The efficient energy supply of plus energy districts (PED) continues to pose a major challenge. PED are characterised by the fact that they have a positive energy balance over the course of a year. However, achieving a high degree of self-sufficiency appears to be even more important in this context. If the timing of energy demand and generation is taken into account PED can have positive effects on the public electricity grid and benefit from increased security of supply in the event of a power failure. Reversible solid oxide cell (rSOC) systems are an interesting technology in this regard. They can be operated in both electrolysis and fuel cell modes. This means that surplus electricity can be converted into green hydrogen and converted back into electricity at a later point in time. In fuel cell mode, heat is released in addition to electricity, which can be used to (partially) cover heating and hot water requirements. Thanks to the good storability of hydrogen, this system makes it possible to store energy seasonally, i.e. to shift it from summer to winter. This can significantly increase the degree of self-sufficiency of PED, where photovoltaic systems, whose energy supply potential fluctuates seasonally, are usually one of the most important sources of energy. However, due to the relatively early stage of development, the economic use of rSOC technology poses a challenge.

The overarching goal of CELL4LIFE was therefore to develop and design economic operating and business models for the use of rSOC technology in neighbourhoods so that PED requirements could be met.

Analyses were carried out at several system levels. On the one hand, bench tests were used to investigate the transient operation of this technology in order to examine the influence of load changes (load gradients and load change rates) in electrolysis (SOEC) and fuel cell operation (SOFC) on performance and subsequently define the optimum operating conditions. To optimise transient reversible operation, a dedicated test bench was developed specifically for these investigations, featuring a control strategy that was also developed as part of the project. In addition, machine learning-supported rSOC optimisation and control strategies were developed to further improve the operation of rSOC systems and minimise the need for costly test bench trials. All these results and findings were incorporated into a techno-economic optimisation model, which was used to dimension the most important components, evaluate various operating models and investigate the optimal integration of rSOC systems into existing PED energy systems.

The results of the test bench experiments have shown that the transient alternating operation between SOEC and SOFC influences cell behaviour, particularly in the initial cycles. In this phase, changes in the low-frequency contributions occur, accompanied by a slight change in ohmic resistance, which is reflected in an initially deviating cell voltage. Over the period studied, degradation is predominantly determined by the ohmic component. Ohmic resistance increases less in alternating operation than in steady-state operation, while a stronger increase than in the first section is observed in the final steam electrolysis section. The polarisation component remains largely constant across all phases and contributes only slightly to the overall curve. Overall, the experiment suggests that alternating operation can cause short-term adjustments in diffusion- and transport-relevant process components, but does not sustainably reduce ohmic-dominated degradation under the conditions investigated.

For ML-based optimisation and control, a neural network was coupled with a genetic algorithm (GA) to evolutionarily optimise SOFC and SOEC operation. The optimisation was carried out according to electrical power and electrical efficiency. These optimisation cases were then used for GA optimisation of the polarisation curve. It was found that the results of GA optimisation of the polarisation curve are associated with large fluctuations, which, if applied directly to the rSOC system, are highly likely to damage the cell and thus the stack. It is therefore recommended not to apply the optimisation results directly, but to include an intermediate step in order to derive control strategies based on the optimisation results.

A control concept based on a cascaded PI lead controller was developed for the new rSOC test bench, in which the inner control loop regulates the oven temperature and the outer control loop regulates the air outlet temperature. This allows the output of the outer controller, which represents the setpoint temperature for the oven, to be easily and intuitively limited before it is used as the setpoint for the inner controller. A PI lead controller can be converted into a (feasible) PID controller, but has the great advantage of being easy and intuitive to parameterise – the two pole positions of the section can be shortened with the zero points of the controller, resulting in a very easy-to-use open-loop transfer function.

The techno-economic assessment has shown that, although rSOC systems are an interesting technology for use in PEQ from a technical and energy perspective, it is not possible to find an economic use case under the current conditions. This is mainly due to the high investment costs and the relatively low electrical efficiency for the production, storage and subsequent reconversion of hydrogen into electricity. However, as rSOC systems are a relatively new technology, it can be assumed that there will be positive developments in the coming years, both in terms of investment costs and achievable efficiency. The sensitivity analyses carried out show that, if developments are positive, rSOC technology could also be an interesting option for use in PED from an economic point of view in the future. The integration of rSOC systems is particularly advantageous if a high degree of self-sufficiency is to be achieved.

As no economic use case for rSOC systems in PED could be found, it is recommended that research initially focus on further developing the rSOC technology itself and, if developments are positive, re-examine the potential for its use in PED.

Project management

4ward Energy Research GmbH

Project or cooperation partners

• Institut für Wärmetechnik, Technische Universität Graz
• Kristl, Seibt & Co. Gesellschaft m.b.H.

4ward Energy Research GmbH
Reininghausstraße 13A
A-8020 Graz
Tel.: +43 (664) 882 518 30
E-mail: markus.rabensteiner@4wardenergy.at
Web: www.4wardenergy.at