What is driving the shift from businesses being passive energy consumers to becoming active participants in the electricity system?

“Energy is no longer simply an overhead that businesses can purchase and forget about. Price volatility, grid constraints, electrification and the financial cost of disruption have made it a strategic operational issue.

“Businesses also have far more control than they did previously. Battery storage, on-site generation, smart meters and increasingly sophisticated energy-management systems allow them to decide when they consume electricity, where it comes from and, in some cases, whether surplus capacity can be offered back to the system. At NatPower, we have a global development pipeline of approximately 30GW across renewable energy and energy infrastructure projects, and we see this shift accelerating, not slowing.

“AI companies, data-centre operators and developers need to treat energy infrastructure as a competitive variable, not a fixed cost. The next generation of data centres will need to be built faster and run at a fraction of today’s energy cost and power infrastructure spend to remain competitive as margins come under pressure from lower-cost international rivals, including Chinese competitors. Shared infrastructure models are how that speed and cost base can be delivered.

“This is not just about securing a lower electricity price. It is about resilience, predictability and operational control. Companies will increasingly need to manage energy with the same discipline they apply to finance, technology and supply chains.”

Which organisations stand to gain most from battery storage, on-site renewables, microgrids and demand-response technology, and which may struggle to justify the investment?

“The strongest business cases tend to be found among organisations with high or concentrated electricity demand, costly peaks, critical operations or some ability to shift when power is used. Manufacturers, ports, logistics hubs, cold-storage operators, large commercial estates and certain data-centre applications are obvious examples.

“AI data centres are the clearest illustration of both the opportunity and the difficulty. Data-centre and AI infrastructure investment is now central to the UK’s growth agenda, but only if capacity can actually be delivered, not just applied for. Time to market is the binding constraint. A self-built substation with grid access already in place removes the single largest delay in bringing a data centre online. Our model is designed to cut that timeline by around four years compared with a standalone connection application.

“The same logic applies to other sectors such as marine electrification. Ports and shipping face similarly large, variable loads as they electrify, with similarly little tolerance for instability that could affect port operations.

“Companies operating electric vehicle fleets, refrigeration, heating, pumping or other controllable equipment may also have considerable flexibility without interfering with their core operations.

“A business should not install a battery simply because the technology is available. For a smaller or low-energy business, a company occupying premises on a short lease or an organisation with very limited flexibility, the economics may be less convincing. Every case has to begin with the site itself: its consumption profile, connection capacity, operational risk and future energy requirements. There is no universal smart-energy package.”

How should business leaders build the financial case for smart energy infrastructure when returns may include avoided outages, reduced peak charges, greater resilience and new revenue?

“Judging an investment on electricity-price savings alone is the biggest mistake I see.

“The calculation should include reduced peak charges, better use of on-site generation, lower exposure to volatile prices, avoided curtailment, potential market revenue and the value of maintaining operations during a disruption. Resilience has a financial value, even if it does not appear as a new line of revenue. For a factory, port or temperature-controlled warehouse, a short interruption can cost far more than several months of electricity savings.

“By sharing energy infrastructure rather than each user building its own, we can cut a data centre’s physical footprint by around two-thirds, speeding up planning consent. GigaPark infrastructure allows an AI data centre to operate at a load factor of 1.25 to 1.30 over a 1.0 grid connection, reducing its request to the grid and allowing a faster connection. The connection can be phased, and capacity increased as a facility scales, without requiring a fresh connection process each time. Combined with the four-year reduction in time to market, that provides a materially faster and smaller route to operational AI capacity than the standalone alternative.

“The broader system benefits are not limited to data-centre operators. Our analysis of the GigaParks portfolio puts avoided wind-curtailment savings at £3.5 billion a year and avoided grid-balancing gas costs at £12.5 billion over a decade. These are costs that would otherwise fall on all UK electricity users, not just large-scale consumers.

“I would build the case on conservative assumptions and avoid relying on market revenues that may not remain available at the same level throughout the asset’s life. Flexibility income should strengthen a fundamentally sound project, not disguise a weak one. Leaders should also calculate the cost of waiting: delayed action can leave a business facing higher connection costs and greater price exposure later.”

What role can AI-powered energy-management platforms play in deciding when a business should consume, store, generate or sell electricity?

“AI can turn a collection of individual energy assets into an intelligently managed system, analysing demand, renewable generation, prices, weather, production schedules and battery condition, then deciding when electricity should be consumed, stored or supplied to the market. That ability to process multiple variables continuously is genuinely valuable.

“But AI is not a fix for bad infrastructure. A badly designed energy system does not become a good one because an algorithm is controlling it. The quality of the data and the underlying hardware still matter more than the software sitting on top of it.

“Transparency, cybersecurity and human oversight are not optional extras. The system must operate within clear limits, protect critical loads and allow the business to intervene when circumstances change. AI should optimise the commercial and operational strategy. It should never be allowed to dictate it mindlessly or serve as an excuse for management to abdicate its responsibilities.”

“Demand response can generate meaningful value, particularly for larger users with controllable loads or storage assets. For many businesses, however, the revenue is not yet simple or predictable enough to justify an investment on its own.

“Returns can change according to the service being provided, market conditions, location and asset availability. Businesses may also encounter different technical rules, contractual arrangements and routes to market. That complexity makes it difficult to forecast income confidently over the full life of an investment.

“The more credible approach is to combine several benefits. Reduced energy costs, peak management and greater resilience should form the core of the business case, with demand-response revenue providing additional value. The opportunity is real, but it needs greater simplicity and clearer long-term signals.”

How can organisations participate in energy markets without disrupting production, customer service or other critical business operations?

“The business must come first. Energy trading comes second.

“Organisations should identify which loads are genuinely flexible and which must always be protected. Refrigeration, vehicle charging, heating, pumping and some industrial processes may be adjustable within agreed limits. Critical production, safety and customer systems should be ring-fenced.

“A GigaPark operates fully integrated with the grid under normal conditions but can also operate independently if the grid goes down, keeping critical loads such as data centres and ports running through a blackout that would otherwise take them offline. A large long-duration battery storage system also offers stability to the grid, helping to prevent a blackout in the first place.

“A GigaPark’s long-duration battery storage balances load and frequency, dynamically reallocating capacity across the site as immediate fluctuations and longer seasonal patterns occur. It uses the full infrastructure to trade energy efficiently between users rather than requiring every participant to overbuild dedicated backup capacity that sits idle most of the time.

“Battery storage can provide flexibility without requiring the underlying business to reduce output. Automated controls can then respond to market signals while maintaining minimum reserves and respecting operational limits.

I would begin with detailed load analysis and a controlled pilot rather than immediately exposing an entire site to a new operating model. Flexibility should be designed around the operation. The operation should never be compromised to chase uncertain market revenue.”

What technical, regulatory and contractual barriers currently prevent more UK businesses from installing storage, establishing microgrids or selling their flexibility to the grid?

“Britain’s electricity system has a queue problem. Ofgem now puts total demand connection applications at 125GW, up from 41GW in under a year, with data centres alone accounting for around 80GW of that. The regulator’s new consultation on commitment fees for speculative projects is a direct response to a system straining under applications it cannot process at the pace AI investment now demands.

“Grid connections remain one of the greatest obstacles. A business can have the capital, technology and a viable site, yet still face a lengthy connection process, uncertain costs or insufficient network capacity.

“Planning, metering requirements, fragmented market arrangements and inconsistent technical standards add further complexity. Existing supply contracts may restrict access to independent flexibility providers, while microgrids raise additional questions around network ownership, licensing and responsibility for system operation.

“We believe a model built around shared, consented energy infrastructure, rather than each site fighting its own planning and connection battle, is better suited to UK conditions.

“Britain does not have a shortage of businesses willing to invest. Government, Ofgem and NESO recognise many of these barriers. The priority now is delivery: a clearer, faster route from feasibility and connection through to operation and market participation.”

As the UK develops its clean-flexibility plans, what practical steps should businesses take now to become more resilient, lower their energy costs and capture emerging opportunities?

“Start with the data, not the technology.

“Businesses need to understand when they consume electricity, where their peaks occur, which operations are critical and which loads could be moved safely. They should then consider how electrification of transport, heating or production could change demand over the next five to ten years.

“That analysis will identify sensible first steps, which may include improving efficiency, reviewing tariffs, reducing peaks or assessing storage and on-site generation. Early engagement with the network operator, energy supplier and potential flexibility partners is essential because connection capacity may determine what is commercially achievable.

“Rather than a data centre or port applying for its own grid connection and waiting in the same queue as everyone else, it can connect to shared local energy infrastructure that already has consented, transmission-level grid access and actively manages the power quality and balancing services required by both the grid and the site.

“Any proposal should then be tested under conservative assumptions, including lower-than-expected market revenues and changes in future energy use. Companies should act now, but invest intelligently. Waiting for perfect certainty is not a strategy. Neither is buying technology before establishing the operational and financial case.

“The term “grid partner” must not become an attractive label attached to a business case that does not work.

“Businesses can play an important role in balancing a more renewable electricity system, but they cannot be expected to absorb unnecessary complexity or invest against unpredictable commercial signals. Participation has to work for the business as well as for the grid.

“DESNZ, Ofgem and NESO have an opportunity to simplify their own task. Giving priority to strategic, shared-infrastructure projects such as GigaParks, a model that any developer can replicate rather than a one-off, is a direct way to cut the connection queue and separate viable projects from the “zombie” applications currently clogging it.

“More broadly, government should not shy away from leading the design of this new energy model, the Internet of Energy, and positioning the UK at the forefront of a sector set to attract trillions of pounds of global investment over the next decade.

“The Treasury and the Department for Business and Trade have a direct interest in this too. Infrastructure that removes the grid as a barrier to entry will determine whether the next wave of data-centre and industrial investment lands in the UK or goes elsewhere.

“The UK does not lack technology, private capital or appetite. What it needs is a simpler and more predictable route from investment to delivery. Get those conditions right and smart power can reduce costs, strengthen business resilience and support a cleaner, more secure electricity system.”

Stefano D.M. Sommadossi, Founder & Chief Executive Officer, NatPower UK.

A serial entrepreneur with 25+ years of experience in sustainability investments and cutting-edge digital services in four continents. Currently leading NatPower UK, NatPower Marine, and Sea.Energy. Founded and led NextEnergy Capital (sustainable energy fund manager, currently $3.0bn+ AUM), WiseEnergy ($5.6bn in clean energy assets operated), Nexcentrica (sustainability business incubator and investment company), and four energy transition companies in clean energy generation and electric vehicle charging services. Founder and trustee of Evogea, a foundation that incubates and finances high-impact sustainable projects to turn communities into leaders in sustainability and the green economy. Former advisor to governments, fund managers, lenders and large listed corporations in sustainable investments, energy, marine, automotive, telecoms and media. Former contract professor at Bocconi University and Senior Partner at The European House Ambrosetti.