UNDERGROUND COAL GASIFICATION MODELING

PROJECT REPORT

MATHEMATICAL MODELING OF UCG

Submitted by

AASHISH CHANDRA (CH20M002)

MASTER OF TECHNOLOGY

Under the guidance of

Prof. PREETI AGHALAYAM

DEPARTMENT OF CHEMICAL ENGINEERING

INDIAN INSTITUTE OF TECHNOLOGY MADRAS

CHENNAI 600036

AUGUST-JANUARY 2022

1|Page
 TABLE OF CONTENTS

CONTENTS PAGE NO

Introduction 3

The analytical study 3

Equation involved in this study 3

Mathematical analysis methods 4

Two -dimensional geometry 4

Meshing 4

Solution model 4

Results 5

Results of simulation attached in this file

2|Page
INTRODUCTION
Underground Coal Gasification is a non-
traditional, in-situ combustion process for
converting coal into product gases. In this
process coal is combusted and the produced
syngas which basically contains CO2, H2, CO
and CH4 is extracted to the surface with the help
of drilled wells.
In this study, a 2D mathematical model which
was dynamic in order and based on kinetic
reaction rates was created and investigated in
ANSYS FLUENT to evaluate numerical
results. The mathematical model prepared in
the Geometry Modeler of the ANSYS
Workbench consists of the inlet and outlet
streams and a coal reactor part where the
reactions take place. There is a total of 6 species
involved in this work mentioned as O2, CO2,
CO, H2O, H2 and N2. For these six chemical
species there are six chemical reactions
involved from reaction 3 to 8 as shown below:
Coal  Coal (dry) + H2O …… (1)
Coal (dry)  Volatile Matter + Char …… (2)
CO + ½ O2  CO2 ……. (3)
H2 + ½ O2  H2O …. (4)
CO + H2O = CO2 + H2 ……. (5)
C + O2  CO2 ……… (6)
C + CO2  2 CO ………… (7)
C + H2O  CO + H2 ...... (8)
1. THE ANALYTICAL STUDY
A two-dimensional mathematical model along
with its solution geometry having the
dimensions of an UCG reactor was developed
using the ANSYS WORKBENCH and ANSYS
FLUENT software environment. Developing a
Mathematical model in two dimensions and
working on it for its solution geometry is less
complex and sufficient to analyse the effects of
gasification parameters.
A 2D model is first created using the ANSYS
Workbench and then a grid is created using
ANSYS Meshing. The Computational Fluid
Dynamics model created using the software
environment is used to investigate the thermal
and chemical heat and mass transfer
interactions between the source and parameters.
2. EQUATIONS INVOLVED IN THIS
STUDY
+ div (ρ 𝑢⃗ ) = Sm ………. (1)
( ⃗) ⃗
+ div (ρ 𝑢𝑢)⃗ = div (μ grad 𝑢⃗) + Su …. (2)
( )
+ div (ρ e 𝑢⃗ ) = div (⋋ grad T- P 𝑢⃗) + Se
…… (3)
( )
+ div (𝜌 𝐶 𝑢⃗ ) = div (𝜌 𝐷 grad 𝐶 ) +
Sα …... (4)
Equation 1 = Continuity equation
Equation 2 = Momentum equation
Equation 3 = Energy equation
Equation 4 = species transport equation
Where:
ρ = Density of fluid, Kg/m3
P = Pressure (Pa)
𝑢⃗ = Velocity vector of fluid element, m/s
μ = Viscosity of fluid, Pa. s
e = total energy related to unit mass of fluid,
Kj/Kg
⋋ = Thermal conductivity, W/Mk
𝐷 = Diffusion coefficient, m2/s
T = Temperature, K
𝐶 = Concentration of species α in mixture,
kmol/m3
3|Page
Sm, Su, Se, Sα are the source term in the above
equations, which is associated with mass,
momentum, energy and transport of species.
3. MATHEMATICAL ANALYSIS
METHODS
All the above equations are Partial Differential
Equations which are usually referred to as
PDE’s and the solution or analysis of PDE in
using mathematical analysis can be done using
two approaches which are:
1. Deterministic approaches
2. Stochastic approaches
A deterministic approach is one in which the
output is completely dependent on the input
parameters whereas the stochastic method is the
one in which the results are based on principle
of statistics and the output values may not be
the same for each input value. The three main
methods to determine the results over numerical
analysis are:
3.1 Finite difference method:
PDE to be satisfied at nodes.
3.2 Finite volume method:
PDE to be satisfied over a set of controlled
volumes which are referred to as cells.
3.3 Finite element method:
The third or the final method which involves the
use of test functions and the definite integration
of the integration for boundary limits that
define the whole domain.
The method is used in our study for numerical
analysis of the 2 D model is Finite Element
Analysis.
4. TWO- DIMENSIONAL GEOMETRY
A 2D solution geometry was prepared using the
ANSYS Workbench software and the geometry
included three different sections viz. inlet,
outlet and the reactor. The dimensions of all the
sections were predefined in the ‘Dimension’
section of the modelling part of the Workbench.
4.1 MESHING
The next step after the successful formation of
the two-dimensional geometry on the ANSYS
Design Modeler is Meshing. As discussed
above in the various methods of numerical
analysis, the calculations are made at the
interconnected points known as nodes and the
overall geometry obtained along with these
nodes is called a mesh. The number of nodes
and elements in a specific geometry can be set
which is visible in the Statistics section of the
Details window. The number of nodes can be
changed by changing the sizing values in
accordance with the actual dimensions of the
two-dimensional geometry. The different
sections of the geometry need to be assigned
with a particular name in order to define the
calculations. The named sections in our study
are as follows:
1. Inlet
2. Walls
3. Base
4. Outlet
Generating mesh is the next step and this
completes the first half of our study.
4.2 SOLUTION MODEL
One of the main steps in the numerical analysis
of the two-dimensional geometry representing
Underground Coal Combustion is Initialization
& Calculations. These calculations in our study
were done on ANSYS FLUENT (Fluid Flow
(FLUENT) Parallel Fluent) and the models
used in the setup are as follows:
1. Energy model
2. Viscous model (k-epsilon)
3. Radiation model
4. species Transport model
4|Page
5. Discrete phase model value Absolute Convergence
criteria status
The materials used in this study were the
Continuity 0.000997124 0.001 Yes
common combustion materials including coal
x-velocity 0.0003259046 0.001 Yes
and volatile air.
y-velocity 0.0001966869 0.001 Yes
We set boundary condition and initial condition Energy 6.64256e-08 1e-06 Yes
and the we do iteration to obtained results. K 0.000324398 0.001 Yes
5. RESULTS Epsilon 0.0003963062 0.001 Yes
Vol 0.0001663359 0.001 Yes
The graph shows the simulation of the UCG 2D O2 0.0001663359 0.001 Yes
model for a set of 300 iterations. The number of CO2 0.0001663358 0.001 Yes
iterations can be varied as and when a stability H2O 0.0001663771 0.001 Yes
in the results is visible which in this graph can CO 0.0001663359 0.001 Yes
be seen. The main approach behind this study is P1 3.443843e-08 1e-06 Yes
to satisfy a set of Partial Differential Equations
which are mainly the:

1. Energy equation

2. Momentum equation

3. Continuity equation

4. Species transport equation

All the above equations are based on and find

applications in Computational Fluid Dynamics.
Apart from the Scaled Residual Graph that
shows the variation of all the residuals with the
iterations, a set of Contour graphics can also be
obtained with the simulation. The different
contour plots are:

1. Mass imbalance

2. Discrete phase variables

3. Velocity

4. Turbulent kinetic energy

5. Strain rate

6. Mesh

All the results including graphs and contour

are attached in this file.

I got convergence at 309 iterations.

5|Page
6|Page
PREDICTION OF CAVITY GROWTH RATE DURING UNDERGROUND COAL
GASIFICATION USING REGRESSION ANALYSIS:
During underground coal gasification, whereby coal is converted to syngas in-situ, a cavity is
formed in the coal seam. The cavity growth rate (CGR) or the moving rate of the gasification
face is affected by controllable factor like operation pressure, gasification time, geometry of
UCG panel and uncontrollable factor like coal seam properties. The CGR is usually predicted
by mathematical models and laboratory experiments, which are time consuming analysis and
expensive.
So, a group of researchers have developed a new simple model for CGR by using non-linear
regression analysis.
NON-LINEAR REGRESSION ANALYSIS:
Non-linear regression is a method for building a non-linear model of relationship between the
dependent variable and a set of independent variables. Unlike traditional linear regression,
non-linear regression can estimate a model with arbitrary relationships between dependent
and independent variables. The researches develop two non-linear regression equations were
developed using SPSS (statistical software).
EQUATION 1:
𝟐𝟖.𝟎𝟕𝟔
CGR = 0.076-0.32* log (CA) + 2.181* (CM) – 215.102* 𝒉 + 0.0000825*(OP) –
399.834* (𝑪𝑷)𝟏𝟗𝟖.𝟗𝟓
R^2= 0.79 (Regression coefficient)
EQUATION 2:
𝑪𝑪𝑽 𝟎.𝟓
𝟎.𝟎𝟎𝟒∗ 𝑫 ∗(𝑶𝑷)𝟎.𝟐𝟔𝟓 ∗(𝑪𝑴)^𝟎.𝟏𝟖𝟓
CGR= , R^2 = 0.85
𝑪𝑽𝑴∗(𝑪𝑨)𝟎.𝟐𝟓 ∗(𝑪𝑷)𝟎.𝟏𝟓𝟗

Where,
CGR= Cavity growth rate (m/day)
OP = Operating pressure (KPa)
CP = Coal permeability (Darcy)
h = Coal seam thickness (m)
CA = Coal ash (0.01 %)
CM = Coal moisture content (0.01 %)
CCV = Coal calorific value (kCal/kg)
D = Depth of coal seam (m)
CVM = Coal volatile matter (0.01 %)
Equation 2 will give good result as compared to equation 1as regression coefficient of
equation 2 is high as compared to equation 2

1|Page
VALIDATION OF ABOVE EQUATION:
Due to high regression coefficient of equation 2, we are using equation 2
Results from field trails
UCG site Seam depth (m) Operating pressure CGR (m/day)
(kPa)
El Tremedal, Spain 580 5000 1.2
RM1, USA 130 500 0.65
Secunda, South 160 140 0.45
Africa
PSC, USA 60 430 0.82
Chinchilla, Australia 140 1100 0.5
Hanna I, USA 80 150 0.7
Mecsek Hills, 600 5100 0.5
Hungary

RESULT FROM THE EQUATION:

For Tremedal, Spain
OPERATING PRESSURE SEAM DEPTH CGR
5000 580 1.259
4000 250 1.807
2000 1000 0.752
7000 80 3.7069

Error = 4.68%
For RM1, USA
OPERATING PRESSURE SEAM DEPTH CGR
500 130 0.6586
600 200 0.5572
700 250 0.5192
101.325 1000 0.1555

Error = 1.305%

According to literature appreciable error is 15%

2|Page
CONCLUSIONS:
We can use the equation 2 for the prediction of Cavity growth rate as we are getting error
when we compare with the result of field trials vs predicted result, we are getting error less
than 15% which acceptable according to literature.

3|Page
 Ansys Fluent Simulation Report

Analyst AASHISH

Date 3/14/2022 06:35 AM

Table of Contents
1 System Information
2 Geometry and Mesh
2.1 Mesh Size
2.2 Mesh Quality
2.3 Orthogonal Quality
3 Simulation Setup
3.1 Physics
3.1.1 Models
3.1.2 Material Properties
3.1.3 Cell Zone Conditions
3.1.4 Boundary Conditions
3.1.5 Reference Values
3.2 Solver Settings
4 Run Information
5 Solution Status
6 Plots
7 Mesh
8 Contours

System Information

Application Fluent

Settings 2d, pressure-based, species, standard k-epsilon

Version 21.2.0-10201

Source Revision fcb749f05e

Build Time May 28 2021 13:55:16 EDT

CPU AMD Ryzen 5 5500U with Radeon Graphics

OS Windows

Geometry and Mesh

Mesh Size

Cells Faces Nodes

6402 13001 6600

Mesh Quality

Name Type Min Orthogonal Quality Max Aspect Ratio

ucg_reactor Mixed Cell 0.56430513 4.9644227

Orthogonal Quality

Simulation Setup

Physics

Models

Model Settings

Space 2D

Time Steady

Viscous Standard k-epsilon turbulence model

Wall Treatment Standard Wall Functions

Heat Transfer Enabled

Radiation P1 Model
 Model Settings

Species Reacting

Coupled Dispersed Phase Enabled

Material Properties
 Fluid

 air

Density 1.225 kg/m^3

Cp (Specific Heat) 1006.43 J/(kg K)

Thermal Conductivity 0.0242 W/(m K)

Viscosity 1.7894e-05 kg/(m s)

Molecular Weight 28.966 kg/kmol

Standard State Enthalpy 0

Standard State Entropy 194336 J/(kgmol K)

Reference Temperature 298.15 K

Absorption Coefficient 0

Scattering Coefficient 0

Scattering Phase Function isotropic

Thermal Expansion Coefficient 0

Refractive Index 1

Speed of Sound none

 Solid

 aluminum

Density 2719 kg/m^3

Cp (Specific Heat) 871 J/(kg K)

Thermal Conductivity 202.4 W/(m K)

 Mixture

 coal-volatiles-air

Mixture Species names

Reaction finite rate/eddy dissipation

Mechanism reaction mechs

Density incompressible ideal gas

Cp (Specific Heat) mixing law

Thermal Conductivity 0.0454 W/(m K)

Viscosity 1.72e-05 kg/(m s)

Mass Diffusivity constant dilute appx

Absorption Coefficient 0

Scattering Coefficient 0

Scattering Phase Function isotropic

Refractive Index 1

Speed of Sound none

Cell Zone Conditions

 Fluid

 ucg_reactor
 Material Name coal volatiles air

Specify source terms? no

Specify fixed values? no

Frame Motion? no

Participates in radiation yes

Laminar zone? no

Porous zone? no

Reaction Mechanism mechanism 1

Activate reaction mechanisms? yes

Surface-Volume-Ratio [m^-1] 0

Boundary Conditions
 Inlet

 oxidant_inlet

Reference Frame Absolute

Mass Flow Specification Method Mass Flow Rate

Mass Flow Rate [kg/s] 5

Total Temperature [K] 400

Supersonic/Initial Gauge Pressure [Pa] 0

Direction Specification Method Direction Vector

Component of Flow Direction (x,y) (1, 0)

Turbulent Specification Method Intensity and Viscosity Ratio

Turbulent Intensity [%] 5

Turbulent Viscosity Ratio 10

Specify Species in Mole Fractions? no

External Black Body Temperature Method Boundary Temperature

Internal Emissivity 1

 Outlet

 product_outlet

Backflow Reference Frame Absolute

Gauge Pressure [Pa] 0

Pressure Profile Multiplier 1

Backflow Total Temperature [K] 300

Backflow Direction Specification Method Normal to Boundary

Turbulent Specification Method Intensity and Viscosity Ratio

Backflow Turbulent Intensity [%] 5

Backflow Turbulent Viscosity Ratio 10

External Black Body Temperature Method Boundary Temperature

Internal Emissivity 1

Specify Species in Mole Fractions? no

 Backflow

constant False

constant False

constant False

constant False
 constant False

Backflow Pressure Specification Total Pressure

Build artificial walls to prevent reverse flow? no

Average Pressure Specification? no

Specify targeted mass flow rate no

 Wall

 wall

Wall Thickness [m] 0

Heat Generation Rate [W/m^3] 0

Material Name aluminum

Thermal BC Type Heat Flux

Heat Flux [W/m^2] 0

Wall Motion Stationary Wall

Shear Boundary Condition No Slip

Internal Emissivity 1

Wall Roughness Height [m] 0

Wall Roughness Constant 0.5

Species Boundary Conditions Zero Diffusive Flux/Zero Diffusive Flux

Convective Augmentation Factor 1

 base

Wall Thickness [m] 0

Heat Generation Rate [W/m^3] 0

Material Name aluminum

Thermal BC Type Heat Flux

Heat Flux [W/m^2] 0

Wall Motion Stationary Wall

Shear Boundary Condition No Slip

Internal Emissivity 1

Wall Roughness Height [m] 0

Wall Roughness Constant 0.5

Species Boundary Conditions Zero Diffusive Flux/Zero Diffusive Flux

Convective Augmentation Factor 1

Reference Values

Area 1 m^2

Density 1.225 kg/m^3

Depth 1m

Enthalpy 0 J/kg

Length 1m

Pressure 0 Pa

Temperature 288.16 K

Velocity 1 m/s

Viscosity 1.7894e-05 kg/(m s)

Ratio of Specific Heats 1.4

 Yplus for Heat Tran. Coef. 300

Reference Zone ucg_reactor

Solver Settings
 Equations

Flow True

Turbulence True

vol True

o2 True

co2 True

h2o True

co True

Energy True

P1 True

 Numerics

Absolute Velocity Formulation True

 Under-Relaxation Factors

Pressure 0.3

Density 1

Body Forces 1

Momentum 0.7

Turbulent Kinetic Energy 0.8

Turbulent Dissipation Rate 0.8

Turbulent Viscosity 1

vol 1

o2 1

co2 1

h2o 1

co 1

Energy 1

P1 1

Discrete Phase Sources 0.5

 Pressure-Velocity Coupling

Type SIMPLE

 Discretization Scheme

Pressure Second Order

Momentum Second Order Upwind

Turbulent Kinetic Energy First Order Upwind

Turbulent Dissipation Rate First Order Upwind

vol Second Order Upwind

o2 Second Order Upwind

co2 Second Order Upwind

h2o Second Order Upwind

co Second Order Upwind

 Energy Second Order Upwind

 Solution Limits

Minimum Absolute Pressure [Pa] 1

Maximum Absolute Pressure [Pa] 5e+10

Minimum Temperature [K] 1

Maximum Temperature [K] 5000

Minimum Turb. Kinetic Energy [m^2/s^2] 1e-14

Minimum Turb. Dissipation Rate [m^2/s^3] 1e-20

Maximum Turb. Viscosity Ratio 100000

Run Information

Number of Machines 1

Number of Cores 1

Case Read 14.231 seconds

Iteration 36.314 seconds

AMG 22.092 seconds

Virtual Current Memory 0.220024 GB

Virtual Peak Memory 0.227177 GB

Memory Per M Cell 11.689

Solution Status

Iterations: 309

Value Absolute Criteria Convergence Status

continuity 0.000997124 0.001 Converged

x-velocity 0.0003259046 0.001 Converged

y-velocity 0.0001966869 0.001 Converged

energy 6.64256e-08 1e-06 Converged

k 0.000324398 0.001 Converged

epsilon 0.0003963062 0.001 Converged

vol 0.0001663359 0.001 Converged

o2 0.0001663359 0.001 Converged

co2 0.0001663358 0.001 Converged

h2o 0.0001663771 0.001 Converged

co 0.0001663359 0.001 Converged

p1 3.443843e-08 1e-06 Converged

Plots

Residuals

Residuals

continuity
x-velocity
1.00e+0
y-velocity
energy
k
1.00e−2
epsilon
residuals

vol
o2
1.00e−4
co2
h2o
co
1.00e−6
p1

0 50 100 150 200 250 300

iterations

Mesh

mesh-1

Contours
contour-9

contour-8

contour-7
contour-6

contour-5
contour-4

contour-3
contour-2

contour-1