PhD Defense: Carnot batteries for heat and power coupling: Energy, Exergy, Economic and Environmental (4E) analysis by Antoine LATERRE

With the transition to intermittent renewable energies, the need for energy storage is expected to grow significantly. Moreover, the technologies required for this transition will lead to a substantial rise in demand for strategic materials, making their rational and optimal use essential. It is therefore crucial to develop various storage technologies, each tailored to meet specific needs. In this context, Carnot batteries, which store energy in the form of heat and convert it to electricity using heat pumps and heat engines, emerge as true Swiss Army knives for multi-energy systems, particularly benefiting from the heat-to-power coupling. However, for Carnot batteries to effectively contribute to the energy transition, optimal thermodynamic and economic configurations must still be identified, both in terms of use case and technology. Specifically, medium-temperature systems (< 150°C) have been scarcely investigated within literature. To broadly explore this challenge, this thesis proposes a three-pronged approach:

• Thermodynamic optimisation, aimed at identifying the configurations that deliver maximum storage efficiency while aligning with diverse technological preferences;
• Techno-economic optimisation, based on two case studies, to characterise and understand how to improve their financial performance and their contribution to energy systems;
• Environmental optimisation, to identify the distinctive environmental benefits of Carnot batteries compared to more popular electro-chemical batteries.

Importantly, the work proposes several methodological innovations. Regarding thermodynamic design, the 'Modeling to Generate Alternatives' approach was applied for the first time to the optimisation of thermodynamic cycles, in order to investigate a wide range of Carnot battery configurations and identify the necessary trade-offs. Regarding techno-economic optimisation, Linear Programming was used, with the goal of optimising both the design and operation of the system. Lastly, this work reports one of the first Life Cycle Assessment of a Carnot battery, to challenge the widely held belief of reduced impact compared to electro-chemical batteries.

The results show that identifying optimal designs is not straightforward, as power-to-power efficiency (affecting operational costs) and electrical energy density (affecting capital expenditure) are generally conflicting objectives. Additionally, depending on design parameters such as storage temperature, cycle regime, or fluid selection, not all technological preferences can be simultaneously satisfied. Trade-offs in performance must therefore be discussed to identify the design best suited to the intended integration. The economic analysis reveals that, under conservative capital cost assumptions, Carnot batteries are financially viable—though not competitive with other storage technologies. Two cases are specifically examined: (i) Electricity storage to improve energy self-sufficiency in a data centre fed by a photovoltaic system, and where waste heat is recovered by the Carnot battery; (ii) Integrated heat and power storage in residential applications. For both applications, opportunities for improving the Carnot battery design are thoroughly discussed, pointing towards enhanced performance potential. Finally, the environmental analysis shows that Carnot batteries do not provide a substantial benefit in terms of greenhouse gas emissions compared to lithium-ion batteries. However, they offer a distinct advantage in terms of mineral resources consumption.

In conclusion, although Carnot batteries suffer from lower techno-economic performance compared to technologies such as electro-chemical batteries, they remain a flexibility option worth considering for heat and power coupling, and for reducing reliance on strategic materials.

Jury members :

• Prof. Francesco Contino (UCLouvain, Belgium), supervisor
• Prof. Vincent Lemort (ULiège, Belgium), supervisor
• Prof. Nicolas Moës (UCLouvain, Belgium), chairperson
• Prof. Sylvain Quoilin (ULiège, Belgium)
• Prof. Ward De Paepe (UMons, Belgium)
• Dr. Diederik Coppitters (UCLouvain, Belgium)
• Prof. Burak Atakan (Universität Duisburg-Essen, Germany)

Teams Link :