Integrated Interpretation of Wireline and 3D Seismic Data To Delineate Thin Oil-Producing Sands in San Jorge Basin, Argentina Carlos Torres-Verdı ´n, SPE, U. of Texas at Austin; Raghu K. Chunduru, Shell Intl. E&P; and Alberto G. Mezzatesta, SPE, Baker Atlas Summary Hydrocarbon-bearing sand units in San Jorge basin, Argentina, exhibit a wide range of spatial variability and thickness. In view of this, a typical well is planned to vertically intersect as many sand units as possible. There are several problems faced by the petro- physicist in assessing whether a given sand unit should be perfo- rated: (a) discrimination between oil- and water-bearing sands is not trivial because of very low-salinity water, (b) there exist sub- stantial vertical variations of effective porosity within an indi- vidual sand because of shale laminations, and (c) it is often im- possible to assess lateral extent away from the well. We have successfully addressed most of these difficulties with an interpre- tation procedure centered on the 2D inversion of wireline array- induction data. Two-dimensional inversion of array-induction data is necessary for the accurate estimation of shallow and deep re- sistivities, as well as the invasion length in light of significant shoulder-bed effects. This procedure has been complemented with the use of borehole nuclear magnetic resonance (NMR) data to provide estimates of effective porosity within individual sand units. Finally, we have made use of geostatistical inversion of 3D post-stack seismic data to estimate the lateral extent of hydrocar- bon-bearing sands laterally away from wells. We present several application examples that yield results consistent with borehole testing and production data. Introduction Located in the heart of Patagonia, and extending from the Atlantic Ocean to the Andean foothills, San Jorge basin accounts for 32% of the hydrocarbon production in Argentina (see Fig. 1). The ori- gin and subsequent geological evolution of the basin are caused by the rift process responsible for the opening of the Atlantic Ocean in early Jurassic times. Accumulation of terrigeneous sediments continued until well into early Cretaceous times. 1 Clastic deposi- tion in the hydrocarbon-producing zone is characterized by thick shale laminations of lacustrine and flood-plain origin, interspersed with much thinner and laterally sparse sand units that today serve as hydrocarbon reservoirs. The relatively small concentration of sand units in the sedimentary column is explained by their ephem- eral fluvial origin, which could only account for effective clastic accumulations between 0.5 and 15 m (but predominantly thinner than 4 m). Starting in early Cretaceous times, Andean tectonism caused yet another significant perturbation of the sedimentary col- umn in the form of finely laminated deposits of pyroclastic origin (tuffs) associated with intermittent pulses of volcanic activity. These pyroclasts ranged anywhere from thin ash beds to welded tuffs several meters thick. The presence of tuffs significantly al- tered the original petrophysical properties of existing sand units. Subsequent structural deformation also adversely modified the al- ready marginal porosity and permeability of the sands and caused extensive fracture damage to the existing tuff units. In this paper, we focus our attention on oil-producing sand units within the Bajo Barreal formation (mid- and upper-Cretaceous age), buried at depths between 1200 and 2900 m. Fig. 2 is an outcrop of the Bajo Barreal formation showing a representative sand/shale/tuff sequence with bed thicknesses similar to those en- countered in the hydrocarbon-producing zone. A close-up view of an individual sand unit, shown in Fig. 3, exhibits a fluvial cross- bedding structure and is bounded by clastic debris and tuff units above and below. The petrophysical interpretation of well-log data in San Jorge basin is not trivial. Not only are shale and tuff laminations in sand units difficult to detect and quantify, but the low salinity of connate water also makes it hard to differentiate water- from oil-bearing sands with their resistivity signatures alone. To compound this problem, fractured tuff units and associated debris have a tendency to develop resistivity and spontaneous potential (SP) responses similar to those of permeable sands. Sampling of formation fluids also has evidenced substantial variations in the salinity of connate water, making it extremely difficult to assess fluid saturations by traditional means, even with the help of core data. The presence of shale and tuff debris in sands causes significant variations in the sand’s effective porosity and permeability. Several approaches have been put forth to solve this problem. Most notably, Khatch- ikian and Breda 2 have used array-induction data to differentiate oil-bearing from water-bearing sands, as well as to predict fluid production in tuffaceous sandstones. Their procedure makes use of the ratio between the short and long spacings of array-induction measurements to characterize both the mud-filtrate-invasion pro- cess and the formation fluids. Corbelleri et al. 3 and Solanet et al. 4 have reported the use of NMR measurements to perform petro- physical evaluations of tuffaceous sands in San Jorge basin. The proposed method has been used successfully to predict effective saturations and porosities in highly laminated sands; it also has been used to predict production enhancement with hydrofracturing operations. A thorough technical review of these petrophysical projects in San Jorge basin shows that when properly interpreted, array-induction measurements can provide a quantitative indica- tion of types of fluid and mud-filtrate invasion in permeable sands, and that NMR measurements can provide an assessment of effec- tive porosities and fluid saturations. The majority of strategies used to interpret array-induction data are based on the assumption of a linear relationship between a perturbation in the model parameters (e.g., shallow and deep elec- trical resistivities, length of mud-filtrate invasion) and the ensuing perturbation in the measurements (e.g., voltage induced in the coils at one or more sampling frequencies). Weighting functions, or kernels, governing this linear relationship are synthesized with specific linear combinations of the raw induction data. Before this, a correction to the raw induction data usually is applied to mini- mize borehole and electromagnetic skin effects. The electrical re- sistivity curves that result from such a linear estimation process selectively emphasize the shallow-to-deep annular electromagnetic Copyright © 2003 Society of Petroleum Engineers This paper (SPE 87304) was revised for publication from paper SPE 62910, first presented at the 2000 SPE Annual Technical Conference and Exhibition, Dallas, 1–4 October. Original manuscript received for review 20 February 2001. Revised manuscript received 2 July 2003. Paper peer approved 5 September 2003. 369 December 2003 SPE Reservoir Evaluation & Engineering