Supermarket Cooling Flexibility Estimation

• Published in: 2024 International Conference on Smart Energy Systems and Technologies (SEST) pp. 1 - 6
• Main Authors: Kotu, Lasya Priya, Rorvik, Ella-Lovise H., Rajasekharan, Jayaprakash, Lindberg, Karen B.
• Format: Conference Proceeding
• Language: English
• Published: IEEE 10-09-2024
• Subjects: Cooling; Data analytics; Delays; Demand response; flexibility; Machine learning; Maximum likelihood estimation; Power markets; Refrigeration; Renewable energy sources; Safety; smart buildings; supermarket cooling; Temperature measurement
• ISSN: 2836-4678
• DOI: 10.1109/SEST61601.2024.10694477

With the increased penetration of renewable energy in the Norwegian national grid, finding new sources of flexibility is of great importance. Supermarkets can provide flexibility as the food items inside the refrigeration and freezer cabinets can store energy due to their inherent thermal inertia. The overall goal of the work is to predict the flexibility of the supermarket to participate in the power markets. This paper presents real-world control experiments to estimate the flexibility of the supermarket's cooling machine. The flexibility offered is calculated by comparing the cooling machine's altered power consumption to the baseline power consumption. The baseline is estimated using different machine learning (ML) techniques and the flexibility calculated by the ML methods is compared to the flexibility calculated by a simple naive persistence model. From the experiments, the supermarket cooling machine offers maximum flexibility of 6.51-8.27 kW, depending on the machine learning technique used for estimation, accounting for 60-75% of the average power consumption during a demand response event. Future experiments will focus on mitigating the delays encountered during the initial experiments to improve the duration of the flexibility available.