KMS Technologies – KJT Enterprises Inc.

Strack, K. – M., Chunduru, R., Frenkel, M. A., Mezzatesta, A. G., Zhang, Z.

Well Log Inversion Review: Limits & Possibilities

1999

Presentation

Society of Exploration Geophysicists Taos NM, Summer Research

WALS97670a

Well Log Inversion Review:Well Log Inversion Review:

Limits & PossibilitiesLimits & Possibilities

Taos 1999Taos 1999K.-M. Strack, K.-M. Strack, KMS Technologies, R. , R. ChunduruChunduru, M.A., M.A.

FrenkelFrenkel, A.G. , A.G. MezzatestaMezzatesta, Z. Zhang, , Z. Zhang, Baker Atlas

OutlineOutline•• ObjectivesObjectives

•• Introduction: tools and methodsIntroduction: tools and methods

•• Practical implementationPractical implementation

•• Case historiesCase histories

•• Conclusions Conclusions

WALS97670b

PastPast PresentPresent FutureFutureModelingModeling

ImagingImagingModelingModeling

InterpretationInterpretation

Field LogsField Logs

EvaluationEvaluation

ImagingImagingModelingModeling

InterpretationInterpretation

Field LogField Log

EvaluationEvaluation

Field LogsField Logs

InterpretationInterpretation

IntegrationIntegration

EvaluationEvaluation

LogsLogs

ProcessingProcessing

KMS99001c

Designer wells Designer wells & reservoirs& reservoirs

Cost & value Cost & value domaindomain

The Paradigm Shift: log analysisThe Paradigm Shift: log analysis

Logging methods statusLogging methods status

• Nuclear: Mote Carlo routine, parametric in earlyphase

• Acoustic: inversion in toolbox; parametric modelsstarting, mostly non parametric

• Electrical: inversion in technology transfer phase• Borehole seismics: well advanced, 3D almost

commercial

Modeled

0 4000

REC

EIVE

R O

FFSE

T (M

eter

s)

TIME (Microseconds)

Comp. Shear: Fast Slow

3.35

4.42

TIME (Microseconds)1000 5000

Observed

Fast Slow

Monopole Shear Splitting: A Stress IndicationMonopole Shear Splitting: A Stress Indication

Acoustics logs: inversion as toolAcoustics logs: inversion as tool

(after Tang 1999)

Oil - Resistivity relationship: boreholeOil - Resistivity relationship: borehole

Rxo Shale R

Oil

Tight

Oil

1/2 Oil1/2 Water

Water

t

′′V

V

′V

• Objectives

•• Introduction: tools and methodsIntroduction: tools and methods

• Inversion methodology

• Practical implementation

• Case histories

• Conclusions

WALS97670c

OutlineOutline

WALS961104e-II

Receiver

Groundloop

Transmitter

InductionInduction LaterologLaterolog

Combination PossibilitiesCombination Possibilities

focusing

measure

monitor

+

-

+

+

+

+

-

-

-

-

HDLL

′′V

V

′V

Joint: Induction & GalvanicJoint: Induction & Galvanic

HDIL

WALS97201c

RRmmRRt, n-1t, n-1

BHDBHD

RRxo, n-1xo, n-1 RRt, n-1t, n-1

RRt, n+1t, n+1

RRt, nt, n RRxo, nxo, n LLxo, nxo, n

boreholeboreholeLLxo, n-1xo, n-1

RRt, nt, n

RRt, n+1t, n+1

2-D Earth Model2-D Earth Model

RRxo, n+1xo, n+1 LLxo, n+1xo, n+1

WALS961104p-I

Geological formation

Logs

Earth model describedby parameters(resistivity distribution)

AWS951357

Measurements

WESTERN ATLAS

WESTERN ATLAS

Logging processLogging process

KS92

Geological formation

Logs

Earth model describedby parameters(resistivity distribution)

AWS951357

?

Modeling

Inversion

Modeling process for logsModeling process for logs

KS92

Parameter subspace 2

Parameter subspace 1

Initial Guess

Nonlinear OptimizationNonlinear Optimization

WALS 9710021dd-IWALS 9710021dd-I

Data AcquisitionData Acquisition Data ModelingData ModelingFieldLogs

ModelLogs

Acquisition

Response

Excitation Acquisition

Response

Simulation

WALS98208u

Inversion: Process flowInversion: Process flow

AWS95727b-I

from The Log Analyst, 1994

invasion radiusinvasion radius

100 0.2 200

est. riest. ri

est. Rtest. Rt

est. Rxoest. Rxo

true Rxotrue Rxo

true Rttrue Rt

DLLDLL

DILDIL

0 in. ohm-m

riri

100

150

200

ft2-D Inversion Results

2-D Inversion2-D Inversion Results Results

OutlineOutline• Objectives

• Introduction: tools and methods

•• Practical implementationPractical implementation

• Case histories

• Conclusions

Rtban Induction log exampleRtban Induction log exampleRtban Induction log example

300X320

X340X360

DLL

Rxo

Rt

2-D2-D

Rxo

DLL

RIPM

Rt

X300X320

X340X360

DLLDLL

DLLDLL

RxoRxo

RxoRxo

RtRt

RtRt

LxoLxo

LxoLxo DPILDPILDPILDPIL

1-D1-D

Inversion: 1D versus 2DInversion: 1D versus 2D

WALS98209e

OutlineOutline• Objectives

• Introduction: tools and methods

• Practical implementation

•• Case historiesCase histories

• Conclusions

WALS97670e

1-D Statistics1-D Statistics1-D Statistics

LxoLxo

X500

RtRt

RxoRxo

WALS96745b-I

Rxo Error Bounds Rt Error Bounds

0.2 200Lxo0 50

Lxo Error Bounds

Rxo

Importances

2000.2

Rt

WALS96745c-I

Water

Porosity0 25Percent

Perm0.1 1000

PercentPermSW

Gamma Ray0 100

(API)App. grain den

2.5 8.5(gm/cc)

0 100Percent

Volume0 100

Lime

Sand

ShaleHydrocarbon

Immovable HC

X500X500CRA

(Rt from 1-D inversion)CRACRA(Rt from 1-D inversion)

WALS96745d-I

Hydrocarbon

Immoveable HC

WaterPorosity0 25Percent

Perm0.1 1000

PercentPermSW

Gamma ray0 100

(API)App. grain den

2.5 8.5(gm/cc)

0 100Percent

Volume0 100

Lime

Sand

Shale

Water

Porosity0 25Percent

Perm0.1 1000

PercentPermSW

0 100(API)

2.5 8.5(gm/cc)

0 100PercentVolume0 100

Lime

Sand

ShaleHydrocarbon

Immovable HC

X500

CRA(Rt from 2-D inversion)CRACRA(Rt from 2-D inversion)

App. grain den

Gamma Ray

RxoRxo

WALS979516c

RsRsMLLMLL

Dual Laterolog underestimating oil reserves

Dual LaterologDual Laterolog underestimating oil reserves

RdRdLxoLxo

calipercaliper

RtRt

Gamma rayGamma ray

RtRt

SFLSFL

TBRTTBRT

TBRT inversion& SFL data

TBRT inversion& SFL data

TBRT & SFL datasynthetics fit

TBRT & SFL datasynthetics fit

Thin bed analysisThin bed analysis

WALS974012WALS974012

Reserves estimate Reserves estimate DPIL vs. HDILDPIL vs. HDIL

Net Pay and Saturation Analysis

DPIL HDIL

Reservoir Thickness: 270 ftNet Pay (ft) 103.6 130.1Net Pay 38.4% 48.2%Por. Feet 15.4 ft 18.9 ftHyd. Feet 7.4 ft 9.2 ft

HDIL data allowed 24%

more OIIP be booked.

HDILHDIL

DPILDPIL

HDILHDILDPILDPIL

Step change through hardwareStep change through hardware

WALS98209g

2 ft VRM vs 2D Inversion

OIIP = (A*h)*Por*(1-Sw)A= 160 acreAssume 7,758 API Bbl/acre-foot1 Bbl = $22

2 ft VRM Curves 2D Inversion

h 67.00 ft 74.75 ftPor 25.8% 25.9%

(1-Sw)69.5% 67.3%

OIIP 14,912,427 Bbl 16,173,193 Bbl

Value 328 M$ 356 M$

HDIL VRMHDIL VRM

2D inversion2D inversion

volumetricsvolumetrics

porositiesporosities

SwSw

WALS97201g

Step change through softwareStep change through software

XX

Comparison of HDLL and MLL/DLLComparison of HDLL and MLL/DLL

Rt Rt

Gamma rayGamma ray

calipercaliper

LxoLxoRxoRxo

MLL MLL

LLsLLs

LLd LLd

DLL DLL HDLL HDLL

WALS98209j

Focused & Raw Data InversionFocused & Raw Data Inversion

Rt Rt

Gamma rayGamma ray

calipercaliper

SPSP

RxoRxo

MLLMLL

MLLMLLRxo-sRxo-s

RtRt

Rxo-m Rxo-m

RawRawfocussedfocussed

WALS97201d

Reservoir Analysis HDILINV MPRINV JOINT

OIIP Estimates 10% 0.7% 12%

Reservoir Analysis HDILINV MPRINV JOINT

OIIP Estimates 10% 0.7% 12%

XX00XX00

X020XX40

Estimation of OIIP: HDIL & MPREstimation of OIIP: HDIL & MPR

XX00XX80

X020XX40

100 0100 0 50 0 100 0 Sw SwHDIL MPR

Por. Vol.

Reservoir Analysis HDILINV MPRINV JOINT

OIIP Estimates +40% -1.3% +43%

Reservoir Analysis HDILINV MPRINV JOINT

OIIP Estimates +40% -1.3% +43%

Estimation of OIIP: HDIL & MPREstimation of OIIP: HDIL & MPR

HDLL

′′V

V

′V

Joint: Induction & GalvanicJoint: Induction & Galvanic

HDIL

WALS97201c

Rt (joint)Rt (joint)Joint Inversion ofJoint Inversion ofHDLL & HDIL DataHDLL & HDIL Data

Rt (HDLL)Rt (HDLL)

ImportancesImportancesGamma rayGamma ray

calipercaliperWALS97201b

HDLL & HDIL Data FitHDLL & HDIL Data Fit

Gamma rayGamma ray

calipercaliper

First diff.First diff.Sub volt.Sub volt.

WALS97201a

Formation Resistivity - LWD &Formation Resistivity - LWD &WirelineWireline

• Ability toenhance bothLWD & Wirelineresistivity data

• More resistivezones translateinto more OIP

Joint inversion of MPR & HDIL dataJoint inversion of MPR & HDIL data

JointMPRHDIL

KMS99002f

OutlineOutline• Objectives

• Introduction: tools and methods

• Practical implementation

• Case histories

•• Conclusions Conclusions

WALS97670e

Log inversion benefits: Log inversion benefits:

• Better delineated oil bearing zones

• More accurate formation parameters

• Risk analysis parameters

• New ways for data integration & upscaling

• BUT: Sensitivities, Uncertainties???

⌦CLOSING THE LOOPWALS97670f

Closing the loopClosing the loopInversion

PresentationGround truth

Statistics /Analysis

Evaluateanalysis

IntegrationGeology

Petrophysicist

Inversion expert Geophysicist

Geologist

YES

NOdiscard

NO

YES/NO

NOdiscard

YES

$$value

$$value

WALS974011b

AcknowledgementsAcknowledgements

D. Beard, R. Busch, M. Eckard, G. Itskovitch, V. Koelman, A. Fabris, T. Fischburn, J. McDougall, M. B, Rabonovich,L.A. Tabarovsky, T. Tamarchenko, R. Truman

and others

data from:ADCO, Agip, Chevron, PDO, Shell, Texaco

KMS Technologies – KJT Enterprises Inc. 6420 Richmond Ave., Suite 610

Houston, Texas, 77057, USA Tel: 713.532.8144

info@kmstechnologies.com

Please visit us

http://www.kmstechnologies.com//