Earth models for early exploration stages · Geostatistical Seismic Inversion algorithm - GSI (Soares et al. 2007; Caetano, 2009)). Use of geological analogs for extraction of a priori

ANNUAL MEETING MASTER OF PETROLEUM ENGINEERING

Ângela Pereira

PhD student

[email protected]

3/May/2016 Instituto Superior Técnico

Earth models for early

exploration stages

1

Motivation

Methodology

Case Study

Seismic inversion results

Conclusions

3/May/2016 Instituto Superior Técnico 2

Outline


How?

Combination of information from

Geological analogs and Geophysical

data in a geoestatistical seismic

inversion procedure.

ENMC ENMC

Motivation

Prospect characterization and uncertainty assessment of unexplored areas or in

early stages of exploration.

Geostatistical Seismic Inversion algorithm - GSI (Soares et al. 2007;

Caetano, 2009)).

Use of geological analogs for extraction of a priori distributions of Acoustic

Impedance (AI) (e.g. well-logs from nearby wells).

Use of a geological conceptual model based on seismic interpretation.


Methodology


Global approaches (Soares et al., 2007; Caetano, 2009)

i. Generation and perturbation of entire cube of parameters (acoustic properties) .

ii. Optimization method (genetic algorithms) to assure the convergence of the iterative procedure.

Geostatistical Seismic Inversion - GSI


The distributions of AI should be representative of the expected lithofacies and also of the relation between the

different litho-stratigraphic units.

Seismic Interpretation

Geological Analogs

(wells-logs from nearby wells)

GEOLOGICAL MODEL

Acoustic Impedance distributions

extracted from geological analogs

Dt Rhob AI

Geological analogs

Regional Geology


Zone 1

Zone 2

Zone 3

DEFINITION OF

GEOLOGICAL ZONES

SEISMIC SEISMIC INTERPRETATION GEOLOGICAL MODEL

Geological model with 3 different lithofacies

based on seismic interpretation

Grid size: 198 x 279 x 190

Geological model


Zone 1

Zone 2

Zone 3

GEOLOGICAL MODELA PRIORI DISTRIBUTION OF ACOUSTIC IMPEDANCE (AI)

Each a priori AI distribution, for the different geological zones were extracted directly from geological

analogs (well-logs from nearby wells). They correspond to the expected lithofacies for each zone.

GEOLOGICAL

ANALOGS(nearby appraisal well)

AI

Dt Rhob AI

Geological analogs

Acoustic Impedance - AI

sands with shales marlsshales


*

Compare

seismicConvolution

Seismic

Synthetic

Local correlation coefficient calculation and

Selection of best fit models

Best local

correlation coef.Best AI model

Use best models as secondary variable for co-simulation.

Iterate until the convergence of the method.

Stochastic

Simulation of

AI (DSS)

AI distributions extracted

from geological analogs

Geological model N(AI)

Workflow

Forward model

Methodology

The methodology proposed was applied to an unexplored area, where the potential

reservoir is expected to be in a turbidite system.


3D post-stack seismic volume, with sampling interval 2ms

3 appraisal wells outside the area used as analogs (w1, w2, w3)

Case study


Well-logs used as analogs

Well-logs available: Sonic,

Density, Acoustic Impedance

(computed from Sonic and

Density logs)

Dt Rhob AI Dt Rhob AI Dt Rhob AI

Case study


Geological model with 3 lithofacies zones

Tree different a priori Acoustic Impedance (AI) distributions extracted from

the geological analogs.

Spatial continuity pattern expressed by a variogram model obtained from

seismic data.

Statistical wavelet extracted from seismic data.

The algorithm was run with 6 iterations in each one were generated 32 models

of Acoustic Impedance.

Seismic Inversion Parameterization


Evolution of Global Correlation

Coefficient


REAL SEISMIC SYNTHETIC it#6 CC=0.85

seis

mic

synthetic it#6

max

min

max

min

Correlation between real and synthetic seismic


Difference in amplitude between real and synthetic seismic

REAL SEISMIC SEISMIC - SYNTHETIC

Zone 1

Zone 2

Zone 3

GEOLOGICAL MODEL

Transition zones have higher

diferences in amplitude

SYNTHETIC it#6max

min

max

min

max

min

Seismic amplitude


LOCAL CC it#6 BEST LOCAL CC it#6REAL SEISMIC SYNTHETIC it#6max

min

max

min

1

0

Local Correlation Coefficient - CC


Zone 2

Potential reservoir – turbidite system

CC=0.90HORIZON SLICE

SEISMIC

HORIZON SLICE

SYNTHETIC it#6REAL SEISMIC

seis

mic

synthetic it#6

max

min

max

min

max

min

Correlation between real and synthetic seismic

Zone 1

Zone 2

Zone 3

GEOLOGICAL MODEL

A PRIORI DISTRIBUTION OF ACOUSTIC IMPEDANCE

sands with

shales

marlsshales


MEAN AI it#6HORIZON SLICE

BEST AI it#6

SIMULATED AI it#6

Zone 1

Zone 2

Zone 3

GEOLOGICAL MODEL

max

min

max

min

max

min

Acoustic Impedance model

Low acoustic impedance values correspond to high

porosity areas


It is important that the acoustic impedance distributions extracted from

geological analogs should be representative of the expected lithofacies and

also of the relation between the different litho-stratigraphic units.

The Geostatistical Seismic Inversion methodology using geological analogs

proved to be a valuable tool to be applied on unexplored areas or in early

stage of exploration for prediction of the subsurface geology and for prospect

characterization.

Conclusions


Partex Oil and Gas and Repsol for allowing the use of the data and give

permission to show the results.

Schlumberger for the academic licenses of Petrel®

CGG for the academic licenses of Hampson-Russel software.

Thank you

Acknowledgements

Caetano, H. [2009] Integration of Seismic Information in Reservoir Models: Global Stochastic

Inversion, PhD thesis, Instituto Superior Técnico, University of Lisbon.

Pereira A., Nunes R., Azevedo L., Guerreiro L. & Soares A. [2016] Seismic inversion using

analogs for reservoir characterization and uncertainty assessment in early stages of exploration.

Search and Discovery, Proccedings of International Conference and Exhibition AAPG/SEG 2016

(Accepted).

Pereira A., Nunes R., Azevedo L., Guerreiro L. & Pereira M.J. [2015] Global Stochastic Inversion

Using "Analogs-wells" and Zonal Distributions - Application to an Unexplored Area. Proceedings

of EAGE 2015 Conference - Petroleum Geostatistics.

Soares, A., Diet, J.D. & Guerreiro, L. [2007] Stochastic Inversion with a Global Perturbation

Method, EAGE Conference on Petroleum Geostatistics, Extended Abstracts.


References

Documents

Earth models for early exploration stages · Geostatistical Seismic Inversion algorithm - GSI (Soares et al. 2007; Caetano, 2009)). Use of geological analogs for extraction of a priori