water saturation estimation by using pattern search_support vector regression in sarvak formation
water saturation estimation by using pattern search_support vector regression in sarvak formation
Mohammad Reza Cheraghchi1 Hasan Amiri Bakhtiyar2
1) Department Of Petroleum Engineering, Omidiyeh Branch, Islamic Azad University, Omidiyeh, Iran
2) National Iranian South Oil Company , Ahvaz , Iran
IOI Code
:
XYZG-ZABAF
Publication :
2nd International Conference on Science and Engineering(2icesconf.com)
Abstract :
Water saturation is one of the most key parameters at Petroleum Engineering in order to estimate the reservoir’s hydrocarbon volume and also reduction in oilfield economic investment risk. The primary attempts due to estimate formation water saturation were accomplished by Archie with well logging data in clean sandstone formation which the result was developed as a theoretical equation for carbonate reservoirs. In carbonate reservoir, because of the excessive heterogeneous that exists, there are numerous errors occur result in assuming this parameters are constatnt. On the other hand, that is impossible to calculate for the whole length of the well due to time-consuming and cost and also losing a part of the core. In this study, by using support vector regression model, water saturation with ordinary well logging in a case study of Sarvak formation was estimated and compared with the results of core analysis. The performance of this model was optimized by menas of pattern search method. Also, a cross-validation approache was implemented to reduce the proposed model’s error. As comparison of model studied shows that support vector regression model hase much better performance than Archie equation.
Keywords : water saturation well logging archie model support vector regression pattern search sarvak formation