Optimal Synthesis of Liquefaction Cycle for LNG FPSO Considering Operating Conditions

Abstract

The liquefaction process system is regarded as primary among all topsides systems in liquefied natural gas floating, production, storage, and offloading (LNG FPSO). The liquefaction process system condenses the separated and pretreated natural gas into LNG whose volume takes up about 1/600th the volume of the natural gas. Moreover, the liquefaction process system typically accounts for 70% of the capital cost of the topside process system and 30% to 40% of the overall cost of LNG FPSO. The cycles of liquefaction process have seven main equipment parts: compressor, condenser, expansion valve, evaporator, phase separator, common header, and tee. Many types of liquefaction cycles are determined according to their respective synthesis. In this paper, two methods for the optimal design of the liquefaction cycle were studied. First, for the dual mixed refrigerant (DMR) cycle, which was examined for possible application to LNG FPSO, the optimal operating conditions such as pressure, temperature, volume, flow rate, and composition of the refrigerant at the inlet and outlet of each equipment piece, were determined to minimize the required amount of power for the compressors. To obtain the optimal operating conditions, a mathematical model of the DMR cycle was formulated. In the mathematical model of the DMR cycle, the operating conditions were defined as design variables. For the equipment-related constraints, the following essentials were used: mass, energy and entropy balance, compressor efficiency, isobaric condition, isothermal condition, separation condition for the phase separator, mixture process at a constant composition, conservation condition of the output temperature, and minimum temperature approach at heat exchanger. This mathematical model was an indeterminate system in which the number of design variables is larger than the total number of equality constraints. To obtain the optimal operating conditions, the minimization of the required power for the compressors at the cycle was defined as an objective function, while the mathematical model can be regarded as the optimization problem. The optimization problem was a constrained optimization problem and it had 107 design variables, 91 equality constraints, and 11 inequality constraints. To consider the inequality constraints, the original objective function was modified by using the augmented Lagrange multiplier method, a kind of penalty function method. To consider the equality constraints, the system of nonlinear equations for the 91 design variables was solved by using the bisection method and Newton-Raphson method, and the values of the design variables were calculated. The values of the design variables were used to calculate the original objective function. By using these methods, the constrained optimization problem was transformed into an unconstrained optimization problem with 16 free variables and the modified objective function. Finally, the calculation results showed that the required power within the obtained conditions decreased by 0.98% compared with the corresponding value of the past relevant study. Second, this study proposed the generic model of the liquefaction cycle to represent its various types including the DMR cycle. The optimal operating conditions for 27 feasible liquefaction cycle synthesized from the generic model were determined. Based on the results of the optimal operating conditions, the optimized liquefaction cycle was selected and proposed. In addition, the optimal operating conditions of the optimized liquefaction cycle from the synthesis were compared with those of the DMR cycle. The calculation results showed that the required power for the proposed optimal liquefaction cycle in this study decreased by 1.2% compared with that of the DMR cycle.