How AI is advancing the energy transition to net-zero
By Erin Kalejs
Artificial intelligence (AI) might not be the first tool that comes to mind when thinking about the energy transition or green energy. Which is why it might surprise you to learn that the energy industry is investing heavily in machine learning for purposes such as, sensor-connected power plants, wind turbines that speak to each other and smart grids balancing electricity supply and demand. In fact, these solutions are already being used all around the world.
AI for Good’s webinar, “Can AI and digital tech accelerate the energy transition?“ explored how AI is helping the energy transition meet the challenge and goal of a lower-carbon future.
“AI can enable greater prediction and optimization, and these are already well-established applications in all areas,” explains Amy Challen, General Manager for Artificial Intelligence at Shell. She reminded everyone to set realistic expectations “I don’t think any of us here would claim that AI and digital technologies are really the only solution to the energy transition. The solutions are going to be physical and chemical technologies along with a profound shift in the way we do things, but I do believe that AI and digital will be really important enablers. ”
She added that this is especially evident “through increasing the energy efficiency of everything we currently do, in both production and consumption.”
It all starts with data
AI is of course not possible without data and digitization as Catherine Newman, chief executive officer of Limejump pointed out.
“The big challenge I think we all know with AI is, it’s about the quantity and quality of the data required to actually do something productive. It’s a massive challenge,” says Newman. “We can’t optimize the grid without having access to the data that tells us how it’s operating in real time,” she added.
This is why it is imperative to get the general public along with private companies to provide accessible data to help accelerate the energy transition. As Ed Mitchard, Chief Technical Officer & Co-founder at Space Intelligence explained, the public should be involved with monitoring land cover change of their local environment.
“We have all these people around the world with this brilliant piece of technology that has a GPS in it. So many people have smartphones now, we should be taking advantage of those because they capture really rich spatial information on what’s there at that point in time”
Another issue that was discussed in the session was transitioning from a demand led system to a supply-led system in the future.
To find out more about how blockchain is supporting renewable energy watch the AI for good session here
“As energy consumption has grown over the last century, we’ve just built more and more infrastructure. We put more power plants in place, more gas and oil. But as we move and transition to renewables, what we see happening is that obviously renewables are only available at specific movements. And it’s very difficult to store energy. By and large we’re not going to be able to store all the energy that we generate from renewables and make them available at the times we need them,” explained David Boundy, Chief Technology Officer and General Manager Europe, at Innowatts.
Which is why, the real challenge he adds is: “how do we change our grid to be a supply led grid?” For this to be achieved Boundy explained “we need to be able to plan around the consumption side, and then we also need to control what’s happening in the supply side so in essence it’s about identifying which loads can be moved to which times to match with the supply.”
Boundy added, “There is actually a lot of load that can be moved whether it’s heating or cooling of a home or when an electric vehicle is charged. All of those different things can be managed in an intelligent way so that they can optimize and be aligned to the grid, the opportunity arises when it’s automated and that’s where were trying to get to.”
To find out more about AI and the energy transition watch the AI for good session here
Building ‘green AI’
The environmental impact of doing AI in itself, is another important issue to address.
Mitchard said “launching a satellite involves a massive burst of carbon and then storing that data and running deep learning AI models also burns carbon. Many data centers now use renewable energy because it’s the cheapest energy in many other parts of the world.” However overall, he insisted that we must “be mindful of and work to reduce computing work and not run unnecessary code. Businesses should be sensible about what they do and that will reduce their AI processing and emissions.”
On the environmental impact of AI, “it’s not just about carbon, any respectable AI company needs to look at their supply chain and understand how their supply chain is providing them with their services. For example, how is the data center dealing with electronic waste and recycling? Whilst data centers are putting pressure on the grid today, they can also be used to help ensure that renewables are being optimized on the grid,” said Boundy.
One big problem with the current energy system is there are no accurate price signals for clean energy producers which is why a lot of industry experts are looking at zonal pricing – a pricing zone to exchange energy within a market area. “If we really want to try and drive these very fundamental changes that enable ourselves to use machine learning. Zonal pricing is definitely one of the favored options on the table,” Newman explained.
Boundy agreed with this point by saying “ultimately we need to have more granular pricing which will invoke a change in behavior that can be supported by automation and AI.. For example, today there are times where there’s negative pricing in the wholesale energy market if that was transferred through to the consumers would that drive a change in behavior? I think it would.”
Boundy also spoke about the risk that inadequate pricing could exacerbate fuel poverty for people who have little access to renewables.
Using blockchain technology to take action
Along with AI, blockchain is also playing an important role in the energy transition by helping integrate in a fully transparent way distributed energy resources such as batteries and solar panels into the grid. Blockchainenables the traceability of renewable energy products.
In an AI for Good webinar about renewable energy and blockchain technology, Natasa Marangou, Blockchain Product Manager, at Shell said the company’s focus is on “supply chains and the energy transition. We really see blockchain technology as this neutral, industry wide infrastructure that can enable trusted and seamless exchange of information between parties in a supply chain.”
“Ultimately, it’s about energy optimization and orchestration at large scale, that’s where blockchain can really benefit from AI to provide that optimization,” believes Marangou.
Blockchain can also help accelerate the energy transition by making it possible “to track sustainable energy and certificates from the origin through every stage and transaction” says Marangou. This means an “additional layer of certainty and trust” can be given to users that the electricity they are purchasing is in fact from renewable sources.
Combining blockchain technology with AI will create many opportunities and solutions for the energy transition, as Marangou explains, “on the demand side, I can see us using AI to identify those structural gaps in the temporal mismatch between renewable power supply and demand to inform procurement strategies. On the supply side we can use AI to perform decisions on how to get to 100% green energy and for better temporal-spatial matching of renewable power generation.”
This webinar highlights that digital technology can have a positive impact in reducing global emissions of greenhouse gases. But a lot more innovation is required to harness at scale AI and Blockchain technology in the digital backbone of a decarbonized energy sector. Follow AI for Good Summit to stay on top of these evolutions.