Function (biology)

Psychology

Flooding (psychology)

Geotechnical engineering

Production (economics)

Reservoir simulation

Engineering

Economics

Hydrogeology

Psychotherapist

Well control

Mathematics

Evolutionary biology

Biology

Mathematical optimization

Macroeconomics

Drilling

Net present value

Computer science

Petroleum engineering

Petrophysics

Porosity

Mechanical engineering

Computational Geosciences

In a recent paper, we developed a physics-based data-driven model referred to as INSIM-FT and showed that it can be used for history matching and future reservoir performance predictions even when no prior geological model is available. The model requires no prior knowledge of petrophysical properties. In this work, we explore the possibility of using INSIM-FT in place of a reservoir simulation model when estimating the well controls that optimize water flooding performance where we use the net present value (NPV) of life-cycle production as our cost (objective) function. The well controls are either the flowing bottom-hole pressure (BHP) or total liquid rates at injectors and producers on the time intervals which represent the prescribed control steps. The optimal well controls that maximize NPV are estimated with an ensemble-based optimization algorithm using the history-matched INSIM-FT model as the forward model. We compare the optimal NPV obtained using INSIM-FT as the forward model with the estimate of the optimal NPV obtained using the correct full-scale reservoir simulation model when performing waterflooding optimization.

Waterflooding optimization with the INSIM-FT data-driven model