Paper Conference

Proceedings of Building Simulation 2021: 17th Conference of IBPSA


A low-order semi-physical borefield model for optimal control applications

Iago Cupeiro Figueroa 1, Lieve Helsen 1,2
1 KU Leuven, Belgium
2 Energyville, Belgium

Abstract: Accounting for the borefield long-term dynamic behavior in optimal control applications can lead to substantial savings. To that end, a borefield controller model that covers the long-term time scale (i.e., interactions between boreholes) is necessary. However, the direct use of an analytical model may result in a computational burden increase due to the high number of required states. This paper presents a novel loworder resistance-capacity (RC) network model that accounts for the borefield long-term dynamic behavior based on the physical information provided by its characteristic thermal response. It is shown that the mean square error (MSE) of the borefield outlet fluid temperature remains below 0.015 K2 in two example borefields using a total number of states below 15. The validation experiment revealed that attention should be paid to the short-term dynamics since the MSE indicator captures the error of the g-function over the full time horizon and as a consequence large short-term dynamic errors in the fluid temperatures are not detected.
Keywords: Control-oriented modeling, borefield modeling, model validation
Pages: 1928 - 1934