National Grid’s $1.75bn Joulent investment brings AI data centre power into its growth portfolio

National Grid plc, listed on the London Stock Exchange under the ticker NG., has agreed to invest $1.75 billion for a 35% stake in Joulent LLC through its commercial arm, National Grid Ventures. The transaction gives National Grid exposure to dedicated power infrastructure for data centres and other large electricity users whose requirements are developing faster than conventional grid connections can accommodate. Joulent’s initial project is the 2.67 GW Project Kilby in West Texas, being developed with Chevron Corporation to supply a Microsoft Corporation-operated data centre under a 20-year power purchase agreement. The investment expands National Grid’s addressable market beyond its regulated network businesses and into contracted generation serving artificial intelligence infrastructure. The central tension is whether predictable long-term demand can compensate for the construction, financing and minority-governance risks associated with entering a rapidly scaling but still developing power model.

Why does National Grid’s Joulent investment represent more than another United States power project?

The strategic importance of the Joulent transaction lies in the type of electricity demand National Grid is pursuing. The company is not simply adding another conventional generation asset to National Grid Ventures. It is positioning itself within the emerging market for dedicated power systems designed around artificial intelligence data centres, cloud infrastructure and other large industrial loads.

Electricity demand from these customers is unusually concentrated, time-sensitive and continuous. A conventional utility connection may require years of transmission upgrades, substation development, interconnection studies and regulatory approvals. Data centre developers, however, are often making multibillion-dollar technology investments on schedules that cannot wait indefinitely for the wider grid to catch up.

Joulent’s proposed model combines co-located generation, battery storage, renewable integration, electrical infrastructure and what the company calls Across-the-Meter connections. The commercial proposition is to supply power close to the customer while retaining a longer-term pathway to grid connection and potential electricity exports.

For National Grid, the attraction is that this model uses capabilities it already understands. These include high-voltage infrastructure, system balancing, project development, complex engineering and long-duration asset management. Yet it allows the company to apply those skills to electricity demand growth that sits outside the traditional regulated utility investment cycle.

That distinction matters because regulated networks typically earn returns approved by regulators and recovered from a broad customer base. Dedicated data centre infrastructure can provide higher growth and potentially attractive contracted cash flows, but the risks are more concentrated around particular projects, counterparties, construction schedules and technology customers.

National Grid is therefore not abandoning its regulated model. It is using National Grid Ventures to add a selective commercial growth layer alongside it.

How does Project Kilby convert artificial intelligence electricity demand into contracted infrastructure revenue?

Project Kilby is the foundation of the Joulent investment case. The West Texas facility is expected to provide approximately 2.67 GW of electricity through a phased and modular development structure. Joulent holds the project through a 50-50 partnership with Chevron Corporation, while the output is intended for a Microsoft Corporation-operated data centre campus under a 20-year power purchase agreement.

The long-term contract materially changes the project’s risk profile. A new gas-fired power facility developed for merchant electricity markets would be exposed to wholesale power prices, fuel spreads, dispatch frequency and competition from other generators. Kilby instead has a named customer with a very large and persistent electricity requirement.

The 20-year term potentially creates the kind of contracted revenue visibility that infrastructure investors prefer. It can support project financing, procurement commitments and long-term operating planning, provided the contract contains commercially viable pricing, fuel-cost treatment, availability obligations and credit protections.

Those details have not been disclosed publicly. The existence of a power purchase agreement therefore improves bankability, but it does not by itself reveal the expected return on National Grid’s investment.

Microsoft’s participation is nevertheless strategically significant. Data centre projects need confidence that power infrastructure will be available when computing capacity is ready. Power developers need confidence that the customer will require the contracted electricity for long enough to justify substantial upfront capital expenditure. A long agreement between the project and a major technology company helps align those two investment cycles.

Project Kilby is described as being at an advanced stage, with critical equipment secured, including GE Vernova turbines, and engineering, procurement and construction capacity reserved. First power is targeted for 2028, while a final investment decision is expected during 2026.

This means the project is considerably more advanced than an early-stage concept, but it is not yet an operating asset. Equipment reservations reduce supply-chain uncertainty, particularly when turbine manufacturing capacity is constrained, but they do not remove permitting, construction, commissioning or cost risk.

The next major value transition will occur when the partners take the final investment decision and provide greater clarity on total project cost, funding structure and construction milestones.

What does the 35% Joulent stake give National Grid, and which risks remain outside its control?

National Grid’s position is a strategic minority investment rather than a full acquisition. A 35% shareholding gives it meaningful economic exposure and potentially substantial influence, but Joulent will remain an independent company.

That structure allows National Grid to participate without assuming the entire funding requirement or development risk of Joulent’s platform. It also brings together complementary partners. Joulent contributes its development pipeline and large-load power model, Chevron contributes energy and project capabilities, and National Grid contributes infrastructure and electrical-system expertise.

The minority structure also limits control. Project decisions, future fundraising, development priorities and capital allocation will depend on shareholder agreements and governance arrangements that have not been disclosed in detail.

National Grid’s interest in Joulent should also not be confused with direct 35% ownership of Project Kilby. Joulent itself participates in Kilby through a 50% partnership with Chevron. National Grid’s economic exposure to the project is therefore mediated through its minority ownership of Joulent and the project-level partnership.

This layered structure can be efficient because it spreads capital and expertise across several parties. It can also make financial outcomes harder for investors to model. Cash generated at the project level may first be used for debt service, operating costs and reserves before distributions reach Joulent and ultimately National Grid.

The $1.75 billion figure requires similar caution. National Grid has agreed to invest that amount for the stake, but the announcement does not provide enough information to determine whether all capital will be transferred immediately, contributed in stages or linked to development milestones.

It would therefore be misleading to treat the commitment as an instant cash outflow or to divide $1.75 billion by 35% and assume the result represents a simple, independently validated valuation of Joulent. The economic meaning will depend on transaction terms, funding timing and whether part of the capital is earmarked for Project Kilby or the wider pipeline.

Can dedicated data centre generation ease grid delays without transferring costs to existing consumers?

One of Joulent’s central claims is that dedicated infrastructure can serve large new loads without placing the full cost of their growth onto existing electricity customers.

That issue is becoming increasingly important. Large data centres can require electricity volumes comparable with industrial complexes or small cities. When those loads connect directly to public networks, utilities may need to construct new substations, transmission lines, generation resources and reliability infrastructure.

The allocation of those costs can become politically and commercially contentious. Households and existing businesses may resist paying higher network charges to support infrastructure primarily required by technology companies.

Co-located generation offers one response. By developing power close to the data centre, Joulent could reduce dependence on immediate large-scale grid expansion. Battery systems and advanced controls could help manage fluctuations, maintain reliability and coordinate future interaction with the wider network.

However, dedicated power is not independent of the wider energy system. Project Kilby will still require fuel supply, permits, water or cooling arrangements, emissions compliance, land, transmission coordination and emergency planning. A facility of 2.67 GW is major infrastructure even when most of its electricity is consumed by one nearby customer.

The project’s gas-fired generation also creates a long-duration exposure to natural gas availability, price structures and environmental policy. A 20-year customer agreement reduces demand risk, but the commercial protections around fuel costs will be important. If fuel-price movements are passed through to the customer, project margins may be more stable. If the project retains significant commodity exposure, returns could vary more widely.

Renewable generation and battery storage may improve flexibility and reduce the emissions intensity of the overall power solution, but National Grid has not disclosed the eventual technology mix or the percentage of Kilby’s output expected to come from each source.

The project’s success will therefore be judged not only by speed. It will also be judged by reliability, cost competitiveness, emissions performance and whether its infrastructure model genuinely limits cross-subsidisation by existing grid customers.

Does National Grid have enough balance-sheet capacity for Joulent alongside its £70 billion investment programme?

National Grid enters the transaction from a position of strong operating growth but unusually high capital intensity. For the year ended March 2026, the company reported underlying operating profit of £5.7 billion, up 9% at constant currency, while underlying earnings per share rose 8% to 78 pence. Capital investment increased 21% to £11.6 billion, and group asset growth accelerated to 10.9%.

The financial results demonstrate why National Grid can consider an investment of this size. Its regulated businesses generate substantial recurring operating cash flow, supported by long-lived electricity and gas network assets.

The balance sheet is nevertheless carrying the weight of an expanding infrastructure programme. Net debt reached £44.2 billion at the end of March 2026, compared with £41.4 billion a year earlier. Cash generated from operations increased to £7.9 billion, but capital expenditure, dividends and financing costs resulted in a £6 billion net cash outflow before disposals and other adjustments.

National Grid has committed to invest at least £70 billion across its businesses during the five years to March 2031. The company said the Joulent investment would be incremental to that programme and funded through available balance-sheet headroom.

That wording is important. Joulent does not appear to be replacing investment in regulated networks. It represents additional capital allocation into a commercial opportunity.

The strategic logic is understandable, but investors will expect discipline. Regulated network spending comes with relatively visible return frameworks. Commercial investments need to justify the higher uncertainty through better growth, contracted economics or diversification benefits.

The Joulent commitment represents a manageable percentage of National Grid’s overall investment capacity, but it is large enough to require clear reporting. Investors will need to see whether returns from Joulent compensate for the additional debt, equity exposure and non-regulated risk.

The transaction becomes more compelling if National Grid can use one platform investment to participate in several contracted projects rather than relying entirely on Kilby. Joulent has identified a multi-gigawatt future pipeline, but those projects remain potential opportunities rather than current operating assets.

What are National Grid shares signalling about the Joulent strategy and wider capital demands?

National Grid shares traded around 1,231 pence during the July 21 session, down approximately 0.4% from the previous close. The stock was about 1.4% lower than its July 14 close, while remaining roughly 0.8% above its June 22 level. Its 52-week range stood between 1,000.5 pence and 1,428.5 pence, placing the shares around 14% below their annual high and approximately 23% above the low.

The market performance points to broadly neutral near-term sentiment rather than a decisive rerating around Joulent. National Grid shares declined on July 1 when the investment was announced, although that movement cannot be attributed solely to the transaction because utilities are also influenced by bond yields, political developments, regulatory expectations and wider market conditions.

Investor caution would be understandable. Joulent expands National Grid’s growth exposure, but it also arrives when the company is already undertaking the largest investment programme in its history and carrying more than £44 billion of net debt.

The shares’ position above the 52-week low suggests investors still assign value to National Grid’s asset growth, dividend profile and regulated earnings. Their discount to the annual high indicates that the market is not treating infrastructure growth as risk-free.

The Joulent investment is unlikely to change earnings materially before Kilby reaches construction and begins delivering power. Near-term valuation will therefore remain more closely linked to National Grid’s regulated performance, interest costs, capital delivery and regulatory settlements.

Over time, Joulent could become a more significant valuation driver if it converts its project pipeline into contracted operating assets. Until then, the stake is better viewed as a strategic growth option supported by one advanced anchor project.

Which milestones will prove whether National Grid’s AI power strategy can create durable returns?

The first milestone is the final investment decision for Project Kilby. That decision would confirm that the partners are prepared to commit full construction capital based on the project’s contracts, permits, cost estimates and expected returns.

The second is disclosure of the funding structure. Project-level debt could limit the amount of equity required from Joulent and its shareholders, but interest costs, covenant requirements and construction guarantees would influence the economics.

Permitting and site progress will provide the next evidence. A large West Texas generation complex requires coordinated approvals, fuel infrastructure, electrical systems, environmental compliance and construction sequencing.

Equipment delivery will also matter. Securing turbine capacity is commercially valuable, but the project must still receive, install, integrate and commission those units on schedule.

Microsoft’s data centre construction must advance in parallel. Dedicated power infrastructure creates value only when the customer’s computing campus is ready to absorb electricity under the contracted timetable.

The 2028 first-power target will be the clearest operational test. A phased launch could begin revenue generation before the full 2.67 GW is completed, potentially reducing the period during which capital is invested without corresponding cash flow.

Beyond Kilby, National Grid will need to show that Joulent’s wider pipeline is real, differentiated and capable of securing customers on similarly robust contracts. One successful project would validate the platform. Several successful projects could make the investment strategically important within National Grid Ventures.

The opportunity has improved because National Grid has secured exposure to a named customer, an advanced project and experienced partners. What remains unresolved is whether the platform can deliver large-scale power quickly without compromising capital discipline.

The investment case would strengthen with a final investment decision, transparent project financing and visible construction progress. It would weaken if costs rise, the timetable slips or the future pipeline remains dependent on uncontracted proposals.

National Grid’s next proof point is therefore not another announcement about artificial intelligence demand. It is evidence that Project Kilby can move from contracted ambition to financed construction and first power in 2028.

What should investors take away from National Grid’s $1.75 billion Joulent investment?

• National Grid Ventures has agreed to invest $1.75 billion for a 35% minority stake in Joulent LLC.
• Joulent’s anchor asset is the 2.67 GW Project Kilby power facility in West Texas.
• Project Kilby is being developed through a 50-50 partnership between Joulent and Chevron Corporation.
• The project will supply a Microsoft Corporation-operated data centre under a 20-year power purchase agreement.
• Critical turbine equipment and engineering capacity have been reserved, but the final investment decision remains expected during 2026.
• The investment gives National Grid exposure to artificial intelligence-driven electricity demand outside its regulated utility businesses.
• National Grid reported £5.7 billion of underlying operating profit and £11.6 billion of capital investment for FY26.
• Net debt increased to £44.2 billion, making capital discipline and project-level funding important to shareholder returns.
• National Grid shares remained close to their level one month earlier as of July 21, suggesting investors are awaiting operating evidence.
• Final approval, financing, construction progress and first power in 2028 are the next measurable tests.