Pore Space Estimation, Upscaling and Uncertainty Modelling for Multiphase Properties
Author: subsurfaceAI

This paper reports results from a sensitivity study aimed at understanding and evaluating uncertainty in a pore-to-field workflow. A methodology to identify and quantify uncertainty at different scales and to propagate, or upscale, this information to the reservoir simulation model has been developed and tested.  A case study is presented for a North Sea oil reservoir.

The pore-to-field workflow can be summarized as follows: On the basis of representative cores from the formations in question, lithofacies models are developed using a process based geo-modelling tool, which mimics real sedimentary geometries. The modeling software generates porosity and permeability distributions for the individual rock types (i.e., pore types) using correlated stochastic fields. The porosity is modeled by a truncated normal distribution and permeability is assumed to be lognormally distributed. Input parameters for this stochastic modelling are typically based on observations of permeability and porosity from core plugs, qualitative interpretation of the lithology from core slabs taken from representative wells, and mini-permeability measurements. The different rock types within each lithofacies are populated with multi-phase flow properties constructed using pore-scale network modelling. Effective (upscaled) flow parameters on the lithofacies scale are then obtained through two-phase, steady-state upscaling. These effective parameters are inserted into the geological model, and upscaling from the geological model to field-scale simulation model provides the desired field-scale relative permeability and capillary pressure curves.

Alf B. Rustad, Thomas G. Theting, and Rudolf J. Held, Statoil Hydro

Presented at  the SPE/DOE Symposium on Improved Oil Recovery, 20-23 April 2008, Tulsa, Oklahoma, USA

Order the complete paper from SPE