EPA/AA/CTAB/PA/81-12
TECHNICAL REPORT
A Review of the Compatibility of Methanol/Gasoline Blends
with Motor Vehicle Fuel Systems
by
Robert J. Garbe
May, 1981
NOTICE
Technical reports do not necessarily represent final EPA decisions or
positions. They are intended to present technical analyses of issues using
data which are currently available. The purpose in the release of such
reports is to facilitate the exchange of technical information and to inform
the public of technical developments which may form the basis for a final
EPA decision, position or regulatory action.
Control Technology Assessment and Characterization Branch
Emission Control Technology Division
Office of Mobile Source Air Pollution Control
Office of Air, Noise and Radiation
U.S. Environmental Protection Agency
2565 Plymouth Road
Ann Arbor, Michigan 48105
-------
Scope and Emphasis
This literature review of the compatibility effects of methanol/gasoline
blends on vehicle systems has been undertaken as part of a larger effort by
EPA to evaluate the request for a waiver of the Section 211(f) fuel additive
regulations submitted by Anafuel Unlimited on February 20, 1981. This
waiver has been requested by Anafuel for Petrocoal, an oxygenated, unleaded
gasoline blend containing up to 12Z methanol, up to 6Z C-4 alcohols (not
identified) and up to 0.033 g/gal but not less than 0.023 g/gal of a
proprietary compound claimed by Anafuel to be a corrosion inhibitor.
Within the two week time period allowed for the preparation of this report,
it has been impossible to obtain copies of literature which were not
immediately available. Fortunately, a fairly comprehensive collection of
references was located in-house and few important references are missing
from this review. Another limitation to this report is that no
compatibility information is available on the specific composition for which
_a waiver is requested, other than what has been supplied by the requestor.
It is known that several interested parties are or will soon be conducting
bench or fleet scale compatibility tests of Petrocoal,(1,2)* and the results
of these tests may be important. However, since the results from these or
other tests are not yet available, they are not included in this review.
Other general limitations have been applied to this review to simplify and
streamline this report. Only methanol/gasoline blends containing 10-15%
methanol are included in this review. Special emphasis has been given to
the effects of inhibitors and C-4 alcohols on the compatibility of methanol
/gasoline blends in motor vehicles.
In order to best fit the specific purpose of this report and yet provide a
sufficient amount of general information for background, this report is
structured to fulfill two objectives. The first objective is to review in
general terms the overall information on compatibility of
* Numbers in parentheses indicate references listed in back of this report.
-------
-3-
meChanel/gasoline blends. The second objective is to apply the existing
pieces of information to the Petrocoal blend in order to provide an estimate
of the most likely problem/benefit areas of this blend. A related part of
this objective will be a discussion of the data gaps which need to be filled
in order to evaluate more fully the compatibility of Petrocoal as a motor
fuel.
Conclusions
Bench tests on typical automotive fuel systems materials of construction
have indicated a number of sensitive materials in both metal and
non-metal categories.
Presence of inhibitors and/or C, alcohols may have a beneficial effect
on automotive metals compatibility as measured in bench tests.
Inhibitors have not shown any beneficial effects on non-metals.
The bench tests have generally not shown good correlation with vehicle
tests, which have appeared to indicate less problems than expected based
on the bench tests.
- A wide variation in experience is evident between different researchers
on the gross effects (performance, not emission) of continued use of
raethanol/gasoline blends. The extent of fuel related vehicle problems
has varied for fleet tests over 1 year of duration.
- Very little data are available on the emissions durability of vehicles
using methanol/gasoline blends. What data there are, are generally
favorable, but are too sparse to support a conclusion.
Relative to the Petrocoal blend, it is apparent that this blend has
compositional characteristics (according to manufacturers' specs)
preferable from a vehicle compatibility viewpoint, to a blend using
-------
-4-
only methanol. However, there appears to be no available data, either
in the published literature or supplied by Anafuel, which would
conclusively demonstrate that Petrocoal would be safe (from a emission
control standpoint) to operate in currently available motor vehicles
over long time periods.
- Relative to the Petrocoal blend several data gaps must be filled before
the compatibility issue is resolved.
1) Emissions durability to 50,000 miles should be demonstrated.
2) Long term observed and controlled vehicle tests are probably
needed.
3) Bench data on water separation sensitivity and the corrosive
effects of water contaminated blends are needed.
4) More complete data are needed on non-metals such as nitrile
for Petrocoal blends.
(^
Review of Compatibility Literature
This section of the report provides a general overview of the compatibility
information on methanol/gasoline blends. The effect of automotive fuel in
general and methanol in particular on automotive fuel systems can be broken
into two areas depending on the materials of interest; metals or
non-metals. Further stratifications of the data are presented to pertain to
the particular characteristics of Petrocoal. These categories are concerned
with inhibitors and higher alcohols relative to how their presence impacts
the compatibility of methanol/gasoline blends. Finally, there is a brief
section on emissions durability over extended mileage. Since the Anafuel
waiver request is discussed in later sections of this report, and since the
data submitted by Anafuel have not been published, it will not be reviewed
in this part of the report.
-------
-5-
Several general reviews have been written which deal with the compatibility
of tnethanol/gasoline blends (3,4,5,6,7). These studies indicate that a wide
variation exists in the results of compatibility studies on raethanol
blends. Many different types of automotive materials have been tested under
a variety of conditions. The following sections will discuss the results of
these investigations on the different automotive materials of construction.
Metals
A variety of metallic parts have been identified as potentially sensitive to
degradation by methanol/gasoline blends in the range of 10-15% methanol.
The most frequently referred to metals are (a) terne steel or terne plate,
which is used in fuel tanks, (b) magnesium and (c) aluminum, which are used
in carburetors and fuel punp bodies (4).
Terne steel, which is sheet steel that is hot dipped in a tin-lead solution
to retard corrosion, is almost exclusively used in current automotive fuel
tanks except for some foreign models, where zinc coatings are used (1,2,4).
Bench scale tests by Leng (8) on pure methanol have shown severe degradation
or dissolution of the lead/tin coating of terne plated steel. A report by
Poteat (9) showed accelerated corrosion of terne plate when a 10% methanol
blend was compared to Indolene as a base. But both Indolene and the
methanol blend corroded terne steel at a small fraction of the rate of pure
methanol(See Table 1).
Table 1
Uniform Corrosion of Terne Steel in Various Fuels
2
Fuel Corrosion (mg/decemeter )
•
Indolene 0.031
10% Methanol/90% Indolene (Dry) 0.173
10% Methanol/90% Indolene (Water Saturated) 0.261
Methanol 4.34
-------
-6-
Uniforra corrosion leads eventually to a removal of the protective terne
lining the fuel tank, which in turn leads to accelerated corrosion of the
fuel tank steel itself. Presence of water contamination leads not only to
more uniform corrosion but tends to increase the tendency toward pitting
corrosion which can lead to perforation of a fuel tank in a short period of
time (9). Keller (6) presented data indicating that little corrosion of
terne plate occurred except with dry pure methanol.
Fleet test results using terne metal in the gas tanks have not shown any
catastrophic failure of the tanks due to corrosion by gasoline/methanol
blends under 152 methanol (10,11,12,13,14). However, Lindquist, et al. (13)
indicated that the conduction of spurious galvanic currents by rear mounted
fuel pumps can lead to rapid fuel tank failure. No other instances of this
type of failure in fleet tests have been found in the literature on this
failure mode but the acceleration of corrosion due to the presence of
electrical currents in methanol/gasoline blends has been reported from bench
scale experiments by other investigators (6,9).
To expand on the issue of galvanic corrosion in methanol/gasoline blends,
gasoline is known to be a relatively good electrically insulating liquid due
to the general non-polar nature of the constituent hydrocarbons. Methanol
on the other hand is >very polar and conducts electricity much better (9).
Therefore, the presence of methanol in a fuel blend would be expected to
increase the tendency and extent of galvanic corrosion (6,9).
Magnesium and aluminum have also been identified as potential problem metals
with respect to use of methanol/gasoline blends. Both of these materials
have exhibited extensive corrosion potential in bench scale tests (6,8,9).
Poteat (9) reported specific corrosion problems (pits) with aluminum at the
interface of separated, water contaminated methanol/gasoline blends.
Magnesium tended to dissolve in pure methanol and was significantly corroded
in methanol/gasoline blends (6). Recommendations for metals to be avoided
when raethanol is used as a motor fuel (pure or in blends) include magnesium,
cadmium, antimony, lead and alloys rich in these metals (8,10).
-------
-7-
The non-metals used in fuel systems consist of a large number of different
polymers, elastomers, rubbers, etc. which have been responsible for the most
reported failures of vehicles fueled with methanol/gasoline blends (10,13).
Other investigators have shown no problems in operation of methanol/gasoline
blends (8). Bench scale testing of noa metals has uncovered a large nunber
of sensitive non-metals which are used in automotive fuel systems. Some
materials have shown consistent adverse reactions to methanol/gasoline
blends. The materials which are particularly sensitive have been identified
as natural rubber (not used in current motor vehicles), polyurethane (used
as fuel lines in some vehicles), cork gasket material, leather, polyester
bonded fiberglass laminate, PVC, and certain other plastics (polyamides and
nethylmethacrylate) (6,8,15,16). Other materials have shown esentially
complete resistance to degradation in methanol/gasoline blends. These
resistant materials are Buna N and Neoprene rubber, polyethylene, nylon and
polypropylene (6,8,15,16). A very large class of non-metals appear to be
affected inconsistently in various literature references. Nitrile was found
to be resistant to methanol/gasoline blends by Leng (8) but to be highly
sensitive in bench scale tests by Abu-Isa (15) and Cheng (16). Viton, a
fluorocarbon elastomer used widely in fuel systems, shows a great variation
in effects, which has been at least partially attributed to slight
variations in the composition of the Viton itself (13).
It is very important to note that some non-metals such as epichlorohydrin,
fluorocarbon elastomers and nitrile appear to be most sensitive to inter-
mediate blends of methanol/gasoline and less sensitive to pure gasoline or
pure methanol (15,16). The reason for this behavior can be explained by an
examination of the solvent or solubility parameter for different solutions.
For some elastomers the solubility parameters are closest to the solubility
parameter of the mixed fuel, and "like dissolves like" in the jargon of the
chemist (15).
-------
-8-
In-service failures of oethanol/gasoline blends due to non metals have been
caused by a variety of problems. Duncan et al. (13) reported very few
problems in 3 years of operation on methanol/gasoline blends. Among the
problems reported were the failure of polyurethane fuel lines, swelling of
an expanded foam carburetor float, softening and swelling of needle valve
tips, disintegration of in-line fuel filters and failure of one acceleration
pump diaphragm. However, it is not known whether vehicles used in this pro-
gram are sufficiently similar to current U.S. vehicles to extrapolate the
results. Nierhauve (10) indicated that 13 cars in a fleet test exhibited
numerous problems that illustrate that these cars are not suited to run on
methanol blends (152 methanol) without modifications. This fleet test
shoved problems similar but more wide spread than the New Zealand tests (13).
Based on informal contacts with investigators at domestic automobile
manufacturers, it was learned that there is widespread concern about the
possible negative impact of methanol/gasoline blends in current U.S.
vehicles. Acceleration pump cup diaphragms were identified as a particular
concern in addition to other elastomeric materials and sensitive metals.
Inhibitors and Higher Alcohols
Corrosion inhibitors and higher alcohols have been investigated as a partial
or total means of addressing the negative effects of methanol/gasoline
olends as automotive fuel (6,17,18). Corrosion inhibitors are used in
conventional gasolines to retard corrosion in metal fuel systems
components. Higher alcohols, especially C4 iso or tertbutyl alcohol, have
been tried as a way to prevent or minimize phase separation in water con-
taminated blends. The prevention of phase separation would have definite
benefits for overall driveability as well as in corrosion of water sensitive
components such as aluminum (6). There are very little data on the overall
effect of using both of these methods at the same time on gasoline/methanol
blends. Codling (17) reports favorable results using a proprietary blend of
inhibitors (PROMAX 8027 and PROAL) to eliminate pure ethanol corrosion when
used at the refinery gate in amounts of approximately 0.5 g/gallon. It is
not known what effect this formulation would have on methanol/gasoline
-------
-9-
blends. Bench scale tests performed by Keller (6) using a methanol/higher
alcohol/gasoline blend with and without corrosion inhibitors showed es-
sentially no benefits due to the inhibitors. Inhibitors are not known to
exert any effects on non-metals, due to methanol/gasoline blends (6).
Higher alcohols have been well investigated for their effects on the water
sensitivity of methanol/gasoline blends (18). Data generated by Svahn (18)
showed that 1% isobutanol added to a 15% methanol/gasoline blend orovided a
stable blend up to 0.4% water at +5°C. Data generated by Keller (6) con-
firmed Svahn1s data above and concluded that a non-specific blend of
C -C, alcohols, as would be produced by certain methanol production
processes, would be effective as a co-solvent to minimize water separation
in methanol/gasoline blends.
Emissions Durability
A prime consideration in evaluating the overall compatibility of a fuel
additive or fuel blend is the effects the blend will have on total emissions
over the vehicles' useful life compared to operation with gasoline. The
currently accepted way to gauge this effect is to operate the vehicle for
50,000 miles according to federal specifications contained in CFR 40 Part
86. However, this type of test specification is designed to provide an ac-
celerated view of a vehicle's emission durability over a 50,000 mile
period. Normally, this type of program may take only 6 months to complete
rather than the 5 or so years in consumer use. Thus, if a fuel blend has
long term effects on a vehicle (taking place after first 6 months) different
Chan gasoline, a program of this type may not uncover it. Therefore, other
information on long term vehicle operation and bench scale tests are im-
portant to the overall evaluation of a new gasoline additive or blend.
Very little data on methanol/gasoline blends are available on the emissions
durability using the federal test procedures for 50,000 miles of operation.
Stamper et al. (14) presented data for seven 1977 and 1978 vehicles for
50,000 miles of operation. These data did not indicate any emissions in-
creases which could be attributed to compatibility problems with the
methanol/gasoline blends. Crowley et al. (19) reported on a 4 vehicle fleet
test (1975 vehicles) operated under similar but not identical conditions to
-------
-10-
that of Stamper. This program used a blend of 92 methanol, 17. isopropanol
and a dispersant additive package in unleaded gasoline. The results of this
program agreed with the previous program, showing no compatibility related
emissions increases over 50,000 miles of operation over a 5 month time per-
iod.
Observations on Petrocoal
With the information presented in the previous sections of this report and
with the information presently available on the composition of Petrocoal,
some estimates can be made about the compatibility of Petrocoal. The com-
position of Petrocoal has been given as up to 122 methanol, up to 62 of
unidentified C4 alcohols and up to 0.033 g/gal of a proprietary inhibitor.
The C4 alcohols included in the Petrocoal additive may be sufficient to
minimize phase separation in the field if appropriate caution is used in its
distribution. The inhibitor in the Petrocoal blend, although unidentified,
may decrease the corrosion potential of the blend, but other than a possible
beneficial effect on sensitive metals which has not been indicated in the
literature for inhibitors in general, no other beneficial effects are
expected. In particular, the Petrocoal blend should have essentially the
same degradative effects on elastomers that other 10%-152 methanol gasoline
blends do. The investigators at automobile manufacturers which were
informally contacted view the potential elastomer compatibility problems as
a key concern in this waiver evaluation.
Observations on Anafuel Waiver Request
Up to this point in the report, the compatibility data presented by Anafuel
have not been discussed. This section of this report intends to discuss the
compatibility data that Anafuel presented in light of the overall litera-
ture, and to highlight data gaps which may be important to the overall
evaluation of Petrocoal.
-------
-11-
The Anafuel submissions consist of the main February 20 request to the EPA
Administrator and a supplementary report prepared by Geoff J. Gerraaine PhD.
It has been assumed that the compatibility information contained in the
February 20 request is a summary of the data in the supplementary report.
Therefore, only the supplementary report will be reviewed.
As a general comment, the compatibility report submitted by Anafuel seemed
to be of high quality and the experiments were reported in a professional
manner. The data presented indicate no significant or even noticeable
problems with the Petrocoal blends in the metals and elastomers tested.
However, these results appear to conflict with the published literature and
also have several omissions and inaccuracies that should be pointed out.
With respect to metals, the existing literature do not support the pos-
sibility of an inhibitor that can eliminate the corrosive and degradative
effects of a 10% methanol/gasoline blend on metals. However, since the
corrosive effects of a low percentage methanol/gasoline blend have not been
shown to be a major problem with metals tested in the Anafuel compatibility
tests, it is possible that metal corrosion may not be a negative factor with
Petrocoal. Some success has been shown with ethanol in developing inhibi-
tors and thus it may be possible to have some effect on methanol/gasoline
blends with a proprietary inhibitor.
The brief statement in the Anafuel report on the possible presence of
galvanic corrosion is strongly supported by the literature. Methanol and
methanol/gasoline blends do conduct electricity much better than gasoline.
The impact of this tendency on actual vehicle operation is not known, but
results from current fleet tests have not identified galvanic corrosion as a
problem.
The report points out that zinc is now used to plate gasoline tanks in
current vehicles (since 1975). This statement is incorrect as terne plate
is still used exclusively in domestic and most foreign vehicles. In fact,
the samples tested in the Anafuel report would actually be terne plate
-------
-12-
if they came from an American vehicle (such as Che GM Monza). Magnesium
should have been tested as a particularly susceptible metal which, because
of its light weight, is being used in more vehicle applications. 1C is
possible that the carburecor metal tested in the Anafuel report is some
magnesium alloy but no definite information could be obtained from the manu-
facturers (according to Anafuel). Water has been indicated as both a bad
and good actor in metal corrosion studies with methanol, and the presence of
at least some water in the gasoline distribution network is endemic. Also,
water condenses in motor vehicle gasoline tanks. However, Anafuel submitted
no data on water-contaminated blends. The metal compatibility results with
the pure additive (60-70 methanol) are suspect based on existing literature
and should be verified. Most references indicated at least some noticable
corrosion of these metals with methanol at this percentage regardless of
inhibitors.
With respect to non-metals, there is little existing literature that would
collaborate the data presented in the Anafuel waiver request. The
elastomers tested by Anafuel were noted in the existing literature as being
the most resistant to attack by methanol but even these materials tended to
show some effects with oiethanol/gasoline blends. Many more elastomers must
be examined before conclusions on Petrocoal and elastomers are made.
Nicrile rubber, as the report indicates, should be tested with Petrocoal as
well as fluorocarbon elastomers. Specific problem parts should be tested
also including accelerator pump plungers and in-line fuel filters. Water
contaminated blends were not tested and should be tested. It is important
to have a complete picture on the compatibility of Petrocoal with vehicle
systems before a waiver is granted. Controlled fleet studies should perhaps
be run, such as is being done in California for pure methanol vehicles. In
this way, the extent of compatibility related vehicle emissions increases
and operational difficulties can be ascertained.
-------
-13-
REFERENCES
1. R.J.Garbe, Private communication with General Motors Representative,
Warren, Michigan 4/29/81.
2. R.J.Garbe, Private communication with Ford Motor Company Represent-
ative, Dearborn, Michigan, 4/29/81.
3. D.L. Hagen, Methanol: Its Synthesis, Use as a Fuel, Economics and
Hazards; SAE Paper 770792.
4. D.J. Bologna, Corrosion Considerations in Design of Automotive Fuel
Systems; SAE Paper 789020.
5. N.E. Gallopoulos, Alternative Fuels for Reciprocating Internal
Combustion Engines - A Literature Review; GMR Report #2537.
6. J. Keller, Methanol Fuel Modification for Highway Vehicle Use; DOE
Report £HCP/W3683-18.
7. R.R. Adt, et al. , Methanol-Gasoline Fuels for Automotive Transportation
- A Review; Nov. 1974.
S. I.J. Leng, Fuel Systems for Alcohol-Corrosion and Allied Problems; 4th
International Symposium on Alcohol Fuels Technology.
9. L.E. Poteat, Compatibility of Automotive Fuel Systems Materials with
Methanol Gasoline Fuels.
10. Methanol-Gasoline Blended' Fuels in West Germany. Specification and
Early Field Experience; B. Nierhauve 4th International Symposium on
Alcohol Fuels Technology.
11. G.R. Cassels, Third International Symposium on Alcohol Fuels Technology,-
B P New Zealand, Experience with Methanol/Gasoline Blends.
-------
-14-
12. J.C. Ingaraells, et al.,Methanol as a Motor Fuel or a Gasoline Blending
Component. SAE Paper 7501123.
13. J. Duncan, et al., The MK15 Blend Test Programme of the New Zealand
Liquid Fuels Trust Board; 4th International Symposium on Alcohol Fuels
Technology.
14. K. Stamper, Fleet Trials Using Methanol/Gasoline Blends; 4th Inter-
national Symposium on Alcohol Fuels Technology.
15. Ismat A. Abu-Isa, Effects of Mixtures of Gasoline with Methanol and with
Ethanol on Automotive Elastomers; GMR Report-3137.
16. C.W. Cheng, Effect of Gasohol and Alcohols on Elastomeric Materials',
Detroit Rubber Group Meeting October 18, 1979.
17. Victor J. Codling, The Development and Testing of an Anti-Corrosive
Additive for Alcohol Fuels; 4th International Symposium on Alcohol
Fuels Technology.
18. L.C. Goran Svahn, Specifications of Alcohol Motor Fuels; 4th
International Symposium on Alcohol Fuels Technology.
19. A.W. Crowley, et al., Methanol Gasoline Blends Performance in Laboratory
Tests and in Vehicles, SAE Paper 750417.
-------
-15-
Other Literature Not Cited
1. H. Quadflieg, et al., Objectives and First Results of the German Federal
Alcohol Fuels Project; 4th International Symposium on Alcohol Fuels
Technology.
2. R.K. Pefley, et al., A Feedback Controlled Fuel Injection System can
Accomodate any Alcohol-Gasoline Blend; 4th International Symposium on
Alcohol Fuels Technology.
3. W.H. Baisley, et al., Emission and Wear Characteristics of an Alcohol
Fueled Fleet Using Feedback Carburetion and Three-Way Catalysts; 4th
International Symposium on Alcohol Fuels Technology.
4. A. Nersisian, Resistance of Rubber Compounds to Gasoline-Methanol
Blends; Dupont Internal Report.
5. Effects of Alcohol-Containing Fuels on Spark Ignition Engine Wear;
DOE/DODIAG E (49-28)-1021.
6. H.J. Stevens, Corrosion of Aluminum and Magnesium Alloys in
Alcohol-Water Solutions, Evaluated using the Random Assignment
Statistical Analysis.
7. W. Lee, et al., Volkswagen Methanol Program and the Results of Vehicle
Fleet Test; Symposium of Alcohol as Alternate Fuels for Ontario.
November 1976.
8. E.E. Wigg, et al., Metkanol as a Gasoline Extender - Fuel Economy,
Emissions, and High Temperature Driveability; SAE Paper 741008.
9. Emissions and Compatibility Effects of Gasolhol, MTBE, and TBA; ECTD
Technical Report December 1978.
-------
-16-
10. J. Jennaine, Compatibility of Petrocoal with Automotive Fuel System
Metal and Polymeric Materials, April 1981.
11. K. Stamper, 50,000 Mile Methanol/Gasoline Blend Study - a Progress
Report; Third International Symposium on Alcohol Fuels Technology.
12. L.G. Goran Svahn, Metanol/Gasoline Mixtures in Four-Stroke OHO Engines;
Third International Symposium on Alcohol Fuels Technology.
13. G. Terzoni, et al., Improvement of the Water Tolerability of
Methanol-Gasoline Blends; Third International Symposium on Alcohol Fuels
Technology.
14. E. Earl Graham, New Zealand's Mathanol-Gasoline Transport Fuel
Programme; Third International Symposium on Alcohol Fuels Technology.
------- |