A hybrid assistant history matching algorithm considering tracer studies

This paper presents a hybrid history matching algorithm based on separable natural evolution strategies (sNES) and Adjoint-based tracer studies gradient. We demonstrate the effectiveness of the approach conditioning porosity and permeability fields on the historical production data and information about tracer studies. Tracer studies and fluid flow in the reservoir is numerically modeled. This approach to the joint assistant history matching problem significantly reduces the number of reservoir simulator runs with similar quality to the resulting models.