Enhanced Reservoir Description via Areal Data Integration and Reservoir Fluid Geodynamics: A Case Study From Deepwater Gulf of Mexico

Abstract

Accurate reservoir characterization is vital for effective decisions made throughout the life cycle of an oilfield reservoir, including management and development. Of all the components of reservoir description, hydraulic connectivity carries the highest amount of uncertainty, where inaccurate connectivity evaluation often results in production underperformance. Shortcomings are faced when applying conventional approaches of connectivity assessment. Seismic surveys are not always sufficient to evaluate lateral connectivity as detected faults can be transmissive or partially transmissive, while some faults are below the detection limits of seismic amplitude measurements. Vertical connectivity represents another uncertainty, where pressure measurements and well logs are often either unable to detect the baffles along oil columns or cannot assess whether detected baffles are relevant seals or flow diverters. Although conventional downhole fluid analysis (DFA) workflows have proven effective in delineating reservoir connectivity, enough DFA data are not always available, and with added complexity, uncertainties arise. Additionally, while equilibrated asphaltene gradients, measured through DFA probes, imply connectivity, ongoing reservoir fluid geodynamics (RFG) processes, such as current hydrocarbon charging, can preclude equilibration in a connected reservoir. Thus, a comprehensive assessment approach, that utilizes all available data streams, is needed to overcome the significant spatial complexity associated with moderately and heavily faulted reservoirs. In this paper, we employed our recently introduced interpretation workflow to evaluate the connectivity of a heavily faulted reservoir in the deepwater Gulf of Mexico. The field was divided into five investigation areas penetrated by 12 wells. Areal downhole fluid analysis (ADFA) was applied to assess local connectivity leading to reservoir-scale connectivity. Through integrating fluid/dynamic and rock/static data, each data type provided insights that were pieced together to enhance consistency and reduce uncertainty. Analyzed data included pressure-volume-temperature (PVT) reports, pressure surveys, well logs, and geochemistry. The study resulted in a verifiable connectivity description where faults, previously regarded as sealing, were classified into sealing or partially transmissive faults; unresolved faults were detected. Fault-block migration was detected, and fault throw was estimated; asphaltenes behavior was used to deduce original field structures prior to faulting. We also examined RFG processes to investigate oil biodegradation, where an asphaltene clustering trend was observed, causing high oil viscosities toward the bottom of one sandstone. A correlation was then derived and successfully implemented to estimate oil viscosity.