By Peter Witten, AAU Communication and Public Affairs
Illustration: David Graff

As global temperatures rise, cities in tropical and subtropical regions face a vicious cycle: extreme heat drives massive energy consumption for cooling, which in turn increases CO₂ emissions.

The research project SWiM aims to develop intelligent, scalable, and sustainable district cooling systems that could reduce energy consumption in megacities in Asia and Africa by up to one-third.

The project is a collaboration between Aalborg University, Aarhus University (AU), and Nanyang Technological University (NTU) in Singapore. It is funded by the Grundfos Foundation with DKK 60 million.

Professor Kim Guldstrand Larsen leads the work package Operation: Control, Optimization and Interoperability with Dr. Sivanand Somasundaram from NTU as co-lead.  Other collaboraters include Wan Man Pun, Chau Yuen from NTU and Professor Rafael Wisniewski from AAU.

This work package will develop methods and algorithms to ensure real-time management of energy consumption and production using intelligent control strategies, while adapting to occupant behavior, environmental conditions, and regulatory requirements for optimal performance in hot and humid climates. 

In particular methods for optimal control in terms of energy use, CO₂ emissions, comfort and operation will be developed, with possible integration of renewable energy sources and reduction of peak energy demand using demand-side management techniques.

Kim Guldstrand Larsen will leverage on recent breakthrough results on Hierarchical Multi-Agent Safe Reinforcement Learning together with Wan Man Pun and Rafael Wisniewski. This task focuses on developing robust and safe machine learning methods to manage cooling operations in real time across both building-scale hybrid systems and district-level networks.

Professor Torben Bach Pedersen’s task is Sector Coupling of the Electricity and Cooling Systems. He is the task leader, with Associate Professor Chau Yuen from NTU and Professor Amjad Anvari-Moghaddam from AAU as collaborators.

A key challenge is to enable optimization and trading (also known as sector coupling) across the two energy vectors of cooling and electricity by exploiting flexibilities in both. This task operates at the building and district levels, within the timescales of operational planning and real-time control. System-wise, it spans the IoT and local control layer as well as the data, analytics, and global optimization layer.

Torben Bach Pedersen will also investigate cooling flexibility modeling, estimation, aggregation, and optimization, answering the research question "How can the flexibilities in cooling systems be utilized for joint optimization with electricity and cold production, storage, and demand to minimize CO₂ emissions and cost?"

Answering this research question enables the utilization of the immense flexibility potential in cooling systems, such as storing cold in both the cooling network and dedicated storage units. Specifically, this task will extend AAU’s award-winning FlexOffer technology to utilize the flexibilities in cooling assets and systems, building on existing work in electricity and heating.