Control-oriented model of a solar community with seasonal thermal energy storage: development, calibration and validation

Abstract

The development of a control-oriented model of a solar community with seasonal storage (the Drake Landing Solar Community) is investigated. The proposed approach, intended to facilitate the development and testing of control strategies and targeting an actual predictive control implementation, is based on grey-box models, and enables the prediction of the system state (temperatures at key locations). This paper discusses the concept of state update procedure (whereby the system state is periodically corrected with measurements), which plays a fundamental role for control purposes. Firstly, the DLSC is presented and both operation and monitoring system are described. Secondly, a simplified model is developed for each sub-system: district and solar loops, short-term (water tanks) and seasonal (borehole) thermal energy storage, and existing operation rules are encoded. Finally, the model is calibrated and validated by using measurements at 10-min intervals over two years of operation (2015–2016, 2016–2017) and accurately predicts the system performance.