Prediction of PVT properties in crude oil systems using support vector machines

Jawad Nagi, Sieh Kiong Tiong, Syed Khaleel Ahmed, Farrukh Nagi

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

2 Citations (Scopus)

Abstract

Calculation of reserves in an oil reservoir and the determination of its performance and economics require good knowledge of its physical properties. Accurate determination of the pressure-volume-temperature (PVT) properties such as the bubble point pressure (Pb) and the oil formation volume factor (Bob) are important in the primary and subsequent development of an oil field. This paper proposes Support Vector Machines (SVMs) as a novel machine learning technique for predicting outputs in uncertain situations using the ε-Support Vector Regression (ε-SVR) method. The objective of this research is to investigate the capability of SVRs in modeling PVT properties of crude oil systems and solving existing Artificial Neural Network (ANN) drawbacks. Three datasets used for training and testing the SVR prediction model were collected from distinct published sources. The ε-SVR model incorporates four input features from the datasets: (1) solution gas-oil ratio, (2) reservoir temperature, (3) oil gravity and, (4) gas relative density. A comparative study is carried out to compare ε-SVR performance with ANNs, nonlinear regression, and different empirical correlation techniques. The results obtained reveal that the ε-SVR once successfully trained and optimized is more accurate, reliable, and outperforms the other existing approaches such as empirical correlation for estimating crude oil PVT properties.

Original languageEnglish
Title of host publicationICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment
Subtitle of host publicationAdvancement Towards Global Sustainability
Pages1-5
Number of pages5
DOIs
Publication statusPublished - 01 Dec 2009
Event2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability, ICEE 2009 - Malacca, Malaysia
Duration: 07 Dec 200908 Dec 2009

Publication series

NameICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability

Other

Other2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability, ICEE 2009
CountryMalaysia
CityMalacca
Period07/12/0908/12/09

Fingerprint

Support vector machines
Crude oil
Temperature
Oil fields
Gas oils
Learning systems
Gravitation
Physical properties
Neural networks
Economics
Testing
Gases
Oils

All Science Journal Classification (ASJC) codes

  • Renewable Energy, Sustainability and the Environment
  • Environmental Engineering

Cite this

Nagi, J., Tiong, S. K., Khaleel Ahmed, S., & Nagi, F. (2009). Prediction of PVT properties in crude oil systems using support vector machines. In ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability (pp. 1-5). [5398681] (ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability). https://doi.org/10.1109/ICEENVIRON.2009.5398681
Nagi, Jawad ; Tiong, Sieh Kiong ; Khaleel Ahmed, Syed ; Nagi, Farrukh. / Prediction of PVT properties in crude oil systems using support vector machines. ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability. 2009. pp. 1-5 (ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability).
@inproceedings{0f596892aac945019b2c54cb60fc46d8,
title = "Prediction of PVT properties in crude oil systems using support vector machines",
abstract = "Calculation of reserves in an oil reservoir and the determination of its performance and economics require good knowledge of its physical properties. Accurate determination of the pressure-volume-temperature (PVT) properties such as the bubble point pressure (Pb) and the oil formation volume factor (Bob) are important in the primary and subsequent development of an oil field. This paper proposes Support Vector Machines (SVMs) as a novel machine learning technique for predicting outputs in uncertain situations using the ε-Support Vector Regression (ε-SVR) method. The objective of this research is to investigate the capability of SVRs in modeling PVT properties of crude oil systems and solving existing Artificial Neural Network (ANN) drawbacks. Three datasets used for training and testing the SVR prediction model were collected from distinct published sources. The ε-SVR model incorporates four input features from the datasets: (1) solution gas-oil ratio, (2) reservoir temperature, (3) oil gravity and, (4) gas relative density. A comparative study is carried out to compare ε-SVR performance with ANNs, nonlinear regression, and different empirical correlation techniques. The results obtained reveal that the ε-SVR once successfully trained and optimized is more accurate, reliable, and outperforms the other existing approaches such as empirical correlation for estimating crude oil PVT properties.",
author = "Jawad Nagi and Tiong, {Sieh Kiong} and {Khaleel Ahmed}, Syed and Farrukh Nagi",
year = "2009",
month = "12",
day = "1",
doi = "10.1109/ICEENVIRON.2009.5398681",
language = "English",
isbn = "9781424451456",
series = "ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability",
pages = "1--5",
booktitle = "ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment",

}

Nagi, J, Tiong, SK, Khaleel Ahmed, S & Nagi, F 2009, Prediction of PVT properties in crude oil systems using support vector machines. in ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability., 5398681, ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability, pp. 1-5, 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability, ICEE 2009, Malacca, Malaysia, 07/12/09. https://doi.org/10.1109/ICEENVIRON.2009.5398681

Prediction of PVT properties in crude oil systems using support vector machines. / Nagi, Jawad; Tiong, Sieh Kiong; Khaleel Ahmed, Syed; Nagi, Farrukh.

ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability. 2009. p. 1-5 5398681 (ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability).

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

TY - GEN

T1 - Prediction of PVT properties in crude oil systems using support vector machines

AU - Nagi, Jawad

AU - Tiong, Sieh Kiong

AU - Khaleel Ahmed, Syed

AU - Nagi, Farrukh

PY - 2009/12/1

Y1 - 2009/12/1

N2 - Calculation of reserves in an oil reservoir and the determination of its performance and economics require good knowledge of its physical properties. Accurate determination of the pressure-volume-temperature (PVT) properties such as the bubble point pressure (Pb) and the oil formation volume factor (Bob) are important in the primary and subsequent development of an oil field. This paper proposes Support Vector Machines (SVMs) as a novel machine learning technique for predicting outputs in uncertain situations using the ε-Support Vector Regression (ε-SVR) method. The objective of this research is to investigate the capability of SVRs in modeling PVT properties of crude oil systems and solving existing Artificial Neural Network (ANN) drawbacks. Three datasets used for training and testing the SVR prediction model were collected from distinct published sources. The ε-SVR model incorporates four input features from the datasets: (1) solution gas-oil ratio, (2) reservoir temperature, (3) oil gravity and, (4) gas relative density. A comparative study is carried out to compare ε-SVR performance with ANNs, nonlinear regression, and different empirical correlation techniques. The results obtained reveal that the ε-SVR once successfully trained and optimized is more accurate, reliable, and outperforms the other existing approaches such as empirical correlation for estimating crude oil PVT properties.

AB - Calculation of reserves in an oil reservoir and the determination of its performance and economics require good knowledge of its physical properties. Accurate determination of the pressure-volume-temperature (PVT) properties such as the bubble point pressure (Pb) and the oil formation volume factor (Bob) are important in the primary and subsequent development of an oil field. This paper proposes Support Vector Machines (SVMs) as a novel machine learning technique for predicting outputs in uncertain situations using the ε-Support Vector Regression (ε-SVR) method. The objective of this research is to investigate the capability of SVRs in modeling PVT properties of crude oil systems and solving existing Artificial Neural Network (ANN) drawbacks. Three datasets used for training and testing the SVR prediction model were collected from distinct published sources. The ε-SVR model incorporates four input features from the datasets: (1) solution gas-oil ratio, (2) reservoir temperature, (3) oil gravity and, (4) gas relative density. A comparative study is carried out to compare ε-SVR performance with ANNs, nonlinear regression, and different empirical correlation techniques. The results obtained reveal that the ε-SVR once successfully trained and optimized is more accurate, reliable, and outperforms the other existing approaches such as empirical correlation for estimating crude oil PVT properties.

UR - http://www.scopus.com/inward/record.url?scp=77949584632&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77949584632&partnerID=8YFLogxK

U2 - 10.1109/ICEENVIRON.2009.5398681

DO - 10.1109/ICEENVIRON.2009.5398681

M3 - Conference contribution

SN - 9781424451456

T3 - ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability

SP - 1

EP - 5

BT - ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment

ER -

Nagi J, Tiong SK, Khaleel Ahmed S, Nagi F. Prediction of PVT properties in crude oil systems using support vector machines. In ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability. 2009. p. 1-5. 5398681. (ICEE 2009 - Proceeding 2009 3rd International Conference on Energy and Environment: Advancement Towards Global Sustainability). https://doi.org/10.1109/ICEENVIRON.2009.5398681