Effects of fuel preheat temperature on soot formation in methyl linolenate co-flow diffusion flames

Abstract

The objective of this study was to investigate the mechanism of soot formation in biodiesel by analyzing the combustion of individual components. The paper presents a numerical analysis of the effect of preheat temperatures on nucleation rates, coagulation rates, and soot volume fraction in methyl linolenate (MLe) co-flow flame. In this work, Moss-Brooke’s soot model and a reduced kinetic mechanism containing 177 chemical species and 2904 chemical reactions were used to simulate the pyrolysis and combustion of MLe. A laminar jet flame with inlet velocities of 0.4 m/s was studied. The preheat temperature of the fuel was varied between 300 and 450 K. The burner walls were stationary and no-slip conditions were applied. The pressure outlet had Neumann boundary conditions and the tangential velocity was set to zero at the wall. It was established that an increase in fuel preheat temperatures causes an increase in nucleation rates and the amount of soot due to accelerated fuel pyrolysis, improved diffusion, acceleration from buoyancy, and earlier formation of PAHs. It was discovered that increasing the fuel preheat temperature had a greater impact on soot formation along the centerline than on the wing.

Keywords:

• Methyl linolenate
• jet-flame
• co-flow
• soot formation
• biodiesel
• preheating

REVIEWING EDITOR:

• D T Pham
• University of Birmingham
• United Kingdom

SUBJECTS:

• Heat Transfer
• Thermodynamics
• Chemical Engineering

1. Introduction

High population growth and environmental degradation are some of the key factors that demand a search for a substitute fuel (Mao et al., 2020). Biodiesel is one of the fuels that can be used as an alternative because it is environmentally friendly and can be synthesized from renewable materials. Biodiesel consists of five major components. The constituents are methyl linolenate (MLe), methyl linoleate (MLi), methyl oleate (MO), methyl palmitate (MP), and methyl stearate (MS), the proportions of these methyl esters vary depending on the type of feedstock (Hoekman et al., 2012).

Soot contributes significantly to particulate emission from hydrocarbon combustion (Johnson and Joshi, 2018). Soot in flame is caused by incomplete combustion of hydrocarbons due to insufficient oxygen required to convert fuel to carbon dioxide and water vapor (Vinayagam et al., 2021). To reduce soot emissions, the effect of controlling parameters on soot formation has been extensively researched (Mazzei et al., 2017), e.g., composition of fuel (Gu et al., 2016), temperature (Sun et al., 2017) and pressure (Sarnacki and Chelliah, 2018).

A complex chemical process (gas-phase process) and a physical process combine to form soot (soot particle dynamics). The formation of soot begins with pyrolysis and oxidation, followed by combination and cyclization reactions that result in the formation of polycyclic aromatic hydrocarbons (PAHs) (Frenklach and Wang, 1994). Polycyclic aromatic hydrocarbons nucleate as they grow, resulting in the formation of soot. Following nucleation, the primary soot particles congeal and develop a surface (Mazzei et al., 2017). Surface growth is caused by the condensation of polycyclic aromatic hydrocarbons and hydrogen abstraction carbon addition (HACA) surface growth mechanism. This expansion is responsible for an increase in the size and mass of soot particles (Celnik et al., 2009). Primary particles collide to form particle aggregates that resemble chains. Because the soot is formed at high temperatures, it is constantly accompanied by oxidation. O₂ and OH both play important roles in oxidation (Westbrook et al., 2009).

Taylor et al. (2014) studied the impact of preheating the flame on soot creation in a two-dimensional laminar diffusion flame involving ethylene diffusion flame. Detailed gas-phase reaction mechanism and complex thermal and transport properties are employed during simulation. Two-equation soot model was applied to simulate the soot formation together with detailed gas-phase chemistry. The numerical outcomes were compared with experimental results which revealed that the effect of flame preheating significantly affects the soot formation predictions. Fuel preheat temperatures can have a significant impact on soot loading in flames (Bai et al., 2016). Wu et al. (2013) studied the emission and performance of preheated croton megalocarpus oil in a generator. The authors noted that the exhaust temperatures were higher than the in-cylinder temperatures. There was a decrease in particulate matter emission when the oil was preheated due to improved atomization. Similarly, Karabektas et al. (2008) reported that preheated cottonseed oil methyl ester produced higher combustion temperatures and reduced CO₂ emission. This is because heating decreases the viscosity of biodiesel which leads to improved oxidation. Kodate et al. (2021) study on emission and performance of preheated Dhupa oil biodiesel established that preheating reduced the smoke opacity due to improved combustion.

Chu et al. (2021) studied the effects of reactant preheat in ethylene coflow flames using both experimental and numerical approach. Ethylene flames were established under different reactant preheat temperatures (both for air and fuel), with values of 300 K (low temperature), 473 K (medium temperature), 673 K (high temperature) and 713 K (ultra-high temperature). Soot volume fractions, primary particle diameters and the temperature of the soot in the flame were measured during experimental analysis. The authors reported that both experimental and numerical data indicated that soot formation was promoted by higher soot surface growth. Numerical investigation revealed that PAHs adsorption, which is a function of PAH concentration, becomes important at increased preheated temperatures as its mass contribution increases from 50 to 70%. This was attributed to early pyrolysis of the fuel. Konsur et al. (1999) conducted a study to examine the effect of fuel preheating on soot structure in a laminar jet flames. The authors reported that fuel preheating resulted to higher soot concentration along the central axis of the flame while reducing the formation of soot in the outer annular area of flame.

Significant progress has been made in the development of soot models. There are three types of soot models, according to Kennedy et al. (1997): empirical, semi-empirical, and detailed soot models. To develop empirical soot models, experiments based on phenomenological relationships between soot formation rates and combustion conditions are used e.g., lift-off model. Semi-empirical models attempt to integrate some aspects of the chemistry and physics of the phenomenon e.g., Moss-Brooke soot model, whereas detailed models are based on basic combustion chemistry and aerosol dynamics theory e.g., kinetics and physical models.

Previous research has focused on the performance and emissions of preheated biodiesel and biodiesel blends in various combustion devices. Various types of biodiesels have been used to study the effects of preheating, though the effects of preheating major biodiesel components on soot formation is still unclear. As a result, the purpose of this research is to investigate the effects of preheating major biodiesel components on soot formation by determining the rate of soot nucleation and soot volume fraction in a preheated Methyl linolenate co-flow jet flame. The rates of nucleation, coagulation and soot volume fraction were simulated using the semi-empirical Moss-Brooke soot model within the ANSYS FLUENT software. Insight on the effects of preheating biodiesel components on soot formation aids in understanding the soot formation in biodiesel individual components.

2. Material and method

2.1. Fuel sample

This study focused on one of the major biodiesel methyl esters: methyl linolenate (C₁₉H₃₈O₂). MLe was chosen as a fuel because it is a major component of biodiesel, has a high number of carbon atoms, and a high number of carbon double bonds when compared to other components.

2.2. Numerical modelling

In this work, the governing equations for a 2-D axisymmetric configuration was used to simulate the jet flame in ANSYS FLUENT. The velocity components v, density, pressure p, temperature T, and mass fraction Yi, I = 1, 2, …, N are the dependent variables. The independent variables are the distance (z) from the inlet along the axis and radius (r) from the center. The co-flow configuration is represented by mass, momentum, species, and energy conservation Eqs. (1)–(5) (Shkiro et al., 1978).

Mass conservation: (1) $$\frac{1}{r}\frac{\partial }{\partial r}\left(r\rho u\right)+\frac{\partial }{\partial z}\left(\rho u\right)=0,$$(1) where v and u are radial and axial velocities, z and r are axial and radial coordinates, and is the density mixture.

Momentum equations:

Axial component: (2) $$\rho v\frac{\partial u}{\partial r}+\rho u\frac{\partial u}{\partial r}=-\frac{\partial p}{\partial z}+\frac{1}{r}\frac{\partial }{\partial r}\left(r\mu \frac{\partial u}{\partial r}\right)+2\frac{\partial }{\partial z}\left(\mu \frac{\partial u}{\partial z}\right)-\frac{2}{3}\frac{\partial }{\partial z}\mathrm{ }\left[\frac{\mu }{r}\frac{\partial u}{\partial r}\mathit{\text{rv}}\mathrm{ }\right]‐\mathrm{ }\frac{2}{3}\frac{\partial }{\partial z}\left(\mu \frac{\partial u}{\partial z}\right)+\frac{1}{r}\frac{\partial }{\partial r}\left(r\mu \frac{\partial u}{\partial r}\right)+\rho g_z.$$(2)

Radial component: (3) $$\rho v\frac{\partial u}{\partial r}+\rho u\frac{\partial u}{\partial z}=-\frac{\partial p}{\partial r}+\frac{\partial }{\partial z}\left(\mathrm{r}\mu \frac{\partial u}{\partial r}\right)+\frac{2}{r}\frac{\partial }{\partial r}\left(r\mu \frac{\partial u}{\partial r}\right)-\frac{2}{3}\frac{1}{r}\frac{\partial }{\partial r}\mathrm{ }\left[\mu \frac{\partial }{\partial r}\left(\mathit{\text{rv}}\right)\right]+\frac{1}{r}\frac{\partial }{\partial r}\left(r\mu \frac{\partial u}{\partial z}\right)+\frac{\partial }{\partial z}\left(\mu \frac{\partial u}{\partial r}\right)-\frac{2\mu v}{r^2}+\frac{2}{3}\frac{\mu }{r^2}\mathrm{ }\left[\frac{\partial u}{\partial r}\mathit{\text{rv}}\mathrm{ }\right]+\frac{2}{3}\frac{\mu }{r}\frac{\partial u}{\partial z},$$(3) where μ is the mixture’s dynamic viscosity; pressure is denoted by p and gravitational acceleration along the axis denoted by $g_z.$ (4) $$c_p\left(\rho v\frac{\partial T}{\partial r}+\rho u\frac{\partial T}{\partial z}\right)=\frac{1}{r}\frac{\partial }{\partial r}\left(r\lambda \frac{\partial T}{\partial r}\right)+\frac{\partial }{\partial z}\left(\lambda \frac{\partial T}{\partial z}\right)-\sum _{K=1}^{\mathit{\text{KK}}}\left[\rho c_{p,k}{Y}_k\left(v_{k,r}\frac{\partial T}{\partial r}+v_{k,z}\frac{\partial T}{\partial z}\right)\right]-{\sum }_{K=1}^{\mathit{\text{KK}}}{h}_k{w}_k{\dot{w}}_s-\rho c_{p,s}{Y}_{s\mathrm{ }}\left(v_{\mathit{\text{Ts}},r}\frac{\partial T}{\partial r}+v_{\mathit{\text{Ts}},z}\frac{\partial T}{\partial z}\right)-h_s{w}_{s\mathrm{ }}{\dot{w}}_s.$$(4)

Energy conservation equation: (5) $$\rho v\frac{\partial Y_k}{\partial r}+\rho u\frac{\partial Y_k}{\partial r}=\frac{1}{r}\frac{\partial }{\partial r}\left(r\rho Y_k{u}_{k,r}\right)-\frac{\partial }{\partial z}\left(\rho Y_k{u}_{k,z}\right)+w_k{\dot{w}}_{k}k=1,\dots \dots .\mathit{\text{KK}},$$(5) where $c_p$ is the specific heat of the mixture at constant pressure; T is the temperature; and the thermal conductivity of the mixture is denoted λ; $c_{p,k}$ denotes the species of kth heat capacity at constant pressure; ℎ_k is the species’ specific enthalpy. $c_{p,s}$ denotes soot specific heat. Y_s is the soot mass fraction; v_Ts,z and v_Ts,r are the axial and radial soot particle velocities, respectively; h_s is the soot specific enthalpy; w_s is the soot molecular weight; and ${\dot{w}}_s$ is the rate of soot production.

Equation of state: (6) $$p=\rho \mathrm{\text{RT}}\sum _{k=1}^{\mathit{\text{KK}}}\left(\raisebox{1ex}{$Y_k$}\!\left/ \!\raisebox{-1ex}{$w_k$}\right.\right),$$(6) where Y_k is the kth mass fraction of the species, and the kth axial and radial species velocities are denoted by v_k,z and v_k,r, respectively; w_k represents the kth molecular weight of the species. The total number of gaseous species is KK, and ${\dot{w}}_k$ is the kth molar production rate of the species per unit volume.

2.2.1 Soot model

The Moss-Brooke soot model is a tool for analyzing soot formation in relation to factors such as fuel type, fuel composition, and oxidant type. The model was used to compute soot volume fraction, nucleation and coagulation rates in MLe flame. This model was chosen because it is compatible with non-premixed combustion models (Patki et al., 2014). Furthermore, the study was limited to the semi-empirical Moss-Brookes model because it is easier to use and provides a better understanding of the soot formation mechanism. Benzene (C₆H₆) is one of the major soot precursors used in studying soot formation. Benzene breaks down into intermediate species such as acetylene (C₂H₂). Lignell et al. (2007) reported that acetylene (C₂H₂), ethylene (C₂H₄) and propylene(C₃H₆) were major intermediate species during combustion of palm oil methyl esters while C₂H₂ was found on both rich side and lean side of the flame. Therefore, in this study C₂H₂ was used as a precursor for convenience of modelling using ANSYS FLUENT. The model calculates mass and nuclei concentration using Eqs. (7)(7) $$\frac{\partial }{\partial t}\rho Y_{\mathrm{\text{soot}}}+\Delta \bullet \left(\rho v{Y}_{\mathrm{\text{soot}}}\right)=\Delta \bullet \left(\frac{{\mu }_t}{{\sigma }_{\mathrm{\text{soot}}}}\Delta Y_{\mathrm{\text{soot}}}\right)+\frac{\mathit{\text{dM}}}{\mathit{\text{dt}}},$$(7) and (8)(8) $$\frac{\partial }{\partial t}\rho b_{\mathrm{\text{nuc}}}^{\mathrm{*}}+\Delta \bullet \left(\rho v{b}_{\mathrm{\text{nuc}}}^{\mathrm{*}}\right)=\Delta \bullet \left(\frac{{\mu }_t}{{\sigma }_{\mathrm{\text{nuc}}}}\Delta Y_{\mathrm{\text{nuc}}}\right)+\frac{1}{N_{\mathrm{\text{norm}}}}\frac{\mathit{\text{dN}}}{\mathit{\text{dt}}},$$(8) (Smooke et al., 1991). (7) $$\frac{\partial }{\partial t}\rho Y_{\mathrm{\text{soot}}}+\Delta \bullet \left(\rho v{Y}_{\mathrm{\text{soot}}}\right)=\Delta \bullet \left(\frac{{\mu }_t}{{\sigma }_{\mathrm{\text{soot}}}}\Delta Y_{\mathrm{\text{soot}}}\right)+\frac{\mathit{\text{dM}}}{\mathit{\text{dt}}},$$(7) (8) $$\frac{\partial }{\partial t}\rho b_{\mathrm{\text{nuc}}}^{\mathrm{*}}+\Delta \bullet \left(\rho v{b}_{\mathrm{\text{nuc}}}^{\mathrm{*}}\right)=\Delta \bullet \left(\frac{{\mu }_t}{{\sigma }_{\mathrm{\text{nuc}}}}\Delta Y_{\mathrm{\text{nuc}}}\right)+\frac{1}{N_{\mathrm{\text{norm}}}}\frac{\mathit{\text{dN}}}{\mathit{\text{dt}}},$$(8) where $\rho$ is the soot density, Y_soot is mass fraction of soot$,{\sigma }_{\mathrm{\text{soot}}}$ is the soot prandtl number, ${\sigma }_{\mathrm{\text{nuc}}}$ is the nuclei transport prandtl number, m is mass concentration of soot (kg/m³), μ_t denotes the dynamic viscosity coefficient, b^∗_nuc is the concentration of nuclei (particles × 10⁻¹⁵/kg) = $\frac{N}{\rho N\mathrm{\text{norm}}},$ ν is the mass stoichiometry and N is the number density for a soot particle (particles/m³) N_norm = 1015 particles. The immediate rate of soot production which depends on coagulation and gas phase nucleation is given by Eq. (9)(9) $$\frac{\mathit{\text{dN}}}{\mathit{\text{dt}}}=\left[{C_{\alpha }{N}_A\left(\frac{X_{\text{prec}}P}{\mathit{\text{RT}}}\right)}^l\mathrm{e}\mathrm{\text{xp}}\left(-\frac{T_{\alpha }}{T}\right)\right]-\left[C_{\beta }{\left(\frac{24\mathit{\text{RT}}}{{\rho }_{\mathrm{\text{soot}}}}\right)}^{\frac{1}{2}}{d_p}^{\frac{1}{2}}{N}^2\right].$$(9) . (9) $$\frac{\mathit{\text{dN}}}{\mathit{\text{dt}}}=\left[{C_{\alpha }{N}_A\left(\frac{X_{\text{prec}}P}{\mathit{\text{RT}}}\right)}^l\mathrm{e}\mathrm{\text{xp}}\left(-\frac{T_{\alpha }}{T}\right)\right]-\left[C_{\beta }{\left(\frac{24\mathit{\text{RT}}}{{\rho }_{\mathrm{\text{soot}}}}\right)}^{\frac{1}{2}}{d_p}^{\frac{1}{2}}{N}^2\right].$$(9)

The model constants are C_β, C_α and l. Where Avogadro number is (NA = 6.022045 × 10²⁶ kmol ⁻¹) and X_prec is precursor mole fraction. $T_{\alpha }$ is the soot inception. P represent pressure and R denotes molar constant. The mean particle diameter of soot is d_p and soot density considered to be 1800 kg/m³.

2.2.2. Boundary conditions

A 2-D computational domain with a radius of 0.045 m and an axial length of 0.5 m as shown in Figure 1 was considered. Laminar jet flame with inlet velocities at 0.4 m/s was considered in this analysis. The domain was set large enough to ensure a fully developed flame. A uniform mesh along the r and z direction was generated. A finer mesh with 373,280 elements was adopted for the study.

Figure 1. Computational domain.

Uniform velocity was set at inlet co-flow air stream with O₂ and N₂ mass fraction set at 0.233 and 0.767, respectively. Species mass fraction, velocity, pressure and temperature at the inlet were quantified using Dirichlet boundary conditions. Preheat temperature was varied between 300 K and 450 K while pressure was maintained constant at 1 bar. This preheat temperature range was chosen to ensure that the highest preheat temperature was above MLe flash point, which is 386 K. The burner walls were considered stationary with no-slip conditions. Neumann boundary conditions were applied at the pressure outlet and the tangential velocity was set to zero at the wall. The ANSYS Fluent package, integrated into the ANSYS Workbench software (version 2022 R1), was utilized to solve the equations for a steady state combustion in a non-premixed, laminar co-flow diffusion flame. The energy, momentum, continuity, and species equations were all solved using the pressure-based solver in ANSYS Fluent.

The chemical reaction used was based on the FAME mechanism from the CRECK modeling group, which includes 2904 chemical reactions and 177 species (Ranzi et al., 2014). The FAME mechanism was chosen for its ability to display the physio-chemical properties and chemical structure of methyl linolenate. This model was validated by using experimental data from key FAME components (Battin-Leclerc, 2008).

3. Results and discussion

The computation domain was divided into smaller sections using the quadrilateral method, creating nodes and elements. The mesh distribution at the inlet of the domain is illustrated in Figure 2. Various mesh sizes were tested to determine the best mesh size for improved accuracy. The mesh test included 373,280, 320,000, 93,944, and 46,977 elements and the mesh refinement were done equally along the z and r axis. It was found that there was very little difference in temperature profiles when using a mesh size of 373,280 elements compared to 320,000 elements. So, a mesh size of 373,280 elements was ultimately chosen for the study.

Figure 2. Section of mesh distribution used at the inlet.

Figure 3 shows axial profiles of flame soot temperature at different preheat temperatures. The temperature of the flame increases gradually until it reaches the highest point right after ignition, as shown by profiles. It was observed that the soot temperature in the central region increases with increase in fuel preheat temperature. When the fuel was heated to 450 K before combustion, the computed flame temperature was observed to be particularly high at 2260 K. The model’s accuracy was determined by comparing its results to data found in previous studies. The temperature profiles found in this study were found to be consistent with the results from Chun et al. (2022) which reported a temperature range of 2000–2500 K for biodiesel methyl esters combustion. The main limitation of this study was the lack of experimental data on combustion of MLe fuel for validation. Though there is no experimental data, the model flame temperatures were validated using biodiesel methyl esters temperature range because MLe is one of the major biodiesel components. Therefore, this model can be used to study the effect of preheating MLe on the rate of soot nucleation and soot volume fraction. Uncertainty and limitation of the simulation in this study are associated to the chemical kinetic model and soot formation model adopted. Moss-Brookes soot model that was used in this study limited simulation of soot formation and analysis. The model could only account for C₂H₂ as the only soot precursor. In addition, the model also assumes that the shape of soot particles is spherical and have a monodisperse size distribution.

Figure 3. MLe temperature profiles for different preheated temperatures.

Soot nucleation refers to the process of how particulate matter forms from molecular precursors in the gas phase (Frenklach and Wang, 1990). In Moss-Brooke’s soot model, acetylene (C₂H₂) was considered as a major soot precursor during the computation of the nucleation rate. Figures 4 and 5 represents nucleation rates plots and nucleation rate contours of MLe in flames respectively. The plots display a visual representation of soot nucleation rates along the axis at different preheat temperature while the contours display a visual representation on nucleation rates in the entire flame. The figures show that there was increased nucleation at the centerline of MLe flame. The nucleation rate was found to be higher at a temperature of 450 K than a temperature of 300 K. According to the results, increasing preheat temperature from 300 to 400 K increased the nucleation rate by 15%. The main reason for the high rate of fuel pyrolysis and the rapid formation of PAHs is a high initial temperature. The rate of nucleation increased as the preheat temperature increased, as a result of increased MLe pyrolysis and the formation of PAHs. Mahmoud et al. (2019) studied the effects of fuel inlet boundary conditions on formation of aromatic species. The authors reported that preheat temperature increased the formation of aromatic species which promotes the nucleation process of soot.

Figure 4. MLe Nucleation rates for different preheat temperature.

Figure 5. Images of nucleation rates at different preheat temperature.

When two small soot particles come into contact, they combine and become a larger particle, a process called coagulation. It is believed that this results in the formation of a bigger spherical particle, while still maintaining the overall volume of soot (Maricq, 2007). Figures 6 and 7 represents coagulation contours and coagulation rate plots of MLe in flames respectively. The plots display a visual representation of coagulation rates along the axis at different preheat temperature while the contours display a visual representation on coagulation rates in the entire flame. The coagulation rates increased steadily to maximum at 0.1 m from the inlets. As the preheat temperature increased, the rate of soot particle combination decreased. When the preheat, temperature was set to 450 K, fewer soot particles combined compared to a preheat temperature of 300 K. An increase in preheat temperature resulted to decreased coagulation rates. According to results, computed coagulation rates decreased by 6% when the preheat temperature was raised from 300 K to 450 K. This is due to higher preheat and flame temperatures lead to more frequent collisions among soot particles, leading to fewer smaller particles. Similarly Sato et al. (1991) reported that an increase in preheat temperature resulted in a decrease in clumping of soot particles.

Figure 6. Images of coagulation rates at different preheat temperature.

Figure 7. MLe coagulation rates profiles at different preheat temperature.

Understanding the mechanisms of soot production and how soot is distributed within combustion systems are crucial for various aspects of combustion research. Figures 8 and 9 represents soot volume fraction plots and soot volume fraction contours of MLe in flames respectively. The plots display a visual representation of soot volume fraction along the axis at different preheat temperature while the contours display a visual representation of soot volume fraction in the entire flame. At a preheat temperature of 450 K, the formation of soot occurs slightly earlier and later than at other preheat temperatures. A higher soot volume fraction was observed at a preheat temperature of 450 K, while the lowest soot volume fraction was recorded at 300 K. An increase in preheat temperature resulted to increased soot volume fraction. According to results, computed soot volume fraction increased 32% when the preheat temperature was raised from 300 K to 450 K. It was discovered that increasing the fuel preheat temperature had a greater impact on soot formation along the centerline than on the wing. Similar observations were reported by Lou et al. (2018); Chun et al. (2010). This is believed to be related to the earlier onset of soot inception along the centerline. A similar conclusion was reached by Khosousi et al. (2015), that increasing the preheat temperature slightly increased the production of soot precursor PAHs in large hydrocarbon flames. However, this change also influenced the buoyancy-induced acceleration, which determined the amount of time available for soot growth. As a result, extending the available time promoted soot formation.

Figure 8. Images of soot volume fraction at different preheated temperatures.

Figure 9. Soot volume fraction at different preheat temperatures.

4. Conclusion

The study examined the effect of preheating fuel on nucleation rates, coagulation rates, and soot volume fraction in MLe co-flow diffusion flame, with initial fuel temperatures ranging from 300 to 450 K. The main findings of this study are as follows:

• An increase in fuel preheat temperature leads to an increase in the rate of fuel pyrolysis and quick formation of PAHs. Exploring the formation of PAHs is very significant in studying soot formation because it gives insight towards achieving more regulated and environmentally friendly combustion.

• As the preheat temperature increases, the coagulation rate also reduces due to the high flame temperatures resulting in more frequent collisions among soot particles, which causes fewer smaller particles.

• An increase in fuel preheats temperature results in higher temperatures and increased amount of soot due to increased diffusion and acceleration caused by buoyancy. This affects the amount of time available for soot to form.

Acknowledgement

The authors gratefully acknowledge Dedan Kimathi University of Technology, Kenya for the support provided during this study.

Disclosure statement

No potential conflict of interest was reported by the authors.