Fresnel solar cooling plant for buildings: Optimal operation of an absorption chiller through inverse modelling

Abstract

Increasing comfort conditions in buildings imply higher energy demands. However, these needs can be mitigated by solar cooling solutions. These systems, such as absorption chillers, are complex and require stable operation, with strict control to maximise the solar fraction and minimise gas consumption. This is incompatible with the variability of renewable resources, so they are often coupled with auxiliary gas systems. Although gas-free operation is possible if these systems are optimally controlled, they would require special supervision. This paper aims to develop an experimental validation of an inverse model to manage an absorption chiller coupled with a solar cooling plant. To know its real behaviour, long-term experiments have been performed using this plant, which consists of a linear Fresnel solar collector and an auxiliary natural gas boiler. The inverse model is used as a predictive control tool to decide the auxiliary boiler commands of the absorption chiller to optimise its operation: maximum cooling production by minimising gas consumption and maximising solar contribution. It has been identified with data from two weeks and validated with data from one summer month. Results show that the model estimates, on a time base of fewer than 30 min, are acceptable with errors of less than 5%. In addition, the maximum error of the estimated seasonal COP and the renewable fraction are less than 6% per day. Therefore, the results prove the usefulness of the proposal as a predictive control for optimal operation. Furthermore, it could be used as a baseline for preventive maintenance. If the proposed model is used for optimal management of the absorption chiller, the thermal efficiency of the plant increases significantly, doubling the solar contribution. As a result, the gas consumption of the solar cooling plant is halved and the total cost of air conditioning the building decreases by 16%. © 2022 The Authors