This article was published in our nuclear briefing, Nucleus.
In new nuclear development, the choice between freehold and long leasehold is often framed as a tension between investment appeal and operational control.
Freehold carries obvious presentational advantages, particularly when seeking international investment, and is the form of ownership most readily recognised by nations and their financial institutions. It looks like absolute ownership, gives an indefinite right to use and occupy the land, and is often presented as the optimal position for attracting investment. But there is a trade-off. Where neighbouring land uses, such as existing nuclear plants or data centres, create operational and regulatory sensitivity, it can be critical to retain control over compliance with restrictive and positive covenants.
That is the real issue. In nuclear projects, the land structure is less about what sounds most complete in theory and more about what remains enforceable through construction, operation, defuelling and decommissioning.
A long lease, also known as a virtual freehold, can be underestimated in this debate. In a new nuclear context, it is not merely a right to occupy land for a very long period - it is a governance tool. Controls on alienation, occupation, use, repair, access and compliance sit in the primary instrument tied directly to the leasehold estate itself. That matters where one site must remain a safe and predictable neighbour to another, or where a legacy or adjoining station still has licensing, revenue or decommissioning sensitivities of its own. The lease structure also better preserves the concept of a single accountable neighbour, even if there are occupational interests beneath it.
A freehold route can, with robust drafting, replicate many of those protections, but usually only by spreading them across a more elaborate suite of documents: the transfer, deeds of covenant, title restrictions, indemnities, options, guarantees, financing interfaces and, in some cases, priority and intercreditor arrangements.
More moving parts create more points of weakness. Future dealings with the land become harder to police. Restrictive covenants can lose their enforceability. Multiple documents must be read together; some are not registrable and are prone to being lost. Record-keeping becomes critical. Some protections depend on the creation of contractual chains rather than proprietary rights; positive covenants do not run with land in English law, meaning they do not automatically bind successors. Deeds of covenant giving rise to enforceable contractual relations must be used to preserve the original intent.
Insolvency emerges as the clearest dividing line between the two tenures. The freehold structure can be more problematic in insolvency than a leasehold transaction from the perspective of those seeking to protect the rights and obligations enshrined in the title. A lease can be disclaimed by a liquidator but in a nuclear context that is unlikely once the plant has gone critical. The path to resolution likely drives the insolvency practitioner down the special administration route, preserving the leasehold instrument. A freehold can also be disclaimed but the consequences for a nuclear plant are deeply uncertain and unprecedented. Such issues go to the heart of how a nuclear land deal behaves under stress.
Deeds of covenant and indemnities may be exposed to disclaimer as onerous contracts. A call option to recover land for decommissioning may be vulnerable to insolvency or to enforcement of security. Funders may resist restrictions that limit their disposal options, although security is likely to be taken higher up the corporate stack. A single-asset project vehicle can intensify those concerns, particularly where regulated support, however robust, may not answer every question for the full life of the asset, through to ultimate land clearance and de-licensing for the next non-nuclear use.
The decommissioning point is especially important because nuclear land planning is unusually durationally long-tailed, particularly if there has been some unexpected contamination event. Under a leasehold model, the parties can manage that uncertainty more cleanly because the tenant’s rights and the landlord’s reversion sit in one structure. Under a freehold model, the answer may depend on a future right to call land back, perhaps for nominal consideration, at a time when tax, valuation, security enforcement and the precise footprint required may all be contested. That is workable in principle, but harder to make resilient in practice. Equally, depending on the wider context, a landowner may not in fact want the land back!
The distinction between licensed and offsite land also deserves attention. Site licence controls and Office for Nuclear Regulation (ONR) consent requirements provide meaningful protection for licensed land. But not every strategically important parcel will necessarily be licensed. Offsite land may still matter for access, coastal protection, shared infrastructure or the operational integrity of an adjoining operation. In that setting, private law controls do more of the heavy lifting. A freehold structure can impose restrictive covenants and transfer controls, but these are unusual in a freehold context, making them potentially unattractive to funders and more cumbersome to administer than lease-based controls.
None of this means freehold should be dismissed. Many issues are broadly neutral or largely drafting-dependent between the two alternatives. Positive covenants can be supported by deeds of covenant and title restrictions. Restrictive covenants are orthodox in principle, even if later challenge to their enforceability can never be ignored entirely. They endure for so long as they protect an interest in the land with the benefit of that restriction, and isn’t that what matters?
Nuclear liability and contamination migration are not points of real divergence between the two structures, given the tight statutory regulation of those issues and the fact they are not ultimately determined by reference to the proprietary interests of the polluter.
For any new nuclear project, a project-specific residual risk analysis is essential. It should be prepared against the background of the adjoining-site relationship, the specific stakeholder arrangements, the financing model (RAB or other incentive-based instruments), and typically the unknown identity and risk appetite of future investors. The fair, market-facing conclusion is not that freehold is unsuitable for new nuclear, but that it demands more from the document architecture and leaves less margin for error if the project or its stakeholders run into difficulty. Correcting project inception assumptions, such as land boundaries, is harder under a freehold structure.
For many new nuclear developments with adjoining sensitive land use, long leasehold has the stronger claim as the more resilient approach. Freehold may look simpler from the outside. In reality however, nuclear projects test land structures over decades, not at signing. The better structure is the one that keeps the protection package whole when ownership changes, financing is enforced, licences evolve and decommissioning finally arrives. On that test, control often matters more than the rhetoric or vanity of ownership.
