4D Seismic Monitoring of a CO2-EOR Demonstration Project in a Desert Environment: Acquisition, Processing and Initial Results

Smith, Robert (Geophysics Technology, EXPEC Advanced Research Center) | Bakulin, Andrey (Geophysics Technology, EXPEC Advanced Research Center) | Jervis, Mike (Geophysics Technology, EXPEC Advanced Research Center) | Hemyari, Emad (Geophysics Technology, EXPEC Advanced Research Center) | Alramadhan, Abdullah (Geophysics Technology, EXPEC Advanced Research Center) | Erickson, Kevin (Geophysics Technology, EXPEC Advanced Research Center)

OnePetro 

Abstract

Saudi Aramco recently started the company's first CO2-EOR demonstration project in an onshore carbonate reservoir. Time-lapse (4D) seismic has proven to be a valuable reservoir management tool for monitoring the areal expansion of CO2 plumes in many similar projects around the world. However, the complex and dynamic nature of the near surface encountered in the desert environments of the Middle East results in high levels of 4D noise. This noise, coupled with the weak 4D signal expected from injection into a stiff carbonate reservoir, makes mapping the time lapse signal very challenging. The objective of this project was to develop a highly repeatable system capable of detecting small reservoir changes related to CO2 injection to enable the plume expansion to be tracked over time.

Achieving highly repeatable seismic data requires specialized seismic acquisition and dedicated processing. A novel acquisition system using buried receivers was adopted to reduce 4D noise resulting from near-surface variations. To minimize the non-repeatability inherent in using surface sources, a differential GPS guidance system was implemented to ensure high positioning accuracy. Even with these acquisition efforts, a fit-for-purpose 4D processing workflow was necessary to further reduce differences between surveys.

Despite the challenges faced, outstanding data repeatability has been achieved, with mean NRMS values of less than 5% for data acquired during the same season. This level of repeatability is comparable to data acquired in marine 4D surveys and has resulted in the detection of the small 4D signal caused by CO2 injection. Frequent monitor surveys, with one full survey acquired every four weeks, shows the CO2 plume growing over time with increasing injection volume. While the observed CO2 plume largely correlates to available engineering data, discrepancies have been identified when compared with the predicted seismic response based on the reservoir simulation model. This indicates that 4D seismic can be used to constrain the reservoir model, yielding a better history match and improved predictions to enable more informed engineering decisions to be made.

This is the first successful application of seismic monitoring of a carbonate reservoir in an area renowned for poor seismic data quality. To overcome the challenges, a novel hybrid acquisition system using buried sensors and surface sources was developed. Advances in the seismic processing workflow were also required to bring the 4D noise down to a level that enabled detection of the CO2 injection.