In-Situ Residual Oil Saturation to Gas from Time-Lapse Borehole Gravity

Abstract

Summary Three time-lapse borehole gravity surveys were performed one year apart in the dedicated data gathering and monitoring well of the Rabi oil-rim. For the first time, time-lapse density logs derived from Borehole Gravity Meter (BHGM) runs are used to estimate residual oil saturation in-situ. This is also the first time that the BHGM is used to determine interval density at one meter spacing. This paper reviews the planning of the BHGM surveys, explains why it became necessary to reduce the survey intervals to one meter, highlights the enhancements that made such measurements possible, and presents the first residual oil saturation results. BHGM density is derived from the gravity gradient measured between two survey depths. The accuracy of the BHGM density is therefore governed by the accuracy of both gravity and depth measurements. The uncertainty of the gravity difference between two survey depths was significantly reduced through sensor improvements and new processing techniques. Depth-related errors were practically eliminated using a "Shuttle Sonde" to accurately position the gravity sensor within intervals up to 2.5 m. The time-lapse technique is immune to structural effects which can bias single BHGM surveys. In addition, the BHGM is latched into grooves built into the glass-reinforced Plexiglas (GRP) casing to ensure that the stationary tool is positioned exactly at the same depths from year to year. The reproducibility of BHGM density measurements between different runs is + 0.015 g/cm3 (one-, i.e. one standard deviation confidence level) over one meter intervals. This level of accuracy had only been achieved over three meter intervals prior to the present survey, thereby suggesting a threefold increase in the vertical resolution of the BHGM density measurement. As a result, a first estimate of residual oil saturation to gas was determined from BHGM data after only two years to within 10 su (one-) on a one-meter survey interval. The remaining oil saturation to water estimate derived from time-lapse BHGM surveys is in close agreement with residual oil saturation to water interpreted from time-lapse PNC logs. Introduction The Rabi Field. The 1.8 billion barrel Rabi field is located in Gabon's dense rain forest, 60 km inland from the Atlantic Ocean, and just South of the Equator. The 1100-m deep unconsolidated sandstone reservoir is contained within a complexly faulted, l3x4 km anticline, and is trapped at the base of a system of thick salt domes. The 46-m oil rim is overlain by a medium-sized gas cap and underlain by a large aquifer. The reservoir sands are clean and generally coarse-grained, with permeabilities ranging from 1 to 5 D, and porosities ranging from 24 to 32 pu. Formation water is salt-saturated. Determination of Residual Oil Saturation In-Situ. The motivation to accurately determine residual oil saturation to water and to gas in-situ for input into a five-sector full-field reservoir simulation model was presented earlier in the context of the optimisation of the Rabi Phase II incremental field development plan. Time-lapse Pulse Neutron Capture (PNC) logging proved very effective in the Rabi environment to assess residual oil saturation to water at the sought accuracy of 5 su (all uncertainties quoted in this paper are "one- ", i.e. one standard deviation confidence level). A 10 su accuracy on residual oil saturation to gas was desired. Such high accuracy could not be achieved from PNC measurements, because the capture cross-section contrast between oil and gas is about 5 times lower than the contrast between water and oil. The PNC measurement is, therefore, significantly less sensitive to gas saturation changes than it is to water saturation changes, with a corresponding loss of accuracy in the estimation of residual oil saturation to gas. P. 855