Article

pubs.acs.org/EF

Determination of Syngas Premixed Gasoline and Methanol

Combustion Products at Chemical Equilibrium via Lagrange
Multipliers Method
Osman Sinan Süslü*,† and Ipek Becerik†,‡
†
Energy Institute and ‡Chemistry Department, Istanbul Technical University, Maslak, Iṡ tanbul 34469, Turkey

ABSTRACT: This article investigates theoretically the products from adiabatic combustion of synthesis gas reformed out of
aqueous methanol with a variable gasoline−methanol mixture assuming chemical equilibrium. Previous research has generally
focused on the combustion of a given fixed fuel composition, whereas in this article the composition of the fuel is varied by three
different parameters: the amount of water in the aqueous methanol used as the synthesis gas feedstock, the amount of methanol
in the gasoline−methanol mixture, and the ratio of synthesis gas energy to total fuel energy. The effects of these parameters on
the adiabatic flame temperature and combustion product distribution are determined. The method used in this article allows the
above parameters and the air/fuel ratio parameter to vary in equation sets derived using Lagrange undetermined multipliers in
order to determine the combustion temperature and products with a computer solution. Stoichiometric combustion of synthesis
gas reformed out of an equimolar water−methanol mixture causes a decrease in the NO ratio by 20% compared with methanol
and 40% compared with gasoline under equilibrium conditions, whereas the adiabatic flame temperature decreases by only 50
and 100 K, respectively. Stoichiometric combustion of synthesis gas reformed out of neat methanol causes a higher adiabatic
flame temperature and higher NOx emissions than stoichiometric combustion of gasoline or methanol. However, neat-methanol-
based synthesis gas has a higher power density and extends the lean flammability limit of the engine more efficiently than
aqueous-methanol-based synthesis gas because the reforming enthalpy of neat methanol is higher than that of aqueous methanol.
Hydrogen-rich synthesis gas fuels should be combusted with lean air/fuel ratios at part load to increase the thermal efficiency,
combustion stability, and control emissions simultaneously. If the road load increases, the engine charge’s energy density and
motor octane number will be increased by substitution of synthesis gas with liquid fuel while decreasing the air/fuel ratio
simultaneously.

1. INTRODUCTION Although these exhaust gases possess substantial availability,

An alternative to three-way catalytic converters, the current this energy is not recovered for further use in state-of-the-art
standard for emission reduction in spark ignition (SI) engines, automotive technology. The resulting exhaust heat could
is the modification of the initial combustion process by using increase the energy level of a fuel by reforming the fuel to a
lean mixtures.1 Moreover, stable combustion of leaner mixtures hydrogen-rich synthesis gas. The prerequisite is that the fuel’s
decreases the engine charge’s energy density for an efficient reforming reaction take place at a temperature lower than the
operation at part load without throttling the flow of engine exhaust gas temperature. The lower the temperature of the
charge. reforming reaction, the higher will be the reactor’s heat
Hydrogen’s wide flammability limit and high flame speed recovery and syngas output.
allow ultralean combustion, reducing exhaust emissions. The Compared with alkanes and higher alcohols, methanol has a
exhaust emissions of an SI engine were shown to be reduced by higher oxygen-to-carbon ratio, allowing the catalytic decom-
addition of pure hydrogen to gasoline.2 However, on a position of the neat fuel vapor to CO and H2 at relatively low
commercial level, hydrogen stored on board poses a number temperatures. Methanol’s steam reforming reaction causes a
of problems, such as hydrogen infrastructure, mechanical less reducing reaction environment that retards coke formation
problems involved with refueling, limited driving range, high compared with other commercial hydrocarbons. The resulting
weight involved with the size of the storage cylinder, and safety lower propensity to form coke preserves the catalyst from
concerns related to carrying a high-pressure cylinder of deactivation due to solid carbon formation and thereby
hydrogen.3 prevents a premature reactor shutdown.5 Moreover, steam
Instead of pure hydrogen storage on board, partial oxidation reforming of methanol consumes less energy to produce 1 mole
of gasoline with air generates H2-rich synthesis gas (syngas), of hydrogen compared with commercial fuels such as
which when added to gasoline reduces engine exhaust compressed natural gas (CNG), liquefied petroleum gas
emissions as well.4 However, partial oxidation is not considered (LPG), gasoline, and ethanol. Given the same exhaust gas
to be attractive in terms of efficiency because the product gas amount and temperature, the lower energy requirement of a
has a lower calorific value than the original feedstock as a result
of the exothermic reaction.1 Received: November 7, 2013
The exhaust gases of SI engines for automotive propulsion Revised: January 19, 2014
leave the engine at a temperature between 900 and 1000 K. Published: February 11, 2014

© 2014 American Chemical Society 2076 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

reformer increases the output of a reactor in contact with an not stored in a separate tank onboard, such an ethanol solution
exhaust gas heat exchanger.6 cannot be combusted with air in the engine efficiently. When
Besides steam reforming, methanol can also be decomposed the temperature difference between the SI engine exhaust gas
to syngas for use as a motor fuel in order to reduce automobile temperature and the ethanol reforming temperature is
emissions.7 The endothermic decomposition of methanol (eq considered, the reaction enthalpy for complete ethanol steam
1) has been demonstrated to produce H2-rich syngas as an reforming to H2 and CO2 (ΔHr = 173.46 kJ/mol) is too high to
engine fuel.8,9 The decomposition products of methanol enable be recovered out of exhaust gas.6
an SI engine to run with an increased compression ratio (14:1), Low-temperature steam reforming of an equimolar ethanol
increasing the thermal efficiency.10 Several patents for solution to a mixture of H2, CH4, and CO2 is a promising
generating hydrogen-rich syngas out of methanol to combust reaction pathway for ethanol steam reforming via exhaust gas
the syngas in an automotive engine have been published.11,12 recovery. A copper-plated Raney nickel catalyst is stable and
active enough to reform ethanol below 300 °C.17 Neat ethanol
CH3OH ⇄ CO + 2H 2 ΔHr1° = 90.36 kJ/mol (1) or gasoline-blended ethanol (E85) can be decomposed to CO,
Cu/Cr-based catalysts are stable in methanol decomposition, H2, and CH4 without water in the fuel using the same catalyst,
but they are still gradually deactivated because the methanol and the liquid fuel containing no water can be used as a direct
decomposition products constitute a reducing reaction environ- fuel too.18 Although the H2 selectivity of low-temperature
ment.13 The higher the amount of CO in the product gas, the ethanol decomposition is not as high as that in methanol
higher is the propensity for carbon deposition to deactivate the decomposition, its reaction enthalpy (ΔHr = 49.1 kJ/mol) is
catalyst because CO in the product gas can be reduced to lower.
carbon by hydrogen (eq 2) or even by itself (eq 3). The main obstacle to the use of ethanol as a sustainable
Furthermore, these reductions to carbon are exothermic engine fuel is its higher cost of commercial production
reactions, the selectivity for which is increased by the compared with methanol. The opportunity cost of renewable
comparably lower decomposition temperature of methanol. ethanol production is higher since it requires input such as corn
or sugar cane, which are in high demand as human food
CO + H 2 ⇄ C(s) + H 2O ΔHr2° = −131.3 kJ/mol reserves. On the other hand, methanol can be produced from
(2) widely available agricultural, forest, and municipal waste as a
renewable feedstock that can be mixed with lower-cost inputs
2CO ⇄ C(s) + CO2 ΔHr3° = −172.5 kJ/mol (3) such as coal and natural gas as fossil feedstocks.19
Reforming of aqueous methanol solutions preserves the If the water-to-methanol molar ratio (w) in the methanol
catalyst better than the decomposition of neat methanol solution is less than unity, one can assume that all of the steam
because steam evaporated out of the solution can oxidize is reduced to hydrogen by methanol according to eq 4. The rest
methanol directly to CO2 (eq 4). The presence of H2O and of the methanol not oxidized by steam decomposes to CO and
CO2 in the reaction environment shifts the thermodynamic H2 according to eq 1 if there is enough heat for 100%
equilibrium from carbon toward CO since both are products of conversion. The dilution of methanol with water increases the
the above reduction reactions. On the other hand, the potential hydrogen selectivity and catalyst stability while decreasing the
for heat recovery is decreased because of the lower reaction overall reaction enthalpy, product gas energy density, and
enthalpy of steam reforming compared with the reaction methanol dehydration selectivity to higher hydrocarbons. If
enthalpy of the decomposition reaction. reactions other than steam reforming and decomposition are
neglected,20 the product gas composition as a function of w (for
CH3OH + H 2O ⇄ CO2 + 3H 2 ΔHr4° = 49.19 kJ/mol 0 < w < 1) can be calculated as w·(eq 4) + (1 − w)·(eq 1),
(4) leading to eq 5a:
Methanol can be reformed with steam to give a H2-rich gas at CH3OH + w H 2O ⇄ wCO2 + (1 − w)CO + (2 + w)H 2
a temperature higher than 160 °C with an Al2O3-supported (5a)
CuO/ZnO catalyst.14 If an equimolar steam−methanol mixture
is reformed by a Cu/Cr catalyst (80 mol % Cu) at 235 °C and Likewise, the overall reaction enthalpy corresponding to eq 5a
atmospheric pressure, the CO2 selectivity reaches 90%.3 This can be calculated as w·ΔH°r4 + (1 − w)·ΔH°r1, leading to eq 5b:
selectivity corresponds to molar concentrations of around 74% ° = (90.36 − 41.17 ·w) kJ/mol
ΔHr5a (5b)
H2 and 1% CO, while the rest of the product gas is mainly CO2.
H2-rich syngas can also be generated from aqueous ethanol The current work will illustrate the equilibrium products of
with a Co catalyst supported on Al2O3 or SiO2 at 400 °C.15 the combustion of an evaporated liquid fuel mixture with
MgAl2O4 is an alternative support to suppress ethanol aqueous-methanol-based syngas produced by the overall
dehydration to ethylene, which leads to catalyst deactivation reaction in eq 5a. The liquid fuel consists of a variable
due to coke deposition.16 Noble metals such as Ru, Pt, and Ir methanol−gasoline mixture. The amount of water in the
are added as promoters to increase catalyst stability and H2 aqueous methanol solution varies the syngas composition
selectivity. Complete conversion of ethanol requires reactor according to eq 5a. The amount of hydrogen in the syngas
temperatures as high as 500 °C. Such a high reaction increases with dilution of the aqueous methanol until aqueous
temperature challenges a reforming reactor heated with engine methanol becomes stoichiometric (w = 1), where the hydrogen
exhaust gas to produce the amount of syngas required to run output is maximized. However, before the determination of the
the engine by syngas alone. combustion products, the impact of syngas generation on the
The stoichiometry of ethanol steam reforming to H2 and system efficiency and the energy density of the fresh charge
CO2 requires 3 moles of water per mole of ethanol in the must be estimated.
solution. This requirement decreases the energy density of the Methanol’s calorific value and heat of vaporization are 676.2
energy stored on board. If water required to reform ethanol is and 37.9 kJ/mol, respectively. Because methanol is stored in
2077 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

liquid form onboard, efficiency calculations must be based on The mean lower heating values of neat-methanol-based and
liquid methanol’s calorific value of 638.3 (=676.2 − 37.9) kJ/ stoichiometric aqueous-methanol-based syngas are equal to
mol. For each mole of methanol reacted, the approximate 255.5 [=(2 × 241.8 + 283)/3] and 181.4 (=3 × 241.8/4) kJ/
amounts of liquid methanol energy contained in 2 moles of H2 mol of fuel, respectively. The denominator of the expression for
and 1 mole of CO are 76% (=2 × 241.8/638.3) and 44% HG represents the specific molar amount of the engine charge
(=283/638.3), respectively.21 based on 1 mole of fuel inducted (denoted below as n0T). In
The resulting heat recovery via the endothermic decom- order to determine this amount, the stoichiometric oxygen is
position reaction (eq 1) (90.36/638.3 = 14%) and vaporization divided by the mole fraction of oxygen in the atmosphere
of liquid methanol (37.9/638.3 = 6%) is 20% of the calorific (0.21), which gives the stoichiometric air. Its product with λ
value of liquid methanol. While the decomposition reaction provides the amount of air inducted into the engine. To this
enthalpy can only be recovered from exhaust heat, the amount of air, unity is added in order to account for the
vaporization enthalpy can also be recovered from the engine amount of evaporated fuel. The specific molar amounts of neat-
coolant because methanol’s boiling point (64.7 °C) is lower methanol-based and stoichiometric (w = 1) aqueous-methanol-
than the standard coolant temperature (90 °C). based syngas at λ = 1 are equal to 3.38 and 2.79 mol of gas/mol
If stoichiometric aqueous methanol (w = 1) is reformed to of fuel, respectively. The ratio of the mean LHV to the specific
syngas according to eq 5a, the product gas contains molar amount is the calorific value of 1 mole of air−fuel
mixture, which is equal to 75.6 and 65 kJ/mol for neat-
approximately 114% (=3 × 241.8/638.3) of the liquid methanol
methanol-based and stoichiometric aqueous-methanol-based
energy, resulting in a 14% higher calorific value. The heat
syngas, respectively.
recovery via the endothermic steam reforming reaction The standard mixture heat value (HG° ) is defined by assuming
accounts for only 8% (=49.2/638.3) of liquid methanol’s that the cylinder state at IVC is equal to the standard state (p°
calorific value. Because neither CO2 nor H2O have calorific = 1 atm, T° = 298.15 K) and replacing the molar density at IVC
value, they are not taken into account. (ρ) with the standard molar density (ρ° = 0.041 mol/L):
Even though heat recovery due to vaporization remains
unchanged (at 6%) with respect to w, vaporization of water in LHV
HG° = ρ° ·
aqueous methanol requires additional heat input. Furthermore, [λ ·(ost /0.21) + 1]
the boiling point of an equimolar aqueous methanol solution
The use of ρ° is a fair estimate, as the fresh charge is not
(73.5 °C) is higher than that of neat methanol but still less than
throttled or supercharged. The formula for H°G no longer
the standard engine coolant temperature. The theoretical contains engine-specific data and enables comparison between
reactor efficiency of neat-methanol-based syngas appears to be different fuel options. Evaluation of this formula for a
7.4% higher than that of aqueous-methanol-based syngas. stoichiometric air−fuel mixture results in standard mixture
However, the production of neat-methanol-based syngas heat values of 3.45 kJ/L for isooctane and 3.39 kJ/L for
challenges the catalyst stability much more than the production methanol, while the standard mixture heat values of neat-
of aqueous-methanol-based syngas. methanol-based and stoichiometric aqueous-methanol-based
Engines fueled with a H2−CO mixture have good efficiency syngas decrease to 3.09 and 2.66 kJ/L, respectively.
at low indicated mean effective pressure (IMEP) equivalent to When λ is doubled to form a lean mixture, the decrease in
engine torque. However, according to engine tests, a H2−CO the standard mixture heat value is much more pronounced for
mixture could not achieve the same maximum IMEP compared liquid fuels than for syngas. At λ = 2, HG° drops to 1.74 kJ/L for
with engines fueled with gasoline or liquid methanol because isooctane, 1.81 kJ/L for both methanol and neat-methanol-
the energy density of the gaseous reformate is lower than that based syngas, and 1.62 kJ/L for stoichiometric aqueous-
of the liquid fuel itself.22 The system efficiency is increased at methanol-based syngas. When λ is increased further, the values
the cost of a decrease in maximum power. of HG° for hydrogen-rich gaseous fuels decrease further but
Although the power output of fuels with lower energy remain higher than that of gasoline or methanol at the same λ.
density can be improved by increasing the engine compression This shows that the maximum power produced by syngas may
ratio, this alternative is limited by the combustion stability, exceed those of liquid fuels when lean (λ > 2) air−fuel mixtures
which depends on a fuel’s motor octane number. The energy are combusted.20
densities of different fuel options are quantified by the mixture If the engine’s kinetic parameters are adjusted properly, the
heat value (HG), which represents the energy density of an peak cylinder pressure during combustion represents the
engine’s fresh charge. The higher the mixture heat value, the engine’s maximum power output. In an experimental analysis
higher is the maximum power produced by the engine. of an SI engine working at 1800 rpm and an absolute air
The mixture heat value is defined as the ratio of the engine pressure of 61.5 kPa, λ was increased from stoichiometric to
charge’s total energy (Qt) to the engine’s displacement volume 1.36. The peak cylinder pressure decreased by 24.9% for
(Vd) (eq 6). The mixture heat value of a neat fuel (in units of combustion of lean gasoline and by 18.4% for combustion of
kJ/L) can be determined from the fresh charge’s molar density 2.5% ethanol-based syngas-blended gasoline, while the
(ρ) after intake valve closure (IVC), the relative air/fuel ratio indicated thermal efficiency increased from 36% to 38.6% for
(λ), the mean lower heating value (LHV), and the lean gasoline and to 40% for lean syngas-blended combustion.23
stoichiometric oxygen (ost), as shown in eq 6.19 These The peak engine brake thermal efficiency of an SI engine at
parameters will be defined and determined for any gaseous 1400 rpm and an absolute air pressure of 61.5 kPa increased
from 30.55% at λ = 1.13 to 31.95% at λ = 1.25 when pure H2
fuel mixture in the next section.
was blended with methanol at a volume fraction of 3%. The
Qt LHV brake mean effective pressure was increased by about 19.3% at
HG = = ρ· λ = 1.39 for the same amount of H2 blending.24 Besides
Vd [λ ·(ost /0.21) + 1] (6) increasing the maximum engine brake thermal efficiency and
2078 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

brake mean effective pressure, blending of methanol with H2 eq 6 and rp in eq 7. Substituting Qp + Qd for Qt in eq 7 and
also shifted the maximum engine brake thermal efficiency to rearranging gives eq 8, which shows that Qd is determined by rp
higher λ. and Qp.
The higher efficiency and higher engine peak pressure in lean
combustion of syngas-blended gasoline and hydrogen-blended 1 − rp
methanol can be explained by the higher flame and diffusion Q d = Q p·
rp (8)
speed of CO and H2 compared with gasoline and methanol.
The higher flame and diffusion speed initiate more complete
Mixing of liquid fuel and methanol-based syngas can alter the
combustion of the air−fuel mixture in the engine because CO
power produced according to road load. The air/fuel ratio must
and H2 stimulate the formation of O and OH radicals,
improving the chain reaction in the fuel combustion process.25 be decreased to form a near-stoichiometric mixture for a proper
High engine pressure may cause combustion instabilities load response during increasing road load. The syngas amount
during combustion of fuels with higher syngas or H2 blending has to be substituted by liquid methanol or gasoline
due to a higher radical concentration. Because of the high flame simultaneously to approach the maximum engine power
speed, very low requirement of ignition energy, and small produced with liquid methanol or gasoline only. Liquid fuel
quenching distance of hydrogen, attempts to burn either can be injected into the intake port or into the engine to
hydrogen-enriched fuels or hydrogen itself are plagued by increase the mixture heat value.
abnormal combustion of hydrogen in the form of preignition, In addition to power adjustment, substituting methanol or
backfiring, and knocking. These problems are exacerbated gasoline in place of syngas at higher loads also results in more
particularly when the relative air/fuel ratio is dropped to stable combustion thanks to the higher motor octane numbers
stoichiometric in order to increase the engine power.1 of both methanol and gasoline compared with hydrogen.
Furthermore, because of the high flame speed of H2, its Methanol increases the mixture’s average motor octane
addition tends to increase the combustion temperature, which number, leading to stable combustion of a near-stoichiometric
increases NO emissions under stoichiometric SI combustion.26 hydrogen−methanol−air mixture. Also, addition of alcohol
On the other hand, the NO, HC, and CO emissions of an blends to gasoline boosts the octane number while reducing
idling SI engine at 790 rpm all decreased when λ was increased CO and NOx emissions.32
from stoichiometric to 1.3 while H2 was blended with gasoline To decrease the engine power, liquid fuel must be partially
at a volume fraction of 3% simultaneously.27 Hydrogen-rich substituted with syngas. Simultaneously, the air/fuel ratio must
syngas mixtures may be more appropriate in lean burn be increased and adapted to the premixed fuel ratio to prevent
applications where the combustion temperature is moderated engine knock. The power can be further decreased if the syngas
by excess air. Such mixtures could also serve in “dual-fuel” amount is also throttled, until the lean flammability limit of
engines that operate under compression ignition using a pilot hydrogen is reached. This keeps the manifold pressure constant
injection of diesel fuel.28 Because syngas is generated from to run the engine unthrottled with hydrogen, regulating the
methanol, a dual-fuel SI engine operated with methanol and power output by varying the syngas flow and thereby excess
methanol-based syngas does not require storage of two air.33 An onboard decomposed methanol engine can be
different fuels.
operated with a wide-open throttle over a large torque range
The cold start of an SI engine with methanol is very difficult
at a practical maximum air excess of around 4.21 This method of
to perform below 5 °C because of difficulties in preparing an
ignitable gaseous air−fuel mixture due to the higher boiling power regulation eliminates pumping losses of an SI engine,
point and heat of vaporization of methanol compared with although a homogeneous charge is combusted in the engine.
gasoline.21 The production of decomposed methanol onboard
the vehicle for use as a cold-start fuel is a possible solution. 2. DETERMINATION OF COMBUSTION PRODUCTS AT
After heat-up of the engine and reformer is complete, an CHEMICAL EQUILIBRIUM
appropriate amount of produced gas can be stored in a small, The theoretical complete combustion of an oxygenated fuel
moderately pressurized metal flask for the next few cold with excess air (λ > 1) is shown in eq 9. The relative air/fuel
starts.29 The method of starting an engine with a battery is ratio must be higher than unity to minimize the amount of
based on a similar strategy of storing energy from a previous incomplete combustion products.
steady run to be used in the next transient cold start.
Mixing of liquid fuel and hydrogen-rich syngas is also a
strategy to reduce emissions of diesel engines. Exhaust gas fuel Cc HhOo + ost ·λ(O2 + 3.76N2)
reforming in diesel engines has been studied experimentally as a h
→ cCO2 + H 2O + ost [(λ − 1)O2 + λ ·3.76N2]
way to assist premixed charge compression ignition (PCCI) 2 (9)
operation. By substitution of part of the main fuel with
hydrogen-rich gas, NOx and smoke emissions can be reduced The stoichiometric amount of O2 required for theoretical
simultaneously.30 The premixed fuel ratio (rp) is the ratio of the complete combustion of an oxygenated hydrocarbon fuel to
calorific energy of the premixed fuel (syngas) (Qp) to the total CO2 and H2O (ost, in units of mol of O2/mol of fuel) is
fuel energy of the fresh charge (eq 7):31 determined in eq 10 as a function of the fuel’s chemical
composition. The amount of oxygen inducted into the engine
Qp with fresh charge is the actual oxygen (O2,in), which is equal to
rp =
Qt (7) the product of ost and λ. The excess oxygen and unreacted
nitrogen are added to the amount of CO2 and H2O to
The sum of Qp and the direct fuel energy (Qd) is equal to the determine the total amount of complete combustion products
total fuel energy Qt used to determine the mixture heat value in (nC, in units of mol of gas/mol of fuel).
2079 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

Table 1. Enthalpies and Free Enthalpies of Formation and Polynomial Coefficients for the Heat Capacity
Δhf° (J/mol) Δgf° (J/mol) A 103·B (K−1) 106·C (K−2) 10−5·D (K2)
H 218000 203290 2.50 0 0 0
O 249170 231770 2.52 −0.026 0.0076 0.108
OH 39460 34280 2.88 0.891 −0.127 0.556
N 472680 455510 2.52 −0.021 0.0069 −0.017
NO 90291 86550 3.48 0.784 −0.154 −0.308
H2 0 0 2.97 0.696 −0.065 0.440
H2O −241818 −228572 3.00 2.299 −0.361 0.397
CO −110525 −137169 3.28 0.893 −0.173 −0.189
CO2 −393509 −394359 5.45 1.530 −0.295 −1.728
O2 0 0 3.79 0.513 −0.062 −0.602
N2 0 0 3.17 0.946 −0.179 −0.048
C8H18 −208750 16260 4.11 70.57 −22.21 0
CH3OH −200660 −161960 2.21 12.22 −3.450 0
C 0 0 1.77 0.771 0 −0.867

h o enough and the temperature of the burnt mixture is high

ost = c + − enough to satisfy the ideal gas relationship. For an ideal gas i,
4 2
the molar free enthalpy (gi) is equal to the chemical potential
h (μi) of the reacting ideal gas i. The molar free enthalpy of
nC = c + + (4.76·λ − 1) ·ost
2 (10) reacting gas i at standard pressure, g°i (T), is determined from
the standard molar enthalpy of formation (Δhf,i° ), the standard
The standard reaction enthalpy of combustion is determined
molar entropy (sT°° ), the change in molar enthalpy Δhi(T), and
in eq 11 in terms of the standard formation enthalpies of the
the change in molar entropy Δsi(T) as a function of the
° ), CO2 (Δh°f,CO2) and H2O (Δh°f,H2O). The absolute
fuel (Δhf,CHO
reaction temperature T (eq 14):
value of the standard combustion enthalpy (ΔH°c ) is the lower
heating value (LHV) of the fuel, which can be inserted into eq gi°(T ) = Δh°f , i + Δhi(T ) − T ·[sT°°, i + Δsi(T )] (14)
6 to determine the mixture heat value.
The standard molar entropy (s°T°) is the ratio of the difference
h between the standard molar enthalpy of formation (Δhf°) and
ΔHc◦ = ◦
c·Δhf,CO2
◦
+ ·Δhf,H 2O
◦
− Δhf,CHO the standard molar free enthalpy of formation (Δgf°) to the
2
standard temperature (T°), as shown in eq 15, where the
LHV = |ΔHc◦| (11) changes in molar enthalpy and entropy are also defined.
2.1. Fundamentals of the Chemical Equilibrium Δhf,°i − Δg f,°i
Calculation. The assumption of chemical equilibrium to sT°°, i =
determine the composition of the combustion products is a T°
good approximation for performance estimates in engines. T
During combustion at elevated temperatures, intermediate Δhi(T ) = ∫T ° Cp,i dTi
products such as H2, CO, and NO and radicals such as H, N,
OH, and O are formed along with the complete combustion T Cp , i
products in eq 9. The products of the reaction in eq 12 are the
Δsi(T ) = ∫T ° Ti
dTi
(15)
major species formed during combustion assuming chemical
equilibrium. The molar heat capacity at constant pressure (Cp), which is
required for the determination of the molar enthalpy and
Cc HhOo + ominλ(O2 + 3.76N2) entropy changes, can be assumed to be constant for small
temperature differences. However, a polynomial function such
⇄ nCO2CO2 + n H2OH 2O + n N2 N2 + nO2O2 + nCOCO
as in eq 16, in which R is the universal gas constant (8.3145 J
+ n H2 H 2 + nH H + nN N + nOO + nOHOH mol−1 K−1), can be used to estimate the heat capacity as a
function of the reaction temperature because the temperature
+ nNO NO (12) change during combustion is very high.35
The products in eq 12 are formed by means of the Cp , i(T ) = R ·(Ai + Bi ·T + Ci·T 2 + Di ·T −2)
simultaneous multiple reactions depicted in eq 13: T
1 1
CO + O2 ⇄ CO2 , H 2 + O2 ⇄ H 2O
⇒ Δhi(T ) = ∫T ° Cp,i(Ti) dTi (16)
2 2
Gasoline is a mixture of several hydrocarbons with high
2H ⇄ H 2 , 2O ⇄ O2 , H + OH ⇄ H 2O knock resistance. In this article, gasoline is idealized and
2N ⇄ N2 , N2 + O2 ⇄ 2NO (13)
represented by isooctane. The coefficients A, B, C, and D for
carbon, methanol, and isooctane (Table 1) were obtained from
The gas species that make up the working fluids in internal several sources36−38 and are valid below 1500 K. Although the
combustion engines can usually be treated as ideal gases.34 The combustion temperature is much higher than 1500 K, these
partial pressures of the gases in the unburnt mixture are low fuels are oxidized or dissociated before the combustion
2080 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

temperature is reached. The polynomial coefficients of the f g = {0, 0, 0, 0, 0, x H2 , 0, xCO , xCO2 , 0, 0, 0, 0, 0}T
other gases are applied to the same polynomial type and were
obtained by least-squares optimization of the gases’ enthalpy 1−w w 2+w
xCO = , xCO2 = , x H2 =
values from a different source.39 Values of Δhf° and Δgf° for the 3+w 3+w 3+w
products and reactants during combustion are listed in Table (19)
1.35
The premixed fuel vector fp is the product of the premixed
For each variable listed in the column headings of Table 1,
fuel’s mole fraction xp (0 ≤ xp ≤ 1) and the product gas vector
the corresponding values for the gases were collected into a
fg (eq 20):
column vector. Instead of entering a scalar value for a single gas
into any of the formulas presented above and doing the same f p = x p· f g (20)
calculation for each gas, the column vector was entered into the
formulas embedded in a code written with Mathematica. The The liquid fuel is composed of isooctane and methanol. The
equations of mass conservation, equilibrium, and energy were alcohol content of the liquid fuel stored in the fuel tank (0 ≤ al
determined by parameters and algebraic formulas entered into a ≤ 1) determines the liquid fuel vector (fl) (eq 21):
code that determines the product distribution numerically. f l = {0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, al, 1 − al, 0}T (21)
The mass conservation during combustion can be
represented by the conservation matrix (A) depicted in Table The product of the direct fuel’s mole fraction (xd = 1 − xp)
2. Each gas i present among the combustion products is and the liquid fuel vector determines the direct fuel vector fd
(eq 22):
Table 2. Conservation Matrix fd = (1 − xp) ·f l (22)
After fp from eq 20 and fd from eq 22 are substituted into eq
18, the fuel vector f becomes a linear combination of the
product gas vector and the liquid fuel vector (eq 23):
f = xp·f g + (1 − xp) ·f l (23)
The sum of the components of the fuel vector must be unity
because all of the components of the reactant and product
vectors are based on 1 mole of fuel. The unity condition is
satisfied because the sums of the components of fl and fg equal
represented by a column of the matrix. Each entry aki shows the unity, too.
number of atoms of the element listed in row k (C, H, O, N) in n n n
the formula of gas i. Mass conservation provides one equation ∑ fg,i =1∧ ∑ fl,i =1⇒ ∑ fi =1
for the conservation of each element. i=1 i=1 i=1
The four conservation equations of the elements C, H, O, The mole fraction of the premixed fuel (xp) must be
and N are derived by setting the matrix product of the determined as a function of the premixed fuel ratio rp. The
conservation matrix (A) and the reactant vector (n0) equal to LHVs of the direct and premixed fuel mixtures must be
the matrix product of A and the product vector (n). The calculated to determine xp from rp. The LHV of an oxygenated
reactant vector can be determined as a function of parameters fuel is given by eq 11. The numbers of atoms of carbon (c) and
such as the relative air/fuel ratio (λ), the water content of the hydrogen (h) in the formula of each gas are listed in the first
methanol solution (w), the premixed fuel ratio (rp), and the row (A[1]) and second row (A[2]) of A, respectively. The
alcohol content of the direct fuel (al). The product vector n is standard formation enthalpy vector (Δh°f ) is shown in the first
the solution of chemical equilibrium calculation and satisfies eq column of Table 1. The standard combustion enthalpy vector
17. (ΔHc°) is determined by inserting the carbon and hydrogen
An 0 = An (17)
row vectors of A into eq 11, as shown in eq 24:
1
2.2. Determination of the Reactant Vector. The ΔH◦c = A[1]·Δhf,CO
◦
+ ◦
A[2]·Δhf,H 2O
− Δh◦f T
2
2 (24)
reactant vector is composed of the fuel vector (f) and the air
vector (a). The fuel vector affects the air vector according to The lower heating value vector (LHV) is the negative of the
the complete combustion reaction in eq 9. The fuel is standard combustion enthalpy vector because the LHV is the
composed of gaseous premixed fuel and liquid direct fuel. absolute value of the standard combustion enthalpy and the
Consequently, the fuel vector is represented as the vector sum latter quantity is always negative:
of the premixed fuel vector (fp) and the direct fuel vector (fd) LHV = |ΔHc◦| ∧ ΔHc◦ < 0 ⇒ LHV = −ΔH◦c (25)
(eq 18). Both fp and fd must be determined to calculate the fuel
and air vectors. The dot products of the lower heating value vector with the
premixed fuel vector, direct fuel vector, and fuel vector
f = f p + fd (18) determine the premixed fuel, direct fuel, and total fuel energies,
respectively, based on 1 mole of fuel inducted. Multiplying
The premixed fuel is a product gas generated by the reaction these scalar products by the molar amount of the fuel inducted
of methanol with water and is composed of CO, CO2, and H2. into the engine until IVC (nf) determines the corresponding
The product gas vector (fg) (eq 19) is determined by the energies Qp, Qd, and Qt consumed during an engine cycle (eq
stoichiometric coefficients of the reaction in eq 5a: 26):
2081 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

Q p = (LHV·f p) ·n f Scheme 1. Flow Diagram of the Proposed Model

Q d = (LHV·fd) ·n f
Q t = Q p + Q d = (LHV·f) ·n f (26)
The definition of the premixed fuel ratio in eq 7 is used to
derive a relationship between rp and xp. This determines xp as a
function of rp and the mean LHV ratio of liquid to gaseous fuel
(d), as defined in eq 27. The parameter d is a function of w and
al. The value of xp determined using eq 27 can be inserted back
into eq 23 to evaluate f.
Qp LHV·f p rp·d
rp = = ⇒ xp = ,
Qt LHV·f 1 + rp·(d − 1)
LHVl LHV·f l
where d = = = f (w , al)
LHVg LHV·f g (27) mixture. Although the diagram shows two separate tanks
The stoichiometric oxygen amount for the combustion of an containing the premixed fuel and direct fuel, both fuels may
oxygenated fuel is determined in eq 10. The coefficients of C, also be supplied from the same source.
H, and O constitute a vector. Its product with the conservation The premixed fuel may be generated from pure methanol or
matrix generates the stoichiometric oxygen vector (ost) (eq 28). a mixture of aqueous methanol and gasoline. If gasoline is in the
If a component of ost is negative, the corresponding gas is an premixed fuel, the above calculation must consider the amount
oxygen source. of unreacted gasoline leaving the reformer. The storage of
premixed and direct fuel in separate tanks in Scheme 1 should
ost = {1, 14 , − 12 , 0}A (28)
demonstrate the effects of the amount of methanol in the direct
fuel and the amount of water in the premixed fuel
independently.
The stoichiometric amount of oxygen required for the Equation 33 shows how the mixture heat value HG can be
combustion of the fuel mixture (O2,st) is the dot product of ost evaluated for an arbitrary fuel mixture by inserting Qt from eq
and f (eq 29): 26 into eq 6 and replacing Vd in eq 6 by n0T/ρ, since the molar
O2,st = ost·f (29) density is given by ρ = n0T/Vd:
Qt LHV·f
The actual amount of oxygen inducted (O2,in) is the product of HG = = ρ·
the relative air/fuel ratio (λ) and the fuel mixture’s Vd n0T (33)
stoichiometric oxygen amount (O2,st). The amount of nitrogen
As in the case of eq 6, the mixture heat value can be based on
inducted into the engine (N2,in) is determined according to its
the standard state by replacing the molar density ρ with the
fraction in the atmosphere:
standard molar density ρ°:
0.79
O2,in = λ ·O2,st , N2,in = ·O2,in LHV·f
0.21 (30) HG° = ρ° ·
n0T
The air vector based on 1 mole of fuel (a) contains only O2
and N2 as components (eq 31): If the direct fuel is injected into the engine in the liquid phase,
the volume of the liquid fuel can be neglected, and the control
a = {0, 0, 0, 0, 0, 0, 0, 0, 0, O2,in , N2,in , 0, 0, 0}T (31) variable i will not include the fuel as a component in the
reactant vector (i.e., i ≤ 11).
The reactant vector (n0) is the sum of the fuel and air vectors 2.3. Determination of the Product Vector. The reactant
(eq 32): vector in eq 32 can be inserted into eq 17. The components of
n0 = f + a (32) n0 are shown explicitly in Table 3. The product vector (n)
shown in Table 3 is substituted by the product of the total
For a more exact calculation, the amount of exhaust gas left in molar amount of combustion products (nT) and the mole
the cylinder from the previous cycle could be added to the sum. fraction vector (y). The components of the mole fraction
The sum of the reactant vector components is equal to the vector are the mole fractions of the combustion products (yi).
molar amount of the fresh engine charge (n0T) which is also the The values of yi are between 0 and 1, which are easier to
denominator of eq 6.
14 Table 3. Product and Reactant Vectors Setting up the
n0T = ∑ n0,i Conservation Equation System
i=1

Scheme 1 shows a flow diagram of the model presented in

this work. The vectors defined above are represented as gas
streams in this diagram. The mixing of liquid fuel with hot
reformate evaporates the liquid fuel, cools the hot reformate,
and facilitates the formation of a more homogeneous fuel
2082 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

calculate numerically compared with the components of n, reaction pressure. The equilibrium constants can be determined
which may be any positive real number. An additional equation as functions of temperature using the least-squares method.40
must be set up to determine the additional unknown nT. This The seven reactions in eq 13 provide seven equations of the
equation expresses the fact that the sum of the components of above type. In addition, the five conservation equations in
the mole fraction vector is unity. The fuels and carbon are Table 4 are added to the equation system. The eleven
assumed to be fully consumed, so their corresponding amounts components of the mole fraction vector and nT in Table 4
(the last three components of the product and mole fraction represent a total of 12 unknowns in total. Therefore, the
vectors) are set equal to 0. equation system can be solved numerically for a given reaction
The matrix multiplication in Table 3 is carried out in Table 4 temperature.41 However, the solution for multiple reactions
for the case of pure isooctane without premixed fuel to derive does not lend itself to standardization that allows a computer
the conservation equations for the elements C, H, O, and N. solution, especially if the reaction temperature is to be
determined by adding the energy equation to the equation
Table 4. Conservation Equation System for Combustion of system.
Pure Isooctane (al = 0, rp = 0, xp = 0) An alternative equilibrium criterion applied to multiple
reactions is the minimization of the system’s total free enthalpy.
This method offers a standard solution procedure based on the
method of Lagrange undetermined multipliers.42 This criterion
has hardly been applied to combustion calculations where the
fuel properties vary with the premixed fuel ratio and the liquid
fuel composition. Moreover, the number of different gases
assumed to be present in the burning mixture can be increased
for a more realistic equilibrium calculation. Instead of reactions,
The initial value of nT in the numerical calculation is set equal this method assumes only the presence of different gases in the
to the molar amount of complete combustion products nC reaction environment.
defined in eq 10. For the fuel mixture f defined in eq 18, this The free enthalpy of formation for a gas (ΔG°f ) is the free
amount can be determined from the dot products of the row enthalpy change for the reaction in which the gas is formed
vectors A[1] and A[2] with f, which give the molar amounts of from its elements. The reactants of the formation reaction are
CO2 and H2O, respectively (eq 34). Excess oxygen and the elemental substances from which the gas is formed (C, H2,
unreacted nitrogen are added to this sum of the dot products, O2, N2), and the product is the gas for which ΔG°f is to be
as was done in eq 10. determined. The free enthalpies of each gas and its constituent
elements in the formation reaction must be determined
1
nC = A[1]·f + A[2]·f + (λ − 1)·O2,in + N2,in according to eq 14. Instead of evaluating ΔGf° for each gas
2 (34) separately by inserting the free enthalpies into eq 35, the ΔG°f
The products of the combustion reaction in eq 12 are formed values of all of the gases are embodied in the vector of the
through multiple reactions that are assumed to be in system’s free enthalpy change (ΔGf°). The definition of ΔGf° in
equilibrium. The standard free enthalpy change for a reaction, terms of the conservation matrix’s row vectors and the free
ΔG°(T), is defined as the difference between the standard free enthalpies of the elements, an additional contribution of this
enthalpies of the products and reactants weighted by their work, is shown in eq 37:
stoichiometric coefficients. If signed stoichiometric coefficients
g H° (T )
(νi negative for reactants and positive for products) are used, ΔG°f (T ) = g °(T ) − gC°(T ) ·A[1]T − 2
·A[2]T
ΔG°(T) can be written as the sum of the standard free 2
enthalpies weighted with νi (eq 35). For the calculation of the gO° (T ) g N° (T )
standard free enthalpies g°i (T) according to eq 14, each product − 2
·A[3]T − 2
·A[4]T
and reactant is assumed to be a pure substance at the standard- 2 2 (37)
state pressure and at the reaction temperature T.
n
where g°(T) is the vector of standard free enthalpies of the
gases.
ΔGi°(T ) = ∑ νi·gi°(T ) ΔGf° in eq 37 is evaluated as a column vector on the left-
i=1 (35)
hand side of Table 5. The corresponding formation reactions
The value of ΔG°(T) at the end-state temperature
determines the equilibrium constant K(T) for that reaction Table 5. Vector of the Free Enthalpies of Formation
(eq 36):
K (T ) = e−ΔG°(T )/ RT (36)
The equilibrium constant depends on temperature and
provides a relationship among the partial pressures (pi) of the
gases, assuming that they are ideal:
n
n ⎛ p ⎞∑i=1 νi
K (T ) = ∏ ⎜ i ⎟
i=1 ⎝
p° ⎠
The partial pressures can be replaced with the corresponding
mole fractions using the fact that pi = p·yi, where p is the
2083 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

are depicted on the right-hand side of the table. The dU = dQ + dW ⎫

components of the column vector generated by eq 37 reflect ⎪
the free enthalpies of formation reactions of the gases according dH = dU + p dV + V dp ⎬ ⇒ dH = dQ + V dp
⎪
to eq 35. The free enthalpies of formation of C, H2, O2, and N2 dW = − p dV ⎭
are zero because they are elemental substances.
The condition of chemical equilibrium in the method of the adiabatic ⇒ dQ = 0 ⎫ ⎪

Lagrange undetermined multipliers is shown in eq 38. The ⎬ ⇒ dH = 0 ⇒ H0 = HT

isobaric ⇒ V dp = 0 ⎭ ⎪

(40)
derivation of this equation is shown in ref 35. The components
λk of the vector of undetermined multipliers λ = (λC, λH, λO, The initial enthalpy (H0) is the sum of the initial thermal
λN) are the undetermined multipliers corresponding to the enthalpy (n0·hT0) and the initial chemical enthalpy (n0·Δhf°),
elements C, H, O, and N that constitute the gases in the while the final enthalpy (HT) is the sum of the final thermal
reaction environment. The fuels and carbon are assumed to be enthalpy (n·hT) and the final chemical enthalpy (n·Δh°f ) (eq
fully consumed according to the product vector in Table 3, and 41). The molar thermal enthalpies are based on the standard
therefore, the control variable i varies from 1 to 11. temperature T° and can be calculated using the molar heat
capacities at constant pressure (Cp). If the reactants have
⎛ p⎞ 4 different initial temperatures, the initial temperature T0 must be
ΔGf,°i + RT ln⎜⎜yi · ⎟⎟ + ∑ λk ·aik = 0 substituted with an initial temperature vector T0.
⎝ p° ⎠ k=1 (38)
T0

Equation 38 can be solved for the mole fractions yi of the

H0 = n 0·(Δh°f + h T0), h T0 = ∫T ° Cp(T ′) dT ′
combustion products (eq 39): T
HT = n·(Δh°f + h T), hT = ∫T ° Cp(T′) dT′ (41)
p° −[ΔGf,°i + (λ A)i ]/ RT
yi = ·e
p (39) The energy expression in terms of mole fraction vectors (eq
42) is obtained by inserting the enthalpies in eq 41 into eq 40:
When the reaction temperature and pressure are inserted into
eq 39, the undetermined multipliers will remain as the only n0T ·[y0·(Δh°f + h T0)] = nT ·[y·(Δh°f + h T)],
14
unknown parameters.
The mole fractions determined in eq 39 are inserted into the n0T = ∑ n 0i , y0 = n 0/n0T
conservation equations in Table 3. After the resulting 5 × 5 i=1 (42)
equation system has been solved numerically for the four After the mole fraction vector y in eq 39 is inserted into both
undetermined multipliers and the molar amount of combustion eq 42 and the conservation equations derived in Table 3, eq 42
products nT, the undetermined multipliers are determined. must be solved simultaneously with the conservation equations.
These parameters can be inserted back into eq 39 to evaluate The resulting 6 × 6 equation system can be solved for the four
the mole fraction vector. The product of nT and y is equal to undetermined multipliers, the molar amount of the combustion
the product vector. The product vector components display the products nT, and the reaction temperature T, which is equal to
molar amounts of the combustion products based on 1 mole of the adiabatic flame temperature at constant pressure (Tp). The
fuel. undetermined multipliers and the flame temperature must be
2.4. Determination of the Adiabatic Flame Temper- inserted back into eq 39 to determine the mole fraction and the
ature. The final temperature of the products in an adiabatic product vectors.
combustion process is called the adiabatic flame temperature. 2.4.2. Adiabatic Flame Temperature for a Constant-
This is the reaction temperature T presumed for the calculation Volume Process. A constant-volume change of state is called an
of the product vector. Constant-volume and constant-pressure isochoric process. During an isochoric combustion, the volume
combustion are two ideal processes of an adiabatic combustion. of the combustion chamber (V) remains constant. Con-
sequently, the reaction pressure of the isochoric process (pV)
In reality, the pressure and the volume change simultaneously
has to be determined with an equation of state. The ideal gas
during combustion in the engine’s combustion chamber.
equation is an appropriate equation of state for internal
The higher the flame speed of an air−fuel mixture, the faster
combustion. Applying the ideal gas equation to the initial state
the combustion is finished. Thus, a high flame speed indicates (subscript 0) and the final state (subscript V) of the adiabatic
that a constant-volume process describes the adiabatic flame isochoric combustion allows pV to be determined (eq 43). The
temperature better than a constant-pressure process because reaction pressure depends on the flame temperature of the
the change in volume is small during a fast combustion. adiabatic isochoric combustion (Tv).
Because most fuels have a finite flame speed and the
combustion takes tens of crank angles, the real change of TV nT
p0 ·V = n0T ·R ·T0 , pV ·V = nT ·R ·TV ⇒ pV = p0 · ·
state during combustion is somewhere between constant- T0 n0T
pressure combustion and constant-volume combustion. (43)
2.4.1. Adiabatic Flame Temperature for a Constant- The final combustion pressure pV can be inserted into eq 39
Pressure Process. For an adiabatic constant-pressure process of instead of the pressure p of the constant-pressure process. The
a steady-state open system, the enthalpy (H) remains constant. mole fraction vector for the constant-volume process is then
The initial state of the system is the state of the reactants, determined as a function of undetermined multipliers, the
identified by the subscript 0. The final state of the system is the molar amount of combustion products nT, and the adiabatic
state of the products, identified by the subscript T. flame temperature TV (eq 44):
2084 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

Figure 1. Isothermal mole fractions of combustion products (y) from (left) this work and (right) Heywood, 1988.34

p° T0·n0T −[ΔGf,°i + (λ A)i ]/ RTV energy equation for the constant volume process (eq 47). The
yi = · ·e calculation with internal energies is reduced to the calculation
p0 TV ·nT (44)
with enthalpies, which were defined in eq 42.
The internal energy (U) of a closed system remains constant
during an adiabatic constant-volume process. The enthalpies in n0T ·[y0·(Δh°f + h T0) − R ·T0]
eq 42 are substituted with internal energies to determine the
= nT ·[y·(Δh°f + h T) − R ·T ]
energy equation of a closed system’s constant volume process
in eq 45. ⇒ H0 − n0T ·R ·T0 = H − nT ·R ·T (47)
dU = dQ + dW ⎫ ⎪ The mole fraction vector of the constant volume process
⎬ ⇒ dU = dQ − p dV from eq 44 can be inserted into eq 47 and into the conservation
⎭
⎪
dW = − p dV
equation system in Table 3. The conservation equations and eq
adiabatic ⇒ dQ = 0 ⎫ 47 constitute a 6 × 6 equation set to be solved for the four
undetermined multipliers, the isochoric adiabatic flame temper-
⎪

⎬ ⇒ dU = 0 ⇒ U0 = UT
isochoric ⇒ pdV = 0 ⎭ ⎪
ature, and nT simultaneously. The undetermined multipliers,
TV, and nT must be inserted back into eq 44 to determine y and
U0 = UT ⇒ n 0·(Δu°f + u T0) = n·(Δu°f + u T) (45) n.
The molar thermal internal energy (uT) can be determined
by integrating the differential du = CV dT, where CV is the 3. RESULTS
molar heat capacity at constant volume, which is equal to the The equation systems presented and the computer model
difference between the molar heat capacity at constant pressure written according to the method in this work were tested by
and the universal gas constant (i.e., CV = Cp − R). The comparing the results with results of a reference. The
subtraction of the scalar R from the vector Cp is mathematically comparison can be easily illustrated by presenting the results
not correct. Therefore, in eq 46 the scalar R is substituted with of this work and the results from the reference in graphical
the product of the scalar R and the unit vector (I) in order to form side by side. The reference analyzes combustion of
subtract R from each component of the vector. isooctane.
T0
Following the side-by-side comparison, the effects of the
u T0 = ∫T ° (Cp(T ′) − R ·I) dT ′ premixed fuel ratio and the direct fuel’s alcohol content on both
the flame temperature and combustion products were analyzed.
T The calculation of exhaust gas content depends also on how
uT = ∫T ° (Cp(T′) − R·I) dT′ fast the fuel is burned by the flame front and how fast the gas
cools during expansion. The flame speeds of different
⇒u T0 = h T0 − R ·(T0 − T °) ·I combustible gases must be taken into account. Nevertheless,
the distribution of combustion products derived from an
u T = h T − R ·(T − T °) ·I assumption of chemical equilibrium must also be given in order
Δu°f = Δh°f − R ·T °·I (46)
to determine the exhaust gas composition.
3.1. Verification of the Presented Model by Compar-
Substitution of the vectors for the internal energy of ison to the Results of a Reference. The mole fractions of
formation (Δuf°) and the initial (uT0) and final (uT) molar isooctane combustion products at a pressure of 30 atm and a
thermal internal energies from eq 46 into eq 45 determines the temperature 2750 K are depicted as functions of equivalence
2085 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

Figure 2. Adiabatic flame temperatures Tp and TV and the isochoric combustion pressure pV from (left) this work and (right) Heywood, 1988.34

Figure 3. Effect of the premixed fuel ratio on the adiabatic flame temperature for (left) λ = 1 and (right) λ = 2.

ratio (ϕ = 1/λ) in Figure 1. In the right panel of Figure 1, the isochoric combustion and increased to pV according to the
results of a reference34 using a model developed at NASA43 are prevailing parameters.
shown. In the left panel of Figure 1, the results of the model Figure 2 shows that an increase in the equivalence ratio from
presented here are shown. Table 4 shows the conservation zero to unity, corresponding to a decrease in the air/fuel ratio
equation system for this evaluation. from infinity to unity, increases the adiabatic flame temper-
The model in the reference assumed the radical N not to be atures because the amount of excess air decreases. Increasing ϕ
present in the reaction environment. For the purpose of further (ϕ > 1) decreases the flame temperatures because of
comparison, the product vector’s fourth component was set to incomplete combustion of the fuel. However, the maximum
0 (yN = nN = 0) in our model. An increase in the equivalence temperatures are reached when the air−fuel mixture is slightly
ratio (i.e., a decrease in λ) promotes combustion products (H2, richer than stoichiometric. Both Figure 1 and Figure 2 show
that the differences between the results of this work (on the
CO) and radicals (H) that contain elements coming from the
left-hand sides) and those of the reference (on the right-hand
fuel. On the other hand, a decrease in the equivalence ratio
sides) are negligible.
promotes combustion products and radicals that contain
3.2. Adiabatic Flame Temperatures. The adiabatic flame
elements coming from the air (O, NO, O2). temperature significantly affects both outputs of the engine:
The adiabatic flame temperatures at constant pressure (Tp) performance and exhaust gases. Methanol’s and isooctane’s
and constant volume (TV) according to eqs 42 and 47 are adiabatic flame temperatures for a constant pressure process are
shown in Figure 2. In addition, the combustion pressure at shown in Figure 3 as functions of the premixed fuel ratio. The
constant volume (pV) according to eq 43 is also shown. The left and right panels of Figure 3 are valid for relative air/fuel
results of this work (Figure 2 left) are almost identical to the ratios of 1 (stoichiometric) and 2, respectively.
results of the reference (Figure 2 right). The results are valid for In both of the charts in Figure 3, the amount of alcohol in the
isooctane without premixed fuel at an initial temperature of T0 direct fuel is shown on the primary vertical axis, where the fuel
= 700 K. The combustion pressure was 10 atm for isobaric consists of direct fuel only (rp = 0). M0 and M100 indicate
combustion. This pressure was also the initial pressure of the direct fuels containing only isooctane and methanol,
2086 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

respectively, while M85 indicates a fuel with a methanol mole by drawing a line parallel to the line of the corresponding water
fraction of 85%. The amount of water in the premixed fuel is amount of the charge temperature at 700 K. This line should
shown on the secondary vertical axis in these charts. The begin at the methanol amount corresponding to the increased
secondary vertical axis is located at rp = 1, indicating that the charge temperature (e.g., 900 K).
fuel consists of premixed fuel only. The effect of the relative air/fuel ratio on the adiabatic flame
The mole fraction of the premixed fuel (xp) is determined as temperature is shown in Figure 4. Except for the curves marked
a function of rp in eq 27. The use of xp instead of rp could have
simplified the equilibrium calculation. However, the function
containing rp is inserted into the fuel vector instead of xp. The
adiabatic flame temperature varies linearly with respect to rp,
and the resulting curves are lines connecting the two points on
the primary and secondary vertical axis of the chart.
Consequently, the choice of the parameter rp instead of xp is
proven to be more appropriate for presenting the results.
According to the Figure 3 (right), the adiabatic flame
temperature of a lean mixture (λ = 2) always increases with
respect to the premixed fuel ratio. The comparison of the two
charts shows that the flame temperature decreases almost
linearly with increasing water content in the premixed fuel
because the amount of inert CO2 in the premixed fuel increases
linearly with the increase in water amount. In contrast to the
water content, the direct fuel’s alcohol content does not have a Figure 4. Effect of the relative air/fuel ratio on the adiabatic flame
linear effect on the flame temperature. The flame temperature temperature.
decreases with increasing alcohol content for near-stoichio-
metric mixtures and increases with increasing alcohol content
for lean fuel mixtures. with M0 and M100, all of the results in the figure are valid for
For a given near-stoichiometric fuel mixture, there exists a fuel mixtures consisting of indirect fuel only (rp = 1). The flame
water content w of the premixed fuel for which the flame temperature decreases with respect to λ for any given parameter
temperature remains constant with respect to the premixed fuel because the amount of inert N2 and excess O2 in the
ratio. The resulting functions are horizontal lines representing combustion products increases with λ. The flame temperature
flame isotherms at this water content. For a stoichiometric M85 for constant-volume combustion is about 250 K higher than
air−fuel mixture the flame isotherm is at 2520 K, located at w = that for constant-pressure combustion. The y axis is situated at
0.612. λ = 0.9 to show the parameters.
In order to reach the maximum power of the engine, the air/ Figure 4 also shows the effect of the charge temperature and
fuel ratio must be decreased to its minimum value above pressure for λ = 1 and λ = 1.5. For these results, λ on the x axis
stoichiometric. When the air/fuel ratio is decreased toward is just a parameter increased from 1 to 1.8 while the air/fuel
stoichiometric, the amount of the premixed fuel must be ratio becomes a constant parameter. T0 and p are multiplied by
decreased to prevent engine knock resulting from a low motor λ to vary T0 from 700 to 1260 K (=1.8 × 700 K) and p from 10
octane number of hydrogen in the premixed fuel. If the water to 18 atm (=1.8 × 10 atm) while holding w equal to unity.
amount of the premixed fuel corresponds to the one of the The increase in the charge pressure from 10 to 18 atm has
direct fuel’s flame isotherms, the adiabatic flame temperature only a marginal effect on the flame temperature. The adiabatic
remains constant with respect to the premixed fuel ratio when flame temperature increases almost linearly from 2431 to 2722
supplementing premixed fuel with direct fuel. Therefore, the K for λ = 1 when the charge temperature is increased from 700
flame isotherm could help define the water amount of the to 1260 K. The linearity of the flame temperature with respect
premixed fuel for a given alcohol amount of the direct fuel. to the charge temperature explains how the results for T0 = 900
While the piston moves to the top dead center, the flame K in Figure 3 (left) can be deduced from the results for T0 =
propagates from the plugs toward the piston and cylinder walls, 700 K. The results for different charge temperatures can be
increasing the pressure of both the burnt gas and unburnt derived by drawing lines parallel to that for the corresponding
charge. While the flame penetrates into the unburnt charge, the water amount of the premixed fuel at 700 K because the
charge temperature increases because of the rise in pressure and adiabatic flame temperature increases linearly with respect to
further compression of the piston. Because the initial T0, w, and rp.
temperature of the charge increases during flame propagation, Figure 5 shows the amounts of combustion products as a
it will have an impact on the adiabatic flame temperature as function of rp while the parameters w, λ, al, p, and T0 are held
well. Therefore, the left chart in Figure 3 contains an additional constant. The combustion products NO (Figure 5 left) and CO
group of results valid for a charge temperature of 900 K (Figure 5 right) decrease almost linearly with increase in the
identified by this temperature. premixed fuel’s water amount, similar to the adiabatic flame
When the premixed fuel ratio of the M100 fuel mixture is temperature. The water in the premixed fuel is reduced to CO2
varied from 0 to unity at a charge temperature of 900 K, the during reforming of methanol to H2 in the reactor. CO2 is an
flame temperature increases only 3 K when w = 0.776. The inert gas and decreases the adiabatic flame temperature and the
M100 fuel mixture with w = 0.776 is a flame isotherm at 700 K. amounts of both NO and CO indirectly. The shapes of the two
Consequently, the additional results for M100 at 900 K are charts are similar to that for the adiabatic flame temperature,
almost parallel to the results of the same direct fuel at 700 K. showing that the flame temperature has a stronger effect on the
The results for a different charge temperature can be estimated combustion products compared with other parameters.
2087 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

Figure 5. Effect of the premixed fuel ratio on the isobaric adiabatic combustion products for (left) NO and (right) CO.

The combustion of isooctane at λ = 1 produces 4500 ppm mixtures. As the fuel mixture gets leaner, the water content of
NO and 18 000 ppm CO in the combustion products at the premixed fuel has a greater impact than the alcohol content
chemical equilibrium according to Figure 5. On the other hand, of the direct fuel on the fraction of combustion products.
an SI engine working at 1800 rpm and an absolute air pressure Figure 6 shows the effect of the relative air/fuel ratio on the
of 61.5 kPa caused tailpipe emissions of 2800 ppm NO and combustion products. The results for NO and CO are shown
2300 ppm CO at λ = 1.23 The combustion products NO and
CO are further reacted to complete combustion products
during the expansion process after combustion in the engine.
The decrease in the amount of CO during expansion from 18
000 to 2300 ppm is much more pronounced than the decrease
in the amount of NO from 4500 to 2800 ppm because the rate
of NO reduction to N2 is lower than the rate of oxidation of
CO at temperatures during expansion.
When ethanol-based 2.5% syngas was added to gasoline, NO
emissions and CO emissions were higher than for gasoline
combustion at the same λ, but a decrease in both emissions was
possible when λ was increased to 1.35.23 However, syngas must
be added to the charge to lean out the charge further.
According to Figure 5, decomposed neat methanol (w = 0)
causes higher CO and NO amounts in the combustion
products than combustion of gasoline as well. On the other Figure 6. Effect of λ on the NO and CO fractions in the combustion
hand, the increase in syngas volume fraction in the charge of a products.
hybrid syngas−gasoline engine from 0% to 1.84% at λ = 1.2
decreased the NO emissions in exhaust gases from 2047 to on the primary (left) and secondary (ight) vertical axes,
1499 ppm in an experimental study.44 respectively. While the CO amounts for methanol (M100),
The amount of NO in the combustion products decreases to isooctane (M0), reformed dry methanol (w = 0), and reformed
2900 ppm from 4600 and 3600 ppm for isooctane and stoichiometric methanol (w = 1) decrease monotonically with
methanol, respectively, when an equimolar water−methanol λ, the NO amounts have a maximum between λ = 1.2 and 1.3.
mixture (w = 1) is reformed in the reactor and used instead of The trend in the NO amount in the combustion products of
liquid fuel (rp = 1) according to Figure 5. Such substitutions of M0, M100, and reformed stoichiometric methanol intersect at λ
gasoline and methanol results in 37% and 19% reductions of = 2, similar to the left panel in Figure 5, while the combustion
NO in the combustion products, respectively. Thus, products of reformed dry methanol remain higher because of its
combustion of syngas reformed out of more aqueous methanol higher adiabatic flame temperature.
causes lower NO and CO emissions even at near-stoichio- The different flame speeds of methanol, gasoline, CO, and
metric. Increasing λ further for load adjustment can decrease H2 are not taken into account in the results presented above.
the emissions further. Substitution of gasoline with methanol Hydrogen has a higher flame speed than other fuels such as
results in a 22% reduction of NO in the combustion products gasoline and methanol, so combustion of stoichiometric syngas
assuming constant-pressure combustion. mixtures could be better described with constant-volume
The flame isotherms for M0, M85, and M100 from the left combustion. On the other hand, gasoline has a lower flame
chart in Figure 3 are also shown in both panels of Figure 5. The speed, and its constant-volume combustion may result in the
lines representing the flame isotherms in Figure 3 show a prediction of a too high adiabatic flame temperature.
decrease in the amounts of NO and CO in the combustion Constant-pressure combustion was assumed in order to
products when the premixed fuel ratio is increased from zero to compare different fuels on the same basis. Therefore,
unity. When the air/fuel ratio is doubled (λ = 2), the results for stoichiometric mixtures with rp near unity may have a higher
different alcohol amounts in the direct fuel overlap. The direct adiabatic flame temperature than predicted because constant-
fuel’s alcohol content does not affect the NO amount for lean volume combustion causes a flame temperature 200−400 K
2088 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

higher than the assumed constant-pressure combustion article could be interesting for decision makers in engine design
according to Figure 2. in the automotive industry and for catalyst manufacturers, oil
The higher flame temperature will also have a negative refineries, fuel distributors, and all NGOs and governments that
impact on NO emissions. In fact, experimental works show an advocate for lower-CO2-emitting vehicles.
increase in NOx emissions if liquid methanol is substituted with
neat-methanol-based syngas while the air/fuel ratio is held
constant.45 Consequently, from a perspective of emission
■ AUTHOR INFORMATION
Corresponding Author
control, a high premixed fuel ratio is only applicable for lean *E-mail: osmansinan@superonline.com.
combustion at part load, which is also a prerequisite for
Notes
combustion stability. Increasing the premixed fuel ratio and
The authors declare no competing financial interest.

■
relative air/fuel ratio simultaneously accomplishes a higher
engine and system efficiency at part load, which decreases CO2
ACKNOWLEDGMENTS
and other noxious emissions based on produced energy.
The authors are grateful for the support of Istanbul Technical
4. CONCLUSION University.
In research related to onboard fuel reforming, the purpose is to
increase the thermal efficiency via recovery of the exhaust heat
of an internal combustion engine and decrease emissions by
■ NOMENCLATURE
Symbols
enabling combustion of leaner mixtures. Hydrogen as the main a = air vector (mol/mol of fuel)
product of reforming allows stable combustion of ultralean A = conservation matrix
mixtures in the engine. Despite its lower charge density, aki = entry of A at row k and column i
hydrogen-rich syngas can be mixed with liquid gasoline or al = mole fraction of alcohol in the direct fuel
methanol to increase the energy density of the engine charge. c = amount of carbon in a molecule
On the other hand, the high flame speed of hydrogen causes Cp = molar heat capacity at constant pressure (J mol−1 K−1)
higher NOx emissions and combustion instability when syngas CV = molar heat capacity at constant volume (J mol−1 K−1)
is combusted with a near-stoichiometric air/fuel ratio. d = mean LHV ratio of liquid to gaseous fuel
The method of Lagrange undetermined multipliers has been f = fuel vector (mol/mol of fuel)
applied to the combustion of hydrogen-rich syngas premixed fd = direct fuel vector (mol/mol of fuel)
with a methanol−gasoline mixture. The method of calculation fg = product gas vector (mol/mol of fuel)
in this article is independent of engine parameters, valid only fl = liquid fuel vector (mol/mol of fuel)
for chemical equilibrium conditions, and presents a well fp = premixed fuel vector (mol/mol of fuel)
organized, well-defined, and easy to follow programming code g°i = molar free enthalpy of gas i based on standard pressure
that takes different parameters related to the fuel’s composition (J/mol)
into account. This is possible because the programming code h = amount of hydrogen in a molecule
expresses gas, liquid fuels, air, reactants, and products as H = enthalpy of the engine charge (J)
vectors. Moreover, the calculation results are expressed as HG = mixture heat value (kJ/L)
vectors. Final exhaust emissions from the tailpipe and thermal H°G = standard mixture heat value (kJ/L)
efficiency can only be determined given specific geometric and K(T) = equilibrium constant for a reaction
cycle parameters of an engine. LHV = lower heating value vector (kJ/mol)
This method can determine the amount of water in the LHV = lower heating value (kJ/mol)
aqueous methanol that is needed for a given specific alcohol LHVg = mean lower heating value of gaseous fuel (kJ/mol)
amount in the liquid fuel in order to maintain a stable flame LHVl = mean lower heating value of liquid fuel (kJ/mol)
temperature regardless of the premixed fuel ratio for a specified n = product vector (mol/mol of fuel)
air/fuel ratio. The results of the article also indicate that with nC = amount of theoretical complete combustion products
increasing amount of water in the aqueous methanol, the (mol/mol of fuel)
adiabatic flame temperature and combustion products such as nf = molar amount of fuel combusted during an engine cycle
CO and NO are reduced. (mol)
When an equimolar water−methanol mixture is reformed to n0 = reactant vector (mol/mol of fuel)
syngas, as the premixed ratio increases, the NO and CO ratio of n0T = molar amount of engine’s fresh charge (mol/mol of
the combustion products decreases regardless of other fuel)
parameters. Combustion of syngas reformed out of an nT = molar amount of combustion products (mol/mol of
equimolar water−methanol mixture causes a decrease in the fuel)
NO ratio by 20% compared with methanol and 40% compared N2,in = amount of nitrogen inducted into the engine (mol of
with gasoline under equilibrium conditions, whereas the N2/mol of fuel)
adiabatic flame temperature decreases by only 50 and 100 K, o = amount of oxygen in a molecule
respectively. This decrease in the NO and CO ratio causes a ost = stoichiometric oxygen vector for theoretical complete
reduction in exhaust emissions since the initial values of the combustion (mol of O2/mol)
NO and CO fraction among postcombustion exhaust emissions O2,st = stoichiometric oxygen for complete combustion of the
are lowered. fuel mixture (mol of O2/mol of fuel)
In future research, the addition of kinetic data on combustion O2,in = amount of oxygen inducted into the engine (mol of
reactions and geometric parameters of an engine to this O2/mol of fuel)
programming code will enable the calculation of tailpipe p = reaction pressure (Pa)
exhaust emissions, indicated power, and efficiency as well. This pi = partial pressure of gas i (Pa)
2089 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091
Energy & Fuels Article

p° = standard pressure (1 atm = 101 325 Pa) p = variable of an isobaric process

pV = isochoric combustion pressure (Pa) T = variable based on the reaction temperature T
Q = heat exchanged between the engine and its surroundings V = variable of an isochoric process

■
(J)
Qd = direct fuel energy (J) REFERENCES
Qp = premixed fuel energy (J)
Qt = total fuel energy (J) (1) Jamal, Y.; Wyszynski, M. L. Int. J. Hydrogen Energy 1994, 19,
557−572.
rp = premixed fuel ratio (2) May, H.; Gwinner, D. Motortech. Z. 1981, 42, 125−130.
R = universal gas constant (8.3145 J mol−1 K−1) (3) Lindström, B.; Pettersson, L. J. Int. J. Hydrogen Energy 2001, 26,
sT0° = standard molar entropy (J mol−1 K−1) 923−933.
T = reaction temperature (K) (4) Finegold, J. G. Int. J. Hydrogen Energy 1978, 3, 83−104.
T0 = temperature of unburnt charge (K) (5) Haji, S.; Malinger, K. A.; Suib, S. L.; Erkey, C. Fuels and Fuels
T° = standard temperature (298.15 K) Processing. In Fuel Cell Technology: Reaching towards Commercializa-
Tad = adiabatic flame temperature (K) tion; Sammes, N., Ed.; Springer: London, 2006; pp 165−202.
Tp = combustion temperature of an isobaric adiabatic process (6) Süslü, O. S.; Becerik, I. Energy Fuels 2009, 23, 1858−1873.
(K) (7) Pefley, R. K.; Saad, M. A.; Sweeney, M. A.; Kilgore, J. D. SAE Pap.
TV = combustion temperature of an isochoric adiabatic 1971, 719008.
(8) Voecks, G. E.; Dawson, S.; Houseman, J. Operation of a catalytic
process (K)
methanol decomposition reactor for vehicular use. Proc. 4th Int. Symp. on
u = molar internal energy of the gas in the engine (J/mol) Alcohol Fuels Technology, B-9, Guaruja, Brazil, 1980.
U = internal energy of the gas in the engine (J) (9) Finegold, J. G.; Karpuk, M. E.; McKinnon, J. T. Demonstration of
Vd = displacement volume (L) Dissociated Methanol as an Automotive Fuel: System Design. Proc. Int.
w = water-to-methanol molar ratio in aqueous methanol Symp. Alcohol Fuels Technol., 4th 1980, B-11.
W = work done by the engine (J) (10) Finegold, J. G.; Karpuk, M. E.; McKinnon, J. T.; Passamaneck,
xi = mole fraction of gas i in the product gas R. Demonstration of Dissociated Methanol as an Automotive Fuel:
xp = premixed fuel’s mole fraction System Performance. Conference of the American Section of the
yi = mole fraction of gas i International Solar Energy Society, Philadelphia, 1981.
y = mole fraction vector of combustion products (11) Lindström, O. Fuel Treatment for Combustion Engines. U.S.
y0 = mole fraction vector of fresh engine charge Patent 3,918,412, 1975.
(12) Lindström, O. Procedure for the Operation of Combustion
yad = mole fraction vector of adiabatic combustion products
Engine. U.S. Patent 4,244,328, 1981.
ϕ = equivalence ratio (13) Cheng, W. H. Appl. Catal., B 1995, 7, 127−136.
Δgf° = standard molar free enthalpy of formation (J/mol) (14) Christiansen, J. A. J. Am. Chem. Soc. 1921, 43, 1670−1672.
ΔGf°(T) = standard change in free enthalpy for the (15) Batista, M. S.; Santos, R. K. S.; Assaf, E. M.; Assaf, J. M.;
formation reaction of a gas (J/mol) Ticianelli, E. A. J. Power Sources 2004, 134, 27−32.
Δh(T) = change in molar enthalpy (J/mol) (16) Profeti, L. P. R.; Ticiannelli, E. A.; Assaf, E. M. Appl. Catal., A
ΔG°(T) = standard free enthalpy change of a reaction (J/ 2009, 360, 17−25.
mol) (17) Morgenstern, D. A.; Fornango, J. P. Energy Fuels 2005, 19,
Δh°f = standard molar enthalpy of formation (J/mol) 1708−1716.
Δh°f,CHO = standard molar enthalpy of formation of an (18) Sall, E. D.; Morgenstern, D. A.; Fornango, J. P.; Taylor, J. W.;
oxygenated hydrocarbon (J/mol) Chomic, N.; Wheeler, J. Energy Fuels 2013, 27, 5579−5588.
(19) Dong, Y.; Steinberg, M. Int. J. Hydrogen Energy 1997, 22, 971−
ΔHc° = standard combustion enthalpy (kJ/mol) 977.
ΔH°r = standard reaction enthalpy (kJ/mol) (20) Suslu, O. S.; Becerik, I. J. Fuel Cell Sci. Technol. 2013, 10,
Δs(T) = change in molar entropy (J mol−1 K−1) No. 031001.
λ = relative air/fuel ratio (21) Pettersson, L.; Sjöström, K. Combust. Sci. Technol. 1991, 80,
λ = vector of Lagrange undetermined multipliers 265−303.
μ = chemical potential of a gas (J/mol) (22) Anthonissen, E.; Wallace, J. J. Dissociated Methanol Engine
ν = signed stoichiometric coefficient of a gas in a reaction Testing Results Using H2−CO Mixtures. Proc. Intersoc. Energy Convers.
ρ = molar density of the engine charge (mol/L) Eng. Conf., 18th 1983, 839091.
ρ° = standard molar density (0.041 mol/L) (23) Dai, X.; Ji, C.; Wang, S.; Liang, C.; Liu, X.; Ju, B. Int. J. Hydrogen
Energy 2012, 37, 14624−14631.
Abbreviations (24) Ji, C.; Zhang, B.; Wang, S. Int. J. Hydrogen Energy 2013, 38,
CNG = compressed natural gas 7490−7498.
LPG = liquefied petroleum gas (25) Conte, E.; Boulouchos, K. SAE Pap. 2004, 010972.
IMEP = indicated mean effective pressure (26) Li, H.; Karim, G. A. Int. J. Hydrogen Energy 2005, 30, 1491−
1499.
IVC = intake valve closure
(27) Ji, C.; Wang, S. Int. J. Hydrogen Energy 2012, 37, 3938−3944.
PCCI = premixed charge compression ignition (28) Boehmann, A. L.; Le Corre, O. Combust. Sci. Technol. 2008, 180,
SI = spark ignition 1193−1206.
STP = standard temperature (T° = 298.15 K) and pressure (29) Ozawa, M.; Hirota, T. Starter for Alcohol Engine. European
(p° = 1 atm) Patent 0022876-A1, 1981.
syngas = synthesis gas (30) Ishii, H.; Koike, N.; Suzuki, H.; Odaka, M. SAE Pap. 1997,
970315.
Indices
(31) Tsolakis, A.; Megaritis, A. Int. J. Hydrogen Energy 2005, 30, 731−
ad = adiabatic 745.
i = gas component i, column i of the conservation matrix A (32) DeFilippo, A.; Chin, G. T.; Chen, J.-Y. Combust. Sci. Technol.
k = row number of an entry in the conservation matrix A 2013, 185, 1202−1226.

2090 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091

Energy & Fuels Article

(33) De Boer, P. C. T.; McLean, W. J.; Homan, H. S. Int. J. Hydrogen

Energy 1976, 1, 153−172.
(34) Heywood, J. B. Internal Combustion Engine Fundamentals;
McGraw Hill: New York, 1988.
(35) Smith, J. M.; Van Ness, H. C.; Abbott, M. M. Introduction to
Chemical Engineering Thermodynamics, 6th ed. in SI Units; McGraw-
Hill: Singapore, 2001.
(36) Spencer, H. M. Ind. Eng. Chem. 1948, 40, 2152.
(37) Pankratz, L. B. Thermodynamic Properties of Elements and
Oxides. Bull.U.S., Bur. Mines 1982, 672.
(38) Kelley, K. K. Contributions to the Data on Theoretical
Metallurgy: XIII. High Temperature Heat Content, Heat Capacity and
Entropy Data for the Elements and Inorganic Compounds. Bull.
U.S., Bur. Mines 1960, 584.
(39) JANAF Thermochemical Tables; NBS-37; National Bureau of
Standards: Gaithersburg, MD, 1971.
(40) Olikara, C.; Borman, G. L. SAE Pap. 1975, 750468.
(41) Ferguson, C. R.; Kirkpatrick, A. T. Internal Combustion Engines:
Applied Thermosciences, 2nd ed.; John Wiley & Sons: New York, 2001.
(42) Sutter, D. H. J. Chem. Educ. 1996, 73, 718−732.
(43) Shevla, R. A.; McBride, B. J. Fortran IV Computer Program for
Calculation of Thermodynamic and Transport Properties of Complex
Chemical Systems; NASA Technical Note TN D-7056; NASA Lewis
Research Center: Cleveland, OH, 1973.
(44) Ji, C.; Dai, X.; Wang, S.; Liang, C.; Ju, B.; Liu, X. Int. J. Hydrogen
Energy 2013, 38, 11169−11173.
(45) Yamaguchi, I.; Takishita, T.; Sakai, T.; Ayusawa, T.; Kim, Y. K.
SAE Pap. 1985, 852201.

2091 dx.doi.org/10.1021/ef4022009 | Energy Fuels 2014, 28, 2076−2091