Fuzzy Ranking of Gas Exploitation Opportunities in Mature Oil Fields in Eastern Venezuela

Abstract

TX 75083-3836, U.S.A., fax 01-972-952-9435. AbstractIncreased demand for natural gas in Venezuela has led to an intensified search for additional supply from mature oil fields in Eastern Venezuela. The search has involved the ranking of oil reservoirs that could be converted to gas producing reservoirs. Unreliable reservoir and historical production data introduced high uncertainties to the ranking based on gas reserves only. A fuzzy model was therefore constructed and used for ranking 45 reservoirs. Eight input variables were used to evaluate each of the reservoirs. The variables included remaining petroleum reserves, remaining gas reserves, gas cap volume, oil production per psi, gas production per psi, flow capacity, storage capacity and the distance to existing gas transmission lines. Membership functions were developed for each variable and 20 fuzzy rules were used to capture the nonlinear relationship amongst the input and output variables. The output variable was represented by 5 membership functions and presented in a 0 -1 range after defuzzification. The reservoirs were then ranked on the basis of their output range values.The results from the fuzzy model were compared to those obtained from a conventional methodology using a Decision Tree -Monte Carlo model. In this conventional model, the 8 input variables were normalized and represented as events in a decision tree. The variables were then weighted by using Monte Carlo simulation to generate the mean probability of each event occurring. A rolling netback calculation produced an output value for each reservoir. The output values were used to rank the 45 reservoirs. The paper concludes that although the two models handled the uncertainties of the impact of the 8 input variables on the output variable, the fuzzy model best capture the vagueness in the non-linear relationships between these variables.