Perpustakaan Digital ITB

Asphaltene deposition can cause serious problems in the oil industry. It can cause problems for reservoir, wellbore, and production facilities. Asphaltene deposition can cause permeability reduction in the reservoir, reduces oil productivity in afflicted wells, and in worst case, deposition in a massive amount inside tubing can stops the wells from flowing. These problems will affect the economic aspect of the field. To mitigate the asphaltene deposition problems, researchers have developed several models and correlations for predicting asphaltene deposition. This paper aimed to make an expert system to predict the degree of asphaltene deposition problem. This expert system was build using MATLAB. Four models available for predicting asphaltene deposition was combined with fuzzy logic to generate an expert system that can evaluate the degree of asphaltene deposition problem. The models used in this program is the refractive index model, solubility model, SARA (Saturates, Aromatics, Resin, Asphaltene) Analysis, and Colloidal Instability Index correlation. The output value of these models was then used as input in the fuzzy interference system. The input is fuzzified using a membership function. Membership function correlates the input with the degree of membership. Each input has its own membership function that determines its degree of membership depending on the asphaltene deposition model. The degree of membership is then evaluated using IF-THEN rules evaluator. In this study, nine IF-THEN evaluators were used. These evaluators will result in fuzzy output values for each rule. The final output from this program is the degree of asphaltene deposition problem, which value can be low, medium, or high asphaltene deposition problem. To validate this expert system, the set of data that has a very high and very low asphaltene deposition problem was generated and used in this study. This kind of data was selected because their character can be easily detected as high and low asphaltene deposition problem. From the validation, the expert system shows good results. For low data, the degree of asphaltene deposition problem is 28.3% and for high data the degree of asphaltene deposition is 70%. The capability of this expert system to measure the degree of asphaltene deposition can be used as an early judgment of the enhanced oil recovery decision making whether to do or not to do enhanced oil recovery.