In English

Routine for Automatic Optimisation of Heat Exchanger Networks with the Matrix Method

Claes Andersson
Göteborg : Chalmers tekniska högskola, 2001.
[Examensarbete på avancerad nivå]

In this master thesis is a tool has been developed which can be used to determine the optimum design of a heat exchanger network. The optimisation is performed for a specific energy saving predefined by the designer. By the use of pinch technology for identification of possible energy saving for an existing network a method called The Matrix Method can be used for retrofit design of the network. A helpful tool for such a study is the computer program Matrix.xla. The program allows the designer to manually select locations for heat exchangers to construct a network which satisfies the energy saving. To build up an optimal network, i.e. the least expensive, often several iterations are needed which demands a lot of time. Besides this, in a retrofit study often many different global temperatures are studied and for each one a number of heat recovery levels. The result of this work is an optimiser implemented to the matrix program. The Matrix Method Optimiser automatically searches and presents the optimal solution for a certain energy saving, above and below pinch separately. The routine uses a branch and bound strategy to search for possible solutions among all available heat exchanger combinations. Once a solution is found the cost for it is used as upper bound in the rest of the search. The solution stored in memory when all combinations are checked is the solution with the lowest annual cost. Together with info about the heat exchangers included in the result the optimum cost is displayed on screen. Instead of manually selecting matches it is from now on possible to use an automatic routine in order to find the lowest cost solution for a heat exchanger network at a specific energy saving. This saves a lot of time for the user and also guarantees the optimum result.



Publikationen registrerades 2007-12-28. Den ändrades senast 2013-04-04

CPL ID: 63966

Detta är en tjänst från Chalmers bibliotek