History Matching Reservoir Models With Many Objective Bayesian Optimization

ABSTRACT Reservoir models for predicting subsurface fluid and rock behaviors can now include upwards of billions (and potentially trillions) of grid cells and are pushing the limits of computational resources. History matching, where models are updated to match existing historical data more closely,...

Full description

Saved in:
Bibliographic Details
Main Authors: Steven Samoil, Clyde Fare, Kirk E. Jordan, Zhangxin Chen
Format: Article
Language:English
Published: Wiley 2024-12-01
Series:Applied AI Letters
Subjects:
Online Access:https://doi.org/10.1002/ail2.99
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1846126939931148288
author Steven Samoil
Clyde Fare
Kirk E. Jordan
Zhangxin Chen
author_facet Steven Samoil
Clyde Fare
Kirk E. Jordan
Zhangxin Chen
author_sort Steven Samoil
collection DOAJ
description ABSTRACT Reservoir models for predicting subsurface fluid and rock behaviors can now include upwards of billions (and potentially trillions) of grid cells and are pushing the limits of computational resources. History matching, where models are updated to match existing historical data more closely, is conducted to reduce the number of simulation runs and is one of the primary time‐consuming tasks. As models get larger the number of parameters to match increases, and the number of objective functions increases, and traditional methods start to reach their limitations. To solve this, we propose the use of Bayesian optimization (BO) in a hybrid cloud framework. BO iteratively searches for an optimal solution in the simulations campaign through the refinement of a set of priors initialized with a set of simulation results. The current simulation platform implements grid management and a suite of linear solvers to perform the simulation on large scale distributed‐memory systems. Our early results using the hybrid cloud implementation shown here are encouraging on tasks requiring over 100 objective functions, and we propose integrating BO as a built‐in module to efficiently iterate to find an optimal history match of production data in a single package platform. This paper reports on the development of the hybrid cloud BO based history matching framework and the initial results of the application to reservoir history matching.
format Article
id doaj-art-56d3b5d1cff345e68ce1c4a3f71d84f3
institution Kabale University
issn 2689-5595
language English
publishDate 2024-12-01
publisher Wiley
record_format Article
series Applied AI Letters
spelling doaj-art-56d3b5d1cff345e68ce1c4a3f71d84f32024-12-12T07:26:42ZengWileyApplied AI Letters2689-55952024-12-0154n/an/a10.1002/ail2.99History Matching Reservoir Models With Many Objective Bayesian OptimizationSteven Samoil0Clyde Fare1Kirk E. Jordan2Zhangxin Chen3Schulich School of Engineering University of Calgary Calgary CanadaIBM Research—Europe, IBM Daresbury UKIBM Research, IBM Cambridge Massachusetts USASchulich School of Engineering University of Calgary Calgary CanadaABSTRACT Reservoir models for predicting subsurface fluid and rock behaviors can now include upwards of billions (and potentially trillions) of grid cells and are pushing the limits of computational resources. History matching, where models are updated to match existing historical data more closely, is conducted to reduce the number of simulation runs and is one of the primary time‐consuming tasks. As models get larger the number of parameters to match increases, and the number of objective functions increases, and traditional methods start to reach their limitations. To solve this, we propose the use of Bayesian optimization (BO) in a hybrid cloud framework. BO iteratively searches for an optimal solution in the simulations campaign through the refinement of a set of priors initialized with a set of simulation results. The current simulation platform implements grid management and a suite of linear solvers to perform the simulation on large scale distributed‐memory systems. Our early results using the hybrid cloud implementation shown here are encouraging on tasks requiring over 100 objective functions, and we propose integrating BO as a built‐in module to efficiently iterate to find an optimal history match of production data in a single package platform. This paper reports on the development of the hybrid cloud BO based history matching framework and the initial results of the application to reservoir history matching.https://doi.org/10.1002/ail2.99Bayesian optimizationhistory matchingreservoir simulation
spellingShingle Steven Samoil
Clyde Fare
Kirk E. Jordan
Zhangxin Chen
History Matching Reservoir Models With Many Objective Bayesian Optimization
Applied AI Letters
Bayesian optimization
history matching
reservoir simulation
title History Matching Reservoir Models With Many Objective Bayesian Optimization
title_full History Matching Reservoir Models With Many Objective Bayesian Optimization
title_fullStr History Matching Reservoir Models With Many Objective Bayesian Optimization
title_full_unstemmed History Matching Reservoir Models With Many Objective Bayesian Optimization
title_short History Matching Reservoir Models With Many Objective Bayesian Optimization
title_sort history matching reservoir models with many objective bayesian optimization
topic Bayesian optimization
history matching
reservoir simulation
url https://doi.org/10.1002/ail2.99
work_keys_str_mv AT stevensamoil historymatchingreservoirmodelswithmanyobjectivebayesianoptimization
AT clydefare historymatchingreservoirmodelswithmanyobjectivebayesianoptimization
AT kirkejordan historymatchingreservoirmodelswithmanyobjectivebayesianoptimization
AT zhangxinchen historymatchingreservoirmodelswithmanyobjectivebayesianoptimization