Kyrylo Leus

I have been thinking about batteries that were discussed as “support” for renewables for a long time. That framing feels outdated. Storage is becoming the system layer that decides whether clean power can actually work at scale. Solar and wind can produce cheap electricity. But batteries decide whether that electricity can be shifted, traded, stabilised, and used when the grid actually needs it. The battery market is now moving into a new phase: not just more capacity, but more strategic capacity. Paired with solar. Used for grid services. Built around congestion and price signals. I think that the most interesting climate-tech companies may not be the ones producing energy. They may be the ones making clean energy dispatchable. What matters more for the next 5 years — adding more renewable generation, or making existing renewable generation more usable?

What do you think about this shift? Microsoft, Google, Amazon, and Meta are backing a new initiative to pilot climate technologies inside data centres. That makes sense. Data centres have intense needs such as power, cooling, water, backup systems, reliability, efficiency, real-time monitoring. This makes them perfect laboratories for climate tech. If a technology can work inside a data centre, under strict uptime pressure, it may be much easier to trust elsewhere. But there is also a tension. Are data centres helping scale climate solutions? Or are climate solutions being used to justify more data centre growth? Both can be true. So should data centres be seen as a climate-tech problem, a climate-tech platform, or both?

Lithium-ion has dominated the storage conversation. But AI-driven power demand may create space for another category: long-duration storage. A recent project in Switzerland plans to use vanadium flow batteries for a large AI data-centre energy system. That is interesting because flow batteries are not trying to win the phone battery race or the EV battery race. They are built for a different job: longer duration many cycles lower fire risk grid stability heavy infrastructure use As AI loads grow, storage near data centres may need to be safer, longer-lasting, and more flexible than standard short-duration batteries. So AI demand could accidentally accelerate the market for long-duration storage. What do you think? Will data centres become one of the first major commercial markets for flow batteries?