On this interview, we characteristic Stefan Niessen, Head of Expertise Area Sustainable Vitality & Infrastructure at Siemens Expertise, as he sheds mild on the intersection between tutorial analysis and industrial innovation within the realm of sustainable power and infrastructure. With in depth expertise in main technological developments, Mr. Niessen shares insights into the challenges and alternatives in integrating rising applied sciences into present power grids. This interview is a part of AI Time Journal’s AI for Vitality Initiative.
In your twin function, how do you steadiness the calls for of main technological developments in sustainable power with the theoretical and academic elements of your professorship?
We now have three levels of maturity through which we deal with these numerous subjects. One is pure college analysis as a professor the place I can work on subjects the place changing right into a product is 5 years or longer away. Then we have now publicly funded tasks through which I could take part both as a professor or as a Siemens worker with my group, permitting us to co-create with different universities, prospects, and rivals on pre-development subjects. This co-creation mode is nice as a result of any new expertise can solely work if deployed by an ecosystem of companions. The third diploma is inside Siemens tasks the place my group does precise product growth work for enterprise items, guiding a promising expertise from the preliminary college levels to an precise product stage the place it may be successfully deployed in actual life.
Are you able to focus on a particular undertaking or innovation below your management at Siemens that exemplifies the mixing of AI into sustainable power methods?
One instance is the multimodal power storage and suppleness concept I discussed, which requires sensible interplay between grid operators, grid customers, and renewable electrical energy injectors to synchronize consumption with era. The grid operator must estimate how loaded the grid is, which requires AI for forecasting future demand for warmth, cooling, traction, and electrical energy costs and renewable era. Knowledge cleansing and high quality assurance is one other AI aspect, as everytime you measure knowledge, points come up that want detecting and correcting. Grid state estimation, processing knowledge from completely different sources for an summary of the grid’s working state, is the final aspect.
What rising applied sciences or strategies in power buying and selling do you imagine maintain essentially the most promise for enhancing sustainability globally?
I’ve sympathy for power buying and selling, being a part of the group that established the European Vitality Alternate. Nonetheless, we’re not but utilizing the complete potential additional down the voltage ranges the place sensible, internet-connected units like charging stations, warmth pumps, batteries, and PV inverters exist however should not but actively buying and selling, although they may. To reply your query, I imagine automated native power buying and selling inside grid boundaries is a key rising expertise. We can’t rely that grid capability will all the time be absolutely out there, so sensible injection is required. We now have field-tested how this could work in tasks like pebbles in southern Germany and with the Viennese metropolis utility.
Contemplating your work in mobility, what are the largest challenges and alternatives for integrating electrical automobiles into present power grids?
The most important problem is low voltage grid capability, as these grids weren’t designed to attach many charging stations. Sometimes buried underground in Europe, quickly growing capability is tough in comparison with the US the place you possibly can add one other line on poles. This implies we should purchase time by making the grid interplay more and more sensible to synchronize charging with native PV era and decrease grid load. Additional variables like warmth pumps, stationary batteries, industrial processes, and buildings offering flexibility as thermal power storage can contribute.
Together with your multimodal focus, how do you strategy optimizing these methods for effectivity and sustainability with out compromising reliability and accessibility?
I might argue it truly helps, as decarbonized methods robotically imply extra decentralized constructions with inherent redundancy benefiting reliability. In Germany, there are over 2 million decentralized electrical energy mills, principally rooftop PV. These multimodal couplings convey flexibility by way of inherent storage performance, as any client in a position to delay consumption supplies flexibility, enhancing reliability.
Might you elaborate on any latest conversion part developments considerably impacting power infrastructure sustainability?
No single expertise solves the power drawback, however warmth pumps can now present increased temperatures as much as 150°C or extra at industrial scale, permitting use of waste warmth sources. When fuel costs spiked, many industrial websites invested in warmth pumps to make the most of beforehand wasted decrease temperatures. Developments in battery supplies like sodium-ion as an alternative of lithium, utilizing considerable sodium, is also a breakthrough even when power density is decrease. This mature expertise already has automobiles manufactured in sequence in China. I nonetheless count on extra battery surprises within the subsequent two years because it stays a extremely lively analysis discipline.
How have lifecycle assessments influenced sustainable undertaking design and growth at Siemens?
We do in depth lifecycle assessments, aiming for all our merchandise to have environmental product declarations by 2026, going past regulatory necessities. My group develops the methodology to more and more robotically generate lifecycle assessments for {hardware} merchandise, for software program and repair choices.
In transitioning cities and areas to sustainable power, how essential is AI’s function in forecasting, planning and managing these advanced methods?
At present, it’s not essential, however sooner or later AI will help quickly discover believable assumptions when creating decarbonization roadmaps for cities or areas. You want assumptions on future consumption profiles for infrastructure that doesn’t exist but, and AI can contribute believable assumptions.
Lastly, reflecting in your profession, what recommendation would you give younger engineers and researchers aspiring to considerably influence sustainable power and infrastructure?
I might advise younger engineers to try to really perceive fundamentals, because the legal guidelines of physics don’t change. Sooner or later, resilience would possibly develop into as vital as sustainability. The multimodal, sector-coupled power methods benefiting decarbonization are additionally advantageous for safety of provide, so these subjects are converging. Engineers additionally are likely to overlook understanding prices and creating a real curiosity in enterprise fashions is extraordinarily useful.