Study on the model experiment and numerical simulation

for underground coal gasification

Lanhe Yang*

College of Mineral Resources and Geosciences, China University of Mining and Technology, Xuzhou, Jiangsu Province 221008, China

Received 22 August 2002; accepted 4 August 2003; available online 21 September 2003

Abstract

The gas production process in underground coal gasification is closely linked to the temperature distribution and seepage conditions of the

gasifier. In this paper, mathematical models on the underground coal gasification in steep coal seams are established according to their

storage conditions and features of gas production process. Additionally, the paper introduces ways to determine model parameters and the

control volume method to solve the model. On the basis of the model experiment, the calculation results are analyzed. From the distribution

of temperature field, the calculation value is slightly higher than measurement one. The relative errors of all measuring points are virtually

within 10%. According to the simulated calculation results, with the increase of the length for the gasification channel, the heating value of

the gas improves. However, behind the reduction zone, it increases with a smaller margin. The influence of the temperature field on the

heating value for the gas is noticeable. Due to the effect of temperature, in high temperature zone, the change gradient of the measured value

of the concentration field for the gas compositions is larger than that of calculated value. The simulated results of the pressure field in the

gasification panel 22 and 75 h after the ignition indicate that the relative calculation errors are 5.00–14.29 and 9.68–17.24%, respectively,

with a corresponding pressure drop rate of 9.5 and 11.97%. On the whole, the simulation value conforms with the experiment value, which

supports the numerical simulation on the temperature field, concentration field and pressure field of underground coal gasification in the steep

coal seams is correct. Thus, it lays a necessary and scientific theoretical foundation for further quantitatively studying the process of

underground coal gasification and forecasting the change patterns.

q 2004 Elsevier Ltd. All rights reserved.

Keywords: Underground coal gasification; Steep coal seam; Temperature field; Concentration field; Pressure field; Mathematical model

1. Introduction

Thanks to the basic theoretical research on the technique

of underground coal gasification and its increasing appli-

cation and promotion as well as the joint efforts made by the

workmates home and abroad, great progress has been made

in the technique of this country and it is developing into a

new discipline. Because the gas production process

completely depends on the temperature conditions and

convection conditions of gasifier, the research of tempera-

ture field, concentration field and pressure field has always

been the focus of the researchers in the profession. As of the

beginning of 1980s, a large number of experimental and

theoretical studies on the technique of underground coal

gasification had been made in many countries such as the

Former Soviet Union, Belgium, France, Germany and

the United States, etc. [1–5], resulting in great progress in

this field, which contributes a lot to the popularization and

application of the underground coal gasification. While, for

the last two decades, except China, all other countries in the

world has discontinued the research on this field. Over the

past decades, research on the gasification of different coal

seams, ranging from model experiment, field experiment to

numerical simulation [6–8], was made. Large number of

data and a lot of experiences are accumulated [9–12].

According to the data, the storage conditions of coal seams

in this country are diversified. Steep coal seams account for

8.3% [13]. In old and abandoned coalmines, high-angle coal

seams are becoming more and more [14,15]. In order to

further study the regularities of the gasification process in

the steep coal seams, the mathematical models on the

temperature field, concentration field and pressure field of

the underground gasification, are established. The numerical

solution is found. Through the model experiment,

0016-2361/$ - see front matter q 2004 Elsevier Ltd. All rights reserved.

doi:10.1016/j.fuel.2003.08.011

Fuel 83 (2004) 573–584

www.fuelfirst.com

* Tel.: þ86-516-3885-762; fax: þ86-516-3888-682.

E-mail address: lhyang2053@sina.com (L. Yang).

the reasonability of the mathematical models is verified,

which provides an important theoretical basis for the

quantitative analysis and scientific guidance of the process

of underground coal gasification.

2. The features of the underground gasification

in the steep coal seams

The underground gasification in the steep coal seams

eminently features the following three aspects:

(1) The thickness of steep coal seams is much smaller

than its slant height and strike length; therefore, the

temperature conduction and retort gas movement in

the coal seams can be regarded as a matter of 2D

plane. Meanwhile, the heat absorption of aqueous

phase change and coal pyrolysis, together with the

heat losses in peripheral rocks, can be held up as

different minus source items in heat transfer

equations. Likewise, the height and width of the

gasification channel are far smaller than its

length, its field model can be thought of as 1D

matter, as shown in Fig. 1. The contribution of the

output of retort gas in the coal seams to the gas

compositions in the channel can be thought of as

source items.

(2) With the consumption of coal, due to the dead weight

and the effect of protruding pressure of retort gases,

the unburned coal falls onto the gasification panel,

which leads to a percolation-patterned channel with

comparatively big porosity [16]. Assuming the

porosity is the same everywhere.

(3) Because the coal seams are thin, the downside

and upside of the reaction zone can be thought of

as floor and roof. There is lime-ash layer under the

reaction zone. Therefore, there is material flow only

at the upper boundary.

According to the features mentioned above, the simpli-

fication of the control equation set of the gasification

process in percolation-patterned channel and the model

equation set of dry and pyrolytic model gives the

mathematical models on the gasification process of steep

coal layers, where, only seven kinds of compositions are

taken into account in gas phases, i.e. O2, CO, H2, CH4, CO2,

N2, H2O(g). Six kinds of compositions are taken into

consideration in solid phases. They are C, H, O, N, ash

matter, and water in coal. The corresponding relationship

between the compositions of gas phase, compositions

of solid phase, temperature and the order number is shown

in Table 1.

3. Mathematical models

3.1. Model of seepage gasification field

of gasification channel

Theequationof theconservationofgasphasecomposition:

f›ðCgiÞ

›t¼

›

›xfCgiUg þ CgD

›ygi

›x

� �þX

gjiRj þ bpi r

pi ði ¼ 1–7; j ¼ 9–14Þ ð1Þ

where f is the porosity; Cgi; the mole concentration of

composition i (mol/m3); Ug; the velocity of gas phase (m/s);

Cg; the overall concentration of gas phase (mol/m3); D; the

effective diffusion coefficient (m2/s); t; time (s); ygi; the mole

percent of the composition i; bpi ; r

pi ; the weighing and

measures coefficient of pyrolysis reaction and reaction rate of

the composition i (mol/m3 s); gji; the weighing and measures

coefficient of chemical reaction; and Rj is the reaction rate of

chemical reaction j (mol/m3 s).

The equation of the conservation of overall gas phases

f›Cg

›t¼ 2f

›ðCgUgÞ

›xþXX

gjiRj þX7

i¼1

bpi r

pi ð2Þ

The equation of the conservation of the compositions for

solid phase

ð12fÞ›Csi

›t¼ð12fÞ

›ðCsiUsÞ

›x

2X

gjiRjAMi ði¼ 9–14; j¼ 1–7Þ ð3ÞFig. 1. Physical model of the underground gasification in the steep coal

seam.

Table 1

The corresponding relationship between the compositions of gas phase,

compositions of solid phase, temperature and the order number

No. 1 2 3 4 5 6 7 8

The compositions of gas

phase and temperature

O2 H2O(g) CO2 CO H2 CH4 N2 Tg

9 10 11 12 13 14 15

The compositions of solid

phase and temperature

C H O N Ash H2O Ts

L. Yang / Fuel 83 (2004) 573–584574

where Csi is the concentration of the composition i for solid

phase (kg/m3) and AMi; the atomic (molecular) weight of

the composition i for solid phase (kg/mol).

The equation of the conservation of overall solid phases

ð1 2 fÞ›Cs

›t¼ 2ð1 2 fÞ

›Cs

›x2XX

gjiRjAMi ð4Þ

where Cs is the concentration of solid phase (kg/m3).

Assuming that the homogeneous reaction heat is exerted

on the gas phase, while nonhomogeneous reaction heat is

exerted upon the gas and solid phases, so the equation of

energy [17] is as follows.

Gas phase:

fX7

i¼1

CgiCpgi

›Tg

›t¼2f

X7

i¼1

Cpgi Cgi

›ðUgTgÞ

›x

� �

þCgDTg

›2ygi

›x2þfg

X4

i¼1

RiDHi

!

þR5DH52a1A1ðTg2TsÞ2Qg ð5Þ

where Tg is the temperature of gas phase (K); Cpgi; the

specific heat of the composition i for gas phase

(kJ/mol K); fg; the distribution coefficient of nonhomo-

geneous reaction heat; DHj; the chemical reaction heat

(kJ/mol); R5; the rate of homogeneous reaction (5) (the

transformational reaction of CO) (mol/m3 s); DH5;

the reaction heat of homogeneous phase reaction (5)

(the transformation reaction of CO) (kJ/mol); a1; the

overall heat exchange coefficient of solid and gas phases

(kJ/m3 s K); A1; the contact area between solid phase and

gas phase (m2/m3); and Qg is the heat losses of gas

phase (kJ/m3 s).

Solid phase:

ð12fÞX14

i¼9

CsiCps

›Ts

›t¼2 ð12fÞ

X14

i¼9

CsiCps

›ðUsTsÞ

›x

þ›

›xKs

›Ts

›x

� �þð12 fgÞ

X4

j¼1

ðRj DHjÞ

þa1A1ðTg 2TsÞ2Qs ð6Þ

where Ts is the temperature of solid phase (K); Cps; the

specific heat of solid phase (kJ/kg K); and Qs is the heat

losses of solid phase (kJ/kg s).

The equation of gas phase flow

›Pg

›t¼ as

›

›x

1

a0 þ b0Ug

›Pg

›x

!þ Wg

" #ð7Þ

where Pg is the gas pressure (Pa); as; the coefficient of

conduction pressure (m2/s); a0; b0; the coefficients related to

the permeability and porosity of media; and Wg is the

source-sink item.

The equation of gas state

Pg ¼ CgRTg ð8Þ

where R is the constant of gas (J/mol K).

The moving speed of flame working face is the amount of

coal consumed per unit area [18], i.e.

Us ¼V

A0

ð9Þ

where V is the amount of coal consumed within a period of

time (m3/s) and A0 is the cross section area of gasification

channel (m2).

Eqs. (1)–(7) can be generalized into the following:

›F

›t¼

1

b2

›

›xfj þ gjs 2 a

›h

›x

� �þ S

� �ð10Þ

The initial condition

Fiðx; 0Þ ¼ Fi;0 ð11Þ

The boundary condition

Fið0; tÞ ¼ Fi;0 ð12Þ

›FiðL; tÞ

›x¼ 0 ð13Þ

3.2. The model on the temperature field of coal seam

and the flow field of the retort gas

The model on the temperature field

rCp

›T

›t¼

›

›xl›T

›y

� �þ

›

›yl›T

›y

� �

2 ET 2X7

i¼1

bpi r

pi CpgiT Tðx; y; 0Þ ¼ T0 ð14Þ

2l›Tð0; y; tÞ

›x¼ a1½Tð0; y; tÞ2 Tin� ð15Þ

2l›TðL; y; tÞ

›x¼ a1½TðL; y; tÞ2 Tout� ð16Þ

2l›Tðx; 0; tÞ

›y¼ a2½Tðx; 0; tÞ2 Tg� ð17Þ

2l›Tðx;H; tÞ

›y¼ a3½Tðx;H; tÞ2 T0� ð18Þ

where T is the temperature of coal seam (K); Cp; the

specific heat of coal seam (kJ/kg K); r; the density of coal

seam (kg/m3); l; the coefficient of heat conductivity for

coal seam (kW/m K); E; heat losses coefficient

(kW/m2 K); T0; the initial temperature of coal seam (K);

a1; the heat convection coefficient (kW/m2 K); Tin; the

temperature of the gasification agent at the inlet (K); Tout;

the temperature of gas at the outlet (K); a2; the heat

convection coefficient, when the mass flow exists

L. Yang / Fuel 83 (2004) 573–584 575

(kW/m2 K); and a3 is the heat losses coefficient

(kW/m2 K).

The model on the flow field (quasi-stable state)

0 ¼›

›x

KPs

mRT

›Ps

›x

� �þ

›

›y

KPs

mRT

›Ps

›y

� �þX7

i¼1

bpi r

pi ð19Þ

Psð0; yÞ ¼ Pin ð20Þ

PsðL; yÞ ¼ Pout ð21Þ

Psðx; 0Þ ¼ PN ð22Þ

Psðx;HÞ ¼ P0 ð23Þ

where Ps is the pressure of coal seam (Pa); m; the dynamic

viscosity (Pa s); K; the permeability (D); Pin; Pout; PN; the

pressure at the inlet, outlet and gasification channel,

respectively (Pa); and P0 is the pressure of natural coal

under the ground pressure (Pa).

4. The discretization and solution of the control equation

In view of the nonlinearity of the above control

equation set and the strong coupling between equations, it

is difficult to solve the model in analytical method. As a

result, numerical solution is adopted. In the paper, we use

control volume method [19], which belongs to finite

difference method in the form. In methodology, it is a

kind of discretization having no difference from finite

element method. This method is designed to keep the

balance of integral within control volume and uses nodes

to represent control volume. The discretization of the

domain to be solved normally consists of even network

and uneven network. According to the features of the

problem to be addressed in this paper, the spatial

distribution of each variant is the function of time.

Thus, when the solution is made, it will be divided in the

form of even network. Along the x-axis, the network

division of channel is identical with that of coal seam, as

shown in Fig. 2.

Eqs. (14)–(19) have the same form. Selecting one of

them gives the general form. Take the discretization of the

equation of temperature field as a proxy [20].

Taking the derivative of the internal control volume P

during the interval from t to tþ Dt in Eq. (14)ðn

s

ðe

w

ðtþDt

trCp

›T

›tdt dx dy

¼ðtþDt

t

ðn

s

ðe

w

›

›xl›T

›x

� �dx dy dt

þðtþDt

t

ðe

w

ðn

s

›

›yl›T

›y

� �dy dx dt

2ðtþDt

t

ðn

s

ðe

wEþ

X7

i¼1

bpi g

pi Cpgi

!T dx dy dt ð24Þ

Assuming the variants change in a stepped way

according to time and space in the unstable items; the

variants in the diffusion items change according to space in

a piecewise linear way, to time in a stepped way; in the

linear source items, the variants change according to time

and space in a stepped way, thus, the above integral of each

equation can be expressed as the following.

The left side is:

rCpðTP 2 TP0ÞDx Dy ð25Þ

The first item on the right side is:

le

TE 2 TP

dx

2 lw

TP 2 TW

dx

� �Dy Dt ð26Þ

The second item on the right side is:

ln

TN 2 TP

dy

2 ls

TP 2 TS

dy

" #Dx Dt ð27Þ

The third item on the right side is:

ðSC þ SPTPÞDx Dy Dt ð28Þ

where

SC ¼ 2 E þX7

i¼1

bpi r

pi Cpgi

!TP0 ð29Þ

SP ¼ 2 E þX7

i¼1

bpi r

pi Cpgi

!p

TP ð30Þ

Substitute the above equations back into Eq. (24), and

simplify it into the commonly used form of discretization

equation:

aPTP ¼ aETE þ aWTW þ aNTN þ aSTS þ bT ð31Þ

where the specific expression of each coefficient is as

follows:

aP0 ¼rCP0 Dx Dy

Dt; bT ¼ SC Dx Dy þ aP0 TP0 ð32Þ

aE ¼ le Dy=dx; aW ¼ lw Dy=dx ð33Þ

aN ¼ ln Dy=dy; aS ¼ ls Dy=dy ð34ÞFig. 2. The network division of the domain to be solved.

L. Yang / Fuel 83 (2004) 573–584576

aP ¼ aE þ aW þ aN þ aS þ aP0 2 SP Dx Dy ð35Þ

In the above equations, superscript ‘0’ denoting the

feature of point P has the same feature with that of the

previous period; le; lw; ln; and ls are heat conduction

coefficients in the interfaces of control volume, which are

the function of temperature. According to the temperature of

nodes, the heat conduction coefficients lE; lW ; lN ; and lS

can be obtained. The heat conduction coefficients at the

boundary of control volume can be obtained through

compromise averaging method, thus

le ¼2lElP

lE þ lP

; lw ¼2lWlP

lW þ lP

;

ln ¼2lNlP

lN þ lP

; ls ¼2lSlP

lS þ lP

ð36Þ

The establishment of the nodal equation at the boundary.

The four boundaries of the coal seams to be gasified are: the

left boundary x ¼ 0; the right boundary x ¼ L; the left and

right boundaries exchange heat through convection with

gasification agent and gas all the time. Thus, it belongs to

the third kind of boundary condition; the lower boundary

y ¼ 0; is mainly the heat convection between the air current

with high temperature in the gasification channel and wall

plane of the channel, belonging to the third kind of boundary

condition; the upper boundary y ¼ H; will be considered

under the following three circumstances: the first, assuming

the temperature is fixed. It is initial temperature and belongs

to the first kind of boundary condition; the second, there is

heat radiation at a constant heat current, which belongs to

the second kind of boundary condition; the third, in the

model experiment, because the gasifier is exposed to the air,

it is also regarded as the third kind of boundary condition.

As for the first kind of boundary condition, because the

boundary node temperature TB is known, it is not necessary

to add the additional boundary nodal equation, but directly

substitute the boundary node temperature into the algebraic

equation of neighboring node.

As for the treatment of the second and the third kinds of

boundary conditions, the additional source item method is

adopted in this paper. Thus, we can substitute the boundary

conditions into the source items close to the boundary node

in the discretization equation. Remove the discretization

equation of the boundary node B; which can reduce the

dimension of the equation set and increase the solution

speed. According to the control body close to the boundary

(Fig. 3), the relational form between the close boundary

node P and the neighboring node can be obtained:

aPTP ¼ aETE þ aBTB þ aNTN þ aSTS þ b ð37Þ

where B is boundary node, and

aB ¼lB Dy

dx

ð38Þ

qB ¼ aBðTB 2 TfÞ ð39Þ

Subtracting aBTP simultaneously from the two sides of

Eq. (37) gives

ðaP 2 aBÞTP ¼ aETE þ aNTN þ aSTS þ aBðTB 2 TPÞ þ b

ð40Þ

From the heat conduction formula, we know

aBðTB 2 TPÞ ¼ qB Dy ð41Þ

When the second kind of boundary condition is given, qB

is a given value. Substituting Eq. (41) back into Eq. (37) and

simplifying it, we get

a0PTP ¼ aETE þ aNTN þ aSTS þ b0 ð42Þ

where

a0P ¼ ðaP 2 aBÞ ¼ aE þ aN þ aS þ aP0 2 SP Dx Dy ð43Þ

b0 ¼ qB Dy þ SC Dx Dy þ aP0 TP0 ¼ qB Dy þ b ð44Þ

When the third kind of boundary condition is given, i.e.

the heat convection coefficient a and the fluid temperature

Tf are given, from heat balance, we obtain

qB ¼ aðTf 2 TBÞ ¼lBðTB 2 TPÞ

dx

ð45Þ

Eliminating TB gives

qB ¼Tf 2 TP

1

aþ

Dx

2lB

ð46Þ

Substitute Eqs. (46) and (47) back into Eq. (43), and

arrange it

a0P ¼ aP 2aB þ

Dy

1

aþ

Dx

2lB

b0 ¼ bþTf Dy

1

aþ

Dx

2lB

ð47Þ

As for the turning node involving two boundaries, the

similar method mentioned above can be adopted to treat it.

Fig. 3. Boundary control volume.

L. Yang / Fuel 83 (2004) 573–584 577

5. The determination of major model parameters

5.1. The heat exchange coefficient a1 between gas

flow and solid particles

The reaction between gas and solid mainly occurs in the

gas holes of solid phase. The effect of nonhomogeneous

reaction heat is exerted on the solid phase. When the

oxidization reactions occur, the solid phase will transfer

heat to gas phase. While the reduction and drying occur, the

gas phase will transfer heat to solid phase.

The heat exchange between gas and solid consists of the

heat convection and radiation. Especially in the period of

high temperature, heat transfer in the form of radiation plays

a major role. Furthermore, when the diameter of the

particles is comparatively large, the heat conduction

among themselves should be taken into consideration.

According to Ref. [21], a1 can be determined through the

following formula:

a1 ¼1

1

aL þ a0S

þ1

aK þ aS

ð48Þ

where, aL; and a0S are the heat convection coefficient and

heat-exchange by radiation coefficient through the lime-ash

layer, respectively (kW/m2 K), and aK ; and aS are the heat

convection coefficient and heat-exchange coefficient by

radiation through the convert film, respectively (kW/m2 K).

5.2. The heat exchange coefficient a2 of wall plane

The percolation-patterned gasification channel, the heat

exchange coefficient between gas flow and the wall plane of

channel can be obtained through the following formula [22]:

a02

Cpgug

Pr2=3 ¼0:6Re1=2 ðRe , 40Þ

0:2Re20:2 ðRe . 40Þ

(ð49Þ

where Pr is Prandtl number and Re is Reynolds number,

Re ¼ rgdPug=m; rg is the density of gas phase (kg/m3).

Because Eq. (49) does not take into account the influence

of mass flow opposite to the direction of heat transfer at the

wall plane on the heat exchange coefficient, the existence of

mass flow (the dry distillation gas) decreases the gradient of

temperature in the proximity of wall plane, which, as a

result, reduces the heat transfer coefficient. Therefore, on

the basis of the establishment of a simple membrane model,

Eq. (49) is revised in Ref. [23], which gives the following

expression:

a02

a2

¼lnð1 þVÞ

Vð50Þ

where a02 is the heat exchange coefficient without mass flow

(W/m2 K); a2; the heat exchange coefficient with compara-

tively big mass flow (W/m2 K); and V is the ratio of the heat

in the channel brought by mass flow and the overall heat

transferred to wall plane through heat convection.

5.3. Heat conduction coefficient l

On the condition of the combustion and gasification of

coal seams, the overall heat conduction coefficient of media

consists of heat conduction and heat radiation, whose

expression is as follows:

l ¼ f1ðTÞ þ f2ðTÞ

where f1ðTÞ is the function of heat conduction coefficient of

the coal body (kW/m K) and f2ðTÞ is the function of

radiation and heat exchange coefficients between coal

chunks (kW/m K).

5.4. Permeability K

According to the seepage experiment, the permeability of

coal is the function of temperature [24] and increases with

the rise in temperature. The relationship between them can

be expressed as follows

K ¼ 0:2286 þ 0:01041T þ 0:0001786T2 ð51Þ

5.5. The specific heat Cp

The specific heat of coal is related to its compositions and

temperature [25,26]. In this paper, the specific heat formula

of the gas-fat coal adopted can be expressed as follows

Cp ¼X

½An cosðnpT =9Þ þ Bn sinðnpT=9Þ� ð52Þ

where the sampling of the coefficients An and Bn is as

follows:

A0¼1:69 A1¼20:60A2¼0:40 A3¼20:17 A4¼0:027

A5¼20:037A6¼0:046 A7¼20:05 A8¼0:04 A9¼20:012

B0¼0:00 B1¼0:333 B2¼0:017 B3¼20:131B4¼0:054

B5¼0:003 B6¼0:007 B7¼20:012B8¼0:026 B9¼0:00

The solution flow of numerical model on the computer is

shown in Fig. 4.

6. Model experiment

The model platform is 7.5 m long by 2 m high by 1 m

wide, consisting of base and lid, shown in Fig. 1. Under the

bottom of the gasifier a line of hydraulic jacks was installed,

which makes the gasifier rest in any angle. The hearth is so

spacious that it can be injected with other materials, which

are used to simulate coal seams with different thicknesses.

On the sides of the gasifier there are a number of gas inlets,

outlets, and slip casting holes used to simulate various

L. Yang / Fuel 83 (2004) 573–584578

gasifiers and study different modes of air pumping and

gasification channel with different lengths. A circular air

(steam) injection pipeline is equipped in the platform in

hope for supplying the air (steam) in a forward-inverse way

with moving pumping points. The gasifier adopts hydraulic

insulation technology. Over the roof of the gasifier is

installed a piece of pedal pushed by hydraulic push rod,

which imposes a certain pressure on the coal seam through

the pedal in the course of gasification, in order to simulate

the pressure on the coal seam.

The simulated coal seam in the gasifier is 6.8 m long by

1.1 m high by 0.25 m wide, with the angle of 68 8C. The

properties of the coal sample used in the model experiment

are shown in Table 2.

Fig. 4. The box diagram of the numerical calculation program.

L. Yang / Fuel 83 (2004) 573–584 579

The gasification panel is 6.0 m long by 1.0 m high by

0.25 m wide, and the initial equivalent diameter of

gasification channel is 70 mm.

The system of the model experiment is shown in Fig. 5.

The pipeline system is designed as armillary circuit.

Through the reversing valve, we can pump the air and

steam negatively and positively. Coal gas cleaner is mainly

to remove the tar from the gas and lower the temperature of

the resulting gas. In order to measure the temperature

distribution in the gasifier accurately, we deploy the

temperature-measuring points in an intense way. In the

gasification panel, 22 rows of temperature-measuring points

are buried, with 7 each row. The number totals 154. The

temperature-measuring elements adopt the strictly standar-

dized NiCr–NiSi thermal couples. The data are collected by

data acquisition system automatically and displayed on the

screen of the computer and recorded. The resulting gas will

be analyzed in gas phase chromatographer, which can tell

the contents of different compositions.

After the experimental system has been prepared, before

the formal ignition, the cold test should be carried out,

whose main purpose is to measure the leakage rate of the

gasifier and check whether all the system is in a good

working state or not.

Suppose the flowing capacity of the air supplied at the

inlet is Q1; the flowing capacity of air discharged at the outlet

is Q2; the quantity of wind leaked in the gasifier is

DQ ¼ Q1 2 Q2; thus the leakage rate of the gasifier is

h ¼ DQ=Q1

The experimental results of the cold test are shown

in Table 3.

From Table 3, we can see that the leakage rate generally

falls below 8.00%, with the maximum of only 10.16%. With

the increase in the air quantity blasted, leakage rate tends to

increase but still within the tolerance of losses, which is the

basic condition for the smooth run of the model experiment.

First of all, ignite the gasifier by means of an electrically

fired igniter, then, blast air with small quantity, sample the

gas from the outlet every 5 min, analyze its compositions

and heating value, half an hour later, the interval of

sampling is prolonged to 10 min, increase air blast quantity

gradually, the interval of sampling is so long as 1 h 15 min

later. When the heating value of the gas reaches 1300 kJ/m3,

the ignition process is considered successful. According to

the measurement results of the temperature, after the stable

temperature field in the gasifier is formed, the interval of

collecting temperature data at all the temperature-measuring

points is 1 min.

The model experiment lasts for 117 h, undergoing four

phases, namely, gasification preheating (24 h), stable gas

production (58 h), intermittent air pumping (18 h) and

inverse air pumping (17 h). The operating conditions are:

the gasification agent, air; the pressure at the inlet, 2400 Pa;

and the flowing capacity at the inlet, 20 m3/h. The results of

the experiment are shown in Figs. 6(a)–9(a), 10 and 11.

7. The analysis of calculation results

The calculation results of numerical simulation are

shown in Figs. 6(b)–9(b), 10 and 11.

Comparing Figs. 6 and 8, we can find that the calculation

results of temperature field virtually conform with the

measurement results. Except the measuring points in the

combustion zone, where the relative errors between

calculated value and measured value of temperature are

comparatively high (certain points, over 20%), the relative

errors of other points are no more than 14%, majority of

which are within 10%, which completely meets the precision

Table 2

Properties of the coal sample studied

Proximate analysis (wt%) Ultimate analysis (daf, wt%) Heating value (MJ kg21)

Moisture Ash Volatile matter C H O N S

4.18 7.61 23.08 72.72 4.71 8.32 1.13 1.33 28.14–29.31

Fig. 5. The system of the model experiment.

Table 3

The data of the cold test (m3 h21)

Modes of air supply Forward air supply Inverse air supply

Q1 36.4 32.1 27.3 21.5 40.0 33.8 24.7 19.5

Q2 32.7 29.5 25.5 20.6 36.1 31.6 22.9 18.8

DQ 3.7 2.6 1.8 0.9 3.9 2.2 1.8 0.7

h (%) 10.2 8.1 6.6 4.2 9.7 6.5 7.3 3.6

Average (%) 7.3 6.8

L. Yang / Fuel 83 (2004) 573–584580

requirements [27] of numerical simulation on the tempera-

ture field. From the distribution of temperature field, the

calculation value is a bit bigger than measurement one. The

reasons are as follows: first of all, heat losses coefficient E in

mathematical models is determined, on the basis of

composite structure calculation formula of heat transfer

coefficient for the fixed conditions (temperature, wind

velocity). In the process of experiment, due to the natural

ventilation or forced ventilation, the heat losses coefficient

increases, which contributes to the slight drop of the

temperature. Then, during calculation, the heat conduction

coefficient of coal seams is not the one of body coal but

the equivalent heat conduction coefficient taking the

influence of convection and heat radiation into consider-

ation. The coal seams in the model experiment are piled up

by various sizes of coal chunks. Though the interstices are

filled with coal powder, its holes are far bigger than the real

one. Therefore, heat exchange is not a single form of

heat conduction. With the effect of convection intensifying,

the surface heat conduction coefficient of the coal

seam improves, which influences the precision of calcu-

lation results.

Because the reaction rate heightens rapidly with the

increase of temperature, as a result, due to the impact of

temperature, the change gradient of the compositions

concentration for measurement value in high temperature

zone is bigger than that of calculation value (Figs. 7 and 9).

According to Figs. 7 and 9, with the expansion of

gasification channel, the heating value of the gas increases

gradually. However, behind the reduction zone, the extent

of increase decreases. The influence of the temperature field

on the heating value of the gas is remarkable. In the first

period, in spite of the comparatively long gasification

channel, due to the low temperature in the oven, the heating

value of the gas is relatively low; in the second period, the

temperature of the oven increases in the comparatively

shorter gasification channel, but the heating value of

resulting gas is higher than that in the first period. Therefore,

maintaining a temperature field with high temperature and

Fig. 6. The temperature field in the gasification panel 22 h after the ignition.

Fig. 7. The concentration field in the gasification channel 22 h after the ignition.

L. Yang / Fuel 83 (2004) 573–584 581

comparatively longer gasification channel is conducive to

the stability and increase of the heating value for the gas.

Figs. 10 and 11 demonstrate that, the simulated

calculation value of the pressure for gas flow basically

conform with the measured value. The comparative error in

the first period between calculation value and measurement

value is 5.00–14.29%, with its average drop rate for the

pressure of fluid 9.5% (þ15.2, 213.1%). From Fig. 11, we

know that, 75 h after the commencement of gasification, the

error of the simulated calculation for pressure is 9.68–

17.24%, with its average drop rate of the pressure of gas

along gasification channel 11.97% (þ21.39, 217.33%). It

can be concluded that, with the prolonging of gasification

time, the extent of the pressure drop of fluid increases and

the calculation error becomes bigger. The major reasons are

the following, at the very beginning of the experiment, the

gas moves along the free gasification channel with little

resistance on the fluid and small fluctuation of pressure.

Accordingly, the drop rate and calculation error are

comparatively low. With the development of gasification,

the top coal layer over the gasification channel, due to

the effect of high temperature, expands, inbreaks and

falls onto the gasification channel because of dead

weight, which fills the gasification channel with loose

coal chunks. Thus, the free gasification channel becomes

percolation-patterned. The resistance over the gas move-

ment increases dramatically, so does the extent of the drop

rate of pressure. The curve of the experiment takes on wild

fluctuation. Considering the changes of movement con-

ditions and the determination of major model parameters,

such as conduction pressure coefficient, permeability as well

as the involvement of human factors and empirical factors in

the calculation results in a certain error in the parameters

calculation, which causes oscillation in the value of

numerical simulation.

In short, the simulated results support that the calculation

value conforms with the measurement value well, which

shows that the establishment of mathematical models on

Fig. 8. The temperature field in the gasification panel 75 h after the ignition.

Fig. 9. The concentration field in the gasification channel 75 h after the ignition.

L. Yang / Fuel 83 (2004) 573–584582

the temperature field, concentration field and pressure field

in the gasifier, the determination of parameters, the analysis

and treatment of boundary conditions and the solution

method are correct. This provides necessary theoretical

basis and scientific guidance for the comprehensive

quantitative study and production practice of underground

coal gasification.

8. Conclusions

(1) According to the storage conditions and the features of

gasification process for the steep coal seams, on the

basis of model experiment, the mathematical models

on the gasification process of steep coal seams are

established. The simulated results demonstrate that the

calculation value can virtually conform with the

measurement one, which supports that numerical

simulation on the temperature field, concentration

field and pressure field is reasonable in the under-

ground gasification of steep coal seams on the

experimental condition.

(2) The numerical simulation shows that, in high tempera-

ture zone, the calculation value of the temperature field

is a bit bigger than measurement value; the change

gradient of the measurement value of the concentration

of various compositions for the gas is bigger than that

of the calculation value; temperature field has a great

influence on the heating value of the gas. The

heating value of the gas increases with the rise in

the temperature of gasifier. The ideal temperature

field with high temperature is conducive to the

improvement of the gas and the stability of the

gasification process.

(3) According to the calculation results, the relative error

between the calculation value and measurement one of

the fluid pressure and its drop rate increase gradually

with the gasification process. The change in the

seepage condition of gasification channel is mainly

responsible for the comparatively big calculation error.

(4) The numerical calculation results basically demon-

strate the real change patterns of the temperature field,

concentration field and pressure field in the process of

underground gasification of steep coal layers. Though a

certain error exists between the calculation value and

experiment one, however, according to the specific

combustion and gasification conditions of coal seams,

choosing appropriate parameters or optimal parameters

with the help of inverse calculation, the above

mathematical models can be fully applied to pro-

duction practice and to predict the change patterns of

‘three fields’ of the process of underground coal

gasification.

Acknowledgements

This work was supported by the National Natural Science

Foundation of China (Ratification No. 59906014,

50276066). The technical contributions of professor Yu Li

and Dr Liang Jie are gratefully acknowledged by the author.

References

[1] Kpekmomo EB. Coal Chem Ind 1993;6:61–5.

[2] Chandelle V. Collect Transl Works Mining 1992;13:5–7.

[3] Beyer LG. Collect Transl Works Mining 1989;10:1–5.

[4] Guntermann K. Collect Transl Works Mining 1988;9:6–10.

[5] Liang J. PhD Thesis, China University of Mining and Technology;

1997.

[6] Yang LH. J China Univ Mining Technol 2000;29:142.

[7] Liu SQ. J China Univ Mining Technol 2000;29:606.

[8] Guntermann K, Gudenau HW, Franke FH. Proceedings of the 12th

Annual Underground, Coal Gasification Symposium, Washington,

DC, 24–28 August. ; 1986. p. 207–15.

[9] Mortazavi HR, Emery AF, Corlett RC. Proceedings of the 12th

Annual Underground Coal, Gasification Symposium, Washington,

DC, 24–28 August. ; 1986. p. 252–64.

[10] Francis TW, Mead SW, Daniel HS. Underground Coal Gasification:

The State of the Art 1983;79:154–9.

[11] William BK, Robert DG. Underground Coal Gasification: The State

of the Art 1983;79:129.

[12] Guo CW. Mining World 1994;15:3.

[13] Sun JH. China Energy 2001;5:19.

[14] Yang LH. PhD Thesis, China University of Mining and Technology;

1996.

Fig. 10. The pressure field in the gasification channel 22 h after the ignition.

Fig. 11. The pressure field in the gasification channel 75 h after the ignition.

L. Yang / Fuel 83 (2004) 573–584 583

[15] Liu SQ, Liang J, Yu L. Coal Conversion 1999;22:50.

[16] Yang LH. In: Song DY, editor. Study on the seepage combustion

methods in underground coal gasification. Xuzhou: China University

of Mining and Technology Press; 2001. p. 61–3.

[17] Fan WC, Wan YP. In: Hu SH, editor. The model on movement and

combustion and its calculation. Hefei: China University of Science

and Technology Press; 1992. p. 201–7.

[18] Wu RY. In: Huang W, editor. Coal gasification. Xuzhou: China

University of Mining and Technology Press; 1988. p. 173.

[19] Guo KL, Kong XQ, Chen SN. In: Jiang YY, editor. Calculation heat

transfer. Hefei: China University of Science and Technology Press;

1988. p. 75–91.

[20] Liu SQ. PhD Thesis, China University of Mining and Technology;

2000.

[21] Xiang YQ, Zhang LP, Xiang P, Wei JF. Coal Gas Heat 2001;21:24.

[22] Yagi S, Wakao N. AichE J 1959;5:80–5.

[23] Smith DM, Williams FL. Fuel 1984;63:259.

[24] Encinar JM, Beltran FJ, Ramiro A, Gonzalez JF. Fuel Process Technol

1988;55:225–8.

[25] Qizhi N, Williams A. Fuel 1995;74:108.

[26] Xiang YQ. Coal Gas Heat 1995;15:10.

[27] Yang LH. In: Song DY, editor. Underground coal gasification

project. Xuzhou: China University of Mining and Technology

Press; 2001. p. 184.

L. Yang / Fuel 83 (2004) 573–584584