Petrophysics – The SPWLA Journal of Formation Evaluation and Reservoir Description · 2009 · 16 citations · 0 references
EngineeringGeological EngineeringPetroleum EngineeringFinite Element AnalysisReservoir CharacterizationReservoir EngineeringDirectional DrillingWell ModelingFault IdentificationGeologic ModelingHorizontal Well AnalysisHorizontal Well ModelingSubsurface Reservoir CharacterizationThree-dimensional Reservoir ModelingThree-dimensional Resistivity ModelingWell Placement
This paper presents a new highly efficient three-dimensional (3D) EM modeling library for 3D reservoir characterization workflows using LWD directional resistivity measurements. The parallelized library is part of a high-performance computing environment, fully integrated with reservoir models. The library includes finite-difference two-and-a-half-dimensional (2.5D) and 3D frequency domain solvers, a 3D static solver and a general purpose 3D finite-element solver, each targeting a different application. It can be used in pre-job planning to allow a better understanding of measurements, assisting in geosteering in complex scenarios and for detailed post-job analysis for improved reservoir characterization in high-angle and horizontal wells. We present sensitivity analyses of directional LWD EM measurements in complex scenarios including: nearby boundaries and faults; crossbedding; and invasion shape and size. Multiple effects can cause inconsistencies in model-based real-time interpretation and may affect accuracy of inversion results, such as distance to boundaries and formation resistivities. We show examples of subseismic fault identification during drilling and interpretation is refined based on 2.5D modeling. Directional EM responses are reconstructed by manually adjusting fault and layering dips and formation resistivities. We also present an example from a deepwater turbidite reservoir to demonstrate the use of 3D modeling. We jointly interpret LWD directional EM data with laterolog LWD measurements and images to estimate the fault azimuth and to refine the model built in real-time, reducing the uncertainty in structure interpretation.