Sensitivity Analysis for Joint Inversion of Production and - - PowerPoint PPT Presentation

Sensitivity Analysis for Joint Inversion

• f Production and Time-Lapse

Seismic Data of Norne Field

Amit Suman and Tapan Mukerji

SCRF Annual Meeting May 4-5, 2011

4-5 May 2011 SCRF 2

Introduction

Well log data Core data 3D Seismic Production data 4D Seismic data Reservoir management Reservoir forecasting Static data Dynamic data

4-5 May 2011 SCRF 3

Joint Inversion Loop

Generate multiple models Evaluate misfit

.

Reservoir Model Observed flow and seismic response Predicted flow and seismic response

4-5 May 2011 SCRF 4

Norne Field Segment E

• Southern part of Norwegian sea
• Five prime zones

– Garn – Not – Ile – Tofte – Tilje

• 3 producer and 2 injector wells
• Gas

Oil and water

4-5 May 2011 SCRF 5

Data Available

• Well logs
• Horizons
• Well data
• Oil , gas and water flow rate
• BHP (Bottom hole pressure)

4-5 May 2011 SCRF 6

Time Lapse Seismic Data

Near, Mid , Far and Full 3D seismic Q – Marine Seismic 2001 2003 2004 2006

4-5 May 2011 SCRF 7

Seismic Amplitude on Top

2001 2003 2004 2006

4-5 May 2011 8

Dynamic Dynamic Dynamic Dynamic modeling modeling modeling modeling Δ Δ Δ Δ Pressure Pressure Pressure Pressure Δ Δ Δ Δ Saturation Saturation Saturation Saturation Production Production Production Production data at time t data at time t data at time t data at time t Rock physics Rock physics Rock physics Rock physics modeling modeling modeling modeling Velocity at Velocity at Velocity at Velocity at time t time t time t time t Seismic data Seismic data Seismic data Seismic data at time t at time t at time t at time t Optimize Optimize Optimize Optimize mismatch mismatch mismatch mismatch Update Update Update Update parameters parameters parameters parameters

Motivation

4-5 May 2011 SCRF 9

“what are the sensitive parameters in joint inversion loop?”

4-5 May 2011 SCRF 10

Sensitivity Parameters

• Relative permeability
• Pore compressibility
• Rock physics models
• Saturation scale
• Porosity and permeability models

4-5 May 2011 SCRF 11

Porosity Model Generation

Structural Structural Structural Structural modeling modeling modeling modeling Generate Generate Generate Generate variogram variogram variogram variogram SGSIM SGSIM SGSIM SGSIM

• 168168

cells

4-5 May 2011 SCRF 12

Selection of Porosity Models

Multi-dimensional scaling (MDS) A set of N earth models can be uniquely mapped in any dimension less than or equal to N using a pair wise distance between them

(Caers et al 2009) Euclidean distance

4-5 May 2011 SCRF 13

Selection of Porosity Models

100 Porosity Models MDS and Clustering 3 Porosity models selected for Analysis

4-5 May 2011 SCRF 14

Relative Permeability

1 1 2 2 1 2

Oil and water relative permeability

4-5 May 2011 SCRF 15

Pore Compressibility

0.47 P95 0.17 P5 0.30 Mean 0.34 P95 0.16 P5 0.25 Mean 0.34 P95 0.16 P5 0.27 Mean

GARN ILE TOFTE TILJE

0.52 P95 0.15 P5 0.27 Mean All compressibilities are in the order of 10-9(Pa-1) (6.89 X 10-6 Psi-1 )

Spatial variability in pore compressibility can play an important role in time lapse seismic modeling

(Suman et al,2009)

4-5 May 2011 SCRF 16

Rock Physics Models

Velocity – Porosity trends can be established using well data but there are uncertainties away from the wells

• Cemented sand

model

• Uncemented sand

model

1 2 3 4 5 6 0.1 0.2 0.3 0.4 0.5 Cementing vs. Sorting Trends Vp Porosity clean cementing trend Suspension Line (Reuss Bound) sorting trend

New Deposition

Mineral point

(Avseth et al., 2005)

4-5 May 2011 SCRF 17

Saturation Scale

• Seismic velocities depend on fluid saturation as

well as saturation scale

• Reservoirs with gas are very likely to show patchy

behavior

Sengupta ,2000

4-5 May 2011 SCRF 18

Experimental Design

Model 3 Model 2 Model 1 Porosity Patchy Uniform Saturation behavior Uncemented Cemented Rock physics models High Low Relative permeability 5X10-10 (Pa-1) 3X10-10 (Pa-1) 1.5X10-10 (Pa-1) Pore Compressibility

4-5 May 2011 SCRF 19

Dynamic Dynamic Dynamic Dynamic modeling modeling modeling modeling (1997 (1997 (1997 (1997-

• 2001)

2001) 2001) 2001) Δ Δ Δ Δ Pressure Pressure Pressure Pressure Δ Δ Δ Δ Saturation Saturation Saturation Saturation Cumulative oil Cumulative oil Cumulative oil Cumulative oil production in production in production in production in 2001 2001 2001 2001 Rock Rock Rock Rock physics physics physics physics modeling modeling modeling modeling Seismic Seismic Seismic Seismic velocity in 1997 velocity in 1997 velocity in 1997 velocity in 1997 and 2001 and 2001 and 2001 and 2001

Methodology

Compare Compare Compare Compare Compare Compare Compare Compare Difference in Difference in Difference in Difference in seismic seismic seismic seismic velocity velocity velocity velocity

4-5 May 2011 SCRF 20 Cemented sand model Uncemented sand model

Sensitivity to rock physics model Sensitivity to relative permeability Sensitivity to porosity model

L1 Norm of change in seismic velocity after 4 years

Sensitivity of 4D Seismic

4-5 May 2011 SCRF 21

Sensitivity to porosity model Sensitivity to relative permeability

Sensitivity of Production

4-5 May 2011 SCRF 22

Conclusions and Future Work

• Rock physics model is the most sensitive parameter

(amongst those tested) for seismic modeling

• Relative permeability and Porosity models are second

and third most sensitive parameters for seismic modeling

• Relative permeability is the most sensitive parameter for

flow response

• Sensitivity to pore compressibility and saturation scale

are under investigation

• Other parameters to be tested (variogram ranges, poro-

perm relations)

• The result of this study will be used in joint inversion of

time-lapse and production data of Norne field