Connecting computation to experiments
In the coming months, the integration of the Center for Computational Energy Research (CCER) into EIRES as its fifth focus area will be completed. CCER’s scientific director Peter Bobbert shares the story of the past, present and future of computational energy research in Eindhoven.
‘Some ten years ago the board of the (then) Department of Applied Physics asked me to investigate the possibility to start a collaboration with DIFFER on computational sciences,’ Bobbert recollects. At that time, two developments came together: Shell and NWO had just established a large scale joint research program on Computational Sciences for Energy Research, and the FOM research institute DIFFER at Rijnhuizen was planning on moving from Rijnhuizen to the TU/e campus. ‘That is when the idea arose to launch a joint research center on computational sciences for energy research, and on 21 June 2017, CCER was officially launched.’
Just before that, Vianney Koelman, then Vice President Computational R&D at Shell Bangalore, India, had decided the time had come for him to retire from his career in industry. He was the perfect candidate to act as CCER’s first scientific director, since he had visited Eindhoven several times as a representative of Shell to get acquainted with the people and ongoing projects within the scope of the NWO/Shell program. ‘We were very fortunate that he said yes when we asked him,’ Bobbert says.
‘I am happy that we managed to find a way for all three partners to come together and strengthen each other’
Foster synergies
During the first few years of its existence, CCER had DIFFER as a home base, he tells. ‘We organized our seminars there, which turned out to be excellent occasions to physically connect computational scientists to experimentalists working on energy systems.’ In the meantime, in the summer of 2020, EIRES was launched. ‘Since it makes no sense to have three different entities focusing on similar topics in splendid isolation on the same campus, we explored the possibilities of integrating CCER into EIRES, while at the same time making sure that DIFFER could stay involved. I am happy that we managed to find a way for all three partners to come together and strengthen each other,’ Bobbert says.
To the outside world, CCER will keep presenting itself as a separate entity. However, in an organizational sense it will become a fifth focus area under the EIRES umbrella, acting as a cross cutting theme across the four other focus areas. ‘Together with Suleyman Er from DIFFER, I will join the management team of EIRES as a focus area leader,’ Bobbert explains. The CCER seminars have physically moved to the Disruptor building, where EIRES is located. ‘That is an excellent location to foster social dynamics and expand the community,’ he comments.
Broader outreach
Over the past six years, CCER has not only been building a community of computational scientists in Eindhoven, Bobbert emphasizes. ‘Together with NWO and with support from Shell, we also established the Applied Computational Sciences (ACOS) conference to bring together Dutch academic and industrial researchers in applied computational science in the broadest sense.’ This series of conferences is succeeded by the Dutch Computational Sciences (DUCOMS) Day and computational scientists in the Netherlands have organized themselves around the National Agenda Computational Sciences. ‘This way, the stimulus Shell aimed to provide to the field with its initial NWO research program certainly has paid off.’
With the integration of CCER into EIRES, Bobbert hopes to broaden the scope of the computational energy research in Eindhoven. ‘We have already established links with the heat storage researchers, which culminated in a joint project called Wax+ where we are looking for novel composite materials that can store and release heat as efficiently and fast as possible. We want to expand this type of intensive collaboration between theory and experiment to other topics as well. For example, we are now reaching out to the metal fuel people, to see how our models can help them understand and improve the oxidation and reduction cycle of the iron powder.’
Power of AI
In terms of computational technologies, CCER has the ambition to expand upon the machine learning and artificial intelligence expertise that the center began to build under Koelman. ‘Computational sciences are indispensable when it comes to developing new types of solar cells, batteries, or energy networks. However, the complexity of the problems we work on is immense. We are working on multiscale problems, both in terms of time and length scales. Artificial intelligence is of great added value there.’ Bobbert gives some examples. ‘At CCER, we have a strong track record when it comes to quantum calculations. By using machine learning techniques we are now able to scale up our atomistic models and bridge the gap between the nanoscale and the continuum. This way, we will be able to develop large scale models for the behavior of things like plasmas in a fusion reactor or flows in hydrogen reactors, based on a fundamental understanding of what happens at the smallest of scales. And we can also use this type of artificial intelligence to study the behavior of large scale energy systems, for example by using digital twins of entire energy networks. Artificial intelligence enables us to do more with the same computational power, thus opening up a vast new array of possibilities for energy research.’
