Predictive Rule-Based Control Strategy for Optimizing the Operation of Solar District Heating Plants
Abstract
For large-scale solar district heating plants, there is often the choice either to provide solar heat to on-site consumers or to feed it into a district heating grid. Plant operators get a better price if they sell the heat to the on-site consumers instead of to the grid. Current state-of-the-art control strategies typically decide on the basis of temperature thresholds on the mode of operation: if the heat should be stored for selling it later to the consumers or it should be fed into the district heating grid. Such strategies, however, can lead to frequent rapid mode switches throughout the day and sometimes the storage is loaded insufficiently, so that heat has to be bought back from the grid. If, the other way around, the storage tank is loaded to a higher extent than needed, this leads to increased storage losses. To address these problems, this contribution presents a predictive rule-based control strategy that takes information on the predicted future conditions into account. By doing so, it ensures that the storage is only loaded to an extend which can be sold to the on-site consumers, thus reducing storage losses, increasing efficiency and maximizing monetary profit for heat sales.