The optimization of natural gas liquefaction processes

Abstract

An optimization method is presented for the cost minimization during the design of process engineering plants. The mathematical model of process plants, the general formulation of an optimization problem and the numerical method are illustrated by a natural gas liquefaction process.

Special requirements of design optimization yield discontinuous functions arising in the objective function and the constraints. The process model, a large system of nonlinear equations (flowsheet equations) with discontinuous functions and sparse Jacobian matrix, is included by the equality constraints of the optimization problem. A feasible path SQP (Sequential Quadratic Programming) method has been developed by which the special structure of the problem is taken into account. The new approach leads to a significant cost reduction which is illustrated by some computational results.