Separator Pressure Optimisation and Cost Evaluation of a Multistage Production Unit Using Genetic Algorithm

Abstract

Once reservoir fluid is brought to the surface it is separated to discrete streams of oil and gas in production unit. This should be conducted with consideration of maximizing oil production and to decreasing total costs as well. The regulation of separators pressure is a key point to achieve the highest oil recovery in this process. In addition, estimating the optimum number of separation stages, would augment the design of production plant economically. Adjusting separator pressure is usually conducted by employing PVT data which are provided through flash or equilibrium separation tests. In case of experimental tests unavailability, an alternative method could be empirical correlations. However, such tentative methods contain inaccuracy and are also time consuming which may limit their utility. In this study the issue was addressed by applying genetic algorithm (GA) as well as flash calculation through a case study on one of Iranian oil field. It has been tried to optimize set point pressures of each stage and enhance liquid share of the fluid with respect to minimized gas oil ratio (GOR). Prior to this, four scenarios were considered for determining the number of separation stages. To determine the prime scenario, a cost evaluation was implemented which leaded to the selection of four-separator scenario as the most economical case.

Applied Process Simulation-Driven Oil and Gas Separation Plant Optimization Using Surrogate Modeling and Evolutionary Algorithms

“…In refs. [2,7,9,45] an improvement in terms of increased liquid production is explicitly stated compared to a non-optimized liquid production. The improvement observed in the present study is estimated using the difference between the base case and optimized profit cf.…”

“…By optimizing the pressure in the first three separators, it was found that the oil production could be increased by approximately 2% and 8%, for crude and gas condensate, respectively. Motie et al [9] made a comprehensive study investigating the optimum separator pressure in a multistage separation train, studying the effect of the number of stages, both in terms of operating conditions, but also in terms of an NPV analysis in order to investigate to which extent the added cost of additional equipment for additional separation stages can be justified. The optimization of separator pressures was carried out by means of a genetic algorithm.…”

“…Common for [2][3][4][5][6][7][8][9] is the lack of a compression system providing condensate recycle streams i.e., these studies assume a simple straight-through process with the number of controllable variables normally not exceeding 2 to 5.…”

Applied Process Simulation-Driven Oil and Gas Separation Plant Optimization using Surrogate Modeling and Evolutionary Algorithms

“…2% and 8%, for crude and gas condensate, respectively. Motie et al [9] made a comprehensive study investigating the optimum separator pressure in a multistage separation train, studying the effect of the number of stages, both in terms of operating conditions, but also in terms of an NPV analysis in order to investigate to which extent the added cost of additional equipment for additional separation stages can be justified. The optimization of separator pressures was carried out by means of a genetic algorithm.…”

“…Common for refs. [2][3][4][5][6][7][8][9] is the lack of a compression system providing condensate recycle streams i.e. these studies assume a simple straight through process with the number of controllable variables normally not exceeding 2 to 5.…”