Title | ||
---|---|---|
A Novel Data Analytics Framework for the History Matching Problem for Reservoir Simulation in Up-Stream Petrochemical Industry |
Abstract | ||
---|---|---|
Matching the historical production of well and field with the simulated output is considered to be a technically challenging problem for the up-stream petrochemical industry. Even though a majority of industry solutions rely on the RMS analysis, it is widely accepted in the industry that RMS based solutions fail to provide any meaningful insight to the problem and fail to identify phenomenon like “Early Breakthrough” and “Late Breakthrough” and so on. In this paper, we discuss the “History Matching” problem and provide an overview of the data analytics based solution framework we developed in conjunction with a leading oil exploration company based in the middle-east. As opposed to the RMS based analysis, the core of our solution is based on the “Dynamic Time Warping” `(DTW), a proven technique for time series classification and temporal similarity matching exercises. We use the DTW distances scores and output from the DTW analysis to formulate the features that are provided as input to the machine learning algorithms for classification of simulation models. |
Year | DOI | Venue |
---|---|---|
2020 | 10.1109/SysCon47679.2020.9275909 | 2020 IEEE International Systems Conference (SysCon) |
Keywords | DocType | ISSN |
Dynamic Time Warping,Machine Learning,Artificial Intelligence,History Matching,Reservoir simulation | Conference | 1944-7620 |
ISBN | Citations | PageRank |
978-1-7281-5366-7 | 0 | 0.34 |
References | Authors | |
0 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Sumit Kumar Bose | 1 | 0 | 0.34 |
Ben Amaba PE | 2 | 0 | 0.34 |
Stephen Lord | 3 | 0 | 0.34 |