Power policy
Policymakers globally are grappling with constrained power supply, with data centres competing for capacity against factories, offices, hospitals and homes.
Recent reforms in the UK, for example, have sought to address the significant backlog of grid connection requests by introducing the “Connections Reform”. A key tenet is the introduction of gateways, set by the National Energy System Operator (NESO). To secure a position in the queue, a project needs to meet certain Gate 2 criteria to demonstrate it is sufficiently “ready”.
Distribution network operators (DNOs) have also been granted new “Connection Milestone Management” powers. Often referenced as ‘use-it-or-lose-it’ provisions, DNOs can monitor development progress against agreed milestones and ultimately terminate connection agreements for stalled projects to free up capacity. This makes land speculation much more challenging, given the need to continue ordering long lead-time equipment and investing in connection works to demonstrate progress.
The European Union is addressing similar challenges. In December 2025, the European Commission announced the European Grids Package to reform Europe’s power grid and reduce supply bottlenecks. One core policy document is the ‘guidance on efficient and timely grid connections’ (C/2025/6703). The guidance suggests network operators adopt processes to rationalise the allocation of capacity (e.g. by using ‘first ready, first served’ or ‘milestone-based, milestone-enforced’ prioritisation frameworks). It also considers prioritising certain energy user groups based on ‘economic criteria’ and ‘policy objectives’ (including the deployment of data centres). The detail is left for EU Member States to implement, and it will be important to keep abreast of developments.
Spain, for example, has some of the strictest rules, including development milestones that must be met within specific deadlines following the grant of a grid access permit (with the final milestone being the execution of an access agreement for at least 50% of the allocated capacity within five years).
In addition, capacity reservation payments apply from the grant of the grid access permit until the facility becomes operational. Further regulatory requirements are expected in relation to renewable energy sourcing, usage and generation, as well as reporting metrics such as Power Usage Effectiveness (PUE), water usage effectiveness and energy reuse factors.
The interaction between ‘use-it-or-lose-it’ regulatory regimes and the broader business plan is critical: satisfaction of the relevant milestones needs to dovetail with planning, permitting and the broader construction and development plan.
Portability and cost
Connection structure and portability
Another key consideration is the identity of the contracted applicant. If the land is transferred, will the purchaser automatically benefit from the connection agreement, or will it need to be assigned or novated? Any consent process may cause delay and introduce requirements for guarantees or other forms of surety to support delivery. It is therefore preferable, where possible, to address the portability of power at the outset when structuring the project.
Multi-building sites and power distribution
Multiple data centre buildings may be constructed on a single site. Where these are disposed of separately, security of power becomes critical.
If separate connections are not feasible, common ownership of a shared substation may be required. Alternatively, plots may rely on onward distribution of power across the site – although this can introduce additional regulatory considerations. In certain jurisdictions, a contractual obligation to supply power, even under a private onward-supply arrangement, may bring parties within regulatory regimes intended for licensed energy distributors.
Cost considerations
The current and forecast cost of power in relevant geographies will also be key to assessing viability. Legal, commercial and technical advice should be sought to obtain market comparisons and understand the availability of local subsidy regimes, as these will be relevant considerations for prospective occupiers and factors that ultimately go to marketability.
Planning and permitting
After power, permitting is the next biggest development hurdle. Regimes vary significantly between jurisdictions, and sometimes even between local councils or municipalities. What tends to be a common theme, however, is a byzantine process requiring local expertise.
As data centres become more visible to the public, governments and planning authorities are giving increased scrutiny to permitting decisions. This creates both opportunities and challenges.
On the one hand, policymakers recognise the role of data centres in supporting economic growth and technological advancement. The UK Government’s “Delivering AI Growth Zones” paper, for example, sets out a broader vision for technology-led growth. Planning reforms such as the Infrastructure Planning (Business or Commercial Projects) (Amendment) Regulations provide a pathway for certain data centre projects to be treated as “nationally significant infrastructure”, potentially allowing them to proceed via a national planning process.
However, the scope of these reforms remains to be clarified, and their practical impact will depend on implementation.
At the same time, national strategy does not guarantee smooth local delivery. There are increasing instances of local authorities refusing consent for data centre developments, often driven by community concerns or competing demands for land and power (such as residential development). Even where approvals are granted, they may be subject to detailed and sometimes onerous conditions relating to construction and operation – including noise, water usage, cooling systems, biodiversity and flood risk.
We are also seeing increasingly stringent requirements in areas with high data centre density, such as mandatory renewable energy usage levels and obligations to reuse waste heat in local communities.
Understanding the permitting status of a site, as well as the broader regulatory environment and associated compliance costs, is therefore central to any valuation or investment decision.
How is powered land valued?
Powered land in Europe can broadly be categorised into four distinct market types, defined more by risk and liquidity than geography, each of which is priced differently:
(1) FLAPD markets (Frankfurt, London, Amsterdam, Paris and Dublin): At one end of the spectrum sit the core FLAPD markets, where power is severely constrained, demand is high, and assets are treated as quasi-infrastructure. Here, powered land can command €800k - €1.4m per megawatt (MW) (grid) and residual value is firmly assumed in any valuation.
(2) Growth markets: Next are “growth” markets such as Milan, Madrid or Warsaw, where hyperscale demand is there, but is less proven, and power is more available. These are typically underwritten at €400k – €900k per MW, with more reliance on pre-leasing to unlock value.
(3) AI / power-led markets: This third category comprises geographies where land and power are abundant but location (proximity to demand – and low latency) is less critical, for example the Nordics, parts of Spain and the north of the UK. Here, investors often assume little or no residual value and apply higher return thresholds. Powered land here will trade more like a development optionality at €200k – €700k per MW until risk is removed.
(4) Edge or latency-driven urban markets: This final category sits slightly apart — with smaller sites embedded in cities, valued more like real estate with power overlay than scalable campuses. The same headline MW is priced very differently depending on how much demand certainty, power security and exit liquidity the market provides.
The timing of power availability is a key determinant of value, with a premium attaching to sites capable of delivering capacity in the near term (typically 2028–2029). This reflects acute capacity constraints facing hyperscalers, who are actively seeking to secure medium-term growth pipelines.
By contrast, power availability beyond 2030 carries additional risks, including changes in technology (particularly AI hardware efficiency), shifts in demand, evolving regulatory frameworks and the emergence of new supply in alternative markets. As a result, longer-dated sites are often heavily discounted.
Power availability post-2030 may be priced more akin to logistics or agricultural land than a near-term data centre opportunity. A megawatt delivered in 2028 can be materially more valuable than one delivered in 2032.
Ultimately, powered land is no longer simply a real estate play – it is a function of timing, power certainty and regulatory execution. As a result, value is increasingly determined not just by location, but by how quickly and credibly power can be secured and delivered.
Capital and control
Before they can consume power, data centres have a significant appetite for capital. Powered land strategies often lend themselves to joint venture and co-investment structures, where one party contributes land and another provides capital, with development and operational expertise sitting with one or both parties. Development costs are typically drawn down in tranches to ensure efficient capital deployment. The parties will need to decide how this capital is called – typically, but not always, contributions will be required in proportion to the parties’ equity contributions.
The structuring of the capital stack is also important. For example, preferential equity could be used to give holders a priority return. A real estate investment-style carried interest or “promote” structure may be used, for example where one party also has development or operational management responsibilities, with the potential to significantly increase value. Where multiple sites are undergoing development, promote might be measured portfolio-wide or at an asset-specific level (possibly with clawback where an underperforming site drags the Internal Rate of Return (IRR) back below the hurdle).
Where investors join forces to execute a powered land strategy, it’s also important to align at the outset on the ultimate exit strategy. In an ideal world, the platform will permit liquidity at a building, campus and platform level in order to maximise flexibility and value realisation. Designing this flexibility at the outset requires collaboration across all workstreams – real estate, legal, power and regulation, structuring, planning and permitting.
In Part 1 and 2 of this series, we introduced powered land as an emerging asset class and the policy, planning and regulatory landscape shaping data centre development. In the final article, we will examine how delays to grid connection are causing developers to consider alternative power solutions, namely co-located power generation and private wire networks, and the risks and challenges of these arrangements from a financing, regulatory and structuring perspective for developers, investors and lenders.
Written by Peter McCabe and Jack Shand.
