STUDI PERBANDINGAN METODE OPTIMISASI ALOKASI LAJU INJEKSI GAS-LIFT DI LAPANGAN K

• Main Author: Aghni Febrianti, Nurul
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/42807
• Institution: Institut Teknologi Bandung

Gas-lift is one of artificial lift methods commonly used in the upstream petroleum industry. The artificial lift requires continuous evaluation and optimization to evaluate the existing production system and to ensure optimal production. In the "K" field, an onshore oil field located in South Sumatra Province, oil wells are produced one of them using a gas-lift as artificial lift due to low reservoir pressure and high water cut. The data used in this study is taken from the test data of four gas-lift wells. Gas lift optimization aims to allocate a limited amount of injection gas to produce the maximum oil production rate. The optimization starts with evaluating the artificial lift performance that represented by its performance curve (GLPC) as the basis of the optimization method, where GLPC will describe the response of the well to the gas-lift injection rate. The allocation of the gas-lift injection rate is then optimized to obtain the aktual optimum value by using the Nonlinear Programming (NLP) method, Simplex Linear Programming (SLP), and analytical model. The result of each method will be compared and evaluated. The results of the SLP method for gas-lift optimization show a potential 49.5% increase from initial oil production by reallocating gas injection rates. The NLP method shows a yield of 48.5%, the potential for increased oil production. Whereas using analytic model showed higher results, a 60% increase in oil production. This is because synthetic data used for analytic model in the form of quadratic polynomials in which the quadratic polynomial equation does not represent the entire aktual data. Based on Ricky et al's paper. "A" Simple-Effective-Efficient "Analytical Model For Multi-Well Gas Lift Allocation Optimization", the limit for using this analytical model is that R2 from GLPC is more than 95%. Whereas some wells have R2 less than 0.95. For the three methods above are traditional methods that do not consider back pressure effect on wellhead pressure, gas-lift valve opening pressure and the availability of casing operating pressure to inject gas-lift rate at specific amount, so the results of gas-lift injection rate allocation optimization obtained are still felt quite large compared to the initial production of oil. The NLP method is the best method compared to the SLP method and analytical models to optimize gas-lift. This is because the synthetic data used in optimization is quite close to the aktual data for each well. So that the optimum value obtained is more reliable than other methods. Production network modeling software can used to obtain better optimization results because it considers several operational aspects in optimization compared to traditional methods.