Applying an optimized low risk model for fast history matching in a giant oil reservoir

Mojtaba Karimi, Ali Mortazavi, Mohammad Ahmadi

Research output: Contribution to journalReview article

Abstract

In this paper, the latest approaches for automated history matching (AHM) were applied to a real brown field having 14 active wells with multiple responses (production rate, bottom hole pressure and well block pressure) located in the south of Iran. A modified support vector machine was employed to create a proxy model incorporated based on design of experimental. Thereafter, all model parameters were adjusted to reproduce the observed history within the created proxy model. Accordingly, the proposed proxy model was successfully constructedusing1086samplesbasedonanR 2 coefficientofabout0.9forthetrainedandtestdataset.Finally,theprocess was optimized by two main algorithms to reach the best solutions, which are genetic and particle swarm optimization.

Original languageEnglish
Pages (from-to)84-89
Number of pages6
JournalKuwait Journal of Science
Volume46
Issue number1
Publication statusPublished - Jan 1 2019

Fingerprint

Bottom hole pressure
Particle swarm optimization (PSO)
Support vector machines
Oils

Keywords

  • Cubic centered face
  • Fast history matching
  • Least square support sector
  • Optimization

ASJC Scopus subject areas

  • General

Cite this

Applying an optimized low risk model for fast history matching in a giant oil reservoir. / Karimi, Mojtaba; Mortazavi, Ali; Ahmadi, Mohammad.

In: Kuwait Journal of Science, Vol. 46, No. 1, 01.01.2019, p. 84-89.

Research output: Contribution to journalReview article

@article{f43b651c4ce842edad12d5f49bea7e78,
title = "Applying an optimized low risk model for fast history matching in a giant oil reservoir",
abstract = "In this paper, the latest approaches for automated history matching (AHM) were applied to a real brown field having 14 active wells with multiple responses (production rate, bottom hole pressure and well block pressure) located in the south of Iran. A modified support vector machine was employed to create a proxy model incorporated based on design of experimental. Thereafter, all model parameters were adjusted to reproduce the observed history within the created proxy model. Accordingly, the proposed proxy model was successfully constructedusing1086samplesbasedonanR 2 coefficientofabout0.9forthetrainedandtestdataset.Finally,theprocess was optimized by two main algorithms to reach the best solutions, which are genetic and particle swarm optimization.",
keywords = "Cubic centered face, Fast history matching, Least square support sector, Optimization",
author = "Mojtaba Karimi and Ali Mortazavi and Mohammad Ahmadi",
year = "2019",
month = "1",
day = "1",
language = "English",
volume = "46",
pages = "84--89",
journal = "Kuwait Journal of Science",
issn = "2307-4108",
publisher = "Kuwait University",
number = "1",

}

TY - JOUR

T1 - Applying an optimized low risk model for fast history matching in a giant oil reservoir

AU - Karimi, Mojtaba

AU - Mortazavi, Ali

AU - Ahmadi, Mohammad

PY - 2019/1/1

Y1 - 2019/1/1

N2 - In this paper, the latest approaches for automated history matching (AHM) were applied to a real brown field having 14 active wells with multiple responses (production rate, bottom hole pressure and well block pressure) located in the south of Iran. A modified support vector machine was employed to create a proxy model incorporated based on design of experimental. Thereafter, all model parameters were adjusted to reproduce the observed history within the created proxy model. Accordingly, the proposed proxy model was successfully constructedusing1086samplesbasedonanR 2 coefficientofabout0.9forthetrainedandtestdataset.Finally,theprocess was optimized by two main algorithms to reach the best solutions, which are genetic and particle swarm optimization.

AB - In this paper, the latest approaches for automated history matching (AHM) were applied to a real brown field having 14 active wells with multiple responses (production rate, bottom hole pressure and well block pressure) located in the south of Iran. A modified support vector machine was employed to create a proxy model incorporated based on design of experimental. Thereafter, all model parameters were adjusted to reproduce the observed history within the created proxy model. Accordingly, the proposed proxy model was successfully constructedusing1086samplesbasedonanR 2 coefficientofabout0.9forthetrainedandtestdataset.Finally,theprocess was optimized by two main algorithms to reach the best solutions, which are genetic and particle swarm optimization.

KW - Cubic centered face

KW - Fast history matching

KW - Least square support sector

KW - Optimization

UR - http://www.scopus.com/inward/record.url?scp=85062540133&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85062540133&partnerID=8YFLogxK

M3 - Review article

VL - 46

SP - 84

EP - 89

JO - Kuwait Journal of Science

JF - Kuwait Journal of Science

SN - 2307-4108

IS - 1

ER -