E.ON using IBM quantum cloud to optimise renewables – www.computing.co.uk

Energy company E.ON is working with IBM on using quantum computing to optimise the delivery of renewable energy.

In an online media briefing last week, Victoria Ossadnik, chief operating officer - digital at E.ON SE, said the need to transition away from fossil fuels is an urgent but highly complex task.

Both supply and demand are much less predictable: renewables are affected by the weather, and heat pumps and electric vehicles (EVs) use electricity rather than gas or petrol, requiring grid upgrades and new cabling to support them. Increasingly, the cost of delivering electricity is outstripping that of generation as infrastructure creaks under the strain.

E.ON is one of Europe's largest operators of energy networks and infrastructure, running power stations, renewable generation facilities, power grids and EV charge points across Europe. The firm has a target to educe emissions by 55 per cent by 2030, and currently connects more than 40,000 renewable assets each year to a grid infrastructure that was designed for much more predictable workloads. These new sources are making the grid more and more complex and more vulnerable to failure, Ossadnik said.

Adding to this complexity is the emerging peer-to-peer energy market, where households and businesses can trade power generated from rooftop directly to consumers.

Managing decentralised power networks with potentially millions of small sources, their output varying with the weather, means staying on top of a huge number of variables. There is a limit to what even the largest supercomputer can do in this scenario, and there are some requirements that are simply beyond the bounds of classical computers, however powerful.

Optimising complex systems is a key strength of quantum computers, which is why, even though the technology is in its infancy, E.ON is getting in on the bottom rung.

It may be early days, but according to IDC analyst Stefano Perini, the quantum computing market was worth more than $700 million in 2021 and is expected to be $3.5 billion by 2025. Not huge numbers, perhaps, but an impressive rate of growth. Currently, this growth is being driven by financial services and manufacturing, but with the need for rapid decarbonisation it makes sense that energy companies get on board too.

"All these industries are united by the need to address extremely important innovation challenges around their business and operating models," Perini said. "We can expect that quantum computing will be one of the important enablers of the energy transition."

Quantum computers will play a complementary role to high performance computing (HPC) and AI rather than replacing them, said Ossadnik. Their ability to try multiple calculations at once (rather than having to iterate through the alternatives) makes them ideal for solving problems like grid balancing and optimal cable routing, where hundreds of thousands of variables may come into play. And while they are not suitable for big data, they can be very useful in designing better predictive algorithms and ML models that can incorporate many more dimensions.

Quantum computers will help ask better questions of the data, Ossadnik added, for example where to optimally lay cables in a big, crowded, historic city like Hamburg or London.

"We need to put much, much, much more cable into the ground, but you need to find out what do you really need which kind of infrastructure. This is a super complex system, because of this highly volatile demand and highly volatile supply, and with a quantum computer we can give it hundreds of thousands of little bits of information and find out which model is supported best."

Early tests using IBM's quantum cloud services have already helped improve these infrastructure models, but potential applications go further. For example, designing algorithms that can provide the right nudge factor' to ensure behaviours align with the availability of power.

"What happens when everyone charges their cars and runs their heat pumps on a cold, windless night? The next morning the city just won't have anybody going to work because unfortunately the cars are not charged. This shouldn't happen, and that means we need to control this in a super smart way. In a highly democratic society how do we do this? We need to learn which algorithm is best, how we can incentivise people to behave in the right way and to protect the energy resources in the short term, and this is where we would use quantum computing."

Heike Riel, head of science & technology and lead of IBM Research Quantum Europe, believes the technology will have a crucial role in developing other cleaner technologies. Since, at the sub-atomic level, all matter is governed by quantum physics, the ability of quantum computers to model the behaviour of atoms and molecules will be far greater than anything currently possible. This has important implications for battery design; the complex chemical processes that occur in batteries are still poorly understood.

"If we better understand all these processes we can better optimise the batteries, which means that we can reduce the weight of electric cars because you need less battery to store the same energy," she explained.

Riel offered other cases where the ability to better understand matter at the atomic scale could advance the sustainability cause. For example, the race is on to find materials that can efficiently adsorb CO2 so it can be removed from the atmosphere, and to find alternatives to energy-intensive industrial practices such as the Haber-Bosch process used to produce ammonia for fertilizer - fertilizer production consumes 3 per cent of global power.

So, tasks are already being lined up for the nascent technology to tackle, but with most current devices requiring operating temperatures around 10 milli-Kelvin ("100 times colder than outer space"), is there a risk that the energy required to refrigerate them negates the gains? Not at all said Riel, pointing out that the IBM Summit supercomputer has a peak power of 15 MW, while IBM's current quantum computers run on a 36 kW supply.

"In general it's really much lower power. Also you have to keep in mind you can solve questions in the future which you cannot even solve with a classical computer - even if you make it bigger and bigger and bigger it would just consume more and more and more power."

Excerpt from:

E.ON using IBM quantum cloud to optimise renewables - http://www.computing.co.uk