Abstract
For big variation of formation water salinity in Chang 8 stratum, Triassic, northwestern Ordos Basin, China, low resistivity contrast exists between oil layers and water layers. In order to increase the accuracy of log interpretation, accurate formation water salinity is a vital part. Based on the petrophysical theory, this paper summarizes and improves two methods to estimate formation water salinity. Firstly, reservoir resistivity-porosity cross plot method is introduced for oil-water layers and water layers. To be specific, resistivity and porosity log values of target reservoir are added to the cross plot. Data points, which are closest to the origin of coordinates, are selected as water layer ones. Then, formation water salinity is calculated by Archie formula. Secondly, shale water salinity is approximately regarded as formation water salinity. Because shale water salinity estimation is a nonlinear problem with small sample sets and there is no theoretical equation, Least Squares Support Vector Machine (LSSVM) is used for shale water salinity prediction. 9 parameters are extracted from lithology, resistivity and porosity log curves, among which, 5 are optimized as sensitive parameters by Principal Component Analysis (PCA). The effectiveness and reliability of resistivity-porosity cross plot and improved SVM method are tested by 23 formation water chemical analysis data. The average relative error of the former method is 19.79%, while that of the latter 27.57%. In addition, formation water salinity of another 50 wells are calculated by the two methods. Based on them, a salinity plane distribution map is drawn by Geomap software. In high salinity area, producing wells gather. Thus, one possible origin of formation water salinity variation is proposed. High salinity water moves into reservoirs with oil from source rock, which leads to high water salinity. In ultra-low permeability clastic reservoir with near source accumulation, formation water salinity probably varies significantly because of oil migration and accumulation. Furthermore, layers with oil often have higher formation water salinity, which is the main cause of low resistivity oil layers. Thus, the accurate formation water salinity calculated by the improved methods, will play an important role in the evaluation of low resistivity contrast oil layers and water layers.
