An efficient probabilistic assisted history matching tool using Gaussian processes proxy models: Application to coalbed methane reservoir

Sachin Rana, Turgay Ertekin, Gregory R. King

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

1 Citation (Scopus)

Abstract

Currently used assisted history matching algorithms such as differential evolution, particle swarm optimization etc. usually requires a large number of numerical simulation runs in order to converge to acceptable solutions. If each numerical simulation takes a long time to complete, these algorithms become inefficient. In this research, a new assisted history matching tool is presented that can provide multiple solutions of history matching in much less number of numerical simulations. The proposed tool uses Gaussian process based proxy models to provide fast approximate forward solutions which are used in Bayesian optimization to find history match solutions in an iterative manner. The uncertainty in history match solutions is quantified via MCMC sampling on trained forward GP model. In order to check for independence and convergence of the collected MCMC samples, auto-correlation plots and Geweke z-score diagnostic tests are also performed. The converged MCMC samples are then used to quantify uncertainty in EUR of gas reserves via a forecasting GP model. The proposed methodology is successfully applied to a synthetic heterogeneous coalbed methane reservoir. The code and data for this case study is also available online for future studies. The results show that history matching can be performed in approximately four times less number of numerical simulation runs as compared to state of the art differential evolution algorithm. Besides, the P50 estimate of EUR is shown to be in close agreement with truth values for the presented case study.

Original languageEnglish (US)
Title of host publicationSPE Annual Technical Conference and Exhibition 2018, ATCE 2018
PublisherSociety of Petroleum Engineers (SPE)
ISBN (Electronic)9781613995723
StatePublished - Jan 1 2018
EventSPE Annual Technical Conference and Exhibition 2018, ATCE 2018 - Dallas, United States
Duration: Sep 24 2018Sep 26 2018

Publication series

NameProceedings - SPE Annual Technical Conference and Exhibition
Volume2018-September

Other

OtherSPE Annual Technical Conference and Exhibition 2018, ATCE 2018
CountryUnited States
CityDallas
Period9/24/189/26/18

Fingerprint

Computer simulation
Autocorrelation
Particle swarm optimization (PSO)
Sampling
Coal bed methane
Gases
Uncertainty

All Science Journal Classification (ASJC) codes

  • Fuel Technology
  • Energy Engineering and Power Technology

Cite this

Rana, S., Ertekin, T., & King, G. R. (2018). An efficient probabilistic assisted history matching tool using Gaussian processes proxy models: Application to coalbed methane reservoir. In SPE Annual Technical Conference and Exhibition 2018, ATCE 2018 (Proceedings - SPE Annual Technical Conference and Exhibition; Vol. 2018-September). Society of Petroleum Engineers (SPE).
Rana, Sachin ; Ertekin, Turgay ; King, Gregory R. / An efficient probabilistic assisted history matching tool using Gaussian processes proxy models : Application to coalbed methane reservoir. SPE Annual Technical Conference and Exhibition 2018, ATCE 2018. Society of Petroleum Engineers (SPE), 2018. (Proceedings - SPE Annual Technical Conference and Exhibition).
@inproceedings{bc34abdc22b34385b77bd821973166c9,
title = "An efficient probabilistic assisted history matching tool using Gaussian processes proxy models: Application to coalbed methane reservoir",
abstract = "Currently used assisted history matching algorithms such as differential evolution, particle swarm optimization etc. usually requires a large number of numerical simulation runs in order to converge to acceptable solutions. If each numerical simulation takes a long time to complete, these algorithms become inefficient. In this research, a new assisted history matching tool is presented that can provide multiple solutions of history matching in much less number of numerical simulations. The proposed tool uses Gaussian process based proxy models to provide fast approximate forward solutions which are used in Bayesian optimization to find history match solutions in an iterative manner. The uncertainty in history match solutions is quantified via MCMC sampling on trained forward GP model. In order to check for independence and convergence of the collected MCMC samples, auto-correlation plots and Geweke z-score diagnostic tests are also performed. The converged MCMC samples are then used to quantify uncertainty in EUR of gas reserves via a forecasting GP model. The proposed methodology is successfully applied to a synthetic heterogeneous coalbed methane reservoir. The code and data for this case study is also available online for future studies. The results show that history matching can be performed in approximately four times less number of numerical simulation runs as compared to state of the art differential evolution algorithm. Besides, the P50 estimate of EUR is shown to be in close agreement with truth values for the presented case study.",
author = "Sachin Rana and Turgay Ertekin and King, {Gregory R.}",
year = "2018",
month = "1",
day = "1",
language = "English (US)",
series = "Proceedings - SPE Annual Technical Conference and Exhibition",
publisher = "Society of Petroleum Engineers (SPE)",
booktitle = "SPE Annual Technical Conference and Exhibition 2018, ATCE 2018",
address = "United States",

}

Rana, S, Ertekin, T & King, GR 2018, An efficient probabilistic assisted history matching tool using Gaussian processes proxy models: Application to coalbed methane reservoir. in SPE Annual Technical Conference and Exhibition 2018, ATCE 2018. Proceedings - SPE Annual Technical Conference and Exhibition, vol. 2018-September, Society of Petroleum Engineers (SPE), SPE Annual Technical Conference and Exhibition 2018, ATCE 2018, Dallas, United States, 9/24/18.

An efficient probabilistic assisted history matching tool using Gaussian processes proxy models : Application to coalbed methane reservoir. / Rana, Sachin; Ertekin, Turgay; King, Gregory R.

SPE Annual Technical Conference and Exhibition 2018, ATCE 2018. Society of Petroleum Engineers (SPE), 2018. (Proceedings - SPE Annual Technical Conference and Exhibition; Vol. 2018-September).

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

TY - GEN

T1 - An efficient probabilistic assisted history matching tool using Gaussian processes proxy models

T2 - Application to coalbed methane reservoir

AU - Rana, Sachin

AU - Ertekin, Turgay

AU - King, Gregory R.

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Currently used assisted history matching algorithms such as differential evolution, particle swarm optimization etc. usually requires a large number of numerical simulation runs in order to converge to acceptable solutions. If each numerical simulation takes a long time to complete, these algorithms become inefficient. In this research, a new assisted history matching tool is presented that can provide multiple solutions of history matching in much less number of numerical simulations. The proposed tool uses Gaussian process based proxy models to provide fast approximate forward solutions which are used in Bayesian optimization to find history match solutions in an iterative manner. The uncertainty in history match solutions is quantified via MCMC sampling on trained forward GP model. In order to check for independence and convergence of the collected MCMC samples, auto-correlation plots and Geweke z-score diagnostic tests are also performed. The converged MCMC samples are then used to quantify uncertainty in EUR of gas reserves via a forecasting GP model. The proposed methodology is successfully applied to a synthetic heterogeneous coalbed methane reservoir. The code and data for this case study is also available online for future studies. The results show that history matching can be performed in approximately four times less number of numerical simulation runs as compared to state of the art differential evolution algorithm. Besides, the P50 estimate of EUR is shown to be in close agreement with truth values for the presented case study.

AB - Currently used assisted history matching algorithms such as differential evolution, particle swarm optimization etc. usually requires a large number of numerical simulation runs in order to converge to acceptable solutions. If each numerical simulation takes a long time to complete, these algorithms become inefficient. In this research, a new assisted history matching tool is presented that can provide multiple solutions of history matching in much less number of numerical simulations. The proposed tool uses Gaussian process based proxy models to provide fast approximate forward solutions which are used in Bayesian optimization to find history match solutions in an iterative manner. The uncertainty in history match solutions is quantified via MCMC sampling on trained forward GP model. In order to check for independence and convergence of the collected MCMC samples, auto-correlation plots and Geweke z-score diagnostic tests are also performed. The converged MCMC samples are then used to quantify uncertainty in EUR of gas reserves via a forecasting GP model. The proposed methodology is successfully applied to a synthetic heterogeneous coalbed methane reservoir. The code and data for this case study is also available online for future studies. The results show that history matching can be performed in approximately four times less number of numerical simulation runs as compared to state of the art differential evolution algorithm. Besides, the P50 estimate of EUR is shown to be in close agreement with truth values for the presented case study.

UR - http://www.scopus.com/inward/record.url?scp=85059740229&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85059740229&partnerID=8YFLogxK

M3 - Conference contribution

AN - SCOPUS:85059740229

T3 - Proceedings - SPE Annual Technical Conference and Exhibition

BT - SPE Annual Technical Conference and Exhibition 2018, ATCE 2018

PB - Society of Petroleum Engineers (SPE)

ER -

Rana S, Ertekin T, King GR. An efficient probabilistic assisted history matching tool using Gaussian processes proxy models: Application to coalbed methane reservoir. In SPE Annual Technical Conference and Exhibition 2018, ATCE 2018. Society of Petroleum Engineers (SPE). 2018. (Proceedings - SPE Annual Technical Conference and Exhibition).