Rapid history matching of petroleum production from well logs and 4D seismic via Machine Learning techniques in the Norne Field, offshore Norway
Jones Ebinesan A , Greg Smith
A * and Ritu Gupta A
A Curtin University, Bentley, WA 6102, Australia.
The APPEA Journal 63 S227-S231 https://doi.org/10.1071/AJ22093
Accepted: 10 March 2023 Published: 11 May 2023
© 2023 The Author(s) (or their employer(s)). Published by CSIRO Publishing on behalf of APPEA.
Abstract
Predicting oil and gas reservoir behaviour requires multiple property modelling using various formulae, relationships and empirical techniques, which is time-consuming and often ineffective in precisely capturing non-linear dependencies. Artificial Intelligence (AI) or Machine Learning (ML) techniques can build time-series models for modelling dynamic reservoir properties such as water, oil and gas saturation, and pressure, thus capturing changes caused by hydrocarbon production. Here, 4D (time lapse) seismic surveys have been used to model the changes in water saturation using AI techniques such as multi-linear regression, multi-variable kriging and random forest. Statistical testing of the resulting 3D reservoir models using R-Squared, RMSE (root-mean-square error) and MAPE (mean absolute percentage error) indicated the random forest technique gave the best results and stratification had a negligible effect. Increasing the training size set from 10% to 80% improved the statistics as expected though the rate of improvement falls rapidly above a training dataset size of 40%. This indicates that 3D models with good accuracy could be obtained even with limited data. Similar techniques can be run to build 3D time-series pressure models and the results can be used for improved history matching, forward estimation of production data and estimation of reserves.
Keywords: artificial intelligence, AI, 4D seismic, Machine Learning, multi-variate kriging, Norne Field, random forest, reservoir simulation.
Jones Ebinesan has around 24 years’ experience in the IT industry, working in different technical roles at Dell, Tesco, CBH and NEC. Currently he is a sessional academic at Curtin University School of Mathematical Sciences, and Lecturer at the Kaplan Business School, teaching Probability, Statistics, Actuarial Science and Data Science. In his ‘spare time’ he researches the use of AI for reservoir simulation as part of his PhD in Mathematics and Earth Sciences. |
Gregory Smith is Adjunct Professor in Earth and Planetary Sciences, Curtin University. He has 40 years’ experience in Petroleum Geology, Geophysics and Geochemistry involving technical, research and managerial positions at Exxon, ARCO, BHP, Woodside/Shell and Herman Research Laboratory, including major exploration discoveries, field developments and production in Australia and overseas. Greg undertakes research with Honours, Masters and PhD Geology and Mathematics students interpreting seismic, logs, core and production datasets, for burial and thermal modelling, machine learning, organic petrology and diagenesis. Greg’s group has several research and consulting projects with industry and government. Member of AAPG, PESA, TSOP, ICCP, past corporate AusIMM and ASA. |
Ritu Gupta is an Associate Professor of Statistics and Data Science Coordinator at Curtin University in the School of Electrical Engineering, Computing and Mathematical Science and is a senior member of Curtin Centre for Optimisation and Decision Science. Ritu has 25 years of experience working in academia and industry. Ritu leads consulting and research projects in data science and predictive analytics for government and industry covering resources, defence and sports. She develops software for reserves estimation, conducts industry short courses and statistical audits, and supervises several PhD students. |
References
Biau G, Scornet E (2016) A Random Forest Guided Tour. TEST 25, 197–227.| A Random Forest Guided Tour.Crossref | GoogleScholarGoogle Scholar |
Breiman L (2001) ‘Random Forests.’ (University of California, Berkeley, CA)
Breiman L, Friedman J, Olshen R, Stone C (1984) ‘Classification and regression trees.’ (Routledge: New York, USA)
Dadashpour M, Rwechungura R, Eka S (2009) Norne Field case. In ‘ISAPP Symposium’. (NTNU)
Deutsch C, Journel A (1997) ‘GSLIB’, 2nd edn. (Oxford University Press: New York, USA)
Ho T (1995) Random Decision Forests. In ‘Proceedings of the 3rd International Conference on Document Analysis and Recognition’, Montreal, August 1995. Vol. 1, pp. 278–282. (IEEE)
International Association for Mathematical Geology (1991) ‘Studies in Mathematical Geology. Vol. 3’, 2nd edn. (Ed. RA Olea) (Oxford University Press)
Majani S (2018) ‘Production Optimization using Reservoir Recovery Techniques.’ (NTNU)
Makariou D, Barrieu P, Chen Y (2021) A random forest based approach for predicting spreads in the primary catastrophe bond market. Insurance: Mathematics and Economics 101, 140–162.
| A random forest based approach for predicting spreads in the primary catastrophe bond market.Crossref | GoogleScholarGoogle Scholar |
Na-udom A (2007) ‘Experimental design methodology for modelling response from computer simulated experiments.’ (Curtin University)
Norwegian Petroleum (2022) Norne. Retrieved from https://www.norskpetroleum.no/en/facts/field/norne/
Pan W, Torres-Verdín C, Duncan I, Pyrcz M (2022) Reducing the Uncertainty of Multi-Well Petrophysical Interpretation from Well Logs via Machine-Learning and Statistical Models [Preprint]. Earth ArXiv
| Reducing the Uncertainty of Multi-Well Petrophysical Interpretation from Well Logs via Machine-Learning and Statistical Models [Preprint].Crossref | GoogleScholarGoogle Scholar |
Rossiter D (2018) ‘Co-kriging with the gstat package of the R environment for statistical computing.’ (Cornell University)
Shahkarami A, Mohaghegh S (2015) Assisted History Matching Using Pattern Recognition Technology. Paper SPE-173405-MS, presented at the SPE Digital Energy Conference and Exhibition, Texas, USA, March 2015. (Society of Petroleum Engineers)
Tjia D (2016) ‘Statistical Methods for History Matching of Hydrological Model.’ (Curtin University)
Wang M, Feng D, Li D, Wang J (2022) Reservoir Parameter Prediction Based on the Neural Random Forest Model. Frontiers in Earth Science 10, 888933
| Reservoir Parameter Prediction Based on the Neural Random Forest Model.Crossref | GoogleScholarGoogle Scholar |
