8/3/2019 26060

1/10

~SPE 26060Permeability Evaluation in Heterogcmecw Formations Eiocldy of Petroleum EngineersLking Geophysical Well Logs and Geological InterpretationsS. Ameri, D, Molnar, S, Mohaglmgh, K. ArninianW8st Virginia UniversitySPE M-embersCopy rbhl 1003, SOCI* IY of PcI roM.bn Ewlneam Inc ,Tti BP wu pmpwul for pwnonwtloa at ttm SW W-mm RwIwJ81 MOMIIXI held In AI)clwrEua, Alaokm 20.28 M#y lBLt3,WA pap+r ww dtcwd (w wwattoth by w) SW PrqIram Commitlea fcilowlcu rwhw of Irdawnttirm cc+tdnocl In an @&tract submitted by ttw $tdhorlc), (k+utmtsof tho pe+.w, a prmwtted, IWVOnot k+en mvbwe.d by ttM SLci eIy of Potrolmim Engirw.m sod U* a~}oct to coIrBc licm by tha aut lmrlol. Tfw mat add, a pfwnted, doesnot c4ws9st iy rat lat cay posit ion of the Societ y of Patrolawn Em2i*rm 119of f imrq w membwsc f%wa p remed qt WE meet itwe am Eubject to po!dlcatlon r rnviowby Edi!wid Cwnmitt- of ttw Sochty o! %!roleum Emalrwwo, Pwmiwion to copy i o !eauictw! w an *tract of no mom ttmn 300 words. Illtmtmtim mby not be cop ied,The * tract dw+dd conts in conapicuout w+uwwludomant of whmo am! by wtmmttn PJ+MWla pwe.nhd. Writ . Llb+tr im, $ (W, P.O,Box 833938, I ll chwdsom (X 75003 .3836 U,S,A, Tolw(, Y30f#t39 3PEDAI. ,

ABSTRACTPermeability is the key parameter forreservoir characterization. The permeability

of the formation is usually evaluated fromthe cores and/or well tests. It shou!d benoted that cores and well test data areoftwr only trvai!able from few wells in areservoir while the well logs are availablefrom the majority of th~ wells. Therefore,the ovaiuation of permeability from well logdata represents a significant technical asweli as economic advantage.

The evacuation of permaabiiity inheterogenwrs formation from well iog datahowev~r represents a diffim.rit and cornpiexprobiem. CXmeraiiy, a simple corrwiationbotwmm pwme~biiity and porosity cannotbe cieveioped in h~terogenous forrnatiwl.The goai of this study has been to dovciopa goncraiized mothudoiogy to dotmmirro t,hopcrmoabiiity of a hetwogcrwous formationutilizing geophysical wcli logs as wcii asother gooiogicai information,

Granny Crcok Field in West Vir{linioiws boon soioctod as tim ttudy area ii] t,!]i$paper. Ti~is fieid has producod oii from Riginjun Formation sinca oariy 1900s. Timwatmrfiooding qmration was initiatmi in1970s and mrrrr.wtiy is in ilrogross. Wciiiog data ara availtlblo on substantialnurnimr of weils. Corn samples ark? aisoavai!nbio from ~afnw wolis. Corm snmplos

and the well iogs were anaiyzed todetermine permeability, porosity and watersaturation. The resuits of core and ioganaiysis were corirpiemented by goologicaiinterpretations to deveiop a correlationbetween permeability and weil iogresponses. The resuits presentmd in thispaper could serve as a guidciine forcorrelating permmbiiity with weil iogsresponses in heterogeneous formations. Asystematic and synergetic approach whichintegrating data from various weli iogs asweii as depositional, iithoiogicai andsedimentoiogicai interpretations has beendeveioped to cwaluate the permeability.iidTRCIDLJCTIC)N

Reservoir ciwractarization play acriticai role in appraising ttw f~conomicsuccess of rosarvoir managwncnt anddovoiopmcnt motilods. Tim prediction ofpormaabi~ity distribution is tho most criticnlaspect of reservoir (:i~aroctt:riz:iti otl. Noariyali rwmrvoirs show somo dogroo Llfiloterogrmaity duo to tiw contrastirlgIitiloiogiflsf (ii!~orw$;is, or :;(3Liirn[?t7toit){~ic;:]lcornp!oxity. Ci][lrtictc)riztltioil ofhotwog[3noous rcsorvoirs is H cornploxPmbiomom probiem stems from tho factti-)at sufficiorrt data to mmurot.oiy predictpwmcabiiity distribution is not umialiyavail ai)io,The permonbiiity of tim formntirmis usuaily ovaillatod frum ttm corw+ and/orprowwrw transirmt tmsts. (hros ;]nd wnli

8/3/2019 26060

2/10

2 PERMEABILITY EVALUATION IN HEWROGENCOUS FORMATIONS SPE26CJ60USING GEOPHYSICAL WELL LOGS AND GEOLOGICAL INTERPRETATIONS.. .

test data however, are available only fromfew walls in a reservoir, At the satna time,the geophysical logs from the majority, ifnot all, of the wells in the reservoir areusually available. Consequently, thoevaluation of permeability from well logdata represents a significant technical aswell as economic advantage.

Prediction of permeability fromporosity begins with the relationshiphf3tween core porosity and corepermeability, which is then generalized bycalibration of WQIIlogs so that permeabilitycan be prf Jicted from log porositythroughout the reservoir. The attempts topredict permeability from well log data hasgenerally been in the form empiricalcorrelations between permeability,porosity, and water saturation (1). Thistechnique has been usad with somesuccess in sandstone (2) and carbonatereservoirs (3,4). However, the exitingcorrelation are mainly for homogeneousformations that have fairly constantporosity and grain size.

fhQ objective of this study is toinvestigate the feasibility of evaluatingperm~ability distribution for a heterogeneousreswvoirs utilizing geophysical wall logs aswell as geological interpretations. GrannyCreak field in West Virginia has beenselected as tho study area, The producir~ghorizon in the Granny Croak fioid is thoupper Pocono Big lnjun sand of LowwMississippifin age that is charactwizod bysovw Iwttlrogmwity.

The Granny Creek oil fir!ld is locatedapprwximatdy 25 miles northc[~st ofCharlrxNxm, West Virginia (sw.1 Figuro 1).Tlm field is located structurally WI thonorthwest flank of a syrwlirm which strikm+

N 15-20 degrees east to S 15-20 degreeswest,Th8 oil accumulation was partiallycontrolled by porosity and pmrmabilityvariations as WGII as W.ructumi Theproducing horizon irr the Granny Creek fieldis ttw upper Pocono Big Injun sand ofLower Mississippian age. The Pocono f3igInjun sandstone has been subdivided intothree informal members (SW Figure 2) thatcorrespond to grain-size distribution andbulk density variations(5,5):(1) A member: the upper coarse-grainedsandstone and conglomerate (low densit)of the channel facies with generally goodporosity and permeability.(2) B member: the underlyingcoarse -grained sandstone andconglomerate (high density) with poorporosity and permeability.(3) C rnembw: the basal fine-grcrinedsandstone (low density).

Laterally, the C member consists ofprqrading tongues, numbered from oldestto youngest, respectively, as Cl, C2, andC3 within Granny Creek field. The Cmember and its tongues represent a faciesdeposited in a deltaic riwr-mouth barenvironment and the subfacies we thedistal and proximal parts of the bar, furtherdistinguish~d by whether dominated bymarirw or fluvial processes (sew Figurf3 2).~ho best porosity and permeability inGranny Creek fitdd occur morw consistentlyin tho proximal inouth-fm facies of tho Bigltljun(6).

The digonesis pkiywf an rmpr.:ciailyimportant rolo ir} contritmtir~g to tht;hctorogwwity of ttm Big Injun reservoir inGranny Cr!~ok field, particularly in twins ofporosity and parrrwahility. Tho initialporosity and porrneabilit.y of tho A and Bmembars p~ObiJbly, W:J:> hi~}t, bhtrxrno[ltation of tile Elmember during burii; lproduced a digwmtic faci~s. Tlm rni~infactor in porosity proservatiorl in tho

8/3/2019 26060

3/10

1

( SPE2tXlENl AMERI,S.,MOLNAR, C).,MOHAGHEGH,S., AMINIAN,K. 3proximal mouth-bar interval was thedevelopm~nt of well-formed chloritecoatings that restricted quartz cementation.A combination of those factors came intoplay to lower porosity of distal mouth-barfacies to some extent and drasticallyreduce permeability (6).

The field was developed ovw aperiod of nearly 30 years beginning in1916. Production hascontinwd throughoutmow of the field until the present day. Thefield is roughly five miles long, has amaximum width of a little over two miles,and has a total productive mea of about3,000 acres . The crude oil in GrannyCreek field is a paraffin base, PennsylvaniaGrade oil. It has a viscosity of 3.14 cp atatmospheric pressure and 7!5F,, and aliquid gravity of 45.4API at 60F. Total oilproduction is estimated to be between6,500,000 and 6,750,000 barreis.

The waterfiooding operations inGranny Creek Fieid were initiate during the1970s and ewiy 1980s. The waterfloodhas been rnoderateiy successful, However,the waterflood ar~ai and verticai sweepefficiencies have been poor due to theheterogeneous nature of the formation, Atertiary recovery COz piiot project wasconducted b~ginning in 1976. Because ofthe extremeiy hdwogeneous nature of thereservoir formation, iess that 4 percent oftlm injacted CXlz entwod the pattcm. Eventhis .srnaii amount was respcmsibie for thoproduction of over 4000 barreis of oii fromwithin tho pattern. This mcovwy wasconsidered very good under thecircumsttinces. A rninitcst COZ project wasconductod in a part of the same patternstwerai years iator. A wnaii amount ofaclditiormi oii was producod. The C(IZ fioodhas not bwn expfindcd hccnuse of pooroconomic$.

METHODOLOGYThe objective of this study is topredict pwmeabiiity of Big injun Formation

in Granny Creek Fieid from weii iog data.Due heterogeneous nature of the Big injunFormation, it was necessary to divido thoformation into practicai subunits (or zones)that intwnaiiy show a trend in permeabilityvariation and homogeneity with respect tofacies content. Subsequently, thepermeability variations in each zones wiilbe studied as function waii log data toinvestigate the existence of a correlationbetween permeability and iog data.

The resuits of the whoie (or fuiidiameter) core anaiysis on 7 centraiiyio~ated weiis in the Granny Creek fieidwere avaiiabie (see Figure 1). Gamma Ray,induction, and Density Logs were coiiectedfor aii these weiis. The zones wereidentified and deiineate based on geologicalinterpretations (see Figure 2). Ti-w resuitsof the whole core and iogs armiysis forvarious zones (based on stratigraphy,iithofacie, and depositional environments)were studied to deveiop correlationsbetween permeability, porosity, watersaturation, depositiormi environmerrt,andpore type. However, satisfactorycorrelations couid not be deveioped. Timinabiiity in finding a correlation betweenpermoabiiity and iog data was contributedto insufficient accuracy of the whoie commmiysis, This is not to say that tilephysicai measurements are irwccurate.However, tht3 fuli diarnctw coro anaiysisrosuits represent the avarngo rockproperties over tile intwvai of study, As arcmuit, the whoic coro ar~aiysis iws atendency to ignore the rlipid cimngrx inrock propwties that arw common toiwtwogr]nous formations,

in order to aiicviato tim mmragingprobiwn witi~ whoio mm? anaiysis , two

8/3/2019 26060

4/10

4 PERMEABILITY EVALUATION IN HETEROGENEOUS FORMATIONS SPE26060USING GEOPHYSICAL WELL LOGS AND GEOLOGICAL INTERPRETATIONS .

wells, were selected for detailed plug (orconventional) core analysis, The twoinvestigation wells, as shown in Figure 1,are located cm the most easterly (well15-1 110) and the most westeriy (woII15-11 34) sides of the field. Figures 3 and4 iliusrrate the stratigraphic/lithologicinterpretations for the two investigatiorlweils. lt is apparent that thestratigraphic./lithologic interpretations arenot similar. For example, zone C3 in the1 !5-1110 well corresponds to the three bardepositional environments, In the 15-1134well, zone C3 is absent and zones C2 andCl correspond to these same deposits.

Core pltg were taken atapproximately 6-inch intewals throughoutthe length of the core in both investigationwells. The permeability and porosity valuesof the core plugs were measured in thelaboratory. The comparison of themeasured porosity values for core plugsand the porosity values evaluated fromwell logs indicated the need for som~adjustments to overcome the inherentinadequacies in coring and core handlingtechniques. In other words, the core wasmoved up or down to provide a goodmatch between porosity values determinedfrom the core and log analysis. Figure 5shows the comparison of core and logdetermined porosity va!ues for well 15=1 1 1 0 ,

lho measured permeability valuesfor the primary immstigation WC!ISand thelog response values of bulk density,rwsistivity, and gaimna ray aro illustr:~tmi inFigures 6 tind 7. A.ctuol log response valueswore solcctsd in order to miniirrize anyassumptions that are mwdod fordetermining porosity and water saturationfroin well logs. The permeability val~ics asfunction well log responsos wem studiedfor previously de finr}dstriltigraptlic/li ttlol{~[]i[; zcmns. Altlw{qh,

the results of data evaluation indicated ageneral trend might exist betweenperrneahility and bulk density but, thescatter in data points was significailtenough to preclude possibility ofdeveloping a correlation. This failure todevelop correlations can be mainlycontributed to qualitative nature ofgeological interpretations relative todepositional environmei7ts, grain-size,stratigraphy, and iithology. In other words,the boundaries of the various zones areapproximately defined. Therefore, it isnecessary to integrate the geologicaldescriptions of the various zones,geophysical well log responses and th~trend of tlw permeability variations in ord~rto define ttie zones quantitatively.

The comparison of log responsesand permeability values on the twoinvestigation wells indicated similarities onpermeability variation ancj the responses indensity, induction, and gamma ray logsfrom well to well. As a result, severalzones were delineated in terms of logresponses and annotated as Garnrna Ray-Induction-Density (G. I.D.) zones 1,Transition, and 2, Zones 1 and 2 werefurther subdivided into 1A ,1 El, 2A and 2B.As illustrated in Figures 6 and 7, %one]Abegins with the first cross over of inductionand gamma ray log responses arrdterminates wlwn they cross over again.Zone 1El initiates at this socoi~d cross ovorand termirwtrx at the i~ext cross over ofinduction aiKJ gamintr ray responses. Thetransition zone starts at the last cross ovorand coiltinucs as density and induction logresponses follow a decreasing trrm-f whilogamma ray response increases aild thoridccreasos. Zono 2A is chamctmrizm.f byrciativoly constant induction afld gainmaray log responses. When thu induction tmdgamina ray log rospoi~ses begii~ to divwgozoiv? 2R begins and continues to tho end oftho coro, Fi!]~ircs 3 and 4, coinpt]re the

8/3/2019 26060

5/10

SPE26060 AMEF?I,S., MOLNAR,D., MOHAGHEGH,S.,AMINIAN, K. 5G.I.D zones, as defined in this study, forthe two investigation w6!Is with othergeological interpretations,

RESULTS AND DISCWSSIC9NFigures 8,9, and 10 illustrate the

permeability values for the twoinvestigation wells that are plotted versuskMlk density for G.I. D. zone 1, transition,and zone 2, Although there are few datapoints in zone 1, a trend can be postulated.lhe data in th~ transition zone, asexpected, demonstrates lack U:permeability - bulk density correlation. Theexistence of a well defined relationshipbetween permeability and bulk density forzone 2 is apparent in Figure 10.

The data available from the 7 wellswith whole core analysis, that are locatedbetween the two investigation wells, wereutilized to evaluate the applicability of thecorrelations dc veloped in the previoussection. The G.I.Q. zones were firstdelineated utilizing the !og responses oneach well. Figure 11 illustrates the logresponses for one of these welis (15-1107). As it can be seen, the establishedtrends of the log responses are prasant inthis well (and ail other wells). Therefore,the various G.I.D. zones can be readilyrklineated for each well, Figure 12 and 13compare the results for zones 1 and 2 withthe establis!md correlation from the twoinvestigation wells. Figure 12 indicmte%t.hatthe whole core anzr[y.sis data have thosame genr.val trend (siopa} as cornpwc totho correlation for zorlf? 1 but the dampoints aro shifted toward higherpermeability values. Wwcral factorscontribute to this situation, First, ttwcorrelation for zonel is cfevelopecl fromIirnitmd data and is thoroforo unreliable atthis point. Second, tlm perrnrmbility valuasdetermined from whole com analysis tend

to be somewhat optimistic. It should befurther noted that zone 1 M the lessproductive and/or unproductive part of theformation so it is of less interest. Figure13, however, illustrate a fairly goodagreement between the data points and thecorrelation. Considering the nature ofwhole coro analysis data, the closeagreement can substantiate the reliabilityof this correlation.CONCLLBIOIW

The followings conclusions werereac;hed in this stt.rdv:1. It is plausibfe to deveiop a correlationfor predicting the permeability from welllog responses in heterogeneous reservoirs.2. Whole core analysis is not sufficientlyaccurate for developing correlations inheterogeneous formations.3. It is necessary to use detailedconventional core (F.eg) analysis irrheterogeneous formations.4. It is necessary to integrate thegeological interpretations, geophysical welllog r~sponses and the trm~d of thepwrrwabi!ity variations in order to dividethe formation into the zones for correlationpurposes.REFERENCES1. Timur, A.: AI? investigation ofPerrnoabilityr Porosity, and WaterSatwotio~ I Rclatiorwhip for $%rtvlstoncRwwrvoirs, Tlm Log Analyst, VOI.9, No.4,196[1.2. Wubor, K. J., and VanGcutw,L.C.:Frnmowork for (lmstrllcting ClosticReservoir Simulation model, JPT, Vol.42, No.10, 1990.

8/3/2019 26060

6/10

6 PERMEAfllLITY EVALUATION IN HETEROGENEOUS FORMATIONS SPE26060USINO GCOPNYSICAL WELL LOGS ANO GEOLCGICAL tNTERPFiETATioNs..- ..-.2~. Lucia,F.J. : Pet.rophysical ParameterEstimated from Visual Description ofCarbonate Rock : A Field Classification ofCarbonate Pore Space,JPI_, Vol. 35, No.3, 1983.4. Licia, F.J., and Fogg, G.E.:Geologic/Stochastic mapping ofheterogeneity in a carbonate reservoir, JPT, Vol. 42, No. 10, 1990.

5, Dcma!dson, A., et :]1.: Measuring andPredicting Reservoir Heterogeneity irlComplex Deposystcms, Annual Report,DOE/BC/14657-7, August 1992s6, Donaldson, A,, et al,: Measuring andPredicting Reservoir Heterogeneity inComplex Deposystems, Annual Report,130E/BC/14657, August 1993 (In print).

_-.. .A.-1 3_. .C3C2c 1, .........IONNJ

.-------

[1IiAN__-.HPB.,.__...-MIPB.. . ,--,.-ml? I.-...-J.. -

COMSE.. . _ - ..- - -FINE........---

8/3/2019 26060

7/10

. SPE26060

DE PM[ t Z ONE E NV I RONME NTTHOFAC I EGM)I 890

1900

1010

1920

1n30u......El..( x lu...HAN. .F I P1 3MI P BDIST :1.OARSE.F J I NE. . . , - ..1A 1..--.-IEI--..-TRAN$ I UON- - .2A..2 B.. . .l l T HOL OGI C/ ST MT ! GMPHI CNT EHP I I 7 A T I ON

DE PT H, ,Z ONE[ I ! r ! l o l i l f l iI 1 } I OF ACI Ei l l ) ,2195

2205

2215

2225 Ii

-.-..B-...c d. . -- .c 1. .h.---NAN1.FIPEI.--.MIP9NH...L..COAN,FNIE... .. ....II............B...-T I I N$ I I I O- - 2A2 6. .

L l r t l OL OGI C/ ST WT l GWt 4 1 CNTERPRETAT I ONFigure 3. Geological In terpretat ions for Well 151110 Figure 4. Geological Intorprotations for Wdl 15-1134

Iirjuto h. (h-w Iormity w. I O(I Porosity for W(:I I 15-1110

8/3/2019 26060

8/10

*,000 -

1oo-

lo -_DFN,,Ty#. //--ESISTIVI?Y ~--GAMMA RAY PERMEABILITY

0.1 -

10

/ - .~[:~(J,~y4.--/-.mlwm 4- GAMMA HAY=

1 ,t)m 1,890 1 ,09!; 1 ,!300

i ! , l !JO

--. . .. _____TRAN

I

2AQ-- -%.------------q+ @ aa% [ M c.)u ~, ~ ( . ; . , p m-; --r-----------=8 %n .f-

2.7

---2.6

2.6 -ii>.~,:m2!U!

-- 2.4 ~i:3Crl

. . ~,:j

LJ- ., ;~.~1 ,$lo!l 1,910 1,915 1,920 1 ,!?:?5 1 ,s:30 1,935

DEPT1{ , ft

Fiqur~ G. (km PIw Prmvmbility and I.og Wxponscs for Well 15-111()

,.... . _,-___ -_-fii 1TI?AN

], . l , . , . . , , . , , . l,,,!,,.,

-.....-.-_-_._-...-...2?/4

r - .,,.-...,-,.-,,,-..-,._,....-.

-.. .....------..... ,-..--- ....

rl

l li:! ,!.?00 :$!,20!; f.!,2! 1(J ; !, :! 1! ; $,,2?,)0 :! ,9:!!) :!,; !:10

8/3/2019 26060

9/10

q

._.-..-.. .. .

\a.

COR II E: A TI ON* WEL L 1 11 0 Z ON E 1 AM WELL 1t t020Nl ? 18. WELL 1134 Zofit i I t !

\

l., .. ... -.:

..

N.

.. L.w.u.i- l * . . + . Wy . . k . y c . y w , , , , + J . A2.7

.. . . ... ......,.---.--..-~ -., ..~.. . . . .-. _1,,=~~ ,,,.,. wmL, ,,*----------

l :igur t3 8. Ikjrmcal)ility vs. Bulk KleIlsity for 7.ont: 1[:igLjre $), porf[)~;;]l)ility V S. [ 31 jl k [)msity for fransilinn ~Oll[!

-i!

.

- .-... .-.. .-,... .. .. .. . . . . .. .. . . . . ... . ... .. .-

[,

-- n o? l n KL AT< r . >N- WEL L * I 1 0 x c I N r mm

\

M WEL L . , , . ZOWW q -.!J . . Ws l , l . , , * : ONK . A= , I ?. M Wu t . 1. , , s . Z o ? dt l a e- . - . . . _ _ _ _ _ _ _ _ . . . . . . . . . .: , . ,I:q M:3-e,..%

\

*, m

,1 ~ w

\

M

1

8/3/2019 26060

10/10

VI.H~: -- t::

SPE26060

1,000--- = . . . .. .--...-..-..-- -.-------=-.. .. .. ...... ..-..- ..1A ~ lB ~ r IT -2.7TFWJJ 2A 213 :,l/ \ l1 dL .J,+--- -. :?.6L-d

u I(JLJ ] 1/ .1..2.4g

1Ou

:1..,..::. , -,,~...~~~,Ty .,.-- />. =- fwslswrlY-=- ,/.~~-GAMMA HAY4Kl pERMEA~lL/ry

&J . L . L ~ - U- , - . L . t L J - J - J . . +Q. Q - + . 1 . . L. L . L . , . . . 4 - A . . . L - . . . p. . l . . l . . , _, ~ , : ?2,11s 2,120 2,125 2,100 2,1:1!, 7,140 :?,14L !2,1so 2,1 $!; 7,160

DFYI-}1,tIiwro 11, Whoffr Gm Pornml)irity and Log f?m~wrws for Well 1 ii-l101

\.,,,,

II o

.- ....-......-.--------- CORRELATIONQ WI{OL. C C(2RC ZONt: 1

. . . ...-. .. ...... . . ..- .. . . .

\, ,,

,, 0

,, ,> v

/

.) o II,,,l? ;,

m ,>

,, ,,b Z*,>

,) ,,,41

,! ,,,)

( ,,

,,

!!; ,l !, i ,,..; !,,, ,,l l ,,!,. !l ,, l ,l ,,l !! l I l !,l j l l l

1 l x) .- .- .. ......-. - -----~.-----...---------.-------- .-1- CO RIICLATION, ] Wll(ll.f< CX)l+l: ZONI1 2!\

1. . . .. . . . ..... . . . ..

\i, ., ,,,+,

,, 1) ,,,> t,\i