Datapages, Inc.Print this page

Prediction for Carbonate Reservoir by High-Resolution Electric Image Log Facies Analysis: Example from the Ordovician Strata in Tz Area, Tarim Basin, Northwest China

Zhong, Guangfa 1; Haijun, Yang 2; Chengwen, Xiao 2; Xinzhong, Qi 2; Guanjun, Hui 1; Xiuli, Guo 2
1 School of Ocean and Earth Science, Tongji University, Shanghai, China.
2 Tarim Oilfield Branch Company, China National Petroleum Corporation, Kuerle, China.

Prediction of carbonate reservoirs is difficult because of their strong heterogeneity, which resulted from entangled depositional, diagenetic, tectonic and occasionally karstification processes. Characteristics and distribution of carbonate lithofacies are required for a successful prediction of carbonate reservoirs. Carbonate lithofacies, however, are usually analyzed by traditional geological methods such as core inspection and thin-section observation. These methods depend on the availability of cores, and are incapable or not efficient in cases of analyzing un-cored wells or well intervals. Formation micro-resistivity image logging, as a method of very high resolution (up to 5 mm) and being capable of measuring in any desired well interval, provides efficient compensation for traditional core-based facies analysis.

In this study, the method of carbonate image log facies analysis was described on the basis of analyzing electric image logs from 20 wells in TZ area, Tarim Basin, northwest China. An image log facies is defined as the overall characteristics of a rock unit on borehole electric images that differentiate the unit from others around it. Based on color and texture of the electric images, fifteen classes of carbonate image log facies were classified, which could be further grouped into three categories: massive or uniform, bedded, and patchy facies. Calibration with core and traditional well log data made it possible to a satisfactory degree to reveal the geological significance of each image log facies. Statistics showed that good reservoirs existed mostly in three image log facies, i.e. the dark (low resistive) patchy facies, dark (low resistive) uniform facies and dark (low resistive) thick-bedded facies. Thickness percentage of these three favorable reservoir image log facies was used to predict the occurrence of promising reservoirs in the study area. The predicted results, with accuracy of about 66.7%, were in good agreement with the real drilling dada, indicating the robustness of our method.


AAPG Search and Discovery Article #90090©2009 AAPG Annual Convention and Exhibition, Denver, Colorado, June 7-10, 2009