Estimation of pseudo relative permeability curves for a heterogeneous reservoir with a new automatic history matching algorithm

• Development of a new automatic history matching algorithm to estimate pseudo curves.
• Applying a log transformation to ensure the monotonicity of pseudo curves.
• Comparing the performance of generated pseudo curves with dynamic pseudo functions.
• Investigating the effect of pseudo curve estimation uncertainty on the oil recovery.
• Examining the predictability of the newly generated pseudo curves.

Reliable relative permeability curves are important for successful reservoir simulation and modeling. These curves may be available on fine scale models. However, detailed descriptions demand a huge amount of memory and processing time. Therefore, it is needed to upscale core-size relative permeability curves measured by laboratory experiments to a scale suitable for the various types of simulation. This upscaling is accomplished with pseudo functions. In this work, a new automatic history matching algorithm based on B-spline representation was applied to estimate pseudo relative permeability curves. The performance of the generated pseudo curves were compared with four dynamic pseudo functions for a 3D heterogeneous water-flooding case. In addition, the effect of the pseudo function estimation uncertainty on the cumulative oil production was investigated after 10 years of production. Results revealed that relative permeability curves obtained from history matching gives the best performance among other pseudos and this methodology can be used for reservoir simulation over periods beyond that of the original match.