Font Size: a A A

Global optimization of chemical process networks for resource recovery and power generation

Posted on:2003-08-25Degree:Ph.DType:Thesis
University:University of California, Los AngelesCandidate:Martin, Lealon LeCorteFull Text:PDF
GTID:2468390011984539Subject:Engineering
Abstract/Summary:
Prior works on the HEN/MEN minimum total annualized cost (TAC) problem have failed to establish useful optimality properties for these networks. In our work, we develop rigorous proofs, using a variational induced minimization (VIM) technique, which establish conditions under which stream bypassing and self-recycling are not necessary at the HEN/MEN global TAC optimum. The computational limitations of existing optimization-based synthesis methodologies, for general process networks, have prevented the global optimization of such networks to (late. The Infinite DimEnsionAl State-space (IDEAS) approach to process network synthesis overcomes these limitations, by formulating the general process network synthesis problem as an infinite convex (linear) program whose local solutions are guaranteed to be globally optimal. Because of its wide-ranging applicability, process design via the IDEAS approach is expected to have a revolutionary impact on the area of chemical process design. The strength of IDEAS as a synthesis and design tool is demonstrated in power generation case studies. Here we propose IDEAS power cycle designs, consisting of distribution (splitting and mixing) and unit operation (turbine, expansion valve, pump, and heat exchanger) networks, optimally configured to achieve a minimum cost/area objective. It is shown that IDEAS is able to deliver a supercritical ammonia design with a solution that approaches, within four percent, a “perfect” cycle that employs an ideal (imaginary) fluid.
Keywords/Search Tags:Process, Networks, IDEAS, Global, Power
Related items