Power consumption in refrigeration systems - Modeling for optimization

Refrigeration systems consume a substantial amount of energy. Taking for instance supermarket refrigeration systems as an example they can account for up to 50-80% of the total energy consumption in the supermarket. Due to the thermal capacity made up by the refrigerated goods in the system there is a possibility for optimizing the power consumption by utilizing load shifting strategies. This paper describes the dynamics and the modeling of a vapor compression refrigeration system needed for sufficiently realistic estimation of the power consumption and its minimization. This leads to a non-convex function with possibly multiple extrema. Such a function can not directly be optimized by standard methods and a qualitative analysis of the system's constraints is presented. The description of power consumption contains nonlinear terms which are approximated by linear functions in the control variables and the error by doing so is investigated. Finally a minimization procedure for the presented problem is suggested.