Fig: 1- Location Map of study area

Fig: 2- Showing the bad hole effect on the logs

JMBR-43

Bad Hole effect

Fig:- 1

Feasibility study for discriminating Litho-fluid type using elastic logs estimatedthrough rock physics modeling to reduce uncertainty in seismic inversion study

– A case study

Bisht, B.S., Jhaldiyal Beena*, Singh RBN, Sanjay Singh ,GEOPIC (ONGC Dehradun), India*Email: jhaldiyal_b@ongc.co.in, birsingh.bisht@gmail.com

KeywordsFeasibility, Petrophysics, rock-physics, petro-elastic modeling, seismic reservoir characterization

AbstractQuantification of the relation between geological processand geoscientifice data is the most important aspect of anygeophysical study. Most of the time, intensive cost andextensive time consuming geophysical studies carried outwithout prior feasibility study conclude with ineffectiveresults. Since all the geophysical studies are observationbased and data intensive therefore a detail analysis of alltypes of data linked with the study is required. The ultimateobjective of any geophysical study is to delineate areservoir body which is interesting from hydrocarbon pointof view. The problem becomes much easier if the area ishaving sufficient number of well log data and othergeological information. Since the log data is recorded atvery high resolution therefore its ability to identify differentgeological process is much more than any of the recordeddata. But at the same time, during drilling process log datagets affected by borehole rugosity, invasion, mud cakeformation, salinity, temperature & pressure etc. Sometimesthe logs could be entirely missing or not usable due to badhole conditions. The analysis made on such unreliable datamay lead to wrong identification & estimation of litho fluidtype. To overcome these situations various type ofempirical relations, methods and model have beenformulated. These techniques will help to re-estimate thevarious responses of subsurface for geologicalinterpretation.Once the data has been conditioned for interpretation it isfurther subjected to feasibility study to assess theachievability of the project objective, whether variouslithologies with different type of fluids can bediscriminated at the log scale or not. For this , various typeof cross plots i.e between P-imp Vs Density and Vp/Vswith litho fluid color on Z-Axis are generally generated.

This paper highlights the important and deciding role offeasibility study to carryout seismic inversion study in thearea where no litho-fluid discrimination is possible throughcrossplots generated by conventional log sets, therebyresulting in inconclusive results in post-stack seismicinversion study. This study further highlights that even asingle well with proper DSI data can guide us throughfeasibility study, whether or not to proceed for rockphysicsmodeling to be use in Pre-stack seismic inversion.

This paper illustrates role of feasibility study on thescenario modeling of rock properties as applied in the studyof Oil & gas field Jambusar in the Cambay basin ofWestern onland field of ONGC, India. Fig-1Seismic inversion study requires set of elastic logs (Vp,Vsand Density) as input which are obtained through rockphysics modeling, which uses petrophysical evaluatedresults as input. Thus to achieve best results appropriateconstituent volumes, porosity and saturation are neededwhich is possible only through use of well conditioned logdata.WorkflowThe following sequential workflow has been adopted forthis process:Log data conditioningIn general log data is effected due to bad hole condition,Invasion effect, Salinity of the formation water and sometimes with tool calibrations also as shown in Fig-2 below.

11th Biennial International Conference & Exposition

Feasibility study for discriminating Litho-fluid type using elastic logs estimated through rock physics modeling

JMBR-43 (Density Comparison)Fig:- 2

JMBR-47 & 35Fig:- 19

Fig: 3A- Showing raw dataFig: 3B- Showing conditioning effect on data

Fig: 4- Petrophysical output of well A& B

Fig: 5- Cross plot of P-Imp & Density with Lito-fluids

Fig: 6- Cross plot of P-Imp & Vp/Vs with Lito facies

In certain cases quality of the log data is reasonable but insome intervals important logs are missing. Synthetic logsneed to be generated by multiple linear regressiontechniques based on the other logs to fill the missing datagap. Once the data gap is filled necessary environmentalcorrections are applied along with the appropriate filter forspike and noise removal. This conditioned data is nowsuitable for further analysis. Thus log conditioning plays avery crucial role in the data analysis and to get themeaningful output Fig-3: A & B

Petrophysical AnalysisThe objective of the petrophysical interpretation is totransform well log measurements into reservoir propertiesi.e. porosity, saturation, permeability, mineral componentvolumes etc. These parameters are responsible for oil/gasestimation and production. For determination of rockpetrophysical properties an adequate logging suit isnecessary which can measure the desired propertyaccurately. Once the Petrophysical model has been fixed, itwill be applied for estimation of the reservoir parametersi.e. Effective Porosity (Ф), Water saturation (Sw) andVolume of Clay (VCL) using multi-mineral inverseoptimization technique. This technique takes into accountthe effect of conductive or nonconductive, heavy minerals,radioactive minerals and different clay contents reported incore studies.Fig-4 shows the petrophysical output of thetwo wells taken into study.

Feasibility StudyThe feasibility studies are to assess the adequacy of the datato meet the objective of the project. The following stepswere applied for this study:1. The reservoir with fluid and non- reservoir wereidentified on the basis of petrophysical analysis of log data.2. The reservoir with desired fluid and non- reservoirwere differentiated with the help of elastic properties offormation. These elastic properties were obtained fromelastic logs viz Rhob, Vp and Vs, which are responsiblefor seismic response.3. Finding a relationship between petrophysicalproperties and elastic parameters of the formation bygenerating various kinds of cross plots.

Cross plots were taken between P_imp & Density for manywells of the study area. Unfortunately none of the well datagave any clue for the demarcation of the litho fluid inP_Impedance range as shown in Fig-5. All lithologies werefalling in the same acoustic range, therefore based upon thisanalysis, Post Stack Seismic inversion study will lead toinconclusive results. Now, to assess whether Prestackinversion is suitable to carry out in this area, DSI data wasrequired and hence it was decided to record DSI in a wellwhich may lead to carryout rockphysics modeling in otherwells where shear sonic is not available, for prestackseismic Inversion.

Based upon the recorded shear sonic a cross plot of P-Imp& Vp/Vs was taken which clearly showed the litho-Fluid

11th Biennial International Conference & Exposition

Feasibility study for discriminating Litho-fluid type using elastic logs estimated through rock physics modeling

Fig. 8: Well to seismic tie and Wavelet extraction

Fig. 9: inverted P-impedance, S-impedance and Vp/Vs

Fig-7: Shows the output of Petrophysics andRockphysics

demarcation in the Vp/Vs domain in the range of 1.55~1.65as shown in Fig-6Once the above relationship has been established with somekind of discriminator in acoustic impedance domain, thenthis can lead to forward modeling and synthetic loggeneration through rock physics modeling.

Rockphysics:

Rock-physics transform petrophysical results into elasticproperties that can be used for seismic interpretation. Thiscomplementary nature of Petrophysics and Rock physicsrequires a tight integration for seismic reservoircharacterization. Well log data plays a crucial role in thisprocess of integration. In general, most of the timespetrophysical evaluation, petro -elastic modeling andsynthetic to seismic tie are done separately prior tointegration. This introduces uncertainty and inconsistencyacross the geoscientific data. Ideally, Petrophysics andRock physics modeling should be an integrated process thatcan produce a greater consistency between all the data andlead to reduced uncertainty.The Rock Physics uses a phase drive mixing method orcombination of methods to combine elastic properties ofthe minerals and fluids that predict the measured elasticlogs i.e. density, P-velocity and S-velocity. The derivationof a modeling is complex task with iteration loop involvingrock physics, log conditioning and petrophysical analysis.The minerals and pores space are mixed and then the fluidsare filled, finally the other fluids are also introduced intothe porous mineral via Xu-White modeling.The Petrophysical model out put i.e. Volume of minerals ,Volume of clay, total porosity fluid saturation has beentaken into the rockphysics model to estimate the elasticlogs i.e. Vp, Vs and Density through velocity Mix functionusing Xu-White approximation. Fig- 7 shows the integratedoutput of petrophysics and rockphysics.

Application in Seismic Inversion

Rock physics modeled (RPM) logs were used for well toseismic tie & Wavelet extraction and interpretation of the

lithology with fluids Fig.8 & 9. The wavelets of all thewells under study were similar in characteristics and stablefor all angle stacks. Synthetic to seismic correlation wasalso good for all the angle stacks.

Discussion of results:

1.Feasibility Study:

Feasibility study at log level is essential beforecarrying out inversion study.

DSI log data helps to assess the feasibility forrockphysics modeling and prestack seismicinversion.

Feasibility study opens a way to carryout Pre-Stack seismic inversion study where Post stack isnot able to give conclusive results.

2. Rockphysics Modeling The rock physics modelled logs are free from

invasion and borehole effects.

11th Biennial International Conference & Exposition

Feasibility study for discriminating Litho-fluid type using elastic logs estimated through rock physics modeling

Cross plots of Pimp versus Vp/Vs shows cleardemarcation of gas.

Rock physics modeling and Pretsack Seismicinversion has been carried out in this field.

The modeled logs have been used for seismicreservoir characterisation

2. Inversion Study:

Gas sand polygon selected from cross plot ofextracted P-impedance and Vp/Vs of the wellsand is used to delineate most probable oil / gassand.

Inversion results at unknown well shows goodmatch with well log P- impedance and Vp/Vs.

Conclusion:

The elastic logs generated by Rock-Physics model havebeen used for well to seismic tie and low frequency modelbuilding. The petrophysical interpretation has thus beentransferred into elastic domain to understand thecorresponding seismic response in the entire area. Thedistribution of the reservoir sand at and away from the welllocation could be delineated from this simultaneousinversion. This approach has given an effective lead to theseismic reservoir characterization guided by rockphysicsmodel.

Acknowledgement:The authors are indebted to Shri Anil Sood, GGM-HOI-GEOPIC, Dehradun, India for providing the technical inputand guidance for writing this paper. Thanks are due to ShriDr. Hari Lal, Head-INTEG, GEOPIC, Dehradun, India forproviding all kind of support for this work. The authors aregrateful to the concern geoscientist of this domain forproviding necessary input during the project work andwriting this paper. We thank ONGC management forallowing us to submit this paper in 11th BiennialInternational Conference & Exposition of SPG-2015

The views expressed in this paper are solely of the authorsand do not necessarily reflect the view of ONGC

References

Serra Oberto, Well Logging. Vol. 3 Well Logging andReservoir Evaluation

An effective inclusion-based rock physics model for asand–shale sequence

Mark Sams1* and Thomas Focht 2

Mavko, G., Mukerji ,T. and Dvorkin, J. [2009] The rockphysics handbook: tools for seismic analysis in porousmedia. Cambridge University Press.

Joel Walls, Jack Dvorkin, Matt Carr, Well Logs and rockphysics in seismic reservoir characterization

Xu, S., and White, R.E., 1995, A new velocity model forclay-sand mixtures: Geophysical Prospecting 43, 91-118.

Jeff Baldwin, Tightly Integrating Petrophysics, RockPhysics in Single Model Generates Improved Results

11th Biennial International Conference & Exposition