The global energy industry finds itself approaching a crucial crossroad. For the first time in half a century, oil accounts for less than 30% of global energy consumption. Low-carbon sources now exceed 40% of electricity production. Yet, despite record renewables, carbon emissions are still rising, coal plants remain operational and global electricity use grows at 3% annually.
We’re caught in a green consumption paradox. We’re building renewable capacity faster than ever, but the pace of our energy consumption exceeds even that.
For professionals such as engineers and urban planners that are collaborating with the energy industry, this presents a critical challenge for truly future proofing the built environment and energy infrastructure.
AI’s impact on power systems
This issue has been supercharged by the increasing availability and capability of artificial intelligence (AI) technologies that promise to revolutionise how we manage energy. AI is also poised to consume staggering amounts of it, so we must be ready to support it. The International Energy Agency projects that global electricity consumption from AI and data centres will more than double by 2030.
The UK’s AI Energy Council meeting last month underscored this challenge. The UK must support a 20-fold increase in computing capacity over the next five years. With £2bn committed to grid upgrades, the instinctive response is to build more infrastructure. But this traditional approach of builds, additional pylons and grid connections is not only extremely slow, but extremely costly.
The real challenge isn’t just about generating more electricity; it’s about managing when and how we use it. As we continue to move towards electrified processes in most areas of society, we’re adding hundreds of millions of new electricity-hungry devices to our grid. These devices must adapt to an increasingly variable electricity supply as renewable generation fluctuates with weather patterns.
This is where AI must be shifted from problem to solution. Modern AI agents can orchestrate millions of smart devices without compromising user comfort. Air conditioners, heat pumps, electric vehicles and batteries can shift consumption patterns automatically. When renewable generation is abundant, these systems increase consumption and alternatively, when supply drops, they can reduce demand simultaneously.
Re-engineering the energy sector with AI
The path forward centres on intelligent technology and artificial intelligence, creating systems for automated device management in real-time. Despite scepticism about AI’s potential, several crucial technologies are emerging to address energy challenges effectively.
Arguably the most critical energy management solutions harnessing AI is demand response technology, which is a truly flexible, immediately deployable solution that reduces energy demand and redistributes unused power when demand peaks and prices are high.
When enough demand response devices are connected on this cloud-based network, they can optimise energy consumption across power grids at scale, reducing strain during peak usage periods without relying on environmentally harmful and costly backup power facilities, while also providing stability when renewable energy generation experiences temporary shortfalls.
By helping to balance the grid, this can, in turn, enable energy savings and have a positive impact on energy costs. We have seen this come to fruition recently through our work with University of Wales Trinity Saint David, which secured 13% in energy savings after installing demand response tech into its student accommodation.
Demand response’s critical role in the electrification age
The International Energy Agency has recognised our entry into an “electrification age”. Within this period, AI’s role in managing worldwide energy consumption becomes crucial. Even as AI increases demand, it must also be a tool for managing it and the energy transition’s success ultimately depends on these technologies.
The scale of demand-side flexibility possibilities is remarkable. Research from the National Energy System Operator (Neso) indicates that by 2030, buildings alone could provide up to 10-12GW of adaptable power demand, representing approximately 25% of current UK peak electricity requirements. This capacity will expand significantly as electrification accelerates across the built environment. Combined with essential battery storage infrastructure, the UK will rapidly acquire sufficient demand-side flexibility to address up to 20% of total annual electricity needs, enabling a fully decarbonised power grid.
Realising this potential means buildings, their electrical appliances and the equipment these use for energy management must become smarter and better connected. While AI does present a double-edged sword, the property owners, landlords and facility operators that integrate AI-enabled demand response technologies into their energy systems will be the key to unlocking energy savings and in turn, cost savings, not only for the individuals directly impacted, but also for all energy consumers.
But everyone has a role to play here. Crucially, it will take energy players and the property and design ecosystem, particularly those with roles at the crux of built environment and energy infrastructure design, to help scale the adoption of these critical solutions at the pace needed to meet our global net zero commitments.
Randall Bowen is UK managing director at Voltalis