Deep neural network application for 4D seismic inversion to changes in pressure and saturation: Optimizing the use of synthetic training datasets

Geophysical Prospecting - Tập 68 Số 7 - Trang 2164-2185 - 2020
G. Côrte1, Jesper Dramsch2, Hamed Amini3,1, Colin MacBeth1
1Heriot-Watt University, Edinburgh EH14 4AS UK
2Technical University of Denmark, Copenhagen, Denmark
3Currently at Aker BP, Stavanger, Norway

Tóm tắt

ABSTRACTIn this work, we tackle the challenge of quantitative estimation of reservoir dynamic property variations during a period of production, directly from four‐dimensional seismic data in the amplitude domain. We employ a deep neural network to invert four‐dimensional seismic amplitude maps to the simultaneous changes in pressure, water and gas saturations. The method is applied to a real field data case, where, as is common in such applications, the data measured at the wells are insufficient for properly training deep neural networks, thus, the network is trained on synthetic data. Training on synthetic data offers much freedom in designing a training dataset, therefore, it is important to understand the impact of the data distribution on the inversion results. To define the best way to construct a synthetic training dataset, we perform a study on four different approaches to populating the training set making remarks on data sizes, network generality and the impact of physics‐based constraints. Using the results of a reservoir simulation model to populate our training datasets, we demonstrate the benefits of restricting training samples to fluid flow consistent combinations in the dynamic reservoir property domain. With this the network learns the physical correlations present in the training set, incorporating this information into the inference process, which allows it to make inferences on properties to which the seismic data are most uncertain. Additionally, we demonstrate the importance of applying regularization techniques such as adding noise to the synthetic data for training and show a possibility of estimating uncertainties in the inversion results by training multiple networks.

Từ khóa


Tài liệu tham khảo

Aki K., 1980, Quantitative Seismology: Theory and Methods V1

10.1111/1365-2478.12071

Amini H.(2014)A pragmatic approach to simulator‐to‐seismic modelling for 4D seismic interpretation. PhD Thesis Heriot‐Watt University.

Amini H.(2018)Comparison of Xu‐White simplified Xu‐White (Keys and Xu) and Nur ’s critical porosity in shaley sands. 80th EAGE Annual meeting Copenhagen Denmark Extended Abstracts We A11 08.

Amini H.andMacBeth C.(2015)Calibration of rock stress‐sensitivity using 4D seismic data. 77th EAGE Annual meeting Madrid Spain Extended Abstracts We A11 08.

Ayzenberg M.andLiu S.(2014)Saturation and pressure inversion ‐ from 4D seismic to reservoir model updating. 76th EAGE Annual meeting Amsterdam Netherlands Extended Abstracts We G102 16.

Bergstra J. Yamins D.andCoxD.D.(2013)Making a science of model search: hyperparameter optimization in hundreds of dimensions for vision architectures.30th International Conference on Machine Learning ICML Atlanta Georgia USA 115–123.

10.1162/neco.1995.7.1.108

Blanchard T.D.(2012)Inherent uncertainties in 4D AVO and the implications on pressure and saturation inversion. 74th EAGE Annual meeting Copenhagen Denmark Extended Abstracts Y020.

Blanchard T.D.andThore P.(2008)Introducing prior information to pressure and saturation inversion: the key for success?83rd SEG Annual meeting Houston Texas USA Extended Abstracts 4971–4975.

10.1190/tle36030234.1

Coleou T. Roustiau A. Machecler I. Ayzenberg M. Fayemendy C. Skjei N. et al. (2013)4D Petrophysical seismic inversion ‐ case studies. 75th EAGE Annual meeting London UK Extended Abstracts We 17 03.

Côrte G. MacBeth C.andAmini H.(2019)North Sea field application of 4D Bayesian inversion to pressure and saturation changes. 81st EAGE Annual meeting London UK Extended Abstracts Tu_P04_09.

10.1111/1365-2478.12037

10.1016/j.petrol.2013.03.017

Dramsch J. Corte G. Amini H. Lüthje M.andMacBeth C.(2019a)Deep learning application for 4D pressure saturation inversion compared to Bayesian inversion on North Sea Data. Second EAGE Workshop Practical Reservoir Monitoring We PRM 11.

Dramsch J. Corte G. Amini H. MacBeth C.andLüthje M.(2019b)Including physics in deep learning ‐ an example from 4D seismic pressure saturation inversion. 81st EAGE Annual meeting London UK Extended Abstracts WS10_05.

10.1144/GSL.SP.1995.090.01.03

Floricich M.(2006)An engineering‐consistent approach for pressure and saturation estimation from time‐lapse seismic data. PhD Thesis Heriot‐Watt University.

Gassmann F., 1951, Über die Elastizität poröser Medien, Vierteljahrsschrift der Naturforschenden Gesellschaft in Zürich, 96, 1

Kingma D.P.andBa J.(2015)Adam: a method for stochastic optimization.International Conference for Learning Representations San Diego California USA.

Kingma D.P.andWelling M.(2014).Auto‐encoding variational Bayes.Proceedings of International Conference for Learning Representations Banff Canada.

Kingma D.P., 2015, Variational dropout and the local reparameterization trick, Advances in Neural Information Processing Systems, 28, 2575

10.1144/0050645

10.1190/1.1444973

10.1190/1.1707070

10.1111/j.1365-2478.2006.00536.x

10.1190/1.1437977

Omofoma V.(2017)The quantification of pressure and saturation changes in clastic reservoirs using 4D seismic data. PhD Thesis Heriot‐Watt University.

10.1111/1365-2478.12718

10.1111/j.1365-2478.1987.tb00856.x

Srivastava N., 2014, Dropout: a simple way to prevent neural networks from overfitting, Journal of Machine Learning Research, 15, 1929

10.1190/geo2011-0267.1

10.1190/1.3549756

Wong M.Y.(2017.Pressure and saturation estimation from prm time‐lapse seismic data for a compacting reservoir. PhD Thesis Heriot‐Watt University.

Xue Y. Araujo M. Wang K. Lopez J. Kumar G.andBrew G.(2019)Machine Learning to reduce cycle time for time‐lapse seismic data assimilation into reservoir management. 81st EAGE Annual meeting London UK Extended Abstracts We_R07_08.

10.1029/2019JB018408

Thông tin
Thông tin xuất bản

Geophysical Prospecting

Tập 68 Số 7

2164-2185

Thông tin tác giả