Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data

J. S. Dramsch, G. Corte, H. Amini, M. Lüthje, C. MacBeth

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

In this work we present a deep neural network inversion on map-based 4D seismic data for pressure and saturation. We present a novel neural network architecture that trains on synthetic data and provides insights into observed field seismic. The network explicitly includes AVO gradient calculation within the network as physical knowledge to stabilize pressure and saturation changes separation. We apply the method to Schiehallion field data and go on to compare the results to Bayesian inversion results. Despite not using convolutional neural networks for spatial information, we produce maps with good signal to noise ratio and coherency.

Original languageEnglish
Title of host publication2nd EAGE Workshop Practical Reservoir Monitoring 2019
PublisherEAGE Publishing BV
ISBN (Electronic)9789462822849
DOIs
Publication statusPublished - 1 Apr 2019
Event2nd EAGE Workshop Practical Reservoir Monitoring 2019 - Amsterdam, Netherlands
Duration: 1 Apr 20194 Apr 2019

Conference

Conference2nd EAGE Workshop Practical Reservoir Monitoring 2019
CountryNetherlands
CityAmsterdam
Period1/04/194/04/19

Fingerprint

North Sea
learning
saturation
inversions
signal-to-noise ratio
train
seismic data
signal to noise ratios
inversion
sea
gradients
method
calculation

ASJC Scopus subject areas

  • Geophysics

Cite this

Dramsch, J. S., Corte, G., Amini, H., Lüthje, M., & MacBeth, C. (2019). Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data. In 2nd EAGE Workshop Practical Reservoir Monitoring 2019 EAGE Publishing BV. https://doi.org/10.3997/2214-4609.201900028
Dramsch, J. S. ; Corte, G. ; Amini, H. ; Lüthje, M. ; MacBeth, C. / Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data. 2nd EAGE Workshop Practical Reservoir Monitoring 2019. EAGE Publishing BV, 2019.
@inproceedings{f7da20d7a07744cd9f1d61034122ad82,
title = "Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data",
abstract = "In this work we present a deep neural network inversion on map-based 4D seismic data for pressure and saturation. We present a novel neural network architecture that trains on synthetic data and provides insights into observed field seismic. The network explicitly includes AVO gradient calculation within the network as physical knowledge to stabilize pressure and saturation changes separation. We apply the method to Schiehallion field data and go on to compare the results to Bayesian inversion results. Despite not using convolutional neural networks for spatial information, we produce maps with good signal to noise ratio and coherency.",
author = "Dramsch, {J. S.} and G. Corte and H. Amini and M. L{\"u}thje and C. MacBeth",
year = "2019",
month = "4",
day = "1",
doi = "10.3997/2214-4609.201900028",
language = "English",
booktitle = "2nd EAGE Workshop Practical Reservoir Monitoring 2019",
publisher = "EAGE Publishing BV",
address = "Netherlands",

}

Dramsch, JS, Corte, G, Amini, H, Lüthje, M & MacBeth, C 2019, Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data. in 2nd EAGE Workshop Practical Reservoir Monitoring 2019. EAGE Publishing BV, 2nd EAGE Workshop Practical Reservoir Monitoring 2019, Amsterdam, Netherlands, 1/04/19. https://doi.org/10.3997/2214-4609.201900028

Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data. / Dramsch, J. S.; Corte, G.; Amini, H.; Lüthje, M.; MacBeth, C.

2nd EAGE Workshop Practical Reservoir Monitoring 2019. EAGE Publishing BV, 2019.

Research output: Chapter in Book/Report/Conference proceedingConference contribution

TY - GEN

T1 - Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data

AU - Dramsch, J. S.

AU - Corte, G.

AU - Amini, H.

AU - Lüthje, M.

AU - MacBeth, C.

PY - 2019/4/1

Y1 - 2019/4/1

N2 - In this work we present a deep neural network inversion on map-based 4D seismic data for pressure and saturation. We present a novel neural network architecture that trains on synthetic data and provides insights into observed field seismic. The network explicitly includes AVO gradient calculation within the network as physical knowledge to stabilize pressure and saturation changes separation. We apply the method to Schiehallion field data and go on to compare the results to Bayesian inversion results. Despite not using convolutional neural networks for spatial information, we produce maps with good signal to noise ratio and coherency.

AB - In this work we present a deep neural network inversion on map-based 4D seismic data for pressure and saturation. We present a novel neural network architecture that trains on synthetic data and provides insights into observed field seismic. The network explicitly includes AVO gradient calculation within the network as physical knowledge to stabilize pressure and saturation changes separation. We apply the method to Schiehallion field data and go on to compare the results to Bayesian inversion results. Despite not using convolutional neural networks for spatial information, we produce maps with good signal to noise ratio and coherency.

UR - http://www.scopus.com/inward/record.url?scp=85066458069&partnerID=8YFLogxK

U2 - 10.3997/2214-4609.201900028

DO - 10.3997/2214-4609.201900028

M3 - Conference contribution

BT - 2nd EAGE Workshop Practical Reservoir Monitoring 2019

PB - EAGE Publishing BV

ER -

Dramsch JS, Corte G, Amini H, Lüthje M, MacBeth C. Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North sea data. In 2nd EAGE Workshop Practical Reservoir Monitoring 2019. EAGE Publishing BV. 2019 https://doi.org/10.3997/2214-4609.201900028