| Summary: | Design for process flexibility is an industrially important topic so it is not surprising that it is attracting much research work. Given the size of the problem it is also not surprising that workers have concentrated on heat exchanger networks, which can be considered to be a self-contained sub-problem. Unfortunately, recent research has suffered from a number of major drawbacks. Problem formulation has often been unrealistic. Proposed procedures tend to be 'clinical' rather than practical. Academic research has often been conducted without proper consideration of the industrial environment. Very few research results have been tested on full scale industrial problems. In this thesis a new problem formulation and new solution techniques are presented. They have been designed to fulfil the needs of industry. In problem formulation it is recognised that the specification of flexibility is primarily an economic problem. The amount of flexibility industry will demand is a function of how much it costs. The methodology developed here allows both, the flexibility cost and the existing trade-off between flexibility needs, capital costs and energy costs to be fully evaluated. Flexibility problem formulation is mainly based on Multiple Base Cases. Different plant operating scenarios are set and a design found that is able to satisfy each case. This is a formulation that has found a wide acceptance in industry. To be industrially practical, process design techniques must be intelligible to the non-specialist. The techniques developed here are simple and straightforward and give insight. Two new design techniques have been developed. The first of these is 'Downstream Paths'. These are used to identify and evolve the network structures that permit cost effective flexibility. The second technique is 'Sensitivity Tables'. These can be used to analyse the performance of a given structure. In addition they determine the cost effective modification to elements of the structure which provide the specified flexibility needs. The technique is rapid, simple to apply and easy to repeat. Consequently, many scenarios can be screened without much effort and a cost profile developed in order to evaluate the cost trade-off described above. The approach proposed in this thesis, involving the new problem formulation and solution techniques, has been applied to a number of case studies of industrial scale. These case studies have covered problems as diverse as catalyst deactivation, fouling, and plant debottlenecking. One major spin-off from the work is a new approach to the design of heat exchanger networks subjected to fouling. Large potential savings have been identified.
|
|---|
