Control strategy optimization of a Stirling based residential hybrid system through multi-objective optimization
View/ Open
Date
2020-02-20Author
Bengoetxea Larrea, Aritz
Fernández Andrés, Marta
Pérez Iribarren, Estibaliz
González Pino, Iker
Las Heras Casas, Jesús
Ercoreca González, Aitor
Metadata
Show full item record
Energy Conversion and Management 208 : (2020) // Article ID 112549
Abstract
Hybrid systems for space heating and Domestic Hot Water (DHW) production are an attractive option for buildings to decrease their CO2 emissions. In this research, the operation variables of an installation, composed of a Stirling engine, a condensing boiler and a thermal storage tank, were optimized to supply heating and DHW demands of a virtual detached house. For this purpose, in an experimental installation operated with common control set points, the hybrid system was tested during the week of highest demand. Then the installation was modelled in TRNSYS, calibrated and validated against the experimental data and, finally, different multi-objective optimizations were carried out on the model to optimize the operation set points. The results obtained show that solely by optimizing the control variables of the calibrated model of the actual installation gives a reduction of 7% in cost and an improvement of 3.9% in exergy efficiency. Thus, it can be concluded that simply optimizing control variables in this type of hybrid systems can lead to low cost reductions. By reducing the thermal losses of the calibrated model from the experimental 17% of energy consumed to 5% and then optimizing it, a reduction of 17% in cost and an improvement of 23% in exergy efficiency were obtained. Different model insulation levels were tested, and two interesting conclusions were found from this analysis: until the level of transmission losses is below 5% of the energy consumed, the optimized operation conditions lead to a negligible use of the thermal storage, after which it increases drastically. On the other hand, it can be concluded that the cost optimization that optimizes the adiabatic model of the analysed hybrid system is insensitive to the selected control variables