US 12,066,586 B2
Lithofacies guided core description using unsupervised machine learning
Yunsheng Li, Dhahran (SA); Weihua Wang, Dhahran (SA); and Maan A. Hawi, Dhahran (SA)
Assigned to Saudi Arabian Oil Company, Dhahran (SA)
Filed by Saudi Arabian Oil Company, Dhahran (SA)
Filed on Apr. 18, 2022, as Appl. No. 17/722,964.
Prior Publication US 2023/0333277 A1, Oct. 19, 2023
Int. Cl. G01V 11/00 (2006.01); E21B 47/00 (2012.01); E21B 49/00 (2006.01); G06N 3/02 (2006.01); G06N 3/088 (2023.01); G06N 20/00 (2019.01)
CPC G01V 11/00 (2013.01) [E21B 49/005 (2013.01); E21B 47/00 (2013.01); E21B 49/00 (2013.01); E21B 2200/20 (2020.05); E21B 2200/22 (2020.05); G06N 3/02 (2013.01); G06N 3/088 (2013.01); G06N 20/00 (2019.01)] 17 Claims
OG exemplary drawing
 
1. A computer-implemented method, comprising:
obtaining, with one or more hardware processors, well log curves for training wells in a each respective zone of a formation comprising multiple zones;
preparing, with the one or more hardware processors, the well log curves using zone data to limit the prepared well log curves to each respective zone;
training, with the one or more hardware processors, a machine learning model with input data comprising the prepared well log curves for each respective zone of the multiple zones, wherein the trained machine learning models output predicted lithofacies for wells in each respective zone of the multiple zones;
predicting, with the one or more hardware processors, lithofacies for wells in the each respective zone using the trained machine learning models;
grouping, with the one or more hardware processors, the wells in the each respective zone based on the predicted lithofacies; and
updating, with the one or more hardware processors, core descriptions of the wells in each respective zone of the multiple zones based on the grouped wells.