The Greatest Upgrade in a Century: Takeaways from the Energy Tech Summit

Last week I spent a few days in the Basque Country speaking with leaders from across the energy sector about the challenges of the next decade and how we can move forward to the clean power systems of the 2030's, creating value for the environment, society and the economy along the way.

AI Might Break the Energy System, but it also Might Save It

It wouldn't have been a technology conference without a long discussion on pros and cons of AI. While energy companies are using AI for standard day to day activities, like any business - routine tasks, etc. Utilities have a long way to go before they can actually utilise the technology. Some of the reasons for this are in historical data management:

• Many systems are still on premise, putting as limit on the amount of processing that can be done at any one time
• Companies are using data lakes and other outdated tech, which are not set up for modern AI technologies
• Very little data is collected in terms of real time or time series data, limiting the types of solutions that can be put in place

Data and software infrastructure within the distribution and transmission network require a massive overhaul to enter the age of AI and unlock the solutions that Deep Tech offers. This is likely to include:

• Greater smart meter data availability & data management - currently unavailable in most countries (for either technical or regulatory reasons)
• Increased Meter installations - more ways to collect data from across the network and feed it back to provide actionable insight and time series data (which is rare today in the energy sector)
• Scale-able data processing, allowing increasingly challenging problems to be addressed
• Secure cloud infrastructure - energy networks have and will always be targets for physical and digital attacks. It is crucial to ensure that the growth in data driven energy systems don't create new risks. Networks are already looking at energy security in a post quantum world

While there is a long way to go, there are plenty of people looking to capitalise on this market. Companies in the US, Europe and Asia are currently developing AI chips for the energy sector, citing this as a major market.

On the other side of the challenge, AI is incredibly power hungry. With the huge demand for things like LLMs, compute power requirements are doubling every 100 days and certain countries in Europe are now pausing the construction of data centres, without necessary investment in supporting energy infrastructure. The demand for AI is unlikely to decrease in the coming years, and so there are growing requirements from energy networks and data infrastructure providers to work together to ensure that there is sufficient base load, while also working to increase the energy efficiency of compute. There was a lot of support for using Arm and RISC-V computing capability to reduce energy consumption.

One potential technology solution on the horizon is quantum. Quantum computing works with different principles to traditional computing, providing capability to work on far more complicated problems, with a broader range of inputs, outputs and provide significantly quicker responses to grand challenges. There are also studies that indicate that 7 Gt CO2e/year by 2035.

The distribution network evolution presents the biggest opportunity in a generation

Energy networks across the world represent some of the most complex human made systems ever built. However these systems are, in many cases, more than 100 years old and designed for large, centralised dispachable power plants, with unidirectional power flows.

The energy transition brings billions of small, distributed, intermittent virtual power plants, with bidirectional power flow, which put increasing pressure on our distribution networks.

Challenges today as a result of the ageing grid include delays to electrification and renewable projects, around 15% of renewable energy loss due to grid limitation - power is being created, it's just difficult to move if efficiently.

What this means is that, while there is a supply of flexible energy ready to engage, most of the value is still not realised due to grid limitations, leaving economic opportunities on the table and leading to significant curtailment. The flexibility iceberg, presented by Plexelgrid, shows this nicely.

As distribution networks evolve and provide the opportunity for mass participation in the energy system, we can expect to see the largest shift in more than 100+ years, comparable to the advent of smart phones in terms on new business models, revenue streams, innovations. This is backed with more than $600Bn in annual global investments in distribution networks by 2030.

New entrants are building the next phase of energy infrastructure

These changes are being driven by the start ups, working in collaboration with networks and infrastructure operators, which means that innovations from the finance, healthcare and materials markets, as well as first of a kind innovations are enabling the future of energy to be secure, resilient, clean, affordable and smart.

At Cambridge Consultants we're seeing this first hand through the Climate Tech SuperCluster Accelerator where we're supporting 11 early stage technology companies to develop deep tech innovations that support the energy transition and unlock the technical and business opportunities that are emerging from grid upgrades.

What is clear is that the opportunity to develop and scale innovation in this sector has never been greater, the opportunity for investors and start ups is huge and the next decade will see the largest infrastructure transformation in living memory.