New Development Strategy using Automatic Well Placement by Artificial Intelegent Application

Abstract

Finding new infill well location was classically generated in reservoir simulation with trial- and-error manner of well placement. The process was usually consumed plenty of time in linear relationship with the reservoir model size. For time efficiency, a new method was implemented using the latest technology of reservoir simulator that was enhanced with artificial algorithm for forecast optimization and uncertainty quantification. This has been proven to optimize the well placement simulation in automatic way.

New strategy was introduced to include artificial intelligence and programming logic to enable user in rules modification within the simulation. Differential evolution optimization algorithm was used to find the global optimum scenario that resulted best well location as well as the coordinate and trajectory. Scripting was used to define well constraints and optimum oil and gas production rate setting as the objective function. Afterwards, a quality- check on the result will be required. More iteration with additional rules may be applied to achieve the most optimum result.

This is considered as an invention in reservoir simulation practices. Application can be varied based on the field specification. The implementation has been conducted successfully to several Pertamina reservoir models with higher time efficiency than before.