Limited and unreliable data has been the bane of reservoir characterization. This is especially true for fields which were developed before the 70's, when field development was more art than science. Producers struggling with re-engineering such fields have to evaluate every shred of available information to optimize modern production operations. This is a common problem faced by most operators today.
In this Department of Energy (DOE) sponsored project, we characterize one such field, the Sulimar Queen, using recent technology coupled with conventional reservoir engineering. The Sulimar Queen - located in southeastern New Mexico - is a part of the Permian shelf. The producing formation is the upper Queen (Shattuck member) at 2000 ft. The original database comprised some pre-70's gamma ray and neutron logs, and production data from 35 wells. One well (well 1-16) contained modern gamma ray, resistivity, and porosity logs. In addition, a single core was available in this well. A successful characterization of the Queen was done based on this meager data set.
The methodology entailed a multidisciplinary team to characterize the Sulimar Queen. On recommendations from the team geologists, additional data was collected from the Sulimar Queen outcrops and other adjacent fields. Transient tests were conducted and advanced core analyses were performed on the single available core.
In this paper, we discuss the integration of all the assimilated data, which spans from the micro-scale (thin sections) to megascale (outcrops), to characterize the Sulimar Queen. Applications of new artificial intelligence tools, such as fuzzy-logic, to correlate thin-section data with permeability were developed. The use of geostatistics and simulated annealing to build simulation-scale property distributions are discussed. In addition, we present simulation scenarios based on honoring the dynamic production data with automatic history-matching, and 3D visualizations. The final version of the conditioned Sulimar Queen model includes a previously undetected gas cap.
In conclusion, a geologic model of the Sulimar Queen honoring the observed data is presented. This geologic model incorporates all major depositional events and their effects on reservoir properties.
An understanding of the reservoir structure and developing characteristic measures for heterogeneity classification are essential to maximize oil production from producing reservoirs. The physical phenomena associated with oil recovery have been relatively well understood for some time. Nevertheless, there have been disappointing gaps between model predictions based on laboratory and field tests, and the actual production of oil. The scarcity of detailed reservoir data has contributed to such failures. A unique opportunity for a detailed and integrated study existed at the Sulimar Queen, which was available as a research field.
One goal of the Sulimar Queen study was to use the integrated data to explain its waterflood success and to share the insight gained in this study with operators re-engineering reservoirs similar to the Sulimar Queen. The reservoir characterization methodology, resulting from history-matching the primary production to predict secondary recovery performance, can be readily utilized by oil and gas producers.
The challenge of the Sulimar Queen project was to develop a reasonable reservoir model using limited and old data - a common problem with reservoirs in the Permian Basin. Typically, old well logs (gamma ray-neutron perforating logs) and production history are the only data available for this maturing area. The methodology developed to deal with such data will be briefly explained, addressing each data type in an ascending order of scale.
Sulimar Queen: A Typical Old Field. The Sulimar Queen unit is typical of many old fields lacking high-quality reservoir data. In most cases, this situation prevents an operator from investigating the future potential of such fields. For example, an infill drilling program would require at least the spatial mapping of key reservoir properties.