Neural Network with Genetic Algorithm Prediction Model of Energy Consumption for Billing Integrity in Gas Pipeline

Along the development of oil and gas industry, missing data is one of the contributors that restrains in analyzing and processing data task in database. By monitoring and maintaining using metering system, the reliability and billing integrity can be ensured and trustworthy can be developed between distributors and customers. In this context, PETRONAS Gas Berhad (PGB) as a gas distributor and an existing system in Nur Metering Station, Kulim is held responsible to evaluate the energy consumption from the sales gas produced. The system is standalone that consists of measuring equipment including pressure transmitter and temperature transmitter, turbine meter, gas chromatography and flow computer but does not have any reference system to verify its integrity. Customers are being charge according to the amount of energy consumption calculated and any error in calculation will cause loss of profit to the company and affect PETRONAS’s business credibility. In the future, it is such a vital to have an ideal analysis in order to maintain the sustainability. In this paper, several techniques will be discuss and selected including neural network prediction model, least square vector regression and combination of either two methods mentioned before with genetic algorithm as preferable technique to indicate the missing data. The model that has been selected based on its evaluation will predict the missing data and compare it with the results of the existing metering system to ensure the reliability and accuracy of the system. The billing integrity between oil and gas company especially PETRONAS and the customers could be maintained and in the future if the project is expanded, it will have the potential of saving of millions of dollars to Malaysian oil and gas companies.