Files

Abstract

A comparative study has demonstrated that the formerly developed Pinch Analysis (PA) based targeting and design methods outperform (in both heat recovery and cost-effectiveness) the simplified combinatorial design methods proposed by other authors (namely the Omnium Verfahren and the Permutation Method). The PA based methods provide valuable insight in the problem, but aren’t suitable for automated design and optimization. PinchBATCH results therefore in the following advances: 1) a strategy (including guidelines and quantitative criterias) is proposed to restrict the analysis of batch processes to meaningful time slices. Likely variations of the schedule are taken into account; 2) practical constrains limiting the repipe and resequence of heat exchangers are identified and formalized under a matrix format. The formalization of all these constraints is needed for automated design methods to actually focus on feasible solutions; 3) a strategy based on genetic algorithms (GAs) is proposed for the synthesis of direct batch heat exchangers networks (i.e. without heat storage). The strategy features a decomposition into an upper level, which optimizes the overall network, and a lower level which search for the best use of the heat exchangers across time slices; 4) for indirect heat recovery using heat storage, a heuristic, PA inspired approach for targeting and simplified design is proposed and demonstrated. It uses TAM composites and includes a method to identify the minimum number of heat storage units (HSUs) as function of the heat recovery. Most generally, local minimums are to be found just before a storage pinch appears, and the search strategy systematically removes one bottleneck at a time, taking advantage of the available degrees of freedom. A simple graphical tool to assess opportunities to decrease the required capacity of HSUs by rescheduling of streams is proposed. A model suitable for a GA based optimization is also proposed, featuring a decomposition into two levels; 5) detailed specifications of the organization and of the features to be implemented in a software tool to address the heat integration of batch processes have been produced. The very next steps will be concerned with the implementation and the validation of the GA based synthesis approaches.

Details