to impose
control of any fuels and fuel additive for the purpose of water quality protection.
Requires the Administrator of the Environmental Protection Agency to ban the use of
MTBE in motor fuel within four years of this Acts enactment.
The State Governor should be authorized, upon notification of the EPA Administrator,
to waive oxygen content requirements for reformulated gasoline other than those
regarding oxygen content to be reformulated gasoline. This Requires regulations to: (1)
ensure that toxic air pollutant emissions reductions achieved under the reformulated
gasoline program are maintained in such States; and (2) establish performance
standards.
The amendments require the Administrator to: (1) carry out tests in order to evaluate
health and environmental effects of the use of fuel and fuel additives. Furthermore, tests
should be carried out in order to study the effects of using MTBE and other ethers which
may be used to replace MTBE; (2) to publish analysis showing changes of the air quality
resulting from implementation of the Act; (3) to complete a model which reflects the
effects on the emissions that are related to the characteristics or components of the fuel
used during 2005.
The Act eliminates the waiver that allows higher RVP limits for ethanol blended
gasoline.
Allows State implementation plan revisions that apply conventional gasoline
prohibitions to non classified areas.
The Act also authorizes grants to MTBE merchant producers to assist in conversion of
production facilities to the production of other fuel additives. Authorizes appropriations
(US Senate 2001).
2.4. Fuel Ethanol in Japan
It has been difficult to find information on investigations carried out in Japan concerning the
introduction and use of ethanol in either neat or blended forms. One reason for this is that
interest in ethanol has been low in Japan. Another is that the few relevant reports that have been
found have nearly all been written in Japanese.
Page 15
However, Japanese interest in alternative fuels has been concentrated for many years on natural
gas or liquid fuels that could be produced from natural gas, such as methanol and DiMethyl
Ether (DME). In recent years Japan has also shown increasing interest in so-called synthetic
fuels, such as paraffin (synthetic diesel) and (to some extent) alkylate (synthetic gasoline) fuels
produced from natural gas (and maybe in the future from gasified biomass) by the so-called
Fischer-Tropsch technique.
In the last few decades methanol has been the main Japanese alternative to gasoline. However,
according to some unconfirmed reports and personal communication with Jan Lindstedt, The
Swedish Bioalkohol fuels Foundation, BAFF, Japan has started to introduce ethanol and is
currently planning to increase its use.
One of the main requirements linked to the introduction of PROALCOOL was that there should
be close coordination amongst the authorities and other parties involved, namely the Ministry
of Agriculture and sugarcane planters, the Ministry of Science and Technology and research
centres, the Ministry of Industry and Commerce, the automobile industry, Ministry of Mines and
Energy, PETROBRS (state owned oil company), the fuel distributors, and the gas stations, the
Ministries of Finance and Planning and, last but not the least, the automobile owners. There
was also a requirement that subsidies should be used to stimulate the production of cars to be run
on alcohol, and the relaxation of tax on industrialised products is seen to have been effective in
this context.
An important benefit of the program is that it has provided stable employment for
approximately 500 000 workers in the sugar cane plantations and similar numbers in the
alcohol/ethanol production industry and other activities connected to the use of ethanol fuels
such as transportation, blending etc.
There have also been considerable benefits in terms of greenhouse gas reductions since ethanol
produced in Brazil is exported to and used in many countries, especially India, Japan, China,
Thailand and Australia. A certain amount is also exported to Sweden and blended in gasoline.
Page 16
.
Figure 2.1. Growth in ethanol production (anhydrous and hydrous) in Brazil from 1975/76 to
2003/4.
Unfortunately, few papers prepared in Brazil and written in English have been found. Therefore,
the effects that ethanol-gasoline blends with up to 25 % ethanol contents have had on the
drivability, deterioration and wear of vehicles and last (but not least) the emissions and air
quality in urban areas are not clear.
According to Szwarc, the use of ethanol as a blending component has increased over the years
and the content of ethanol in Brazilian gasoline blends has increased from 4.5 % by volume in
1977 to 25 % in 2002, as shown in Figure 2.2. However, today more than 4.2 million cars are
ethanol-powered according to data and information presented by Online TEFL. Of these, about
40 % are passenger cars and in total the 4.2 million vehicles annually consume approximately 14
million m3 of ethanol (Ethanol Curriculum Online TEFL Module 5, 2005).
Page 17
30
25
20
15
10
5
0
1977 1979 1981 1985 1998 1999 2002
YEAR
Figure 2.2. Changes in the maximum ethanol content of ethanol-gasoline blends in Brazil,
1977-2002 (Szwarc, 2004).
No actual data from emission tests in Brazil have been found, but the following observations
were made by Szwarc (2004) at the workshop in Buenos Aires.
China
In 2001 the Chinese State Development Commission launched an ethanol program and after
careful consideration issued quality standards for denatured ethanol and ethanol blended
gasoline. In the same year Beijing (China) also organised a World fuel ethanol congress (2001).
It is well known that there is an urgent need to improve the air quality in China, especially in
Beijing, and when it launched the program the Chinese government announced that China plans
to spend nearly $12 billion on a program to cut smog and pollution in Beijing by 2008. This
emphasises one of the reasons for the large-scale introduction of ethanol, and when launching
the program it was also said that China may soon become an ethanol industry leader... The
program for the Congress also indicates that this event was considered a platform for initiating
use of ethanol as an automotive fuel on a broader scale (Beijing World Fuel Ethanol Congress
2001).
Page 18
In China ethanol is commonly produced from corn and sorghum (Nan et al., 1994). According to
the Beijing Times, China is pushing the use of ethanol as fuel by constructing a plant in the
Henan province that will produce 300 000 tons of fuel per year (Beijing Times, 2002). A
headline in the Beijing Times also stated that China Promotes Ethanol-Based Fuel in Five
Cities. Otherwise, this should be changed to The Beijing Times has also reported that China is
promoting the use of ethanol-based fuel in five cities. These five cities are Zhengzhou, Luoyang
and Nanyang in Henan province and Harbin and Zhaodong in Heilongjiang province, northeast
China (Beijing Times, 2002a). In order to force motorists to use fuel ethanol, all vehicles
carrying a licence plate starting with Yu A have to use ethanol-based fuel according to an
order by the Zhengzhou Municipal Government. The specification of ethanol-based fuel is not
clear, but in another paper it is said that ethanol will be blended with lead-free gasoline in a 1:9
ratio (US Commercial Service, 2003). In the province of Liaoning new regulations state that car
owners must use ethanol as a fuel for their vehicles. If they do not switch to ethanol they will be
fined between 5,000 yuan (US$600) and 30,000 yuan (US$3,600) since "Ethanol fuel can play
an important role in easing consumption of traditional petrol and protecting the environment"
according to the Liaoning Development and Reform Committee (Peoples Daily, 2004).
India
In India, like other countries around the world, there was a considerable shortage of oil and sharp
rises in crude oil prices in the 1970s, prompting interest in fuel ethanol. In India too, plans have
been developed to use ethanol-gasoline blends (Uppal, 2002; Bhanot and Chaudhari, 2003), and
even ethanol-diesel oil blends (Acharya et al., 2002).
Based on recommendations by a committee for the development of alternative fuels for motor
vehicles, trials were carried out in New Delhi in 1991. The test fleet included 93 vehicles and
comprised cars, vans and jeeps. The fuels used were blends of 5 % and 10 % ethanol in gasoline.
The results of the field trials are summarised in the following quotation:
There seems to have been a lull in the development of fuel ethanol in India during the 1990s.
However, according to the available literature considerable activity is now underway.
Since the authorities in India found that the agriculture sector needed support and that air quality
had to be improved the Ministry of Petroleum and Natural Gas Resolution started to examine
issues related to the introduction of ethanol blended gasoline. According to the specifications set
by the Bureau of Indian Standards for Gasoline up to 5 % ethanol can be blended with gasoline
(The Gazette of India, 2002). In addition, Winrock International India* initiated the formation of
*
Winrock International India (WII) is a non-profit organization working in the areas of natural resource
management, clean energy and climate change.
Page 19
an Ethanol Coalition of India in 2000 to promote the development of fuel ethanol (Mishra,
2002).
Thailand
In the autumn of 2002 Thailand hosted the 14th ISAF International Symposium on Alcohol
fuels, entitled The Role of Alcohol Fuels in Meeting the Energy, Environmental and Economic
Needs of the 21st Century. At this symposium many papers prepared in Thailand were
presented, some of which gave information on the development of fuel ethanol production in
Thailand. In addition a number of relevant papers reporting investigations and other activities in
Thailand have been found in Internet searches.
In November 2001 the RFA reported that the Government of Thailand had approved the use of a
10 % ethanol/gasoline blend in order to increase the production of ethanol in the country. One
reason why the government wants to increase the content of ethanol in gasoline is the need to be
less dependent on imported oil and to promote the use of domestically produced energy carriers
(Renewable Fuels Association, 2001). It has also been reported that there is an attempt in
Thailand to replace MTBE by ethanol (World Association of Beet and Cane Growers, 2001).
One way to stimulate this replacement is seen to be a plan to decrease the excise tax on ethanol
mixtures.
According to the literature there are five or six main potential sources of ethanol in Thailand,
including sugar cane, molasses and cassava (Ethanol Industry in Thailand, 2004) and sweet
sorghum (Thanonkeo et al., 2002). The scope for producing ethanol from sweet sorghum has
been investigated via a process presented by Thanonkeo et al . In addition, production of ethanol
from bagasse and rice straw has been studied by Siwarasak and Wirivutthikorn (2002).
This is also reflected in a paper presented by Dr. Kemp (the Australian Minister for the
Environment and Heritage), where it is stated that "The major automobile manufacturers have
advised my Department that they accept the use of 10 per cent ethanol blends and that such
blends will not affect vehicle warranties," (Ministry for the Environment and Heritage, 2003)
According to the paper and other reports from Australia there is interest in blends with higher
alcohol contents, and investigations have been carried out with blends of up to 20 % ethanol in
gasoline and some conclusions from these investigations will be discussed below (Orbital
Engine Company, 2003; 2004). The main results from these tests have been presented in
comprehensive summaries in the cited reports.
In 2004, the ethanol production capacity in Australia amounted to 135 000 m3. Of this, around
55 000 m3 was used for blending with gasoline. About 30 35 000 m3 was exported for other
purposes and the rest was used domestically, but not as an automotive fuel. There are plans to
increase the production capacity of fuel grade ethanol incrementally by about 270 000 m3, in
total, by building three new production plants. An interim production target will be half of this
Page 20
amount. However the production of 270 000 m3 of ethanol, less than the 360 million litres
required to supply ethanol for an E10 blend in Queensland alone. (Boswell, 2004).
Today, ethanol is produced from sugarcane and wheat, but the possibility of using other
feedstocks (such as barley, corn and sorghum) is discussed. There is also interest in producing
ethanol from other biomaterials, such as wood, i.e. lignocelluloses (Cheung et al., 2003).
The discussion about fuel ethanol in Australia has heavily focused on two issues: future gasoline
quality standards and labelling of the gasoline pumps.
In September 2000 a paper entitled Setting National Fuel Quality Standards was released by
the Natural Heritage Trust on the Proposed Standards for Fuel Parameters (Gasoline and
Diesel) and the Revised Commonwealth Position. One of the main positions taken by the
Commonwealth was that the oxygen content in gasoline must be higher than 2.7 %, as proposed
in the standards for gasoline, to allow the continued use of a 10 % ethanol-gasoline blend already
available on the market. The proposal by the Commonwealth was that an exemption should be
made for ethanol so as the standard could be the following when referring to Summary of
revised Commonwealth proposal for fuel quality standards where the specification for gasoline
is: Oxygen content: 2.7 % (max) with an exemption for ethanol blends up to 10 % (Natural
Heritage Trust, 2000).
Noting the Commonwealth position, Environment Australia published a paper entitled Setting
the Ethanol Limit in Petrol which officially invited the public to send in their views on the limit
for ethanol blends in gasoline (Environment Australia, 2002).
After years of discussions, in which representatives of the petroleum industry, the Australian
Automobile Association (AAA), various gasoline suppliers and the insurance company NRMA*
(which covers cars inter alia), have been involved (the authors of this report conclude that
NRMA is not willing to cover the risk for damage to cars that may be caused by a high
percentage of ethanol in gasoline). The outcome of these discussions prevented Environment
Australia (and thus the Government), from allowing higher alcohol contents than 10 % in blends
with gasoline. NRMA has also been conducting research on ethanol gasoline blends used in
motor vehicles (Australia Automobile Association, 2002; 2002a; NRMA, 2003) and found that
there could be problems with higher alcohol contents. A study conducted by Apache Research
Ltd. for NRMA showed that 10 % ethanol could be accepted, but the use of a 20 % blend led to
increased wear (see section 3.1, below).
All those who were strongly opposed to higher ethanol contents than 10 % supported the 10 %
limit and, in fact, welcomed the opportunity to use gasoline with that of blended ethanol. The
AAA says that they support ethanol contents in gasoline of up to 10 %, and the policy of
labelling ethanol blends at the gasoline pump. The opinion of the NRMA has been that no higher
ethanol content than 10 % should be allowed until comprehensive research shows that it will
not damage fuel systems and engines (Australian Automobile Association 2002; NRMA, 2002;
*
National Roads and Motorists Association
Page 21
2002a). In a paper presented in 2003, Shell notes that the Australian Institute of Physics* (AIP)
has been taking positive steps towards achieving the Governments [350 000 m3] target for bio
fuels by participating in various activities related to the further introduction of bio fuels,
including ethanol. The position of the AIP is that the introduction of ethanol under the
government policy should be viable if the consumers feel confidence in its use as a fuel (Shell,
2003). Early in 2003 BP announced that it had been delivering a 10 % blend of ethanol in
gasoline, but was going to stop producing it since consumer confidence in ethanol blended
gasoline was low (BP, 2003). In December 2003 BP announced that it will limit its marketing to
a 10 % blended regular unleaded gasoline in Queensland after the middle of December 2003.
However, Dr George Nicolaides of BP, admitted that ethanol blended gasoline fuels have a role
to play in Australia as a renewable fuel and as an octane enhancer (BP, 2003a).
The actions taken by the petroleum industry, the AAA and NRMA obviously created pressure on
Environment Australia and the Australian government since it resulted in the Minister for the
Environment deciding to set a 10 % limit for ethanol blends in gasoline in April 2003. He also
announced that he was going to appeal to the State Governments to require labelling at fuel
pumps delivering ethanol blends (Federal Government, 2003; Federal Government, 2003a).
The decision taken by the Minister for the Environment is explained in papers released by the
Ministry. One of the papers states that ethanol blends for the Australian market means gasoline
that, as tested in accordance with the Fuel Standards (Gasoline) Determination 2001, contains
more than 1% ethanol. It is also stated that the ethanol blend may contain up to and including
10% ethanol; and a statement that the ethanol blend is the subject of this standard (Kemp,
2003). In a second paper the standards for some of the fuel parameters are presented and in a
third paper the labelling requirements are listed and it is stated that If you sell an ethanol blend
you will have to display a label (Department of the Environment and Heritage, 2004 and
2004a). When delivering blended fuel from a pump the label must be displayed as close as
practicable to each nozzle that dispenses the ethanol blend and for retail supply special
requirements are to be followed. In Figure 2.3 a replicate of the label is shown.
Preparation of the position paper for the Fuel Quality Standards Act 2000 (Office of
Legislative Drafting, Attorney-Generals Department 2004) has started, and the Department of
the Environment and Heritage has invited stakeholders to send in comments by by close of
business, 18 February 2005 (Department of the Environment and Heritage, 2004a). A
background paper for Setting a Quality Standard for Fuel Ethanol has been prepared by Hart
Downstream Energy Services (Hart Downstream Energy Services, 2004).
*
Australian Institute of Physics (AIP has gained recognition as a key representative body of Australia's petroleum
industry).
Page 22
Figure 2.3. The Australian label for ethanol blended gasoline. (Department of the Environment
and Heritage, 2004a).
Page 23
Papers from lobby organizations were also often excluded since they tend to be prejudiced and to
present conclusions without giving sufficient information about the basic conditions and data on
which the conclusions are based. When examining reports dealing with blended fuel, wear of the
vehicles and vehicle performance it soon becomes apparent that opinions differ among those
concerned with the use of ethanol and the impact on vehicles of using ethanol. Data on air
quality, health effects and emissions linked to the use of ethanol blended gasoline, are presented
and discussed in two separate sections, 4 and 5, below.
A comment on World Wide Fuel Charter (WWFC) was that WWFC should increase the
maximum oxygen content in gasoline to 3.5% (by mass) in order to allow a blend of 10%
ethanol in gasoline (WWFC, 2002; WWFC Comments, 200)
The response from WWFC is that the WWFC Committee selected a limit of 2.7 % as the general
mass of oxygen in gasoline in order to assure correct operation of the engine. However, a
footnote to this statement indicates that the Committee accepts the use of 10% ethanol in
gasoline if the fuel conforms to the requirements set for the fuel.
Furthermore, the position of WWFC is that the level of 3.5% oxygen content for ethanol blends
in gasoline is too high. There is a WWFC requirement for using co-solvents and inhibitors when
methanol is used.
Due to the maximum oxygen level in gasoline WWFC do not allow an addition of 10% ethanol
to gasoline which may already contain 2% MTBE despite the fact that this is allowed by US
EPA. (WWFC, 2002; WWFC Comments, 200)
Page 24
The Orbital Engine Company in Australia has carried out extensive tests for the Australian
government on vehicles fuelled with ethanol blended gasoline, focusing especially on its effects
on the durability of their components (Orbital Engine Company, 2004)
A Holden Commodore VN, 1990 model, which has an electronic fuel injection system, a
three-way catalyst, a closed loop control system and runs on ULP (unleaded petrol)
gasoline.
A Ford Falcon XE, 1985 model, which has electronic fuel injection, open loop control
systems and runs on LRP (low vapour pressure) gasoline.
A Holden Commodore VK, 1985 model, which has a carburettor and runs on LRP gasoline.
According to the Orbital report the tests followed, as closely as possible, standard SAE protocols
(J1748 for polymeric material, J1748 for metallic material and J1681 for material/component
immersion testing). A 20 vol% ethanol gasoline blend was used as test fuel. The results from the
2000 hour testing and evaluation program can be summarized as follows:
Metallic fuel system: various parts of the fuel system such as the fuel pump, the fuel
injectors, metallic parts of the fuel regulator diaphragm, and the fuel pressure regulator
showed corrosion, tarnishing and pitting.
The fuel tank metal, the fuel sender unit, and the PCV valve/spool showed corrosion and
pitting, tarnishing and the plastic filters were discoloured.
Various parts of the carburettors and associated components of the carburettor-equipped car
showed corrosion, pitting and tarnishing.
The general view derived from studying the international literature is that blends with up to 15 %
ethanol do not have any serious effects on the performance and wear of the vehicle. Many
reports and other documents deal with the use of ethanol, including its effects on performance
and wear, but the most extensive reports focus on flexible fuel vehicles (FFVs), i.e. vehicles
designed to be fuelled with up to 85 % ethanol in gasoline.
Car owner/driver acceptance of a new fuel, e.g. a fuel with a certain level of an alcohol in
gasoline, is highly dependent on the cost of the fuel and the performance of the vehicle.
Concerning drivability: The most important aspect of performance (other than starting) is
acceleration, both from a stop and at highway speeds in order to pass and is a vital issue for
customers according to McLean and Lave (2002). In the Science Fair Projects Encyclopedia
(2004) it is said that both of the ethanol-gasoline variants E10 and "E15", containing 10 % and
15% ethanol in gasoline, respectively, are generally safe for common usage in automobile
engines. Adding a higher percentage of an alcohol like ethanol may affect the performance of the
vehicle, especially its starting and drivability at low temperatures. These effects on the engine
and vehicle may also result in increased emissions, due to sub-optimal operation of the vehicles
emission control systems during the time lag until the engine and catalyst reach the normal
temperature for a continuously running engine. However, ongoing development of engines and
control systems for the fuel and emissions have resulted in considerable improvements related
not only to emission performance but also to drivability. Further advances may solve some of the
problems concerning the cold start parameters of the engines, thus improving their start-ability
and emission performance when used with ethanol-gasoline blends. Although vehicle
performance and wear of the engine and its control systems are linked to a number of different
factors, of varying importance, many significant steps have been taken in recent years which will
reduce these problems.
Page 25
Several methods and devices are available to improve the cold start and emissions during the
warming up period. One recommended for use with FFVs is a specially designed catalyst
presented by the National Renewable Energy Laboratory (NREL). A key feature of the catalyst
tested by NREL is that it is insulated with a Variable-conductance vacuum insulation (NREL,
1996). Another suggestion is to use certain additives in the fuel, but this does not seem to be a
good alternative for ethanol-gasoline blends with low alcohol contents. In Sweden an engine
heater is commonly used to improve the cold starts and warming up of the engine.
Fuel volatility: Adding ethanol to gasoline will increase the volatility, decrease the 50 %
distillation point (T50), and affect both the drivability index and vapour lock protection.
Vapour pressure: This is a measure of front end volatility, and a fuel with extremely high
vapour pressure may cause problems with hot start ability, hot drivability and vapour lock.
Distillation properties: Ethanol in gasoline will reduce the T50 value of the fuel, which may
cause problems with older vehicles in warm weather. It has been shown that later models of fuel
injected vehicles are less sensitive to a reduction in T50 than older cars.
Drivability index: The drivability index is based on the relationship between the distillation
temperature of the fuel and the cold start and warming up parameters of the vehicle. The
following formula can be used to calculate the drivability index (DI):
DI = 1.5 T10 + 3.0 T50 + 1.0 T90
Vapour lock protection index: ASTM D 4814 defines six classes of vapour lock production, as
shown in Table 3.1.
Page 26
Table 3.1. ASTM D 4814 Vapor Lock Protection Class Requirements (Oak Ridge National
Laboratory, 2002).
Vapour Lock Vapour/Liquid (V/L)*#
Protection Class Test Temperature C V/L, max
1 60 20
2 56 20
3 51 20
4 47 20
5 41 20
6 35 20
* At 101.3 kPa pressure (760 mm Hg.)
#The mercury confining fluid procedure of test Method D 2533 shall be used for gasoline-oxygenate blends. Test
Method D 5188 may be used for all fuels. The procedure for estimating temperature-V/L may only be used for
gasoline
(ASTM D 4814e).
The definition according to the ASTM standard ASTM D 4814 is that the Vapour Lock Index,
according to the ASTM standard ASTM D is the ratio of the volume of vapour formed at
atmospheric pressure to the volume of fuel tested in Test Method D 2533 (Renewable Fuels
Association, 2002). As can be seen in Table 3.1, test method D 2533 is to be used for oxygenate-
gasoline fuel blends. The requirement for each of the classes is that the maximum vapour to
liquid ratio formed at the test temperature (TV) must be at most 20 (Oak Ridge National
Laboratory, ORNL, 2002), as shown in the table. The vapour lock protection index is defined as
TV/L20, meaning that higher TV/L temperatures are required for summer grades of ethanol-
gasoline blends and lower TV/L20 temperatures are required for winter grade blends, which are
more volatile. The ORNL has presented two examples, one of which is related to vapour lock
protection class 1, which would have a TV/L20 temperature of 60C while class 6, which is
more volatile would have a TV/L20 temperature of 35C.
In practice this is reflected in the fact that the RVP standards for winter and summer grade
gasoline in Sweden are 95 kPa and 70 kPa, respectively, and that the vapour lock class depends
not only on the climate but also the altitude. A salient issue to consider is whether a 10 %
ethanol content in Swedish summer gasoline would meet the ASTM vapour lock protection
standards. The authors of this report strongly believe that adding 10 % ethanol to Swedish winter
grade gasoline would not create any problems in meeting the ASTM vapour lock protection
standards.
Oxygen content: The standard for the oxygenate content in gasoline is set on a weight basis, as
can be seen in Table 3.2 (Renewable Fuels Association, 2002).
Since there are differences in the density of different grades of gasoline compared to the density
of ethanol the final content of oxygen will vary if the blending volume is fixed to a certain
volume.
Page 27
Water tolerance: It is well known that ethanol has affinity to water, so appropriate measures
must be taken when blending ethanol with gasoline and during the distribution of ethanol-
gasoline blends. According to the RFA the water tolerance of blended fuels is temperature-
dependent, i.e. the tolerance is lower at low temperatures. A 10 % ethanol blend in gasoline will
tolerate approximately 0.5 % water (v/v) at temperatures of 15.5 C or more, while the water
tolerance is 0.3 % (v/v) water at approximately -12 C (Renewable Fuels Association, 2002).
In a report prepared by Krause and finalized by Korotney the possibility of phase separation
occurring in ethanol-gasoline blends contaminated with water has been discussed.. When phase
separation occurs in a blend of ethanol in gasoline, water starts to remove ethanol from the
gasoline and another phase, containing both ethanol and water, forms in addition to the gasoline
and ethanol phase.. The impact of the water-ethanol phase on the engine is greater in a two-
stroke engine than in a four stroke engine, were it may combust in the engine. In a two-stroke
engine the ethanol-water phase will compete with the gasoline-oil mixture and reduce the
lubricating ability of the lubricating oil. However, phase separation normally only occurs in the
presence of liquid water in the ethanol-gasoline blend (Krause and Korotney, 1995).
In a study of a blend of anhydrous ethanol and unleaded gasoline the wear was not found to be
unusually high (US Department of Energy, 1991). However, with 11 % water in ethanol a
significant increase in the wear occurred, since the engine temperature was reduced. Formic acid
and water vapour in the combustion gases during low temperature combustion of alcohols
containing high percentages of water may oxidize metal components in the engine. The acidity
of the combustion gases in combination with the lubricating ability of the lube oil will
significantly increase the wear rate. It is well known that the concentrations of formaldehyde and
acetic acid are relatively high in the combustion gases when methanol or methanol-gasoline
blends are used in a combustion engine. However, certain amounts of these gases are also
formed, in addition to considerable amounts of acetaldehyde and acetic acid, during the
combustion of ethanol or ethanol-gasoline blends.
According to the US DOE, those who are responsible for delivering and blending alcohols in
gasoline are aware of the risks and consequences of contamination of alcohol fuels by water. The
possibility of increased wear of components in the engine is not the only risk, since water in the
fuel can also adversely affect the start ability and driving performance of the engine. Methods for
blending alcohols in gasoline are discussed in section 1.2. However, since the risk for water
contamination of the fuel is a potentially serious problem Chevron in the US has decided to use
ethanol only in areas where appropriate terminal facilities (i.e. where ethanol is properly
distributed) are available, in order to ensure the quality of the blended alcohol-gasoline fuel
(Chevron, 2004).
The RFA has established recommendations for blending ethanol in gasoline, for storage tanks,
for distribution, and at customer delivery points, for protecting pumps and meters. Many of these
recommendations are certainly known by suppliers of ethanol blended fuels. However one of the
recommendations is of great importance for those filling cars with ethanol blended gasoline,
namely that: When first converting to an ethanol program it is advisable to recalibrate meters
Page 28
after 10-14 days to ensure that the change of product has not caused any meters to over-
dispense (Renewable Fuels Association, 2002).
The effects of adding ethanol to gasoline on fuel properties have been discussed by Chandra
Prakash in a report prepared for Environment Canada (Prakash, 1998). She notes, inter alia, the
advantage of using ethanol as an enhancer of the octane number of the blended fuel. As an
example she refers to a comparison (summarized in Table 3.3) of ethanol blended gasoline with
neat gasoline.
Her essential message concerning the octane number is that the high octane rate of ethanol
increases the value of the blended fuel. She also points out that the higher octane number of the
blended fuel confers advantages in terms of fuel efficiency for later models of vehicles since
they are commonly equipped with knock sensors, which retard the ignition timing in the event of
knocking. In such cases the fuel will be more efficiently matched to the engine.
The fact that ethanol has a lower heating value than gasoline may affect the performance of the
vehicle, since the presence of an alcohol in gasoline will lean out the fuel, resulting in some loss
of engine power. According to Prakash this is somewhat offset by the fact that adding ethanol to
gasoline results in a higher volume of combustion gases, which increases the pressure in the
cylinder and thus increases the power efficiency by 1 to 2 percent.
The negative effects of blending ethanol in gasoline in terms of increased volatility, enleanment
(which affects the cold start ability), fuel economy and the water miscibility of ethanol are also
discussed in the cited report (Prakash, 1998). These effects have been discussed in other sections
of the present report, but not the risk for deposits forming in the engine caused by cold starts at
low temperatures. Additives compatible with the engine oil must be used in order to avoid
deposits building up.
Besides pollution from exhaust emissions there are also issues related to the handling of
ethanol/gasoline blends, for instance their storage in underground fuel tanks and the potential
contamination of groundwater due to leaks, which are addressed in a number of reports. Powers
et al. (2001) state that several modelling studies predict that the presence of ethanol in gasoline
will probably increase the BTEX (i.e. benzene toluene, ethylbenzene and xylene) plume from a
leakage. According to Deeb et al. (2002) and Lovanh et al. (2002) simulations indicate that the
benzene plume is likely to increase by 16 to 34% in the presence of ethanol. Neither the true
extent of the potential increase nor the risks for leakage to the ground soil have been well
characterised. In addition there are indications in the reports that biodegradation of benzene is
severely inhibited by the presence of ethanol.
Page 29
Professor Mathur (Delphi) has reported that he has completed a three-year trial on the road using
approximately one hundred government vehicles. He claims that there has not been any problem
related to aspects such as storing the fuel, fuelling the vehicles and other issues (Mathur, 2004).
One interesting observation is that analyses of samples of lubricating oils have detected no
differences in comparisons of the use of ethanol blended fuel and neat gasoline.
A question to consider is whether (and if so to what extent) components in the lubrication oil like
sulphur and other chemicals will react with acidic combustion products. The Alliance of
Automobile Manufacturers (AAM) and the Association of International Automobile
Manufacturers (AIAM) have proposed that additional data on fuel sulphur effects at near-zero
levels of sulphur should be compiled. They also recommended that data should be gathered on
the effects of fuel oxygenates to provide a knowledge base for the California Air Resource Board
(CARB) (AAM/AIAM, 1999).
In a Vehicle Buyer's Guide for Consumers published by the US DOE the owners of FFVs are
advised to check their service manuals to ensure that the right fuel oils are used (US Department
of Energy, 2002). Two of their recommendations for a FFV when using ethanol gasoline blends
are:
Special engine lubricants may be required when fuelling with ethanol. If you are driving a
(FFV), check the owner's manual or consult with the vehicle manufacturer to be sure that you
are using the right engine oil.
When ordering replacement parts for an FFV make sure to let the dealer know you are
fuelling with ethanol.
No such recommendations from the US DOE concerning low ethanol-gasoline blends, i.e. those
with 10 % or less of ethanol, have been found.
Additional service and maintenance instructions to be followed by the service personal and car
owners have not been (and are still not) generally issued for FFVs. However, according to the
National Renewable Energy Laboratory (NREL) car manufacturers have agreed to apply the
same warranties for FFVs as those of vehicles being run on neat gasoline, and the maintenance
Page 30
practises of dealers are very similar for these two types of vehicles. It has also been noted by
Ford in literature concerning their Ford Taurus that no special service instructions are needed.
FFVs have been used in private and government fleets for years. The technology is proven,
and the knowledge base about them is strong. Manufacturers stand behind them with
standard warranties equal to those of gasoline vehicles. Dealer maintenance practices for
FFVs are very similar to those followed for gasoline vehicles (NREL, 2003).
In the information given by Ford Motor Company concerning their FFVs there are: no
special service or maintenance issues with the Taurus Flexible Fuel Vehicle (Ford Motor
Company, 2003).
For vehicles using up to 10 % ethanol in gasoline there does not generally seem to be any need
for special service instructions, and for vehicles available on the US market the use of blends
with up to 10 % alcohol has been accepted by the car manufacturers. This benefit has been noted
by many authors, including Launder (2001) from Michigan State University in a study of the
development of ethanol blends in gasoline in the USA from 1970 onwards. As early as 1980
Chrysler, Ford, and General Motors stated that a blend of 10 % ethanol in gasoline would be
covered by their warranties for their cars on the US market. By 1995, the same year that Ford
started production of FFVs, almost all US car manufacturers were recommending the use of 10
% ethanol in gasoline according to Launder.
According to the CRFAs website this warranty is valid for ethanol concentrations up to 10 %
and no engine modification is needed (www.greenfuels.org/ethanolterms.html).The CRFA also
says that a higher engine compression ratio is advisable for ethanol-gasoline blends above 10 %,
but then the warranty may not be valid since increasing the compression ratio may damage the
engine. Many car manufacturers in Europe and Japan sell vehicles in North America, and thus
the assumption of the authors of the present report is that the design of vehicles on the European
market, including Sweden, may allow use of ethanol-gasoline blends with at least 10 % alcohol.
The most serious reported risk associated with blending an alcohol, especially methanol, in
gasoline is the potential formation of water vapour and formic acid during low temperature
combustion (US Department of Energy, 1991). The proposed solutions of the problem are to use
acid neutralizers in lubrication oils, surface treatment of engine parts and more frequent
replacement of lubrication oil or higher quality synthetic oils or a redesign of conventional
engine lubrication oils. In the report by the US DOE the following materials are listed as
subject to degradation due to high concentrations of alcohols since ethanol may not be
compatible with them:
Lubricating oils.
Terne steel (in gas tanks).
Page 31
Hsieh et al. (2002) highlight the further potential problem, which may especially apply to older
cars, that alcohol tends to react with rubber and they advise that modern cars should be designed
to be compatible with the use of alcohol blended gasoline. Many authors of papers dealing with
alcohol fuels claim that most problems associated with the use of alcohol, especially methanol,
as a fuel for vehicles have occurred in older cars.
It was also noted by the Oak Ridge National Laboratory that Du Pont had found that highly
fluorinated fluorohydrocarbons provided the best resistance to either highly aromatic gasoline or
to ethanol. Hammel-Smith et al. (2002) have reported, following material tests extending over
many years, that these or similar materials may be used in modern vehicle systems. In the report
by Hammel-Smith et al. it is also said that few, if any incidents have been reported on 10%
blends associated with cylinder wall wash.
In the investigation by Minnesota State University (referred to above) the use of a 30 % ethanol-
gasoline blend did not cause more than normal wear to the engines (Bonnema et al., 2004).
Joseph Jr., representing the Brazilian Automobile Manufacturers Association, has presented a
study entitled Vehicular Ethanol Fuel, describing experience with the use of ethanol-gasoline
blends both anhydrous (99.3 %w) and hydrous (93.2 %w) in cars in Brazil (Joseph Jr., 1998).
One aspect discussed was the effects of ethanol blended gasoline relative to neat gasoline.
Brazilian experience suggested that the relative compatibility of the blends with plastic materials
was good and their tendency to cause the formation of gum deposits was low. Another
conclusion was that a 22 % ethanol blend is more corrosive towards metals than gasoline. On the
other hand it was felt that gasoline was as corrosive towards copper strips as a 10 % ethanol
gasoline blend. The conclusion of the Joseph Junior presentation is that use of a 10 % ethanol-
gasoline blend had no apparent detrimental effect on vehicle performance.
In its Industry Guidelines, Specifications, and Procedures the FRA states that o-rings and seals
used in meters for neat ethanol should be designed to withstand ethanol. On the other hand it
states that no accelerated wear has been seen in gasoline meters which are used for ethanol-
gasoline blends (Renewable Fuels Association, 2002). However, it also advises that the meters
should be recalibrated 10-14 days after switching to the distribution of ethanol blended
gasoline, as noted in section 2.3.
Page 32
The Guidebook from the US DOE mentioned above provides information on solutions to
problems related (inter alia) to fuel sampling, one of which may be to check the fuel facilities
since their materials may be incompatible with the use of ethanol. There are indications that one
potential problem is excessive wear of the nozzles and hoses etc. in the equipment used (US
Department of Energy, 2001).
For blends with low ethanol contents (max. 10 %) there has been shown to be no difference
between using them and neat gasoline. The engine wear that occurs when using a 10 % ethanol-
gasoline blend was studied by Apache Research Ltd in Australia, who found that there is no
additional or unusual wear to that normally expected, and that there was no increase in wear or
reduction of TBN (total base number) compared to corresponding parameters when using neat
gasoline (Apace Research Ltd, 1998).
There seems to have been an ongoing agreement between the Australian Institute of Petroleum,
the Federal Chamber of Automotive Industries and the Australian Automobile Association that a
10 % ethanol content in gasoline is acceptable, but not more, because higher ethanol contents
will result in loss of drivability, loss of fuel economy and accelerated wear of engine
components and fuel lines. In addition it will have an effect on the warranty provisions for the
vehicle (Australian Automobile Association, 2002).
The ethanol production industry has not agreed to the ethanol content of ethanol-gasoline blends
being limited to 10 %, and claims that 10 to 20 % ethanol in gasoline has been used in vehicle
fleets since 1992. The industry also made an agreement with a university to study the effects of
higher blends of ethanol in gasoline (Department of the Environment and Heritage, 2002).
Several vehicle manufacturers have stated that their vehicles can be run on ethanol-gasoline
blends with a 10 % ethanol content without any problem. For example, the Nissan Motor Co.,
Ltd, which noted approvingly, in a paper released in 2005, that it has been stated in the News
that the Minister the Minister of the Energy Ministry together with the Thai Automotive
Industry Association and major fuel suppliers such as the Petroleum Authority of Thailand
(PTT), Bangchark Petroleum and Shell announced that the use of gasohol in cars is one way to
reduce fuel imports and promote the production of gasohol in Thailand (Nissan, 2005).
In the cited paper Nissan says that two grades of gasoline are used in Thailand, RON 91 and
RON 95, and that MTBE has been added to both of them to date. The essential message from
Nissan is that the cost of MTBE can be saved by replacing MTBE with ethanol without
decreasing the octane number of the fuel when compared with the gasoline used. Furthermore
Nissan states that Nissan gasoline engines that are equipped with the injection system (EGI) can
run on gasohol without any ensuing problem and that No modification is required to use
gasohol, just open the fuel tank lid and fill it up. You can change back to normal gasoline at
anytime, no need to wait until the tank is empty.
In a series of papers Chevron has discussed, inter alia, issues related to oxygenated fuels and
especially ethanol blends in gasoline (Chevron, 2004), in one of which, entitled Oxygenated
Gasoline, it is stated that in modern motor vehicles equipped with engine control systems that
adjust the air-fuel ratio, oxygenated gasoline will perform well. However, in vehicles with
carburettors or systems for fuel injection that do not control the air-fuel ratio, oxygenate in the
fuel may result in a too lean air-fuel mixture. In addition, the volumetric fuel consumption will
increase by 2 to 3 percent, on average, according to Chevron when oxygenated gasoline is used.
As early as 1978 gasoline containing 10 % ethanol was being marketed in Nebraska, USA.
Page 33
In California the use of MTBE and other ethers has been banned since 2004, and other states are
considering similar bans (Chevron, 2002a). Considering all the available information, from other
sources as well as Chevron, it seems likely that ethanol will soon be the only oxygenate that can
be used in the USA to satisfy federal requirements for oxygen in gasoline.
In the complete report (Chevron, 2002a) Chevron discusses the implications of the observation
that Methanol is Not Ethanol in that they have different physico-chemical characteristics.
According to Chevron:
Blends of gasoline with methanol are more corrosive towards metals and cause more rapid
deterioration of elastomers in the fuel system.
Methanol-gasoline blends are not authorized by many manuals for vehicle owners.
The vapour pressure of a methanol-gasoline blend is significantly higher than that of a
corresponding ethanol-gasoline blend.
Methanol is toxic.
Vapour lock is linked to the vapour pressure of the fuel, which is a measure of the front end
volatility of the fuel. Fuels with extremely high vapour pressure may cause drivability and hot
start problems as a result of vapour lock. Since the addition of an alcohol to gasoline will
Front-end volatility: A term applied to the volatility of the lower boiling-point fractions of gasoline.
Page 34
increase the volatility of the fuel, there is a risk that it will also cause vapour lock to occur in hot
weather or at high altitudes.
A paper published by Chevron recommends that fuels should not exceed specified Vapour Lock
Index (VLI) values in order to avoid problems such as vapour lock and hot fuel. The following
formula has been developed for calculating the VLI based on the vapour pressure (in kPa) and
percent evaporated fuel at 70 C: VLI=10(VP)+7(E70). The normal range for the VLI, according
to Chevron, is 800 to 1250, and protection from vapour-linked problems is greatest at the lower
end of the range (Chevron, 2002).
Page 35
Coln et al. (2001) have compared the urban air levels of organic compounds in Sao Paulo,
Brazil, and the Los Angeles basin, USA. In Sao Paulo it is estimated that ethanol or ethanol-
gasoline fuel blends account for approximately 40% of the total fuel used, but the corresponding
level in the Los Angeles basin is much lower, hence the urban air in Sao Paulo is much more
strongly affected by the use of ethanol as motor fuel.. It was found that the ambient air levels of
volatile organic compounds measured in Sao Paulo were substantially higher than those in Los
Angeles. For instance, mean concentrations of mono ring aromatics, volatile aldehydes and
simple alcohols (mainly ethanol) were 2 - 3, 5 - 10 and 10 - 100 times higher, respectively.
Alkanes (C4-C11, n-alkanes) were slightly increased in Sao Paulo. The ambient levels of organic
compounds in the Los Angeles basin used by Coln in the comparisons were based on
measurements made by the US EPA in the year 1997 (Lonneman, 1998). A study regarding
carbonyls in urban air, collected in Rio de Janeiro, Brazil, revealed that of the 61 carbonyl
compounds measured the most abundant were formaldehyde and acetaldehyde, with mean
concentrations of 10.8 + 4.1 and 10.4 + 4.6 ug/m3, respectively (Grosjean et al., 2002).
Maclean and Lave (2000) have investigated air quality trade-offs from automobiles fuelled by
alternative fuels. They state that comparisons can be misleading if the vehicles used are
dissimilar. One of the fuels considered was E85 i.e. 85% ethanol and 15% gasoline. Aakko and
Nylund (2003) investigated an FFV running on E85 and found that when the temperature was
lowered from 23 C to 7 C formaldehyde and acetaldehyde emissions increased (the latter
approximately four-fold; from 4 to 15 mg/km driven).
Exhausts from mobile sources such as motor vehicles are chemically very complex. The
chemical compounds emitted range from gaseous, liquids to particles (solids) and may also be
distributed between particles and the gas phase depending on the physical properties of the
compounds. Particle emissions consist mainly of carbon particles onto which a variety of
compounds are adsorbed. Up to 10-40 % by weight of the particles can be extracted from the
carbon matrix with organic solvents (National Research Council, 1982). Such an extract is often
referred to as the soluble organic fraction (SOF).
Several factors affect the chemical composition of the exhaust emissions, including the fuel, fuel
quality, engine lubricating oil, engine wear, exhaust after treatment, driving conditions and
ambient temperatures. Ambient temperature driving conditions, which of course depend on
seasonal variations, can have a major impact on emissions, especially for Otto engines, both with
and without three-way catalysts (Almn et al., 1997) at low ambient temperatures from +20 C
down to 20 C or lower. Investigations of light duty diesel engines have indicated that they are
relatively unaffected by the ambient temperature compared to investigated Otto engines. This
Page 36
could be due to the differences in the combustion mechanisms of Diesel and Otto engines.
Aakko and Nylund (2003) investigated exhaust emissions from an FFV running on E85 at
temperatures of +20, 0 and 7 C, and concluded that the emissions of CO, HC and total particles
increased as the temperature was lowered while NOx emissions were only slightly increased.
Another important factor affecting the chemical composition of exhaust emissions generated in
the combustion process in an engine is the service and maintenance history of the vehicle/engine.
Extremely high levels of both regulated and unregulated emissions have been detected in exhaust
emissions from gasoline fuelled vehicles that have been poorly serviced and maintained, i.e.
incorrectly functioning engines or so-called high emitters (Sjdin et al., 2000). High emitters
should be mended or eliminated from the car fleet in general.
Armstrong (1995) has evaluated ethanol as an individual compound and states that it is readily
degraded in the environment, so human exposure to it is anticipated to be very low. Furthermore,
abundant information is available on the metabolism of ingested ethanol, which strongly
suggests that environmental exposure to ethanol will have no adverse health impact on humans.
Average ambient levels of ethanol in Porto Alegre, Brazil, where 17% of the vehicles run on
ethanol have been found to be 12 ppb. Animal studies have found that the Lowest Effect Level
(the lowest level to give a detectable response) of ethanol is 45 ppm, which is approximately
4000 times higher. The dose for a person living in Puerto Alegre might be about 0.5 mg ethanol,
which is a negligible dose.
In section 5.3 a selection of exhaust constituents that may have adverse effects on the
environment or health of animals and humans is listed and discussed.
Page 37
5. EMISSIONS
In an investigation by the Swedish government entitled Alternativbrnsleutredningen* (SOU,
1996) it was proposed that a thorough chemical characterization of exhaust emissions (regulated
and unregulated exhaust constituents) of gasoline blended with alcohols should be carried out.
Furthermore, it was suggested that bio assays, such as mutagenicity tests, TCDD-receptor
affinity tests and neurotoxicity tests, should be included in the risk evaluation. The risk criterion
was that the modified fuel alcohol/gasoline blend should not have any greater potential
environmental and health impact than a baseline gasoline fuel. However, an updated set of
compounds and classes of compounds that should be considered is discussed below.
Table 5.1. Sensitivity of the flame ionization detector towards selected compounds relative to
that of the hydrocarbon n-heptane (C7H16), which is set to 1.00 (Diez, 1967).
Hydrocarbons Alcohols Organic acids
Methane 0.97 Methanol 0.23 Formic acid 0.01
Ethane 0.97 Ethanol 0.46 Acetic acid 0.24
Emission measurements must be carried out as part of the type approval of the vehicle according
to relevant legislation, for example the European regulations (current EU Directive 70/220/EEG
with amendments). There is also a need for emission testing to generate emission factors for use
in emission inventories and air quality studies. Characterisation of emissions from vehicles is
also essential for research purposes. Emission tests are commonly carried out according to
standard procedures to enable data generated at different laboratories or during different projects
to be compared. There may also be a need to generate emission factors for local traffic situations,
*
Investigation of alternative fuels.
Page 38
in which case specific driving patterns (driving cycles) may have to be followed when testing the
vehicles, depending on the regulations in force.
Assume, for example, that a vehicle is run on gasoline and has HC emissions of 0.15 mg/km (i.e.
uncombusted fuel hydrocarbons), and that the same vehicle is run on a gasoline blend and the
measured HC emissions are also 0.15 mg/km (consisting of both uncombusted fuel
hydrocarbons and uncombusted ethanol) in comparative tests.
The standard analytical method used for determining hydrocarbons in motor vehicle exhaust is
to use a Flame Ionization Detector (FID), which can be regarded as a carbon counter. The
relative sensitivity of an FID for each of the hydrocarbons methane, ethane, propane and pentane
is approximately 1 (Dietz, 1967). The corresponding value for ethanol is 0.46. Using these
relative sensitivity factors, and assuming that only hydrocarbons are present in the exhaust the
HC emission factor will be 0.15 mg/km, but if only (100%) ethanol is present in the exhaust the
ethanol emission factor from the vehicle is in reality 0.33 mg/km However, if only ethanol is
present in the exhaust the HC emission measured is still 0.15 mg/km but should be practically
zero as no hydrocarbon is present in the exhaust.
The above discussion highlights the need to distinguish between HC and alcohol contents in
vehicle exhaust, especially when alcohol gasoline blends are used as fuel. An updated method
needs to be developed for HC measurements from a legal perspective. This also means that a
standardized method needs to be developed.
*
Other compounds, such as alcohols/aldehydes, also contribute to the signal from the FID.
Page 39
The US EPA has estimated that more than 20,000 individual chemical compounds are emitted
from diesel fuelled vehicles (US EPA, 1990), approximately 500 of which have been
positively/tentatively identified in the scientific literature. This means that more than 97 % of the
estimated compounds emitted from diesel vehicles are unknown (as are, therefore, their health
effects). However, a selection of unregulated exhaust constituents is considered below that are
considered to be important and to have potential health effects on animals and humans, and thus
should be monitored and reduced in exhaust emissions from automotive sources. Greenhouse
gases and smog-forming compounds are also considered. The list of such compounds is expected
to be updated and modified according to future findings.
Alcohols
When alcohols are used as blending components in gasoline, uncombusted alcohols from the fuel
are emitted in the exhaust in various amounts. These uncombusted alcohols in the exhaust
emissions should be measured since there is a risk that the HC signal from the FID will be
overestimated, leading to estimates of HC emissions from alcohol blended gasoline fuels being
higher than they really are (see Regulated exhaust emissions regarding HC section 5.1).
Aldehydes
Aldehydes are known to be irritants, they may induce allergic reactions and formaldehyde is
considered to be a carcinogen (CARB, 1989). Aldehydes of special interest are formaldehyde
and acetaldehyde. Using ethanol as a blending component in gasoline results in increased
emissions of both formaldehyde and acetaldehyde (Egebck and Bertilsson, 1983). According to
Aakko and Nylund (Aakko and Nylund, 2003) significant increases in aldehyde emissions occur
at relatively low ambient temperatures (-7 C). The effect is expected to be even greater at lower
ambient temperatures than -7 C, especially for ethanol-gasoline blended fuels.
Alkenes
Alkyl nitrites
Methyl and ethyl nitrite are mutagenic (Trnqvist et al., 1983; Wild et al., 1983), hence they are
potential carcinogens. Alkyl nitrites are formed from uncombusted alcohol reacting with
nitrogen oxides (NOx) in exhaust plumes. From vehicles fuelled with methanol/gasoline blends
and methanol/diesel blends methyl nitrite is formed (Johnsson and Bertilsson, 1982). Both
methyl and ethyl nitrite are formed in exhaust plumes from vehicles fuelled with ethanol-
gasoline blends (Egebck and Bertilsson, 1983).
Page 40
Monoaromatics
Monoaromatic compounds emitted from vehicles include benzene, toluene, ethyl benzene and
xylene, which are often collectively called BTEX. According to the California Air Resources
Board (CARB, 1998) benzene is a known human carcinogen and may cause leukaemia through
occupational exposure (Trnqvist and Ehrenberg, 1994). A source of BTEX emissions from
vehicles is the BTEX content of the fuel used, emitted in the uncombusted fuel constituents.
However, BTEX can also be pyrosynthesised in the combustion process from the fuel or
lubricating oil constituents.
Particulate emissions
Particlulate emissions are measured on a weight basis i.e. through gravimetric determination
using a dilution tunnel technique at a diluted exhaust temperature below 52 C. Exposure to
diesel exhausts clearly induces lung tumours in rats. These neoplasms may be caused by the
particle fraction of the exhaust (International Agency for Research on Cancer, IARC, 1983;
1989; Camner et al., 1997). Due to the findings that TiO2 and carbon black particles, with no
genotoxic compounds adsorbed on them (i.e. clean particles) can also give rise to lung cancers
in rats (Pott, 1991; Pott and Heinrich, 1990; Heinrich et al., 1993) interest in particles per se has
increased. Important particle parameters in general are their size, number, surface area and
chemical composition. Studies by Heinrich et al. (1995) indicate that particle size is a very
important parameter. The Institute of Environmental Medicine (Karolinska Institute, Sweden)
has published a report which concludes that it is not currently possible to tell if a non-specific
particle factor or a direct genotoxic effect of material adsorbed on the particles is responsible for
causing lung cancer (Camner et al., 1997). Therefore, the particles emitted should be chemically
analysed with respect to both their size and numbers.
Peroxyacetyl nitrate
In a study by Gaffney et al. (1997) field measurements taken in Albuquerque, New Mexico, were
used to compare atmospheric levels of peroxyacetyl nitrate (PAN) associated with the use of
different fuels. Levels were measured before and after introduction of a 10% ethanol gasoline
fuel blend (>99%) and the cited authors detected increased levels of PAN and aldehydes during
the wintertime. A study which is more valid during summertime conditions in Porto Alegre,
Brazil, (Grosjean et al., 2002) indicates that aromatics and alkenes have a major role, and
acetaldehyde and ethanol minor roles, as precursors to PAN in urban air in contrast to a report
prepared by the Orbital Engine Company (2002) for Environment Australia, which states that
acetaldehyde is particularly important since it reacts with NOx, forming PAN in the atmospheric
photochemical system. In the report it is stated that acetaldehyde emissions increase as the
ethanol content in the blended gasoline fuel increases. This conclusion is supported by the results
of several other investigations.
Environmental Protection Agency (SEPA) for air monitoring programs in 1999, however
published in 2002 (Bostrm et al., 2002). Recommended guideline value concentrations in
ambient air are set to 0.1 ng/m3 for benzo(a)pyrene (B(a)P) and for fluoranthene 2 ng/m3 in
Sweden (Bostrm et al., 2002). So far B(a)P is the only PAH whose concentration in ambient air
will be regulated by European Commission and the limit value will most probably be set to 1
ng/m3. However, at present the European Parliament has not made a final decision about the
limit value according to Kyrklund (Kyrklund, 2004).
Nitrogen dioxide
Nitrogen oxides (NO and NO2) are formed by the oxidation of nitrogen from the air in the
combustion process. An important parameter for the formation of nitrogen oxides is the
combustion temperature i.e. increased combustion temperature results in increased nitrogen
oxide emissions. Nitrogen dioxide (NO2) is a compound that plays a role in respiratory diseases,
especially those of asthmatics and children (Nitschke, 1999). It should be noted that nitrogen
oxides (NOx) are regulated pollutants that are determined jointly, as the sum of NO and NO2
contents rather than as individual components.
Greenhouse gases
Greenhouse gases of most interest are carbon dioxide (CO2), methane (CH4) and nitrous oxide
(N2O), all of which interact with the radiative energy fluxes in the atmosphere, thereby
increasing the average temperature of the earth. In Table 5.2 the Global Warming Potential
(GWP) for selected greenhouse gases is shown. The GWP indicates the relative contribution of a
molecule of a greenhouse gas compared to a molecule of carbon dioxide, for which GWP is set
to one. According to Rodhe (2005), the background concentration in the atmosphere of carbon
dioxide is 375 ppmv (yearly variation, +/- 2 ppmv), while the background concentrations of
methane and dinitrogen oxide (nitrous oxide) are 1.75 ppmv and 0.315 ppmv, respectively. By
multiplying the concentration of a greenhouse gas and its respective GWP, its relative
contribution to the global warming effect can be estimated. Clearly, the most important
greenhouse gas is carbon dioxide, due to its relative abundance in the atmosphere.
Table 5.2. Global Warming Potential (GWP) and the relative contribution to the global warming
effect (RCGWE) of selected compounds.
Compound GWP Concentratio GWP x ppm % RCGWE
n ppmv
Carbon dioxide, CO2 1 375 375 74
Methane, CH4 23 1.75 40.3 7.9
Dinitrogen oxide, N2O 296 0.315 93.2 18
Quinones
A recent publication by Xia (Xia et al., 2004) shows that a quinone-enriched polar fraction from
a diesel particulate extract was more potent than PAH with respect to toxic effects in RAW
264.7 cells macrophage-like cells derived from tumours induced in male BALB/c mice by the
Abelson murine leukemia virus. Quinones are a group of organic compounds that consist of
diketones (carbonyls) which contain oxygen and are present in both diesel particles (Schuetzle et
al., 1981) and ambient air (Cho et al., 2004). Aromatic quinones have previously been identified
and measured in gasoline particulate exhaust extracts (Schuetzle et al., 1981; Alsberg et al.,
1985). The origin of the aromatic quinones is not fully understood if it is related to the gasoline
fuel as uncombusted quinones (initially in the fuel) or if they are formed in the combustion
Page 42
process.. It is well known that fuel properties affect the chemical composition of the exhaust
emitted from engines. For instance, adding ethanol to gasoline increases the emission of
aldehydes (carbonyls). It is possible that increasing the ethanol content of gasoline-ethanol
blends causes similar increases in quinone emissions, but there are no empirical data related to
this issue as yet.
In Table 5.3 mean emission factors are shown from three TWC-equipped cars running on 3, 6
and 10 vol. % ethanol-gasoline blends (Schifter et al., 2001). It is difficult to interpret or to draw
firm conclusions from the data in the table, as they are mean values from three cars. However,
NOx, benzene and acetaldehyde emissions increased with increasing ethanol contents, while CO
and formaldehyde emissions and the ozone formation factor (g O3/g Non Methane Organic
Gases, NMOG) decreased.
Table 5.3. Mean emission factors from three TWC-equipped cars running on 3, 6 and 10 vol. %
ethanol gasoline blends (Schifter et al., 2001).
Emission 3% 6 % Ethanol 10 % Ethanol Ethanol
Ethanol effect
CO, g/km 3.11 2.89 2.97 -
HC, g/km 0.23 0.21 0.22 0
NOx, g/km 0.42 0.42 0.48 +
g O3/g NMOG 3.12 3.09 3.08 -
Benzene, mg/km 7.22 7.38 8.11 +
Butadiene, mg/km 0.83 0.77 0.83 +/-
Formaldehyde, mg/km 1.32 0.78 1.01 -
Acetaldehyde, mg/km 1.12 1.25 1.62 +
In Table 5.4 the mean fuel economy and mean evaporative emissions from the three TWC-
equipped cars running on 3, 6 and 10 % ethanol gasoline blends are shown.
Table 5.4. Mean fuel economy and mean evaporative (Evap) emissions from three TWC-
equipped cars running on 3, 6 and 10 % ethanol-gasoline blends (Schifter et al., 2001).
3 vol % 6 vol % 10 vol % Ethanol
Ethanol Ethanol Ethanol effect
Fuel consumption l/10 km 0.87 0.85 0.87 +/-
Evap. diurnal, g/test 4.14 3.77 4.60 +/-
Evap. hot-soak, g/test 0.46 0.44 0.68 +
Page 43
Regarding emissions of benzene, toluene, ethyl benzene and xylene (BTEX) a general
conclusion is that as the ethanol content of the blended fuel increases the BTEX emissions are
reduced (Orbital Engine Company, 2002) by the dilution of the base gasoline.
In a study carried out in Australia (Orbital Engine Company, 2004) the unregulated emissions
such as BTEX and aldehydes were measured. The emission measurements were made using
fuels comprising of 20% ethanol (E20) in gasoline and neat gasoline. Furthermore, the vehicle
emissions were determined at both 6400 and 80000 km odometer readings. In Figure 5.1 shows
average benzene emissions from five tested cars. The Holden car decreases its benzene
emissions with increased driving distance. The benzene emission is lower when the E20 fuel is
used as fuel and the emission is reduced at increased driving distance. Comparative benzene
emissions from the Hyundai car is that benzene emissions increase as the driving distance
increases which is valid for both fuels tested. Furthermore, the Subaru car has larger benzene
emissions from the E20 fuel compared to neat gasoline. From the Figure 5.1 it can be concluded
that there is a relatively large variation in benzene emissions due to fuel and accumulated driving
distance and vehicle tested. Because of this, it is difficult to make firm conclusions with respect
to benzene emissions vehicle dependency.
6400 km
7
80 000 km
6
5
4
3
2
1
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Figure 5.2 shows average acetaldehyde emissions from all vehicles tested. From the figure it can
be seen that increased accumulate driving distance results in an increased acetaldehyde emission,
however the Toyota car is excluded. This seems to be fuel independent. From the Figure 5.1 it
can be concluded that the there is a relatively large variation in acetaldehyde emissions due to
fuel and accumulated driving distance and vehicle tested. Because of this, it is difficult to make
firm conclusions with respect to formaldehyde emissions vehicle dependency.
Page 44
1,5
0,5
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
HOLDEN FORD TOYOTA HYUNDAI SUBARU
Figure 5.2. Acetaldehyde emissions (mg/km) gasoline versus E20.
Figure 5.3 shows average formaldehyde emissions from all vehicles tested. All of the tested
vehicles had increased formaldehyde emissions with respect to increased accumulative driving
distance. The relative emissions of formaldehyde are vehicle and fuel dependant.
0,9
0,8 80 000 km
0,7
0,6
0,5
0,4
0,3
0,2
0,1
0 0
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Summing up, from Figures 5.1 to 5.3 due to accumulative driving distance the function of the
catalysts is deteriorating which results in increased emissions of benzene, acetaldehyde and
formaldehyde emissions. The gasoline fuel used contains up to 500 ppm sulphur which is
extremely large compared to gasoline available on the Swedish market which is expected to have
an impact on the exhaust emissions.
Page 45
The engines power output depends on the energy content per unit volume fuel injected into its
combustion chamber. The reduction in energy content per unit volume fuel caused by adding
five different percentages of ethanol to a certain blend of gasoline is shown in Figure 6.1.
30
25
20 % EtOH in gasoline
% change of fuel energy
15
Fuel energy content [MJ/l
10
-5
-10
Figure 6.1. The effect of blending ethanol with Swedish Class 1 (MK1) gasoline on the fuels
energy content.
When adding ethanol up to 25 % the energy change may be sufficient for the driver to clearly
detect a change in the engines power* during acceleration of the vehicle. However, vehicles
produced today are designed to adjust the air-fuel ratio in order to compensate for changes in the
fuel. It is well known that vehicles are equipped with systems for adaptive learning or adaptive
memory, and according to information obtained from Web FAQs, Modern adaptive learning
engine management systems control the combustion stoichiometry by monitoring various
ambient and engine parameters, including exhaust gas recirculation rates, the air flow sensor,
and exhaust oxygen sensor outputs (Hamilton, 2004).
*
As the energy content (MJ/l) of ethanol gasoline blends is lower than that of neat gasoline the
fuel consumption will most likely increase for the average driver.
Page 46
definitive percentage for the increase since it depends on a number of factors, of which the
design of the fuel system of the vehicle and engine, the percentage of ethanol in the fuel and the
driving pattern of the vehicle appear to be most important.
An issue to consider is whether the owner of the vehicle will recognize the increase in fuel
consumption that may occur when ethanol is added to gasoline. In a report prepared by the
Orbital Engine Company for Environment Australia concerning an investigation into the use of
20 % ethanol in gasoline it is stated that the increase in fuel consumption ranges from 2.5 % to
7 % depending on the cycle and the vehicle. In the same report Orbital also states that
Increases in fuel consumption of 5% or more are considered to be recognisable to the average
driver (Orbital Engine Company, 2003).
Results presented by Apache Research Ltd. from an investigation of the use of a blend of 10%
ethanol in gasoline show that fuel consumption increased by 2.4 % when driving according to
the US EPA City cycle and by 2.7 % when driving according to the Highway cycle (Apache
Research Ltd, 1998). According to the statements by Orbital this increase in fuel consumption
should not be detectable by an average driver.
In a paper from the Transportation Office of Energy Efficiency in Canada it is said that adding
10% ethanol to gasoline (E10) will result in a blend with an energy content equivalent to 97 % of
the energy content of the base gasoline. Since this decrease in energy content will be partly
compensated by the improved combustion efficiency of the ethanol blended fuel, the overall
increase in fuel consumption when using E10 will be only 2 %. In comparison, increasing the
speed of the vehicle from 100 km/h to 120 km/h will result in a 20 % increase in fuel
consumption (Transport Office of Energy Efficiency in Canada, 2004).
Page 47
A key issue today is whether newly manufactured gasoline-fuelled vehicles equipped with an
efficient emission control system emit more or less harmful substances than vehicles fuelled with
any of the fuels mentioned in the paper from the EPA. It can also be questioned whether any
automotive fuel used today can be defined as a clean fuel since all types of fuels generate
pollution when burned in a combustion engine.
Concerning the use of a blend with a certain content of ethanol in gasoline it is quite clear, as
already mentioned in this report, that there is a risk that emissions, especially of acetaldehyde,
will increase, but those of benzene, toluene, ethylbenzene and xylene (BTEX) may decrease
compared with the use of neat gasoline in the same vehicle. Whether (and if so to what degree)
these changes will be better or worse overall in terms of the air quality and health has been
considered in various studies, especially in the USA/California, and in this report. The most
important issue to clarify is whether, and if so to what extent, the regulated emissions will
change compared with the use of neat gasoline when switching to the use of a blend of
ethanol/gasoline. Hammel-Smith and his colleagues at NREL have indicated that a key aspect
concerning vehicle performance and emissions when using a blend of ethanol/gasoline of up to
17 to 24 % is that fuel control is able to compensate for a high oxygen content in the fuel
(Hammel-Smith et al., 2002).
One of the conclusions to be drawn is that it is generally not certain that a higher blend of
ethanol can be used in all vehicles currently available. Potential barriers in this respect are the
age of the vehicles, their technological standards and their control systems. Electronically
controlled fuel injection systems have long replaced carburettors and they are more advanced
today in that new functions have been incorporated. The use of electronic mapped digital
systems makes it possible to control ignition, fuel injection, emissions and automatic
transmission in addition to other engine variables. According to the NREL these systems have
been important in the evolution of the use of alcohols as automotive fuels (Hammel-Smith et al.,
2002).
When considering the performance and emissions from motor vehicles, especially vehicles using
blends of ethanol in gasoline, the technological status of the vehicle should be considered. It is
accepted that adding an alcohol, i.e. ethanol, in gasoline will enhance the octane rate of the fuel.
On the other hand it has been shown in section 6 of this report that blending ethanol in gasoline
will decrease the energy content of the fuel, resulting in an increase of the volumetric fuel
consumption. In vehicles equipped with advanced engine control systems the ability to exploit
the higher octane rating compensates, to a certain degree at least, for the inevitable drop in the
energy content of the fuel.
When considering the effect of blending ethanol with gasoline on emissions a factor that must be
taken into account is that motor vehicles are sensitive to changes that affect the air/fuel ratio.
This is especially true for vehicles equipped with older types of fuel systems, such as
Page 48
carburettors or open loop fuel injection systems. Newer models of vehicles equipped with
closed-loop fuel injection systems, especially those with advanced closed-loop fuel injection
systems with adaptive learning functions are less sensitive in this respect.
When using alcohol blended gasoline it has been observed that NOx emissions may increase
compared to the use of neat gasoline due to the leaning effect of the alcohol. This effect does not
generally occur when using alcohol blended gasoline in newer models of vehicles.
During the survey of reports presenting data and experience related to the use of ethanol blended
gasoline some data from emission and fuel consumption measurements have been found. In the
following sections these data and experiences are briefly described. When reading the tables and
studying the figures it should be noted that presented data on HC emissions include components
other than hydrocarbons, as discussed in sections 5.2 and 5.3.
Two types of gasoline (AEN* Summer ULP and AEN Perth ULP; designated ULP and PULP,
respectively) were used as base gasolines for blending with ethanol. The difference between the
two types of gasoline, according to Orbital, is as follows: In summary, there are some small but
quantifiable differences to be taken into account when comparing trends over the accumulated
mileage and across some of the vehicle fleet using the ULP stock. The base fuel differences are
however relatively insignificant with respect to distillation and constituency when compared to
the change introduced by the ethanol blending (Orbital Engine Company, 2004).
The investigation presented in the cited report is an extension of earlier tests (Orbital Engine
Company, 2003). The aim of the later study, named 2B, included an assessment of durability
over accumulated mileage and the following measurements and analyses at specific
breakpoints (Orbital Engine Company, 2004):
*
Australian Energy News (AEN)
Page 49
Since the effects of using ethanol-gasoline blends on engine wear have already been considered
in this report discussion here is mainly focused on the emissions. The emission testing included
both regulated and non regulated emissions and the main results were as follows. The emission
testing included measurements of both regulated and non regulated constituents and the tests
were carried out in accordance with the Australian test procedure ADR37/01 for monitoring
emissions from light duty vehicles, which includes the US 75 test cycle. The main results were
as follows
As can be seen in Table 7.1 there was only a small difference in emission levels when comparing
the two fuels at 6 400 km, but the emission performance of E20 was considerably worse than
that of the neat gasoline at 80 000 km.
To elucidate the reasons for the observed deterioration in emissions with accumulated mileage
associated with use of the blends, the cited authors studied factors that could contribute to their
impact on emissions.
Since one of the factors could be the catalyst system, detailed analyses of the individual vehicles
were carried out, especially with respect to loss of catalyst efficiency. The catalyst efficiency
was studied for each of the three phases of the test cycle, of which Phase 1 (the first 505 sec) is
the cold start phase according to the test procedure the vehicle is initially equilibrated to a
room temperature of around 20 to 25 C. The three phases can be seen in Figure 7.1.
Figure 7.1. The three phases of the test cycle followed during the emission test.
Figures 7.2, 7.3 and 7.4 show the difference in the efficiency of the catalyst found when using
the 20 % ethanol blended fuel compared with the use of neat gasoline in each of the three phases.
Page 50
80
%
60
40
20
0
E20
E20
E20
E20
E20
Gasoline
Gasoline
Gasoline
Gasoline
Gasoline
Holden Ford Toyota Hyundai Subaru
Figure 7.2. Phase 1 NOx catalyst efficiency when using E20 compared with neat gasoline after
6 400 and 80 000km driving
100
80
%
60 6400 km
80 000 km
40
20
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
80
%
60
40
20
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Holden Ford Toyota Hyundai Subaru
Figure 7.4. Phase 1 NOx catalyst efficiency when using E20 compared with neat gasoline after
6 400 and 80 000km driving.
There is a considerable difference between the pairs of vehicles. For the Toyota vehicles
there is very little difference in catalyst efficiency between use of E20 and use of neat
gasoline.
There is a difference in catalyst efficiency when comparing the three phases of the test
cycle. As expected, the efficiency was considerably lower during Phase 1.
The catalyst efficiency was somewhat higher, at least for some vehicles, during Phase 2
than during Phase 3.
The most important observation to note is that the deterioration of the catalyst performance
was considerably worse when using E20 than when using neat gasoline.
Another factor to consider when studying data from emission tests is the impact of sulphur in the
fuel. Studies carried out for the American Lung Association of Minnesota have shown that
sulphur in the fuel has a considerable impact on both regulated and unregulated emissions. There
is reason to expect the deterioration of the vehicles tested after 80 000 km driving to be linked, to
some degree, to sulphur in the fuel used in the tests carried out by Orbital. According to Orbital
two grades of gasoline were used in the ethanol blends: unleaded gasoline (ULP) and Perth
unleaded gasoline (PULP). The Australian gasoline specifications at the time of the tests show
that the upper limits for the sulphur contents in ULP and PULP were 500 ppm and 150 ppm,
respectively. A table taken from the report (Table 7.2) suggests that the ULP grade was used in
the neat gasoline tests, but elsewhere it suggests that PULP was also used. Furthermore, it is not
clear whether the same grade of gasoline was used for the ethanol-gasoline blends as that used
for the neat gasoline tests, although it can be assumed that even a sulphur content of 150 ppm in
the base fuel would have had a considerable impact on the emissions.
Page 52
Table 7.2. Emission data from tests with neat gasoline and 20 % ethanol-gasoline blends.
Emissio 6 400 km 40 000 km 80 000 km % Difference
n ULP E20 ULP E20 ULP E20 ULP E20
g/km
Holden THC 0.120 0.081 0.162 0.155 0.147 0.119 22.5 46.9
CO 0.711 0.728 0.800 0.865 0.845 0.907 18.8 24.6
NOx 0.100 0.083 0.106 0.126 0.134 0.208 34.0 150.6
CO2 256.7 267.6 259.3 265.3 256.8 260.7 0.0 -2.6
*
Ford THC 0.126 0.120 0.112 0.142 0.191 0.218 51.6 81.7
CO 2.090 1.710 1.871 2.075 4.731 3.798 126.4 122.1
NOx 0.125 0.265 0.171 0.174 0.457 0.481 265.6 81.5
CO2 255.7 258.8 255.0 253.5 252.4 250.3 -1.3 -3.3
Toyota THC 0.025 0.031 0.030 0.024 0.045 0.027 80.0 -12.9
CO 0.551 0.457 0.873 0.671 1.063 0.895 92.9 95.8
NOx 0.034 0.036 0.039 0.056 0.070 0.047 105.9 30.5
CO2 247.5 248.4 235.1 293.8 235.1 236.6 -5.0 -4.7
Hyundai THC 0.041 0.046 0.049 0.065 0.051 0.112 24.4 143.5
CO 0.304 0.345 0.472 1.155 0.45 1.821 48.0 427.9
NOx 0.15 0.18 0.105 0.488 0.132 0.637 -12.0 253.9
CO2 168.8 172.0 166.4 163.3 166.5 165.8 -1.40 -3.60
Subaru THC 0.050 0.041 0.079 0.084 N/A 0.104 58.0 104.9
CO 0.388 0.288 0.697 0.936 N/A 1.023 79.6 225.0
NOx 0.059 0.043 0.059 0.048 N/A 0.071 0.0 11.6
CO2 255.6 254.8 250.0 254.3 N/A 250.6 -2.2 -0.2
*
AENFO02 results are based on tests conducted at 8 900 km rather than 6 400 km.
AENSU05 results are based on tests conducted at 7 600 km rather than 6 400 km.
In the available edition of the report it is noted that the report has not undergone detailed
technical review by the Environmental Technology Advancement Directorate Canada.
Page 53
Table 7.4. Selected data for the fuels used in the emission tests.
Fuel properties SGC1 10 % 15 % Ethanol 20 % Ethanol
Ethanol
% C by mass 86.7 82.8 81.0 79.3
% H by mass 13.6 13.5 13.5 13.5
% O by mass - 3.7 5.5 7.3
Density [g/ml] 0.7453 0.7497 0.7519 0.7541
1
Summer Grade Commercial baseline tests of fuel.
In this section the results from tests of regulated emissions are presented in the form of figures.
A more complete set of emission data presented in the report by Augin and Graham is given in
the Appendix 1.
The emission tests were carried out in accordance with the US Federal Test Procedure (FTP 75)
and two repeats of each cycle were performed on each vehicle in order to provide a minimal
measure of the repeatability, according to the authors of the report (Augin and Graham, 2004).
The results of the emission tests are shown as means for the five vehicles in Figures 7.5 to 7.9
and Tables 7.5 to 7.7
0,60
0,50
0,40
g/km
0,30
0,20
0,10
0,00
0 % EtOH 10 % EtOH 15 % EtOH 20 % EtOH
Figure 7.5. Mean emissions of carbon monoxide from the five test vehicles according to the
US EPA FTP 75 test procedure
Page 54
0,050
0,045
0,040
0,035
0,030
g/km)
0,025
0,020
0,015
0,010
0,005
0,000
0 % EtOH 10 % EtOH 15 % EtOH 20 % EtOH
Figure 7.6. Mean emissions of hydrocarbons (HC) from the five test vehicles according to the
US EPA FTP 75 test procedure.
0,080
0,070
0,060
0,050
g/km
0,040
0,030
0,020
0,010
0,000
0 % EtOH 10 % EtOH 15 % EtOH 20 % EtOH
Figure 7.7. Mean emissions of NOx from the five test vehicles according to the US EPA FTP
75 test procedure.
125,0
100,0
75,0
50,0
25,0
0,0
0 % EtOH 10 % EtOH 15 % EtOH 20 % EtOH
Figure 7.8. Mean emissions of carbon dioxide from the five test vehicles according to the US
EPA FTP 75 test procedure.
Page 55
1,00
0,90
0,80
0,70
l/10 km
0,60
0,50
0,40
0,30
0,20
0,10
0,00
0 % EtOH 10 % EtOH 15 % EtOH 20 % EtOH
Figure 7.9. Mean fuel consumption of the five vehicles according to the US EPA FTP 75 test
procedure.
Table 7.5. Emissions of aldehydes when using neat gasoline and three different blends of
ethanol.
Formaldehyde Acetaldehyde
Vehicle Honda Insight Grand Am Honda Insight Grand Am
US FTP cycle mg/km mg/km mg/km mg/km
0 % Ethanol 0.019 0.603 0.143 0.298
10 % Ethanol 0.087 0.615 0.273 0.653
15 % Ethanol 0.491 0.516 0.373 0.926
20 % Ethanol 0.665 0.578 0.385 0.814
Table 7.6. Emissions of aldehydes when using neat gasoline and three different blends of
ethanol.
Formaldehyde Acetaldehyde
Vehicle Toyota Echo Honda Civic Toyota Echo Honda Civic
US FTP cycle mg/km mg/km mg/km mg/km
0 % Ethanol 0.534 0.311 0.261 0.099
10 % Ethanol 0.360 0.870 0.379 0.721
15 % Ethanol 0.690 0.416 1.044 0.435
20 % Ethanol 0.671 0.572 1.311 0.597
Table 7.7. Emissions of specific hydrocarbons when using neat gasoline and three different
blends of ethanol.BDL means below the detection limit for the instrument used
Methane Ethylene Acetylene Ethylene
Vehicle Toyota Honda Toyota Honda Toyota Honda Toyota Honda
Echo Civic Echo Civic Echo Civic Echo Civic
US FTP cycle mg/km mg/km mg/km mg/km mg/km mg/km mg/km mg/km
0 % Ethanol 4.848 1.952 4.332 2.032 0.615 BDL* 1.311 0.640
10 % Ethanol 4.338 1.417 4.406 1.554 0.622 BDL 1.212 0.261
15 % Ethanol 4.406 1.212 4.344 1.423 0.640 BDL 1.293 0.553
20 % Ethanol 4.829 1.846 4.375 1.541 0.472 BDL 1.181 0.423
*
BDL, Below Detection Limit.
Page 56
The following conclusions can be drawn from the comparisons of the use of ethanol blended fuel
and neat gasoline:
There are considerable variations in the emission levels between different blends of
ethanol in gasoline and also between the different vehicles, especially when comparing
vehicles with relatively large and relatively small engines.
As expected, CO emissions decrease when replacing neat gasoline with ethanol blended
fuel.
HC emissions amount to less than 0.1 g/km and there is no clear indication whether
blending ethanol in gasoline increases or decreases them.
For all vehicles except the Toyota Echo NOx emissions increase when using a blend of
ethanol in gasoline. For the Toyota vehicle they decrease.
According to the presented data it is not clear whether the tailpipe emissions of CO2
increase or decrease when ethanol is blended with gasoline.
As expected fuel consumption increases when using a blend of ethanol in gasoline.
The presented data provide clear indications that the emissions of formaldehyde increase
when using ethanol blended gasoline.
As expected, the acetaldehyde emissions increase when using ethanol blended gasoline.
The increase found in this investigation is also quite dramatic compared to corresponding
increases found in other investigations discussed in this report.
The emissions of the measured specific hydrocarbons may either decrease or increase when
using ethanol blended gasoline.
Table 7.8. Vehicles selected for emission testing and measurement of fuel economy.
Vehicle Model Engine size Emission Mileage
identifier (litre) Standards km
1&6 Toyota Yaris 1.0 Euro III 22 000
2 Vauxhall Omega 2.2 Euro III 19 000
3 Fiat Punto 1.2 Euro II 51 000
4 VW Golf 1.6 Euro III/IV 21 000
5 Rover 416 1.6 Euro II 117 000
The fuels used for the tests were neat gasoline and a blend of 10% ethanol in gasoline. An
interesting point to note is that the RVP of the neat gasoline and the blended fuel was 60 kPa
66.5 kPa, respectively.
*
AEA Technology is a British technology company.
Page 57
The emission tests were conducted according to the current European test cycle, ECE/EUDC,
and six of the test cycles designed by the Warren Spring Laboratory (WSL), as listed in Table
7.9.
The ECE/EUDC tests were carried out following the defined test procedures. The WSL-tests
(results from the WSL tests are not presented here) were carried out following the standard
procedure for the test laboratory.
It should be noted that only results from ECE/EUDC tests are presented here. In this section the
results from tests of regulated emissions are presented in Figures 7.10 to 7.14. A more complete
set of emission data presented in the report by Augin and Graham is given in Appendix 1.
4
3,5
3
2,5
Cold ECE gasoline
2
1,5 Cold ECE E10
1
0,5
0
CO THC NOX
Figure 7.10. Arithmetic mean of five vehicles tested according to the Cold start ECE emission
test cycle.
0,35
0,30
0,25
0,20 Cold EUDC gasoline
0,15 Cold EUDC E10
0,10
0,05
0,00
CO THC NOX
Figure 7.11. Arithmetic mean of emissions from five vehicles tested according to the EUDC
Cold start emission test cycle.
Page 58
PARTICULATE EMISSIONS
mg/km
10
8
6
4
2 n.d.
0
Cold ECE Cold ECE E10 Cold EUDC Cold EUDC E10
gasoline gasoline
Figure 7.12. Arithmetic mean of particulate emissions in tests according to ECE and EUDC
Cold start emission cycles.
FUEL CONSUMPTION
g/km
100
80
60
40
20
0
Cold ECE Cold ECE E10 Cold EUDC Cold EUDC E10
gasoline gasoline
Figure 7.14. Arithmetic mean of fuel consumption in tests according to ECE and EUDC Cold
start emission cycles.
When comparing the data from the measurements of fuel consumption and emissions of
regulated consitutents, PM and CO obtained using a blend of 10 % ethanol in gasoline with those
obtained using neat gasoline the following conclusions can be drawn
The unregulated emissions are increased for a few components while the emissions are
decreased in most cases. There are significant differences between vehicles, so for some vehicles
the emissions are increased while for others they are decreased. As expected, the emission of
acetaldehyde is increased in most cases. In the following table the numbers of increases (+) and
decreases (-) are presented. Unfortunately, the unregulated emissions were only measured from
three vehicles (plus a repeated measurement on the Toyota Yaris) and there are only four
comparative measurements for benzene and nitrous oxide. As can be seen in Table 7.10,
ammonia and PAH emissions were only measured from three vehicles.
Table. 7.10. Numbers of increases (+) and decreases (-) of unregulated emissions found in the
tests using E10 compared to the tests using neat gasoline.
Since the temperature in Sweden is rather low in the winter - occasionally down to minus 20 C
in the Stockholm area and minus 30 C to minus 40 C in the north of Sweden and the mean
temperature over the year is around plus 7C in the Stockholm area (around 0 C in the north of
Sweden), the influence of the outdoor temperature on the emissions generated when using
blended fuel and neat gasoline has been studied in a number of projects. It is well known that the
use of an alcohol as a fuel in motor vehicles will increase the cold start emissions if the vehicle is
not specially designed, i.e. equipped with an engine heater to improve engine starts. In one
investigation (Egebck et al., 1984) five well-maintained and carefully-checked cars without
catalytic converters and one equipped with a three-way catalyst system were used. The vehicles
were tested with unchanged fuel-air ratio settings (which were set for neat gasoline) with three
different fuels. The fuels used were neat gasoline, 5% ethanol mixed with unleaded gasoline
(E5), and 15% methanol mixed with a refinery-produced gasoline. All fuels were tested at 22 C
and the neat gasoline and ethanol mixture were also tested at +5 and -7C.
Since nearly all light duty vehicles with spark ignition engines in Sweden today are equipped
with a three-way catalyst the results of the emission testing at different temperatures are of
Page 60
interest for this report. A comparison of the emissions from the catalyst-equipped vehicle
adapted for the use of alcohol in neat lead-free gasoline can be seen in Table 7.11.
Table 7.11. Emissions of regulated and non regulated emissions generated when using neat
gasoline and E5 at different ambient temperatures. US-73 test cycle.
As can be seen from the figures in the above table the ambient temperature has a clear impact on
the emissions, especially the particulate emissions and (thus) PAHs. Starting the engine at low
temperatures is known to have a stronger effect on the emissions when using fuels with high
contents of an alcohol than when using neat gasoline.
Important aspects of adding an alcohol to gasoline to consider are its effect not only on the
exhaust emissions, but also on the evaporative emissions and how the fuel is changed by the
addition. The aim of an investigation carried out by Laveskog and Egebck (1999) for the
Swedish Transport and Communications Research Board (KFB) was to study the effect of such
additions on the RVP of the fuel and the evaporative emissions by adjusting the vapour pressure
of the fuel.
Two types of gasoline were ordered from a refinery with vapour pressures (RPV) of 63 kPa and
70 kPa. To elucidate the strength of the effect of adding alcohol on the vapour pressure, samples
were sent to a special laboratory for analysis (see Fuel type, Table 7.12). Both regulated and
evaporative emissions were measured (the latter by the standardised SHED method. Ethanol
blended fuel was compared with neat gasoline in order to determine how addition of ethanol
affects the exhaust emissions and evaporative emissions compared with neat gasoline.
In total, three vehicles were each tested, once, with three test fuels with two different vapour
pressures, yielding a total of 18 emission tests. Of the three cars tested, two were equipped with
a catalytic converter. The two older cars a Volvo 240 GL 1985 model and a Volvo 240 DL
1987 model represented the technology level found in most cars sold in Sweden in the years
following 1987 but only the latter had a catalytic converter.. The age of the oxygen sensors and
catalysts of this car corresponded to approx. 50,000 km driving in traffic and they were designed
to meet the Swedish A12 emission requirements. The other vehicle, equipped with a catalytic
converter, was a Saab 9000T and the engine of this car was upgraded to correspond with a 1996
model, equipped with an adaptive fuel supply system, which automatically adjusts the rate of
injected fuel according to the energy content of the fuel.
In addition to the emissions tests, the oldest car was also tested for evaporative emissions (SHED
test) with the base gasoline (71.5 RVP), base gasoline (64 RVP) + 10% ethanol, and base
gasoline (71.5 RVP) + 25% ethanol. The emission test results and the fuel consumption for the
catalyst-equipped cars (Volvo DL and Saab 9000T) are presented in Tables 7.13 and 7.14.
Page 61
Table 7.13. Results of measurements of exhaust emissions during an urban driving cycle (UDC)
according to US standards. Volvo 240 GL, 1985 model. Base gasoline, 64 kPa and 71.5 kPa.
Car Amount Exhaust RVP CO HC NOX CO2 NMHC Fuel
EtOH% System kPa g/km g/km g/km g/km g/km consump-
tion
l/10 km
Volvo DL 0% EtOH Cat. 64 2.15 0.21 0.17 243 0.18 1.00
10% EtOH Cat. 69 1.82 0.17 0.18 240 0.15 1.03
25% EtOH Cat. 67 1.13 0.12 0.23 235 0.10 1.06
Table 7.14. Results of the measurement of exhaust emissions during an urban driving cycle
(UDC) according to US standards. Saab 9000 CS, 1996 engine and engine management system.
Base gasoline 64 and 71.5 kPa.
Car Amount Exhaust RVP CO HC NOX CO2 NMH Fuel
EtOH% System kPa g/km g/km g/km g/km C consump-
g/km tion
l/10 km
Saab 9000 T 0% EtOH Cat. 64 1.16 0.06 0.14 242 0.05 0.99
10% EtOH Cat. 69 1.06 0.06 0.14 253 0.05 1.08
25% EtOH Cat. 67 1.23 0.05 0.13 250 0.04 1.13
The CO and HC emissions were lower for the Volvo with a catalytic converter (Table 7.13).
NOX emissions increased significantly when the gasoline with a vapour pressure of 64 kPa was
used, but insignificantly with the 10% mixture. Despite the increase, the NOx emissions were
still lower than the emission standards for this car. When gasoline with a vapour pressure of 71.5
kPa was used, all emissions decreased slightly with the 10% ethanol mix. With the 25% ethanol
mix, NOX emissions rose from 0.19 g/km to 0.30 g/km. Efficiency also seemed to increase for
this car, as the increase in fuel consumption for ethanol mixtures does not correspond to the
reduced energy content of the test fuel.
In the car with an adaptive fuel supply system the effect of the ethanol mixture on CO emissions
appears to be small, and the already very low level of HC emissions were almost totally
unaffected (Table 7.14). NOX emissions were low, and showed a tendency to fall further as the
ethanol content in the fuel increased. In the tests with the fuel with an RVP of 71.5 kPa, the
tendency was the same as for the other cars the fuel consumption was lower than expected
when ethanol was added to the gasoline. The comparison of these two cars indicates that a car
with an adaptive emission control system has better emission characteristics than a car without
such a system.
Unfortunately, evaporative emission tests were only carried out on the Volvo 240 GL, 1985
model, which lacked a catalyst and canister for trapping fuel vapour. The results from the tests of
this car are presented in Table 7.15.
Table 7.15. Evaporative losses from a car without evaporation controls, Volvo 240 DL, values
from single measurements using fuels with 0% and 10% ethanol contents.
HC emission g/test Increase in
emissions [%]
Base gasoline, 71.5 RVP 37.2 -
Base gasoline, 64 RVP +10% ethanol 38.9 4.6
Base gasoline, 71.5 RVP +10% ethanol 42.4 14
Since 1999, the American Lung Association of Minnesota has been presenting studies on the
impact on emissions of sulphur in the fuel and this issue has been discussed in a number of
reports, which have shown that the emission rate depends considerably on the fuel used. The aim
of the investigation discussed here was to further study the impact on emissions of the sulphur,
benzene and olefins contents in gasoline. Minnesota gasoline is regulated under the
Antidumping* Requirements introduced by the EPA and therefore it is of interest to study
whether these requirements are fulfilled. The Antidumping Statutory Baseline Fuel Parameters
set by the US EPA are shown in Table 7.16.
In 1999 the EPA introduced more rigorous emission standards, necessitating amendments to the
fuel requirements. Since the sulphur content is one of the most important variables to reduce in
the fuel, the EPA has established new requirements for the sulphur content in gasoline. According
to the Federal Register / Vol. 65, No. 28 / Thursday, February 10, 2000 / Rules and Regulations
the level of sulphur should be reduced, on average, to 15-40 ppm by weight and limited to a
maximum of 80 ppm by 2006, see Table 7.17.
Table 7.17. Gasoline sulphur requirements according to the Federal Register, February 10, 2000.
Gasoline sulphur standards for the averaging period beginning:
January 1, 2004 January 1, 2005 January 1, 2006 and
subsequent
Refinery or Importer N/A 30.00 30.00
Average
Corporate Pool Average 120.00 90.00 N/A
Per-Gallon Cap 300 300 300
The paper by Aulich and Allen of the Environmental Research Center, University of North
Dakota, released by the American Lung Association of Minnesota, reports results from an
evaluation of the impact of fuel sulphur on emissions. Samples of gasoline from three different
gasoline suppliers were analyzed in detail and the impact of the fuel sulphur content was
evaluated by using the resulting fuel data as inputs in the EPAs MOBILE6.2 vehicle emission
*
The purpose of the Antidumping Requirements is to ensure that conventional gasoline is not more polluting than it
was in 1990. See Federal Register / Vol. 65, No. 230 / Wednesday, November 29, 2000 / Rules and Regulations.
.
Page 64
modelling software (US EPA 1993). In all, 23 samples of fuel containing 9.2 to 9.8 % ethanol
were analyzed. The results of this part of the evaluation are summarized in Table 7.18.
Table 7.18. The estimated impact of fuel sulphur content on HC and NOx emissions. Holiday,
Amoco and SA (SuperAmerica) are gasoline companies.
Holiday Amoco SA
HC, g/km 0.069 0.095 0.098
NOx, g/km 0.31 0.50 0.51
Sulphur in Fuel, ppm 57 167 188
Two other components in the fuel that have a significant impact on the emissions are olefins and
benzene. Benzene, a single component of gasoline, is known to have an effect on peoples health
while olefins comprise a group of hydrocarbons that are known to be reactive in the atmosphere
and may also cause deposits in the fuel system of the vehicle. Therefore, there is a need to keep
the contents of these components as low as possible in gasoline.
The results of the evaluation conducted by Aulich and Allen on the impact of olefins and
benzene in the fuel are summarized in Table 7.19.
Table 7.19. The estimated impact of the fuel contents of olefins and benzene on emissions of 1,3
butadiene and benzene. Holiday, Amoco and SA (SuperAmerica) are gasoline companies.
Holiday Amoco SA
1.3 Butadiene, 0.00034 0.00071 0.00050
g/km
Benzene, g/km 0.0035 0.0064 0.0055
Olefins in Fuel, % 6 17 7
Benzene in Fuel, % 0.9 2.3 1.4
In another study presented by the American Lung Association of Minnesota (2003) the five
blends of fuel listed in Table 7.20 were studied to assess the impact of sulphur.
Table 7.20. Average Fuel Properties (American Lung Association of Minnesota, 2003).
The emission tests followed the hot start phase of the FTP-75 test cycle. The results of the
tests carried out on a dynamometer and the data evaluation for the fuels listed in Table 7.19 are
shown in Table 7.21.
Page 65
Table 7.21. The impact of fuel sulphur content on HC and NOx emissions.
HC and NOx emissions for the Holiday fuel were under the detection limits, and the
estimated levels of these emissions had to be based on their respective detection limits.
For the BP fuel HC emissions were consistent with the predictions according to
MOBILE6.2, but the levels were higher than for the other fuels tested.
For the non-ethanol fuel with a sulphur content of 103 ppm, HC and NOx emissions were
similar to those for the BP and SA fuels.
The emissions of HC and NOx were much lower for E85, which had a sulphur content of
just 8 ppm, than for the other fuels.
The measured emissions were significantly lower than the levels predicted by the EPA model,
possibly because the tests were carried out with hot cycles while the EPA predictions were based
on a complete, cold start FTP cycle.
This statement is consistent with findings by Laveskog and Egebck, since their investigation
(discussed in section 7.4) indicated that the lower energy content of the blended fuel was to a
certain degree compensated by its higher combustion efficiency (Laveskog and Egebck, 1999).
However, according to some other sources, e.g. a literature review prepared by the Orbital
Engine Company for Environment Australia, there is a direct proportionality between fuel
economy and the energy content of the ethanol-gasoline blend (Orbital Engine Company, 2002).
In another report Orbital Engine Company has presented results from measurements of fuel
consumption according to the Highway Fuel Economy Cycle (HWFET), as shown in Figure 7.15
(Orbital Engine Company 2004).
Page 66
The measurements of emission and fuel consumption were carried out on five vehicles of
newer models and comprised tests in which both gasoline and ethanol were used in the same
vehicle in order to generate comparative data. Some results of the emission measurements
carried out by the Orbital Engine Company to study the efficiency of catalysts are presented in
section 7.1, and in the following figures (7.16-7.18) the fuel consumption results are presented.
Page 67
FTP 75 Average Fuel Consumption for the Vehicle Fleet over milage.
Fuel Consumption [L/100 km] Gasoline vs E20
14 6400 km
12 80 000 km
10
8
6
4
2
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Holden Ford Toyota Hyunday Subaru
Figure 7.16. Average fuel consumption of the tested vehicles at 6,400 and 80 000 km when
driving according to the cycle city
Highway Average Fuel Consumption for the Vehicle Fleet over milage.
Fuel Consumption [/100 km]
Gasoline vs E20
9 6400 km
8
7 80 000 km
6
5
4
3
2
1
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Holden Ford Toyota Hyunday Subaru
Figure 7.17. Fuel consumption of the tested vehicles at 6,400 and 80 000 km when driving
according to the city cycle
Metro-Highway Average Fuel Consumption for the Vehicle Fleet over milage. Gasoline vs E20
Fuel Consumption [l/100 km]
12
6400 km
10 80 000 km
8
6
4
2
0
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Gasoline
E20
Figure 7.18. Average fuel consumption of the tested vehicles at 6,400 and 80 000 km when
driving according to the city cycle.
Page 68
A research project has been carried out on 15 vehicles of models from different years by a group
at the Minnesota Centre for Automotive Research, Minnesota State University. Two blends of
ethanol in gasoline, E10 and E30, were used during the test program, which was carried out to
measure fuel consumption and emissions, and to analyse drivability characteristics, engine wear
and material compatibility. The vehicles were tested on a chassis dynamometer according to both
the hot 505 section of the standard test cycle and on the road when driven by owners of the cars.
The fuel consumption and energy used per unit distance in the chassis dynamometer tests were
matched to the distance driven in the on-road tests, based on the volumetric consumption. Oil
samples were taken to analyse particles in the oil to evaluate whether accelerated wear of the
engine had occurred. During the on-road tests the vehicles were driven by their owners
(Bonnema et al., 2004).
The drivers were asked to fill in forms, the first of which concerned fuel consumption, while a
second concerned maintenance, including changes to the car, such as changes of tyres or other
parts which could have affected the results of the study. A third form concerned drivability.
Consequently, the drivers of the cars during the field tests had to collect samples of oil for
analysis and to record data concerning:
The results for fuel economy obtained in the Hot tests on the chassis dynamometer are
summarized in Table 7.22.
Table 7.22. Fuel consumption and energy use when tested according to Hot 505 procedures
(Bonnema et al., 2004).
Vehicle E10 E30 % E10 E30 %
l/10 l/10 Diff, MJ/10 MJ/10 Diff.
km km km km
1996 Oldsmobile Acieva 1.01 1.19 +14.66 31.93 34.52 +8.11
1998 Dodge Caravan 1.09 1.24 +12.24 34.22 35.98 +5.16
1997 Chevrolet K3500 1.85 1.97 +6.14 58.33 57.37 -1.66
1994 Buick Regal 1.06 1.09 +2.96 33.28 31.64 -4.92
1997 Chevrolet K 1500 1.69 1.86 +8.99 53.26 54.01 +1.41
1998 Ford F-250 1.84 2.01 +8.74 57.81 58.42 +1.06
1997 Ford Taurus 1.28 1.29 +1.30 40.23 37.60 -6.54
1997 Ford F-150 1.65 1.85 +0.78 51.90 53.65 +3.38
1990 Chevrolet C1500 1.09 1.22 +10.42 34.46 35.50 +3.00
1992 Chevrolet K1500 1.51 1.68 +9.88 47.56 48.70 +2.39
1992 Geo Metro 0.62 0.69 +11.04 19.42 20.15 +3.73
+ indicates an increase in fuel consumption or energy use when comparing E30 with E10.
- indicates a reduction in fuel consumption and energy use when comparing E30 with E10.
It should be noted that the authors of the present report have not had access to appendices to the
cited report, and hence have not been able to verify the interpretation of the data collected.
Page 69
However, the data in the above table indicate that there were differences between vehicles with
respect to the effects of the ethanol blend in gasoline. Fuel consumption increased on a
volumetric basis when E30 was used instead of E10, as expected, and the differences varied
from less than 1% to over 12 %, which was less than expected. Concerning the energy used, the
influence of using ethanol-gasoline blends with higher ethanol contents should be similar to their
effect on fuel consumption, but as can be seen, the table shows a decrease for three of the
vehicles. This may be due either to these vehicles being more energy efficient, or to uncertainties
in the measurement of fuel consumption.
The fact that vehicle studies have found a great diversity of responses to increasing ethanol
content in terms of fuel consumption can be explained in a number of ways. However, a
complicating factor is that the cited studies do not supply specific information on either the
engine specifications or the fuels used for testing, although knowledge of these variables is
essential for fully understanding the data.
For example:
If a test compares a fuel with no ethanol and one with a 10% ethanol content, and it is run in an
area where the base fuel used has a relatively low octane number, the lower energy content of the
ethanol blend may be somewhat compensated by its higher octane rating, provided that the
engines used have advanced ignition (knock control) systems according to Chandra Prakash
(1988).
8. VAPOUR PRESSURE
It is known that adding ethanol to gasoline raises the vapour pressure of the fuel. The chemical
explanation for this is that gasoline consists mainly of hydrocarbons (straight, branched and
aromatic hydrocarbons), which are nonpolar and have different boiling points. In contrast, the
ethanol molecule (CH3CH2-OH) contains a hydroxyl group (-OH) which makes it polar. It is a
well-known fact from solubility investigations that like dissolves like, i.e. polar solvents dissolve
polar solutes and nonpolar solvents dissolve nonpolar solutes. Thus, introducing polar
compounds like ethanol into gasoline (which is nonpolar) makes the gasoline blend more polar
than neat gasoline, and increasingly polar as the ethanol content increases. This results in the
vapour pressure of the ethanol-gasoline blend rising, since nonpolar compounds in the neat
gasoline, with relatively low boiling points (and hence relatively high vapour pressure), will
evaporate as the gasoline blend becomes more polar. In Figure 8.1 the Reid Vapour Pressure
(RVP) (measured, by definition, at 100 F) is shown as a function of the percentage of ethanol
blended with gasoline. The neat gasoline from 1997 (Furey and Jackson, 1977) had a higher
initial vapour pressure than the gasoline from 2002 (Hsieh et al., 2002). The lightest chemical
component of gasoline fuel in general is butane, which is removed to produce low RVP gasoline
(Korotney, 1996). This indicates that the evaporative emissions from ethanol blended gasoline
consist mainly of butane.
RVP, kPa 80
70
60
50
40
30
20
10
0
0 20 40 60 80 100
Volume % ethanol
Figure 8.1. Reid Vapour Pressure (at ~38 C) shown as a function of the percentage of ethanol
blended with gasoline (gFurey and Jackson, 1977; DOE, 1991).
For both of the fuels shown in the figure the highest relative RVP was obtained at a percentage
of ethanol in the neat gasoline of about 10%. Hence, there are strong indications that the
maximum RVP will be obtained at a 5 to 10 percent ethanol content in gasoline, regardless of the
neat gasoline fuels used. Decreasing the initial vapour pressure of the neat gasoline will reduce
the absolute vapour pressure of a 10 percent ethanol blend. Furthermore, in Figure 8.1 the largest
vapour pressure increase (a approximately 10 % increase in the RVP) is observed in the 0 to 5
percent ethanol interval. A compound in gasoline which has a major impact on the RVP is
butane (Korotney, 1996).
Page 71
The compounds in the vapour phase from regular, mid and premium grade gasoline containing
ethanol (3.25 to 9.65 wt%) have been investigated by Harley and Coulter-Burke (2000), in a
headspace analysis in which they concluded that n-butane, n-pentane, 2-methylbutane and
ethanol collectively accounted for more than 50% of the total headspace vapour mass at 38 C.
A large number of paper and reports have focused on the fact that blending ethanol in gasoline
increases the vapour pressure, and this phenomenon has been investigated in a number of
studies, especially in the USA, for example Issues Associated with the Use of Higher Ethanol
blends (NREL 2002). In the early 1980s Furey (GM) reported that even a low level of alcohol
in gasoline could have a high impact on the vapour pressure of the fuel (Furey 1985; 1986;
1990). A commonly quoted figure for the increase in RVP when blending base gasoline
available on the US market with 10% ethanol is 1 psi (a little less than 7 kPa), see below
concerning the 1 psi waiver.
A report prepared for the Swedish Transport and Communications Research Board presented a
limited investigation of two fuels (one vapour pressure-adjusted and the other unadjusted) used
in three cars (see also section 7.4). The increases in vapour pressure found after adding 10 %
ethanol were 5 kPa and 4.5 kPa, respectively.
In many countries, including European countries, RVP is commonly expressed in kPa units,
while psi (pound per square inch) is used in the USA. To facilitate comparisons, a graph for
converting psi to kPa and vice versa (based on a psi to kPa conversion factor of 6.895) and psi to
kp/cm2 (another unit for measuring RVP used in some reports) is shown in Figure 8.2.
110 1,11
100 1,01
90 0,91
80 0,81
kPa kP/cm 2
70 0,71
60 0,61
50 0,51
40 0,41
6 8 10 12 14 16
psi
An important issue in several countries related to the increase in vapour pressure caused by
mixing gasoline with ethanol has been whether or not compensatory adjustments should be made
to the vapour pressure of the base gasoline. If it is not adjusted the RVP may increase above
regulated limits and may result in increased evaporative losses. Since many countries have
regulations concerning RVP failing to meet the standards may even be illegal. In the EU, the
RVP of gasoline is regulated by the European gasoline standard (EN 228).
Page 72
Since a high vapour pressure of the fuel may cause vehicle drivability problems and make hot
starting impossible as a result of vapour lock, another relevant issue is driver acceptance of the
fuel.
According to the literature some countries at least, such as the USA, Australia and Sweden, have
decided to decrease the vapour pressure of the base gasoline used in alcohol blends in order to
meet the existing standards/regulations.
There has been ongoing debate in the USA about whether or not this waiver should be retained.
The US EPA has been arguing that the waiver for ethanol-blended gasoline can be seen as a
barrier hindering the broader introduction of reformulated gasoline. On the other hand, there has
been pressure from the farmers producing ethanol and the Ethanol Fuel Association, since an
increase in the use of ethanol, as a replacement for MTBE, would increase demand for ethanol.
One factor to take into account, especially for the US EPA, is that the use of ethanol blended
gasoline is recommended in CO non-attainment areas.
Documents from the California Air Resources Board (CARB) (CARB, 1998) show that
replacing gasoline with ethanol may clearly reduce CO emissions, even if the gasoline is not
adjusted to comply with RVP requirements (i.e. the 1 psi waiver is applied). Thus, adding
ethanol to gasoline has provided a means to meet CO levels required by the air quality standards
in many of the current, so-called CO non-attainment areas. However, these documents also show
that adding ethanol to gasoline without adjusting the RVP leads to increases in the emissions of
volatile organic compounds (VOCs) and nitrogen oxides (NOx), and in the fuels ozone-forming
potential. Since both VOC and NOx are important emissions in the formation of tropospheric
ozone the 1psi waiver has not been applied to mixtures of ethanol in reformulated gasoline
(RFG), which is primarily intended for use in ozone non-attainment areas to avoid increased
formation of ozone.
The effect of this waiver is discussed in a study released by the EPA that provided part of the
rationale for introducing reformulated gasoline (RFG), i.e. gasoline containing an oxygenate,
which at the time could be either MTBE or ethanol (US EPA, 1993). It was proposed that a
renewable oxygenate should be used, and one of the most controversial questions addressed
when the RFG rule was drafted was how ethanol/gasoline blends should be treated. The EPA
argued that the base gasoline had to be adjusted to ensure that the vapour pressure met the
standards. However, the EPA had to accept that adjusting the RVP could lead to cost penalties,
inter alia, for the ethanol blended fuel. An argument propounded in comments sent to the EPA
was that ethanol would be excluded from the market without a waiver. The opinion among the
rule makers within the EPA was that ethanol is cheaper than MTBE per gallon and that this cost
difference could well compensate for the costs of adjusting the RVP, which are minor compared
to the cost of adding oxygen via oxygenate additions.
In the cited paper (US EPA, 1993) the EPA also discussed the detrimental effect on emissions of
granting the waiver of 1 psi RVP, since blending ethanol in gasoline will increase the vapour
pressure of the fuel, and hence increase VOC emissions. Two letters from EPA staff claim that a
fuel that met RFG requirements in every respect except its RVP and had an RVP of 1 psi over
the limit would increase VOC emissions by about 20 % relative to a baseline gasoline.
Page 73
Furthermore, the EPA argued at that time that ethanol blends with a 1 psi waiver would have an
advantage over RFG and thus might dominate the market (US EPA, 1993).
The estimated reduction in VOC emissions likely to follow the introduction of RFG based on
unblended gasoline is estimated to be 11 %, see Table 8.1. The reduction will be less if ethanol
blended gasoline is used, and considerably less if RVP-adjusted gasoline is not used as a base
according to the EPA estimations. Consequently, the reduction in VOC emissions will diminish
as the ethanol market share increases, as can be seen in Table 8.1.
Table 8.1. Losses of In-Use VOC Emission Control as a Result of the Ethanol Waiver for
RFG*
Ethanol Market Share (%)
0 8 24 30 35
% VOC Reduction Due to 11 10.1 8.4 7.7 7.2
MTBE Share
% VOC Reduction Due to 0 -1.0 -3.1 -3.9 -4.6
Ethanol Share (with distillation)
% VOC Due to Commingling 0 -1.2** -2.4 -2.4 -2.4
Total % VOC Reduction 11 7.9 2.9 1.4 0.2
Percentage Change from 100% N/A 28 74 87 98
MTBE Baseline
* Emission percentage change input values shown are based on those calculated in the letter
from Chester France to Dr. Gary Whitten. Memorandum 6 from Paul A. Machiele, Fuel
Studies and Standards Branch, to Richard D. Wilson, Director, Office of Mobile Sources,
"Update of the Relative Ozone Reactivity of Reformulated Gasoline Blends," June 11, 1993.
** Commingling, assuming half of that at higher market shares based on analysis (US EPA,
1993).
A paper from the US Energy Information Administration section (EIA, 2002) of the US
Department of Energy (DOE) describes, inter alia, the effect of adjusting the RVP of base
gasoline intended for blending with ethanol. A primary feature it displays is that there is a
relationship between the RVP of the base gasoline, and the magnitude of the impact on the RVP
when adding ethanol to the base gasoline is shown in Table 8.2
The paper cited above (EIA, 2002) discusses the energy losses from the gasoline pool due to
removing C4 and C5 hydrocarbons from it in order to meet the RVP requirements of gasoline
without the 1-psi waiver. The study that the paper was based upon was initiated by discussions
about the possible removal of the 1-psi waiver following a request by Senator Bingaman
Page 74
(Chairman of the Senate Committee on Energy and National Resources) presented in a separate
paper (Office of Oil and Gas of the Energy Information, 2002).
It is interesting to note that an adjustment of the base gasoline intended for use with ethanol
blends in order to keep the RVP under the required limits may reduce the energy content of the
gasoline pool to a larger degree than the energy supplied by the added ethanol (EIA, Energy
Information Administration, 2002). However, the magnitude of the reduction in energy content
depends on the RVP requirement and the gasoline used for blending. In the USA, especially
California, where the maximum RVP allowed is 7.0 psi (48.3 kPa), adjustment of the gasoline to
be used for blending with 10 % ethanol will result in a 2.8 % reduction in its energy content. In
contrast, in areas of the USA where the maximum RVP is 9 psi (62.1 kPa) the result of adding 10
% ethanol to RVP adjusted gasoline will be a 4.5 % increase in the energy content. Not adjusting
the base gasoline intended for use in 10 % ethanol blends will result in a 5.1 % increase in the
energy of the gasoline pool if there is a 7.0 psi RVP requirement and a 7.5 % increase if there is
a psi 9.0 RVP requirement.
These two examples could be used as models for calculating the effect of adjusting the base
gasoline for blending, in energy terms, when deciding whether it should be adjusted or not.
However, it should be emphasised that the components (mostly butane) removed from the base
gasoline when adjusting the RVP will generally be used for other purposes and not wasted. The
increase in evaporated hydrocarbon emissions that will occur if the RVP is not adjusted should
be analyzed and discussed when deciding whether these polluting emissions will be
unacceptably high (see table below).
Korotney addresses implications of the facts that the only compound removed from conventional
gasoline is butane and that low RVP gasoline is produced solely to meet the RVP requirements
of 6.8 psi to 8.1 psi, with typical values of 7.0 to 7.2 psi.Since the intentions when designing
reformulated gasoline were not only to meet the RVP requirements, but also to reduce emissions
Page 75
of toxic substances, his conclusions concerning low RVP gasoline, compared with conventional
gasoline are inter alia that:
Another paper released by the EPA (US EPA, 1996) represents a response to a letter from the
American Petroleum Institute (API) is presented (US EPA, 1996)The director of the API, C.J.
Krambuhl, opposed statements of the EPA (Korotney) in their paper about low RVP gasoline.
The main argument by Krambuhl was that the low RVP gasoline is more cost-effective than the
reformulated gasoline. In a short answer to the API, the EPAs Fuels and Energy Division
Director, C.N. Freed, notes that The modest reductions in NOx for Phase I, RFG, and the more
substantial reductions in NOx for Phase II, RFG, provide a means for immediately reducing
ozone without waiting for fleet turnover. He also invited the API to a meeting with the EPA in
order to clarify the latters position. In a report (US EPA, 1999) it is stated that Beginning on
January 1, 2000, the Phase II complex model standards at 40 CFR 80.41(e) and (f) will apply to
all RFG in the gasoline distribution system.
Nevertheless, very few tests have been done in Brazil regarding the environmental effects of
using ethanol as a vehicle fuel. Ethanol is regarded as an environmentally friendly fuel with low
CO2 emissions, and no need for further debate is recognised. The potential problems associated
with increased vapour pressure due to adding ethanol to gasoline, including increased
Page 76
evaporative losses, do not apply in Brazil since they use either neat ethanol (with very low RVP)
or at least 20 % ethanol blended in gasoline; a level at which the increased vapour pressure has
declined more or less back to the original RVP of the base gasoline (Alfred Szwarc, personal
communication, 2004).
In Sweden not many investigations have been carried out on the relationship between ethanol
blends in gasoline, increased vapour pressure and increased evaporative emissions. However,
one such study is described in the report, mentioned earlier, prepared for the Swedish Transport
and Communications Research Board (KFB) by Laveskog and Egebck (1998; see section 7.4).
In another, the vapour pressure and evaporative emissions from E85 (85 % ethanol and 15 %
gasoline), E10 and reference gasoline were studied in a project carried out by Lu-Karlsson
(1999) at the MTC (Motortestcenter of Sweden). Using a simulated Sealed Housing for
Evaporative Determination (SHED) test method, the total amounts of hydrocarbons (HC) that
evaporated were quantified and the composition of the evaporative emissions was chemically
characterized. The evaporative emission components detected were ethanol, benzene, toluene,
C4-C6, alkanes and olefins. The E10 mixture gave the highest HC evaporative emissions.
Compared to gasoline, the E10 gave elevated C4 C6 alkanes and olefin emissions, but reduced
aromatic emissions. Compared to the reference gasoline, the base gasoline with a higher vapour
pressure gave significantly higher C4 C6 alkane and alkene emissions.
At the European level (JCR ISPRA), in 2004 Eucar and Concawe carried out a joint
investigation of evaporative losses and other phenomena associated with blending ethanol in
gasoline. The objectives of the project were:
To clarify the effect of fuel vapour pressure on volatility and evaporative emissions from
modern cars equipped with canisters, representative of the recent European fleet.
To assess the effect of ethanol blending on fuel properties and evaporative emissions
(quantitative and qualitative).
To assess the impact of the cars fuel system technology on the evaporative emissions
(canister aging, metal/plastic tanks).
To provide a firm technical basis for debates on gasoline vapour pressure limits in
relation to ethanol blending for the Fuels Directive Review.
The tests were to be carried out on two base gasolines (one vapour pressure-adjusted and one
unadjusted), and ethanol blends (with 5 and 10 % ethanol contents) of these fuels were to be
examined. However, no official report from the project has been presented as yet.
*
During wintertime in Finland, Norway and Sweden the RVP ranges from 50 to 95 kPa.
Page 77
Where Ptotal represents the total vapour pressure of a solution containing gasoline and ethanol,
gas and eth are the mole fractions of gasoline and ethanol, respectively, and P0gas and P0eth are
the pressures of the pure solutions. As gasoline is a mixture of chemical compounds and not a
pure compound, only a theoretical discussion can be applied. However, in Figure 8.3 the vapour
pressure of a gasoline-ethanol fuel blend is shown as a function of the ethanol mole fraction
(eth), based on the same primary data as used in Figure 8.1. The ethanol mole fraction was
calculated using a theoretical calculated mean molecular weight of gasoline. The theoretical
mole fraction of gasoline in the ethanol-gasoline blend was calculated using information on the
percentage (by weight) of carbon and hydrogen in the gasoline. The estimated weight
percentages used for carbon and hydrogen in gasoline are 87 and 13, respectively, accordingly to
Hsieh et al. (2002). Using these values as input data a theoretical molecular formula of C15H26
can be calculated for gasoline, consisting of 87 percent (w/w) carbon and 13 percent (w/w)
hydrogen. In Figure 8.3 a positive deviation from Raoults law can be seen, which can be
attributed to the ethanol and gasoline interactions in the blend being weaker than the interactions
among the molecules as neat liquids. Molecules in the blend have a higher tendency to reside in
the gas phase, thus increasing the vapour pressure and leading to a positive deviation from
Raoults law, as shown in Figure 8.3.
Page 78
Figure 8.3. Reid Vapour Pressure (gFurey and Jackson, 1977; DOE, 1991) as a function of
the mole fraction of ethanol (eth) blended in gasoline. X Ethanol RVP, gasoline RVP,
* gasoline/ethanol blend of ideal solution.
8.6. Findings
When ethanol is added to gasoline the vapour pressure of the resulting blend increases. The
highest RVP is reached at a blending with approximately 10 % ethanol. Increasing the
ethanol content above 10 % reduces the RVP, and at levels of 20 to 40 % ethanol (depending
on the specification of the gasoline used) the RVP of the blend returns to the original level of
the gasoline. Further increases in ethanol content will further decrease the RVP of the blend
to levels below the original RVP of the gasoline used.
Increased RVP may increase VOC emissions, depending on the type of evaporative emission
control system used. However, no direct correlations between the gasoline vapour pressure
and the exhaust VOC emissions from vehicles have been found.
RVP-adjusted gasoline that is used for blending with ethanol, but still fulfils the RVP
requirements set by EN 228 specifications, will have a lower energy content than gasoline
that is not so adjusted
Increased vapour pressure of a fuel may affect the drivability of the vehicle at high ambient
temperatures, i.e. vapour lock may occur. It may also affect hot engine start ability.
There is not much data from Swedish or European investigations on the effects of increased
fuel vapour pressure on evaporative and tailpipe emissions. This issue is currently being
investigated in an EU project, but no data or reports arising from it have been published to
date.
It is difficult to apply data from USA directly to Swedish or European conditions, since the
gasoline used in the USA is quite different from Swedish/European gasoline standards.
Gasoline standards applied in the USA are designed to solve problems such as poor air
quality in city centres and the formation of tropospheric ozone, which are not always
prioritised in Europe. Furthermore, the data from the USA can be mutually contradictory.
There are very few data on emissions from Brazil regarding vehicles running on neat ethanol
or ethanol gasoline blends. However, since the lowest ethanol content of the Brazilian fuel
blends is 20%, the evaporative emissions from the vehicles are limited.
Page 79
The ISO standard for LCA (ISO 14040, 1997) defines it as a technique for assessing the
environmental aspects and potential impacts associated with a product, by:
LCA can be used to assess and compare, in a cradle to grave perspective, effects of using
different vehicle fuels, such as gasoline, bio ethanol and bio ethanol-blended gasoline. The
results may elucidate the total effects on the climate, environment and health of replacing
gasoline to varying degrees by bio ethanol. However, when using LCA for
assessment/comparison of different situations it is important to apply the same system boundary
and framework conditions consistently. Otherwise, the LCA will not give reliable answers to the
questions posed. For example, in response to a request by the Swedish Alternative Fuel
Committee, Magnus Blinge of Chalmers University of Technology, Gothenburg (Blinge, 1996)
compared about 20 different LCAs related to conventional and alternative vehicle fuels. One of
Blinges main conclusions was that it was not possible to make a good comparison between
these LCAs or to extract relevant average emission data concerning production, vehicle use etc.
from them, largely because of small differences between them in terms of parameters such as the
test vehicles (age and engine/vehicle configuration) and the test cycles used.
The environmental and health effects of emissions from the use of neat gasoline and
various ethanol-gasoline blends, ranging from 5 to approximately 20 % ethanol.
Both vapour-pressure adjusted gasoline (i.e. gasoline fulfilling RVP requirements even
after blending with ethanol) and unadjusted gasoline (with which blends would not meet
such requirements).
Emissions not only of carbon dioxide and/or other greenhouse gases, but also of
regulated emissions and, as far as possible, other unregulated contituents.
Page 81
Potential differences in the emissions associated with various ethanol blending levels in
different types of gasoline (RVP-adjusted and unadjusted).
Changes in emissions from sensitive emission points from harbour/refinery to
vehicle/tailpipe due to use of ethanol-gasoline blends with various ethanol contents and,
if possible, identification of the best point in the chain for blending ethanol into gasoline,
to minimise evaporative losses.
If (and if so how) switching from the use of RVP-adjusted gasoline to unadjusted
gasoline (allowing increases in vapour pressure) in gasoline-ethanol blends would affect
the overall climatic, environmental and health impact.
A further objective of the work was to find out if LCAs, such as those mentioned above, carried
out in the USA and other countries could provide a reasonable basis for predicting life-cycle
changes in emissions that would follow the use of 5 to 20 % ethanol blends with gasoline, and
fuels for use in Otto engines, and their impact (especially on the environment and health), under
Swedish or European conditions. However, it was (and is) essential to remember that basing
predictions on LCAs is likely to be difficult, or even impossible, unless the system boundary and
framework conditions are very similar.
10.2. LCA Studies for Ethanol in Neat Form or Blended with Gasoline
USA
The literature survey revealed that many LCAs have been carried out on the production and use
of ethanol, mostly in the USA, under typical American conditions, e.g.:
The raw materials for the production in the USA are often corn or corn residues.
The vehicle parameters used in the LCA calculations are generally typical for
engines/vehicles sold in the USA 5 to 10 years ago.
The conventional fuel used in the comparisons is generally common US gasoline, which
is not the same as European and less than that Swedish gasoline. In addition, the gasoline
quality differs between regions/states in the USA and the US authorities have established
special gasoline qualities for areas with special air quality problems, the so-called non-
attainment areas (with respect to CO or tropospheric ozone). This means gasoline
qualities with a minimum level of oxygen or a minimum level of added oxygenate
(MTBE or ethanol), gasoline with a maximum level of vapour pressure (low vapour
pressure gasoline, LWG) or gasoline with a maximum vapour pressure and low level of
specific compounds, such as benzene and aromatics, reformulated gasoline (RFG).
Fuel-Cycle Fossil Energy Use and Greenhouse Gas Emissions of Fuel Ethanol Produced
from U.S. Midwest Corn, Michel young, Christopher Saricks and May Wu, Argonne
National Laboratory, US, 1997 (Young et al., 1997).
Effects of Fuel Ethanol Use on Fuel-Cycle Energy and Greenhouse Gas Emissions, M
Wang, C Saricks and D Santini, Argonne National Laboratory, USA (Wang et al., 1999).
However, since these LCAs were based on conditions specific for the USA there are various
difficulties in using them as a basis for predictions related to Swedish and/or European
conditions other than merely giving hints about likely results.
Page 82
Furthermore, the LCAs that have been found mostly focused on emissions of carbon dioxide,
and sometimes other greenhouse gases, but LCAs seldom consider emissions of regulated or
other non-regulated constituents.
Europe
For European conditions General Motors, together with Argone National Laboratory, published
a report entitled Well-to-Wheel Energy use and Green House Gas Emissions of Advanced
Fuel/Vehicle System A European Study The work was carried out with BP, ExxonMobil,
Shell and TotalFinal Elf as active participants, and is a complement to a similar study for North
American conditions published in 2001 (GM 2001). However, as can be seen from the title this
LCA-study only covers greenhouse gases and energy use.
Another LCA related to European conditions was jointly undertaken by CONCAWE, EUCAR
and the EU-Commissions/Joint Research Centre, as reported in Well-to-Wheel Analysis of
Future Automotive Fuels and Power Trains in the European Context (Concawe 2004). This
report also focuses on the emissions of greenhouse gases, as do two other European LCAs on bio
fuels/ethanol. One carried out by Patyk and Reinhardt, of the Institut fr Energie und
Umweltforschung, Heidelberg, Germany, (Patyk and Reinhardt, 2002), entitled Life Cycle
Analysis of Bio Fuels for Transportation used in Fuel Cells and Conventional Technologies
under European conditions. The other, carried out by the French authority for environment and
energy, ADEME, together with Ecobilan, PricewaterhouseCoopers and DIREM, was called
Energy and Greenhouse Gas Balances of Bio Fuels Production Chains in France, and was
published in December 2002 (ADEME, 2002).
Sweden
Several LCAs have been carried out on the production and use of alternative vehicle fuels under
Swedish conditions, focusing on (or at least including) ethanol. In 1997, Magnus Blinge, of
Chalmers University of Technology, Gothenburg, carried out an LCA on different alternative
fuels following a request from the Swedish Transportation Research Board (Blinge et al., 1997).
This LCA, despite being relatively old, and not based on the latest data, information and
technology, still seems to be one of the best. It also covers non-regulated emissions. However, in
addition to being old, the results are primarily valid for Swedish conditions and not applicable to
the rest of Europe.
Another Swedish LCA that is specifically related to Swedish conditions is Agro Ethanols LCA
on the annual production of 50 000 m3 grain-based ethanol at the Agroetanol AB production
plant outside Norrkping. However, this LCA is focused solely on emissions of greenhouse
gases.
Other countries
Comparison of Transport fuels carried out by Beer and co-workers, the Australian
Greenhouse Office, The University of Melbourne (Beer et al., 2002).
Both of these LCAs are concerned with emissions of greenhouse gases and energy efficiency.
In addition, of course, LCAs have been carried out in other parts of the world, including China
(see below), India and Canada, but these have also focused mainly on emissions of greenhouse
gases with clear country-specific framework conditions, largely concerning the raw material.
10.3. Discussion
LCAs on the production, distribution and use of a vehicle fuel such as ethanol is often carried
out with country-specific framework conditions such as the sources and nature of the raw
material used to produce the fuel.
In most of the LCAs carried out to date on ethanol, neat ethanol for FFVs, 85 % ethanol and
15 % gasoline (E85), or ethanol-gasoline blends with low alcohol contents (5 % or occasionally
10 %; E5 or E10), are compared with neat gasoline. LCAs comparing blends with higher ethanol
contents, up to 20 or 30 % for example, are scarce.
Concerning the parameters studied in LCAs on the production, distribution and use of alternative
fuels/ethanol most have mainly focused on their climatic impact, especially emissions of carbon
dioxide and, occasionally, emissions of other greenhouse gases.
LCAs that consider regulated emissions are not common. The investigation by Blinge and co-
workers. (Blinge et al. 1997) is an exception, as is the Economics, Environment and Energy
Life Cycle Assessment of Automobiles Fueled by Bio-ethanol Blends in China, study by
Zhiyuan Hu and coworkers of Shanghai Jiao Tong University, China (Zhiyuan et al., 2004).
Furthermore, there have been few studies related to non-regulated emissions, and they are
difficult to find.
To summarize:
Although it is difficult to find the kind of LCA sought for and described in chapter 4.9.2, there is
at least one example of an LCA in which different types of gasoline were examined that could be
used as a procedural model for LCAs considering ethanol blends of RVP-adjusted and
unadjusted gasoline. The study was performed by Teresa Mata and co-workers of the University
of Porto, Portugal (Mata et al., 2003).
In the cited LCA, a life cycle analysis or assessment compared the potential environmental
impact of various gasoline blends that all met the same octane and vapour pressure
specifications. The system boundaries were set to include the refinery process, but exclude the
fuels use in a vehicle. For the chain from refinery to refueling of the vehicle seven stages were
defined and emission factors were calculated. Several types of gasoline that differed in their
specifications, but fulfilled standard octane number and RVP requirements, were compared in
terms of their potential environmental impact due to hydrocarbon emissions. The results of the
Page 84
study showed that since the blending components such as alkylate, cracked gasoline and
reformate have different octane numbers and vapour pressures, as well as differing in their
potential environmental impact, some gasoline blends are less detrimental for the environment
than others.
10.4. Findings
LCAs are difficult to find, worldwide or in Europe/Sweden that:
Have relevant framework conditions (modern engine systems, relevant fuel specifications
etc).
Consider the emissions not only of greenhouse gases but also of regulated and other non-
regulated constituents.
Compare regulated and non-regulated emissions (tailpipe emissions and evaporative
emissions) from a well-to-wheel perspective and their impact on the environment and
health from the production, distribution and use of gasoline with blends of RVP-adjusted
and unadjusted gasoline with 5 to 20 % ethanol.
A number of LCA:s have been carried out (mainly but not solely in the USA), but they have
applied more or less country-specific framework conditions with respect to the raw material used
for the ethanol production and specific gasoline specifications. They have also often focused (as
have most LCAs on the production, distribution and use of other alternative fuels) on the
emissions of greenhouse gases. From a European or Swedish perspective, LCAs carried out in
the USA or other parts of the world probably differ too much in their framework conditions to
provide a relevant basis for predictions under European/Swedish conditions.
To obtain information concerning whether (and if so, how) evaporative emissions and tailpipe
emissions might differ, from a well-to-wheel perspective, it seems necessary to carry out a new
LCA. Such an LCA could be structured similarly to that of Mata and co-workers (Mata et al.
2003), but with specific Swedish framework conditions.
If possible the LCA should include some kind of evaluation concerning the extent to which
different types of canisters could reduce the evaporative emissions from the vehicle and the
extent to which they would reduce the total volume of emissions.
Page 85
American car manufacturers have agreed to accept the use of blends with up to 10 % ethanol in
gasoline in their vehicles, without changing the warranty conditions. Many car manufacturers in
Europe and Japan sell vehicles in North America. Therefore, the design of currently available on
the European market, including Sweden, may allow the use of blends with at least 10 % ethanol
in gasoline.
A conclusion which can be drawn from the literature study is that newer models of vehicles are
more tolerant to changes in the composition of fuels such as a blend of 10 % ethanol in gasoline.
One problem that may apply to older vehicles is that alcohol tends to react with rubber. Owners
of such cars should refrain from using alcohol blended gasoline. Many authors of papers dealing
with alcohol fuels claim that the main cause of reported problems of using alcohols, especially
methanol, as a fuel for vehicles is that they have often been used in older cars. Further studies of
the vehicle fleet in Sweden may be needed to generate information about the implications of
using gasoline blends with up to 10 % ethanol for older cars which may be less tolerant to them
than newer models.
It is difficult to apply data from investigations carried out in other countries directly to Swedish
conditions since many variables differ from country to country, especially in terms of the fleet of
vehicles, the fuels, emission regulations and certain other factors. In addition, some
investigations evaluated in the present report have given contradictory results. However, updated
exhaust emissions (of both regulated and unregulated compounds) from neat gasoline and
ethanol gasoline blends should be characterized as described in the present report in tests at
both ambient temperature of 20 C (in accordance with current certification procedures) and
lower temperatures, e.g. -20 C.
Start ability and drivability at low ambient temperatures such as -20 C are important variables
to investigate due to the harshness of Swedish winters.
Methods for sampling and analysing exhaust emissions have not been adequately developed for
alternative fuels in general since traditional sampling and analysis methods were developed for
characterizing emissions from traditional fuels such as gasoline and diesel oil. There is a need to
develop and update methods for sampling and analysing vehicle exhaust emissions generated
from the combustion of alternative fuels.
The percentage of ethanol in gasoline blends evaluated should primarily be in the range of 10 %
as this could be the first upper target. Higher than 10 % contents of ethanol in gasoline are
expected to be of more importance after the year 2010.
Investigations and evaluations conducted for the American Lung Association of Minnesota
discussed in the present report have shown that the sulphur content of the fuel has a considerable
impact on both regulated and unregulated emissions. Therefore, data on emissions generated
using fuels with relatively high sulphur contents cannot be directly applied to Swedish
Page 86
conditions, since Swedish gasoline standards have allowed only 10 ppm sulphur since January
2005.
There is a need for updated emission factors for both regulated and unregulated exhaust
emissions from modern vehicles running on neat gasoline and ethanol-blended gasoline.
Besides measurements obtained following standard procedures, tests at low ambient
temperatures (-20 C) should also be included. Emission factors are needed for emission
inventories, which provide a basis for evaluating health and environmental effects.
Local fire stations should be informed about the presence and location of ethanol blended
gasoline filling stations to ensure the use of ethanol-resistant foam in the case of fire. This is
also an important issue to consider if fires occur in vehicles running on these fuels, since their
fuel tanks will contain ethanol/gasoline blends.
Increasing RVP may increase the VOC emissions, depending on the type of evaporative
emission control system used. There is a need for updated emission factors for evaporative
emissions from modern vehicles running on ethanol mixed with gasoline (5 - 10 %),
comparing emissions from the use of both RVP-adjusted and non RVP-adjusted gasoline. In
addition to measurements of evaporative emissions according to the standard method, emitted
hydrocarbons should be analysed to generate data for use in further evaluations. With such data
it will be possible to: (i) evaluate whether the evaporative emission standards are met and (ii) to
evaluate whether the emitted hydrocarbons will have an impact on the environment and health
in Sweden.
Increasing the RVP of a fuel may affect the drivability of the vehicle at high ambient
temperatures, since it may cause vapour lock and may also affect hot engine starts. If a gasoline
with an RVP at or close to the upper limit is mixed (commingled) with a gasoline containing
ethanol, the upper limit allowed for RVP could be exceeded.
The chemical compound in gasoline that contributes most to the RVP is butane. By decreasing or
increasing the butane content in gasoline the RVP can be reduced or increased, respectively.
Very few data have been found on evaporative emissions from Brazil regarding vehicles running
on neat ethanol or ethanol gasoline blends. However, since the lowest ethanol content of
Brazilian fuel blends is 20%, the evaporative emissions from the vehicles are limited.
Page 87
The search for literature on the effects of ethanol blends on regulated emissions showed that very
few data are available on emissions measured when using a blend of 10 % ethanol in gasoline.
Since nearly all vehicles in Sweden are equipped with a three-way catalyst exhaust gas after-
treatment system, data obtained in tests with such vehicles were of particular interest. The
emission data found and discussed in this report are from investigations in Australia, Canada,
England (UK), Sweden and the USA. One main drawback when collecting and studying data
from different investigations is that the number of vehicles tested is often limited.
The Orbital Engine Company has carried out a series of investigations on vehicles fuelled with
ethanol blended gasoline initiated by Environment Australia. Most of these investigations have
focused on a blend of 20 % ethanol in two or three grades of gasoline produced for the
Australian market. However, the investigations carried out by Orbital have been of great value
for this project since a broad spectrum of emission data are presented in the reports and they
discuss many aspects such as fuels, vehicles, catalysts, vehicle performance and wear. Data on
regulated, unregulated and evaporative emissions have been gathered and presented. There is
only one main drawback with the data from the extensive tests involved; the fuel (gasoline) used
for the tests with blended fuel and the neat gasoline had sulphur contents of up to 150 and 500
ppm, respectively, according to available information.
Investigations and evaluations conducted for the American Lung Association of Minnesota have
shown that sulphur in the fuel has a considerable impact on both regulated and unregulated
emissions.
A report from Environment Canada compares emission data from tests on ethanol blended
gasoline and neat gasoline. The ethanol contents in the blended fuels were 10, 15 or 20 %.
When comparing data from the different ethanol blends with data from tests on neat gasoline no
dramatic differences were seen.
Beside the investigations carried out in Australia the emission tests carried out at Harwell,
England, on five vehicles fuelled with a blend of 10 % ethanol in gasoline are of some interest,
firstly because a 10 % ethanol in gasoline blend was used, and secondly because the vehicles
used were models that are also used in Sweden. Generally the emission levels were low and no
significant differences in NOx emissions were found. Emissions of CO and PM were
significantly reduced when the use of blended was compared with neat gasoline.
Only limited data have been found concerning evaporative emissions. However, in some relevant
investigations it has been noticed that emissions via permeation have increased when using
ethanol blended gasoline. This has been especially significant in the USA after the strengthening
of the evaporative emission regulations, since new measures had to be introduced to comply with
the standards. One of these measures was to shift to the use of steel tanks for the fuel and another
was to use new and more reliable tubes in the fuel lines.
In a paper from the Transportation Office of Energy Efficiency in Canada it is said that blending
10 % ethanol in gasoline (E10) will result in the energy content of the fuel being 3 % lower than
the energy content of the base gasoline used for blending. Since this decrease in energy content
will be partly compensated by improved combustion efficiency of the ethanol blended fuel the
Page 88
overall increase in fuel consumption when using E10 will be only 2 %. In comparison,
increasing the speed of the vehicle from 100 km/h to 120 km/h increases fuel consumption by 20
%.
In the present report it has been emphasised that the method used for measuring hydrocarbons
(HC) in the exhaust from motor vehicles using an alcohol blended fuel are not satisfactory from
a legal perspective since the emissions of unburned alcohol are included in the readings of the
instrument (FID). Therefore there is an urgent need to develop an analytical method that can
determine HC emissions which are not affected by other non HC compounds present in the
exhaust.
For regulated emissions the recommendation is that a broader more detailed study of the
emissions, especially in low temperature conditions should be carried out, and tests including
evaporative emissions at low temperatures. The resulting figures could probably be used to
evaluate the environmental and health effects of using ethanol blends.
The emissions of aldehydes (especially acetaldehyde and, to a lesser degree, formaldehyde) and
alcohols from vehicles running on alcohol/gasoline blends are expected to increase. There is no
standard validated method for determining aldehyde and alcohol emissions from motor vehicles,
so (an) appropriate method(s) must be developed (see also the conclusion and recommendation
about HC emissions discussed above).
Due to the gasoline dilution effect of adding ethanol the emissions of benzene, toluene, ethyl
benzene and xylene from blends are lower than those from neat gasoline, which is beneficial
from environmental and health perspectives.
From the literature it is not clear whether there is risk that emissions of quinones from gasoline-
fuelled vehicles will increase as the ethanol content increases in a gasoline/ethanol blend. Thus,
it is recommended that a pilot investigation should be initiated to investigate if quinone
emissions are related to the ethanol content of ethanol-gasoline blends.
There is clear, general interest from a health perspective in particle emissions from vehicles in
the literature examined. It is recommended that particles emitted from vehicles fuelled with
ethanol/gasoline blends should be thoroughly characterised chemically, and this
characterisation should include measurements of their sizes and numbers.
Extensive tests and studies carried out in Australia have found that using an ethanol-gasoline
blend with 20 % ethanol increased the wear and other damage to the fuel system of the vehicles,
especially at high accumulated mileages. It has not been possible to verify whether this is linked
to the fact that the sulphur content of gasoline in Australia is relatively high, since this
possibility was not considered in the reports from Australia. In Sweden the sulphur content in
gasoline has been reduced to a maximum of 10 ppm. A high level of sulphur in the ethanol
blended fuel may affect the engine fuel system since there is a small amount of acetic acid in
neat ethanol (a maximum of 0.0025 % by weight, according to AMSE 1114).
In the light of the findings concerning the engine wear and other observed problems associated
with the use of 20 % ethanol in gasoline the Australian government has decided to limit the
amount of ethanol in gasoline to 10 %. Representatives of the oil industry and car owners also
favoured the 10 % limit.
In contrast to the experiences in Australia, a study conducted in Minnesota (USA), in which two
blends of ethanol in gasoline (one with 10 % and the other with 30 % ethanol) were examined,
found that the ethanol caused no serious wear problems. The project included both laboratory
tests and field trials. Unfortunately, however, no emission data are available from these tests.
In the 1970s and 1980s a quite extensive program was carried out in Sweden to study the effects
of using alcohol blends in gasoline in motor vehicles. No such program has been established to
collect experience and data regarding the use of ethanol blended gasoline since then. Therefore
very few Swedish data are available.
It can be concluded from the literature survey that LCA data with relevant framework conditions
for the use of bio ethanol and bio ethanol mixed gasoline in a Swedish/European perspective are
limited or lacking.
A new LCA based on Swedish framework conditions should focus on ethanol-gasoline blends
with alcohol contents ranging from 5 to 20 %. Furthermore, as well as greenhouse gases
(carbon dioxide, methane and nitrous oxide), regulated and other unregulated emissions should
be included in it. The impact on the environment and health, of the production, distribution and
use of the blends should be considered. Such an LCA could be structured in a similar way to the
LCA published by Mata et al. (2003), but using specific Swedish framework conditions.
Furthermore, both RVP-adjusted and non RVP-adjusted gasoline should be included.
Page 90
12. REFERENCES
Aakko P. and Nylund N-O. Particle Emissions at Moderate and Cold Temperatures Using
Different Fuels. SAE Technical Paper Series SAE 2003-01-3285, Special publication SP-1809,
Warrendale US, (1977).
Acharya, G.K., Malhotra, N.R., Raje, N.R. and Bratnagar, A.K. Indian experiences with the use
of Ethanol Gasoline and Ethanol Diesel Blends. ISAF XIV International Symposium on
Aclcohol fuels: The Rule of Alcohol Fuels in Meeting the Energy, Environmental and
Economic Needs of the 21st Century, (2002).
Alsberg T., Stenberg U., Westerholm R., Strandell M., Rannug U., Sundvall A., Romert L.,
Bernsson V., Petterson B., Toftgrd R., Franzen B. Jansson M., Gustafsson J-. Egebck K-E.
and Tejle G. (1985) Chemical and biological characterization of organic material from gasoline
exhaust particles. Environmental Science & Technology, Vol. 19, 43-50, (1985).
Almn J., Ludykar D. and Westerholm R. Unregulated Emission Factors for Light Duty Vehicles
at Different Driving Patterns and Temperatures. AB Svensk Bilprovning, Motortestcenter,
Rapport MTC, Haninge (1997).
American Lung Association of Minnesota. Twin Cities Gasoline Survey and Comparison
Summer 2003. American Lung Association of Minnesota B Clean Air Fuels Education Alliance
EXECUTIVE SUMMARY (2003).
http://www.cleanairchoice.org/outdoor/pdf/ExeSum_Summer 2003.pdf.
Andress, D. (2000). Air Quality and GHG Emissions Associated With Using Ethanol in Gasoline
Blends. Prepared by: David Andress & Associates, Inc.Prepared for:Oak Ridge National
Laboratory / UT-Battelle, Inc. and Office of Fuels Development Office of Transportation
Technologies Energy Efficiency and Renewable Energy U.S. Department of Energy. Subcontract
11X-SY838, Under Contract DE-AC05-96OR22464. http://egov.oregon.gov/energy
renew/Biomass/docs/FORUM/Andress2000.pdf
Apace Research Ltd. Intensive Field Trial of Ethanol/Petrol Blend in Vehicles, Volume 2,
Appendices A-I Prepared by Apace Research Ltd, (1998).
Armstrong S. Ethanol. Brief Report on its Use in Gasoline. Cambridge Environmental Inc. 58
Charles St. Cambridge, MA 02141 US, (1999).
Augin, K. and Graham, L. The Evaluation of Ethanol Gasoline Blends on Vehicle Exhaust and
Evapoorative Emissions, Phase 1 Updated 10/25/02. ERMD Report # 00-64. Environmental
Technology Centre, Emissions Research and Measurement Division, Environment Canada,
(2004).
Page 91
Aulich T. and Allen L. Twin Cities Gasoline Survey and Comparison-Summer 2002. Final
Report Prepared for: The Clean Air Fuels Education Alliance. Submitted to: Tim R. Gerlach ,
The American Lung Association of Minnesota. Energy & Environmental, Research Center,
University of North Dakota, (2002).http://www.cleanairchoice.org/outdoor/pdf/gas%20survey-
summer02.pdf
Australia Automobile Association, AAA, Response to the Environment Australia issues paper.
Setting the Ethanol Limit in Gasoline, (2002).
http://www.aaa.asn.au/submis/2002/Ethanol_March02.pdf
Australian Government. Department of the Environment and Heritage. Setting the Ethanol Limit
in Petrol, (2002).
http://www.deh.gov.au/atmosphere/ethanol/publications/ethanollimit/background.html
Australian Government. A Literature Review Based Assessment on the Impacts of a 10% and
20% Ethanol Gasoline fuel Blend on Non-Automotive Engines, (2002).
Beer T., Grant T., Brown R., Edwards J., Nelson P., Watson H. and Williams D. Comparison of
Transport fuels, Australian Greenhouse Office, The University of Melborne, (2002).
Beijing Time. China Promotes Ethanol-Based Fuel in Five Cities: Zhengzhou, Luoyang and
Nanyang in central China's Henan province, and Harbin and Zhaodong in Heilongjiang province,
northeast China, (2002a). http://english.people.com.cn/200206/17/eng20020617_98009.shtml.
Blinge M., Arns P-O., Bckstrm S., Furnander .och Hovelius K. Livscykel Analys av
Drivmedel. Chalmers University of Technology. Rapport fr Alternativbrnsleutredningen,
(1997).
Bonnema G., Guse G., Senecal N., Gupta R., Jones B. and Ready K. Use of Mid-range
Ethanol/Gasoline Blends in Unmodified Passenger Cars and Light Duty Trucks. MnCAR/MSU
106. Minnesota Center for Automotive Research, Minnesota State University, (2004).
Bostrm C-E., Gerde P., Hanberg A., Jernstrm B., Johansson C., Kyrklund T., Rannug A.,
Trnqvist M., Westerholm R. and Victorin K. (2001) Cancer risk assessment, indicators and
Page 92
guidelines for polycyclic aromatic hydrocarbons (PAH) in the ambient air. Environmental Health
& Perspectives, Vol. 110, Supplement 3, pp 451 488, (2002).
Camner P., Cothgreave I., Ewetz L., Gustavsson P., Hansson H.-C., Kyrklund T., Ljungquist S.,
Pershagen G., Victorin K. and Westerholm R. Particles in the ambient air as a risk factor for lung
cancer. Swedish Environmental Protection Agency (SEPA) Report 4804, Stockholm, Sweden,
(1997).
CARB, California Air Resources Board, California Environmental Protection Agency, (1989).
CARB, California Air Resources Board, California Environmental Protection Agency, Cleaner
burning gasoline: An assessment of its impact on ozone air quality in California, January,
(1998).
Prakash C. Use of Higher than 10 volume percent Ethanol/Gasoline Blends in Gasoline Powered
Vehicles. Prepared by: Motor Vehicle Emissions & Fuels Consultant forTransportation Systems
Branch Air Pollution Prevention Directorate, Environment Canada, (1998).
Cheung, A., Flanders, K. and Walker, P. Ethanol-Blended Fuel policies: When will the cheap
tricks end? (2003).
Colorado Department of Public Health and Environment. The Impact of a 10% Ethanol Blended
Fuel on the Exhaust Emissions of Tier 0 and 1 Light Duty Gasoline Vehicles at 35F, (1999).
Cho A., Di Stefano E., You Y., Rodriquez C., Schmitz D., Kumagai Y., Migel A., Eiguren-
Fernandez A., Kobayashi T., Avol E. and Froines J. Determination of Four Quinones in Diesel
Exhaust Particles, SRM 1649a, and Atmospheric PM2.5. Aerosol Science & Technology, Vol.
38, 68-81, (2004).
Page 93
Coln M., Pleil J., Hartlage T., Guardani M. and Martins M. Survey of volatile organic
compounds associated with automotive emissions in the urban airshed of Sao Paulo, Brazil.
Atmospheric Environment, Vol. 35, 4017-4031, (2001).
Deeb R., Sharp J., Stocking A., McDonalds S., West K., Laugier M., Alvarez J. Kavanaugh M.
and Alvarez-Cohen L. Impact of Ethanol on Benzene Plume Lenghts: Microbial and Modeling
Studies. Journal of Environmental Engineering, Vol 128, No. 9, 868- 875, (2002).
Department of the Environment and Heritage. Fuel Quality Standards Act 2000. Clean Fuels
Bulletin, Australian Government Department of the Environment and Heritage December
(2003). http://www.deh.gov.au/atmosphere/cleaner-fuels/publications/pubs/december-2003.pdf
Department of the Environment and Heritage. Labelling Requirements from March 2004. Fuel
Quality Standards Act 2000, (2004a). Department of the Environment and Heritage.
http://www.deh.gov.au/atmosphere/ethanol/pubs/labelling.pdf.
Department of the Environment and Heritage. Fuel Quality Standards Act 2000, Clean Fuels
Bulletin, Amendments to the Fuel Quality Standards Act 2000 Fuel Quality Standards. (2004)
http://www.deh.gov.au/atmosphere/cleaner-fuels/publications/pubs/december-2004.pdf
Dietz W. Response Factors for Gas Chromatograpic Analysis. Journal of Gas Chromatograpy,
Vol. 5, 68-71, (1967).
Egebck, K.-E., Bertilsson B.-M. Chemical and Biological Characterizon of Exhaust Emissions
from Vehicles Fueled with Gasoline, LPG and Diesel, SNV PM 1635, ISBN 91-7590-115-3,
(1983).
EIA (2002) Potential Supply Impacts of Removal of 1-Pound RVP Waiver. US Energy
Information Administration section.
http://www.eia.doe.gov/oiaf/servicerpt/fuel/pdf/question7.pdf
Embassy of Brazil. The Brazilian National Programme of Fuel Alcohol (PROALCOOL), (2002).
http://brazilembassyinindia.com/proalcool.htm.
Page 94
Environment Australia. Setting the Ethanol Limit in Petrol. An Issues Paper, (2002).
http://www.deh.gov.au/atmosphere/ethanol/publications/ethanol-limit/pubs/ethanoljan02.pdf.
Environment Australia. A Literature Review Based Assessment on the Impacts of a 20% Ethanol
Gasoline Fuel Blend on the Australian Vehicle Fleet, (2002a).
Ethanol Curriculum Online TEFL Module 5 (2005). Module 5, Ethanol: The Fuel.
http://www.nwicc.com/pages/continuing/business/ethanol/Module5.htm
Ethanol Industry in Thailand. Outlook of Raw Materials for Ethanol Industry in Thailand,
(2004).
Federal Government. Federal government Set 10 Per Cent Ethanol Limit, (2003).
http://www.deh.gov.au/minister/env/2003/mr11apr03.html.
Federal Government. Fuel Quality Standards Amendment Bill 2003. Bills Digest No. 30 2003
04, (2003a). http://www.aph.gov.au/library/pubs/bd/2003-04/04bd030.pdf. [04bd030]
The French Authority for Environment and Energy, ADEME, Ecobilan, Pricewaterhouse
Coopers and DIREM, Energy and greenhouse gas balances of bio fuels production chains in
France, (2002).
Furey, R. and King J. Evaporative and Exhaust Emissions from Cars Fueled with Gasoline
Containing Ethanol or Methyl tert-Bytyl Ether. General Motors Research Laboratories. SAE
paper 800261, February, (1980).
Furey R. and Perry K. Vapour Pressures of Mixtures of Gasolines and Gasoline-Alcohol Blends.
General Motors Research Laboratories. SAE paper 861557, (1986).
Furey R. and Perry K. Volatility Characteristics of Blends of Gasoline with Ethyl Tertiary-Bytyl
Ether (ETBE). General Motors Research Laboratories. SAE paper 901114, (1990).
Furey R. and Jackson M. Exhaust and Evaporative Emissions from a Brazilian Chevrolet fuelled
with Etanol-Gasoline Blends. SAE Technical Paper Series SAE 779008, Warrendale US, (1977).
Gaffney, Jeffrey S.; Marley, Nancy A.; Martin, Randal S.; Dixon, Roy W.; Reyes, Luis G.; Popp,
Carl J. Potential Air Quality Effects of Using Ethanol-Gasoline Fuel Blends: A Field Study in
Albuquerque, New Mexico. Environmental Science and Technology, Vol. 31, 3053-3061,
(1997).
General Motors, Argone National Laboratory, BP, ExxonMobil, Shell and TotalFinal Elf. Well
to Wheel Energy use and Greenhouse Gas Emissions of Advanced Fuel/vehicle systems, (2002).
Page 95
Grosjean D., Grosjean E. and Moreira L. Speciated Ambient Carbonyls in Rio de Janeiro, Brazil,
Environmental Science and Technology, Vol. 36, 1389-1395, (2002).
Hammel-Smith C., Fang J, Powders M. and Aabakken J. Issues Associated with the Use of
Higher Ethanol Blends (E17-E24). National Renewable Energy, (2002).
http://www.nrel.gov/docs/fy03osti/32206.pdf.
Harly R. and Coulter-Burke S. Relating Liquid Fuel and Headspace Vapour Composition for
California Reformulated Gasoline Samples Containing Ethanol. Environmental Science and
Technology, Vol. 34, 4088-4094, (2000).
Hart Downstream Energy Services. Setting a Quality Standard for Fuel Ethanol. Report
Presented to: Department of the Environment and Heritage Submitted by:International Fuel
Quality Center, Hart Downstream Energy Services, (2004).
http://www.deh.gov.au/atmosphere/ethanol/publications/pubs/ifqc-report.pdf.
He B., Wang J., Hao J., Yan X. and Xiao J. A Study on Emissions Characteristics of an EFI
Engine with Ethanol Blended Gasoline Fuels. Atmospheric Environment, Vol. 37, 949-957,
(2003).
Heinrich U., Fuhst R., Dasenbrock C., Muhle H., Koch W. and Mohr U. Long term inhalation
exposure of rats and mice to diesel exhaust, carbon black and titanium dioxide. In Proceedings of
the 5th International Inhalation Symposium on Toxic and Carcinogenic Effects of Solid Particles
in the Respiratory Tract, March 1993, Hannover, Germany, (1993).
Heinrich U., Fahst R., Rittinghausen S., Creutzenberg O., Bellmann B., Koch W. and Leven K.
Chronic Inhalation Exposure of Wistar Rats and two Different Strains of Mice to Diesel Engine
Exhaust, Carbon Black and Titanium dioxide. Inhalation Toxicology 7, 533-556, (1995).
Hsieh W., Chen R., Wu T. and Lin T. Engine Performance and Pollutant Emission of a SI
Engine using Ethanol Gasoline Blended Fuels. Atmospheric Environment, Vol. 36, 403-410,
(2002).
Isaias de Carvalho M., Leal M. and da Silva J. Assessment of greenhouse gas emissions in the
production and use of ethanol in Brazil, Government of the state of Sao Paulo, (2004).
Page 96
Johnsson A. and Bertilsson B-M. Formation of Methyl Nitrite in Engines Fueled with
Gasoline/Methanol and Methanol/Diesel. Environmental Science & Technology, Vol. 16, 106-
110, (1982).
Joseph, H. Junior Vehicular Ethanol Fuel. Technical & Exhaust Emission Experience in
Brazil. Brazilian Automobile Manufacturers Association, (1998).
Kemp I. Fuel Quality Information Standard (Ethanol) Determination 2003. 2003122302, (2003).
http://scaleplus.law.gov.au/html/instruments/0/33/pdf/2003122302.pdf
KFB. Low Contents of Alcohols in Gasoline. (In Swedish: Inblandning av lga halter av
alcoholer i bensin). Report for The Swedish Transport and Communications Research Board
(KFB). Swedish version KFB-Meddelande 1999:4, (1999).
Krause K and Korotney D. Water Phase Separation in Oxygenated Gasoline. Corrected version
of Kevin Krause memo. (1995). http://www.epa.gov/otaq/regs/fuels/rfg/waterphs.pdf
Launder K. From Promise to Purpose: Opportunities and Constraints for Ethanol Based
Transportation Fuels- A Plan B Research Report, Submitted to Michigan State University In
partial fulfilment of the requirements for the degree of Masters of Science, Department of
Resource Development, (2001).
Laveskog, A. and Egebck, K.-E. Addition of Small Amounts of Alcohol to Gasoline, Report in
English prepared for MTC (Motortestcenter, Sweden, (1999).
Lonneman W. Comparison of hydrocarbon composition in Los Angeles for the year 1968 and
1997. Preceedings of the EPA/AWMA Symposium on Measurement of Toxic and Related Air
Pollutants, Carry, NC Air & Waste Management Association, Sewickley, PA, 356-365, (1998).
Lovanh N., Hunt C. Alvarez P. Effect of ethanol on BTEX biodegradation kinetics: aerobic
continuous culture experiments. Water Research 36(15), 3739-3746, (2002).
Lu-Karlsson H. Characterisation of the Evaporative Emissions from E85, E10, Gaoline and
Reference Gasoline Fuels, MTC report, (1999).
McArragher J. Minimum RVP Limits in CEN Gasoline Specification. Shell International, UK,
20/6, (1996).
Market Barriers (2004). Testing Gasoline Containing 20% Ethanol (E20) - Phase 2B Final
Report: Market Barriers to the Uptake of Bio fuels Study. Department of the Environment and
Heritage
Mata T., Smith R., Young D. and Costa C. Life Cycle Assessment of Gasoline Blending
Options, University of Porto, Portugal, (2003).
Ministery for the Environment and Heritage. Federal Government to Set 10 Per Cent Ethanol
Limit. (2003;2004). http://www.deh.gov.au/minister/env/2003/mr11apr03.html.
Minister for the Environment and Heritage. Fuel Quality Standards Act 2000 labelling
requirements from 1 March 2004, (2004)
http://www.deh.gov.au/atmosphere/ethanol/publications/labelling.html
Mishra S. India Ethanol Coalition Joining hands for the future fuel, (2002).
http://www.cleantechindia.com/eicnew/India.htm
Nan L., Best G., Coelho de Carvalho Neto C..Integrated energy systems in China - The cold
Northeastern region experience. Food and Agriculture Organization of the United Nations,
Rome, (1994). http://www.fao.org/docrep/T4470E/T4470E00.htm.
National Research Council. Diesel Cars Benefits, Risk and Public Policy. National Academy
Press: Washington DC, (1982).
Natural Heritage Trust. Setting National Fuel Quality Standards, Proposed Standards for Fuel
Parameters (Gasoline and Diesel) Revised Commonwealth Position, (2000).
http://www.deh.gov.au/atmosphere/cleaner-fuels/publications/pubs/fuel0900.pdf
National Ethanol Vehicle Coalition. Why do ethanol vehicles require a certain type of motor oil?
Does this refer to dedicated ethanol vehicles only? (2004).
http://www.e85fuel.com/faqs/motoroil.htm
Nitsche M. Respiratory Health Effects of nitrogen dioxide Exposure and Current Guidelines.
International Journal of Environmental Health Research, Vol. 9, 39 53, (1999).
NREL. Keeping the Heat on Cold-Start Emissions NREL's Insulated Catalytic Converter Slashes
Automotive Air Pollution. Technology Brief, (1996).
http://www.nrel.gov/vehiclesandfuels/energystorage/pdfs/techbr.pdf
NREL. Issues Associated with the Use of Higher Ethanol Blends (E17-E24), (2002).
NRMA. Keep Strict Limit on Ethanol: Ethanol levels in gasoline must not exceed 10 per cent
until research proves there is no negative effect on engines or driving quality, the National Roads
and Motorists' Association Limited said today, (2002).
http://www.mynrma.com.au/releases2002_021011_1.asp
NRMA. NRMA to Federal Government: Act Now on Ethanol. The Chief Executive of NRMA
Motoring & Services, Mr Rob Carter, said he was angered tonight by Federal Cabinet's refusal to
control ethanol levels in petrol, the central issue threatening cars belonging to NRMA's two
million members in NSW, (2002a). http://www.mynrma.com.au/releases2002_021218_1.asp
Nyns, OK and SDAB. Distribution of Blend Fuels. Final report on trials with distribution and
storage of methanol-blended leadfree gasoline, (1983)
Oak Ridge National Laboratory. Fuel Specifications and Fuel Properties issues and Their
Potental Impact on the Use of Ethanol as a Transportation Fuel. Phase III Project Deliverable
Report Ethanol Project, (2002). http://biodiesel.pl/uploads/media/FUEL SPECIFICATIONS
AND FUEL PROPERTY ISSUES AND THEIR POTENTIAL IMPACT ON THE USE OF
ETHANOL AS A TRANSPORTATION FUEL.pdf
Office of Legislative Drafting. Attorney-Generals Department Fuel Quality Standards Act 2000
Act No. 153 of 2000 as amended. This compilation was prepared taking into account
amendments up to Act No. 42 of 2004, (2004).
http://scaleplus.law.gov.au/html/pasteact/3/3418/pdf/FuelQualityStand2000.pdf
Office of Oil and Gas of the Energy Information (2002). Supply Impacts of an MTBE Ban
http://www.eia.doe.gov/oiaf/servicerpt/fuel/pdf/question1.pdf
Orbital Engine Company. A Literature Review Based Assessment on the Impacts of a 20%
Ethanol Gasoline Fuel Blend on the Australian Vehicle Fleet. Report to Environment Australia,
November, (2002).
Orbital Engine Company. Market Barriers to the Uptake of Bio fuels Study. A Testing Based
Assessment to Determine Impacts of a 20% Ethanol Gasoline Fuel Blend on the Australian
Passenger Vehicle Fleet Report to Environment Australia March 2003, (2003).
http://www.deh.gov.au/atmosphere/ethanol/publications/2000hours-vehicle-fleet/pubs/2000-
hours-vehicles.pdf.
Orbital Engine Company. Market Barriers to the Uptake of Bio fuels Study Testing Gasoline
Containing 20% Ethanol (E20). Phase 2B Final Report to the Department of the Environment
and Heritage, (2004). http://www.deh.gov.au/atmosphere/ethanol/publications/biofuels-
2004/pubs/phase2b-final-report.pdf
Patyk A. and Reinhardt G. Life Cycle Analysis of Bio Fuels for Transportation Used in Fuel
Cells and Conventional Technologies under European Condition, Institute fr Energie und
Umwelts Forschung, Heidelberg, Germany, (2002).
Page 99
Pott, F. and Heinrich, U. Relative Significance of Different Hydrocarbons for the Carcinogenic
Potency of Emissions from Various Incomplete Combustion Processes. IARC Sci. Publ. 104:
288-297, Lyon, France, (1990).
Powers S., Alvarez P. and Rice D. Increased Use of Ethanol in Gasoline and Potential Ground
Water Impacts. Chapter 1, UCRL-AR-145380, (2001).
Prakash, C (1998) Use of Higher than 10 volume percent Ethanol/Gasoline Blends In Gasoline
Powered Vehicles, Motor Vehicle Emissions & Fuels Consultant, prepared for Transportation
Systems Branch Air Pollution Prevention Directorate Environment Canada.
Reading A., Norris, J., Feest A. and Payne E. Ethanol Emissions Testing, E&E/DDSE/02/021
Issue 3, AEAT Unclassified, (2002).
http://www.dft.gov.uk/stellent/groups/dft_roads/documents/page/dft_roads_032580.pdf
Renewable Fuels Association. Thailand Endorses Ethanol to Reduce Oil Use, Help Farmers,
(2001). http://www.ethanolrfa.org/ereports/er112001.html
Renewable Fuels Association. Fuel Ethanol Industry Guidelines Specifications, and Procedures.
RFA Publication # 960501 Revised May 2002, (2002). http://www.pei.org/FRD/rfa_pub.pdf
Schifter I., Vera M., Diaz L., Guzman E., Ramos F. and Lopez-Salinas E. Environmental
Implications on the Oxygenation of Gasoline with Ethanol in the Metropolitan Area of Mexico
City. Environmental Science & Technology, Vol. 35, 1893-1901, (2001).
Schuetzle D., Lee F. and Prater T. The Identification of Polynuclear Aromatic Hydrocarbon
(PAH) Derivatives in Mutagenic Fractions of Diesel Particulate Extracts. International Journal of
Environmental Analytical Chemistry, Vol 9, 93-144, (1981).
Schuetzle D., Siegl W., Jensen T., Dearth M., Kaiser E., Gorse R., Kreucher W. and Kulik E.
The relationship between Gasoline Composition and Vehicle Hydrocarbon Emissions. A Review
of Current Studies and Future Research Needs. Environmental Health & Perspectives, Vol. 102,
3-12, (1994).
SDAB (1983). Effects of Gasoline, M15, Methanol and Ethanol on Plants, Soil and Aquatic
Organisms (in Swedish). Swedish Motor Fuel Technology Co for the National Swedis Board for
Energy Source Development. Sekab (2004). http://www.sekab.se/word/Etanol99.pdf
Sjdin ., Westerholm R., Almn J. and De Serves C. Emissions of regulated and non-regulated
compounds from high- and low-emitting gasoline light-duty vehicles. Transport and Air
Pollution, 9:th Symposium Avignon, ISBN 2-85782-533-1.Vol 1. pp 135-142, (2000).
STU. Alcohols and alcohol blends as motor fuels. Vol. II A och Vol.II B, State-of-the-Art
report. Chapters 1-7 och Chapters 8-14, Prepared by The Swedish Motor Fuel Technology Co.
(SDAB), Information 640-1986. (1986).
STU (1987). Project M100 A Test With Methanol-Fuelled Vehicles in Sweden. Prepared by
SDAB. STU information no. 640-1987.
The Swedish National Energy Administration. Motor Alcohols in Sweden (in Swedish). Statens
Energiverk 1986:7, (1986).
The Gazette of India (2002). Ministry of Petroleum and Natural Gas Resolution New Delhi, 3RD
September, 2002 No. P-45018/28/2000-C, (2002).
http://www.ethanolindia.net/ethanol_govt.html
Trnqvist M., Rannug U., Jonsson A. and Ehrenberg L. Mutagenicity of methyl nitrite in
Salmonella typhimurium. Mutation Research. Vol. 117, 47-, (1983).
Page 101
Trnqvist M. Kautiainen A., Gatz R. and Ehrenberg L. Haemoglobin adducts in animals exposed
to gasoline and diesel exhaust. 1. Alkenes. Journal of Applied Toxicology. Vol. 8, 159-170,
(1988).
Trnqvist M., Segerbck D. and Ehrenberg L. The rad-equivalence approach for assessment
and evaluation of cancer risks, exemplified by studies of ethylene oxide and ethane. In: Human
carcinogen exposure, biomonitoring and risk assessment. Edited by: R. Garner, P. Farmer, G.
Steele and A.Wright. Oxford IRL Press. 141-155, (1991).
Trnqvist M. and Ehrenberg L. On Cancer Risk Estimation of Urban Air. Environmental Health
& Perspectives. Vol. 102, 173-182, (1994).
Uppal, J. (2002). Indian Ethanol Program The Road Ahead! Seminar on Opportunities &
Prospects for Use of Ethanol, India-Brazil Partnership. Winrock International India.
http://www.ficci.com/ficci/media-room/speeches-presentations/2002/aug/aug14-ethanol-jai.htm
US Commercial Service. The Significance of Developing Ethanol Gasoline and the Pilot Use of
Ethanol Gasoline. American Embassy, Beijing Vol. 2 No. 137, (2003).
http://www.buyusa.gov/china/en/ccb030703.html
US Department of Energy. Guidebook for Handling, Storing, & Dispensing Fuel Ethanol,
(2002). http://www.agriculture.state.ia.us/pdfs/ethguide.pdf.
US EPA. Cancer Risk from Outdoor Air Toxics. Vol., Washington DC, (1990).US EPA. Clean
Fuels: An Overview, EPA 400-F-92-008 US Environment Protection Agency, Office of Mobile
Sources, (1992). http://www.epa.gov/otaq/consumer/06-clean.pdf
US EPA (1996). Agency Correspondence Regarding RFG and Low RVP Gasoline
http://www.epa.gov/otaq/regs/fuels/rfg/api-let.pdf
US EPA (1999). Control of Air Pollution from New Motor Vehicles: Proposed Tier 2 Motor
Vehicle Emissions. Standards and Gasoline Sulfur Control Requirements.
http://www.epa.gov/otaq/regs/ld-hwy/tier-2/nprm/t2phila2.pdf
US Senate. A bill to amend the Clean Air Act to address problems concerning methyl tertiary
butyl ether, and for other purposes. Bill Summary & Status for the 107th Congress, (2001).
http://thomas.loc.gov/cgi-bin/bdquery/z?d107:SN00950:@@@D&summ2=0&
Wang M., Saricks C. and Santini D. Effects of Fuel Ethanol Use on Fuel-Cycle Energy and
Greenhouse Gas Emissions, Argonne National Laboratory, US, (1999).
Page 102
Wild D., King M., Gocke E. and Eckhardt K. Study of Artificial Flavoring Substances for
Mutagenicity in the Salmonella/microsome, Basc and micronucleus tests. Food and Chemical
Toxicology 21(6), 707-719, (1983).
World Association of Beet and Cane Growers (2001). THAILAND APPROVES PLAN TO
REDUCE TAX ON ETHANOL. http://www.ifap.org/wabcg/faxsheet0101.html.
WWFC, World Wide Fuel Charter. June 2002, European Automobile Manufacturers
Assocoation (ACEA), Alliance of Automobile Manufacturers, Engine Manufacturers
Association (EMA) and Japan Automobile Manufacturers Association (JAMA), (2002).
Xia T., Korge P., Weiss J., Li N., Venkatesen M. and Sioutas C. Quinones and Aromatic
Chemical Compounds in Particulate Matter Induce Mitochondrial Dysfunction: Implications for
Ultrafine Particle Toxicity. Environmental Health Perspectives, Vol. 112, No. 14, 1347-1358,
(2004).
Young M., Saricks C. and Wu M. Fuel-Cycle Fossil Energy Use and Greenhouse Gas Emissions
of Fuel Ethanol Produced from US Midwest Corm, Argonne National Laboratory, US (1997).
Page 103
13. ABBREVIATIONS
API American Petroleum Institute
B(a)P benzo(a)pyrene
BRON Blending Research Octane Number
BTEX benzene toluene, ethyl benzene, xylene
CARB California Air Resource Board
CH3CH2-OH Ethanol
CO Carbon monoxide
CONCAWE The Oil Companies European Association for Environment, Health and
Safety in Refining and Distribution
DME Dimethyl ether
DOE Department of Energy
EIA US Energy Information Administration
EFI Electronic Fuel Injection
Eth Ethanol
EU European Union
EUCAR European Council for Automotive Research and Development
Evap Evaporative
E85 15% conventional gasoline and 85% ethanol
14. APPENDIX 1
14.1. Emission tests in Canada
The following five tables, A1 to A5, list emission data generated during the tests carried out in
Canada and presented in section 7.2 of the main part of the report. As can be seen there is a
considerable variation in emission levels from vehicle to vehicle. It is particularly noteworthy
that the Grand Am and Silverado show higher levels of emissions than the other vehicles
Table A1. Emissions of CO when using neat gasoline and three different blends of ethanol.
Vehicle Grand Honda Insight Silverado Echo Civic
Am
US FTP cycle CO g/km CO g/km CO g/km CO g/km CO g/km
0 % Ethanol 0.920 0.205 0.963 0.280 0.305
10 % Ethanol 0.808 0.218 0.740 0.261 0.273
15 % Ethanol 0.709 0.186 0.938 0.224 0.205
20 % Ethanol 0.777 0.143 0.740 0.242 0.155
Table A2. Emissions of HC when using neat gasoline and three different blends of ethanol.
Vehicle Grand Honda Insight Silverado Echo Civic
Am
US FTP cycle HC g/km HC g/km HC g/km HC g/km HC g/km
0 % Ethanol 0.056 0.027 0.075 0.050 0.025
10 % Ethanol 0.068 0.027 0.068 0.050 0.021
15 % Ethanol 0.056 0.024 0.087 0.049 0.017
20 % Ethanol 0.060 0.024 0.068 0.051 0.021
Table A3. Emissions of NOx when using neat gasoline and three different blends of ethanol.
Vehicle Grand Honda Insight Silverado Echo Civic
Am
US FTP cycle NOx g/km NOx g/km NOx g/km NOx g/km NOx g/km
0 % Ethanol 0.118 0.026 0.068 0.087 0.030
10 % Ethanol 0.131 0.033 0.081 0.087 0.032
15 % Ethanol 0.124 0.049 0.068 0.068 0.039
20 % Ethanol 0.124 0.068 0.081 0.068 0.033
Table A4. Emissions of CO2 when using neat gasoline and three different blends of ethanol.
Vehicle Grand Honda Insight Silverado Echo Civic
Am
US FTP cycle CO2 g/km CO2 g/km CO2 g/km CO2 g/km CO2 g/km
0 % Ethanol 249.2 101.9 303.9 161.6 179.6
10 % Ethanol 244.3 105.0 305.8 162.8 176.5
15 % Ethanol 247.4 103.2 307.0 162.8 176.5
20 % Ethanol 248.6 100.1 307.0 161.0 175.3
Page 106
Table A5. Fuel consumption when using neat gasoline and three different blends of ethanol.
Vehicle Grand Honda Insight Silverado Echo Civic
Am
US FTP cycle FC FC FC FC FC
l/10 km l/10 km l/10 km l/10 km l/10 km
0 % Ethanol 1.063 0.435 1.293 0.701 0.764
10 % Ethanol 1.077 0.463 1.346 0.720 0.783
15 % Ethanol 1.102 0.464 1.352 0.731 0.790
20 % Ethanol 1.144 0.460 1.410 0.739 0.803
In Tables A6, A7 and A8 the emission data and fuel consumption of interest are presented; the
measurements of regulated emissions and fuel consumption in Table A6, and those of
unregulated emissions in Tables A7 and A8. As can be seen there is considerable variation in
emission levels from vehicle to vehicle. It can also be seen that some vehicles show an increase
in emissions when ethanol blended fuel is used, while others show a decrease.
15. APPENDIX 2
15.1. Evaporation Test of Two Ethanol Blended RON 95 Summer Gasolines
Separate report from AVL MTC AB.
Evaporation Test of Two
Ethanol Blended RON 95
Summer Gasolines
Magnus Henke
2005-04-29
MTC-AA-0024 Rev. 4
TABLE OF CONTENTS
1 Summary ......................................................................................................................................... 3
2 Introduction ..................................................................................................................................... 4
3 Experimental ................................................................................................................................... 5
3.1 Test Facilities ......................................................................................................................... 5
3.2 Testing equipment.................................................................................................................. 5
3.3 Fuels....................................................................................................................................... 6
3.3.1 Specifications..................................................................................................................... 6
3.4 Blending & Storage................................................................................................................ 8
3.5 Test procedure........................................................................................................................ 8
3.5.1 Conditioning:..................................................................................................................... 8
3.5.2 Test procedure in the VT shed........................................................................................... 8
3.5.3 Purging of the VT shed...................................................................................................... 9
3.6 Calibration ............................................................................................................................. 9
4 Results ........................................................................................................................................... 10
4.1 Changes over time in Hydrocarbon levels (FID)................................................................ 10
4.2 Changes over time for other components (MS).................................................................. 11
4.2.1 Ethanol............................................................................................................................. 11
4.2.2 Butane.............................................................................................................................. 12
4.2.3 Benzene ........................................................................................................................... 13
4.2.4 Toluene and Xylene......................................................................................................... 14
4.2.5 MTBE .............................................................................................................................. 15
4.3 Comparison of final values, Bar graphs............................................................................... 16
4.4 Vapour pressure analyses..................................................................................................... 19
5 Indexes .......................................................................................................................................... 20
5.1 List of Pictures ..................................................................................................................... 20
5.2 List of Figures and charts..................................................................................................... 20
5.3 List of Tables ....................................................................................................................... 20
5.4 List of Bar graphs................................................................................................................. 20
AVL MTC 2
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
1 SUMMARY
The evaporative emissions from two summer gasoline fuels (with Reid Vapour Pressures,
RVPs, of 63 kPa and 70 kPa, respectively) blended with low percentages of ethanol (0%, 5%,
10% and 15%) have been measured in a VT Shed. For reference purposes E85 (85% ethanol)
was also measured in the same manner.
In the tests, a half-filled container prepared with an opening for the gaseous compounds to
pass through was placed on its side in the shed at an ambient temperature of 45 C. Each fuel
was stabilized at 0C at the beginning of the test and had a final temperature of about 40C.
Total hydrocarbon (THC) emissions were monitored at appropriate intervals through the use
of a Flame Ionization Detector (FID) installed in the VT Shed and a number of components of
interest of the same test fuel were sampled throughout the tests with a mass spectrometer
(MS).
The FID data showed that evaporated THC levels ranged from 200 ppm for the unblended
gasoline with the low (63 kPa) RVP, to 340 ppm for a 10% mix of ethanol in the high RVP
(70 kPa) gasoline.
Test results showed that the RVP rises sharply when ethanol is first added to the fuels and
peaks at ethanol contents between 5 and 10%
Butane was the main component that vaporized during the tests, as expected since the vapour
pressure of commercially supplied gasoline is primarily adjusted to meet regulated standards
by adjusting butane levels.
The evaporation of ethanol was influenced both by its content in the tested blends and by the
consequent changes in vapour pressure (RVP), and thus showed slightly different evaporation
patterns from the other components.
The addition of ethanol also caused the evaporation of all the remaining components to
increase by approximately the same factors as the butane
Evaporation from the blends with the high RVP (70 kPa) base gasoline was stronger than the
evaporation from the gasoline with the lower RVP (63 kPa), although there was some overlap.
AVL MTC 3
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
2 INTRODUCTION
To meet the requirements of the European Directive designed to promote the use of biofuels
(2003/30/EC) Sweden has to substantially increase the production or imports of renewable
fuels. Legislation has to make it easier, and in some cases possible, to market and use these
hybrid fuels.
To meet these goals in a short time steps have to be taken to ensure that the introduced fuels
are usable by the current vehicle fleet. The only plausible way to reach this goal is to increase
the amount of these fuels blended in todays conventional fuels. Previous experience suggests
that alcohols should be added to gasoline for this purpose.
Methanol and ethanol are the main alternative fuels that have been tested on a relatively large
scale, both as blended and pure fuels. This report focuses on the use of ethanol.
One problem related to the use of ethanol in gasoline is the increased vapour pressure caused
by the alcohol when mixed in relatively low proportions with gasoline. Blends of around 10%
are of particular concern today.
In the study reported here, the partial vapour pressures of ethanol and selected components of
gasoline were measured in tests with blends of two base gasolines (with Reid Vapour
Pressures, RVP, of 63 kPa and 70 kPa, respectively) and ethanol contents of 0%, 5%, 10%
and 15%.
AVL MTC 4
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
3 EXPERIMENTAL
All tests were performed using a gas-proof test container (not the fuel container mentioned
below) in which the evaporative behaviour of whole cars is normally tested. It is called a VT
shed as both its volume and temperature are controlled. A standard component of this kind of
equipment is a Flame Ionization Detector (FID) for measuring the total emitted hydrocarbons.
This instrument, along with an Air Sense Mass Spectrometer, was used for the tests.
AVL MTC 5
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
3.3 Fuels
The base fuels used were identical mixtures, except for their contents of the Reid Vapour
Pressure (RVP)-regulating compound butane, a larger quantity of which had been removed
from the 63 kPa RVP fuel. Adjusting the RVP in this way is a standard procedure when
commercially preparing gasoline that is to be blended with ethanol, to ensure that the vapour
pressure of the final blend does not exceed specified limis.
3.3.1 Specifications
AVL MTC 6
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
AVL MTC 7
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
The test fuels were blended by personnel with a very good understanding of the test and the
difficulties associated with it. Particularly close attention was paid to obtaining a uniform start
temperature and making sure that no fuel was exposed to temperatures that might cause early
vaporization. Therefore blending was performed in an area conditioned to a maximum of 0C.
For the same reason, the fuel was stored in the conditioned area until just before the test. On
average the transportation and preparation of the previously blended fuel before the tests took
only 3 min. By the time the test was started the temperature of the gasoline had risen on
average by 1.5C.
3.5.1 Conditioning:
The fuel was brought down to sub-zero temperatures before being poured into the special test
container and taken into the test facility. In the test container the temperature of the fuel was
monitored continuously. When the desired temperature was reached the container was taken
into the VT shed and the test was started.
A specially prepared fuel container conditioned at 0C was placed in the container whilst the
VT shed itself was set to maintain a steady 40C throughout the test. To make sure a uniform
concentration was present throughout the entire volume of the VT shed, a small floor-
mounted fan directed away from the fuel container acted as an air mixer.
In the fuel container a small hole was drilled, which was plugged until the start of the test,
when the container was set on its side, the hole was opened and a thermocouple was plugged
in to monitor the fuel temperature during the test.
AVL MTC 8
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
To ensure that the VT shed was rid of all residual vapour before a new test was started it had
to be purged. This is a lengthy procedure which involves taking the shed to its maximum
temperature for approximately an hour.
A graph displaying changes in temperature over time of all the tested blends is presented
below (Figure 1). The rising traces represent the fuel container temperature and the steady
traces the VT shed air temperature.
Temperatures of the Fuels and Shed Cell Air During the tests
50
45
35
Degrees Celcius
30
0
Time, 2 hrs
Figure 1 Temperature of the Fuels and VT Shed cell Air during the tests
3.6 Calibration
Both of the instruments used for measuring (the FID and MS) during this project were
calibrated using certified pre-mixed gases to ensure quality. All calibration gases were
delivered by Air Liquide and traceable in accordance to the standard of NIST.
AVL MTC 9
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
4 RESULTS
Changes over time in the air levels of the measured constituents as the tests progressed are
here presented in chart form to facilitate interpretation of the measurements.
The final concentrations at the end of the tests are also presented in bar graph format at the
end of this report to facilitate comparisons.
400
350
R63-10
250 R63-15
R70-0
R70-5
200
R70-10
R70-15
150
E85
RVP 63 (0-15%
100
E85
50
0
Time, 2 hrs
Figure 2 Changes over time in Total Hydrocarbon Concentration (FID) in the VT Shed
AVL MTC 10
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
4.2.1 Ethanol
Unsurprisingly, the ethanol content of the vapour proved to increase with increasing ethanol
content in the test container to some extent. However, the relationship was not linear. Two
factors contribute to this pattern. Naturally, the vapour pressure is positively correlated to the
ethanol content of the blend in the container. In addition, however, ethanol enhances the
vapour pressure according to a relationship that is not linear and peaks at relatively low
ethanol contents of around 10% in ethanol-gasoline blends. This is shown here since the
vapour contents (ppm) were virtually the same for both the 5% and 10% blends, while those
of both the 0% and the E85 blends were considerably lower.
14,0
Ethanol (fl.avg.) 70-10 C2H5OH-45
RVP63 (5-15)
Ethanol (fl.avg.) 70-15 C2H5OH-45
12,0
Ethanol (fl.avg.) E85
8,0
E85
6,0
4,0
RVP63-0, RVP70-0
2,0
0,0
Time, 2hrs
AVL MTC 11
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
4.2.2 Butane
Butane is the main compound used to regulate the vapour pressure of gasoline. It is removed
to reduce the vapour pressure when required, for instance in summer gasoline, which is used
in higher ambient temperature ranges. The gasoline with the higher Reid Vapour Pressure
(RVP) index was therefore expected to have a higher content of vaporizing butane. This
expectation was confirmed in the tests, as shown below. In addition, the butane vapour
pressure-enhancing effect of the ethanol blended in the base gasolines was probably maximal at
contents between 5 and 10%.?
40,0
20,0
E85
0,0
Time, 2 hrs
AVL MTC 12
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
4.2.3 Benzene
Benzene is an aromatic hydrocarbon with both ozone formation and carcinogenic potential, so
its emission levels are strictly restricted. Aromatics, however, are widely used to substitute
lead as an anti-knock compound in gasoline.
Benzene is also created during the combustion process in engines.
0,65
Benzene (fl.avg.) 63-0 C6H6-78
0,60
RVP 70-5 (max)
Benzene (fl.avg.) 63-5 C6H6-78
RVP 70-0
Benzene (fl.avg.) 63-10 C6H6-78
0,55 RVP 63-5
Benzene (fl.avg.) 63-15 C6H6-78
0,50
RVP 63-0
Benzene (fl.avg.) 70-0 C6H6-78
Concentration of Benzene, ppm
0,25
0,20
0,15
0,10
E85
0,05
0,00
Time, 2hrs
Max concentration increase compared to base levels: approx. 20% (with E5).
AVL MTC 13
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
2,0
Toluene (fl.avg.) 70-10 C7H8-92
Toluene (fl.avg.) 70-15 C7H8-92
Toluene (fl.avg.) E85
Poly. (Toluene (fl.avg.) E85)
1,5
1,0
0,5
0,0
Time, 2hrs
0,4
0,2
0,0
Time, 2hrs
Max concentration increase compared to base levels: approx. 20% (with E5)
AVL MTC 14
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
4.2.5 MTBE
0,40
0,15
0,10
0,05
0,00
Time, 2 hrs
Max concentration increase compared to base levels: approx. 20% (with E5)
AVL MTC 15
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
25
20
15
10 Ethanol RVP63
Ethanol RVP70
5 Ethanol E85
0
0% 5% 10% 15% E85
Blend of Ethanol
140
120
100
80
Butane RVP63
60
Butane RVP70
40
Butane E85
20
0
0% 5% 10% 15% E85
Blend of Ethanol
AVL MTC 16
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
0,7
0,6
0,5
0,4
0
0% 5% 10% 15% E85
Blend of Ethanol
2,5
1,5
Toluene RVP63
1 Toluene RVP70
Toluene E85
0,5
0
0% 5% 10% 15% E85
Blend of Ethanol
AVL MTC 17
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
1,4
1,2
1
0,8
0,6
Xylene RVP63
0,4
Xylene RVP70
0,2
Xylene E85
0
0% 5% 10% 15% E85
Blend of Ethanol
0,4
0,35
0,3
0,25
MTBE RVP63
0,2
MTBE RVP70
0,15
MTBE E85
0,1
0,05
0
0% 5% 10% 15% E85
Blend of Ethanol
AVL MTC 18
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
A small amount of each of the gasoline and ethanol blends was sampled and sent for analysis
to Saybolt Co. The results are shown in Bar graph 7. The samples were packed in gas-proof
containers, provided by the company, to prevent any components escaping and thus changing
the properties of the sample before analysis.
90
80
70
60
RVP 63
50
RVP 70
40
E85
30
20
10
0
0% 5% 10% 15% E85
Bar graph 7. Results from the Reid Vapour Pressure analyses by Saybolt
AVL MTC 19
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00
MTC-AA-0024 Rev. 4
5 INDEXES
AVL MTC 20
Armaturvgen 1, P.O. Box 223, SE-136 23 Haninge. Tel 08-500 656 00