Consider The Gas Lift And Phase Transitions Of The Well Production Optimization

During production of oil well production system, the pressure drops gradually. Oil and gas will put through reservoir, well bore and surface line before they were sent to separator. Production system has a lot of assembly to make pressure reduce. Since the pressure drop of oil and gas through any assembly is relative to flow rate, we usually choose assembly to control production rate. Therefore, the output of oil well gets the restriction of assembly. It is very important to determine reasonable size and the option of assembly, for instance the oil tube size, separator pressure, choke size, surface line size etc. Due to interact among each assembly, pressure change of any assembly will inevitably lead to the pressure change of entire system. Therefore, research into the optimum parameter combination among each assembly have important actual meaning for increasing production rate of oil well.The traditional production systems analysis of oil well can only handle one node every time, a reasonable calculation can be gotten but is not the optimum one. Multivariate optimization method can reach the optimum combination including every variable.After extensive investigation domestic document, based on the outcome of oil well production system prediction and optimize designing, this paper has established a production forecast model of oil well concerning phase change in well bore and gas lift well.This model which is taking NPV as objective function, selecting tubling diameter, separator pressure, gas injection rate and the etc as decision variable can define optimum production parameter combination to make the value of objective function maximum.We developed a software for optimization oil well production system by using multivariate optimization method in this paper, which is composed of polytope algorithm, genetic algorithm and etc. This software is used for analog calculations of gas lift wells in an oil field. The result indicates that the model and algorithm in this paper is correct and reliable.