Validation of Fault Permeability and Transmissibility Algorithms Using Production Histories

Kachi Onyeagoro¹, Steve J. Naruk¹, Frans F. van der Vlugt¹, Dick Eikmans³, Ellen Zijlstra³, Martin De Keijzer³, Steve Jolley³, Mette Kristensen³, Paul Reemst³, Charles Tixier², and Richard King^4
1Shell, Bellaire, Houston, TX
²SEPCO, New Orleans, LA
³Shell Europe, Aberdeen, United Kingdom
⁴Shell Malaysia, Sarawak, Malaysia

Fault compartmentalization can of course have a major impact on well performance. In some fields, the production data show that the faults are actually not significant baffles. In others, however, production shows that the faults are major baffles, retaining thousands of psi production-induced pressure differences without breaking down, even in cases where the fault displacement is much less than the reservoir thickness, and the faults appeared to be non-sealing prior to production. Historically, industry has been unable to predict these differences in cross fault flow, and in reservoir simulations it has been modeled by ad-hoc history-match tuning parameters. Given this uncertainty in fault modeling, and the current business need to simplify and accelerate reservoir modeling, the current trend is to simply ignore small intra-reservoir faults, potentially leading to significant production forecast downgrades.

The industry-standard SGR-based fault permeability and transmissibility equations yield demonstrably too much cross fault flow when tested in reservoir simulation models for fields having extended production histories. Conversely, the same functions may yield too little cross fault flow in some high permeability reservoirs recently developed in deep water West Africa. Consequently, Shell has developed new fault permeability and transmissibility functions that include variables for the initial reservoir permeability, the effects of cataclasis and the formation of phyllosilicate-fault-framework rocks, and the fault rock capillary entry pressure. “Close-the-loop” tests of the new function in fields with extended production histories have yielded good history matches in multiple wells using the same input parameters.

AAPG International Conference and Exhibition, Cape Town, South Africa 2008 © AAPG Search and Discovery