Datapages, Inc.Print this page

Analyzing Effects of Architecture and Two-Phase Flow Properties of Normal Faults

 

Berg, S.S., E. Øian, N. Fredman, S.L. Semshaug, T. Skar, A. Braathen, University of Bergen, Bergen, Norway

 

Effects of faults on fluid flow are often difficult to predict. Ranking of uncertainties asso­ciated with fault modeling is therefore a key issue in the development of new workflows in the petroleum industry. The aim of this contribution is first to consider possible flow effects of faults with different architectures and two-phase flow properties, and secondly to illus­trate a framework for analyzing uncertainties associated with flow in faults.

Three theoretical scenarios of faults in a production setting are considered: (1) Homogenous, strongly water wet faults, where injection of water increases the sealing capacity. (2) Heterogeneous, water-wet faults where less deformed rock bodies act as con­duits to flow. (3) Homogenous, strongly oil wet faults acting as barriers to flow of water. The scenarios are based on qualitative measures for initial fluid saturations, permeability, capil­lary pressure, and wetting properties, and illustrate how multi-phase flow properties add to the general complexity of faults.

A numerical and geological framework for analyzing relative effects of fault complexity and multi-phase flow properties on fluid flow has been developed. The framework is based on the hierarchical structure of faults and enables analyses of structures on different scales. The scales are defined relative to the grid cell sizes, and are represented by different upscal­ing techniques. The framework is illustrated through a simple fault zone model with a cen­tral fault core surrounded by subsidiary faults and deformation bands. The hierarchical framework allows for comparison of upscaling procedures as well as sensitivity analyses of structural features on different scales and with different flow properties. It provides a tool for delimiting the relative effects of structural uncertainty.