Title: Machine learning-based estimated ultimate recovery prediction and sweet spot evaluation of shale oil

Authors: Lufeng Zhan; Junwen Hu; Shaoyong Wang; Kun Wang; Bincheng Guo; Xuan Yang

Addresses: Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China ' Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China ' Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China ' Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China ' Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China ' Research Institute of Petroleum Exploration and Development (RIPED), PetroChina, Beijing 100083, China

Abstract: In this paper, we propose a machine learning-based, multi-discipline integrated evaluation workflow to evaluate sweet spots. We predict the estimated ultimate recovery (EUR) map and evaluate sweet spots at different oil price scenarios in the study area. The results show that the correlation coefficient between well EUR and predicted EUR is 0.9247. At the oil price of $40, $50, and $60/bbl, the sweet spot areas are 3.31 km², 27.75 km², and 51.61 km², and the total economically recoverable reserves are estimated to be 2.46 × 10⁵ t, 14.02 × 10⁵ t, 26.91 × 10⁵ t respectively. It is concluded that machine learning model is an excellent way to auto learn the relationship between complex shale reservoir variables with EUR. It is innovative to show the distribution of sweet spots according to oil price. This workflow can be wildly used in the shale oil exploration and development to evaluate investments and optimise well placement. [Received: August 26, 2020; Accepted: April 19, 2021]

Keywords: shale oil; sweet spot; machine learning; discounted cash flow; DCF; estimated ultimate recovery; estimated ultimate recovery; EUR; prediction; visualisation.

DOI: 10.1504/IJOGCT.2022.122089

International Journal of Oil, Gas and Coal Technology, 2022 Vol.30 No.1, pp.1 - 17

Received: 25 Aug 2020
Accepted: 19 Apr 2021
Published online: 08 Apr 2022 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article