Yann Lagalaye¹, Eric Chaput¹, Thierry Rives¹
(1) Elf Exploration Production, Pau, France

Abstract: Segmented faults: Dynamic behavior in well test simulations

Well testing is commonly used to infer the presence of lateral barriers like sealing faults. Indeed, in classical well test interpretations, faults are modeled as simple homogeneous linear reduction of transmissibility. Nevertheless, geological observations indicate faults are very often segmented leaving a path to flow. The availability of well testing to discriminate different types of faults has been investigated. Four synthetic cases have been modeled with a dynamic flow simulator: infinite continuous sealing and partially sealing faults, finite sealing fault and segmented fault. The well test signatures are very similar in all cases and differentiation on a well test interpretation is difficult. When the reduction of transmissibility due to partially sealing fault is high (inferior to 0.001) or when the relay zone of a segmented fault is far from the well, the derivative signature is similar to the sealing fault case. Small reduction of transmissibility for partially sealing faults or short distance between well location and relay zone of a segmented fault induce differences of the derivative compared to the sealing fault case. In both cases, a Linear Composite three Zones model fit well with the simulation curves. Meanwhile, the calculated distance to the fault plane, or to other limits, will be wrong if the segmented fault is not identified. Consequences for production strategies could be dramatic in particular for multiphase reservoirs, where the possibility of very localized flow within relay zones can lead to gas or water breakthrough. Case studies illustrate possible misinterpretation of well test without geological input.

AAPG Search and Discovery Article #90914©2000 AAPG Annual Convention, New Orleans, Louisiana