Methods and tools for assessment of dynamic aspects in the design of more efficient industrial energy systems

Important results from the project

In MoveDyne, control methods for multivariable systems have been developed and improved. A proposed approach describes how these could be combined with existing methods and tools to design energy efficient heat exchanger networks with good controllability. Methods were tested on examples from Preem. This identified development needs and led to proposals for better control of these processes. The results supports derivation of an improved controllability measure. Cooperation in the project led to strengthened competence through exchange of experiences.

Expected long term effects

Theoretical studies resulted in the derivation of an optimal LQ control scheme for multivariable systems with wireless communication over lossy channels. Moreover, the project has illustrated difficulties when applying traditional RGA-based controllability measures to heat exchanger networks. As alternatives to RGA, gramian-based interaction measures were studied and an improved method for scaling interaction matrices was derived. A model generator that randomly generates a large set of linear multivariable systems was developed to enable comparison of control methods.

Approach and implementation

The project is based on research on control methods for multivariable systems. These are put in context by conceptualizing a complete methodology for design of energy efficient heat exchanger networks with good controllability. Several examples based on Preem’s processes have been considered, resulting in recommendations for further research as well as for control of the processes. Cooperation between academia, industry and different disciplines has secured relevance, strengthened the competence of the project members and favoured collaboration in this and future projects.