SPE Live Distinguished Lecturer Series: Waterflood Optimization by Data Analytics on Mature Fields Accelerate the Field Developing Process from Months to Weeks

Wednesday, February 15, 2023 | 08:00:AM - 08:30:AM CT

Add to Calendar 02/15/2023 08:00 AM 02/15/2023 08:30 AM America/Chicago SPE Live Distinguished Lecturer Series: Waterflood Optimization by Data Analytics on Mature Fields Accelerate the Field Developing Process from Months to Weeks https://streaming.spe.org/spe-live-distinguished-lecturer-series-waterflood-optimization-by-data-analytics-on-mature-fields-accelerate-the-field-developing-process-from-months-to-weeks

Currently, 70 percent of global oil is being produced from mature fields. The waterflood process is widely used to improve the oil recovery of mature fields. Unlocking the potential of existing mature assets and the optimization of waterflooding can be challenging using traditional dynamic modelling workflows due to high well count, years of historical production data (extensive data set), complex stratigraphy, etc.; however, these challenges create opportunities for applying data-driven techniques. In addition, increased attention towards reducing carbon emissions makes data-driven techniques more attractive as they are computationally-light techniques.First, this work reviews classical analytics for studying the waterflood process; after that, machine learning techniques are used to optimize rates of water injector wells for maximizing the production efficiency. Next, an innovative hybrid physics-guided data-driven method is presented to accelerate field development and locate the remaining oil (LTRO) process from months to weeks. The results of this new workflow are validated against outputs of the numerical simulation and 4D seismic info. Furthermore, the comparison of post drilling results and predictions of the new physics-guided data-driven workflow is presented for a field located in the Middle East. During this SPE Live, the SPE distinguished lecturer, Babak Moradi, demonstrates that it is now possible to deliver digital LTRO projects, capturing the full uncertainty ranges, including complex multi-vintage spatial 4D datasets, and providing reliable non-simulation physics-compliant data-driven production forecasts within weeks. This presentation addresses the challenge of reducing mature fields study resource intensity and therefore, time and costs whilst maintaining a high degree of fidelity.
https://streaming.spe.org/spe-live-distinguished-lecturer-series-waterflood-optimization-by-data-analytics-on-mature-fields-accelerate-the-field-developing-process-from-months-to-weeks
Currently, 70 percent of global oil is being produced from mature fields. The waterflood process is widely used to improve the oil recovery of mature fields. Unlocking the potential of existing mature assets and the optimization of waterflooding can be challenging using traditional dynamic modelling workflows due to high well count, years of historical production data (extensive data set), complex stratigraphy, etc.; however, these challenges create opportunities for applying data-driven techniques. In addition, increased attention towards reducing carbon emissions makes data-driven techniques more attractive as they are computationally-light techniques.First, this work reviews classical analytics for studying the waterflood process; after that, machine learning techniques are used to optimize rates of water injector wells for maximizing the production efficiency. Next, an innovative hybrid physics-guided data-driven method is presented to accelerate field development and locate the remaining oil (LTRO) process from months to weeks. The results of this new workflow are validated against outputs of the numerical simulation and 4D seismic info. Furthermore, the comparison of post drilling results and predictions of the new physics-guided data-driven workflow is presented for a field located in the Middle East.
During this SPE Live, the SPE distinguished lecturer, Babak Moradi, demonstrates that it is now possible to deliver digital LTRO projects, capturing the full uncertainty ranges, including complex multi-vintage spatial 4D datasets, and providing reliable non-simulation physics-compliant data-driven production forecasts within weeks. This presentation addresses the challenge of reducing mature fields study resource intensity and therefore, time and costs whilst maintaining a high degree of fidelity.

Tags: , , ,

SPE Live Distinguished Lecturer Series: Waterflood Optimization by Data Analytics on Mature Fields Accelerate the Field Developing Process from Months to Weeks