Energies: Alcohol Fuels For Spark-Ignition Engines: Performance, E To High Blend Rates For Ternary Mixtures

energies
Article
Alcohol Fuels for Spark-Ignition Engines:
Performance, Efficiency, and Emission Effects at Mid
to High Blend Rates for Ternary Mixtures
James W. G. Turner 1, * , Andrew G. J. Lewis 1 , Sam Akehurst 1 , Chris J. Brace 1 ,
Sebastian Verhelst 2 , Jeroen Vancoillie 2 , Louis Sileghem 2 , Felix C. P. Leach 3 and
Peter P. Edwards 3
1 Institute for Advanced Automotive Propulsion Systems, University of Bath, Bath, Somerset BA2 7AY, UK;
A.G.J.Lewis@bath.ac.uk (A.G.J.L.); enssa@bath.ac.uk (S.A.); enscjb@bath.ac.uk (C.J.B.)
2 Department of Electromechanical, Systems and Metal Engineering, Campus UFO, Ghent University, T4,
Sint-Pietersnieuwstraat 41, B-9000 Gent, Belgium; Sebastian.Verhelst@UGent.be (S.V.);
vancoillie.jeroen@gmail.com (J.V.); Louis.Sileghem@UGent.be (L.S.)
3 Departments of Chemistry and Engineering Science, University of Oxford, Oxford OX1 3PJ, UK;
felix.leach@eng.ox.ac.uk (F.C.P.L.); peter.edwards@chem.ox.ac.uk (P.P.E.)
* Correspondence: j.turner@bath.ac.uk; Tel.: +44-1225-383466

Received: 21 October 2020; Accepted: 25 November 2020; Published: 3 December 2020
Abstract: This paper follows on from an earlier publication on high-blend-rate binary gasoline-alcohol
mixtures and reports results for some equivalent ternary fuels from several investigation streams.
In the present work, new findings are presented for high-load operation in a dedicated boosted
multi-cylinder engine test facility, for operation in modified production engines, for knock performance
in a single-cylinder test engine, and for exhaust particulate emissions at part load using both the
prototype multi-cylinder engine and a separate single-cylinder engine. The wide variety of test
engines employed have several differences, including their fuel delivery strategies. This range of
engine specifications is considered beneficial with regard to the “drop-in fuel” conjecture, since the
results presented here bear out the contention, already established in the literature, that when specified
according to the known ternary blending rules, such fuels fundamentally perform identically to their
binary equivalents in terms of engine performance, and outperform standard gasolines in terms of
efficiency. However, in the present work, some differences in particulate emissions performance in
direct-injection engines have been found at light load for the tested fuels, with a slight increase in
particulate number observed with higher methanol contents than lower. A hypothesis is developed to
explain this result but in general it was found that these fuels do not significantly affect PN emissions
from such engines. As a result, this investigation supplies further evidence that renewable fuels can
be introduced simply into the existing vehicle fleet, with the inherent backwards compatibility that
this brings too.
Keywords: alcohols; gasoline-alcohol blends; ternary blends; renewable fuels; e-fuels
1. Introduction
The affordability of transport for the end-user is the key factor in the success of the motor vehicle,
which has accelerated mankind’s economic development since its inception; cheap transportation is a
driver of economies worldwide and this was not possible before the internal combustion engine (ICE).
This state of affairs has arisen solely because of its cost-effectiveness and the resulting affordability and
utility of the vehicles it is fitted to, coupled to the fact that liquid fuels make the storage and distribution
of the energy that they require similarly cost-effective. This overall transport system cost-to-utility
Energies 2020, 13, 6390; doi:10.3390/en13236390 www.mdpi.com/journal/energies

Energies 2020, 13, 6390 2 of 31
Energies 2020, 13, x FOR PEER REVIEW 2 of 32
ratio has
million given rise
vehicles weretomanufactured,
a situation where, for the first
representing time in 2016,
a year-round almost
rate of three95 million
vehicles vehicles
per second,were and
manufactured,
one representing
billion passenger a year-round
and commercial rate ofhad
vehicles three vehicles
been built per second,
by 2010 [1].and one billionin
Furthermore, passenger
a recent
and commercial
study, Gutzmer ofvehicles had been
INA-Schaeffler built by data
presented 2010showing
[1]. Furthermore, in a recent
that car production study,
rates Gutzmer of
are accelerating
INA-Schaeffler
and analysis ofpresented
his datadata showing
yields that car production
an estimate that the second rates are accelerating
billion vehiclesand analysis
will have of his
been
data yields an estimate that the second billion vehicles will have
manufactured in the 2021–2022 time frame [2]. He estimates the ratio of ICE-only, plug-in hybrid been manufactured in the 2021–2022
time frame
electric [2]. He
vehicles estimates
(PHEVs) and the ratio ofelectric
battery ICE-only, plug-in
vehicles hybrid
(BEVs), andelectric
analysisvehicles
of his(PHEVs)
figures andshows battery
that
electric
if there vehicles (BEVs), shift
is no paradigm and analysis
in customerof hisbehavior
figures shows
then threethat ifbillion
there vehicles
is no paradigm
will have shift in customer
been built by
behavior
2030 then three
[2]. Figure billionthe
1 presents vehicles
data ofwill have in
Gutzmer been built
these by 2030
terms; Figure[2].2 shows
Figure that1 presents
even with the such
data anof
Gutzmer inview
optimistic these ofterms;
the take Figure
up of2pure
shows that the
BEVs, even ICEwithwillsuch an optimistic
continue to dominateview light-duty
of the takepropulsion
up of pure
BEVs,
for theto
years ICE will continue
come, and that to evendominate
if whatlight-duty
might be propulsion
termed “peak for engine”
years to come,
occursand that even
in 2020, then ifstill
whatin
mightnearly
2030 be termed “peak combustion
as many engine” occurs in 2020,
engines then
will bestill
madein 2030
as nearly
were in as 2014,
many whencombustionthey engines
will be
will be made asatwere
manufactured in of
a rate 2014,
2.66when they will
per second. be manufactured
While there are many at aestimates
rate of 2.66 in per
the second.
literature, While
this there
data
has been chosen because INA-Schaeffler is one of the Tier-1 suppliers with a great interest in the
are many estimates in the literature, this data has been chosen because INA-Schaeffler is one of
Tier-1 suppliers
acceleration with a great interest
of electromobility, and in so the
they acceleration
can be assumed of electromobility,
to present aand so they
scenario canabe
with assumed
rapid take
to present
up a scenario
of electric vehicles with anda rapid take up of substitution
corresponding electric vehicles and corresponding
of ICEs. Plainly, despite substitution of ICEs.
various political
Plainly, despite various
pronouncements concerning political pronouncements
the demise of the ICE,concerning
the automotive the demise
industry ofandthe ICE, the automotive
its observers believe
industry
that they andwillits observers
continue thebelieve that they
foreseeable willother
future; continue the foreseeable
estimates by industry future; other put
observers estimates
a globalby
penetration of PHEVs
industry observers putand BEVspenetration
a global combined of at PHEVs
slightly and less BEVs
than 20% by 2030
combined at [3]. One less
slightly mightthan therefore
20% by
suggest
2030 [3].thatOne reports
mightof the impending
therefore suggestdeath of the ICE
that reports appear
of the to be greatly
impending deathexaggerated
of the ICE appear [4]. to be
greatlySynergistic
exaggerated with[4].the vehicle manufacturing industry developing over time, the fuel supply
industry has kept
Synergistic withpace. Becausemanufacturing
the vehicle there was noindustry fossil hydrocarbon
developing over fuel time,
supply theindustry
fuel supply then, early
industry
engines
has kept usedpace. bio-sourced
Because therefuels was and the hydrocarbon
no fossil switch to fossil fuelfuels
supply was made then,
industry early early
on, despite
engines earlyused
concerns
bio-sourced that demand
fuels and thewould switch to outstrip supply
fossil fuels was[5]. made Cheapness of the early
early on, despite energy store that
concerns bothdemandin the
infrastructure
would outstripand aboard
supply [5].the vehicle is of
Cheapness onethe factor
energy in the affordability
store both in theofinfrastructure
the system, and andtheaboard
favorable the
economics
vehicle is one of liquid
factor in fuel
thesupply has given
affordability of the rise to a situation
system, and the where
favorabletheeconomics
authors’ analysis
of liquidoffuel
the supply
global
use of oil rise
has given suggests that it iswhere
to a situation used continuously
the authors’ analysisat a volume of the flow
global rate
useequivalent
of oil suggeststo 7.4%
thatof that
it is usedof
Niagara
continuouslyFalls.at a volume flow rate equivalent to 7.4% of that of Niagara Falls.
100 5000
Cumulative Cars Manufactured / [Millions]
Cars Manufactured per Year / [Millions]
80 4000
60 3000
40 2000
20 1000
0 0
2005 2010 2015 2020 2025 2030 2035
Year / [-]
BEVs HEVs and PHEVs ICE-only Cumulative Cars Manufactured
Figure 1.
Figure 1. Projected
Projected production
productionvolumes
volumesofof cars split
cars into
split different
into propulsion
different typestypes
propulsion (data(data
adapted from
adapted
that given
from in [2]).inBEV:
that given [2]). battery electricelectric
BEV: battery vehicle; HEV: hybrid
vehicle; electricelectric
HEV: hybrid vehicle;vehicle;
PHEV: PHEV:
plug-inplug-in
hybrid
electric vehicle. ICE: internal combustion engine.
hybrid electric vehicle. ICE: internal combustion engine.
Energies 2020, 13, 6390 3 of 31
100 5000
Cumulative Cars Manufactured / [Millions]

Cars Manufactured per Year / [Millions] 80 4000
60 3000
40 2000
20 1000
0 0
2005 2010 2015 2020 2025 2030 2035
Year / [-]
ICE (all) Cumulative Cars Manufactured
Figure 2. Data from Figure 1 reworked to show vehicles with an internal combustion engine fitted
Figure 2. Data from Figure 1 reworked to show vehicles with an internal combustion engine fitted
(i.e., pure ICE and hybrid vehicles). The values for cumulative cars built is the same as on Figure 1.
(i.e., pure ICE and hybrid vehicles). The values for cumulative cars built is the same as on Figure 1.
Note that in 2030, nearly as many engines are predicted to be built as in 2014, and the value of 84
Note that in 2030, nearly as many engines are predicted to be built as in 2014, and the value of 84
million engines per year still represents 2.66 engines per second. ICE: internal combustion engine.
million engines per year still represents 2.66 engines per second. ICE: internal combustion engine.
It should be remembered that the customer is de facto the only financial input to the whole system
It should be remembered that the customer is de facto the only financial input to the whole system
of vehicle manufacture and transport energy supply; all of the other stakeholders, being the vehicle
of vehicle manufacture and transport energy supply; all of the other stakeholders, being the vehicle
manufacturers (OEMs), fuel companies and governments, take this money out of the system either as
manufacturers (OEMs), fuel companies and governments, take this money out of the system either
sales or taxation. If the consumer can no longer afford transportation, then a very worrying state of
as sales or taxation. If the consumer can no longer afford transportation, then a very worrying state
affairs will ensue.
of affairs will ensue.
While it is increasingly true that in many urban and extra-urban driving situations, modern
While it is increasingly true that in many urban and extra-urban driving situations, modern
production IC engines and exhaust after treatment systems actually clean the air (insofar as their
production IC engines and exhaust after treatment systems actually clean the air (insofar as their
tailpipe emissions of hydrocarbons (HCs), carbon monoxide (CO) and oxides of nitrogen (NOx) are
tailpipe emissions of hydrocarbons (HCs), carbon monoxide (CO) and oxides of nitrogen (NOx) are
significantly lower than the background concentration) [6,7], a larger global problem is that of the
significantly lower than the background concentration) [6,7], a larger global problem is that of the
emission of fossil carbon dioxide (CO2 ). At present, the global atmospheric CO2 concentration is
emission of fossil carbon dioxide (CO2). At present, the global atmospheric CO2 concentration is
increasing by at least 3 ppm per year, despite global action being taken to limit it. While all possible
increasing by at least 3 ppm per year, despite global action being taken to limit it. While all possible
steps need to be taken to move from fossil fuels to other energy carriers, as mentioned above it seems
steps need to be taken to move from fossil fuels to other energy carriers, as mentioned above it seems
to be accepted as fact that an ICE has to operate on such fuels. This is plainly not the case. Referring
to be accepted as fact that an ICE has to operate on such fuels. This is plainly not the case. Referring
to the above comments on the dawn of motoring, when vehicles used biofuels because there was no
to the above comments on the dawn of motoring, when vehicles used biofuels because there was no
alternative and consequently there was no meaningful fossil carbon footprint in agriculture, it is an
alternative and consequently there was no meaningful fossil carbon footprint in agriculture, it is an
uncomfortable fact that early ICE vehicles were more carbon neutral than a BEV is today. Unfortunately,
uncomfortable fact that early ICE vehicles were more carbon neutral than a BEV is today.
due to economics alone, society made the switch to fossil fuel to ensure the energy supply kept pace
Unfortunately, due to economics alone, society made the switch to fossil fuel to ensure the energy
with the burgeoning demand. It is not fanciful to state that before this point, when it operated on
supply kept pace with the burgeoning demand. It is not fanciful to state that before this point, when
biofuels, the early ICE vehicle was probably the most environmentally-friendly mechanized transport
it operated on biofuels, the early ICE vehicle was probably the most environmentally-friendly
system we have had or will ever have. What we now desperately need to find an answer to is the real
mechanized transport system we have had or will ever have. What we now desperately need to find
problem: that we presently fuel the majority of vehicles on fossil oil.
an answer to is the real problem: that we presently fuel the majority of vehicles on fossil oil.
Alcohols are some of the most widely used alternative fuels in the automotive market. The fact
Alcohols are some of the most widely used alternative fuels in the automotive market. The fact
that they can be made fully renewably and blended with gasoline is potentially a useful tool in the
that they can be made fully renewably and blended with gasoline is potentially a useful tool in the
drive to reduce the release of fossil CO2 associated with transport, which is one of the most difficult
drive to reduce the release of fossil CO2 associated with transport, which is one of the most difficult
energy consumption modes to decarbonize. This is especially true when one considers the cost of
energy consumption modes to decarbonize. This is especially true when one considers the cost of
BEVs for all of the stakeholders in transport, and particularly for the consumer who (if the present
BEVs for all of the stakeholders in transport, and particularly for the consumer who (if the present
economic model stays in place for the foreseeable future) appears to be expected to pay significantly
economic model stays in place for the foreseeable future) appears to be expected to pay significantly
more for vehicles than he or she presently does. While electricity may be cheap on a cost-per-MJ basis
Energies 2020, 13, 6390 4 of 31
more for vehicles than he or she presently does. While electricity may be cheap on a cost-per-MJ
basis at the moment, and BEVs are approximately three times more efficient in their energy utilization
than combustion engines, the price of such vehicles remains cost-prohibitive for many people, and is
likely to remain so (especially considering total cost of ownership) even if future stretch targets for
automotive traction battery costs are met [8].
Although, for the consumer, electrical energy is cheap in relation to transportation (liquid) fuel
costs, there is of course no guarantee that this will continue to be the case. The application of life-cycle
analysis (LCA) may also see increasing numbers of questions raised over the ecological impact of
electrification, especially since many of the minerals and elements used in the manufacture of traction
batteries and motors are extremely rare (for example cobalt, neodymium, dysprosium, and to a lesser
extent lithium). There is also the danger of energy security concerns being swapped for concerns
over commodity security—although it is true that, unlike the use of fossil fuels in combustion-engine
vehicles, commodity security affects only the construction of EVs and not necessarily their use. There are
moral concerns regarding how cobalt in particular is obtained, too. Therefore, due to the fact that
alcohols can either be used as pure fuels in their own right or blended in many ways with existing
fossil hydrocarbon fuels [9–24], there is, in the authors’ opinion, considerable merit in continuing to
investigate them as a potential future automotive energy vector.
As stated, this paper follows on from a previous work in which the performance of several
different mid- and high-blend alcohols were tested and compared to the performance of regular and
premium gasolines in a variety of test engines [9]. A key tenet in the previous paper was that the
binary blends tested, being either gasoline-ethanol or gasoline-methanol mixtures, were configured
with the same stoichiometry. The paper showed that such fuels perform essentially identically in
any given combustion system, at least in terms of engine torque, power, and brake specific fuel
consumption (BSFC) at the same spark advance, and they could also reduce emissions of oxides
of nitrogen dramatically; however, particulate matter reduction (tested in a direct-injection engine)
was found to be more blend dependent, with that finding being presented for the first time there.
This new work conducts similar tests for so-called ternary blends, configured from gasoline, ethanol,
and methanol such that they have the same stoichiometry as the binary blends discussed in the previous
work. Blends made from these components are termed ‘GEM’ blends here.
The present work utilizes fundamentally the same test engines and equipment as the previous
investigation. While the ternary blend results are presented for the first time here, some reference
will be made to the previous tests where pertinent, since many of the binary gasoline-alcohol blends
reported there have the same stoichiometry as the GEM fuels investigated here; in fact in the space of
possible blends at any given stoichiometry, the binary ones actually represent limit cases. This will be
returned to later. This paper is constructed in the following manner, similar to how the authors dealt
with the subject of binary blending of alcohols in gasoline [9]:
Section 2 discusses the rationale for and a method of introducing methanol into the bulk transport
fuel pool, namely iso-stoichiometric blend formulation. This helps to set the context for work presented
in the rest of this paper.
Section 3 discusses the testing of iso-stoichiometric GEM blends in the “Ultraboost” extreme
downsizing 4-cylinder demonstrator engine at the University of Bath. The very high brake mean
effective pressure (BMEP) that this engine is capable of operating at means that the data shown here
represents testing of the hypothesis of identical performance from any blend configured according to
the rules to a level not published before. Furthermore, the interaction with exhaust gas recirculation
(EGR), already known to have excellent synergies with alcohol fuels, is shown. Some particulate
emissions results are reported here, pertinent to the high-load operating conditions reported.
Section 4 discusses tests conducted at Ghent University using three different engines: two
port-fuel-injection naturally-aspirated engines—one a production 4-cylinder and the other a
single-cylinder research engine configured with a high compression ratio (CR)—and one multi-cylinder
Energies 2020, 13, 6390 5 of 31
direct-injection engine. The blends tested correspond to the stoichiometry of E85 and in terms of knock
all of the blends perform identically and fundamentally the same as E100.
Section 5 presents data gathered at the University of Oxford in which the particulate emissions
performance of GEM blends is investigated in a single-cylinder test engine at light load. These findings
have not been presented before, and how they reinforce the findings from the binary blend tests
conducted in the first part of this investigation is discussed.
Finally, this paper ends with a Conclusions section.
2. A Strategy for Introducing Methanol into Transport Fuel by Configuring Ternary Blends as
Drop-In Fuels
The two “light alcohols”, i.e., those with no isomers, ethanol and methanol, have for years been
blended with gasoline to create fuels for SI engines, with E85 and M85 being examples of binary blends
which have been commercialized for several years. For these fuels, the number represents the volume
percentage of the alcohol (abbreviated to E and M) in admixture with gasoline. In fact all of the alcohols
are miscible with gasoline to a greater or lesser extent and this gives rise to the potential to create
multi-component blends to a target stoichiometry, provided this value is within an upper and lower
target range.
The ternary—or three component—blends used to in this paper are all mixtures of gasoline and the
light alcohols ethanol and methanol. Collectively, they are referred to as ‘GEM’ blends; and specifically
“GXX EYY MZZ”, where XX, YY and ZZ are the volume percentages of the three individual components,
respectively. The process by which such blends can be formulated and how they can offer a route to
increased penetration of renewable energy in the fuel pool has been discussed before [17–20,22,24].
Both engine and vehicle tests of such blends have been reported, the latter showing that when ternary
blends are configured with the same stoichiometric air-fuel ratio (AFR) as E85 (this nominally being a
blend of 85% ethanol and 15% gasoline by volume)—approximately 9.7:1—then they are undetectable
by a flex-fuel vehicle engine management system when such a vehicle has already been calibrated to
be capable of taking any proportion of gasoline and E85 put into its fuel tank [17,18]. This is essentially
due to near-identical volumetric energy content (the volumetric lower heating value (LHV)) in the
blends as well their producing a near-identical response from a flex-fuel vehicle’s ethanol concentration
sensor. Furthermore, in [18], it was shown that increasing the methanol content actually improved the
cold startability of the vehicle, providing a further mechanism to keep increased renewable energy
content in the fuel pool all year round.
All of these previous reports show that if followed, these ternary blending rules permit a method
of simply creating fuel blends equivalent to any original binary gasoline-ethanol mixture. Recently,
Pearson and co-workers [24] have defined more completely the blending rules for such an approach,
and have shown the effect of water on AFR and volumetric LHV if this is left in ethanol (which
could be very beneficial from a production energetics point of view). (Note that unlike all the other
alcohols, methanol does not form an azeotrope with water, and consequently simple distillation is
sufficient to dry it fully. By contrast, ethanol cannot be dried by simple distillation beyond 4.4% v/v
water. In the case of the distillation of the water formed in a production process utilizing CO2 to
synthesize methanol, the process can utilize the waste heat generated within the exothermic synthesis
process itself, supporting the contention that once the feed stocks have been obtained using sustainable
processes, all downstream processes and products can be maintained as fully sustainable too.) Pearson
and co-workers show that blending for constant oxygen content is just as viable as blending for constant
stoichiometry or volumetric energy content, and furthermore that the magnitude of the differences
between the approaches are effectively negligible [24]. Importantly, when such ternary blends are
configured in this manner, all such possible blends have near-identical volumetric LHV, practically the
same octane numbers, and extremely close enthalpies of vaporization (to +/− 2%, the variation arising
from treating the components on a mass basis) [17,18], all properties which need to be the same or
similar for any fuel to capable of being ‘dropped in’ to an existing vehicle fuel and combustion system.
Energies 2020, 13, 6390 6 of 31
It has also been shown that for any given amount of ethanol that can be introduced into the
total fuel
Energies pool,
2020, 13, x the
FORease
PEERof synthesis
REVIEW and hence supply of methanol means that introducing it to make 6 of 32
GEM blends suitable for the vehicles already in the market effectively displaces gasoline [17]. In [17],
authors alsoalso
the authors showshow thatthat
within
withincertain wide
certain wide price
pricelimits,
limits,thetheresulting
resultingblends
blendscancanbebe cheaper
cheaper than
than
gasoline
gasoline onon aa per-unit-energy
per-unit-energy basis; basis; furthermore
furthermore in in [25],
[25], itit was
was shown
shown that that aa taxation
taxation system
system based
based
on
on taxing
taxing units
unitsofofenergy
energyand andapplying
applyinga amultiplier
multiplier based
based onon
thethefossil COCO
fossil 2 impact of supplying the
2 impact of supplying
energy in that form could provide a government with the
the energy in that form could provide a government with the means to encourage means to encourage renewable
renewable energy
energyat
the expense
at the expense of of
fossil fuels
fossil while
fuels whilekeeping
keeping thetheoverall
overall taxation
taxation revenue
revenue gathered
gatheredconstant.
constant.Thus,
Thus, a
framework could be created to encourage the consumer to move away
a framework could be created to encourage the consumer to move away from fossil fuels based on from fossil fuels based on their
being moremore
their being expensive to usetothan
expensive renewable
use than fuels,fuels,
renewable while maintaining
while maintaining economic
economicviability for all
viability forof
allthe
of
stakeholders.
the stakeholders.
For thecase
For the case of stoichiometry
of stoichiometry equivalent
equivalent to E85,totheE85,volumethe relationship
volume relationship
between thebetween
components the
components is shown
is shown in Figure in Figure
3. This 3. This was
was determined determined in previously-published
in previously-published work using
work using a proprietary fuel a
proprietary fuel characteristics tool [18,25], and here one can see that as the
characteristics tool [18,25], and here one can see that as the proportion of ethanol is reduced, so theproportion of ethanol is
reduced, so the rate
rate of increase in the ofproportion
increase in ofthemethanol
proportion of methanol
is faster than that is faster
of thethan that of
gasoline. theisgasoline.
This becauseThisone
is because one volume unit of ethanol is equivalent to a binary gasoline-methanol
volume unit of ethanol is equivalent to a binary gasoline-methanol volume unit with the proportions volume unit with
the proportions
of 32.7:67.3 (this of
then32.7:67.3
sharing (this
thethen sharing the 9:1AFR
9:1 stoichiometric stoichiometric
of ethanol).AFR of ethanol).
100
Proportion of Gasoline or Methanol in Ternary
90
80
70
Blend / [% v/v]
60
50
40
30
20
10
0
0 10 20 30 40 50 60 70 80 90 100
Proportion of Ethanol in Ternary Blend / [% v/v]
Gasoline Methanol
Figure 3.
Figure Representation of
3. Representation of how
how the
the volume
volume proportions
proportions of of three
three blend
blend constituents
constituents change
change for
for aa
targeted stoichiometric
targeted stoichiometric AFR
AFR equivalent
equivalent to to that
that of
of E85
E85 (i.e.,
(i.e., 9.7).
9.7). Here
Here these
these constituents
constituents are
are gasoline,
gasoline,
ethanol, and
ethanol, and methanol,
methanol, andand together
together in
in this
this form
form they
they are
are termed
termed iso-stoichiometric ‘GEM’ ternary
iso-stoichiometric ‘GEM’ ternary
blends. When reading this figure, note that E85 is effectively on the right; where the
blends. When reading this figure, note that E85 is effectively on the right; where the lines end is lines end is the
the
limit of
limit of being
being able
able to
to meet
meet the
the stoichiometric
stoichiometric AFR AFR limit
limit of
of 9.7. The equivalent
9.7. The equivalent where
where no
no ethanol
ethanol isis
present (i.e.,
present (i.e., M56)
M56) is
is on
on the
the extreme
extreme left.
left.
Four such
Four such GEM
GEM blends
blends were
were tested
tested by
by Lotus
Lotus Engineering
Engineering in in aa Saab
Saab 9-3
9-3 BioPower
BioPower production
production
flex-fuel vehicle employing a virtual sensor to infer the alcohol concentrations
flex-fuel vehicle employing a virtual sensor to infer the alcohol concentrations that that it was operating
it was on.
operating
These initial tests showed that all of the blends were invisible to the vehicle’s on-board
on. These initial tests showed that all of the blends were invisible to the vehicle’s on-board diagnostics diagnostics
system which
system which was
was unmodified
unmodified [17],
[17], provided
provided that
that aa certain
certain minimum
minimum level level of
of cosolvent
cosolvent was
was present;
present;
at very low temperatures M56 (the binary gasoline-methanol mixture with the
at very low temperatures M56 (the binary gasoline-methanol mixture with the same stoichiometric same stoichiometric
AFR as
AFR as E85)
E85) was
was found
foundtotoilluminate
illuminatethe themalfunction
malfunctionindicator
indicatorlamp.
lamp.ThisThiswas
was presumed
presumed to to
bebe
due to
due
phase separation because of the absence of any form of cosolvent in this particular
to phase separation because of the absence of any form of cosolvent in this particular fuel blend. fuel blend.
In GEM blends, the function of a cosolvent in mixtures of gasoline and methanol is undertaken
by the ethanol provided some minimum volume is used. This minimum volume was investigated in
later tests in a vehicle certified to a higher emissions standard utilizing a physical sensor for alcohol
concentration. However, in this later series of tests, despite repeated cold soaks and related cold start
tests at −20 °C, [18] a minimum requirement for ethanol was not found. Even M56 (i.e., the binary
Energies 2020, 13, 6390 7 of 31
In GEM blends, the function of a cosolvent in mixtures of gasoline and methanol is undertaken
by the ethanol provided some minimum volume is used. This minimum volume was investigated in
later tests in a vehicle certified to a higher emissions standard utilizing a physical sensor for alcohol
concentration. However, in this later series of tests, despite repeated cold soaks and related cold start
tests at −20 ◦ C, [18] a minimum requirement for ethanol was not found. Even M56 (i.e., the binary
gasoline-methanol mixture) started satisfactorily, suggesting that such GEM blends could help to
displace more gasoline from the fuel pool year round (for a fixed volume of ethanol that is available).
From the vehicle testing program conducted by Lotus, it was deduced that, since all of the vehicles
in the parc do not all use the same alcohol concentration-sensing technology, a minimum ethanol
concentration would in fact be needed to ensure the satisfactory operation of the entire fleet, and that
this minimum level would be in the region of 10% by volume [18]. In turn, this implies that, without
the addition of any other type of cosolvent, the E85-equivalent ternary blend with the maximum
methanol concentration that could be used is G40 E10 M50.
A further important finding from this earlier work was that regardless of the technology used to
sense the alcohol concentration in the fuel system near to the engine, energetic efficiency improvements
on the order of 5% were possible. This was true for any of the ternary blends tested at the E85-equivalent
stoichiometric AFR of 9.7:1 [17,18]. This is important when considering the amount of renewable
energy that would have to be supplied to synthesize such fuels by the electrofuel route as per the
approach discussed by various authors [26–33]. Later sections in this paper will provide further
validation of the efficiency improvements possible.
Issues of material compatibility were investigated in another publication [20], together with
results showing that exhaust emissions do not change significantly as the blends are altered, with
all E85-equivalent ternary blends essentially out-performing gasoline in this respect. However, it
should be noted that particulate and aldehyde emissions were not measured in these tests. Further,
Pitcher et al. [21] investigated the spray morphology of GEM blends in an optical direct-injection
engine, this being important because of its bearing on emissions and fuel consumption in engines
using such fuel introduction technologies. They showed that there was no significant difference in
morphology across the full spectrum of E85-equivalent blends. In the present work, Section 5 builds
on this work and presents new data from GEM blends with regards to their soot-formation tendency,
which is especially important with respect to Euro 6c (and higher) emissions legislation. These tests
were conducted in a different optical engine, and represent the first time such particulate matter data
from such blends has been presented by this group.
Considerable additional work on the GEM concept has been reported by Sileghem and co-workers
in both port-fuel-injection (PFI) and direct-injection (DI) engines [34,35]. All of this work has shown
that in terms of engine and combustion system performance any GEM blend configured according
to the rules established by Pearson and Turner will perform in the same manner using the same
engine control settings. Note that it is possible to produce ternary blends of other alcohols with
gasoline, or even quaternary (or higher number) blends. Similarly, water can be included in such
blends; the physicochemical characteristics of such these blends was investigated [24], although,
practically speaking, phase separation would become an issue with the presence of water, the light
alcohols being especially prone to this when in admixture with gasoline. Other authors have
discussed physicochemical, performance, and blending aspects of binary gasoline-alcohol blend fuels
in detail [36–39], but while it is accepted that challenges exist in blending alcohols with gasoline (such
as high vapor pressure and evaporative emissions effects aboard the vehicle, or phase separation and
blend stability in the distribution system, with materials compatibility issues in both), it is the authors’
belief that these are understood and by no means insurmountable, as evidenced by the amount of
high-blend-rate binary mixtures of either ethanol or methanol that is or has been used around the world
and the number of fully-homologated vehicles that have been sold that are capable of taking them.
From all of this initial work, the inherent compatibility of existing production flex-fuel technology
with ternary blends with stoichiometric AFRs matched to fuels already in the market place has been
Energies 2020, 13, 6390 8 of 31
established. It is therefore believed that a progressive rollout of such fuels, by methanol percentage
and region, could allow staged introduction of methanol as a transport fuel. This has the potential to
meaningfully reduce greenhouse gas emissions, improve energy security, and provide better air quality.
Once the market has accepted the new fuel, it enables widespread use of an energy carrier which could
eventually be completely decarbonized using the electrofuel approach [26–33], with methanol then
becoming a chemically-liquefied form of hydrogen as an extension of the hydrogen economy [40].
At this point in time, a practical liquid fuel that has been fully decarbonized will have been realized.
3. Ternary Gasoline-Alcohol Blend Tests in the “Ultraboost” Extreme Downsizing

Demonstrator Engine
The “Ultraboost” collaborative project, more properly known as “Ultra Boost for Economy”,
was part-funded by the UK Technology Strategy Board (now Innovate UK) and started in 2010. It
sought to demonstrate the potential to provide a vehicle-level CO2 reduction of 35% in an SUV by
employing a heavily downsized (and ‘down-cylindered’) engine. One major target was that this
new engine should be capable of generating the torque curve of the Jaguar Land Rover AJ133 V8
engine. That engine is of 5.0 L capacity and is naturally aspirated (NA). This performance had to be
realized while operating on 95 research octane number (RON) standard pump gasoline [41] (in fact the
project did indeed show that this CO2 reduction target was achievable and in fact greater than this was
demonstrated by vehicle simulation using the dynamometer data gathered).
At the start of Ultraboost, there was no pre-ordained engine swept volume, but the final version
of the engine was of in-line 4-cylinder configuration, with a compound-charging system and 2.0 L
swept volume [41]. However, the first iteration of the design (termed ‘UB100’) was externally boosted
by a charge air handling unit (CAHU) system at the University of Bath facility, where all of the engine
test work, including taking the results reported here, was carried out. Specifications for this engine
are given in Table 1. For detailed information on the other project partners and their responsibilities,
the individual stages of the project, and the test facility, the reader is referred to [41].
Table 1. Specifications of the Ultraboost extreme downsizing demonstrator engine in the

externally-boosted “UB100” configuration, as used for the fuel blends testing reported here [41].
General architecture 4-cylinder in-line with 4 valves per cylinder and double overhead camshafts
All aluminum
Jaguar Land Rover AJ133 cylinder block converted to single-bank operation on the
Construction right-hand side
Siamesed liner pack to facilitate reduced bore diameter
Dedicated cylinder head
Firing order 1–3–4–2
Bore (mm) 83
Stroke (mm) 92
Swept volume (cc) 1991
Pent-roof combustion chamber with asymmetric central direct injection and spark plug
High-tumble intake ports
Combustion system
Auxiliary port-fuel injection
Possible second spark plug position in an under-intake-port location
Compression ratio (:1) 9.0
Chain-driven double overhead camshaft (DOHC) with fast-acting dual
Valve gear continuously-variable camshaft phasers
Cam profile switching tappets on inlet and exhaust
Engine management system Lotus ‘T6’ engine control unit
As well as for its prime aim as being a platform for the development of a self-contained technology
demonstrator package, the engine was used for an extensive series of fuel tests within the program.
For these fuel investigations accurate control of the engine’s plenum temperature and pressure was
Energies 2020, 13, 6390 9 of 31
required and so the CAHU system was used throughout. Cooled EGR was also utilized in some of these
tests, this being provided by a dedicated rig especially developed for the process. This rig took exhaust
gas from the engine and pressurized it for delivery to the intake of the engine, where it was mixed
with the air supply provided separately by the CAHU. This EGR supply system is discussed further
in [41,42]. The temperature of the resulting combined charge was also controlled by the test facility,
and it is this test equipment configuration that is used in the results presented below. The exhaust back
pressure was controlled by the test cell also. The results of most of the other conventional hydrocarbon
fuels tests have been presented in other publications [9,43–47].
From its original work on the ternary blends concept (see Section 2) and as a result of being
a partner in the project, Lotus Engineering requested that E85 and an iso-stoichiometric ternary
alcohol blend be tested within the Ultraboost project, and these fuels were then formulated by Shell;
the specifications of these are shown in Table 2. In this section the ternary blend is henceforth termed
“GEM”. In the present work, the results from these iso-stoichiometric blends will be compared to two
commercial gasolines (95 and 98 RON), and with the project’s control fuel. This control or ‘base fuel’,
was a standard pump gasoline sold as 95 RON fuel, but with some octane giveaway (from 97.0 RON,
see Table 2) as a result of it having to achieve a motor octane number (MON) of 85, which is necessary
for the European market. All of these fuels are also shown in Table 2. The 98 RON gasoline contained
no ethanol, but both of the 95 RON fuels had 5% ethanol and all were production fuels of EN228
standard; as a consequence none of these are considered to be a gasoline-alcohol blend fuel in the
discussion that follows.
Table 2. Specifications of the gasoline, E85, and ternary gasoline-alcohol “GEM” fuels tested in the
Ultraboost extreme downsizing demonstrator engine. All data was supplied by Shell and measured
using standard industry test procedures and equipment.
Alcohol Proportion Volumetric Stoichiometric

Fuel Name RON MON S
and Type LHV (MJ/L) Air-Fuel Ratio (:1)
Base 97.0 85.3 11.7 5% Ethanol 31.29 14.18
95 RON Pump 95.1 85 10.1 5% Ethanol 31.51 14.19
98 RON Pump 98.7 86.5 12.2 No alcohol 31.57 14.45
E85 107.4 89.5 17.9 85% Ethanol 22.65 9.68
23% Ethanol
GEM 106.0 88.1 17.9 22.71 9.78
43% Methanol
Table 2 shows an interesting slight deviation from previous GEM fuel research insofar as the
E85 and GEM did not have RON and MON values with closer alignment. Further, they did not have
the same stoichiometric AFR. Here, they were made from different gasoline blend stocks, which was
also a deviation from previous practice. Thus, before testing began, from their previous experience
a slight difference in performance between these two fuels was expected by the authors; in the
previously-reported work by Lotus Engineering the GEM blends were splash blended using the same
gasoline each time and the achieved stoichiometric AFR range was tighter due to the use of Lotus
Engineering’s proprietary fuels properties calculator (here the stoichiometric AFR range was just over
+/−0.5% for the two fuels; in the previous work by Lotus Engineering, +0.2/−0.4% was achieved across
six blends [17,19,20]).
To illustrate this, Figures 4 and 5 show test data from ignition timing swings for the base fuel, E85,
and GEM at the same operating conditions. These operating conditions and the rationale for adopting
them have been discussed by Remmert et al. [43], together with other information on the testing
conditions. Note that throughout this test campaign, daily checks on performance were maintained
and no issues were raised on drift or repeatability by the researchers operating the engine. In Figure 4,
the engine operating condition is 2000 rpm speed, with intake plenum conditions of 40 ◦ C and 130 kPa
gauge temperature and pressure, respectively. This intake condition is referred to as test point A later in
this paper. For the results in Figure 5, the equivalent conditions are 3000 rpm, 60 ◦ C, and 140 kPa gauge,
Energies 2020, 13, 6390 10 of 31
respectively, and it included 10% cooled EGR; it is termed test point D. It is clear from these figures
that both of these operating points are high-load conditions, and that both of the high-alcohol-content
fuels yield significantly higher BMEP than the base gasoline for the same conditions; furthermore it
should be noted that for all operating points discussed here it was these intake and exhaust boundary
Energies 2020,
conditions 13, were
that x FOR maintained,
PEER REVIEW fixing the air flow rate through the engine at each point, and 10not
of 32the
fuel energy
engine wasflow rate through
operated the engine
at a relative AFR (or(despite the fact thatTest
λ) of 1 throughout). the points
engineBwas
and operated at a relative
C will be discussed
engine was operated at a relative AFR (or λ) of 1 throughout). Test points B and C will be discussed
AFR (or
later. λ) of 1 throughout). Test points B and C will be discussed later.
later.
32.0
32.0
31.5
31.5
31.0
31.0
30.5
30.5
30.0
30.0
/ [bar]
29.5
BMEP/ [bar]
29.5
29.0
29.0
BMEP
28.5
28.5
28.0
28.0
27.5
27.5
27.0
27.0
26.5
26.5
26.0
26.0 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5
-8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5
Spark Advance / [deg BTDC]
Base Fuel E85 GEM
Base Fuel E85 GEM
Figure
4. 4. Ignitiontiming
Ignition timing swings atat test point
point AA for
for E85
E85and
andthe
theGEM
GEMblend
blendcompared with thethe
base
Figure
Figure 4. Ignition timingswings
swings at test
test point A for E85 and the GEM blend compared
compared with
with the base base
95
9595 RON
RON gasoline
gasoline fuel. These results are all at a matched engine speed of 2000 rpm and with 40 °C ◦ C
RON gasolinefuel.fuel.These
These results
results are all at
at aa matched
matchedengine
enginespeed
speedofof2000
2000rpm
rpm and
and with
with 40 40
°C
temperatureand
temperature and 130kPa kPa gaugepressure
pressure in the
the plenum but
butwithout
withoutany
anycooled EGR. Note that thethe
temperature and130 130 kPagauge
gauge pressure inin plenum
plenum but without any cooled
cooled EGR.
EGR. Note
Note that
that the
engine load (i.e., BMEP) varied with spark advance. See text for discussion on accuracy of the results.
engine load
engine load (i.e.,
(i.e.,BMEP)
BMEP)varied
variedwith
withspark
spark advance.
advance. SeeSeetext
textfor
fordiscussion
discussionononaccuracy
accuracy
of of
thethe results.
results.
26.5
26.5
26.0
26.0
25.5
25.5
25.0
25.0
/ [bar]
BMEP/ [bar]
24.5
24.5
24.0
BMEP
24.0
23.5
23.5
23.0
23.0
22.5
22.5
22.0
22.0 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Base Fuel E85 GEM
Base Fuel E85 GEM
Figure
Figure5. 5.Ignition
Ignitiontiming
timingswings
swings at
at test
test point
point DD for
for E85
E85andandthe
theGEM
GEMblend
blendcompared
compared with
with thethe base
base
Figure
9595
RON 5. Ignition
gasoline timing swings at test point D for E85 and the GEM blend compared with the base ◦
RON gasolinefuel.
fuel.These
These results
results are all at
are all at aa matched
matchedengine
enginespeed
speedofof3000
3000rpm
rpmandand with
with 60 60
°C C
95 RON gasoline fuel. These results are all at a matched engine speed of 3000 rpm and with 60 °C
temperature
temperatureand and140140kPa
kPagauge
gaugepressure
pressure in in the plenumwith
the plenum with10%
10%cooled
cooledEGR.
EGR.Note
Note that
that thethe engine
engine
temperature and 140 kPa gauge pressure in the plenum with 10% cooled EGR. Note that the engine
load (i.e.,
load BMEP)
(i.e., BMEP)varied
variedwith
withspark
sparkadvance. See text
advance. See text for
fordiscussion
discussionon
onaccuracy
accuracy ofof the
the results.
results.
load (i.e., BMEP) varied with spark advance. See text for discussion on accuracy of the results.
It Itis isapparent
apparentthat
thathere
herethe
the two
two high-proportion alcoholblend
high-proportion alcohol blendfuels
fuelsdid
didnot notperform
perform quite
quite
It is apparent that here the two high-proportion alcohol blend fuels did not perform quite
identically. They both reached nearly the same maximum BMEP, with E85 being slightly
identically. They both reached nearly the same maximum BMEP, with E85 being slightly higher, but higher,
identically. They both reached nearly the same maximum BMEP, with E85 being slightly higher, but
interestingly the GEM blend generates approximately 0.6 bar more BMEP for any given amount of
interestingly the GEM blend generates approximately 0.6 bar more BMEP for any given amount of
spark advance up to its knock-limited spark advance (KLSA) limit (this being the point at which the
curves end, i.e., at these high BMEP conditions minimum park advance for best torque (MBT) could
not be reached before this point, in spite of the high knock resistance of the fuels). It is presumed that
the reason that E85 can generate slightly more BMEP is the fact that it can travel sufficiently further
the reason that E85 can generate slightly more BMEP is the fact that it can travel sufficiently further
Energies 2020, 13, 6390 11 of 31
but interestingly the GEM blend generates approximately 0.6 bar more BMEP for any given amount of
theEnergies
reason 2020,
that13,E85
x FORcanPEER REVIEWslightly more BMEP is the fact that it can travel sufficiently further
generate 11 of 32up
itsup
spark curvecurve
its spark before it reaches
before it reachesits its
KLSA;
KLSA; this isislikely
this likelyaaresult
result of its higher
of its higherRON RONand andMON MON values
values
up its spark curve before it reaches its KLSA; this is likely a result of its higher RON and MON values
as as
shown in Table 2. The reason that the GEM blend generates more
shown in Table 2. The reason that the GEM blend generates more BMEP at lower values of spark BMEP at lower values of spark
as shown in Table 2. The reason that the GEM blend generates more BMEP at lower values of spark
advance
advance is is
possibly
possiblydue duetotoaacombination
combination of of factors, includingthe
factors, including thefact
factthat
thatthe
theexhaust
exhaust backback pressure
pressure
advance is possibly due to a combination of factors, including the fact that the exhaust back pressure
(EBP)
(EBP) was
wasslightly
slightlylower
lowerforfor the GEM blend
the GEM blendthanthanthe theE85.
E85. This
This is shown
is shown in Figure
in Figure 6 for6test
forpoint
test point
A,
(EBP) was slightly lower for the GEM blend than the E85. This is shown in Figure 6 for test point A,
A,where
whereititcan canalso
alsobebeseenseenthat
thatthethegasoline
gasoline case actually
case actuallyhashas
the the
lowest EBPEBP
lowest of all. of For
all. test
For point D
test point
where it can also be seen that the gasoline case actually has the lowest EBP of all. For test point D
Figure
D Figure 7 shows the same thing, and the shape of the EBP curve for the GEM fuel here can be seen
Figure 77 shows
shows the the same
samething,
thing,and andthetheshape
shapeofofthe theEBP
EBPcurve
curveforforthethe GEM
GEM fuelfuel here
here cancan
bebe
seenseen
reflected
reflected in in theBMEP
the BMEPcurve curveininFigure
Figure5, 5, lending
lending credence
credence to
tothis
this hypothesis;
hypothesis; it itshould
should bebenoted
notedthatthat
reflected in the BMEP curve in Figure 5, lending credence to this hypothesis; it should be noted that
thethe absolutedifference
absolute differencebetween
betweenthe the EBP
EBP values
values for
for test
test point
point DDisismuch
much
the absolute difference between the EBP values for test point D is much smaller than it was for test
smaller
smaller than
thanit it
waswasforfor
testtest
point
point A.A. TheThe difference
difference between
between thethe alcohol
alcohol fuel
fuel cases
cases was
was unfortunatelynot
unfortunately not discovereduntil until afterthe
point A. The difference between the alcohol fuel cases was unfortunately notdiscovered
discovered untilafter after
the
tests tests had been completed.
thehad
testsbeen
had completed.
been completed.
45.5
45.5
45.0
Gauge]
45.0
Gauge]
44.5
44.5
/ [kPa
44.0
/ [kPa
44.0
Pressure
43.5
Pressure
43.5
43.0
43.0
Back
Back
42.5
42.5
Exhaust
Exhaust
42.0
42.0
41.5
41.5
41.0
41.0 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4 5
-8 -7 -6 -5 -4 Spark
-3 Advance
-2 / -1 0
[deg BTDC] 1 2 3 4 5
Base Fuel E85 GEM
Base Fuel E85 GEM
Figure
6. 6. Exhaustback
Exhaust back pressureforfor the data
data presented in
inFigure
Figure4, taken
takenat test point A.A.
SeeSee
text forfor
Figure
Figure 6. Exhaust backpressure
pressure forthethe data presented
presented in Figure 4,4,taken atattest
test point
point A. See text
text for
discussion
discussion on
onon accuracy
accuracy of the results.
discussion accuracyofofthe
theresults.
results.
119.5
119.5
Gauge]
/ [kPaGauge]
119.0
119.0
118.5
Pressure/ [kPa
118.5
118.0
Pressure
118.0
117.5
117.5
Back
ExhaustBack
117.0
117.0
Exhaust
116.5
116.5
116.0
116.0 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20
Base Fuel E85 GEM
Base Fuel E85 GEM
Figure Exhaust
7. 7.
Figure Exhaustback
backpressure
pressurefor
for the
the data
data presented inFigure
presented in Figure6,6,taken
takenatattest
test point
point D.D.
SeeSee text
text forfor
Figure
discussion7. Exhaust
onon back
accuracy pressure for the data presented in Figure 6, taken at test point D. See text for
discussion accuracyofofthe
theresults.
results.
discussion on accuracy of the results.
The difference in plenum pressure and temperature across all of the test results shown in Figures
The difference in plenum pressure and temperature across all of the test results shown in Figures
4–7 was less than 0.3% and 1 °C, respectively, indicating the excellent accuracy with which the CAHU
4–7 was less than 0.3% and 1 °C, respectively, indicating the excellent accuracy with which the CAHU
could hold the intake conditions constant. Note that the valve timing was held constant for each test
condition and as a consequence the difference in back pressure is seen as the main reason for the
Energies 2020, 13, 6390 12 of 31
The difference in plenum pressure and temperature across all of the test results shown in Figures 4–7
was less than 0.3% and 1 ◦ C, respectively, indicating the excellent accuracy with which the CAHU
difference in performance between E85 and the GEM blend, because it would lead to a difference in
volumetric efficiency. It is not believed that the GEM blend would in itself give better performance,
although another difference between the two high-blend alcohol fuels is that different gasoline blend
stock was used for each. This is thought to be one of the reasons why the octane numbers of these
two fuels vary by 1.4 units, too (as shown in Table 2), as well as the difference in stoichiometric AFR.
Whether it is calculated on a mass or volume basis, the E85 and ternary blend will have essentially the
same enthalpy of vaporization, as was discussed separately by Chupka et al. for ethanol blends [48]
and by Turner et al. for GEM blends [17].
Regardless of the above discussion, it is apparent that the two high-blend alcohol fuels allow
the engine to generate significantly more load for the same plenum conditions than when it operates
on the baseline fuel. A total of four different test points (shown in Table 3 and including A and D
mentioned above) were used to compare fuels during this part of the Ultraboost program, and the
rationale for selecting these has been discussed in previous publications [24,43–45]. Salient results for
these conditions are also presented in Figures 8–10.
Table 3. Test operating points for Ultraboost engine testing.
Engine Boost Plenum

Test Load EGR Exhaust Back
Speed Pressure Temperature Notes
Point Condition (%) Pressure (Bar)
(rpm) (Bar Gauge) (◦ C)
A 2000 High 0 1.3 40 0.45
Boost pressure
increased to
B 2000 High 10 1.5 40 0.45
compensate for
presence of EGR
Mimics supercharged
C 3000 Mid-High 10 1.4 60 0.7
operation
Mimics turbocharged
D 3000 Mid-High 10 1.4 60 1.2
operation
Test point A was intended to represent a highly downsized engine operating near to the torque
curve “knee point” but with no cooled EGR being used for knock mitigation. Test point B was
effectively point A with the intake boosted to account for 10% external EGR being used by keeping the
amount of oxygen flowing through the engine constant; the exhaust back pressure was, however, kept
equal between these points. Points C and D are higher speed conditions but with markedly different
back pressure. A supercharged engine condition was mimicked by point C, while point D intended
to be representative of a turbocharged one with higher exhaust port back pressure; both of these use
10% EGR. Since the plenum conditions were the same for points C and D comparison between these
was intended to see what the effect of increased hot residuals retention had on the knock limit. It is
accepted, however, that with these two test points there is a conflating effect of reduced volumetric
efficiency, even though they shared the same, unvarying, valve timing.
Within the discussion above regarding the slight difference in back pressure for the GEM fuel
test, the results shown in Figures 8–10 essentially show the expected response. The high-alcohol
blends generate significantly greater torque at all test points than the two gasolines. At the same time,
the BSFC for the alcohol blends is significantly higher than the two gasolines. However, the brake
thermal efficiency for the alcohol blends is also significantly better: generally 1–2% better in absolute
terms, and approximately 6% better at test point B. Part of the reason for this will be due to the higher
molar expansion ratios of the alcohols versus gasoline, and thus their blends. The importance of this
characteristic has been discussed by Szybist et al. [49] and Nguyen et al. [50].
Energies 2020, 13, x6390
Energies FOR PEER REVIEW 13 of
13 of 32
31
540
540
520
520
500
500
480
480
/ [Nm]
460
/ [Nm]
460
440
Torque
440
Torque
420
420
400
400
380
380
360
360
340
340 A B C D
A Base Fuel B RON E5 Pump
95 C
E85 GEM D
Base Fuel 95 RON E5 Pump E85 GEM
Figure 8. Maximum torque for the base and 95 RON pump gasolines, E85 and the GEM blend, at four
Figure 8. Maximum torque for the base and 95 RON pump gasolines, E85 and the GEM blend, at
Figure 8. Maximum
different torque
fixed-intake for the base
conditions, and 95atRON
operated KLSApump gasolines,
ignition E85For
timing. and description
the GEM blend, at four
of the test
four different fixed-intake conditions, operated at KLSA ignition timing. For description of the test
different fixed-intake conditions, operated at KLSA ignition timing. For description of
operating points, labelled A to D here, see Table 3 and text. Note that the two high-alcohol blends the test
operating points, labelled A to D here, see Table 3 and text. Note that the two high-alcohol blends have
operating points, labelled
have nearly-matched A to D here,and
stoichiometry see enthalpy
Table 3 and
of text. Note that RON
vaporization, the two high-alcohol
and MON, andblends
as a
nearly-matched stoichiometry and enthalpy of vaporization, RON and MON, and as a consequence
have nearly-matched
consequence perform instoichiometry
an essentiallyand enthalpy
identical of
manner. vaporization, RON and MON, and as a
perform in an essentially identical manner.
consequence perform in an essentially identical manner.
400
400
380
380
360
360
340
340
/ [g/kWhr]
320
/ [g/kWhr]
320
300
300
BSFC
280
BSFC
280
260
260
240
240
220
220
200
200 A B C D
A Base Fuel B RON E5 Pump
95 C
E85 GEM D
Figure 9.
Figure BSFCat
9. BSFC atmaximum
maximumtorque
torquefor
forthe
thebase
baseand
and95 95RON
RONpump
pumpgasolines,
gasolines,E85
E85and
andthe
theGEM
GEMblend,
blend,
at four
Figure different
9. BSFC at fixed-intake
maximum conditions,
torque for theoperated
base and at
95KLSA
RON ignition
pump timing.
gasolines,For
E85description
and the of
GEM
at four different fixed-intake conditions, operated at KLSA ignition timing. For description of the test the test
blend,
operating
at four points,
different labelled
fixed-intakeA to D here,
conditions,see Table
operated 3 and
at text.
KLSA Note that
ignition the two
timing. high-alcohol
For blends
description
operating points, labelled A to D here, see Table 3 and text. Note that the two high-alcohol blends of the have
test
nearly-matched
operating points,stoichiometry
labelled A to and enthalpy
D here, of vaporization,
see enthalpy
Table 3 and RONthatand the
MON, and as a consequence
have nearly-matched stoichiometry and of text. Note
vaporization, RONtwo high-alcohol
and MON, andblends as a
perform
have in an essentially
nearly-matched identical manner.
stoichiometry and enthalpy of vaporization, RON and MON, and as a
Energies 2020, 13, 6390 14 of 31
42
Brake Thermal Efficiency / [%] 40
38
36
34
32
30
28
A B C D
Figure 10.
Figure Brakethermal
10. Brake thermal efficiency
efficiency at
at maximum
maximum torque
torque for
for the
the base
base and
and 95
95 RON
RON pump
pump gasolines,
gasolines, E85
E85
and the GEM blend, at four different fixed-intake conditions, operated at KLSA ignition timing. For
and the GEM blend, at four different fixed-intake conditions, operated at KLSA ignition timing. For
description of the test operating points, labelled A to D here, see Table 3 and text. Note that the two
description of the test operating points, labelled A to D here, see Table 3 and text. Note that the two
high-alcohol blends have nearly-matched stoichiometry and enthalpy of vaporization, RON and MON,
high-alcohol blends have nearly-matched stoichiometry and enthalpy of vaporization, RON and
and as a consequence perform in an essentially identical manner.
MON, and as a consequence perform in an essentially identical manner.
The results for this section reinforce that the blending rules for ternary blends discussed
The results for this section reinforce that the blending rules for ternary blends discussed in [17–
in [17–20,22] are entirely borne out. As a counterpoint, it is interesting to note that the work of
20,22] are entirely borne out. As a counterpoint, it is interesting to note that the work of Waqas et al.
Waqas et al. [23] suggested that the octane of the gasoline used for the blend can have an effect in
[23] suggested that the octane of the gasoline used for the blend can have an effect in that a high RON
that a high RON base fuel allows linearity in octane numbers, but that a low RON base fuel yields
base fuel allows linearity in octane numbers, but that a low RON base fuel yields a non-linear
a non-linear behaviors. However, in this section, the octane numbers of the blends were directly
behaviors. However, in this section, the octane numbers of the blends were directly measured.
measured. Nevertheless, Waqas et al.’s observations have obvious implications when splash blending
Nevertheless, Waqas et al.’s observations have obvious implications when splash blending such
such ternary blends. This will be mentioned in later sections.
ternary blends. This will be mentioned in later sections.
Some particulate emissions results are now reported here since they were taken as part of the
Some particulate emissions results are now reported here since they were taken as part of the
Ultraboost project fuels investigation; Section 5 reports further results from a single-cylinder engine.
Ultraboost project fuels investigation; Section 5 reports further results from a single-cylinder engine.
Figure 11 shows the particulate emissions from the base fuel, E85 and GEM fuels at these test conditions.
Figure 11 shows the particulate emissions from the base fuel, E85 and GEM fuels at these test
These were measured using a Cambustion DMS500 sampling at 1 Hz, at steady state, over 30 s and
conditions. These were measured using a Cambustion DMS500 sampling at 1 Hz, at steady state, over
the mean and standard deviation (σ) of the data over that time were taken, with the methodology
30 s and the mean and standard deviation (σ) of the data over that time were taken, with the
being equivalent to that discussed in Section 5 and fully described in [47]. The results show that at
methodology being equivalent to that discussed in Section 5 and fully described in [47]. The results
the four different operating points, the E85 fuel has the lowest PN emissions and the GEM fuel has
show that at the four different operating points, the E85 fuel has the lowest PN emissions and the
higher particulate emissions compared to the base fuel. The high PN emissions from the GEM fuel
GEM fuel has higher particulate emissions compared to the base fuel. The high PN emissions from
are thought to be caused by the rapid evaporation of that fuel upon injection due to the presence of
the GEM fuel are thought to be caused by the rapid evaporation of that fuel upon injection due to the
the high-vapor-pressure methanol. This rapid evaporation then causes locally rich regions to form,
presence of the high-vapor-pressure methanol. This rapid evaporation then causes locally rich
which leads to higher particulate emissions. A fuller analysis of these particulate emissions is included
regions to form, which leads to higher particulate emissions. A fuller analysis of these particulate
in [46,47,51].
emissions is included in [46,47,51].
As a consequence of these tests, one can state that, as expected from previous work, the alcohols
As a consequence of these tests, one can state that, as expected from previous work, the alcohols
in general perform better than normal pump gasoline in terms of KLSA, maximum brake output,
in general perform better than normal pump gasoline in terms of KLSA, maximum brake output, and
and thermal efficiency, and with the expected disadvantage in terms BSFC. The results gathered here,
thermal efficiency, and with the expected disadvantage in terms BSFC. The results gathered here,
allowing for the difference in operating conditions (specifically, the higher back pressure evident when
allowing for the difference in operating conditions (specifically, the higher back pressure evident
the GEM blend was tested), show the potential for GEM blends to perform at least as well as the binary
when the GEM blend was tested), show the potential for GEM blends to perform at least as well as
gasoline-ethanol blends whose stoichiometric AFR they have been matched to, as has been shown in
the binary gasoline-ethanol blends whose stoichiometric AFR they have been matched to, as has been
the authors’ previous work. These results are wholly in line with the findings of the first report on
shown in the authors’ previous work. These results are wholly in line with the findings of the first
this alcohol fuel research using the Ultraboost research engine [9]. However, it must be noted that the
report on this alcohol fuel research using the Ultraboost research engine [9]. However, it must be
Energies 2020, 13, 6390 15 of 31
notedofthat
presence the presence
methanol of methanol
in the GEM fuel leadsintothe GEMperformance
inferior fuel leads towith
inferior performance
respect with
to particulate respect to
number,
and particulate
so more worknumber,
needsand socarried
to be more work
out inneeds to be carried out in this area.
this area.
Figure 11. PN emissions at maximum torque for standard gasoline and two high-alcohol blends at
four Figure 11.fixed-intake
different PN emissions at maximum
conditions. torque
The GEM for standard
blend gasoline
has the highest PNand two high-alcohol
emissions. The errorblends
bars at
four different
correspond to ±σ. fixed-intake conditions. The GEM blend has the highest PN emissions. The error bars
correspond to ±σ.
Nevertheless, it is interesting to reflect that through this approach methanol clearly has a better
blendingNevertheless,
effect with gasolineit is interesting
than ethanol. to reflect
Thethat through for
justification thisthis
approach
assertion methanol clearlyE85
is as follows. has anda better
M56blending
are matched effect
forwith gasoline
the same than ethanol.
stoichiometric AFRThe andjustification
are known from for this assertion
earlier work to is as
have follows. E85 and
essentially
M56 properties,
identical are matched for the latent
including same heat,stoichiometric
volumetricAFR LHV, and
RON,are MON,
knownand from earlier work
sensitivity to have
[17–20,22].
essentially identical properties, including latent heat, volumetric
Previous researchers have discussed that rather than use liquid volume proportions to compare LHV, RON, MON, and sensitivity
[17–20,22].
performance of Previous
such fuels,researchers
using the molar have ratio
discussed
wouldthat rather
provide than use
a better basis, liquid
sincevolume
combustion proportions
occurs to
compare
in the gas phase performance
[11,24]. Theofmolar suchratiosfuels,inusing
these thefuelsmolar ratio calculated
have been would provide using aafuels better basis, since
properties
combustion
calculator occursto
proprietary inthe
theUniversity
gas phase [11,24].
of Bath The [52] molar ratios
and using inlogic
the thesepropounded
fuels have been calculated
in [24], with the using
same a fuels properties
properties calculator
for gasoline used proprietary
throughout to the University
calculations, ofi.e.,
Bathwhich
[52] and is notusingthe the
caselogic propounded
in the actual
fuelsinblended
[24], with theAccepting
here. same properties for gasoline
this limitation, used throughout
the results show that the theethanol
calculations, i.e., which
mole fraction is not
in E85 is the
0.94,case in thethe
whereas actual fuels blended
methanol here. Accepting
molar concentration this limitation,
in M56 the results
is 0.83. Ceteris paribus, show sincethatit the
canethanol
producemole
fraction
the same in E85
effect as a is 0.94, whereas
higher the methanol
molar proportion molarcan,
of ethanol concentration
the lower molarin M56 is 0.83. Ceteris
proportion paribus,issince
for methanol
proofit can
of itsproduce the sameeffect
better blending effectin asgasoline.
a higher molarUsing proportion of ethanolincan,
the same calculator, the the
GEM lower
blendmolartestedproportion
here
for methanol
the ethanol is proof ofmole
and methanol its better
fractionsblending
are 0.24 effect
andin0.63,
gasoline. Using the
respectively, samea calculator,
giving total alcohol in mole
the GEM
blend
fraction of tested here
0.87, i.e., the ethanol
between the two and methanol
binary blends, mole fractions
which are 0.24
is logically what andone 0.63,
wouldrespectively,
expect. giving a
total
Thus,alcohol mole fraction
if one wants primarily ofto0.87, i.e., between
improve performance the two
of a binary blends, which
gasoline-alcohol is logically
fuel, methanol what one
appears
to bewould expect.
the better blend component of the two alcohols. It is believed that this is borne out by the initial
results ofThus, if one wants
the Co-Optima projectprimarily to improve
in the United States. performance of a gasoline-alcohol
This project, which is looking at biofuel fuel,options
methanol
appearsenergy
to improve to be the betterfor
security blend
the component
US and which of the two alcohols.
is looking at mediumIt is believed
blend rates that(below
this is30%),
bornehas out by
foundthethat
initial results
of the of the
various Co-Optima
different chemical project in thetried,
families Unitedthe States.
alcohols This areproject,
the most which is looking
desirable [53]. at
biofuel
Indeed options
in [54], to improve
in terms of RON, energy security for the US
for a given-percentage and which
blend-rate is looking
methanol was at shown
medium to blend
have arates
(below
better 30%),than
response has found
ethanol thatwhenof the various different
expressed up to ~20% chemical familiesFrom
by volume. tried,the the results
alcohols are the most
discussed
heredesirable
at higher [53].blendIndeed
rates, duein [54],
to theinmarked
terms of RON, for
reduction a given-percentage
in the volume of methanol blend-rate
requiredmethanol
to obtainwas
shownengine
the same to haveefficiency
a better response
as ethanol than ethanol
when when expressed
in admixture up to ~20%
with gasoline by volume.
discussed above,From it the
seems results
discussed
self-evident tohere at higherthat
the authors blend rates, of
in terms due to the marked
performance reduction
methanol is ainmuch
the volume of methanol
more favorable required
additive
to obtain the same engine efficiency as ethanol when in admixture
when compared with ethanol for transport fuel. As mentioned earlier, it can also extend the impact with gasoline discussed above,
of it
seems
a given fixedself-evident
volume of to the authors
ethanol (this beingthat limited
in terms byofbiomass
performance methanol
availability, is a much
for example) whenmore favorable
blended
additive when compared with ethanol for transport fuel. As mentioned earlier, it can also extend the
impact of a given fixed volume of ethanol (this being limited by biomass availability, for example)
Energies 2020, 13, 6390 16 of 31
together with gasoline, as has been discussed by Turner et al. [17]. Ternary blending of the three
components can also make the distillation curve more acceptable than binary blending, overcoming
one of the drawbacks listed by the Co-Optima project. As the Co-Optima project moves forward,
methanol has been eliminated as a candidate blend agent, primarily on the grounds of blend stability
at lower concentrations than discussed here. To the authors of this work methanol would seem to be a
preferred choice when considering ternary blends and with engine performance as the sole criterion;
unfortunately there has not been such consideration of ternary blends within Co-Optima. The effect of
ternary blending of gasoline with the light alcohols on engine performance will be further investigated
in the following sections of the present paper.
4. Performance and Emissions of Iso-Stoichiometric GEM Blends in Production Engines and

Blend Knock-Limited Spark Advance Tests
Ghent University has operated several engines using pure methanol in a research stream
investigating the desirable characteristics of methanol combustion in optimized engines. The alcohols
in general have the potential to increase power and efficiency of SI engines [55,56], and their interesting
properties allow different load control strategies to be used as well [57]. These studies have shown that
any potential improvement is dependent upon the method of engine operation, but with this biased
towards the alcohol such benefits can be provided while still maintaining the functional advantage of
flex-fuel capability (i.e. so that gasoline could still be used, albeit with lower performance or efficiency,
when the alcohol fuel is unavailable). To investigate this concept further, three naturally-aspirated
engines were operated on gasoline, methanol and several GEM blends. These were a 4-cylinder PFI
production-type Volvo engine, a single-cylinder research engine based on an Audi PFI engine, and a
4-cylinder Hyundai production engine utilizing DI. Details of these engines are shown in Table 4.
Table 4. Test engine specifications.
Engine Type Volvo 1.8 L Audi/NSU Hyundai

Cylinders 4 in-line 1 4 in-line
Valves 16 2 16
Double overhead Overhead camshaft Double overhead
Valvetrain
camshaft (DOHC) (OHC) camshaft (DOHC)
Bore (mm) 83 77.5 88
Stroke (mm) 82.4 86.4 97
Displacement (cc) 1783 407.3 2360
Compression ratio
10.3 13.1 11.3
(:1)
Fuel system PFI PFI DI
Induction Atmospheric Atmospheric Atmospheric
Engine management
MoTeC M800 MoTeC M4 Hyundai
system
All the measurements presented here were taken at steady-state operating conditions at various
speeds and loads. Stoichiometric operation was adopted in order to enable the use of a three-way
catalyst (TWC). For the Hyundai engine, fitted with DI, tests were conducted with early injection.
This resulted in a homogeneous mixture using a closed-loop feedback to control the air/fuel ratio
to stoichiometric conditions. MBT spark timing was applied, except for knock-limited operating
conditions, where KLSA-2◦ crank angle was used.
The notion that blends of methanol and ethanol with gasoline with the same stoichiometry as E85
can be used in flex-fuel engines has been investigated on the Volvo 1.8 L engine and Hyundai 2.4 L.
The performance, efficiency and emissions of different GEM blends were measured and compared to
Energies 2020, 13, 6390 17 of 31
those on the pure constituent components to evaluate their potential. Additionally, the Audi engine
was used for a preliminary study of the knock behavior of these blends because it had a higher CR.
The blends
Energies 2020, 13, tested in the
x FOR PEER 1.8 L Volvo engine are shown in Table 5. All were splash
REVIEW blended
17 of 32
using a conventional 95 RON pump gasoline. (Note that the blends tested on the Hyundai engine
wereengine was used
very similar infor a preliminary
terms of volume study of the knock
percentages butbehavior
used a of these blendsEEE
Haltermann because it had a higher
certification gasoline.)
CR.
In Table 5, there are two binary blends: Blend A represents an idealized ‘normal’ E85 and Blend D
The blends testedgasoline-methanol
is the iso-stoichiometric in the 1.8 L Volvo engine are shown
equivalent. in Table
This 5. All
has 57% v/v were splash blended
methanol, 1% v/vusing
methanol
a conventional 95 RON pump gasoline. (Note that the blends tested on the Hyundai engine were very
more than in the earlier study of Turner et al. [22]. This difference is because of the stoichiometric AFR
similar in terms of volume percentages but used a Haltermann EEE certification gasoline.) In Table
of the gasoline in this blend. The two remaining blends in Table 5 are true ternary blends. Blend B has
5, there are two binary blends: Blend A represents an idealized ‘normal’ E85 and Blend D is the iso-
half the ethanol ofgasoline-methanol
stoichiometric Blend A, and Blend C has half
equivalent. This that of Blend
has 57% B, and 1%
v/v methanol, in each case, themore
v/v methanol gasoline
than and
methanol
in thecomponent
earlier studyvolumes
of Turnerhas been
et al. [22].adjusted to maintain
This difference the target
is because of the stoichiometry.
stoichiometric AFR of the
gasoline in this blend. The two remaining blends in Table 5 are true ternary blends. Blend B has half
Table 5.ofProperties
the ethanol Blend A, of theBlend
and GEM Cblends tested
has half onof
that theBlend
4-cylinder
B, and1.8inLeach
PFI production engine. and
case, the gasoline
methanol component volumes has been adjusted to maintain the target stoichiometry.
Fuel Blend Blend A Blend B Blend C Blend D
G15 on the 4-cylinder
Table 5. Properties of the GEM blends tested G37
G29.5 1.8 L PFI production G43
engine.
Component ratios (% v/v) E85 E42.5 E21 E0
Fuel Blend M28 B Blend CM42Blend D
M0 Blend A Blend M57
Oxygen content (% m/m) 23.34 G15 G29.5
22.74 G37 22.38 G43 22.54
Component ratios (% v/v) E85 E42.5 E21 E0
Gravimetric LHV (MJ/kg) 29.22 29.48 29.66 29.53
M0 M28 M42 M57
VolumetricOxygen
LHV (MJ/L)
content (% m/m) 22.82 23.34 22.79
22.74 22.38 22.81 22.54 22.59
AFRstoich (kg/kg) LHV (MJ/kg) 9.72 29.22
Gravimetric 9.76
29.48 29.66 9.8 29.53 9.73
Volumetric
∆hvap (kJ/kg) LHV (MJ/L) 762.4 22.82 22.79
762.7 22.81 761.5 22.59 770.8
AFRstoich (kg/kg) 9.72 9.76 9.8 9.73
α (H/C in fuel) 2.83 2.86 2.87 2.9
∆hvap (kJ/kg) 762.4 762.7 761.5 770.8
Specific CO2 emissions (gCOin
α (H/C 2 /MJ)
fuel) 71.69 2.83 71.56
2.86 2.87 71.49 2.9 71.41
Specific CO2 emissions (gCO2/MJ) 71.69 71.56 71.49 71.41
Apparent in this data is the fact that as the gasoline proportion increases so the specific CO2
emissionsApparent
(reportedininthis
gCOdata is the fact that as the gasoline proportion increases so the specific CO2
2 /MJ) decrease; this is because methanol has the lowest carbon intensity in
emissions (reported in gCO2/MJ) decrease; this is because methanol has the lowest carbon intensity
terms of gCO2 /MJ of the three blend components, and more methanol has to be added as the gasoline
in terms of gCO2/MJ of the three blend components, and more methanol has to be added as the
content rises.
gasoline content rises.
4.1. Performance
4.1. Performance
First,First,
the results from
the results from thethe
Volvo
Volvo1.81.8LLengine willbe
engine will bediscussed.
discussed. In In
thisthis engine,
engine, fourfour different
different GEM GEM
blends werewere
blends tested forfor
tested which
which thethebrake
brakethermal efficiency(BTE)
thermal efficiency (BTE) is is shown
shown in Figure
in Figure 12.this
12. In Infigure,
this figure,
the BTE is shown
the BTE is shownfor for
all all
blends
blends atata aconstant
constant load of40
load of 40NmNm(2.82
(2.82
barbar BMEP)
BMEP) overover a speed
a speed rangerange
from from
1500 1500 to 3500
to 3500 rpm. rpm. Within
Within thethe rangeofofexperimental
range experimental uncertainty,
uncertainty,allallof of
thetheiso-stoichiometric
iso-stoichiometric blends
blends
have similar BTEs. Furthermore, this is valid for all of the
have similar BTEs. Furthermore, this is valid for all of the loads tested. loads tested.
Figure 12. BTE of various iso-stoichiometric GEM blends as a function of engine speed for a fixed
Figure 12. BTE of various iso-stoichiometric GEM blends as a function of engine speed for a fixed
brake torque of 40 Nm (2.82 bar BMEP).
brake torque of 40 Nm (2.82 bar BMEP).

Figure
Energies 136390
2020, 13, displays a comparison of the BTEs for an average for the GEM blends with18their of 31
gasoline, methanol,
Figure ethanol
13 displays componentsofatthe
a comparison fixed loads
BTEs forofan40average
and 80 Nm. ThisGEM
for the was performed
blends withfortheir
the
GEM blends
gasoline, because
methanol, their
ethanol results shown
components in
atBTEsFigure
fixedfor
loads12 are within
of 40 and experimental uncertainty. When
Figure 13 displays a comparison of the an average for80 Nm.
the GEM This was with
blends performed for the
their gasoline,
compared
GEM to gasoline, it is clear thatshown
the GEM blends 12
show significant efficiency gains, being close to
methanol, ethanol components at fixed loads of 40 and 80 Nm. This was performed for theWhen
blends because their results in Figure are within experimental uncertainty. GEM
that of
comparedpure ethanol.
to gasoline, The best for
it is clear BTE is
thatinthe methanol,
GEM12 which
blends might be
showexperimentalexpected from
significant efficiency the
gains,results
being already
close to
blends because their results shown Figure are within uncertainty. When compared
discussed.
that of pure ethanol. The best for BTE is methanol, which might be expected from the results already
to gasoline, it is clear that the GEM blends show significant efficiency gains, being close to that of pure
discussed.
ethanol. The best for BTE is methanol, which might be expected from the results already discussed.
Figure 13. Brake thermal efficiency when operating on various fuels as a function of engine speed for
Figure 13. Brake thermal efficiency when operating on various fuels as a function of engine speed for
different
Figure 13.fixed
Brakebrake torques
thermal of 40 when
efficiency Nm (2.82 bar BMEP)
operating (dashed
on various lines)
fuels as aand 80 Nm
function (5.64 bar
of engine BMEP)
speed for
different fixed brake torques of 40 Nm (2.82 bar BMEP) (dashed lines) and 80 Nm (5.64 bar BMEP)
(solid lines).
different fixed brake torques of 40 Nm (2.82 bar BMEP) (dashed lines) and 80 Nm (5.64 bar BMEP)
(solid lines).
(solid lines).
Because
Because ofof the
the near-identical
near-identical LHV LHV of of the
the four
four iso-stoichiometric
iso-stoichiometricGEMGEMblends
blends(see
(seeTable
Table 5)
5) thethe
BSFC
BSFC results
Because do
resultsof not
dothe depend on
notnear-identical mixture composition.
LHV of composition.
depend on mixture This is
the four iso-stoichiometric shown
This is shown in
GEM Figure 14.
blends14.
in Figure Figure
(seeFigure 155) isthe
Table 15 is aa
comparison
BSFC resultsof
comparison do
of the BSFCs
not
the of
depend
BSFCs of the
on components
the mixture
components and
andthe
composition. themean
This of
mean ofthe
is theGEM
shownGEMin blends
Figure for
blends for40
14. and
and80
Figure
40 80Nm.
15 is
Nm. a
Despite
comparison better
Despite better efficiency
of the with
BSFCswith
efficiency of theGEM
GEM blends,
components this
blends, this engine
andengine consumes
the mean ~32%
of the ~32%
consumes more
GEM blends fuel
more fuel by
for by volume
40 volume than
and 80 Nm.than
when
Despite operating
better on GEM
efficiency mixtures
with GEM than when
blends, it
thisis using
engine gasoline.
consumes Because
~32% of the
more properties
fuel
when operating on GEM mixtures than when it is using gasoline. Because of the properties of alcohols by of
volumealcohols
than
this
this difference
when operatingwould
difference on GEM
would be
be expected
mixtures to
expected thanto be smaller
bewhen
smallerit isfor modern
using
for highly-downsized
gasoline.
modern Because of theand
highly-downsized andpressure-charged
properties of alcohols
pressure-charged
direct-injection
this (DI)
difference would
direct-injection engines. With
be expected
(DI) engines. With a full
toabe optimization
smaller
full for modern
optimization of such an engine,
highly-downsized
of such the situation
an engine, the and would
would inin all
pressure-charged
situation all
likelihood be
direct-injectionfurther
(DI) improved.
engines.
likelihood be further improved. With a full optimization of such an engine, the situation would in all
likelihood be further improved.
Figure 14. BSFC of GEM blends as a function of engine speed for a fixed brake torque of 40 Nm (2.82
bar BMEP).
bar BMEP).
bar BMEP).
Energies 2020, 13, 6390 19 of 31

Figure 15. BSFC as a function of engine speed for different fixed brake torques of 40 Nm (dashed
lines)
Figureand
15.80 Nmas
BSFC (solid lines). of engine speed for different fixed brake torques of 40 Nm (dashed lines)
a function
Figure 15. BSFC as a function of engine speed for different fixed brake torques of 40 Nm (dashed
and 80 Nm (solid lines).
lines)
In theand 80 Nm (solid
following lines).
section, the focus will be on the difference between the two binary GEM blends
In the following
as measured on the section, the focus
direct-injected will be on
Hyundai the difference
engine: E85 andbetweenM56. Note the two thatbinary
becauseGEM a blends
differentas
In the
measured
gasoline wasfollowing
on the
used tosection,
direct-injected theHyundai
formulate focus
the GEM will beblends
engine:on theE85fordifference
andthese between
M56.tests,
Note thethatthe two binary
because
composition ofGEM
a different blends
thegasoline
binary
as
was measured
gasoline-methanol on the
used to formulate direct-injected
blendthehas
GEM blends
changed Hyundai
for theseengine:
slightly tests, the
compared E85 toand
compositionM56. of
the blends Note that
the binary
listed because
in Table a different
gasoline-methanol
5, reducing by
gasoline
blend
1%. From was
has the used
changed to formulate
resultsslightly
showncompared the GEM
in Figuresto12–15 blends
the blends for these
listed inthat
it is expected tests,
Table the
the5,results
reducingcomposition
by 1%.other
of every of
From the
the binary
GEM results
blend
gasoline-methanol
shown there
shown in Figures
would blend
12–15 ithas
be aligned changed
is expected
with the slightly
that compared
the results
results of theofbinary to the
every blends
other
blends GEM listed
tested blend inshown
in this Table
study.5, reducing
there wouldby be
1%. From
aligned the
with results
the shown
results of in
the Figures
binary 12–15
blends it is
tested expected
in this that
study. the
First, measurements were performed using E85 at fixed loads of 50, 75 and 150 Nm across a results of every other GEM blend
shown
speedFirst,there
range would
of 1500 be
measurements to aligned
were
3000 with
rpm, thenthethese
performed results
using
were of
E85the atbinary blends
fixed using
repeated loads of tested
M56.50, In in this
75 order
and tostudy.
150 Nm across
produce a speed
the same
range
torque First,
of measurements
1500
output, tovery rpm,were
3000small then performed
these were
adjustments of theusing
repeated E85using
throttle at fixed
valve M56. loads
were of 50,
Inpermitted
order to 75 andthe
produce
with 150
the Nm
twosame across
torque
fuels, a
but
speed
output,
the range
startvery of
of small 1500 to 3000
adjustments
injection rpm,
timing and then these
of theinjection were
throttle valve repeated
pressure wereand using
permitted M56.
ignition with In order
the two
timing to produce
fuels,kept
were but thethe same
thestart
sameof
torque
injection
throughout.output,In very
timing Figure small
and injection adjustments
16, BTE pressure
for the and of ignition
two the
fuelsthrottle
is timing
shown valve
were were
at thekept permitted
the same
different withloads,
the two
throughout.
engine Infuels,
and Figure but
16,
the only
the
BTEstart
significant of
for the injection
two
difference fuels timing
is greater
(i.e., shown and at injection
than the 1%) canpressure
different engine
be seen atand ignition
loads,
150 Nm and
and the timing
2500 rpm,were
only kept
significant
where the same
difference
the error bars
throughout.
(i.e.,
do notgreater
overlap.than 1%) can be seen at 150 Nm and 2500 rpm, where the error bars do not overlap. only
In Figure 16, BTE for the two fuels is shown at the different engine loads, and the
significant difference (i.e., greater than 1%) can be seen at 150 Nm and 2500 rpm, where the error bars
do not overlap.
Figure 16. Cont.

Energies
Energies 2020,
2020, 13,
13, x6390
FOR PEER REVIEW 20
20of
of32
31
Figure 16. BTE as a function of engine speed for different fixed brake torques of (A) 50 Nm, (B) 75
Figure 16. BTE as a function of engine speed for different fixed brake torques of (A) 50 Nm, (B) 75 Nm,
Nm, and (C)
Figure 150 Nm.
and (C)16.
150BTE
Nm.as a function of engine speed for different fixed brake torques of (A) 50 Nm, (B) 75
Nm, and (C) 150 Nm.
As
As stated,
stated, most
most ofof the
the injection
injection parameters
parameters were
were fixed,
fixed, the
the injection
injection duration
duration being
being the
the only
only
variable.
As Figure
stated, 17
most shows
of thethe difference
injection in this
parameters variable
were between
fixed, the E85 and
injection M56. The
duration difference
being
variable. Figure 17 shows the difference in this variable between E85 and M56. The difference is bigger the is
only
bigger than
variable. 1%
Figure for
17 50 Nm
shows and
the 75 Nm for
difference one
in operating
this variablepoint, whereas
between E85 for
and 150 Nm,
M56. the
The
than 1% for 50 Nm and 75 Nm for one operating point, whereas for 150 Nm, the differences are larger differences
difference is
are larger
bigger
but stillthanbut
only1% still only 2.7%
forat50maximum.
2.7% at maximum.
Nm and 75 Nm for one operating point, whereas for 150 Nm, the differences
are larger but still only 2.7% at maximum.
Figure 17. Difference in injection duration of M56 compared to E85 as a function of engine speed for
Figure 17. Difference in injection duration of M56 compared to E85 as a function of engine speed for
different fixed brake torques of 50, 75, and 150 Nm.
different
Figure 17.fixed brake torques
Difference of 50,duration
in injection 75, and 150 Nm.compared to E85 as a function of engine speed for
of M56
4.2. NOx Emissions
different fixed brake torques of 50, 75, and 150 Nm.
4.2. NOx Emissions
Figure 18 compares the four GEM blends tested on the 1.8 L PFI Volvo engine in terms of their
4.2. NOx Emissions
engine-out 18
Figure NOx compares
emissions.the four GEM fueling
Gasoline blends tested on the
generates the1.8 L PFI NOx
highest Volvoemissions
engine in and terms of their
methanol
engine-out
Figure 18
the lowest, NOx
while emissions.
compares
those from Gasoline
the four fueling
the GEM blends generates
blends tested the highest
on the 1.8between
lie somewhere NOx
L PFI Volvoemissions
engine
the two. and methanol
in terms
Since most of NOx the
theiris
lowest,
engine-out whileNOx those from
emissions. the GEM
Gasoline blends
fueling lie somewhere
generates the between
highest
produced by the thermal mechanism (itself highly dependent on the maximum temperature reached) NOx the two.
emissions Since
and most NOx
methanol is
the
produced
lowest,
alcohol fuels by the
while thermal
those
generate from mechanism
lower the GEM
levels of (itself
blends
this highly dependent
because of on
lie somewhere
pollutant thelower
between
their maximum
the two.temperature
adiabatic Since
flamemost reached)
NOx is
temperature.
alcohol
At lowerfuels
produced by the
engine generate
thermaland
speeds lower
mechanism
loadslevels
lower ofNOx
(itselfthis
highlypollutant arebecause
dependent
emissions on theof
believed their
maximum
to be lower
partly adiabatic
temperature
the result flame
ofreached)
elevated
temperature.
alcohol fuels At lower
generate engine
lower speeds
levels and
of loads
this lower
pollutant NOx emissions
because of are
their
levels of internal EGR; this is because with the use of conventional throttling to control the engine’s believed
lower to be
adiabaticpartly the
flame
result of
temperature. elevated
output to a fixed levels
At lower
load, the of
engine internal
vacuumspeeds EGR; this
andintake
in the is because
loadsislower
higherNOx with the
emissions
for the use of
alcohol are conventional
fuelsbelieved
due to the to bethrottling
factpartly to
the
that they
control
result ofthe
have higher engine’s
elevated
BTE. Note output
levels of no
that to knock
a fixed
internal load,this
EGR;
occurred the isvacuum
because
during inwith
the the
intake
the measurements useisof higherallfor
thethe
conventional
using GEM alcohol
blendsfuels
throttling to
and
due to
control the
MBT timing fact that
the engine’s they
could beoutput have
set for to higher
a fixed
every BTE.
loadload, Note that
the vacuum
and engine no knock
speed.inInthe occurred
intake
Figure 18, is during
higher for
generally the measurements
the alcohol
speaking, fuels
the higher
using
due toallthe
the gasolinethefact
GEM that
content blends
they and
haveMBT
the higher thetiming
higher NOxBTE. could
Notebe
output, set no
that forknock
regardless every
of theload and engine
occurred during
proportions ofspeed.
the type In Figure
measurements 18,
of alcohol
generally
using
used to speaking,
allcreate
the GEM the higher
blends
the blend. and MBT
Further the gasoline
timing
research could
in content
this be setthe
areas forhigher
wouldevery the NOx
be load output,
and engine
warranted. regardless
speed. of the
In Figure 18,
proportions of the type of alcohol used to create the blend. Further
generally speaking, the higher the gasoline content the higher the NOx output, regardless of the research in this areas would be
warranted.
proportions of the type of alcohol used to create the blend. Further research in this areas would be
warranted.
Energies 2020, 13, 6390 21 of 31
Figure 18. NOx emissions as a function of engine speed for a fixed brake torque of 40 Nm (2.82 bar
BMEP).
BMEP).
BMEP).
The
Theengine-out
engine-outNOx
NOxemissions
emissionsofofE85
E85and
andM56
M56asasaafunction
functionofofengine
enginespeed,
speed,measured
measuredon onthe
the
DI
DI Hyundai engine, are shown in Figure 19. Here the load is higher, at 75 Nm, and again MBTwas
Hyundai
The engine,
engine-out are
NOx shown in Figure
emissions of 19.
E85 Here
and M56the load
as a is higher,
function of at 75
engine Nm, and
speed, again
measured MBTon the
was
achievable
DI Hyundaithroughout.
achievable throughout. For NOx,
engine, are For NOx,
shown E85performs
inE85
Figureperforms
19. Here slightly
the load
slightly better
is than
better than
M56atM56
higher, although
75 Nm,
although and the bars
again
the error error
MBT bars
was
overlap
overlap for
achievable most
for most points. points.
throughout. For NOx, E85 performs slightly better than M56 although the error bars
overlap for most points.
Figure 19. NOx emissions as a function of engine speed for a fixed brake torque of 75 Nm
Figure 19.BMEP).
(5.29 bar NOx emissions as a function of engine speed for a fixed brake torque of 75 Nm (5.29 bar
BMEP).19. NOx emissions as a function of engine speed for a fixed brake torque of 75 Nm (5.29 bar
Figure
4.3. BMEP).
Knock Behavior
4.3. Knock
FromBehavior
all of the previous work, it is expected that the different GEM fuels would have very similar
4.3. Knock
knockFrom Behavior
behavior [17,19,20].
all of the previousTowork,
investigate this, all the
it is expected thatblends and theGEM
the different individual components
fuels would were
have very
tested
similar
Fromon the
knock Audi-based
all behavior
of single-cylinder
[17,19,20].
the previous To it
work, engine,this,
investigate
is expected but configured
all
that the blends with
different and a the
GEMhigher-than-standard
individual
fuels would have CR
components veryof
13.1:1
were
similar as shown
tested
knock in Table
onbehavior 4
the Audi-based above. The engine
single-cylinder
[17,19,20]. was operated
To investigateengine,
this, allbut as a fixed
theconfigured BMEP
blends andwith of 7.71 bar (i.e., 25 Nm)
a higher-than-standard
the individual components and
theofspeed
CR
were 13.1:1was
tested on fixed
as shown atin
2000
the Audi-based rpm.
Table For each fuel,
4single-cylinder
above. The the
engine ignition
was but
engine, timing
operated was
as
configured advanced
a fixed
withBMEP until an intermediate
of 7.71 bar (i.e., 25
a higher-than-standard
knocking
Nm) condition was obtained. In addition to the RON 95 gasoline
CR of 13.1:1 as shown in Table 4 above. The engine was operated as a fixed BMEP of 7.71 baruntil
and the speed was fixed at 2000 rpm. For each fuel, the ignition timingused as
was the component
advanced 25to
(i.e., an
formulate
intermediate the
Nm) and the speed blends,
knocking a RON
wascondition 98 gasoline
wasrpm.
fixed at 2000 was
obtained. also
For each investigated.
In fuel,
addition The
to the RON
the ignition results
timing95wasof this
gasolineexperiment
useduntil
advanced as theare
an
shown in Figure
component
intermediate 20.
toknocking
formulate the blends,
condition wasa obtained.
RON 98 gasoline
In additionwas to also
theinvestigated.
RON 95 gasolineThe results
used of as this
the
In
experiment Figure
are 20, all
shown the
in GEM
Figureblends
20. have the same ignition timing at incipient
component to formulate the blends, a RON 98 gasoline was also investigated. The results of this knock (determined by
a combination
experiment of audible
are shown knock20.
in Figure and the third derivative of the pressure signal passing a threshold
value). It is therefore reasonable to suppose that they have constant octane numbers. Furthermore,
pure ethanol produces the same result. The GEM blends were configured to be equivalent to E85 which
Energies 2020, 13, 6390 22 of 31
has 93% molar concentration of ethanol as discussed earlier and so this is understandable. Methanol
Energies 2020, 13,superior
still displays x FOR PEER REVIEW
knock resistance
to all of the other fuels. 22 of 32
Figure 20.
Figure Knock-limited
20.Knock-limited spark
spark advance
advance ofGEM
of all all GEM blends
blends and individual
and individual blend components.
blend components. Audi-
Audi-based single-cylinder research engine configured with a higher-than-standard CROperating
based single-cylinder research engine configured with a higher-than-standard CR of 13.1:1. of 13.1:1.
Operating point of 25 Nm (BMEP = 7.71 bar) and 2000 rpm.
point of 25 Nm (BMEP = 7.71 bar) and 2000 rpm.
Since the fuels used in this section were splash blended using relatively high-octane gasoline, this
In Figure 20, all the GEM blends have the same ignition timing at incipient knock (determined
work is considered to support the observations of Waqas et al. that using such a gasoline as a blend
by a combination of audible knock and the third derivative of the pressure signal passing a threshold
component ensures linearity in the knock resistance of the resulting fuel [23].
value). It is therefore reasonable to suppose that they have constant octane numbers. Furthermore,
pure ethanol produces
5. Particulate Emissionsthe same
from GEMresult.
Fuels Thein GEM blends were Spark-Ignition
a Direct-Injection configured to beEngine
equivalent to E85
which has 93% molar concentration of ethanol as discussed earlier and so this is understandable.
In Section
Methanol 3, high-load
still displays PN emissions
superior from the
knock resistance Ultraboost
to all 4-cylinder
of the other fuels. engine operating on GEM
blends
Since the fuels used in this section were splash blended usingfrom
were reported. Here, particulate emissions are presented four GEM
relatively fuels and
high-octane a base
gasoline,
gasoline,
this work measured
is consideredfromtoasupport
single-cylinder DI engineofatWaqas
the observations lighteretengine
al. thatloads.
using This
such was performed
a gasoline as a
because while gaseous emissions have been presented several times for
blend component ensures linearity in the knock resistance of the resulting fuel [23]. GEM blends, PN emission
have not been recorded as much.
5. Particulate Emissions from GEM Fuels in a Direct-Injection Spark-Ignition Engine
5.1. Fuels Tested
In Section 3, high-load PN emissions from the Ultraboost 4-cylinder engine operating on GEM
For this experiment, four GEM fuels were splash blended from the three components. Their
blends were reported. Here, particulate emissions are presented from four GEM fuels and a base
composition was the same as those tested in the first paper published by Turner et al. [17] and is
gasoline, measured from a single-cylinder DI engine at lighter engine loads. This was performed
reproduced in Table 6, together with their predicted dry vapor pressure equivalent (DVPE).
because while gaseous emissions have been presented several times for GEM blends, PN emission
have not been recorded as much.
Table 6. Composition of the GEM fuels tested in the single-cylinder DI engine.
5.1. Fuels Tested Gasoline Ethanol Methanol Predicted DVPE *

Fuel
(% v/v) (% v/v) (% v/v) (kPa)
For this experiment, four GEM fuels were splash blended from the three components. Their
Gasoline 100.0 0.0 0.0 62
composition was the same as those tested in the first paper published by Turner et al. [17] and is
reproduced GEM A 6, together
in Table 15.0with their predicted
85.0 dry vapor0.0pressure equivalent39(DVPE).
GEM B 29.5 42.5 28.0 67
Table
GEM C 6. Composition
37.0 of the GEM fuels
21.0 tested in the single-cylinder
42.0 DI engine.
78
GEM D 44.0
Gasoline 0.0
Ethanol Methanol56.0 Predicted DVPE87 *
Fuel
(%* Prediction
v/v) (% v/v)from a UNIFAC
derived (% v/v)model [58]. (kPa)
Gasoline 100.0 0.0 0.0 62
GEM A 15.0 85.0 0.0 39
The base gasoline for these blends was a CEC RF-02-03-specification gasoline supplied by BP.
GEM
This gasoline has no B
oxygenated29.5 components42.5 present in28.0it, and its properties
67 are shown in Table 7.
GEM C 37.0 21.0 42.0 78
GEM D 44.0 0.0 56.0 87
* Prediction derived from a UNIFAC model [58].
Energies 2020, 13, 6390 23 of 31
Note that in this table the abbreviation DVPE refers to dry vapor pressure equivalent, and IBP to the
initial boiling point in the distillation curve, T90 to the point at which 90% of the fuel mass has boiled
off, and FBP to the final boiling point at which the last components boil off.
Table 7. Composition of the base CEC RF-02-03 gasoline test fuel.
Parameter Unit Result

RON 97.0
MON 87.7
Density @ 15 ◦ C kg/L 0.7439
DVPE kPa 60.0
Aromatics % v/v 29.0
Olefins % v/v 1.8
Saturates % v/v 69.2
Carbon % m/m 86.74
Hydrogen % m/m 13.26
Gross Calorific Value MJ/kg 46.19
Net Calorific Value MJ/kg 43.38
IBP ◦C 28.8
T90 ◦C 166.6
FBP ◦C 200.8
5.2. Expected Fuel Effects on PN Emissions

It has previously been noted [59] that AFRs rich of stoichiometric increase particulate emissions.
This is true of ethanol, but the increase in particulate matter (PM) emissions rich of stoichiometric for
ethanol is very small, and anyway compared to gasoline it is effectively a dramatic reduction; as an
example, E10 has been shown to produce a greater than one order of magnitude reduction in PM
emissions compared with straight gasoline [60]. This is thought to be due to the presence of oxygen in
the fuel molecule which reduces the concentration of intermediate species that are important to the
formation of precursors to soot [61].
As methanol is added to a fuel, its vapor pressure increases, as can be seen by the change in
predicted DVPE in Table 6. High vapor pressures can result in the flash boiling of the fuel on injection,
leading to very poor mixture preparation and concomitantly high levels of PM emissions.
Flash boiling occurs when a liquid jet is injected into a volume with a pressure lower than the
saturation vapor pressure of the liquid. Here the fuel evaporates so quickly that it essentially does so
as it leaves the injector, occurring in an “onion skin” manner, with the outer layers evaporating first.
This is regardless of the composition of the layers and is known to occur in DI engines [62]. The liquid
is then rapidly depressurized, becoming superheated and consequently thermodynamically unstable,
only regaining its stability by flash boiling.
5.3. Test Engine

The single-cylinder engine used for these tests was based on that of the Jaguar AJ133 5.0 L V8
engine [63]. Its specification are shown in Table 8. It is described more fully in [58] and here it was not
used in its optical configuration.
Energies 2020, 13, 6390 24 of 31
Table 8. Single-cylinder DI engine specifications. IVO = intake valve opens; IVC = intake valve closes;
EVO = exhaust valve opens; EVC = exhaust valve closes; ATDC = after top dead center; ABDC = after
bottom dead center; BBDC = before bottom dead center.
Bore (mm) 89
Stroke (mm) 90.3
Cylinder capacity (cc) 562
Compression ratio (:1) 11.1
Fuel injection pressure (bar) 150
IVO (◦ ATDC) 33.5
IVC (◦ ABDC) 61.5
EVO (◦ BBDC) 65
EVC (◦ ATDC) 5
5.4. Particulate Measurements

A Cambustion DMS500 [64] was used to measure PN emissions. While this was not a legislative
test, in order to make these results representative of PN emissions measured during such a test, a Wiebe
filter was used to mimic the effect of the PMP legislatively-compliant protocol [65]. Measurements at
1 Hz were taken for 60 s at steady state, and the mean and standard deviation (σ) of the data over that
time were taken.
5.5. Results
For these experiments, the engine was operated at two fixed indicated mean effective pressure
(IMEP) points (corresponding to low and medium loads) at 1500 rpm as shown as in Table 9. For each
load, open-loop AFR control was used at λ = 0.9 and 1.01. λ = 1.01 was chosen to avoid any small rich
mixture excursions which are known to have a large effect on the PM emissions from non-oxygenated
fuels [59]. Because of this, these rich excursions tend to dominate drive cycle PM results, hence testing
was also conducted at λ = 0.9.
Table 9. Engine operating conditions for GEM fuels and gasoline. BTDC = before top dead center.
IMEP (bar) 2.6 and 4.9

Inlet air temperature (◦ C) 40
Coolant temperature (◦ C) 60
Relative AFR (λ) 0.9 and 1.01
Engine speed (rpm) 1500
Start of ignition (◦ BTDC) 35
Start of injection (◦ BTDC) 280
Figure 21 shows the PN emissions from the GEM fuels at light load (2.6 bar IMEP). In these results,
there is a clear trend of increasing PN emissions for the GEM fuels from blends A to C as the methanol
quantity increases. Confidence in these results is provided by the fact that the experimental sequence
was randomized and a single repeat of GEM D was undertaken with highly repeatable results.
In Figure 21, it also is shown clearly that once a certain level of methanol is passed in the fuel (say
GEM Blend C), there is no further increase in PN emissions. Furthermore the base gasoline values fall
half way in the sequence, i.e., gasoline gives higher levels of PN emissions than GEM B, but lower than
GEM C. At this load point, this in turn suggests that the presence of ethanol reduces PN while methanol
does the opposite, despite a higher proportion of oxygenated components; it is flash boiling that also
explains this. Since the methanol component flashes first, this results in poor mixture preparation and
hence increases PN emissions. As the amount of methanol in the fuel is reduced, its deleterious effect
in this regard is reduced, and instead other factors that cause oxygenate fuels to reduce particulate
emissions become important, causing PN emissions to then be reduced from the gasoline levels.
Energies 2020,13,
Energies2020, 13,6390
x FOR PEER REVIEW 2525ofof3132
Figure 21. PN emissions from GEM fuels at 2.6 bar IMEP. Error bars correspond to ±σ.
In Figure 21, it also is shown clearly that once a certain level of methanol is passed in the fuel
(say GEM Blend C), there is no further increase in PN emissions. Furthermore the base gasoline
values fall half way in the sequence, i.e., gasoline gives higher levels of PN emissions than GEM B,
but lower than GEM C. At this load point, this in turn suggests that the presence of ethanol reduces
PN while methanol does the opposite, despite a higher proportion of oxygenated components; it is
flash boiling that also explains this. Since the methanol component flashes first, this results in poor
mixture preparation and hence increases PN emissions. As the amount of methanol in the fuel is
reduced, its deleterious effect in this regard is reduced, and instead other factors that cause oxygenate
fuels to reduce
Figureparticulate emissions
21. PN emissions frombecome
GEM fuelsimportant, causing
at 2.6 bar IMEP. PNbars
Error emissions
correspondto to
then
±σ. be reduced
Figure 21. PN emissions from GEM fuels at 2.6 bar IMEP. Error bars correspond to ±σ.
from the gasoline levels.
Figure 2222shows
showsthe the
PNPN emissions
emissions fromfrom
BlendBlend A, Blend
A, Blend D and D and gasoline
gasoline at the
at the higher higher
load load
operating
In Figure 21, it also is shown clearly that once a certain level of methanol is passed in the fuel
operating
point point
(4.9 bar (4.9 bar
IMEP). IMEP).
While Whileload
the light the trends
light load trends arethe
are repeated, repeated, the PN of
PN emissions emissions
Blend D of areBlend
now
(say GEM Blend C), there is no further increase in PN emissions. Furthermore the base gasoline
D are now
below thosebelow
of thethose of the
gasoline gasolineItbaseline.
baseline. It is hypothesized
is hypothesized thatpressures
that at higher at higher the
pressures the flash
flash boiling of
values fall half way in the sequence, i.e., gasoline gives higher levels of PN emissions than GEM B,
boiling of ismethanol
methanol is suppressed,
suppressed, allowing thoseallowing
factorsthose
which factors which
generally generally
cause cause
oxygenate oxygenate
fuels to reduce fuels
PN to
but lower than GEM C. At this load point, this in turn suggests that the presence of ethanol reduces
reduce PNintothe
dominate dominate
same way in the
thatsame
they way
do forthat they do for ethanol.
ethanol.
PN while methanol does the opposite, despite a higher proportion of oxygenated components; it is
flash boiling that also explains this. Since the methanol component flashes first, this results in poor
mixture preparation and hence increases PN emissions. As the amount of methanol in the fuel is
reduced, its deleterious effect in this regard is reduced, and instead other factors that cause oxygenate
fuels to reduce particulate emissions become important, causing PN emissions to then be reduced
from the gasoline levels.
Figure 22 shows the PN emissions from Blend A, Blend D and gasoline at the higher load
operating point (4.9 bar IMEP). While the light load trends are repeated, the PN emissions of Blend
D are now below those of the gasoline baseline. It is hypothesized that at higher pressures the flash
boiling of methanol is suppressed, allowing those factors which generally cause oxygenate fuels to
reduce PN to dominate in the same way that they do for ethanol.
Figure 22. PN emissions from gasoline and two GEM fuels at 4.9 bar IMEP. Error bars correspond
Figure 22. PN emissions from gasoline and two GEM fuels at 4.9 bar IMEP. Error bars correspond to
to ±σ.
±σ.
Overall, the results in this section show that using GEM fuels in DI engines will not significantly
Overall, the results in this section show that using GEM fuels in DI engines will not significantly
impact PN emissions versus gasoline, other than possibly to decrease them significantly. The caveat is
impact PN emissions versus gasoline, other than possibly to decrease them significantly. The caveat
that at light load, and when methanol is present in a high proportion, the results suggest that a slight
is that at light load, and when methanol is present in a high proportion, the results suggest that a
increase in PN emissions over those from gasoline may be seen, but as the load increases, then all
slight increase in PN emissions over those from gasoline may be seen, but as the load increases, then
such GEM fuels will then give lower PN emissions. This is likely due to methanol flash boiling at low
all such GEM fuels will then give lower PN emissions. This is likely due to methanol flash boiling at
pressures, and this effect reducing as the BMEP increases. The equivalent conditions are not reached
for the ethanol in the fuel for either test point, so it tends to suppress PN emission formation. Further
research is warranted into this phenomenon and whether, where other alcohols are being considered
as blend components
Figure in future
22. PN emissions fuels
from (for example,
gasoline butanol),
and two GEM they
fuels at 4.9 show any Error
bar IMEP. similar trends.
bars correspond to
±σ.
6. Conclusions
Overall,investigations
Through the results in this sectiondifferent
in several show that using GEM
research fuels
engines, inpaper
this DI engines will not
has shown significantly
that, at mid-to
impact PNrates:
high-blend emissions versus gasoline, other than possibly to decrease them significantly. The caveat
is that at light load, and when methanol is present in a high proportion, the results suggest that a
slight increase in PN emissions over those from gasoline may be seen, but as the load increases, then
all such GEM fuels will then give lower PN emissions. This is likely due to methanol flash boiling at
Energies 2020, 13, 6390 26 of 31
1. In support of earlier work by the authors and other research groups, ternary (three
component) mixtures of gasoline, ethanol, and methanol (or GEM) can be formulated using
an iso-stoichiometric approach to provide a drop-in alternative fuel for any given binary
gasoline-ethanol blend.
2. When so configured, such GEM fuels have matched gravimetric lower heating values, heats of
vaporization, and molar concentrations of reactants in the charge. The latter two characteristics
gives rise to similar molar expansion ratios and extremely close RON and MON values. With these
come very close sensitivities.
3. The improvement in full-load performance of such high-blend-rate fuels in terms of KLSA,
and the associated better BMEP and thermal efficiency, is entirely in line with what is expected for
high-blend alcohols over conventional gasoline fuels. There is also the expected disadvantage in
terms of BSFC due to the reduction in gravimetric energy content in relation to pure hydrocarbon
fuels. These results are consistent with the literature and regardless of EGR being used or the
amount of exhaust back pressure applied to the combustion system (to simulate turbocharged or
supercharged operation of an engine).
4. NOx emissions are shown to be lower with the alcohol blends (and lowest with pure methanol).
Furthermore they are consistent for the iso-stoichiometric formulations, with a suggestion that
within such GEM blends increased methanol content tends to lower NOx emissions further.
5. When tested in a single-cylinder engine, the knock performance of the tested blends is essentially
identical, and at the high blend rate (equivalent to the stoichiometry of E85) the same as pure
ethanol and very close to that of methanol. This result is in line with the previous observations
on the RON, MON, and sensitivity performance of such blends.
6. In terms of particulate emissions, the results presented here show that using GEM fuels is unlikely
to impact the PN emissions from DI engines significantly, and that many of these fuels can
decrease the magnitude observed.
7. GEM fuels with a higher proportion of methanol do tend to generate more PN emissions than
gasoline at low load, which is attributed to higher degrees of flash boiling at lower in-cylinder
pressures. However, this result reverses with increased load and may be due to the change in
the dominance of flash boiling effects, with the presence of oxygen in the fuel then reducing the
number of particulate precursors formed in combustion.
8. At lower loads the presence of a high proportion of ethanol in a blend significantly reduces PN
emissions compared to gasoline. This is again believed to be due to the molecule containing
oxygen; the flash boiling effect is also lower for ethanol than it is for methanol, which is believed
to contribute to this result.
Because methanol can be manufactured using low- and zero-net-carbon pathways (using biological
routes which remove CO2 from the atmosphere via plant growth or by direct air capture and utilization
in a fully synthetic manufacturing route, respectively), there is thus the potential to introduce it into
the fuel pool used by the more than 9 million E85-gasoline flex-fuel vehicles in existence via GEM
ternary blends. The lower price of methanol could also beneficially impact the fuel cost on a per unit
energy basis, which is generally not the case for E85. Furthermore, although only demonstrated here at
a blend rate equivalent to E85, the approach is also viable for any gasoline-ethanol binary blend.
Author Contributions: Conceptualization, J.W.G.T., S.V., F.C.P.L., and P.P.E.; methodology, J.W.G.T., A.G.J.L.,
L.S., and F.C.P.L.; validation, A.G.J.L., J.V., L.S., and F.C.P.L.; formal analysis, J.W.G.T., S.V., L.S., and F.C.P.L.;
investigation, A.G.J.L., L.S., and F.C.P.L.; resources, S.A., S.V., and F.C.P.L.; data curation, J.W.G.T., A.G.J.L., L.S.,
and F.C.P.L.; writing—original draft preparation, J.W.G.T, L.S., S.V., and F.C.P.L.; writing—review and editing,
A.G.J.L., S.A., C.J.B., J.V., L.S., F.C.P.L., and P.P.E.; funding acquisition, S.A., C.J.B., S.V., and F.C.P.L. All authors
have read and agreed to the published version of the manuscript.
Funding: Elements of the reported research took place as part of the “Ultra Boost for Economy” project part
funded by the Technology Strategy Board, now Innovate UK, the United Kingdom’s innovation agency; their
support is similarly gratefully acknowledged. J. Vancoillie and L. Sileghem gratefully acknowledge fellowships
from the Research Foundation—Flanders (FWO09/ASP/030 and FWO11/ASP/056).
Energies 2020, 13, 6390 27 of 31
Acknowledgments: Although too numerous to detail here, the historical support and encouragement of many
individuals and organizations involved in the methanol industry and flex-fuel vehicle production is very
gratefully acknowledged. Their faith in what was initially a purely academic finding concerning the properties
of iso-stoichiometric GEM fuels has been extremely valuable and encouraging throughout the conduct of all
aspects of this research. The partners in the Ultraboost project were Jaguar Land Rover, Lotus Engineering, GE
Precision Engineering, Shell Global Solutions, CD-adapco, the University of Bath, Imperial College London, and
the University of Leeds, and all of their help and contributions to the reported work are gratefully acknowledged.
Conflicts of Interest: The authors declare no conflict of interest.
Abbreviations
ABDC After top dead center
AFR Air-fuel ratio
ATDC After top dead center
BBDC Before bottom dead center
BEV Battery electric vehicle
BMEP Brake mean effective pressure
BSFC Brake specific fuel consumption
BTDC Before top dead center
BTE Brake thermal efficiency
CAHU Charge air handling unit
CO Carbon monoxide
CO2 Carbon dioxide
CR Compression ratio
DI Direct injection
DOHC Double overhead camshaft
DVPE Dry vapor pressure equivalent
EGR Exhaust gas recirculation
EVC Exhaust valve closes
EVO Exhaust valve opens
FBP Final boiling point (at which the last components boil off in the distillation curve)
GEM Gasoline-ethanol–methanol
HC Hydrocarbons
IBP Initial boiling point (at which the first components boil off in the distillation curve)
ICE Internal combustion engine
IMEP Indicated mean effective pressure
ITE Indicated thermal efficiency
IVC Intake valve closes
IVO Intake valve opens
KLSA Knock-limited spark advance
LHV Lower heating value
MBT Minimum spark advance for best torque
MON Motor octane number
NOx Oxides of nitrogen
OHC Overhead camshaft
PFI Port fuel injection
PM Particulate matter
PN Particle number
ppm Parts per million
RON Research octane number
SI Spark ignition
T90 Point at which 90% of the fuel mass has boiled off (in the distillation curve)
TWC Three-way catalyst
∆hvap Heat of vaporization
σ Standard deviation
Energies 2020, 13, 6390 28 of 31
References
1. Gross, M. A planet with two billion cars. Curr. Biol. 2016, 26, R307–R310. [CrossRef]
2. Gutzmer, P. Mobility for Tomorrow—Between the Poles of Electrification. In Proceedings of the 26th Aachen
Colloquium, Aachen, Germany, 9–11 October 2017.
3. Woodward, M. New Market. New Entrants. New Challenges. Battery Electric Vehicles. Deloitte. 2019.
Available online: https://www2.deloitte.com/content/dam/Deloitte/uk/Documents/manufacturing/deloitte-
uk-battery-electric-vehicles.pdf (accessed on 19 June 2020).
4. Leach, F.C.P.; Akehurst, S.; Brace, C.; Busch, H.; Cairns, A.; Davy, M.; Gavaises, M.; Heikal, M.H.; Linne, M.;
Shayler, P.; et al. Road bumps for electric cars (letter). Economist 2017, 413, 16.
5. White, T.L. Alcohol as a Fuel for the Automobile Motor; SAE Technical Paper Series; SAE International:
Warrendale, PA, USA, 1907. [CrossRef]
6. Niizato, T. Future direction of Engine and PHEV technology at Honda. In Proceedings of the 14th International
Conference on Engines for Vehicles, Hokkaido, Japan, 27–31 August 2017.
7. Fraidl, G.; Kapus, P.; Mitterecker, H.; Weißbäck, M. Internal Combustion Engine 4.0. MTZ Worldw. 2018, 79,
26–33. [CrossRef]
8. Powell, N.; Hill, N.; Bates, J.; Bottrell, N.; Biedka, M.; White, B.; Pine, T.; Carter, S.; Patterson, J.; Yucel, S. Impact
Analysis of Mass EV Adoption and Low Carbon Intensity Fuels Scenarios. CONCAWE Report RD18-001538-4;
24 August 2018. Available online: https://www.fuelseurope.eu/wp-content/uploads/Mass-EV-Adoption-
and-Low-Carbon-Fuels-Scenarios.pdf (accessed on 30 November 2020).
9. Turner, J.W.G.; Lewis, A.G.J.; Akehurst, S.; Brace, C.J.; Verhelst, S.; Vancoillie, J.; Sileghem, L.; Leach, F.C.P.;
Edwards, P.P. Alcohol fuels for spark-ignition engines: Performance, efficiency and emission effects at mid to
high blend rates for binary mixtures and pure components. Proc. Inst. Mech. Eng. Part D J. Automob. Eng.
2018, 232, 36–56. [CrossRef]
10. Foong, T.M.; Morganti, K.J.; Brear, M.J.; da Silva, G.; Yang, Y.; Dryer, F.L. The octane numbers of ethanol
blended with gasoline and its surrogates. Fuel 2014, 115, 727–739. [CrossRef]
11. Anderson, J.E.; Leone, T.G.; Shelby, M.H.; Wallington, T.J.; Bizub, J.J.; Foster, M.; Lynskey, M.G.; Polovina, D.
Octane Numbers of Ethanol-Gasoline Blends: Measurements and Novel Estimation Method from Molar
Composition. SAE Technical Paper 2012-01-1274. In Proceedings of the SAE 2012 World Congress, Detroit,
MI, USA, 24–26 April 2012. [CrossRef]
12. Bergström, K.; Melin, S.-A.; Jones, C.C. The New ECOTEC Turbo BioPower Engine from GM
Powertrain—Utilizing the Power of Nature’s resources. In Proceedings of the 28th International Vienna
Motor Symposium, Vienna, Austria, 26–27 April 2007.
13. Bergström, K.; Nordin, H.; Königstein, A.; Marriott, C.D.; Wiles, M.A. ABC—Alcohol Based Combustion
Engines—Challenges and Opportunities. In Proceedings of the 16th Aachen Colloquium, Aachen, Germany,
7 October 2007; pp. 1031–1071.
14. Turner, J.W.G.; Pearson, R.J.; Holland, B.; Peck, B. Alcohol-Based Fuels in High Performance Engines. SAE
Technical Paper 2007-01-0056. In Proceedings of the SAE Fuels and Emissions Conference, Cape Town,
South Africa, 23–25 January 2007. [CrossRef]
15. Wallner, T.; Miers, S.A. Combustion Behavior of Gasoline and Gasoline/Ethanol Blends in a Modern
Direct-Injection 4-Cylinder Engine. SAE Technical Paper 2008-01-0077. In Proceedings of the SAE 2008
World Congress, Detroit, MI, USA, 14–17 April 2008. [CrossRef]
16. West, B.H.; López, A.J.; Theiss, T.J.; Graves, R.L.; Storey, J.M.; Lewis, S.A. Fuel Economy and Emissions of
the Ethanol-Optimized Saab 9-5-Biopower. SAE Technical Paper 2007-01-3994. In Proceedings of the SAE
Powertrain & Fluid Systems Conference and Exhibition, Rosemont, IL, USA, 29 October–1 November 2007.
[CrossRef]
17. Turner, J.W.G.; Pearson, R.J.; Purvis, R.; Dekker, E.; Johansson, K.; Bergström, K. GEM Ternary Blends:
Removing the Biomass Limit by using Iso-Stoichiometric mixtures of Gasoline, Ethanol and Methanol. SAE
Paper Number 2011-24-0113. In Proceedings of the 10th International Conference on Engines and Vehicles,
Capri, Naples, Italy, 11–16 September 2011. [CrossRef]
Energies 2020, 13, 6390 29 of 31
18. Turner, J.W.G.; Pearson, R.J.; Dekker, E.; Iosefa, B.; Dolan, G.A.; Johansson, K.; Bergström, K. Evolution
of alcohol fuel blends towards a sustainable transport energy economy. In Proceedings of the 2012 MIT
Energy Initiative Symposium: Prospects for Flexible- and Bi-Fuel Light Duty Vehicles, Cambridge, MA, USA,
19 April 2012.
19. Turner, J.W.G.; Pearson, R.J.; McGregor, M.A.; Ramsay, J.M.; Dekker, E.; Iosefa, B.; Dolan, G.A.; Johansson, K.;
Bergström, K. GEM Ternary Blends: Testing Iso-Stoichiometric mixtures of Gasoline, Ethanol and Methanol
in a Production Flex-Fuel Vehicle Fitted with a Physical Alcohol Sensor. SAE Paper Number 2012-01-1279.
In Proceedings of the SAE 2012 World Congress, Detroit, MI, USA, 24–26 April 2012. [CrossRef]
20. Turner, J.W.G.; Pearson, R.J.; de Goede, S.B.A.; Woolard, C. GEM Ternary Blends: Investigations into Exhaust
Emissions, Blend Properties and Octane Numbers. SAE Paper Number 2012-01-1586. In Proceedings of the
SAE Powertrains, Fuels and Lubricants Meeting, Malmo, Sweden, 18–20 September 2012.
21. Pitcher, G.; Turner, J.W.G.; Pearson, R.J. GEM Ternary Blends of Gasoline, Ethanol and Methanol: An Initial
Investigation into Fuel Spray and Combustion Characteristics in a Direct-Injected Spark-Ignition Optical
Engine using Mie Imaging. SAE Technical Paper 2012-01-1740. In Proceedings of the SAE Powertrains, Fuels
and Lubricants Meeting, Malmo, Sweden, 18–20 September 2012. [CrossRef]
22. Turner, J.W.G.; Pearson, R.J.; Dekker, E.; Iosefa, B.; Dolan, G.; Johansson, K.; Bergström, K. Extending the
role of alcohols as transport fuels using iso-stoichiometric ternary blends of gasoline, ethanol and methanol.
Appl. Energy 2013, 102, 72–86. [CrossRef]
23. Waqas, M.; Naser, N.; Sarathy, M.; Feijs, J.; Morganti, K.; Nyrenstedt, G.; Johansson, B. Auto-Ignition of
Iso-Stoichiometric Blends of Gasoline-Ethanol-Methanol (GEM) in SI, HCCI and CI Combustion Modes.
SAE Technical Paper 2017-01-0726. In Proceedings of the SAE 2017 World Congress, Detroit, MI, USA,
4–6 April 2017. [CrossRef]
24. Pearson, R.J.; Turner, J.W.G.; Bell, A.; de Goede, S.; Woolard, C.; Davy, M.H. Iso-stoichiometric fuel blends:
Characterisation of physicochemical properties for mixtures of gasoline, ethanol, methanol and water.
Proc. Inst. Mech. Eng. Part D J. Automob. Eng. 2014, 229, 111–139. [CrossRef]
25. Turner, J.W.G.; Pearson, R.J.; Harrison, P.; Marmont, A.; Jennings, R.; Verhelst, S.; Vancoillie, J.; Sileghem, L.;
Pecqueur, M.; Martens, K.; et al. Evolutionary decarbonization of transport: A contiguous roadmap to
affordable mobility using Sustainable Organic Fuels for Transport. In Proceedings of the Institution of
Mechanical Engineers Sustainable Vehicle Technologies Conference, Gaydon, UK, 15–16 November 2012.
26. Pearson RJTurner, J.W.G.; Eisaman, M.D.; Littau, K.A. Extending the Supply of Alcohol Fuels for Energy
Security and Carbon Reduction. SAE Technical Paper 2009-01-2764. In Proceedings of the SAE Powertrain,
Fuels and Lubricants Meeting, San Antonio, TX, USA, 2–4 November 2009. [CrossRef]
27. Pearson, R.J.; Eisaman, M.D.; Turner, J.W.G.; Edwards, P.P.; Jiang, Z.; Kuznetsov, V.L.; Littau, K.A.; di Marco, L.;
Taylor, S.R.G. Energy storage via carbon-neutral fuels made from CO2 , water, and renewable energy. Proc.
IEEE 2012, 100, 440–460. [CrossRef]
28. Verhelst, S.; Turner, J.W.G.; Sileghem, L.; Vancoillie, J. Methanol as a fuel for internal combustion engines.
Prog. Energy Combust. Sci. 2019, 70, 43–88. [CrossRef]
29. Steinberg, M. Production of synthetic methanol from air and water using controlled thermonuclear reactor
power—I. Technology and energy requirement. Energy Convers. 1977, 17, 97–112. [CrossRef]
30. Olah, G.A.; Goeppert, A.; Prakash, G.K.S. Beyond Oil and Gas: The Methanol Economy, 2nd ed.; Wiley-VCH
Verlag GmbH & Co. KGaA: Weinheim, Germany, 2009; ISBN 98-3-527-32422-4.
31. Bandi, A.; Specht, M.; Weimer, T.; Schaber, K. CO2 recycling for hydrogen storage and
transportation—Electrochemical CO2 removal and fixation. Energy Convers. Manag. 1995, 36, 899–902.
[CrossRef]
32. Stucki, S.; Schuler, A.; Constantinescu, M. Coupled CO2 recovery from the atmosphere and water electrolysis:
Feasibility of a new process for hydrogen storage. Int. J. Hydrog. Energy 1995, 20, 653–663. [CrossRef]
33. Weimer, T.; Schaber, K.; Specht, M.; Bandi, A. Methanol from atmospheric carbon dioxide: A liquid zero
emission fuel for the future. Energy Convers. Manag. 1996, 37, 1351–1356. [CrossRef]
34. Sileghem, L.; Coppens, A.; Casier, B.; Vancoillie, J.; Verhelst, S. Performance and emissions of iso-stoichiometric
ternary GEM blends on a production SI engine. Fuel 2014, 117, 286–293. [CrossRef]
35. Sileghem, L.; Ickes, A.; Wallner, T.; Verhelst, S. Experimental Investigation of a DISI Production Engine
Fuelled with Methanol, Ethanol, Butanol and ISO-Stoichiometric Alcohol Blends. SAE Technical Paper
2015-01-0768. In Proceedings of the SAE 2015 World Congress, Detroit, MI, USA, 21–23 April 2015. [CrossRef]
Energies 2020, 13, 6390 30 of 31
36. Shirazi, S.A.; Abdollahipoor, B.; Martinson, J.; Windom, B.; Foust, T.D.; Reardon, K.F. Effects of dual-alcohol
gasoline blends on physiochemical properties and volatility behavior. Fuel 2019, 252, 542–552. [CrossRef]
37. Larsson, T.; Stenlaas, O.; Erlandsson, A. Future Fuels for DISI Engines: A Review on Oxygenated, Liquid
Biofuels. SAE Technical Paper 2019-01-0036. In Proceedings of the SAE 2018 World Congress, Detroit, MI,
USA, 10–12 April 2018. [CrossRef]
38. Shirazi, S.A.; Abdollahipoor, B.; Windom, B.; Reardon, K.F.; Foust, T.D. Effects of blending C3-C4 alcohols on
motor gasoline properties and performance of spark ignition engines: A review. Fuel Process. Technol. 2020,
197, 106194. [CrossRef]
39. Bergthorson, J.M.; Thomson, M.J. A review of the combustion and emissions properties of advanced
transportation biofuels and their impact on existing and future engines. Renew. Sustain. Energy Rev. 2015, 42,
1393–1417. [CrossRef]
40. Züttel, A.; Remhof, A.; Borgschulte, A.; Friedrichs, O. Hydrogen: The future energy carrier. Phil. Trans. R.
Soc. Part A 2010, 368, 3329–3342. [CrossRef]
41. Turner, J.W.G.; Popplewell, A.; Patel, R.; Johnson, T.; Darnton, N.J.; Richardson, S.; Bredda, S.W.; Jackson, R.;
Bithell, C.; Tudor, R.; et al. Ultra Boost for Economy: Extending the Limits of Extreme Engine Downsizing.
SAE Technical Paper 2014-01-1185. In Proceedings of the SAE 2014 World Congress, Detroit, MI, USA,
8–10 April 2014; Volume 7. [CrossRef]
42. Lewis, A.G.J.; Akehurst, S.; Turner, J.W.G.; Patel, R.; Popplewell, A. Observations on the measurement and
performance impact of catalyzed vs. non-catalyzed EGR on a heavily downsized DISI engine. SAE Technical
Paper 2014-01-1196. In Proceedings of the SAE 2014 World Congress, Detroit, MI, USA, 8–10 April 2014;
Volume 7. [CrossRef]
43. Remmert, S.M.; Cracknell, R.; Head, B.; Schuetze, A.; Lewis, A.G.J.; Akehurst, S.; Turner, J.W.G.; Popplewell, A.
Octane Response in a Downsized, Highly Boosted Direct Injection Spark Ignition Engine. SAE Technical
Paper 2014-01-1397. In Proceedings of the SAE 2014 World Congress, Detroit, MI, USA, 8–10 April 2014;
Volume 7. [CrossRef]
44. Campbell, S.; Remmert, S.; Cracknell, R.; Schuetze, A.; Warnecke, W.; Lewis, A.; Akehurst, S.; Turner, J.W.G.;
Patel, R.; Popplewell, A. Fuel Effects in a Downsized, Highly Boosted Direct Injection Spark Ignition Engine.
In Proceedings of the 23rd Aachen Colloquium, Aachen, Germany, 6–8 October 2014.
45. Remmert, S.; Campbell, S.; Cracknell, R.; Schuetze, A.; Lewis, A.; Giles, K.; Akehurst, S.; Turner, J.W.G.;
Popplewell, A.; Patel, R. Octane Appetite: The Relevance of a Lower Limit to the MON Specification in a
Downsized, Highly Boosted DISI Engine. SAE Technical Paper 2014-01-2718. In Proceedings of the SAE 2014
Powertrain, Fuels and Lubricants Conference, Birmingham, UK, 20–23 October 2014; Volume 7. [CrossRef]
46. Leach, F.C.P.; Stone, C.R.; Richardson, D.; Lewis, A.G.J.; Akehurst, S.; Turner, J.W.G.; Remmert, S.; Campbell, S.;
Cracknell, R.F. The effect of oxygenate fuels on PN emissions from a highly boosted GDI engine. Fuel 2018,
225, 277–286. [CrossRef]
47. Leach, F.C.P.; Stone, R.; Richardson, D.; Lewis, A.G.J.; Akehurst, S.; Turner, J.W.G.; Remmert, S.; Campbell, S.;
Cracknell, R.F. Particulate emissions from a highly boosted gasoline direct injection engine. Int. J. Engine Res.
2018, 19, 347–359. [CrossRef]
48. Chupka, G.M.; Christensen, E.; Fouts, L.; Alleman, T.L.; Ratcliff, M.A.; McCormick, R.L. Heat of Vaporization
Measurements for Ethanol Blends Up to 50 Volume Percent in Several Hydrocarbon Blendstocks and
Implications for Knock in SI Engines. SAE Technical Paper 2015-01-0763. In Proceedings of the SAE 2015
World Congress, Detroit, MI, USA, 21–23 April 2015; Volume 8. [CrossRef]
49. Szybist, J.P.; Chakravathy, K.; Daw, C.S. Analysis of the Impact of Selected Fuel Thermochemical Properties
on Internal Combustion Engine Efficiency. Energy Fuels 2012, 26, 2798–2810. [CrossRef]
50. Nguyen, K.-D.; Szybist, J.; Sileghem, L.; Verhelst, S. Effects of molar expansion ratio of fuels on engine
efficiency. Fuel 2020, 263, 116743. [CrossRef]
51. Leach, F.C.P.; Stone, R.; Richardson, D.; Lewis, A.G.J.; Akehurst, S.; Turner, J.W.G.; Shankar, V.; Chahal, J.;
Cracknell, R.F.; Aradi, A. The effect of fuel composition on particulate emissions from a highly boosted GDI
engine—An evaluation of three particulate indices. Fuel 2019, 252, 598–611. [CrossRef]
52. Millwood, D.; Turner, J.W.G. Fuel Properties Calculator; Proprietary Software; University of Bath: Bath,
UK, 2020.
Energies 2020, 13, 6390 31 of 31
53. Sluder, C.S. Estimation of the Fuel Efficiency Potential of Six Gasoline Blendstocks Identified by the U.S.
Department of Energy’s Co-Optimization of Fuels and Engines Program. SAE Technical Paper 2019-01-0017.
In Proceedings of the SAE 2019 Powertrain, Fuels and Lubricants Conference, San Antonio, TX, USA, 22–24
January 2019. [CrossRef]
54. McCormick, R.L.; Fioroni, G.; Fouts, L.; Christensen, E.; Yanowitz, J.; Polikarpov, E.; Albrecht, K.; Gaspar, D.J.;
Gladden, J.; George, A. Selection Criteria and Screening of Potential Biomass-Derived Streams as Fuel
Blendstocks for Advanced Spark-Ignition Engines. SAE Int. J. Fuels Lubr. 2017, 10, 442–460. [CrossRef]
55. Vancoillie, J.; Verhelst, S.; Demuynck, J.; Sierens, R.; Rusu, E.; Pantile, V. Efficiency Comparison between
Hydrogen, Methanol and Gasoline on a Production-Type Four-Cylinder Engine. In Proceedings of the 13th
EAEC European Automotive Conference, Valencia, Spain, 14–16 June 2011.
56. Pearson, R.J.; Turner, J.W.G.; Peck, A.J. Gasoline-ethanol-methanol tri-fuel vehicle development and its role
in expediting sustainable organic fuels for transport. In Proceedings of the IMechE Low Carbon Vehicles
Conference, London, UK, 20–21 May 2009.
57. Naganuma, K.; Vancoillie, J.; Verhelst, S.; Sileghem, L.; Turner, J.W.G.; Pearson, R.J.; Martens, K. Drive cycle
analysis of load control strategies for methanol fuelled ICE vehicle. SAE Technical Paper 2012-01-1606.
In Proceedings of the SAE International Powertrains, Fuels and Lubricants Meeting, Malmo, Sweden,
18–20 September 2012.
58. Leach, F.C.P.; Richardson, D.; Stone, R. The Influence of Fuel Properties on Particulate Number Emissions
from a Direct Injection Spark Ignition Engine. SAE Technical Paper 2013-01-1558. In Proceedings of the SAE
2013 World Congress, Detroit, MI, USA, 16–18 April 2013. [CrossRef]
59. Peckham, M.S.; Finch, A.; Campbell, B.; Price, P.; Davies, M.T. Study of Particle Number Emissions from a
Turbocharged Gasoline Direct Injection (GDI) Engine Including Data from a Fast-Response Particle Size
Spectrometer. SAE Technical Paper 2011-01-1224. In Proceedings of the SAE 2011 World Congress, Detroit,
MI, USA, 12–14 April 2011. [CrossRef]
60. Chen, L.; Braisher, M.; Crossley, A.; Stone, R. The Influence of Ethanol Blends on Particulate Matter Emissions
from Gasoline Direct Injection Engines. SAE Technical Paper 2010-01-0793. In Proceedings of the SAE 2010
World Congress, Detroit, MI, USA, 13–15 April 2010. [CrossRef]
61. Litzinger, T.; Stoner, M.; Hess, H.; Boehman, A. Effects of oxygenated blending compounds on emissions
from a turbocharged direct injection diesel engine. Int. J. Engine Res. 2000, 1, 57–70. [CrossRef]
62. Davy, M.H.; Williams, P.A. The Effects of Flash Boiling on Mixture Formation in a Firing Direct-Injection
Spark-Ignition (DISI) Engine. In Proceedings of the International Congress on Direkteinspritzung im
Ottomotor II (Gasoline Direct Injection Engines), Munich, Germany, 16–17 November 1999; pp. 154–170.
63. Sandford, M.; Page, G.; Crawford, P. The All New AJV8. SAE Technical Paper 2009-01-1060. In Proceedings
of the SAE 2009 World Congress, Detroit, MI, USA, 20–23 April 2009. [CrossRef]
64. Reavell, K.; Hands, T.; Collings, N. A Fast Response Particulate Spectrometer for Combustion Aerosols. SAE
Technical Paper 2002-01-2714. In Proceedings of the SAE Powertrain & Fluid Systems Conference, San Diego,
CA, USA, 21–24 October 2002. [CrossRef]
65. Andersson, J.; Giechaskiel, B.; Muñoz-Bueno, R.; Sandbach, E.; Dilara, P. Particle Measurement Programme
(PMP) Light-Duty Inter-Laboratory Correlation Exercise (ILCE_LD) Final Report; European Commission Joint
Research Centre Institute for Environment and Sustainability: Ispra, Italy, 2007.
Publisher’s Note: MDPI stays neutral with regard to jurisdictional claims in published maps and institutional
affiliations.
© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).

Energies: Alcohol Fuels For Spark-Ignition Engines: Performance, E To High Blend Rates For Ternary Mixtures

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Energies: Alcohol Fuels For Spark-Ignition Engines: Performance, E To High Blend Rates For Ternary Mixtures

Uploaded by

Copyright:

Available Formats

energies

Keywords: alcohols; gasoline-alcohol blends; ternary blends; renewable fuels; e-fuels

Energies 2020, 13, 6390; doi:10.3390/en13236390 www.mdpi.com/journal/energies

BEVs HEVs and PHEVs ICE-only Cumulative Cars Manufactured

Cumulative Cars Manufactured / [Millions]

ICE (all) Cumulative Cars Manufactured

Proportion of Ethanol in Ternary Blend / [% v/v]

3. Ternary Gasoline-Alcohol Blend Tests in the “Ultraboost” Extreme Downsizing

Table 1. Specifications of the Ultraboost extreme downsizing demonstrator engine in the

Alcohol Proportion Volumetric Stoichiometric

Table 3. Test operating points for Ultraboost engine testing.

Engine Boost Plenum

Brake Thermal Efficiency / [%] 40

4. Performance and Emissions of Iso-Stoichiometric GEM Blends in Production Engines and

Table 4. Test engine specifications.

Engine Type Volvo 1.8 L Audi/NSU Hyundai

Energies 2020, 13, x FOR PEER REVIEW 18 of 32

Energies 2020, 13, 6390 19 of 31

Figure 16. Cont.

5.1. Fuels Tested Gasoline Ethanol Methanol Predicted DVPE *

Table 7. Composition of the base CEC RF-02-03 gasoline test fuel.

Parameter Unit Result

5.2. Expected Fuel Effects on PN Emissions

5.3. Test Engine

5.4. Particulate Measurements

IMEP (bar) 2.6 and 4.9

You might also like