Abstract

The increasing demand for cooling in commercial buildings, combined with the growing
deployment of local photovoltaic (PV) systems, has highlighted the need for strategies that
mitigate short-term peaks in net electrical demand. While cooling demand and solar generation correlate well on a seasonal basis, rapid fluctuations in solar irradiance generate temporary peaks in the net load that contribute to demand charges and elevated grid stress. This thesis investigates the potential for reducing such peaks through load shifting of the cooling system, utilising the building’s thermal inertia as a latent flexibility resource.

The study is conducted as a case study of a commercial property in Uppsala, Sweden, equipped with two electrically driven chillers and a PV installation. Operational data with one-
minute resolution from the cooling season of 2024 is used to identify the ten most heavily loaded days. The load shifting problem is formulated as a convex optimisation problem
minimising the maximum net power over the day, subject to a total energy conservation
constraint, and is solved for five time windows between one and fifteen minutes. To verify
physical feasibility, a grey-box thermal model is calibrated against measured data and used to simulate the indoor temperature response under each optimised scenario.

The results show that load shifting consistently reduces the net power peak across all analysed days, with the average reduction increasing from 28 percent at a one-minute window to 56 percent at fifteen minutes. The simulated indoor temperature deviation remains below 0.34°C across all scenarios, well within applicable comfort limits. The economic analysis reveals, however, that the technical potential is only partially realised under the prevailing tariff structure, since the power company’s demand charge is based on hourly averages. Measurable economic savings are obtained only at longer time windows, amounting to approximately one thousand SEK over the summer period. The study thereby identifies a structural discrepancy between the technical and economic load shifting potential and contributes to the discussion of how future tariff structures may better reflect the flexibility inherent in buildings’ thermal inertia.