Analysis of Uncertainty Trend for Estimated Ultimate Recovery Prediction of Shale Gas with Various Production Periods based on Machine Learning

Hyeonteak Oh; Seil Ki; Changhyup Park; Ilsik Jang

doi:10.32390/ksmer.2021.58.5.475

All Issue

2021 Vol.58, Issue 5S Preview Page Next Page

Research Paper

Analysis of Uncertainty Trend for Estimated Ultimate Recovery Prediction of Shale Gas with Various Production Periods based on Machine Learning 머신러닝 기반 생산기간에 따른 셰일가스 궁극가채량 예측의 불확실성 추이 분석

31 October 2021. pp. 475-490

PDF XML

Abstract

On account of the heterogeneity of shale reservoirs, their nano-scale permeability, and the complexity of their flow mechanism, production data analysis using machine learning has been widely used rather than physics-based models such as reservoir simulations. In this study, we performed the estimated ultimate recovery (EUR) prediction of gas wells in the Barnett Shale basin, using random forest (RF) models with various input data. Static data such as reservoir properties, completion information, and well information, and dynamic data such as the initial peak rate of gas, and monthly gas rates for 6 and 12 months, were used as the input data. The EUR prediction errors of the RF models for different input data were compared, and P90/P50/P10 was derived to analyze the change in confidence interval for the uncertainty analysis of the prediction. Conclusively, the usefulness and applicability of the RF model for decision-making processes was confirmed.

Keywords

Shale gas

Machine learning

Random forest

EUR

Uncertainty analysis

셰일저류층의 불균질성, 나노 스케일의 유체투과도 및 유동 메커니즘의 복잡성으로 인해 저류층 시뮬레이션과 같은 물리적 모델 기반 분석보다는 머신러닝을 활용한 생산자료 분석법이 활발히 사용되고 있다. 이 연구에서는 랜덤 포레스트 모델을 활용하여 입력자료에 따른 Barnett Shale의 가스 생산정에 대한 생산성 분석을 수행하였다. 입력자료는 저류층 물성, 완결조건, 생산정 정보 등의 정적자료와 가스의 초기 최대 생산량, 6개월 및 12개월 월별 생산량 등 동적자료로 구성되었다. 다양한 입력자료에 대한 랜덤 포레스트 모델의 궁극가채량 예측오차를 비교하였으며, 예측의 불확실성 분석을 위해 P90/P50/P10을 도출하여 신뢰구간의 변화 추이를 분석한 결과, 랜덤 포레스트 모델의 유용성과 활용성을 확인하였다.

키워드

셰일가스

머신러닝

랜덤 포레스트

궁극가채량

불확실성 분석

References

Ahn, S., Park, C., Kim, J., and Kang, J., 2018, Data-driven inverse modeling with a pre-trained neural network at heterogeneous channel reservoirs, Journal of Petroleum Science and Engineering, 170, p.785-796. 10.1016/j.petrol.2018.06.084

Ahsan, M., Mahmud, M., Saha, P., Gupta, K., and Siddique, Z., 2021, Effect of data scaling methods on machine Learning algorithms and model performance, Technologies, 9(52), p.1-10. 10.3390/technologies9030052

Alizadeh, B., Najjari, S., and Kadkhodaie-Ilkhchi A., 2012, Artificial neural network modeling and cluster analysis for organic facies and burial history estimation using well log data: a case study of the South Pars Gas Field, Persian Gulf, Iran, Computers & Geosciences, 45, p.261-269. 10.1016/j.cageo.2011.11.024

Amr, S., Ashhab, H. E., El-Saban, M., Shietinger, P., Caile, C., Kaheel, A., and Rodriguez, L., 2018, A large-scale study for a multi-basin machine learning model predicting horizontal well production, Paper presented at the 2018 SPE Annual Technical Conference and Exhibition, SPE, Dallas, Texas, USA. 10.2118/191538-MS

BorealisAI, 2021.09.10., https://www.borealisai.com/en/blog/tutorial-8-bayesian-optimization.

Bowker, K., 2007, Developement of the Barnett Shale Play, Fort Worth Basin, AAPG Bulletin, 91(4), p.1-13. 10.1306/06190606018

Breiman, L., 2001, Random Forest, Machine Learning, Kluwer Academic Publishers, Netherlands, p.5-32. 10.1023/A:1010933404324

Geron, A., 2019, Hands-on Machine Learning with Scikit-learn, Keras & Tensorflow, O'reilly, Sebastopol, CA, 260p.

Han, H., 2015, Introduction to Pattern Recognition, Hanbit media, Korea, 242p.

Hastie, T., Tibshirani, R., and Friedman, J., 2008, The Elements of Statistical Learning, Springer, Switzerland, 587p. 10.1007/978-0-387-84858-7_15

He, Q., 2017, Smart determination of estimated ultimate recovery in shale gas reservoir, Proceedings of the SPE Eastern Regional Meeting, Lexington, KY, USA. 10.2118/187514-MS

Iturrar, N. and Parra, J., 2014, Artificial neural networks applied to estimate permeability, porosity and intrinsic attenuation using seismic attributes and well-log data, Journal of Applied Geophysics, 107(1), p.45-54. 10.1016/j.jappgeo.2014.05.010

Jeong, S.H., 2020, Estimation of the Office Rent using the Machine Learning Methods -Focusing on the use of Random Forest, Artificial Neural Networks, Support Vector Machines-, MS Thesis, Hanyang University, Korea, 86p.

Ki, S., Jang, I., Cha, B., Seo, J., and Kwon O., 2020, Restoration of missing pressures in a gas well using recurrent neural networks with long short-term memory cells, Energies, 13(18), 4696, p.1-19. 10.3390/en13184696

Kim, J., Park, C., Lee, K., Ahn, S., and Jang I., 2020, Deep neural network coupled with distance-based model selection for efficient history matching, Journal of Petroleum Science and Engineering, 185, p.1-19. 10.1016/j.petrol.2019.106658

Kothari, S.N., 2011, An unconventional energy resources: shale gas, Paper presented at the Offshore Mediterranean Conference and Exhibition, Ravenna, Italy.

Krasser, M., 2021,09.10., http://krasserm.github.io/2018/03/21/bayesian-optimization.

Liang, Y. and Zhao, P., 2019, A Machine learning analysis based on big data for eagle ford shale formation, Paper presented at the SPE Annual Technical Conference and Exhibition, Calgary, Alberta, Canada. 10.2118/196158-MS

Robertson, S., 1988, Generalized Hyperbolic Equation, SPE, TX, USA, p.1-20.

Shin, H., Lim, J., and Jang, I., 2021, Probabilistic prediction of multi-wells production based on production characteristics analysis using key factors in shale formations, Energies, 14, p.1-30. 10.3390/en14175226

Shin, H.J., 2013, Estimated Ultimate Recovery Prediction using the Oil and Gas Production Decline Curve Analysis and Cash Flow Analysis in Resource Play, MS Thesis, Korea Maritime University, Korea, 56p.

Silversides, K., Melkumyan, A., Wyman, D., and Hatherly, P., 2015, Automated recognition of stratigraphic marker shales from geophysical logs in iron ore deposits, Computers & Geosciences, 77(1), p.118-125. 10.1016/j.cageo.2015.02.002

TodayEnergy, 2021.09.10., https://www.todayenergy.kr/news/articleView.html?idxno=233431.

Vyas, A. and Datta-Gupta, A., 2017, Modeling early time rate decline in unconventional reservoirs using machine learning techniques, Paper presented at the Abu Dhabi International Petroleum Exhibition & Conference, SPE, Abu Dhabi, UAE. 10.2118/188231-MS

Yoon, D.W., Kim, S.M., Park, G.S., Park, H.W., Byeon, J.M., Seo, J.W., Lee, C.W., Jang, I.S., Cho, S.J., and Choi, Y.S., 2018, The Fundamentals of Machine Learning in Resource development, CIR, Seoul, Korea, p.127-135.

Zerrouki, A., Aifa, T., and Baddari, K., 2014, Prediction of natural fracture porosity from well log data by means of fuzzy ranking and an artificial neural network in Hassi Messaoud oil field, Algeria, Journal of Petroleum Science and Engineering, 115(1), p.78-89. 10.1016/j.petrol.2014.01.011

Information

Publisher :The Korean Society of Mineral and Energy Resources Engineers
Publisher(Ko) :한국자원공학회
Journal Title :Journal of the Korean Society of Mineral and Energy Resources Engineers
Journal Title(Ko) :한국자원공학회지
Volume : 58
No :5
Pages :475-490
Received Date : 2021-09-13
Revised Date : 2021-10-06
Accepted Date : 2021-10-26
DOI :https://doi.org/10.32390/ksmer.2021.58.5.475

Journal of the Korean Society of Mineral and Energy Resources EngineersISSN:2288-0291(Print) 2288-2790(Online)한국자원공학회

All Issue