Integration of seismic data

1 - Classification: Internal 2010-06-10

Integration of Seismic Data and Uncertainties in the Facies ModelP. Nivlet*, S. Ng, M.A. Hetle, K. Børset, A.B. Rustad (Statoil ASA),P. Dahle, R. Hauge & O. Kolbjørnsen (Norwegian Computing Center)

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20102 -

Motivation: 3D reservoir modelling

3D reservoir model3D reservoir model

Reservoir Reservoir simulationssimulations

Production dataProduction data

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20103 -

The Snorre field

• Location:

Blocks 34/4 and 34/7 in the Tampen area, in the northern part of the North Sea (191 km2)

• Production start: 1992

• Production

(2009): ~180,000 bbl/day

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20104 -

Motivations: Data integration

3D reservoir model3D reservoir modelWell log data

seismic amplitudes (angle-stacks)

Structure, stratigraphy

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20105 -

Challenges in integrating the data

• Multi-scale issue

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20106 -

Challenges in integrating the data

• Non-unique relationship between seismic amplitudes and geology

• A multivariate problem

2.0

1.7

Vp/V

s

Shale

AI (g/cm3.m/s)6,000 10,000

Sand

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20107 -

The data uncertainty challenge

• Random noise• Acquisition / Processing footprint • Angle Misalignments• Imperfect physical model

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20108 -

Geological setting

1,000 m

•Reservoir depth:

2-2.7 km

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 20109 -

Traditional workflow

Seismic attribute (depth)

Facies model

Reservoir grid (depth)

integration

Well facies+extracted seismic attribute

geometry

conditioning

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201010 -

Proposed workflow

Seismic attribute (depth)

Facies model

Reservoir grid (depth)

integration

Well facies+extracted seismic attribute

geometry

conditioning

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201011 -

Workflow from inversion to facies prediction

Seismic (partial angle-stacks) Inversion

m

Bayesian wavelet extraction

Seismic facies analysis

Vp

Vs

ρ

m

Facies probability

Decreasing probability

of shale

Increasing probability of shale

BCUBCU

SN LL

OWCLunde

SN ML

Lomvi Fm

34/434/4--11

Decreasing probability

of shale

Increasing probability of shale

BCUBCU

SN LL

OWCLunde

SN ML

Lomvi Fm

34/434/4--11

Seismicmodelling

mBG mS mHF

=

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201012 -

Geostatistical seismic inversion

• 1D modelling of seismic amplitudes (Aki&Richards’ model): linear in m=(log(vp ), log(vs ), log))

• Normal distribution of elastic properties m

• Data (e ) stationary uncertainties estimated from analysis of amplitudes

• Prior (m ) stationary uncertainties estimated from well log analysis

nGmd

mm|d = mBG +m

G*(Gm

G* + e

)-1(d -

GmBG )

m|d

= m

-

m

G*(Gm

G* + e

)-1G m

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201013 -

Advantages/limitations of the technique

Lateral correlations

- Different stratigraphy settings

- Grid built from max. 2 horizons

Stationary uncertainty model:

- Global matrix

- Lateral correlations

- Vertical correlations

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201014 -

Inversion result: Elastic properties

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201015 -

Impact on elastic parameter uncertainties

Frequency (Hz)

0 20 40 60

0

-50

0 10 20 30

Prior Posterior uncertainty variation (%)

AI

Vp

Rho

SI

Vs

Vp/Vs

Prior Posterior uncertainty variation AI (%)

Seismic bandwidth (Near)

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201016 -

Inversion results QC

Multivariate correlation (RV ) between band- pass well-logs and inversion results

35% of wells RV > 0.8 33% of wells 0.8 > RV > 0.7 32% of wells RV < 0.7

Well

Inversion

AI SI Rhob

100

ms

Band-pass filtered

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201017 -

Workflow from inversion to facies prediction

Seismic (partial angle-stacks) Inversion

m

Bayesian wavelet extraction

Seismic facies analysis

Vp

Vs

ρ

m

Facies probability

Decreasing probability

of shale

Increasing probability of shale

BCUBCU

SN LL

OWCLunde

SN ML

Lomvi Fm

34/434/4--11

Decreasing probability

of shale

Increasing probability of shale

BCUBCU

SN LL

OWCLunde

SN ML

Lomvi Fm

34/434/4--11

Seismicmodelling

mBG mS mHF

=

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201018 -

Supervised seismic facies analysis

p(m

| Sand)p(Sand | m)

Kernel estimator

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201019 -

Supervised seismic facies analysis

μ

m|d = μm +(I- Σm/d Σm-1)(m

– μm ) + e*

Raw Well logsRaw Well logs

Filtered well logsFiltered well logs

Inversion results at well positionInversion results at well position

Inversion filtered well logsInversion filtered well logs

Different resolution scales

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201020 -

Cross plots: Inversion filtered well logs

Inversion frequency filtered Predicted SAND probability

Vp/V

s

AI (g/cm3 m/s)6,000 10,000

2.0

1.7V

p/Vs

AI (g/cm3 m/s)6,000 10,000

2.0

1.7

0

1

Shale

Sand

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201021 -

Seismic facies analysis: Sand probability results

Sand probability

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201022 -

Inversion results QC: Finding optimal well position

Confidence index (khi2): Vertical sand proportion from well compared with seismic sand probability100

ms

Seismic sand probability section

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201023 -

Facies probability QC

31% of wells: Good confidence 61% of wells: Medium 8% of wells: Bad confidence

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201024 -

Inversion results QC

Potential factors impacting mismatch

Stratigraphic level

Position with respect to OWC

Presence of faults

Average shale proportion

++

+

+

+

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201025 -

3D confidence index• Measurement of prediction

• Weighting function in facies modelling

0

1

WellInversion result

Confidence[0,1]

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201026 -

Proposed workflow

Facies model

Reservoir grid (depth)

integration

Well facies+extracted seismic attribute

geometry

conditioning

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201027 -

Snorre: Average proportion of channelAverage map estimated from 8 realizations

0

1

0

1

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201028 -

Concluding remarks

• Integrated workflow from seismic inversion to consistent seismic constrained facies modelling

• Fast geostatistical inversion approach and facies prediction

• Consistent resolution between inversion results and facies probabilities gives realistic predictions and facies models

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201029 -

Concluding remarks: Further work

• How to refine the upscaling of elastic parameters from well log to seismic scales? How to have a more local approach?

• Constraining observed 4D signals by using predicted facies sand probability (Ayzenberg and Theune, “Stratigraphically constrained seismic 4D inversion” M017, Room 127/128, Wednesday, 9h30)

• Flow simulations of constrained facies models and history matching with 4D for more predictive production prognoses

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201030 -

Acknowledgements

Thanks to Statoil, Norwegian Computing Center and the Snorre partners

Petoro, ExxonMobil Norge, Idemitsu Petroleum, RWE Dea Norge, Total E&P Norge and Amerada Hess Norge

for discussions and permission to publish this work.

72nd EAGE Conference and Exhibition – Barcelona, June 14th-17th 201031 -

Integration of Seismic Data and Uncertainties in the Facies Model

Philippe NivletPrincipal Geophysicist –Petek Tyrihanspniv@statoil.com, tel: +47 958 16 589www.statoil.com

Thank you