Machine learning to reduce cycle time for time-lapse seismic data assimilation into reservoir management

Abstract

Time-lapse (4D) seismic is widely deployed in offshore operations to monitor improved oil recovery methods including water flooding, yet its value for enhanced well and reservoir management is not fully realized due to the long cycle times required for quantitative 4D seismic data assimilation into dynamic reservoir models. To shorten the cycle, we have designed a simple inversion workflow to estimate reservoir property changes directly from 4D attribute maps using machine-learning (ML) methods. We generated tens of thousands of training samples by Monte Carlo sampling from the rock-physics model within reasonable ranges of the relevant parameters. Then, we applied ML methods to build the relationship between the reservoir property changes and the 4D attributes, and we used the learnings to estimate the reservoir property changes given the 4D attribute maps. The estimated reservoir property changes (e.g., water saturation changes) can be used to analyze injection efficiency, update dynamic reservoir models, and support reservoir management decisions. We can reduce the turnaround time from months to days, allowing early engagements with reservoir engineers to enhance integration. This accelerated data assimilation removes a deterrent for the acquisition of frequent 4D surveys.

Deep learning a poroelastic rock-physics model for pressure and saturation discrimination

“…Compared to traditional inversion, neural networks can provide a significant improvement in turnaround time. Xue et al (2019) apply different machine-learning techniques (e.g., neural networks and random forests) for mapping saturation changes by analyzing normalized root-mean square amplitude changes and normalized differences of the reflection coefficient.…”

“…The poroelastic attributes V P , V S , ρ, Q P , and Q S are taken as predetermined, either by inversion or direct measurements from cross-well experiments, serving as observation data on which rock-physics parameters will be calibrated in a similar way as Xue et al (2019). In the present paper, porosity and pressure prior to injection are defined as additional poroelastic attributes affecting the rock physics.…”

BC10 O-North multi-attributes seismic inversion for saturation changes using ML

“…To fill the gap between qualitative interpretation of 4D attribute maps and 4D inversion, several ML and DL based workflows (Cao et al 2017, Xue et al 2019, Côrte et al 2020) are used to predict the maps of reservoir property changes. The ideas are similar in a way that using machine-learning model trained from the synthetic data to replace the nonlinear physics models (rock and fluids models and calculation of seismic reflection coefficients).…”

Deep Learning a Poro-Elastic Rock Physics Model for Pressure and Saturation Discrimination

“…This allows for near-real-time results and therefore improves the operational value of seismic data (Moseley et al, 2018). The poroelastic attributes V P , V S , ρ, Q P , and Q S are taken as predetermined, either by inversion or direct measurements from cross-well experiments, serving as observation data on which rock-physics parameters will be calibrated in a similar way as Xue et al (2019). In the present paper, porosity and pressure prior to injection are defined as additional poroelastic attributes affecting the rock physics.…”