Perpustakaan Digital ITB

Over time, coal bed methane (CBM) needs to be developed more widely to meet gas needs especially in Indonesia. One of method that can be used is to enhance gas by using CO2 injection in producing gas because CO2 gas in Indonesia is quite abundant. In this study of development field, a quick screening process is also needed to determine highest gas cumulative potential. A model to predict gas cumulative obtained is by using proxy analysis to help feasibility study of CO2-ECBM project. This study will focus on make a proxy model to predict cumulative CH4 gas that can be produced as its main function for 10 years production time by considering number of influential parameters. From 24 parameters with 1232 total experiments using Latin Hypercube Sampling (LHS) method, 6 parameters were obtained that had major impact on the cumulative value of CH4 gas. Through this model, sensitivity analysis was applied by using CMG-CMOST software and evaluation of the quality model that has been build will be done and then construct a regression to experimental model to make an equation. Proxy model obtained through the whole process will hel to predict CO2-ECBM based on cumulative CH4 producing gas. This study resulted an equation from regression with best quality with quadratic regression method. Through quadratic regression, value of r-squared is 0.9208 from 6 parameter that gives huge impact over total 24 parameters that has been analyze by using help of tornado chart in Sobol analysis. From these 6 parameters, all of them have an impact in the amount of gas content of methane in CBM reservoir. The average absolute error value is quite good with value 0.1442. After a proxy model has been build, then validation of the equation is needed in order to compare the result of proxy calculation and simulation calculation. The validation process uses another model of CBM data. The result gives satisfy number of cumulative CH4 gas production of the proxy model with value 4137121.134 m3 and the simulation result value is 3777975 m3 with an absolute error 10%. All of these results are expected to be able to help predict cumulative CH4 gas production in CBM development especially in Indonesia. This proxy model is expected to be quite accurate and could be evaluated again for further development.