(736d) Optimal Synthesis of Refrigeration System for Natural Gas Liquefaction

Liquefaction of natural gas (LNG) is an energy-intensive process.  Energy efficiency maximization is the major objective in LNG process design and operation.  A lot of efforts have been done to minimize the energy consumption of the LNG process, most of which use heuristic methods or nondeterministic search methods and thus could not guarantee the global optimal results.  In this paper, a novel methodology for LNG process optimization on energy consumption minimization is presented.  This methodology is based on mathematical programming and deterministic global optimization.  The difficulties of this systematic study include: i) develop process synthesis model for optimal design and operation of LNG process; ii) solving the mixed integer nonlinear programming (MINLP) optimization problem.  iii) develop rigorous process simulation model for optimal results validation.  Also presented are the technologies for obtaining reliable thermodynamic property formulas based on rigorous simulation and the methods for MINLP model simplification.

LINDOGlobal solver in GAMS is employed to solve the optimization problem. The LINDO global optimization procedure employs branch-and cut methods to break an NLP model down into a list of subproblems.  Each subproblem is either solved to find the optimal solution or split into two or more subproblems.  Given appropriate tolerances, after a finite, though possibly large number of steps a solution provably global optimal to tolerances is returned.  The results obtained from GAMS are validated through rigorous simulation.  Based on the methodology, the optimal design and operation conditions of LNG process can be simultaneously achieved.  A case study about a C3MR LNG process is employed to demonstrate the efficacy of the developed methodology.  Comprehensive thermodynamic analysis is also conducted in this paper to give deep understandings of the LNG refrigeration system and optimization procedure.