Residual Oil Saturation (Sorw) is a critical reservoir model parameter for evaluating reserves in the Greater Burgan Field. Past Sorw studies in Greater Burgan Field either looked only at core test data, or only looked at cased-hole log data. None of the past studies considered areal position, different rocktypes, or changes in remaining oil saturation with varying amounts of water sweep. This study includes analysis of Sorw from open-hole water saturation, Time-Lapse PNC data and Special Core Analysis water flood experiments. The majority of the log data in Greater Burgan Field water - swept zones are concentrated in the 3rd sand middle, 3rd sand lower and 4th sand formations. The comparisons of the results from all three methods used in the study to measure remaining oil saturation (ROS) are limited to these reservoirs. Results from these methods were remarkably consistent. All reservoir sand with extensive PNC log data showed that zones encroached by water for 22+ years tend to be at or near residual oil conditions. Measurements in the zones with water encroachment for less than 22 years have about a 50% chance of being incompletely swept. Analysis of the 22+ year data allowed reasonable ranges of Sow were estimated from this data. Investigations of ROS spatial variations in the Magwa, Ahmadi and Burgan sub-fields were made. 3rd sand middle was the only reservoir with both adequate PNC and open-hole coverage in ROS from these three areas in Greater Burgan Field. ROS by rocktype was reviewed in three categories of reservoir rock (excellent, medium and poor quality reservoir) as currently defined by log analysis in Greater Burgan Field. The vast majority of log data occurs in rocktype 1, the highest quality reservoir rock. Only 3rd sand lower formation contained sufficient data in all three reservoir quality rocktype to make valid comparison. Both core flood Tests and PNC Time-Lapse methods also showed no difference in ROS based rocktype. Introduction Background Definitions Residual Oil Saturation (Sorw) is defined as the lowest oil saturation that a reservoir can achieve (technically, a fixed value for a given reservoir and recovery mechanism). With logging tools, we can measure Remaining Oil Saturation (ROS), the oil saturation calculated from a reservoir after it is swept by water due to production. Eventually, these two oil saturations become the same. This study will show evidence to suggest that in Greater Burgan Field, the remaining oil saturation may be changing (lowering) through time, and that true residual oil conditions are frequently not met until a reservoir has been swept for many years. Commonly, the term residual oil saturation is loosely defined as both the true residual oil saturation and the oil saturation that can be measured today with a logging tool. For reserve estimations, the value of the true residual condition is used. When the topic is oil saturation as measured by logging tools, it is also commonly referred to as residual oil, even though the term remaining oil would be more precise. This report will attempt to keep the definitions of Sorw and ROS distinct. ROS Is Important To Reserve Estimation Sorw is a critcal parameter for the accurate estimation of reseves in Greater Burgan Field. Determination of residual oil saturation can provide the basis to refine predictions of a reservoir's recovery factor, and thus possibly increase a field's reserve estimates. There are many ways to measure ROS (and by analogy Sorw) in a dynamic reservoir. Unfortunately, frequently these different methods will not yield consistent answers.

Pulsed neutron measurements are commonly used to locate gas behind casing and quantify steam saturation, but do not always yield desired results. Several parameters are utilized to identify gas and one parameter, the thermal neutron capture cross section, Sigma, is used to compute steam saturation. In the paper we report a mixed experience in identifying gas with these techniques, across fields, tools and vendors. Some parameters have worked well in some cases but have performed poorly in others. The uncertainty in steam saturation, computed using Sigma, is greater than those previously reported elsewhere. Modeling offers insight into the mixed results. It appears that in some cases the PNC-derived Sigma may yield erroneous steam saturation for a variety of reasons, including uncertainties in the input parameters and possibly an inherent nonlinear transport effect that increases as steam saturation increases. An alternative approach based on PNC pseudo-porosity is explored. Calibration of cased-hole tools in gas reservoirs, generic and local, open-hole baseline data and core analysis of complex rocks are essential. Currently, these are either nonexistent or infrequent. Introduction Pulsed neutron capture (PNC) technique, initially used to compute water saturation in high- salinity reservoirs (for example, Dewan, et al., 1973) and in log-inject-log experiments to determine residual oil saturation, was extended to locate gas (Blackburn and Brimage, 1978). It is increasingly being utilized to locate gas in complex conditions and quantify steam saturation in steam floods. In addition, inelastic data, normally used to compute oil saturation, are being used to complement PNC techniques to detect gas. Recent applications in complex conditions include:identification of gas caps to optimize perforation decisions and reduce production of associated gas in West Africa (Badruzzaman, et al., 1997),monitoring steam-chest growth and estimation of steam saturation in steam floods in California and Indonesia (Badruzzaman, et al., 1998; Harness et al., 1998; Zalan, et al., 2003.),locating shallow gas hazards in Gulf of Thailand (Pathanakitchakarnjaroen, et al., 2005) andlocation of pay and identification of swept gas zones in Gulf of Mexico. The technique is being considered to monitor CO2 sequestration in Australia. In addition to PNC techniques, inelastic counts have been utilized as an independent validation of gas behind pipe. Success with pulsed neutron (PN) techniques, involving either the capture or inelastic interactions, has been mixed. Location of gas cap to reduce production of associated gas has generally been successful in Nigeria. Monitoring steam growth with PNC techniques has been successful in steam floods in California and Indonesia. Quantification of steam saturation has been very accurate in California while less so in Indonesia. Locating shallow gas hazards in the Gulf of Thailand has been ambiguous. Pay identification in Gulf of Mexico has been usually clear, but not always.

No abstract

No abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractA logging-while-drilling (LWD) slim compensated longspaced sonic tool that can be run in boreholes as small as 5 3 /4 in. has been found to have useful drilling applications in the Gulf of Mexico. The 4-3 /4-in.-OD tool features a twotransmitter, four-receiver acoustic array. Alternate firings of the two transmitters yield two sets of four waveforms that are processed downhole with a semblance correlation technique to compute compressional and shear slowness ranging from 40 to 180 microseconds/ft. The waveforms are stored in memory for advanced processing at surface. Applications of the sonic information include real-time processing of synthetic seismograms, prediction of pore pressure, computation of acoustic porosity from compressional and shear slowness, and determination of rock mechanical properties for borehole stability analysis. The first of the two studies underscores the accuracy of the LWD slim sonic slowness measurements in predicting pore pressure and detecting the top of supernormal pressure. An LWD slim resistivity tool was run simultaneously with the LWD slim sonic tool, and the resulting resistivity data verified that the slowness measurements indeed served as valid pore-pressure predictors. Operationally, the LWD slim sonic device can be positioned close to the bit, and the measured slowness values can be pulsed uphole in real time during drilling. To enhance drilling efficiency and safety, pore-pressure software at the rig calculates formation pressure and an equivalent mud weight. The second study presents an unusual case in which LWD slim sonic data were used to confirm that a 190-ft section of casing, believed to be cemented in place, slid 90 ft down the wellbore.