1

12th November 2006

Kolkata, India

�Dr Andrew Beath

�CSIRO Exploration & Mining

�Australia

�Dr Cliff Mallett

�Carbon Energy Pty Ltd

�AustraliaSupported by the Australian Government through the Asia Pacific Partnership (AP6) Coal Mining Task Force

Session 2A: Behaviour prediction

Behaviour prediction

One of the key problems with past UCG operations has been the difficulty in understanding what is happening. Many months of data analysis and modelling was required to interpret

results from some experimental trials.

Modern computing allows the opportunity for real-time assessment of the reactor behaviour, if suitable

models can be developed.

UCG reaction processes

Product gas

Drying, Volatile release &

Gasification

Drying & Volatile release

Combustion&

High temperatures

Gas reduction&

Moderate temperatures

Gas equilibrium reactions& Cooling

Drying

Feed gas

Modelling

Coal model Cavity model

Geotechnical model

Regional hydrology model

Process simulation

Coal Model

Dry coal

Wet coal

Diffusion of gasesIN OUT

Char

Water flow

GasThe coal model represents lump coal reacting with a hot gas. Included in the model are:�Reactions�Gas diffusion�Water flow�Drying�Heat transfer�Coal structural changesOutput is used in the cavity model.

Predictions from the coal model-Impact of reactant gas mix and water

0.0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.91:16

1:4

1:1

4:105

101520

2530

3540

4550

Co

al fa

ce r

eces

sio

n ra

te (m

m/h

r)

Water ingress flux (kg/m2/s)

Oxygen:Steam (kg:kg)

2

Predictions from the coal model-Impact of pressure and temperature

1000

1100

1200

1300

1400

1500

1600

1700

1800

1900

1

10

40

1000

50

100

150

200

250

Coa

l fac

e re

cess

ion

rate

(mm

/hr)

Gas temperature (K)

Pressure (atm)

Use of the coal model

�Does not provide standalone predictions relevant to UCG as it neglects many of the gas flow and heat transfer features of real cavities

�Makes spot predictions of coal behaviour under pseudo-steady state conditions to feed into more complex models

�Can be used to predict the general operating regimes that are desirable for efficient gasification

Elements of a cavity model

� Coal & char reactions

� Coal/char structural changes

� Gas flow and reactions

� Water flows and evaporation

� Heat transfer

o Conduction,

o Convection

o Radiation

� Rock & coal breakage and collapse

� Resizing of the matrix with growth

Model of Cavity Growth

Rubble

Coal model

Rock floor

Rock roof

Evaporation front

Gas

Cavity model operation

�� ������������� ������������� ��������������������������

�� ������������������������������������������������������������������������

����������������������������������������������

�� ������������������������������������������������������������������������

��������������������������

����������������������������������������������

������������������������������������������

Specific gas concentration prediction from cavity model

3

Hanna II trial (1975) - Air-blown, Vertical wells

0

10

20

30

40

50

60

155 160 165 170 175 180 185 190 195

Time into test, days

Gas

con

cent

ratio

n, v

olum

e%

Experimental CO2Experimental H2Experimental COExperimental CH4Experimental N2Model CO2Model H2Model COModel CH4Model N2

Rocky Mountain 1 trial (1987-88)-Oxygen-blown, Knife edge CRIP

0

5

10

15

20

25

30

35

40

45

50

10 15 20 25 30 35 40 45 50

Time into test, days

Gas

con

cent

ratio

n, v

olum

e% Experimental CH4Experimental COExperimental H2Experimental CO2Experimental N2Model CH4Model COModel H2Model CO2Model N2

Cavity growth (Experimental) 1

Injection

Injection

Production

Production

Days9 to 53

Days53 to 73

Cavity growth (Experimental) 2

Injection

Injection

Production

Production

Days73 to 93

Days93 to 102

Model performancePredicts accurately:

o Cavity volume changes

o Product gas composition and flow

Hindrances to model performance:

o Requires detailed site information

o Experimentally, the cavity shape was affected by uncontrolled shortening of the ‘CRIP’ and an undetected fault running through the site

Physical site changes

Modelling of other side of UCG, the physical site changes, will be discussed in a later session.

In the cavity modelling, simplified models are used for roof collapse and

hydrological flows and these are ‘tuned’ using output from the more

complex and specialised geotechnical and hydrological models.

4

12th November 2006

Kolkata, India

�Dr Andrew Beath

�CSIRO Exploration & Mining

�Australia

�Dr Cliff Mallett

�Carbon Energy Pty Ltd

�AustraliaSupported by the Australian Government through the Asia Pacific Partnership (AP6) Coal Mining Task Force

Session 2B: Process performance & economic viability

Historical utilisation of UCG gas

�Most Soviet-era UCG produced a fuel gas for use as supplementary fuel in coal-fired boilers, so gas specification was not stringent

�Chinese UCG is commonly used for domestic fuel gas, but has been used for hydrogen production and as a synthesis gas for ammonia production

�Rawlins II trial in the USA demonstrated reliable synthesis gas production and a subsequent (failed) commercial plant at the site was intended to synthesis methane

Modern applications for the UCG

�The product gas can be used as a:

oFUEL

�Electricity production via gas turbines

oSYNTHESIS FEEDSTOCK

�Production of chemicals (eg. fertilisers)

�Synthesis of liquid fuels (Fischer-Tropsch)

Electricity generation

Underground coal gasification can provide a fuel gas that is

suitable for use in modern gas turbines after cleaning.

The power plant design is similar to that of the proposed IGCC plants using mined coal.

Combined cycle electricity generation

GASIFICATION & PROCESSING

GAS CYCLE

Power

STEAMCYCLEPower

Potential for CO2 removal

UCG and gas turbines

�UCG product gas has a different composition for every site and varies significantly from that of entrained flow gasifiers for IGCC systems

�This has an impact on the design of the turbine combustor and the turbine

�Turbines are typically specified on mass flow, so the different gas composition can impact on operation

5

Process efficiencies for combined cycle electricity generation

~45 %IGCC

~37 %Conventional coal

39.8 %UCG with CO2 separation

46.5 %Oxygen-blown UCG

45.4 %Air-blown UCG

EfficiencyProcess

Economics of Power Generation

20

22

24

26

28

30

32

150 300 450 600

Seam depth, m

Ele

ctric

ity c

ost,

US

$/M

Wh

VW Air, 4m seam

CRIP Air, 4m seam

VW Oxygen, 4m seamCRIP Oxygen, 4m seam

VW = Vertical Wells technology & CRIP = Directionally-drilled technology Based on a 200MWe combined cycle plant

Resource utilisation efficiency

An alternative viewpoint is in terms of resource utilisation efficiency, so how does UCG compare with conventional

mining and IGCC in resource utilisation efficiency?

Notes: - Underground mining losses range from 25-80% of the coal- Transport and mining use approximately 5% of the coal energy

Conventional resourceutilisation efficiency

Conventional miningwith IGCC electricity generation

1MW inmined coal

250kW incoal

wastes

Fuel supply750kW in

product coal

415kW inwaste heat

335kW inelectricity

Power generation

Underground coal gasification resource utilisation efficiency

Underground gasificationfor electricity generation

Note: - Underground gasification leaves 10-30% of the coal underground

100kW in wasteslost underground

Processing

1MW of coal

900kW of product gas

385kW inelectricity

380kW inwaste heat

35kW inliquid wastes

Power generationFuel supply

Comments on electricity generation

�A simple option is to use UCG gas in existing coal-fired boiler plants – this will typically be limited to about 30% of the energy input coming from UCG, but allows for very flexible gas compositions

�Modern gas turbines can use UCG gas with minimal cleaning (simple removal of condensates) and can be cover a range of compositions, but efficiencies will vary

6

Liquid fuel synthesis

�UCG is a low cost option for providing gas for Fischer-Tropschsynthesis of liquid fuels

�This is a tempting process to consider due to the high value of the products, but the capital cost of the synthesis plant is very high

Application - Liquid Fuel Synthesis

LIQUID SYNTHESISCOAL GASIFICATION GAS PROCESSING

GTL Plant (SasolChevron)

Reformer$162m

Oxygen$142m

Naturalgas

F-TSynthesis

$176m

GasTreatment

$7m

CO2Removal

$34m

ProductWork-up

$47m

Diesel24,000bbl/d

Gasoline9,000bbl/day

LPG1,000bbl/d

Hydrogen

Steam,Water & Power systems$109m

Site services$180m

Sources: Foster Wheeler & Technip-Coflexip (Qatar plant)

UCG with GTL Plant

Reformer$162m

OxygenBigger

Naturalgas

F-TSynthesis

~Same

GasTreatment

~Same

CO2RemovalBigger

ProductWork-up~Same

Diesel24,000bbl/d

Gasoline9,000bbl/day

LPG1,000bbl/d

Hydrogen

Steam,Water & Power systemsSome changes (more Power)

Site services~Same

UCGLow cost

Economics of Liquid Fuel Synthesis

15

20

25

30

35

40

150 300 450 600Seam depth, m

Syn

thet

ic o

il co

st, U

S$/

BB

L

VW , 2m seamVW , 4m seamVW , 8m seamCRIP, 2m seamCRIP, 4m seamCRIP, 8m seam

VW = Vertical Wells technology & CRIP = Generic directionally-drilled technology Based on a 10,000 bbl/day plant

Comments on liquid synthesis

�The gas specification for this process is much more stringent than for electricity generation and it will be difficult to convince financiers that UCG alone can supply a reliable gas feed

�Large scale UCG with gas blending can maintain constant composition, but may lead to environmental problems

7

Comments

Process simulation is necessary for prospective plants using UCG as:

�Often the plant will have a tight integration between surface and underground operations

�Differences between the UCG gas and conventional gases may have a significant impact on the surface plant operation

12th November 2006

Kolkata, India

�Dr Andrew Beath

�CSIRO Exploration & Mining

�Australia

�Dr Cliff Mallett

�Carbon Energy Pty Ltd

�AustraliaSupported by the Australian Government through the Asia Pacific Partnership (AP6) Coal Mining Task Force

The End