© HZB

The development of a real prototype is time-consuming and expensive. Mathematical models make it possible to better understand the physical and chemical processes in a fuel cell or an electrolyser. A simulation helps to create new approaches and designs in Lobar.

The Centre of Excellence Photovoltaics Berlin, PVcomB for short, at the Helmholtz Centre is investigating integrated photovoltaic electrolysers for the direct conversion of solar energy into hydrogen. With the help of software, the scientists were able to study the operation of the electrolyser and also the heat transfer in the device. Specifically, from the PV to the heat exchanger, where the electrolyte is heated before it enters the electrolyser. “With the software, details of the operation can be understood that cannot be assessed with experimental methods,” explains Erno Kemppainen. The scientist heads the PV to Fuels Technology group at PVcomB. The aim is to explain observations and also to develop predictions. “Particularly detailed models are also easier to build and modify than when similar models have to be programmed from scratch. The finished elements can be put together almost like Lego bricks,” Kemppainen describes.

From a modelling perspective, understanding the effects of electrolyte flow, charge transport, electrochemical reactions and heat transport on each other is essential to be able to understand electrolyser operation. These four main factors must also be taken into account with their feedbacks. But that is ambitious: In general, the most important factors are the catalyst activity and stability as well as the corrosion of each cell component, but with direct coupling with PV, the variable operation of the electrolyser is also a potential problem, Kemppainen explains. “In terms of catalyst activity and the kinetics of the electrolysis half-reactions, the oxygen gas development reaction at the anode of the electrolyser is significantly slower and more difficult than the H2 development reaction at the cathode, making it one of the main bottlenecks of electrolyser operation,” clarifies the scientist.

Analyse phenomena in a coupled manner

For most situations, sufficiently accurate models are available, both for the isolated phenomena and coupled with other phenomena. Precise modulation, however, is very complex: Since electrolysers convert liquid water into H2 and oxygen gas, an accurate, fully coupled calculation model would ultimately have to take into account all gas bubbles in the liquid electrolyte. However, the bubbles are directly or indirectly influenced by all four main factors mentioned above. The amount of bubbles in the electrolyser again depends on the electricity flowing through the electrolyser and the electrolyte flow itself. Therefore, accurate modelling of the gas bubbles cannot really neglect the interaction with one of the main factors, Kemppainen explains. This in turn makes the bubbles perhaps the most difficult detail of electrolyser operation to simulate, he describes. In addition, the operating temperature also has an effect on everything: The electrolysis reaction and the resistance losses also heat the electrolyser, and the bubbles influence the reaction rate and the thermal conductivity of the electrolyte.

The scientists at the ZSW in Ulm also use modelling software. Mathematical models enable researchers and developers to better understand the physical and chemical processes within the fuel cell. They are thus the basis for the development of completely new approaches and represent an important pillar within fuel cell research, from development to system optimisation.

Simulation thus helps to reduce development time and costs through targeted optimisation. In addition to model development, experimental verification is thus guaranteed. The results of this experimentally supported modelling work flow directly into the development work of the researchers at the ZSW. Practically all questions concerning the dimensioning of distribution channels in the bipolar plate up to the complete FC stacks were developed with the help of modelling.[…]

… Read this article to the end in the latest H2-International

Author: Niels Hendrik Petersen