Numerical Investigation of Compatibility of Synthetic Natural Gas with

Conventional Liquefied Petroleum Gas Burners

B Pandey1*, R Ghimire1, N Sapkota2, A Dev1

1
Department of Mechanical Engineering, Kathmandu University, Nepal
2
Metal Cast Pvt. Ltd, Butwal, Nepal
* Corresponding author [email protected]

Abstract. Nepal is seeking carbon-free alternative fuels due to environmental concerns and economic
issues resulting from fossil fuel usage. Carbon-neutral SNG has the potential as an alternative to LPG
gas in domestic cooking, but traditional LPG equipment may not be compatible with SNG.
Computational Fluid Dynamics (CFD) was used to study the compatibility of traditional LPG equipment
with SNG, and it was found that essential optimization is needed to obtain optimal cooking conditions
by SNG combustion. Engineering Equation Solver (EES) was used to determine the optimum inlet
pressure for varying nozzle sizes. Based on the results of the study, it was determined that the optimal
nozzle size for the selected type of burner is 1.15 mm. This particular nozzle size was found to provide
better uniformity of flame temperature and complete combustion, resulting in an average temperature of
1700 ̊K at a fuel inlet pressure of 2.75 kPa. This makes it a potentially better option for cooking
compared to LPG, as it could potentially provide faster cooking times. The results were verified using
the Python CANTERA model. A 40-50% increase in the nozzle orifice size from the traditional LPG
nozzle orifice size is suggested for such conversions.
Keywords: Synthetic Natural Gas, Computational Fluid Dynamics, Engineering Equation Solver (EES),
Combustion, Adiabatic Temperature, Equivalence ratio

1. Introduction

1.1. Background
The transition to cleaner household cooking energy sources in Nepal is a major challenge despite government
and stakeholder initiatives to promote clean cooking. Energy data from 2000 to 2018 shows that about 69% of
households still use solid fuels for cooking [1], which releases black carbon (BC), a potent short-term
climate-forcing agent, and contributes to respiratory diseases, particularly in women and children. Moreover,
such emissions contribute ~34% of global BC emissions [2]. This slow transition of different energy-intensive
sectors to clean fuels in Nepal is due to the high costs associated with clean technologies, which forces people
to use cheap fuel alternatives in various sectors.
Nevertheless, there is a gradual shift in the type of fuel consumption for domestic cooking purposes: from solid
fuels to LPG. According to the Nepal Oil Corporation (NOC), the country consumed 536,028 metric tons of
LPG gas during the fiscal year 2078/79 (2021-22 AD), representing an increase of approximately 11%
compared to the previous year. All the demands for LPG are met by imports from India, NRs 65.55 billion was
spent on importing cooking gas (LPG). Subsidies are being provided to bridge the gap between expensive fuel
prices and customer affordability. As a result of all this government is facing significant losses. Nepal has the
potential for biochemical and electrochemical methane production due to abundant hydropower for hydrogen
production. Reports suggest Nepal can produce 3,135,336 MT of hydrogen from surplus hydropower energy
and use it to generate 630,672 MT of synthetic natural gas annually. Producing this amount of SNG will
capture 1,734,348 MT of carbon dioxide from carbon-intensive industries which otherwise be released into the
atmosphere. [3–5]
Synthetic natural gas (SNG) is a gas consisting of >90% methane with a heating value of 35-55 MJ/m3
(adopted from [6]. Dai et al found that SNG with up to 95% methane concentration had good flame stability in
reference burners, but physical modifications are necessary for domestic cooking stoves in Nepal as the airflow
is difficult to modify. LPG gas in Nepal typically consists of 60% propane and 40% butane, and the
interchangeability of different fuel gases is determined by their Wobbe Index (WI) calculated by dividing the
gross heating value by the square root of the specific gravity as indicated in equation 1[6,7]
𝐺𝑟𝑜𝑠𝑠 𝐻𝑒𝑎𝑡𝑖𝑛𝑔 𝑉𝑎𝑙𝑢𝑒
Wobble Index (WI) = (1)
𝑆𝑝𝑒𝑐𝑖𝑓𝑖𝑐 𝐺𝑟𝑎𝑣𝑖𝑡𝑦

The Wobbe’s index of different gases is as shown in the table below:

Table 1. Wobbe’s Index of different gases. [7]
Family Type of gas Wobbe index range (MJ/Nm3) Wobbe number range
1 Town gas / Syngas 22.5 – 30.0 24.0 – 29.0
2L Natural 39.0 – 45.0
2H 45.5 – 55.0 48.0 – 53.0
3 LPG 73.5 – 87.5 72.0 – 87.0
LPG and SNG with 90% methane equivalence cannot be directly interchanged as LPG has a WI of around 85
MJ/m3, while SNG has a WI of 41 MJ/m3, as shown by Dai et al's experiment. SNG also has a lower density
than air, requiring a larger volume flow rate to produce the same amount of energy as LPG. This can be
achieved by expanding the region perpendicular to the flow or by altering the pressure [8]. CFD was suggested
as the most efficient tool to predict the mixing flow velocity and combustion of a burner. This study utilizes
CFD to determine the most economical and technically optimized condition for using SNG fuel in LPG
burners for domestic cooking. [9,10]

2. Methodology and Material:

Initially, a combustion model was created and was tested in a Bunsen burner to burn LPG. The accuracy of the
model was validated by comparing its results to an experimental paper [9]. To simplify the analysis and save
computational time, a two-step approach was employed. The first step is the cold flow which ensures
pre-mixing and the second step is combustion in the presence of secondary air. The validated combustion
model was then used to simulate a conventional household burner by attaching different inlet nozzle sizes
simultaneously. For this study, a medium size burner (popularly called in the local market as 2 No burner) is
used. The burner is made of brass material and the dimensions are shown in Figure 2. Furthermore, an EES and
CANTERA model was developed to find the optimum inlet pressure and validate final results at a given point
respectively. The different models and simulation steps used are tabulated below.
 Table 2. Solution method used during solving. [11] Table 3. Different models were used in the
simulation. [12,13]
Parameters Assigned scheme
Scheme Used Simple Model Cold Hot
Gradient Second order Flow Flow
Pressure Second order upwind Energy Equation On On
Momentum First Order upwind Turbulence model (k or ε +)
Turbulent Kinetic Energy First Order Upwind On On
Turbulent Dissipation Rate First Order Upwind Species Model (Species
On On
Fuel (Propane or Second order upwind Transport)
Methane) Eddy dissipation model Off On
CO2 Second order upwind (Volumetric response model)
H2O Second order upwind
Energy Second order upwind

3. Development and Validation of the Combustion Model

3.1 Validation of combustion model using Bunsen burner with Literature

The combustion model was developed in ANSYS and the accuracy was validated with an experimental
research paper [14]. This validation was done in the Bunsen burner, as all other domestic burners work on the
same principle. The simulation was performed in two steps. In the first step, cold flow (pre-mixing) was
simulated. The mixing chamber (as shown in Figure 1 a) was modeled in solid works and similar meshes from
previous studies had been tested for convergence. The boundary conditions were set as in Figure 1. The k-ϵ
model was used to account for turbulence [10,15], and the species transport model was turned on to ensure
fuel-air pre-mixing. Different models used for hot and cold flow are tabulated in Table 3. Finally, the CFD data
obtained from part 1 such as the mixing mass flow rates, species mass fraction, and velocities, were used as
input for combustion simulation, using boundary conditions (stated in the reference [14]) as shown in Figure
1(b). The final results of combustion were validated at 260 ml/min (C3H8: O2 0.21084:0.18171), considering
100% propane as LPG. Similar, temperature and flame structure trends were found as shown in Figure 2.

Figure 1 (a) Boundary condition for pre-mixing (b) Figure 2 Validation of combustion model
Boundary conditions for Combustion Zone. using LPG as fuel.

Upon observing the burner ends, it was noted that the temperature of the flame reduces due to the transfer of
heat to the cold walls of the burner. The maximum temperature was found to be at the edge of the inner core, as
the fuel-air mixture received an adequate amount of oxygen to undergo complete combustion. However, at the
center of the flame, the temperature decreases due to the rich mixture from the mixing chamber, which lacks
sufficient oxygen for combustion. The inexact match to the reference values is due to the use of 100% propane
as LPG. This combustion model along with different boundary conditions was used for the further simulation
of the conventional domestic burner.

3.2. Numerical Modeling of Domestic LPG Stoves

3.2.1. Geometry: The complete 3D burner model was prepared in solid works 2019 for the number 2 burner
head size. The geometry includes two parts [10,16]: part I for the premixing of the LPG and air and part II for
the combustion of the premixed fuel in presence of secondary air. The external diameter of the burner is 78 mm
and has 88 holes each of diameter 2 mm. The burner has a center hole for a secondary air intake of 36 mm. The
design was prepared by taking real measurements from a burner as shown in Figure 3.

Figure 3. CAD drawing of the burner head (units in mm) and 3D CAD modeling of the Burner Head

3.2.2. Meshing: The unstructured meshing was done in both parts I and part II taking appropriate inflations at
the air and fuel inlets. The mesh independence test was carried out till a mesh independent condition is
reached.

Figure 4 Mesh Independency for cold flow. Figure 5 Mesh Independency for hot flow.
For part 1: the test involved simulating cold flow with increasing element numbers (516108, 1013225,
1045872, 1115763, and 1276478) and observing the relative results (i.e., Mass fraction of fuel at outlet and
velocity at the outlet). The results of the test are displayed in Figure 4. Similarly, for the second part, another
mesh independence test was carried out on the hot flow. The study involved analyzing the flame temperature
and velocity at the inlet with an increasing number of elements. The results obtained from this analysis are
presented in Figure 5.
In order to reduce the computational time, the mesh with 1.04 M was selected for the cold flow and for the hot
flow the element size of 0.3 mm was selected for complete numerical simulations of the LPG burner at
different operating conditions.

3.2.3. Modeling Schemes and Boundary Conditions: The entire simulation was performed for two gases using
different sizes of inlet nozzle and 2.75 kPa inlet pressure. The simulations were performed first for cold flow
which ensures the premixing of the fuel with the entrained air. Then the results of the premixing at the pressure
outlet are simultaneously used as input for the combustion chamber. The convergence criteria of root mean
square residuals less than 1E-6 were used in convergence control. The detailed modeling scheme and boundary
conditions are tabulated in table 4 and 5 for cold and hot flow respectively.
Table 4 Modeling Scheme and Boundary Condition for cold and hot flow.
Boundary Condition Cold Flow Modeling Scheme Hot Flow Modeling Scheme
Fuel inlet (Inlet condition) Pressure Inlet 2.75 kPa (from EES) Velocity inlet (from part 1)
Air Inlet Pressure inlet 0 Pa (surrounding) Pressure inlet 0 Pa (surrounding)
Outlet Pressure Outlet (Air Gauge Pressure 0 Pa) Pressure Outlet (Gauge pressure 0 pa)
Near wall treatment Adiabatic wall (300 300 ̊K) Adiabatic wall (300 300 ̊K)
Turbulent Model RNG k-ϵ Model Standard k-ϵ Model
Fuel Mass Fraction 1 (Methane or Propane) From cold flow output
Oxygen mass fraction 0.23 (% in the air) 0.23 (% in the air)
Solver Pressure based Pressure based
Combustion model Off Eddy Dissipation Model

Figure 6 Boundary Condition of

Figure 7 Domain of Figure 8 Boundary conditions for
Mixing Chamber
combustion zone combustion domain
The combustion domain was cut into 12 degrees to save computational time and power. Then the boundary
conditions as shown in Figures 6 and 8 are used for further simulation.
3.2.4 EES modeling for optimum inlet pressure: The EES (engineering equation solver) model was developed
to find the relation between inlet pressure and adiabatic flame temperature for LPG and SNG. The results of
the EES simulations are presented in the next section.

4. Results and Discussion:

4.1 Effect of SNG and LPG release pressure on temperature

Figure 9 shows the relation between inlet pressure and adiabatic flame temperature of LPG and SNG. The
analysis was done by creating an Engineering Equation solver (EES) model from the EES library. The flame
temperature shows a steep rise when the inlet pressure is increased from 1 kPa to 3 kPa but after 3 kPa the rise
in flame temperature is not much significant in comparison to the increase in pressure.

Figure 9. Dependencies of LPG flame temperature on inlet Pressure.

4.2 Flame and temperature distribution around the pan.

Figure 10 and 11 compare the flame temperatures of combustion using a 0.5 mm nozzle size with LPG and
SNG. LPG and SNG obtain an average flame temperature of ~1150 ̊K and ~700 ̊K respectively around the pan.
The maximum flame temperature and velocity of the fuel-air mixture are near the inlet holes. SNG being less
dense than LPG, less mass at low velocity (3.9 m/s) enters through the inlet nozzle at 2.75 kPa pressure which
entrains less air through the air inlet ports in the mixing chamber. This mixture entering the combustion
chamber starts burning from the burner wall resulting in high temperature near the burner holes but as the
reaction proceeds, the secondary air from the surrounding makes the mixture further lean resulting in lower
temperature afterward. Hence more fuel mass is needed to carry the uniform high temperature near the pan
wall. It can be pointed out that simulated results are reasonable by comparing them with the temperature
distribution behaviors of the real burner. For example, at the farthest point from the combustion zone near the
pressure outlet (at the top), the temperature is relatively lower.
For the nozzle size of 0.75 mm Figure 12, the velocity of the inlet fuel in the mixing chamber increases up to
13 m/s, near the nozzle which not only ensures a good amount of fuel mass (SNG) to the mixing chamber but
also a decent amount of entrained air due to low-pressure regions near the air inlet ports. But due to high inlet
velocity, the flame jumps off the burner and the pan walls which may lead to inefficiencies.
 Figure 10 Flame structure for LPG Combustion Figure 11 Flame structure for LPG Combustion
@0.5 mm @0.5 mm

In Figure 12 for the 0.75 mm nozzle, despite good flame temperature the flame jump can be seen near the bend
of the pan due to high velocity. In practical applications, this jump can cause flame instability and thermal
losses.
Figure 13 shows the simulation result for a 1 mm nozzle size, due to the moderate fuel inlet velocity, a good
amount of air is entrained in the mixing chamber which shows flame stability and uniform temperature (~1350
̊K) near the pan walls.

Figure 12 Flame structure for SNG combustion Figure 13 Flame structure for SNG combustion
around pan wall 0.75 mm around pan wall 1 mm

As the size of the nozzle is increased to 1.15mm the velocity nearly remains constant but the flame stability
enhanced as in figure 14 and more notably a uniform high temperature (~1700 K ̊ ) is developed near the pan
walls as can be seen in figure 17. The comparison between temperature distribution near the walls of the pan
for 0.75 mm and 1.15 mm is presented in Figures 16 and 17.
 Figure 14 Flame structure for SNG combustion Figure 15 Flame structure for SNG combustion
around pan wall 1.15 mm around pan wall 1.35 mm

Figure 16 Temperature distribution for SNG Figure 17 Temperature distribution for SNG
combustion 0.75 mm (not uniform) combustion 1.15 mm (uniform)
In Figure 18, a summary of the changes in temperature and velocity with nozzle sizes is provided. It
was observed that increasing nozzle size result in higher and more uniform temperature, which is
beneficial for cooking. The combustion performance improved from a nozzle size of 1 mm (or about
40% larger than the conventional LPG nozzle setting) and continued to improve up to a nozzle size of
1.15 mm (approximately 50% larger than the conventional nozzle setting). However, beyond this
point, an increase in velocity led to higher turbulence in the air-fuel mixture than necessary, resulting
in flame instability and flame jumping.
 Figure 18. Dependency of average temperature and inlet velocity on change in nozzle size.
To find out the completeness of the combustion a monitor was set at the pressure outlet of the combustion
chamber where any unburnt methane escaping out was observed. The analysis was done at three different
positions in the flames streamline, i.e., outside, inside, and midpoint. It was observed that the SNG combustion
up to 1.15 mm nozzle size is complete but for 1.35 mm nozzle size mass fraction of unburnt CH4 was 0.001 as
tabulated in table 5. Methane is a greenhouse gas that should not be released directly into the surrounding.
Table 5. Species Residual test at pressure outlet.
Nozzle Mass Fraction Mass Fraction Combustion status at the outlet
Size (mm) (CH4) at the outlet (CO2) at the outlet
0.5 0 0.02 Complete Combustion
Complete Combustion with good temperature
0.75 0 0.09
around the wall.
1 ~0 0.07 Complete Combustion
1.15 ~0 ~0.15 Complete Combustion (good temperature)
1.35 0.001 0.125 Incomplete Combustion at the outlet

4.3 Validation of the simulation results

Since the equivalence ratio plays a vital role in the reaction process. The equivalence ratio is calculated at a
defined point below the pan wall in Ansys as shown in Figure 19. The fuel–air equivalence ratio, ϕ (phi), of a
system, is defined as the ratio of the fuel-to-oxidizer ratio to the stoichiometric fuel-to-oxidizer ratio as
represented by equation 2 [15]. Mathematically for 1.15 mm at a point shown in Figure 19,

𝛟=
(𝑓𝑢𝑒𝑙−𝑡𝑜−𝑜𝑥𝑖𝑑𝑖𝑧𝑒𝑟 𝑟𝑎𝑡𝑖𝑜) ( )𝑒𝑥
𝐹
𝐴 0.18
= 0. 6428
= = (2)
(𝑓𝑢𝑒𝑙−𝑡𝑜−𝑜𝑥𝑖𝑑𝑖𝑧𝑒𝑟 𝑟𝑎𝑡𝑖𝑜)
𝑠𝑡𝑜𝑖𝑐 ( )𝑠𝑡𝑜𝑖𝑐
𝐴
𝐹 0.28

The equivalence ratio of the 1.15 mm nozzle size at the point near the pan wall was analytically calculated to
be ϕ=0.6428 at 2.75 kPa pressure. This calculation is cross verified from the literature [17]

4.3.1 Modeling of SNG and LPG Combustion

A CANTERA modeling was done using references [18] to study the relation between adiabatic flame
temperatures, and equivalence ratio for SNG. Following results were obtained as shown in figure 20.
 Figure 19 Calculation of the Equivalence ratio Figure 20 validation of the temperature at the
(1.15 mm nozzle size) at the point marked equivalence ratio using CANTERA
It is seen that as the equivalence ratio increases from 0.5 to 1 the temperature rises to the maximum but with
further increase in excess air the flame temperature again starts to decrease due to the cooling effect of excess
air. The temperature corresponding to the 64% equivalence ratio is found to be ~1750 ̊K, which was found in
sync with the ANSYS simulation results.

5. Results and Discussion:

The purpose of this research was to replicate the flow and combustion behaviors of LPG burners using SNG as
a replacement for LPG. Initially, it was found that the LPG and SNG are non-interchangeable due to different
Wobbe indexes. After studying the flame structure of free-flowing flame it was found that the highest
temperature occurs at the edge of the inner core due to adequate oxygen availability. During the simulation,
different flame temperatures and flame structures were seen at different nozzle sizes. The stability of the flame
depends widely on the velocity of the fuel and air mixture entering the combustion chamber. Changing the
orifice size affects turbulent intensity, which, in turn, affects the amount of secondary air entrained, leading to
different combustion patterns and distributions. As the nozzle size approaches 1.15 mm, turbulence due to
velocity decreases, resulting in uniform complete combustion near pan walls. Hence it can be concluded
that increasing the inlet nozzle size by 40-50% from conventional LPG orifice nozzle size provides
good flame structure and good temperature distribution for using SNG at an unchanged inlet pressure.

6. References:
[1] Paudel D, Jeuland M and Lohani S P 2021 Cooking-energy transition in Nepal: Trend review Clean Energy
5 1–9
[2] Dang J, Li C, Li J, Dang A, Zhang Q, Chen P, Kang S and Dunn-Rankin D 2019 Emissions from solid fuel
cook stoves in the Himalayan Region Energies (Basel) 12
[3] Fendt S, Buttler A, Gaderer M and Spliethoff H 2016 Comparison of synthetic natural gas production
pathways for the storage of renewable energy Wiley Interdiscip Rev Energy Environ 5 327–50
[4] Thapa B S, Neupane B, Yang H seong and Lee Y H 2021 Green hydrogen potentials from surplus hydro
energy in Nepal Int J Hydrogen Energy 46 22256–67
[5] Becker W L, Penev M and Braun R J 2019 Production of synthetic natural gas from carbon dioxide and
renewably generated hydrogen: A techno-economic analysis of a power-to-gas strategy Journal of Energy
Resources Technology, Transactions of the ASME 141
[6] Raza Shah S A, Ahmed Z, Leghari B A, Laghari W M and Khidrani A 2015 Generation of Electric Power
from Domestic Cooking System Open Journal of Energy Efficiency 04 69–76
[7] Dai W, Qin C, Chen Z, Tong C and Liu P 2012 Experimental studies of flame stability limits of
biogas flame Energy Conversion and Management vol 63 pp 157–61
[8] Ibraheem A and Szodrai F 2019 Numerical Model Analysis of Natural Gas Combustion Burners
International Journal of Engineering and Management Sciences (IJEMS) 4
[9] Suwansri S, Moran J C, Aggarangsi P, Tippayawong N, Bunkham A and Rerkkriangkrai P 2015
Converting LPG stoves to use biomethane Distributed Generation and Alternative Energy Journal 30 38–
57
[10] Wichangarm M, Matthujak A, Sriveerakul T, Sucharitpwatskul S and Phongthanapanich S 2020
Investigation on thermal efficiency of LPG cooking burner using computational fluid dynamics Energy
203
[11] Jangala S S L, Kakumani V S P, Vunnam N S and Sreekanth P S R 2022 A novel brass metal burner
design for conventional LPG stove with convergent holes and swirl flow Mater Today Proc 56 1068–74
[12] P. Aroonjarattham 2016 The Parametric Study of High-Pressure Gas Burner Affect Thermal Efficiency Eng. J
20 33–48
[13] Ogedengbe E O B and Ajibade F D 2017 Improved Burner Efficiency and Fuel Consumption in
Domestic Cooking Appliances Energy and Policy Research 4 29–35
[14] Aruna Devadiga and Prof. Dr. T. Nageswara Rao 2013 Optimizing Bunsen burner Performance
Using CFD Analysis. International Journal of Modern Engineering Research (IJMER) 3 2773–85
[15] El-Mahallawy F, Abdelhafez A and Mansour M S 2007 Mixing and nozzle geometry effects on flame
structure and stability Combustion Science and Technology vol 179 pp 249–63
[16] Datta A, Das M and Ganguly R 2021 Design, Development, and Technological Advancements in Gas
Burners for Domestic Cook Stoves: A Review Transactions of the Indian National Academy of
Engineering 6 569–93
[17] Ko Y C and Lin T H 2003 Emissions and efficiency of a domestic gas stove burning natural gases with
various compositions. Energy Convers Manag 44 3001–14
[18] Abou-Taouk A, Farcy B, Domingo P, Vervisch L, Sadasivuni S and Eriksson L E 2016 Optimized reduced
chemistry and molecular transport for large eddy simulation of partially premixed combustion in a gas
turbine. Combustion Science and Technology18821–39

7. Acknowledgement:

This research was supported by Kathmandu University- Integrated Rural Development Program/Nepal
Technology Innovation Center (KU-IRDP/NTIC) grant funded by Korea International Cooperation Agency
(KOICA). The authors would also like to thank Dr. Bivek Baral and Dr.Biraj Singh Thapa for providing
guidance throughout the research.