Mathematicalmodelingofmethane

flowincoalmatrixusingCOMSOL

ProductionthroughNormalfractures,gasdesorptionanddiffusion1.

1.CompositeEnergy:www.compositeenergy.co.uk/cbmformed.html

TableofContent
Introduction:_________________________________________________________________ 3
GoverningEquations:__________________________________________________________ 5
a)Assumptions:____________________________________________________________________ 5
b)MathematicalEquations:__________________________________________________________ 6
Flowthroughcleats:_______________________________________________________________________6
FlowthroughMicropores:__________________________________________________________________7
c)BoundaryandInitialConditions: ____________________________________________________ 7

3.Formulation:_______________________________________________________________ 8
a)Microporesandcoalmatrix:_______________________________________________________ 8
b)Cleats:__________________________________________________________________________ 9

4.SolutionUsingCOMSOL: ____________________________________________________ 10
a)Microporesandcoalmatrix:______________________________________________________ 10
b)Cleats:_________________________________________________________________________ 11

5.ValidationofModel: _______________________________________________________ 11
a)MicroporesandMatrix:__________________________________________________________ 11
b)Cleats:_________________________________________________________________________ 12

6.ParametricStudy:__________________________________________________________ 13
a)Microporesandmatrix:__________________________________________________________ 13
6.1VaryingDiffusioncoefficient:____________________________________________________________13
6.2VaryingTime: ________________________________________________________________________14
b)Cleats:_________________________________________________________________________ 15
6.3Varyingpermeability:__________________________________________________________________15
6.4Varyingconcentrationatcleatsboundaries:________________________________________________16

7.Conclusions: ______________________________________________________________ 17
AppendixA_________________________________________________________________ 18
References _________________________________________________________________ 19


Introduction:
Almostallgeologicalstratacontaingases.Thesegasesmaybereleasedbyundergroundmining
activities often in concentrations too small to be of concern1. Coalbed methane (the gas
emitted from coal which is primarily methane with minor amounts of heavier hydrocarbons,
carbondioxide,nitrogen,oxygen,hydrogenandhelium)2isachiefcomponentoffiredampina
coalmine3.CBMisviewedasafuelwithmanyenvironmentaladvantagesbecauseofthelower
levelofsulfuroxides,hydrocarbonsandcarbonmonoxidesitreleaseswhencombusted4.The
Methane primarily resides in the phyteral Pores and micro pores, as well as in the adsorbed
stateonthecarboncomplexesinthecoalmatrix5andhencethediffusionrateisverylowatthe
temperature found in mines3. The adsorption potential of coal is awesome, allowing it to
contain very large amount of gas. A 0.45kg of coal can have up to 46,500 m2 of adsorptive
surface area and estimates that the Canadian Elmworth coals in Alberta can contain up to
14m3/tone of free adsorbed material6. Residual gas levels may be as great as 32% for high
volatile bituminous coals7. Methane bearing coals are considered to be a significant gas
resource. Al though coal is a porous medium, permeability is usually quite low and the pore
structure is considerably more complex than the usually found in conventional clastic
reservoirs8. Therefore, the increasing importance of coal seams as gas reservoirs, attention is
beingfocusedonfracturepatternsincoalmatrix9.Theprincipalnaturalfracturespermeability
is the pathways for the flow of gas and water through the cleat systems. Cleats are the
systematicfracturesincoalthatareequivalenttojointsinothersedimentaryrocks10,11.Cleat
systemisoftenthereservoircharacteristicthathasthegreatestinfluenceovertheeconomic
success or failure of coal gas exploration and development program12. Cleats are natural
fracturesincoalthatserveaspermeabilityavenuesforDarcyflowofgasandwatertoawell
bore depressurization13,14. Two cleat sets in orthogonal pattern are designated Face and Butt
cleat.Facecleatsarecommonlyplanar,smoothsidedfracturesthatusuallycomprisethemost
prominent fracture set9. Coalbed permeability may be 310 times greater in the face cleat
directionthanintheotherdirections11.Therearetwophasesofmethaneflowfromcoalmatrix
3

towellbore.Firstly,theadsorbedmethanemustdiffusethroughthe microporesofthecoal
matrixuntilitreachesanaturalfracture(cleat)andsecondlymethaneflowsthroughthecleat
networktothewellboreinresponsetoapressuregradient(DarcyFlow)15.Methaneflowfrom
the coal is dependent on upon the effective permeability of the coal. According to some
mathematicalandcomputermodelsofmethanedrainage,thespacingofthemacrofractures
(cleats)playstheprimaryroleinthetransportofmethanethroughcoal16,17.Sincethecoalhas
verylowpermeabilitycoalbedmethanereservoirgenerallyrequiresstimulationtreatmentfor
increasingtheproductionperwell.TwoCommonmethodsareusedtoincreasetheproduction
viz.a)Increaseinwellborecontactareabyhorizontaldrillingandb)Increasethepermeability
byfracturing8.

Fig1:Flowofmethanethroughthecoalmicrostructure18.

The conventional primary CBM recovery process begins with a production well that is often
stimulated by fracturing to connect the well bore to the coal natural fracture system via an
inducedfracture19.Whenthepressureinthewellisreducedbypumpingwaterfromthewell
byusingan artificiallift mechanism,thepressureintheinducedfractureisreducedwhichin
turn reduces the pressure in the natural coal fracture system. Initially when the operation is
started, water begins to move in the direction of the pressure gradient. When the natural
fracture system pressure drops below the critical desorption pressure, methane starts to
desorb from the primarysecondary porosity interface and is released into the secondary
porositysystem20.Asaresult,theadsorbedgasconcentrationintheprimaryporositysystem
4

near the natural fractures is reduced. This reduction creates a concentration gradient that
results in mass transfer by diffusion through the micro and mesoporosity19. Adsorbed gas
continuestobereleasedasthepressureisreduced.
Inthisstudy,aneffortismadetovisualizethemethaneflowduringdegasificationofcoalseam.
In a virgin coal seam, methane molecules remain absorbed in the micro pores of the coal
matrix. In this two dimensional model, coal matrix is considered as rectangular blocks
surroundedbytwowayconduitsviz.FacecleatsandButtcleats.Itisassumedthatthereisonly
four micropores per coal matrix block. This methane diffuses through micro pores and
howevermeetsthecleats.Thesecleatsallowittoflowuptotheborewell.FEMLABisusedfor
modelingpurpose.

GoverningEquations:
a)Assumptions:
1. Fractures are sufficiently wide and there are no coal particles inside the created
fractures.Iftheyareverysmallthenflowwouldbecomecapillaryflow.
2. Flow of methane inside the cleats is considered as laminar and hence Darcy law is
applicable.
3. Thereisnochangeoftemperatureincoalseamduringthedegasificationprocess.
4. Thewidthandcleatspacingremainsconstantduringthegasflowperiod.
5. Cleatsarestraightandthereisnotortuocityinthem.
6. Coal matrix is in continuous equilibrium with the fracture system i.e. there is no
accumulationofmethaneatthefracturesurface.
7. Thereisnomatrixshrinkageinthenetworkduringdesorptionofmethane.
8. PermeabilityisassumedtobeconstantwhileitvariesaspertheKlinkenbergLaw.

Figure2:Exaggeratedschematicofmethaneflowthroughcoalmatrix.

b)MathematicalEquations:
Atwodimensionalsystemisconsideredforthisproblem.Microporesarerepresentedassmall
circle.Thesearethepotentialsitesformethanediffusion.Methanemoleculesdesorbedfrom
thesesites,diffusethroughthecoalmatrixtilltheyfindacleat.Afterenteringinacleatsystem,
thisflowfollowstheDarcyslawandcomeout.
Flowthroughcleats:

For1Dradialcylindricalflowthecontinuityequationcanbewrittenforagasphasefluid

.(1)

(Neglectingthesetermsforstudy)
Darcylawforafractureflow

....(2)
6

Bysubstitutingequation(2)inequation(1)

(3)

(Neglectingthesetermsforstudy)
FlowthroughMicropores:

Fickslawfordiffusion

...(4)
AdsorptionanddesorptionratesunderequilibriumconditionisgivenbyLangmuirequation

..............................(5)

Thetotalvolumeconcentrationofmatrixgasonapervolumebasis

..(6)

ByputtingthevalueofCinequation(4)

(7)

RealGaslaw

.........................................................................................(8)
c)BoundaryandInitialConditions:
The initial condition is given by as the pressure distribution of the methane gas in the cleat
systematt=0.Thepressuredistributionisuniformthroughoutthecoalseamandisequalto
initial reservoir pressure. Methane gas concentration on the external surface of the matrix
elementsisevaluatedatthegaspressureinthecleats.Themicroporeandmacroporesystems
are coupled by the boundary conditions specified along the cleat walls that the two systems
share.Matrixandfracturetransportequationsarecoupledattheexternalsurfaceofthematrix
elementswherethemicroporegasconcentrationisevaluatedatthegasfracturepressure.

InitialP= andFinal
Variable
Macro pore porosity
Rockdensity(Kg/m3)
SeamTemperatureT(K)
N/m2
DiffusionConstant(m2/s)D
Langmuirspressure

Value
.02
1370
313
10E+5
3.75X10E11
1.2E+6
0.93

Variable
Macroporepermeabilityk(D)
Initialseampressure N/m2
Microporeporosity
K
Langmuirsconstant
GasMolecularweightM

Value
2.47E15
3.4E+6
.025
298
13.9E3
16.043
1

3.Formulation:
Most numerical models developed are based on physical phenomenon. The most commonly
usedmodelisshowninfigure3,wheresideofthesquareshowsfracturingsurface.

Figure3:IllustrationofCoalmatrixunitandmethanedissusion15.
Acoalbedisshowntobemadeupofsmallcubicblocks,orunitsseparated byfractures.The
spacingofthefracturesdetermineshowfarthegashastodiffusebeforereachingthefracture
andthedimensionsofthefracturesdecidethequantityofgasthatcanflowthrough15.

a)Microporesandcoalmatrix:
Asmallunitofcoalmatrixwithside0.005misconsidered.Thisunitisisotropicinnatureand
pressure and temperature is assumed to be constant, though they may vary from point to
point.Theoretically,therearemillionsofbillionsofmicroporespresentincoalunitofthissize.
Fourmicropores(representativeofsameporevolumeofmillionsofbillionsofmicropores)of
8

0.37mmradiuswereconsideredforthisstudy.Therelocationissetasperthesymmetryofthe
block.Assumptionmadeindescribingthismodelhasbeenmentionedpreviouslyinthispaper.
Boundaries of the coal matrix are set at a concentration given by Langmuir isotherm at 1.25
Mpa.Concentrationinsidethecoalmatrixsubdomainisfixedas50mole/m3.Representative
microporeare assumed as constant source of very highmethane concentration(6000 mole/
m3)andtherediffusioncoefficientisveryhighincomparisonofcoalmatrixdiffusioncoefficient
(D=3.75X1011m2/s). In the matrix, diffusion of methane is governed by diffusion constant D.
Coalmatrixboundariesdirectlycommunicatewithfaceandbuttcleats.Withabovementioned
assumptions and nature of the model, the convection and diffusion model under chemical
engineering in mass balance section seems to be best fit. An application mode as shown in
figure4wasdesignedforboundaryconditionspairedwithinitialconditions.
Theboundariesofthematrixareatfixedconcentrationasthemethanemoleculearrivedthere
taken out via cleats. There is no accumulation of gas and pressure remains constant. Gas
constant R, M, V, PL were defined into constant section under option menu. Langmuir
adsorptionequationwasfittedintothescalarexpressionsavailableintheoptionmenu.Since,
thereisnoreactiontakingplaceinthesystemthevalueofRiskeptzero.Also,initialconditions
andboundariesconditionsaredefinedasascalarorconstantexpressionunderoptionmenu.

b)Cleats:
Arectangularelementof.005mX.00001misconsideredasaverticalcleat21(asshowninfigure
5). This unit is isotropic in nature and pressure and temperature is assumed to be constant,
though they may vary from point to point. Permeability for this cleat is defined as a scalar
expressionunderoptionsmenu.

84.4

10

Darcy;

Wherew:widthofthecleat(ThisiscalculatedbyusingBoundarydistancevariable(dist1)

availableinoptionmenu)
z:separationofthecleat
Densityofthegasisalsodefinedasscalarexpression.Boundaries(3and11)ofthecleatsare
keptatapressureof0.1Mpawhileotherboundariesofcleatsarekeptinsulated.Boundary(4
and8)iskeptatequivalentpressureof50mol/m3concentration.
Parameter
Value(MKS)

Density
.68

Porosity
.3

Viscosity
4

1.010

p(t0)

w
5

1.0110

z
5

110

1103


Figure 4: Geometry (up to the scale) drew in COMSOL representing Micropores, coalmatrix
andcleat.

4.SolutionUsingCOMSOL:
Outputswereobtainedfromthesimulationwithabovementionedvalue.

a)Microporesandcoalmatrix:
SurfacemapfordiffusivefluxandconcentrationcontourswereplottedatT=864000s(10days).
surfacethatthevalueoffluxkeepsondecreasingwhilegoingawayfrommicropores.Forthe
same model concentrationcontours for methane were observed. It can be seen from plot
(figure5)thatconcentrationdecreaseswithdistancefromthemicropore.

Figure5:Diffusivefluxsurfacemapandconcentrationcontoursforasinglecoalmatrixunit.
10

b)Cleats:
Surfacemapofvelocityfield(m/s)andpressurecontourswereplottedatT=864000s(10days).
Velocityvariesfromzero(atcoalmatrixbottomend)tohigh(atupperend)asshowninfigure
6.

Figure6:Velocityfieldsurfacemapandpressurecontoursforacleat.

5.ValidationofModel:
Validation is an important part of a model. To validate for the overall solution, the
concentration in the matrix was set to zero. Concentration was plotted across a vertical line
passingthroughthecentreofthemicropores.Thenatureofthecurveiscomparedthecurve
insimilarliterature.

a)MicroporesandMatrix:
The breakthrough profile for concentration was obtained across a line passing through the
centreofthetwomicropores.Theconcentrationforthemicroporeswassettozero.Reverse
fluxpatternwasshownbyfollowingprofile.

11

Figure 7: Surface, showing concentration (arrow showing direction of reverse flow inside the
micropores)andabreakthroughprofileoftotalfluxacrossacrosssection.

Thisconfirmsthevalidityofthemodelofmicropores.Nowconcentrationprofilewasobserved
from source to boundary. This profile matches with the profile found in literature work for
Fick'sdiffusion.

Figure 8: Typical concentration profile of Ficks diffusion (right) obtained from literature and
profileobtainedfrommodel(left).

b)Cleats:
Validationofthecleatsmodelwasdonebychangingthepressuredifferential.Pressure

12

differentialwassetasP,2Pand3P.Darcylawstatesthatvelocityisdirectlyproportionalto
pressuredifferentialwhichisapparentfromFigure9andhencevalidatestheresults.

4.50E01

VelocityVsdP

4.00E01

Velocityattained(m/s)

3.50E01
3.00E01
2.50E01
P

2.00E01

2P

1.50E01

3P

1.00E01
5.00E02
0.00E+00
0.00E+00

1.00E03

2.00E03

3.00E03

4.00E03

5.00E03

6.00E03

Distancefrombase(m)

Figure9:VelocitiesprofileacrossthelengthofthecleatatP,2Pand3P.

6.ParametricStudy:

Parametric study was done by varying the values of diffusion coefficient, concentration of
methaneinmicroporesandconcentrationofmethaneincoalmatrix.

a)Microporesandmatrix:
10

10

15

6.1VaryingDiffusioncoefficient:Diffusioncoefficientwasvariedinarangeof10 (10 ,10 ,

20

10 )andtheconcentration/diffusivefluxprofilewasobserved.

13

Figure 10. Concentration across a line passing through the centre of the two micropores at
T=86400sandD=1010,1015,1020(fromrighttoleftthendown)

It was apparent from the concentration profile that the Diffusion coefficient is the guiding
factorfordiffusionofmethaneincoalmatrix.Ifthevalueofdiffusioncoefficientishighthen
the methane diffuses fast and tries to make the concentration uniform everywhere in the
matrix.

6.2 Varying Time:Breakthroughprofileforconcentrationanddiffusivefluxwasdrawnacross

thecrosssectionlinepassingthroughthecentreofthetwomicroporesatdifferenttimes.
It is clear from Figure 11 that, the concentration of the methane in coal matrix (central part)
keepsonincreasingwithtime.Itcanbeconcludedthatthemethaneconcentrationishigherin
centralpartofthecoalmatrixunit.However,ifonemorefracturecanbeinducedinthecentral
partofthematrixthentherecoveryofmethanecouldbeincreased.Methanegasreceivedat
cleatstravelsuptotheupperboundaryandhenceconcentrationofmethanedecreasessharply
aswegotowardscleatsandremainsconstantinthecleat.

14

Figure11.Totalfluxandconcentrationacrossaline(asshowninfigure10)passingthroughthe
centreofthetwomicroporesatd(T)=1000s.

b)Cleats:
6.3 Varying permeability: Permeability of the cleat was varied by changing the width of the
cleat(2w,3wand4w)inturnvaryingthepermeabilityinmultiplesof(8,27and64).Pressure
Surfaceandvelocityfieldprofilewasdrawn.Herewstandsforwidthofthecleats.Wideningof
the cleat represents high permeability and high permeability allows large velocities. Large
flowingvelocitiesresultsinhigherrateofmethaneextraction.

15

Figure12.Velocityfieldacrossthecleatatw,2wand3winturnchangingthepermeability8,
27and64times(fromrighttoleftthendown)

6.4Varyingconcentrationatcleatsboundaries:Pressurewasvariedas(2P,3Pand4P)atthe
cleatsboundariesandvelocityprofilewasdrawn.HerePstandsforpressureatcleat.Methane
concentration at the coal matrix boundaries increases with pressure (refer equation (8)).
Velocityincreasewithpressurerise(referfigure13).Thisagainsuggeststheenhancedrecovery
of methane. Concentration at the coal matrix boundaries can be increased by increasing the
temperatureofthecoalmatrixunit.

16

Figure 13.VelocityfieldacrossthecleatatdifferentpressuredifferentialsP,2Pand3P(from
righttoleftthendown).

7.Conclusions:
Itwasclearfromthemodelingstudiesthattheamountofmethanediffusedthroughthecoal
matrixperunittimecanbeincreasedbyreducingthedistance(adiffusedatomhastotravel
before finding a Cleat to escape). This will increase the production rate of the methane.
COMSOL seem to provide a reasonable breakthrough profile for concentration and diffusive
fluxbehavior.Sinceagoodfundamentalstudyonparametersarenotreadilyavailable,values
weretobeassumedtogetabreakthroughprofileforconcentration,diffusivefluxandvelocity
field.Itwasclearfromthemodelthattheextractionratefromacoalblockcanbeincreasedby
wideninguptheexistingcleats.Morewidenedcleatsreflectenhancedpermeability.Anyhow,if
pressure of the methane gas can be increased inside the coal unit (in turn at the cleat
boundary)itwillenhancetheextractionrate(increasedvelocity)asseeninthebreakthrough
profile.

Since, no values can be found in literature for such a model, it became quite complicated to
arriveattherealtimesituationwiththismodelandhencemodelisindevelopingstage.Also,
theprogramroutinelycrashes/takeslongertimetosolveforsmallsizegeometriesandvaried
rangesofdata.Allocatedmemoryinthesystemswontallowforfinemeshsizes.

17


AppendixA
Nomenclature:
C=Gasconcentrationincoalmatrix(Kg/m3)sc=standardconditions
D=Diffusioncoefficient(cm2/sec)

=Atinfinitepressure

k=Permeability,md
M=Molecularweight(Kg/kmol3)
P=pressure,psi(kpa)
PL=Langmuirpressureconstant,psi(kpa)
q=Volumetricflowrate,(stdm3/s)
qm=Massflowrate,(kg/s)
R=Universalgasconstant
t=Time(s)
T=Temperature,(K)V
VE=Equilibriumisotherm(Kg/m3)
V=LangmuirVolumeconstant(Kg/m3)
Z=Gascompressibilityfactor
=Viscosity,cP
=Density(Kg/m3)
=Porosity
Subscripts:
a=Macropore
b=Bulk
ai=MacroporeandMicroporeinteraction
E=Equilibrium
i=Micropore
L=Langmuir
r=Radialdirection
18

References
1.CreedyD.P.,SaghafiA.,LamaR.(1997),Gascontrolinundergroundcoalmine.Publishedin
IEACoalResearchJournal,(1997)

2. Markowski A.K. (1998), Coalbed methane resource potential and current prospects in
Pennsylvania.PublishedIn:InternationalJournalofCoalGeology38,(1998)

3.KarnF.S.,FriedelR.A.,ThamesB.M.,SharkeyA.G.(1969),GasTransportthroughsectionsof
solid coal. Published By: Pittsburgh Coal Research Centre Bureau of Mines US Department of
theInterior,(1969)

4.MacDonaldG.J.(1990),Thefutureofmethaneasanenergyresource.PublishedIn:Annual
ReviewofEnergyVolume15Virginia(USA),(1990)

5.HolditchS.A.(1989), EnhancedRecoveryofCoalbedMethanethroughhydraulicfracturing.
PublishedIn:SocietyofPetroleumEngineersTexas(USA),(1989)

6.WymanR.E.(1984),GasresourcesinElmworthcoalseams,Elmworth:CaseStudyofaDeep
BasinGasField,AmericanAssociationofPetroleumGeologistsMemoir,(1984)

7. McCants C.Y., Spafford S., Stevens S.H. (2001), Fivespot Production Pilot on Tight Spacing:
RapidEvaluationofCoalbedMethaneBlockintheUpperSilesianCoalBasin,Poland.The2001
AlabamaCoalbedMethaneSymposium,(2001)

8.SoederD.J.,Theeffectsofoverburdenstressoncoalbedmethaneproduction.PublishedIn:
Techbook,GeologyinCoalResourcesUtilization

9. Laubach S.E., Tremain C.M., Ayers W.B. (1991), Coal Fracture Studies: Guides for coalbed
methaneexplorationanddevelopment.PublishedIn:JournalofCoalQuality,(1991)

10.Kendall P.F., Briggs H. (1933), The formation of Rock Joints and the cleats of coal, Royal
SocietyofEdinburgh,(1993)

11. Mcculloch C.M., Duel M., Jeran P.W., Cleats in Bituminous Coalbed, US Bureau of Mines:
ReportsofInvestigationsNo.7910

12.DhirR.,DernR.R.,MavorM.J.(1991),EconomicandreserveevaluationofCoalbedMethane
Reservoirs:Societyofpetroleumengineers,PaperNo.22024,Texas(USA),(1991)

13.GrayI.(1987),ReservoirEngineeringinCoalseams:ThePhysicalprocessofgasstoragein
movementincoalseams:SocietyofpetroleumengineersReservoirEngineers,(1987)

19

14.GrayI.(1987),ReservoirEngineeringinCoalseams:Observationsofgasmovementincoal
seams:SocietyofpetroleumengineersReservoirEngineers,(1987)

15.HarpalaniS.,SchraufnagelR.A.(1990),Shrinkageofcoalmatrixwithreleaseofgasandits
impactonpermeabilityofcoal.JournalFuel,Volume69,(1990)

16.HarpalaniS.,SchraufnagelR.A.(1990),Cleatsandtheirrelationtogeologiclineamentsand
coalbedmethanepotentialinPennsylvaniancoalsinIndiana.PublishedIn:InternationalJournal
ofCoalGeology,(1990)

17.HarpalaniS.,ZhaoX.(1991),EnergySourcesVolume13,(1991)

18. Gamson P.D., Beamish B.B., Johnson D.P. (1992), Coal microstructure and micro
permeabilityandtheireffectsonnaturalgasrecovery.PublishedIn:FuelVol.72,(1993)

19.MazumderS.,HermentP.V.,BruiningJ.,WolfK.H.A.A.(2001),ApreliminarymodelofCO2
sequestration in coal for improved coalbed methane production. The 2001 Alabama Coalbed
MethaneSymposium,(2001)

20.AndersenM.A.(ChiefEditor),OilfieldReviewSchlumbergerAutumn2003,(2003)

21. Balan H.O., Gumrah F., Assessment of shrinkageswelling influences in coal seams using
rankdependent physical coal properties. Published In: International Journal of Coal Geology,
(2009)

20