Improving local history match using machine learning generated regions from production response and geological parameter correlations

T. Buckle, R. Hutton, V. Demyanov, D. Arnold, A. Antropov, E. Kharyba, M. Pilipenko, L. Stulov

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

Abstract

We present a data driven workflow to improve local history match quality by identifying model regions from correlation between production response and geological modelling parameters for use in an assisted history matching framework. This paper outlines the implementation and results from a large mature field case study. Regions are identified by calculating the partial correlation between individual well production misfits and uncertain geological modelling parameters across 500 models. Wells are then categorised into three groups based on their correlations: positive, negative and insignificant. A probabilistic neural network (PNN) is trained on the location of each well and its group. A map of regions can then be calculated using the PNN. The parameters used to define the region map are then varied separately in each region in an assisted history matching loop. In the full field case study, an 8.8% improvement in oil rate misfit within the positively correlated well group was achieved by regional modification of the net-to-gross multiplier, with no detrimental effect on the other groups match quality. This case study demonstrates the effective identification and utilisation of geologically and dynamically inferred regions which improve the local history match.

Original languageEnglish
Title of host publication4th EAGE Conference on Petroleum Geostatistics
PublisherEAGE Publishing BV
ISBN (Electronic)9789462822962
Publication statusPublished - 2019
Event4th EAGE Conference on Petroleum Geostatistics 2019 - Florence, Italy
Duration: 2 Sep 20196 Sep 2019

Conference

Conference4th EAGE Conference on Petroleum Geostatistics 2019
CountryItaly
CityFlorence
Period2/09/196/09/19

Fingerprint

machine learning
Learning systems
Machine Learning
histories
Neural networks
history
History Matching
Probabilistic Neural Network
multipliers
Partial Correlation
well
oils
Gross
Modeling
Data-driven
Work Flow
modeling
Multiplier
History
parameter

ASJC Scopus subject areas

  • Geophysics
  • Statistics, Probability and Uncertainty
  • Geology
  • Geotechnical Engineering and Engineering Geology

Cite this

Buckle, T., Hutton, R., Demyanov, V., Arnold, D., Antropov, A., Kharyba, E., ... Stulov, L. (2019). Improving local history match using machine learning generated regions from production response and geological parameter correlations. In 4th EAGE Conference on Petroleum Geostatistics [ThPG06] EAGE Publishing BV.
Buckle, T. ; Hutton, R. ; Demyanov, V. ; Arnold, D. ; Antropov, A. ; Kharyba, E. ; Pilipenko, M. ; Stulov, L. / Improving local history match using machine learning generated regions from production response and geological parameter correlations. 4th EAGE Conference on Petroleum Geostatistics. EAGE Publishing BV, 2019.
@inproceedings{caafa97f8c464a45953e89f6987aae33,
title = "Improving local history match using machine learning generated regions from production response and geological parameter correlations",
abstract = "We present a data driven workflow to improve local history match quality by identifying model regions from correlation between production response and geological modelling parameters for use in an assisted history matching framework. This paper outlines the implementation and results from a large mature field case study. Regions are identified by calculating the partial correlation between individual well production misfits and uncertain geological modelling parameters across 500 models. Wells are then categorised into three groups based on their correlations: positive, negative and insignificant. A probabilistic neural network (PNN) is trained on the location of each well and its group. A map of regions can then be calculated using the PNN. The parameters used to define the region map are then varied separately in each region in an assisted history matching loop. In the full field case study, an 8.8{\%} improvement in oil rate misfit within the positively correlated well group was achieved by regional modification of the net-to-gross multiplier, with no detrimental effect on the other groups match quality. This case study demonstrates the effective identification and utilisation of geologically and dynamically inferred regions which improve the local history match.",
author = "T. Buckle and R. Hutton and V. Demyanov and D. Arnold and A. Antropov and E. Kharyba and M. Pilipenko and L. Stulov",
year = "2019",
language = "English",
booktitle = "4th EAGE Conference on Petroleum Geostatistics",
publisher = "EAGE Publishing BV",
address = "Netherlands",

}

Buckle, T, Hutton, R, Demyanov, V, Arnold, D, Antropov, A, Kharyba, E, Pilipenko, M & Stulov, L 2019, Improving local history match using machine learning generated regions from production response and geological parameter correlations. in 4th EAGE Conference on Petroleum Geostatistics., ThPG06, EAGE Publishing BV, 4th EAGE Conference on Petroleum Geostatistics 2019, Florence, Italy, 2/09/19.

Improving local history match using machine learning generated regions from production response and geological parameter correlations. / Buckle, T.; Hutton, R.; Demyanov, V.; Arnold, D.; Antropov, A.; Kharyba, E.; Pilipenko, M.; Stulov, L.

4th EAGE Conference on Petroleum Geostatistics. EAGE Publishing BV, 2019. ThPG06.

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

TY - GEN

T1 - Improving local history match using machine learning generated regions from production response and geological parameter correlations

AU - Buckle, T.

AU - Hutton, R.

AU - Demyanov, V.

AU - Arnold, D.

AU - Antropov, A.

AU - Kharyba, E.

AU - Pilipenko, M.

AU - Stulov, L.

PY - 2019

Y1 - 2019

N2 - We present a data driven workflow to improve local history match quality by identifying model regions from correlation between production response and geological modelling parameters for use in an assisted history matching framework. This paper outlines the implementation and results from a large mature field case study. Regions are identified by calculating the partial correlation between individual well production misfits and uncertain geological modelling parameters across 500 models. Wells are then categorised into three groups based on their correlations: positive, negative and insignificant. A probabilistic neural network (PNN) is trained on the location of each well and its group. A map of regions can then be calculated using the PNN. The parameters used to define the region map are then varied separately in each region in an assisted history matching loop. In the full field case study, an 8.8% improvement in oil rate misfit within the positively correlated well group was achieved by regional modification of the net-to-gross multiplier, with no detrimental effect on the other groups match quality. This case study demonstrates the effective identification and utilisation of geologically and dynamically inferred regions which improve the local history match.

AB - We present a data driven workflow to improve local history match quality by identifying model regions from correlation between production response and geological modelling parameters for use in an assisted history matching framework. This paper outlines the implementation and results from a large mature field case study. Regions are identified by calculating the partial correlation between individual well production misfits and uncertain geological modelling parameters across 500 models. Wells are then categorised into three groups based on their correlations: positive, negative and insignificant. A probabilistic neural network (PNN) is trained on the location of each well and its group. A map of regions can then be calculated using the PNN. The parameters used to define the region map are then varied separately in each region in an assisted history matching loop. In the full field case study, an 8.8% improvement in oil rate misfit within the positively correlated well group was achieved by regional modification of the net-to-gross multiplier, with no detrimental effect on the other groups match quality. This case study demonstrates the effective identification and utilisation of geologically and dynamically inferred regions which improve the local history match.

UR - http://www.scopus.com/inward/record.url?scp=85073114585&partnerID=8YFLogxK

M3 - Conference contribution

BT - 4th EAGE Conference on Petroleum Geostatistics

PB - EAGE Publishing BV

ER -

Buckle T, Hutton R, Demyanov V, Arnold D, Antropov A, Kharyba E et al. Improving local history match using machine learning generated regions from production response and geological parameter correlations. In 4th EAGE Conference on Petroleum Geostatistics. EAGE Publishing BV. 2019. ThPG06