Graphical Abstract

Optimum blending hydrogen ratio with liquid fuel in spray combustion

to reduce emissions of nitrogen oxides
Ahmed Abed Al-Kadhem Majhool, Ammar Muslim Hadi, Faris Alqurashi,
Sara Falih Kareem, Noor M. Jasim, Mujtaba A. Flayyih
 Optimum blending hydrogen ratio with liquid fuel in
spray combustion to reduce emissions of nitrogen oxides
Ahmed Abed Al-Kadhem Majhoola , Ammar Muslim Hadib , Faris
Alqurashic , Sara Falih Kareema , Noor M. Jasimb , Mujtaba A. Flayyihd
a
Department of Mechanical Engineering, University of
Al-Qadisiyah, Al-Qadisiyah, 88, Al-Qadissiya, Iraq
b
Department of Mechanical Engineering, University of Kufa, Kufa, 86, Najaf, Iraq
c
Department of Mechanical Engineering, University of
Bisha, Bisha, 67714, Bisha, Kingdom of Saudi Arabia
d
Biomedical Engineering Department, College of Engineering and
Technologies, Al-Mustaqbal University , Hillah, 51001, Babylon, Iraq

Abstract
The goal of the present study is to build a computational combustion model
and numerically investigate the stoichiometric mixture fraction generated
from blending liquid n-pentane fuel with liquid hydrogen applied in the
form of combustible spray droplets. The mixture fraction-probability density
function is used to construct the combustion model. The role of the mixture
fraction that is generated from blending n-pentane and hydrogen at different
mass fraction ratios. From the numerical results, the increase in H2 blending
in the C5 H12 /H2 mixture increases the temperature and velocity of the flame.
Furthermore, the addition of hydrogen expands the reaction zone to locations
downstream of the flame. The enrichment of the reactive mixture with
H2 reduces the emission of CO. Only in the reaction zone does hydrogen
fuel enrichment decrease the mass fractions of CO and OH. The results
confirm that the PDF-RANS model gives satisfactory agreement with the
experimental data.
Keywords: Spray combustion, Liquid hydrogen addition, PDF combustion
model, Stoichiometric blending, Two phase combustion zones

∗
Corresponding author
1
E-mail address: [Link]@[Link]

Preprint submitted to Elsevier August 28, 2024

1. Introduction
The use of hydrogen as a fuel in industrial applications is an attractive
option due to its unique properties, including its wide flammability range.
The fact that hydrogen can burn in a variety of hydrogen-air mixtures makes
it possible to use lean mixtures at lower temperatures, which lowers the
release of pollutants like NOx [1, 2]. The flammability range of hydrogen
is between 4% and 77% by volume, compared to 4% and 16% for natural
gas. Industrial furnaces [3, 4] are a novel application of the industrial use of
hydrogen in combustion processes. The combustion of hydrogen (or the direct
use of hydrogen-based fuels such as ammonia) offers promising alternatives
to reduce emissions. To reduce the impact of the transition to hydrogen
in industry, it may be possible to develop equipment that can burn both
hydrogen and natural gas. The combustion of hydrogen produces a higher
flame temperature and a different radiation. Hydrogen-based combustion
leads to higher flame velocities [5, 6]. The increased water generation necessitates
adjusting the exhaust gas system, which could increase water separation
capacity and potentially cause corrosion when using a mixture with sulphur-
containing fossil fuels. Hydrogen has a high burning rate compared to carbon-
containing fuels and a non-luminous flame, making optical monitoring difficult.
Using mixtures of hydrogen and ammonia can help with some of these problems
because ammonia burns more slowly and has a clear flame, which also helps
lower NOx (nitrogen oxide) emissions [7]. It may be necessary to use different
types of steel, as not all steels are suitable for hydrogen due to potential
decarbonisation of surface layers or scale removal [8]. Emissions from fossil
fuel combustion and atmospheric damage continue to increase; on the other
hand, the gradual depletion of fossil fuels such as coal, oil, and natural gas
has become one of the most worrying and controversial. Carbon-based fuels,
such as carbon dioxide (CO2 ) and carbon monoxide (CO), contribute to a
continuous increase in greenhouse gas emissions. While CO2 has a significant
impact on global warming, its effect is very low; however, CO has important
indirect effects. That is, it reacts with OH radicals in the atmosphere and
decreases their amount. Since OH radicals act as an inhibitor of decreasing
strong greenhouse gases such as methane [9]. Controlling nitrogen oxide
pollutants and increasing production and energy transfer efficiency are two
important factors in industrial combustion system optimisation [10]. Optimising
the efficiency of liquid burners presents a significant challenge, as an increase
in production and energy transfer invariably leads to the creation of additional

3
pollutants [11]. For the simultaneous optimisation of nitrogen oxides and
energy efficiency, it is necessary to determine the level and dispersion of
the nox pollutants in combustion systems. This approach facilitates the
easy interaction of other nitrogen oxides with combustion variables [12]. In
non-premixed turbulent flames, which are very common in liquid burners,
NOx production, consumption, and transport will cover a wide time and
space along the flow structure. Such a study makes the study of NOx
pollution and flow characteristics inseparable [13, 14]. Unfortunately, current
experiments lack the necessary tools to conduct such investigations. Because
the experiment tools still do not have the ability to calculate and measure
the characteristics of the flow in low time intervals, and inevitably these
calculations can be done and checked through mathematics and modeling.
There are numerous models for studying NOx emission in turbulent flames
[15, 16]. Among them, a widely used method involves considering NOx
emission alongside other flow characteristics in CFD [17, 18]. CFD has a
tremendous ability to predict the properties of fluids in combustion or to
model heat transfer equipment such as furnaces or converters. For example,
Bazooyar et al. [19, 20, 21, 22, 23, 24, 25] demonstrated that compact and
catalytic combustion systems can effectively manage NOx emission. In liquid
combustion, a crucial parameter that has not received much attention is
the interaction between the evaporating particles and the chemically active
environment. The choice of combustion mechanism for biodiesel fuel is
important because it has a complex chemistry and dozens of chemical components.
Fischer and Pitts [26] and Bracora and Ritz [27] have looked at the development
of a number of alternative ways to burn biodiesel fuel, but they have not been
tested. As noted by Bulat [28], one of the biggest challenges in designing and
running industrial burners is making them work with a wider range of fuels.
One of the renewable fuels, biodiesel, serves as a replacement for diesel in
burners. Burning biodiesel in combustion systems can lead to a variety of
issues. For example, it can produce carbon dioxide gas too quickly [29],
raise the levels of formaldehydes [30, 31, 32] and increase the amount of
NOx [33, 34]. Several studies can be used to find out how toxic pollutants
and inlet flow characteristics change, as well as the stoichiometric conditions
and concentration of important parts in the gases that biodiesel combustion
releases. This can help make biodiesel combustion more consistent. In
premixed flames, the results were consistent with the experiments. These
results show an increase in NOx production and a decrease in carbon dioxide
production with biodiesel [35]. Most modelling results for this principle

4
are generally consistent. However, some cases may show slight differences.
The properties of biodiesel and diesel injectors differ. Different models and
experimental test results are available for this fuel spray specification. In
some cases, biodiesel cannot have the same combustion quality as conventional
diesel fuel due to its higher viscosity [36, 37]. Their study, which compared
the combustion of biodiesel mixtures with diesel, reported that there were
no differences between these mixtures when the percentage of biodiesel was
less than 25% by volume. [38]. Rapeseed oil methylester was chosen as
fuel in the examination and included the burner’s characteristics and the
semi-industrial boiler’s operating conditions in its modeling. The numerical
simulation was used to simulate the NOx formation, and the results of the
simulation were compared with the experimental data. The investigation
focused on the diffusion and spread of nitrogen oxides in biodiesel’s previously
non-premixed turbulent flame. Nitrogen oxides (NOx ) typically consist of
a variety of components. In the combustion environment, according to
the conditions, the main components of NOx are nitrogen monoxide and
nitrogen dioxide. Other nitrogen oxides, such as nitrous oxide, although
they have a very long life, will have a tiny and unmeasurable amount [39].
Reducing the environmental impact of power generation fuels and utilising
renewable energy sources is becoming increasingly important. Renewable
energy sources reduce both the greenhouse gas effect and the dependency
on fossil fuels [40]. Internal combustion engines have an important place in
power and emission generation. Spark ignition engines have low thermal
efficiency and soot emissions. However, they produce high HC and CO
emissions. Compression ignition engines, on the other hand, have high
thermal efficiency but produce high amounts of soot and NOx emissions. As
a result of today’s emission standards, engine manufacturers are focussing
their attention on engines that produce fewer emissions and have improved
fuel consumption [41]. One alternative energy source for power generation is
biogas. For many years, studies on the use of biogas in internal combustion
engines have been ongoing. Organic matter typically decays to produce
biogas, a renewable energy source. The main component of biogas is CH4
gas. Furthermore, biogas contains non-flammable gases such as CO2 and
H2S. In order for biogas to be used more efficiently in internal combustion
engines, it can be purified by removing its non-flammable gases [42]. The
production of biogas from the same organic substances, but in different tanks,
leads to low pressure, specific gravity, and large specific volume problems.
High pressures compress biogas to store it in smaller volumes, [43]. Biogas

5
consists of 50–70% CH4 gas. Methane gas is a simple hydrocarbon that
is suitable for internal combustion engines because of its high stability. In
addition, CH4 gas has a much more negative effect on global warming than
CO2 gas. Therefore, using biogas in energy production can reduce the
release of CH4 gas into the atmosphere, thereby positively impacting global
warming [40]. As an engine fuel, biogas offers advantages such as complete
combustion, no lubricating oil dilution, improved exhaust emissions, and
knock resistance. Because of its high knock resistance, the engine can operate
at higher compression ratios [43]. Biogas, which is considered a fuel for
internal combustion engines, has a low energy density due to the CO2 it
contains. Flame spread speed and flammability limits are low. Biogas’ high
self-ignition temperature reduces the possibility of knocking when used as
fuel in spark ignition engines. As a result, it is difficult to use biogas as
the sole fuel in compression ignition engines [44]. Spark ignition engines can
use biogas without significant modifications [45, 46]. Kim et al. [45] found
that NOx emissions went down in all of their tests with a biogas-fuelled
spark ignition engine. They did this by changing the engine’s torque, the
amounts of CH4 and CO2, the amounts of fuel and air, intake air, and fuel
flow rate. They attributed this situation to an increase in the amount of
CO2 in the cylinder at stoichiometric fuel/air ratios. Applications such as
homogeneous charge compression ignition engines [40, 41] and spark plug
compression ignition engines [47, 48] are also included. One of the methods
commonly used in biogas-fuelled compression ignition engines is to ignite
biogas with a small amount of liquid fuel. This method is known as the
dual-fuel mode. In dual-fuel mode, both diesel fuel and biodiesel can serve
as pilot fuel. In this method, biogas mixed with air in the intake channel
is sucked into the cylinders, compressed, and ignited with the pilot fuel.
Similar to the conventional compression ignition engine, the dual-fuel mode
sprays the pilot fuel into the cylinders at the end of compression. Depending
on the operating conditions, the amount of pilot fuel can be adjusted [42].
Jiang et al. [43] found no corrosion or shortening of engine life in biogas-
powered dual-fuel compression ignition engines. They showed that dual-
fuel engines can operate with alternative gaseous fuels, such as natural gas,
liquefied petroleum gas, and hydrogen, in addition to biogas. Engines must
undergo minor modifications to operate in dual fuel mode. However, these
modifications do not affect the operation of the engine with a single fuel. If
desired, a dual-fuel compression ignition engine can be operated with only one
fuel [49]. In a study on a compression ignition engine that ran on natural gas

6
and biogas, Mustafi et al. [50] found that NOx emissions went down compared
to running on a single fuel, while HC emissions went up. CO2 significantly
influenced both the increase and decrease in emissions. Particulate emissions
were lower in dual-fuel operations. Yamasaki et al.[51] reported that NOx
emissions are highly dependent on adiabatic flame temperatures. When
adiabatic flame temperatures decrease, NOx emissions also decrease. However,
when adiabatic flame temperatures increase, NOx emissions increase linearly,
while dioxide levels increase exponentially when temperatures exceed 1920
K. Verma et al. [52] showed that using less pilot diesel fuel makes the
spray area smaller, which makes the combustion more evenly spread out
in the combustion chamber. This lowers the temperature of the combustion,
which in turn lowers NOx emissions. The CO2 and other inert gases in its
content reduce the calorific value and flame spread speed of biogas [53]. This
situation negatively affects the performance and exhaust emissions of internal
combustion engines. Dual-fuel engines have low fuel conversion efficiency,
especially at low engine torques, and produce high exhaust emissions. The
CO2 in biogas reduces NOx emissions by reducing in-cylinder temperatures,
as demonstrated by the EGR effect. HC emissions increase due to a similar
effect[42, 45]. In compression ignition engines, soot consists of carbonaceous
compounds formed as a result of combustion. Rich fuel/air mixtures form in
the central region of the diesel fuel spray, where soot occurs. These clustered
carbon compounds cannot burn because they do not combine with enough
oxygen. The CO2 in the biogas forms free radicals (O and OH), which
enhance particle oxidation and decrease soot emission. However, the rich
mixture formed in the cylinder increases the soot emission [52]. It is thought
that the more carbon elements a fuel contains, the more soot emissions
it will produce during combustion. In addition, chemical structures such
as aromatics and cyclic molecules have a higher tendency to form soot.
Gaseous fuels, such as biogas and natural gas, do not contain these structures.
Liquid fuels are the primary source of soot emissions in dual-fuel compression
ignition engines [50]. Many studies indicate that compression ignition engines
can utilise biogas as an alternative gas fuel. Examining the literature reveals
that researchers have conducted studies using various fuels such as hydrogen,
natural gas, and biodiesel, varying the compression ratios, pilot fuel amounts,
and injection timings. However, it has not found any study that alters the
amount of diesel fuel by reducing the compression ratio, particularly to reduce
NOx emissions. They investigated how the amount of diesel fuel affected
NOx , HC, CO2 , and soot emissions in a biogas-diesel dual fuel compression

7
ignition engine with a reduced compression ratio. No changes were made to
the engine’s electronic control unit, and the original settings were preserved
[54]. In this work, we investigated the formation of NOx during spray
combustion when we blended hydrogen with n-pentane. Throughout the
results, we discovered an optimum ratio for hydrogen blending that yields a
lower emission level, potentially supporting the concept of hydrogen-enriched
combustion.

2. Models and chemistry mechanisms

2.1. Droplets phase
The prediction of the motion of the discrete-phase droplets is done by
integrating the equilibrium force on the droplet, which is calculated in a
Lagrangian reference frame. The following formula for this equilibrium force
equates the droplets at rest with the forces acting on them [55]:

dmd
= −ṁd (1)
dt
dud Fd + Fp
= (2)
dt md
dTd Q̇c + Q̇lat
cp,d = (3)
dt md
where Q̇lat is the latent heat of evaporation and md = πρd d3 /6 is the mass of
the droplet. Considering Abramzon & Sirignano’s proposed model [56], the
instantaneous evaporation rate of the drop is given by the Stefan-Maxwell
equation, in which both diffusion and advection phenomena are computed:
f ln (1 + BM )
ṁd = πdρf Df Sh (4)

Q̇c = hc Ad (Tg − Td ) (5)

Where ρf is the density calculated from ρf = pS MWm /RTS , hc is the
convection heat transfer coefficient, and Ad is the droplet surface area. Df
refers to the vapour diffusivity coefficient, which is calculated from:
α
Df = 3.6059 × 10−3 · (1.8Ts )1.75 · (6)
ps β

8
where α and β are the constants. The vapour pressure at the surface is
determined from:
 
c2 c5
pS = p · exp c1 + + c3 ln TS + c4 TS (7)
TS
The one-third law is used to calculate the temperature at the droplet’s surface
temperature as:
TS = (T + 2Td ) /3 (8)
The Ranz and Marshall correlation [57] is frequently used to evaluate the
heat transfer coefficient as:
h i
g
Nu = 2 + (1 + Red Pr)1/3 max (1, Red )0.077 − 1 /F (BT ) . Pr (9)

For modelling involving liquid spray systems in a heated environment, the

spray is considered the discrete phase of the domain formed by a large number
of droplets that are transported based on Newton’s second law. In this case,
the Lagrangian formulation for this phase gives us the equations for the
drop’s motion and linear momentum. These equations take into account
how important the internal flow is in viscous spheres for figuring out the
fluid dynamic force that is at play as a function of the drag coefficient Cd
[58]. In this work, such a correlation for the Reynolds number is defined as
follows:
18µ Cd Red
Fd = md (u − ud ) (10)
ρd d2 24
where the droplet Reynolds number is :
ρd |ud − u|
Red ≡ (11)
µ
and the pressure gradient is expressed as:
Fp = −Vd ∇p (12)
2.2. Gas phase
The conservation of mass, momentum, energy, and flow species principle
considers a fixed reference volume in the space crossed by a fluid current.
The Navier-Stokes equations for the gas phase are expressed by using an
Eulerian approach as follows [59]:
∂ρ
+ ∇ · [ρu] = Sm (13)
∂t
9
 ∂(ρu)
+ ∇ · [u(ρu)] + ∇p + ∇ · T = SF (14)
∂t
∂(ρE)
+ ∇ · [u(ρE)] + ∇ · [up] + ∇ · [T · u] + ∇ · q = ω̇T + Se (15)
∂t
∂ (ρYm )
+ ∇ · [u (ρYm )] + ∇ · sm = ω̇m + SYm (m = 1, . . . M − 1) (16)
∂t
p = ρRT (17)
The Particle-Source-In-Cell (PSI-Cell) method uses a series of source terms
to infer the influence of the surrounding gas’s carrier phase on the sprayed
liquid droplets. These source terms are defined in the following way:
Nd
1 X
Sm = ṁd (18)
Vc 1
Nd
1 X
SF = − (−ṁd ud + Fd ) (19)
Vc 1
Nd h i
1 X
Se = − −ṁd h (Td ) + Q̇c (20)
Vc 1
SY m = Sm (21)
The turbulence model k - ǫ reported by [57] provides for the closure of the
problem by means of two equations, one relating to the turbulent kinetic
energy, k, while the other to the dissipation rate, ǫ. Using these two quantities,
a spatial scale l = k 3/2 /ǫ and a time scale t = k/ǫ can be obtained. Alternatively,
the following relation provides the kinematic turbulent viscosity:
k2
µt = ρCµ fµ (22)
ǫ
The two transport equations are given below:
   "  1/2 2 #
D(ρk) ∂ µt ∂k ∂k
= µ+ + ρPk − ρ ǫ + 2ν (23)
Dt ∂xi σk ∂xi ∂xi

where Pk is the mean shear production term, defined as:

 
∂Ui
Pk = νt (24)
∂xi

10
   
D(ρǫ) ∂ µt ∂ǫ ǫ ǫ2
= µ+ + Cǫ1 ρ Pk − Cǫ2 ρ (25)
Dt ∂xi σǫ ∂xi k k
where Cµ is a constant. The model constants (C1ǫ , C2ǫ , Cµ , σk , σǫ ) have the
following values.

C1ǫ = 1.44, C2ǫ = 1.92, Cµ = 0.09, σk = 1.0, σǫ = 1.3 (26)

Spray diffusion flames can be modelled as a two-stream mixing problem in

which a fuel mass flow meets an oxygen mass flow. Two parameters are
required to describe the one-step reaction. The mixture fraction Z and its
[Link] transport equation can be derived for the mixture fraction [55].
The sum of all conservation equations of the components is formed. The
chemical source terms then cancel out to zero. Therefore, the resulting
equation for the scalar quantity Z contains no production term and is of
the convection-diffusion type. Unlike the chemical components, the ongoing
reaction does not impact the mixture fraction, allowing only the flow to
transport it. After the averaging described above, the following averaged
transport equation for the mixture fraction results:

∂ρZ̄  
+ ∇ · [ρZ̄v] = ∇ · ρD∇Z̄ − ρv ′′ Z ′′ (27)
∂t
The turbulent exchange term is approximated according to the gradient flow
approach and using the constant assumed turbulent Schmidt number Sct :
µt
−ρv ′′ Z ′′ = ∇Z̄ (28)
Sct
As a result, the modelled transport equation for the mean mixture fraction
is:   
∂ρZ̄ µt
+ ∇ · [ρZ̄v] = ∇ · ρD + ∇Z̄ (29)
∂t Sct
The variance Z ′′2 of the mixture fraction enables statistical statements about
the distribution of Z and can be calculated using a transport equation [59]:

∂ρZ ′′2 h i h i
+ ∇ · ρZ v = ∇ · ρD∇Z − ρv Z
′′2 ′′2 ′′ ′′2 (30)
∂t
Assuming a Lewis number equal to one, the elemental composition and
enthalpy of the gas mixture are a pure function of the mixture fraction

11
Z. If the mixture fraction Z is known, other quantities, such as the mass
fractions Ys or the temperature T , can be determined as a function of Z.
In turbulent combustion, these are often strongly non-linear functions of
the mixture fraction. Average values of the quantities dependent on Z̄ can
be calculated by integrating the probability density function of the mixture
fraction P](Z̄): Z 1
f˜ = f (Z̄)P]
(Z̄)dZ̄ (31)
0
To do this, the P (Z) must be known. However, the process of calculating
the P DF takes a significant amount of time, leading to the assumption of
the function course, also known as the presumed P DF method: The most
commonly used form for the P DF is the approximation with the β function:
  Z̄ α−1 (1 − Z̄)β−1
Pe Z̄, Z ′′2 = R 1 for 0 ≤ Z̄ ≤ 1 (32)
Z̄ α−1 (1 − Z̄)β−1 dZ̄
0

with the parameters:

   
Z̄(1 − Z̄) Z̄(1 − Z̄)
α = Z̄ − 1 , β = (1 − Z̄) −1 (33)
Z ′′2 Z ′′2
which depend on the mean mixture fraction Z̄ and its variance Z ′′2 . The β
function provides a good approximation of experimental curves and is also
mathematically easy to evaluate. The calculation of the spatial distribution
of nitrogen oxide concentration in the spray combustion is done by solving
the transport equation for the mass fraction of NO, which may be expressed
as [59]:  
∂ (ρYN O ) ∂ (ρuj YN O ) ∂ µt ∂YN O
+ = + SN O (34)
∂t ∂xj ∂xj Sct ∂xj
where the source term SN O denotes the rate of formation of NO. In the
combustion process, two different chemical processes lead to the formation
of NO. The Zeldovich mechanism is the main source of nitrogen oxide
during combustion at temperatures above 1800 K. At high temperatures,
a combination of O and N atoms forms NO. The following list outlines the
main reactions that form ”thermal” NO from molecular nitrogen:
O + N2 ↔ N + NO (35)

N + O2 ↔ O + NO (36)

12
The reaction makes a certain contribution when the conditions are near
stoichiometric and in rich fuel-air mixtures:

N + OH ↔ H + NO (37)

The following formulas determine the rates for these reactions based on
experimental data [59]:

kf,1 = 1.8 × 108 e−38370/T (38)

kr,1 = 3.8 × 107 e−425/T (39)
4 −4680/T
kf,2 = 1.8 × 10 Te (40)
3 −20820/T
kr,2 = 3.8 × 10 Te (41)
kf,3 = 7.1 × 107 e−450/T (42)
kr,3 = 1.7 × 108 e−24560/T (43)

Here kf,1 , kf,2 , kf,3 are the rate constants of the forward reactions and kr,1 , kr,2, kr,3
are the corresponding rates of the reverse reactions. The formula that determines
the total rate of formation of NO is:

d[NO]
= kf,1 [O][N2 ]+kf,2 [N][O2 ]+kf,3 [N][OH]−kr,1 [NO][N]−kr,2 [NO][O]−kr,3 [NO][H]
dt
(44)
The calculation of the level of nitrogen oxide emission in the combustion
process is usually made using some assumptions. One of such assumptions is
the assumption of a quasi-equilibrium state for calculating the concentration
of N. The rate of the reaction of nitrogen oxide formation is significant
only at high temperatures since the formation of N requires the rupture
of the strong triple bond of the N2 molecule (dissociation energy of 941
kJ/gmol). This physical phenomenon is represented by the large activation
energy in reaction 1 [eq. 35], which limits the rate of NO formation within the
framework of the Zeldovich mechanism. On the other hand, the activation
energy for the oxidation of N as a result of reaction 2 [eq.36]. In a lean
mixture, for example, when there is enough oxygen, the rate of consumption
of free nitrogen atoms becomes equal to the rate of its formation; therefore,
a quasi-equilibrium state takes place. Thus, the expression for the rate of

13
formation of NO takes the following form:
 
kr,1 kr,2 [N O]2
d[NO] 1− kf,1 [N2 ]kf,2 [O2 ]
= 2kf,1 [O][N2 ]   (45)
dt 1+
kr,1 [N O]
kf,2 [O2 ]kf,3 [OH]

Based on the assumption of equilibrium in the O dissociation-recombination

reaction, we can use one of three methods to calculate the concentration of
atomic oxygen O. The formation rate of NO is significantly slower than the
oxidation rate of hydrocarbon compounds. This implies that the completion
of hydrocarbon compound oxidation forms the majority of the thermal NO.
Consequently, the process of formation of thermal NO can be separated from
the mechanism of the main reactions of the combustion process. The rate of
formation of NO can be determined based on the assumption of equilibrium
in chemical reactions. This assumption significantly simplifies the calculation
of the rate of formation of thermal NO. The following expression provides
the value of the free oxygen atom’s equilibrium concentration in [59]:

[O] = 3.97 × 105T −12 [O2 ]12 e−31090/T (46)

3. Test case and validation Geometry and CFD domain

Fuel n-Pentane
Injection pressure (MPa) 2.67
Injection duration (ms) Steady
Injection quantity (ml) 4
Ambient gas Air
Sauter mean diameter 28µm
Maximum Drop Diameter 79µm
Injector diameter 0.20µm
Spray injection velocity(m/s) 69
Ambient temperature (K) 1255
Ambient pressure (MPa) 3.2,6.3.,9.4
Impingement distance (mm) 50.7

Table 1: Experimental conditions taken from Mao et al. [61]

14
 Table 2: Grid specification
No. cell No. faces No. nodes No. injection cells
case A 5200 10626 5427 5
case B 20800 42052 21253 5
case C 83200 167304 84105 5

Figure 1: Computational model and boundary conditions.

Figure 2: Computational model and boundary conditions.

15
 PDF n(D) D D2
 m−1   m 


m dˆ dˆ 1 2
Rosin-Rammler n n exp − n nΓ 1 + m n2 Γ 1 + m
 ˆ 2
  

exp − (ln(d)−m)
1 2
Log-Normal √
ˆ 2π 2n2 exp m + n2 exp(n2 )exp 2m + n2
dn

Table 3: Probability density fuctions of the droplet size distribution.

Model Constant Standard range Present work

Kelvin-Helmholtz B0 0.61 0.61
B1 1.73-60.0 10.0

Table 4: Breakup model constants.

3.1. grid
Gambit is an integrated preprocessor for CFD analysis. The latter constructs
a geometric model and produces a grid. To facilitate CFD calculations,
the grid discretizes the studied space. The system consists of a set of
grids designed to solve the problem’s equations. This is done to improve
the accuracy of the results produced by the solver. Ansys Fluent requires
a sufficiently refined grid so that the solution is independent of the grid.
The software ”GAMBIT 2.4.6” creates the structured grid. This best offers
extended solutions for the most simple geometries; however, for our case,
we chose the structured grid, which introduces an excess of the number of
cells compared to the other types, like triangle or quadratic cells. Therefore,
additional resources and computation time are required. The study utilizes

Mean air velocity Ug 0.1 − 11.60 m/s

Air mass flow rate ṁg 0.08 − 0.225 kg/s
Air temperature T 1255 K
Air density ρg 1.2 − 5.5 kg/m3
Test pressure pg 3.0 − 9.0 MPa
Air dynamic viscosity µg 1.846 · 10−5 kg/m.s
n-Pentane density ρl 626 kg/m3
n-Pentane dynamic viscosity µl 2.224 g/m.s
n-Pentane surface tension σ 0.01789 kg/s2
n-Pentane volume flow rate Λf 0 − 24.6 mm2 /s

Table 5: Operating conditions and liquid properties

16
Domain Boundary conditions
Fuel+Hydrogen jet Velocity inlet dinjector =0.2 µ m
Vinlet =69 m/s
Xf =1, 0.9, 0.8 and 0.7
Xh2 =0, 0.1, 0.2 and 0.3
Tf =300 K
Th2 =300 K
Oxidant jet Velocity inlet dinlet =0.4 m
Vinlet =11.6 m/s
XO2 =0.21
XN2 =0.79
It =10 %
Tair =1255 K
Confinement wall wall
Symmetry axis Axis
Exhaust gas Pressure-Outlet

Table 6: Boundary conditions

Figure 3: Computational model and boundary conditions.

17
 Table 7: Grid specification
No. cell No. faces No. nodes No. injection cells
case A 5200 10626 5427 5
case B 20800 42052 21253 5
case C 83200 167304 84105 5

an axisymmetric domain with two fluid inlets, a symmetry axis, and walls,
which are considered adiabatic in the absence of a cooling system and non-
adiabatic in its presence. Table 1 shows the grid specification.
The equations described above are nonlinear and coupled partial differential
equations. Numerical methods solve these equations because of their complexity.
The finite volume approach enables domain discretisation and equation resolution
with a large number of unknowns. This method’s principle is to divide the
computational domain into a number of control volumes. Each of these
features a ”main node. The finite volume method integrates, on elementary
volumes of simple shape, the equations written in the form of conservation
laws. It thus naturally provides conservative discrete approximations and
is therefore particularly well suited to the equations of fluid mechanics.
It is simple to implement if the elementary volumes are rectangles. This
work treats each n-pentane droplet as a disperse phase. This work bases
the Discrete Phase Model (DPM) on the Euler-Lagrange approach, which
is the most widely used method for modelling dispersed phases. It solves
the conservation equations for the continuous phase (gas phase). On the
other hand, the forces acting through the carrier medium (gas phase) track
the movements of droplets and particles, viewing them as a set of dispersed
individual particles. The dispersed particles can exchange mass, momentum,
or energy with one another. This study exclusively uses the solid cone injector
model. Using the Rosin-Rammler distribution,
 
D
n(D) = 1 − exp − (47)
Dref

The simulation determines the droplets distribution during liquid disintegration

in the spray process. The entire distribution is assigned to the corresponding
number of discrete intervals. Each interval represents the average diameter of
the particles. For a given diameter, the Rosin-Rammler distribution defines

18
the mass fraction of the droplets as follows:
a1 r −a2 r
n(D) = exp(− ) (48)
r32 r32
The mass flow rate determines the number of droplets injected per unit
time. As a result, the Kelvin-Helmholtz model for primary and secondary
droplet breakup is considered a breakup model, with constants given in
Table 2. Certain laws define the dispersion, evaporation, and combustion
of the droplets and describe the exchange terms. The laws described create
a coupling between the gas phase and the disperse phase. The software
provides a variety of models for this coupling, each with varying considerations
for the interaction between the gas phase and particles, as well as between
the droplets themselves. In this work, we used a so-called two-way coupling,
where droplets do not influence each other, but they can exchange mass,
momentum, and heat with the gas phase. This work used the Upwind method
in a pressure-based equation solver to discretize the convective transport
phenomena. A correction method is used in which the pressure corrections
are calculated using the SIMPLE algorithm (Semi-Implicit Method for Pressure-
Linked Equations). Ansys-FLUENT employs PrePDF as a numerical tool
to produce the probability density function for the chemical reaction tables,
or libraries, utilized by the CFD code’s solver. From a detailed elementary
reaction mechanism, this software solves the numerical integration for the
mean species mass fraction and temperature, which are those of the mass
fractions of all species and that of the temperature as a function of the
mixture fraction Z. It obtains the average mixture fraction Z and its variance
Z from the CFD solver, which allows us to calculate the PDF and do the
integration to find the fractions in the turbulent field.

4. Solution Method
The findings discuss how mixing hydrogen and n-pentane in liquid fuel
affects spray combustion. These analyses are carried out on the liquid spray
droplets injected into the hot channel with the addition of hydrogen at
different blending ratios. The composition of the combustion products is
analysed through the simulation of the concentrations of NOx, O2, OH, and
CO2 along the axis of symmetry and at the reaction zone, where the reaction
zone evolves at the burner outlet. According to the findings, hydrogen
addition plays a very positive role in stabilizing the spray flame without

19
Figure 4: Computational model and boundary conditions.

20
 Fuel n-Pentane
Injection pressure (MPa) 2.67
Injection duration (ms) Steady
Injection quantity (ml) 4
Ambient gas Air
Sauter mean diameter 28µm
Maximum Drop Diameter 79µm
Injector diameter 0.20µm
Spray injection velocity(m/s) 69
Ambient temperature (K) 1255
Ambient pressure (MPa) 3.2,6.3.,9.4
Impingement distance (mm) 50.7

Table 8: Experimental conditions taken from Mao et al. [61]

significantly increasing pollutants in the reaction zone. Indeed, with hydrogen

blending, the intensity of the reaction zones increases with the addition of
hydrogen up to 30% in the fuel, as well as a reduction in NOx emissions. The
reaction zones then move slightly downstream, and NOx emissions increase
again as the percentage of hydrogen in the fuel increases to 30%.

5. Results and discussion

In the context of a comprehensive investigation, the effects of hydrogen
ratios in the chemical reaction on the resulting characteristic turbulent spray
flame, along with their influence on nitrogen oxide formation and its spatial
distribution, are of significant interest. The main influences on nitrogen
oxide formation are the temperature T in the reaction or oxidation zone,
the oxygen content present via the hydrogen combustion blending ratio, and
the chemical reaction (formation mechanism), as detailed in the basics (”1.4
Nitrogen oxide Formation”). It is of interest to see how this behaves under
variable turbulent conditions and to what extent the hydrogen blending ratio
itself has a direct effect on nitrogen oxide formation. This applies both to
the local NO distribution in the reaction zone downstream, far above the
spray flame, and globally at the spray flame structure. The determination of
the distribution of all individually recorded NOx components is carried out
directly in the reaction zone. The local influence of different hydrogen ratios
with the same thermal output of the test burner is particularly interesting.

21
5.1. Temperature
To examine the temperature changes in the combustion process, the
combustion of pure n-pentane fuel is compared with the combustion of n-
pentane fuel enriched with hydrogen at three different blending ratios. These
three cases involve the burning of n-pentane enriched with hydrogen at 10%,
20%, and 30% by mass fraction. Figure 3a shows the spray flame temperature
along the axis of symmetry, with and without the Nox effect. The Nox
model is added to the simulation in order to study the formation of Nox
under the combination of a spray flame and hydrogen addition. It shows
the axial temperature changes in the combustion process. The addition of
hydrogen to n-pentane increases and stretches the spray flame temperature
values around the centre of the axis of symmetry. This is because hydrogen’s
density is low and its lower calorific value is high, which increases the gas flow
rate values in the spray flame. The burner’s axially positioned fuel nozzles
prevent the hydrogen-enriched spray flame from concentrating on the central
axis by increasing the flow rate. Using pure n-pentane as fuel expands the
spray flame into a wide region around the central axis of the combustion
process. This makes the pure n-pentane spray flame shorter. Because of
this, the temperature of the n-pentane-hydrogen is lower in all axial positions
when it is close to the injector than when hydrogen is mixed with other
gases. Furthermore, the calculated temperature values exhibit symmetry in
relation to the central axis. This is because the simulation of the burner used
has a symmetrical structure. According to the findings, hydrogen-enriched
n-pentane flames have a better interaction with combustion products in a
back-pressure combustion channel. The reason for this is that in a back-
pressure combustion channel, the reaction products do not return from the
opposite wall of the combustion zone and surround the combustion flame,
and hydrogen-pentane flames are less wide than pure n-pentane flames. It
is clear that this situation will impact the formation rates and densities of
emissions and intermediate reaction products.
Figure 3b shows the calculated temperature contour plots with different
hydrogen blending ratios. The plots reveal that the numerical calculation
yielded maximum temperature values within the range of x = 0.1 m to 0.4
m. In experimental measurements with 0% H2, the maximum temperature
values are at x = 0.1 m, as shown in Figure 1a. The results indicate that
the temperature falls within the range. In the areas close to the injector, the
temperature values are low because the spray flame is just beginning to form,
which explains the lower rate of NOx formation later. It has been observed

22
 Figure 5: Computational model and boundary conditions.

that the spray flame temperature increases in regions where the air/fuel ratio
approaches the stoichiometric mixture ratio. It is observed that the effective
spray flame length and flame increased with hydrogen blending ratios, but
the spray flame width in the numerical simulation decreased compared to the
0% H2 case. The idea behind exploring the temperature contour plots is to
explain the reaction zones in order to correctly understand NOx generation.
It should be noted that the contour plot of 20% H2 is different from the
ordinary sequence of the development of the spray flame structure, which
leads us to believe that there is an optimum blending ratio for hydrogen
when it is added to the liquid fuel, and this could be considered one of the
constraints that have been used for the blending process. Figure 4 shows
that the enthalpy, axial gas velcocity, and mixture fraction vary along the
axial distance. These figures are used to support the result of the 20% H2
blending ratio, which explains stoichastic behaviour instead of deterministic
one. Figure 4a shows the energy profile for the four cases with the minus sign,
which refers to the source term in the energy transport equation. The 20%H2
blending ratio falls below the 10% and 30%H2 cases, approaching 0%H2, and
Figure 4b illustrates a similar pattern with the axial gas velocity. In Figure
4c, the mixture fraction shows that the 20% H2 case initially sits between
the other two cases at x = 0.05 m, before beginning to decrease towards the
10% H2. That is because the mixture fraction transport equation depends on
the mass transfer’s evaporation rate, which is associated with the chemical

23
reaction in the energy source term.
In addition to the aforementioned investigation, pollutant emissions have
a significant impact on the choice of hydrogen blending ratio for liquid fuel
combustion. Figure 4g displays the carbon monoxide emissions from the
four combustion cases with varying hydrogen blending ratios. All four cases
showed an increase in CO emissions as the hydrogen blending ratio increased,
with the exception of the 20%H2 case, which showed a lower CO emission
than the 10%H2 and 30%H2 cases as the hydrogen ratio increased. This
reduction in CO emissions leads to lower concentrations and mass flows, as
well as a shorter chemical reaction time in the combustion process.

5.2. NO Mass fraction

Figure 1a shows the mass fraction of NO concentrations calculated along
the axis of symmetry. The calculations are based on a constant combustion
air ratio, varying hydrogen blending ratios, and an average flow velocity of
u 0 = 11.6 m/s. The contour plots below illustrate further variations in the
flame structure. The plots use a typical part of the reaction zone to show
how fast chemical reactions happen and how they lead to the formation of
thermal NO. The combustable mixture with the highest hydrogen blending
ratio (30% H2) has lower NO in the reaction zone due to zone shifting, which
increased with axial distance. When the flame is on the verge of extinction at
an axial distance of less than 0.5 m, the concentration of nitrogen monoxide
pollutant will exhibit complete self-similarity.
Figure 1b displays the largest flames that happen when 30% hydrogen
is added, the temperature of the burning process is higher, and the flow
speed of the mixture is increased. These flames also have the highest NO
concentrations in the uppermost calculated positions. Observing that the
20%H2 blending ratio expresses less NO2 than other blending ratios, we
should consider the NOx formation as one of the optimal amounts for hydrogen
addition to any fuel. From this perspective, the nitrogen monoxide pollutant’s
general axial shape, measured at axial distances ranging from 0 to 0.2 m in
the 30%H2 case, exhibits lower values compared to other cases. However,
these values begin to increase as the spray flames move away from the fuel
nozzle.
Areas with low local temperatures will kinetically limit the production of
nitrogen dioxide. Most nitrogen dioxide is a product of the chemical reaction
between nitrogen monoxide and n-pentane ions. This occurs when the spray
flame has disappeared or is far from the reaction zone. The contour plots in

24
Figure 6: Computational model and boundary conditions.

25
 Figure 7: Computational model and boundary conditions.

Figure 3c show the NO mass fraction with different hydrogen blending ratios.
The plots align these ideas with the production of nitrogen dioxide around
the spray flame. The contour plots show the accuracy of the computational
fluid dynamics model in terms of total nitrogen monoxide and local nitrogen
oxide.
H + NO2 → NO + OH (49)

5.3. N2O mass fraction

Figure 5 shows the distribution of nitrous oxide (N2O) pollutant over the
axial distance, confirming the simulation results. Figure 5a reveals that the
spray flame’s nitrous oxide concentration primarily stems from low mixture
ratios, which occur when the hydrocarbons undergo complete destruction and
consumption. Figure 5b shows that N2O production is limited to areas after
the spray flame where the concentration of nitrogen is still high and can react
with oxygen-free radicals. These results are consistent and comparable to
those of the other Nox components. Figure 3c displays the pollutant contour
plots of nitrous oxide at different hydrogen blending ratios. The fuel nozzle
will produce the majority of the nitrous oxide pollutant between 0.2 and 0.8

26
 Figure 8: Computational model and boundary conditions.

meters away. Normally, N2O will form at a much higher space interval than
the other combustion components above, and the conditions for dissociation
of molecular nitrogen are available. In most of these areas, chemical reactions
are in their equilibrium state. The accuracy of N2O simulation is better in the
regions of the combustion channel where the spray flame is present. However,
this model’s accuracy in simulating N2O is high. The presence of induced
nitrous oxide in the combustion mechanism renders the model helpless in
expressing reactions with high chemical speed.

NO + H2 O → N2 O + OH (50)

5.4. NO2 mass fraction

Figure 2 shows the mass fraction of the nitrogen dioxide NO2 along
the axial distance. However, the oxygen concentration in the combustion
chamber also plays an important role in NO2 formation. The maximum
NO2 concentration is recorded at 30%H2 blending ratio, and the gas velocity
is the same for all the cases. Under these conditions, Figure 2b demonstrates
the formation of a significant amount of NO2 in the reaction zone, which,
when combined with excess hydrogen, stimulates the formation of NO2. At

27
 Figure 9: Computational model and boundary conditions.

a higher hydrogen blending ratio, the NO2 decreases, and the concentration
also decreases significantly. Especially at 30%H2, the NO2 emissions drop
to a minimum. However, it is generally evident that, in this case, NO2 only
accounts for a very small proportion of the total NOx emissions. In most
cases, this is less than 4%. Again, the most important finding is that the
20%H2 blending ratio shows less NO2 than the other blending ratios. This
means that the spray flame’s stochastic behaviour is determined by how fast
the droplets evaporate and how concentrated they are.

5.5. Rate NO
Figure 3 shows the rate at which thermal NO is produced. The spray
combustion relies on the production rate of NO and the primary radicals.
It is evident that the production rate drops sharply when the stoichiometry
changes from 1 (combustion stoichiometry) to 0.7 (30%H2) in the reaction
zone. At the same time, the flow of radicals responsible for the formation
of NO atoms decreases, and the flow of radicals leading to the reduction of
NO atoms increases. Several physical factors cause this behavior. Firstly, as
the concentration of hydrogen in the mixture increases, the products of its
dissociation and reactions begin to interact intensively with the precursors

28
 Figure 10: Computational model and boundary conditions.

of NO, N, and O atoms. This results in a decrease in the total rate of

NO production in the central reaction zone. Secondly, adding hydrogen
at a stoichiometry higher than normal combustion results in an increase in
temperature, which is caused by an increase in the mixture’s enthalpy and
additional losses from hydrogen dissociation in the flow field. As is known,
the Zeldovich mechanism is very sensitive to temperature changes, and even
a small change in it leads to an increase in NO generation. Thirdly, the
residence time of NO in the reaction zone is comparable to the reaction
time of its reduction. Thus, the reducing atmosphere leads to additional
suppression of NO production.

5.6. NO density
Figure 10 shows the computation of the NO density to clarify the dependence
of NOX generation in the spray flame zone on the hydrogen blending ratio.
The NO density is defined as the mass per unit volume of NO, N2O, and
NH 3. Figure 10a shows that the 20% H2 has the highest values along the
axial direction due to the lower mass and volume of NO and N2O. In Figure
10b, all the hydrogen cases are smaller than the standard case due to two
factors: firstly, the hydrogen cases have a wider volume and a smaller liquid

29
 Figure 11: Computational model and boundary conditions.

mass compared to the zero blending case.

6. Conclusions

30
 Figure 12: Computational model and boundary conditions.

References
[1] López Ruiz G, Alava I, Blanco JM. Study on the feasibility of the
micromix combustion principle in low NOx H2 burners for domestic
and industrial boilers: A numerical approach. Energy 2021;236:121456.

[2] Cirilli F, Jochler G, Mosconi M, Praolini F. Effects of H2 combustion on

scale growth and steel surface quality in reheating furnaces. Matériaux
& Techniques 2021,109(302).

[3] Ilbas M, Syred N. Computer modelling of combustion and nitrogen

oxides in a small scale two-inlet cyclone combustor. First European
Conference on Small Burner Technology and Heating Equipment 1996;
1:Switzerland 269–278.

[4] Ilbas M, Yılmaz I, Nejat T, Kaplan Y. Hydrogen as burner fuel:

modelling of hydrogen-hydrocarbon composite fuel combustion and NOx
formation in a small burner. International Journal of Energy Research
2005;29(11):973–990.

[5] Hsu CK, Lee CL, Wang CH, Jou CJG. Reduction of energy consumption

31
 and pollution emissions for industrial furnace using hydrogen-rich tail
gas. International Journal of Hydrogen Energy 2014;39:9675–9680.

[6] Zhao Y, McDonell V, Samuelsen S. Assessment of the combustion

performance of a room furnace operating on pipeline natural gas mixed
with simulated biogas or hydrogen. International Journal of Hydrogen
Energy 2020;45:11368–11379.

[7] Xiao H, Wang Q, Shen X, An W, Duan Q, Sunet J. An experimental

study of premixed hydrogen/air flame propagation in a partially open
duct. International Journal of Hydrogen Energy 2014;39(11):6233–6241.

[8] González-Espinosa A, Gil A, Royo-Pascual L, Nueno A,Herce C.

Effects of hydrogen and primary air in a commercial partially-premixed
atmospheric gas burner by means of optical and supervised machine
learning techniques. International Journal of Hydrogen Energy 2020;
45:31130–31150.

[9] Cellek MS, Pınarbaşı A. Investigations on performance and emission

characteristics of an industrial low swirl burner while burning natural
gas, methane, hydrogen-enriched natural gas and hydrogen as fuels.
International Journal of Hydrogen Energy 2018;43(2), 1194–1207.

[10] Jo S, Park S, Kim HJ, Lee JT. Combustion improvement and emission
reduction through control of ethanol ratio and intake air temperature in
reactivity controlled compression ignition combustion engine. Applied
Energy 2019;250:1418–1431.

[11] Eren Vedin A, Dologlu P, Akcin M, Karimova A, Kaplan KB, Sener

A, Karan O. Real time optimization of sour gas processing unit via
sulfur dioxide emissions predictive model. Computer Aided Chemical
Engineering 2024;53:2413–2418.

[12] Darabkhani HG, Varasteh H, Bazooyar B. Carbon Capture Technologies

for Gas-Turbine-Based Power Plants”. First edition, Elsevier 2023.

[13] Boardman RD, Douglas SL. Prediction of nitric oxide in advanced

combustion systems. AIChE Journal 1988;34: 1573–1576.

32
[14] Hill SC, Smoot LD. Modeling of nitrogen oxides formation and
destruction in combustion systems. Progress in Energy and Combustion
Science 2000;26: 417–458.

[15] Alzaharnah I, Habib MA, Elshafei M, Ahmed P. Control of the

boiler swing rate for NO emission minimization. Energy and Fuels
2013;27:6079–6086.

[16] Ferretti G, Piroddi L. Estimation of NOx emissions in thermal power

plants using neural networks. Journal of Engineering for Gas Turbines
and Power 2001;123:465–471.

[17] Pallarés J, Arauzo I, Williams A. Integration of CFD codes and

advanced combustion models for quantitative burnout determination.
Fuel 2007;86:2283–2290.

[18] Bazooyar B, Coomson G, Manovic V, Nabavi SA. Comparative analysis

of ammonia combustion for domestic applications. Journal of Energy
Institue 2023;106:101130–101142.

[19] Aleksandrov YB, Sabirzyanov AN, Yavkin VB. Influence of simplifying a

geometrical model of the gas turbine engine combustion chamber on the
results of numerical modeling. Russian Aeronautics 2017;60(4):575–582.

[20] Zhai X, Ding S, Cheng Y, Jin Y, Cheng Y. CFD simulation with detailed
chemistry of steam reforming of methane for hydrogen production in
an integrated micro-reactor. International Journal of Hydrogen Energy
2010;35(11):5383–5392.

[21] Bazooyar B, Darabkhani HG. Design, manufacture and test of a

micro-turbine renewable energy combustor. Energy Conversion and
Management 2020;213(8):112782–112793.

[22] Bazooyar B, Darabkhani HG. The design strategy and testing

of an efficient microgas turbine combustor for biogas fuel. Fuel
2021;294:120535–120543.

[23] Bazooyar B, Zhu M, Manovic V, Nabavi SA. Direct Numerical

Simulation (DNS) of packed and monolith Syngas Catalytic Combustors
for Micro Electrical Mechanical Systems (MEMS). Energy Conversion
and Management: X 2023;20:100422

33
[24] Bazooyar B, Darabkhani HG. Design and numerical analysis of a 3
kWe flameless microturbine combustor for hydrogen fuel. International
Journal of Hydrogen Energy 2019;44:11134–11144.

[25] Bazooyar B, Ebrahimzadeh E, Jomekian A, Shariati A. NOx formation

of biodiesel in utility power plant boilers. part a: Influence of fuel
characteristics. Energy and Fuels 2014;28:3778–3792.

[26] Fisher EM, Pitz WJ, Curran HJ, Westbrook CK. Detailed chemical
kinetic mechanisms for combustion of oxygenated fuels. Proceedings of
the Combustion Institute 2000;28:1579–1586.

[27] Brakora J, Reitz R. A comprehensive combustion model for biodiesel-

fueled engine simulations. SAE Technical Papers 2013;1: 1099–1119.

[28] Bulat G, Jones WP, Marquis AJ. NO and CO formation in an industrial

gas-turbine combustion chamber using LES with the Eulerian sub-grid
PDF method. Combustion and Flame 2014;161:1804–1825.

[29] Lai JYW, Lin KC, Violi A. Biodiesel combustion: Advances in

chemical kinetic modeling. Progress in Energy and Combustion Science
2011;37:1–14.

[30] Macor A, Pavanello P. Performance and emissions of biodiesel in a boiler

for residential heating. Energy 2009;34:2025–2032.

[31] Bazooyar B, Ghorbani A, Shariati A. Combustion performance and

emissions of petrodiesel and biodiesels based on various vegetable oils in
a semi industrial boiler. Fuel 2011;90:3078–3092.

[32] Barnes CD, Garwood DR, Price TJ. The use of biodiesel blends in
domestic vaporising oil burners. Energy 2010;35:501–505.

[33] Bazooyar B, Shariati A, Hashemabadi SH. Economy of a utility boiler

power plant fueled with vegetable oil, biodiesel, petrodiesel and their
prevalent blends. Sustainable Production and Consumption 2015;3:1–7.

[34] Bazooyar B, Hashemabadi SH, Shariati A. NOX formation of biodiesel in

utility power plant boilers; Part B. Comparison of NO between biodiesel
and petrodiesel. Fuel 2016;182:323–332.

34
[35] Dixit S, Kumar A, Kumar S, Waghmare N, Thaku HC, Khan S. CFD
analysis of biodiesel blends and combustion using Ansys Fluent. Material
Today Proceedings 2019;26: 665–670.

[36] Pereira C, Wang G, Costa M. Combustion of biodiesel in a large- scale

laboratory furnace. Energy 2014;74: 950-955.

[37] Glaude PA, Fournet R, Bounaceur R, Moliere M. Gas turbines and

biodiesel: A clarification of the relative NOx indices of fame, gasoil
and natural gas. Turbomachinery Technical Conference and Exposition
2009;1:419–425.

[38] Cundari L, Komariah LN, Novia IM, Septiana L. Temperature

distribution of biodiesel blends combustion in boiler using CFD-fluent.
International Journal on Advanced Science Engineering and Information
Technology 2016;6:120–127.

[39] Shaahmadi F, Anbaz MA, Bazooyar B. Analysis of intelligent models in

prediction nitrous oxide (N2O) solubility in ionic liquids (ILs). Journal
of Molecular Liquids 2017;246: 48–57.

[40] Kozarac D, Vuilleumier D, Saxena S, Dibble RW. Analysis of benefits of

using internal exhaust gas recirculation in biogas-fueled HCCI engines.
Energy Conversion and Management 2014;87:1186–1194.

[41] Sudheesh K, Mallikarjuna JM. Diethyl ether as an ignition improver for

biogas homogeneous charge compression ignition (HCCI) operation-an
experimental investigation. Energy 2010;35(9):3614–3622.

[42] Feroskhan M, Ismail S. A review on the purification and use of biogas

in compression ignition engines”. International Journal of Automotive
and Mechanical Engineering 2017;14(3):4383–4400.

[43] Jiang CQ, Liu TW, Zhong JL. A study on compressed biogas and
ıts application to the compression ıgnition dual-fuel engine. Biomass
1989;20(1-2):53–59.

[44] Sahoo BB, Saha UK, Sahoo N. Diagnosing the effects of pilot fuel quality
on exergy terms in a biogas run dual fuel diesel engine. International
Journal of Exergy 2012;10(1):77–93.

35
[45] Kim Y, Kawahara N, Tsuboi K, Tomita E. Combustion characteristics
and NO x emissions of biogas fuels with various CO 2 contents in a micro
co-generation spark- ignition engine. Applied Energy 2016;182:539–547.

[46] Sathyamurthy R, Krishnan SAH. Krishnan, investigation on LPG-

Biogas blends in spark ignition engine. International Journal of
Engineering Research in Africa 2015; 14:58–62.

[47] Porpatham E, Ramesh A, Nagalingam B. Investigation on the effect

of concentration of methane in biogas when used as a fuel for a spark
ignition engine. Fuel 2008; 87(8-9):1651–1659.

[48] Crookes RJ. Comparative bio-fuel performance in internal combustion

engines. Biomass & Bioenergy 2006;30(5):461–468.

[49] Polk AC, Gibson CM, Shoemaker NT, Srinivasan KK, Krishnan SR.
Analysis of ignition behavior in a turbocharged direct injection dual fuel
engine using propane and methane as primary fuels. Journal of Energy
Resources Technology-Transactions of the ASME 2013;135(3):1–10.

[50] Mustafi NN, Raine RR, Verhelst S. Combustion and emissions

characteristics of a dual fuel engine operated on alternative gaseous fuels.
Fuel 2013;109:669–678.

[51] Yamasaki Y, Kanno M, Suzuki Y, Kaneko S. Development of an engine

control system using city gas and biogas fuel mixture. Applied Energy
2013;101:465–474.

[52] Verma S, Das LM, Kaushik SC, Tyagi SK. An experimental investigation
of exergetic performance and emission characteristics of hydrogen
supplemented biogas-diesel dual fuel engine. International Journal of
Hydrogen Energy 2018;43(4):2452–2468.

[53] Cheolwoong P, Seunghyun P, Yonggyu L, Changgi K, Sunyoup L, Yasuo

M. Performance and emission characteristics of a SI engine fueled by
low calorific biogas blended with hydrogen. International Journal of
Hydrogen Energy 2011;36(16):10080–10088.

[54] Leykun MG, Mekonen MW. Investigation of the Performance and

Emission Characteristics of Diesel Engine Fueled with Biogas-Diesel
Dual Fuel. Fuels 2022;3:15–30.

36
[55] Majhool AAA, Alqurashi F, Kareem SF, Jasim NM. Stoichiometry
preservation of injection n-pentane/hydrogen binary mixtures on the
combustion characteristics of liquid spray fuel jets. International Journal
of Hydrogen Energy 2024;73:510–523

[56] Majhool AA. Advanced spray and combustion modelling. Manchester

University 2011.

[57] Majhool AA, Jasim NM. Prediction of the initial drop size and
velocity distribution in the cold cryogenic spray. Int. J. Heat Technol
2020;38(3):629–640.

[58] Majhool AA, Hamza NH, Jasim NM. Spray Interface Drag Modeling
Based on the Power-Law Droplet Velocity Using the Moment Theory.
Journal of Applied Mechanics and Technical Physics 2020;61:61–69.

[59] ANSYS Fluent User’s Guide, ANSYS, Inc., release 15.0. edition, 2015.

[60] Majhool AA , Watkins AP. Spray algorithm without interface

construction. Journal of Computational Physics 2012;231:3647–3662.

[61] Mao CP, Szekely JGA, Faeth GM. Evaluation of a locally homogeneous
model of spray combustion. Journal of Energy 1980;4.2:78–87.

[62] Chan SH, Abou-Ellail M. A two-fluid model for reacting turbulent two-
phase flows. J. Heat Transfer 1994;116:427–435.

[63] Versteeg HK, Malalasekera W. An Introduction to Computational Fluid

Dynamics. Pearson, second edition, 2006.

[64] Patankar SV . Numerical heat transfer and fluid flow. Hemisphere

publishing corporation, Taylor and Francis group, New York, 1980.

37