A data centre can have a completion date without having a date for running customer workloads. The distinction is not a technicality: the thing being delivered may be a building for an occupier to fit out, rather than an operating computing facility.
Before treating a completion announcement as evidence of new computing capacity, ask: what was completed, who is responsible for the next part, and which building or phase does the statement cover?
Manor Farm and Wapseys Wood illustrate two different boundaries. At Manor Farm, the issuer separates the powered-shell development from the occupier’s internal fit-out. At Wapseys Wood, the developer distinguishes the first building becoming operational from its subsequent ramp-up and the wider campus coming into use. Both are useful distinctions even while the dates remain forecasts.
Manor Farm: whose completion date is it?
Tritax’s H1 2026 results statement forecasts practical completion at Manor Farm in Q2 2028. Elsewhere in the same document, it describes the proposed letting on a powered-shell basis: the occupier would be responsible for the building’s internal fit-out, while full rent would commence upon practical completion. [1]
Those statements describe different responsibilities. The developer’s practical-completion milestone is not presented as the completion of the occupier’s internal fit-out. It would therefore be wrong to rewrite Q2 2028 as a verified date for live customer workloads.
The commercial qualification matters too. In the retained statement, Tritax says the pre-let is in solicitors’ hands with a potential occupier. We have not inspected an executed lease or independently verified its completion and rent provisions. The rent description is the issuer’s account of the proposed arrangement, not evidence that rent has started or that the occupier is operating.
The distinction also explains why a property milestone should not be used as a shortcut for a computing milestone. Even evidence of rent commencement would answer a different question from evidence of live IT workloads. That is not a criticism of the development model. It is a reason to identify whose delivery obligation a date describes.
The same report states power-delivery tranches of 107 MW in 2027 and a further 40 MW in 2029. Those are attributed supply statements, not measurements of operating IT load. Their dates should not be substituted for the shell-completion date, nor should any of the three dates be treated as proof of completed customer fit-out. [1]
Wapseys Wood: first operation is not the whole campus
The SDC M40 Campus planning timeline reviewed on 1 October envisages submission of a Development Consent Order application in spring 2027, construction beginning in 2028 if consented, and the first data centre becoming operational from approximately 2031, also conditional on consent. These are developer targets, not completed events. [2]
The accompanying campus description makes the delivery boundary more explicit. The first data centre is expected to build up to full capacity over one to two years, while the wider campus would come into operation in phases as each building is completed. [3]
“The campus will be operational in 2031” would lose that distinction. It would turn an approximate first-building target into a whole-campus claim and omit the stated ramp-up. The one-to-two-year period should not be converted into a fixed date for full-campus operation either: the statement describes the first data centre, not the delivery of every building.
Even a future, well-supported report that one building had opened would need that boundary attached. It would not automatically establish that every planned building, or every advertised megawatt, was available to customers.
Ask what each milestone establishes
Rather than compressing delivery into a single “complete” field, keep the following questions separate. They are prompts for reading evidence, not universal contractual definitions or a fixed sequence that every project follows.
- Shell or practical completion
- Which works and contractual scope are described as complete? Does the statement include or exclude the occupier’s internal fit-out?
- Fit-out
- Whose equipment and systems are covered, for which halls or buildings? Is the source describing planned work, installation or an accepted handover?
- Energisation
- What connection or equipment was energised, when, and within what supply boundary? Does that establish anything about customer workloads?
- Commissioning
- Which systems and tests does the claim cover? Who reports the outcome, and what remains outside that scope?
- Operation and ramp-up
- Does the source describe a first service, one hall, one building or the full campus? What evidence supports the claimed operating capacity?
These milestones can overlap or apply separately to different phases. This is not a readiness score, certification checklist or automatic ladder from planning to operation.
The word “commissioned” needs the same care as “completed”. A claim about tested equipment should retain its equipment and test boundary; it should not silently become a claim that the entire campus is serving customers. Equally, an opening announcement should not be turned into a measurement of utilisation unless it provides evidence for that measurement.
Preserve the forecast alongside the milestone
A useful delivery record needs more than a date. It should identify the source, the milestone, the physical and contractual scope, whether the date is a forecast or a reported event, and any stated dependencies.
For Manor Farm, the bounded statement is that the issuer forecast practical completion in Q2 2028 within a powered-shell model that assigns internal fit-out to the occupier. For Wapseys Wood, it is that the developer targeted first-building operation around 2031 if consented, followed by a stated ramp-up and phased campus delivery. Neither statement establishes current operation.
If later evidence reports a completed milestone, preserve the earlier forecast and record what the later source actually establishes. Do not overwrite a first-building event as “campus complete”, or compare a shell forecast with a customer go-live date and call the difference a delay to the same deliverable.
Where independent evidence is missing, say so. Lack of verification is not proof that work has not happened. The purpose is not to downgrade projects; it is to avoid claiming more delivery than the evidence supports.
A completed shell can be a substantial delivery milestone without being an operating data centre. The useful question is not simply whether a project is “done”, but what has been delivered, for whom, within which boundary—and what the source leaves unresolved.
For the related power distinctions, read What does “power secured” actually mean for a data centre? Start with the research guide for the wider series.
Sources and scope
This selective analysis uses evidence retained by 1 October 2026. No fresh external source review or project observation was performed for this draft. The examples describe attributed forecasts and delivery models, not our independent verification of completion, energisation, commissioning or operation.
- Tritax Big Box, H1 2026 results statement (PDF), page 6 for the Q2 2028 practical-completion forecast; page 17 for the powered-shell model, occupier fit-out responsibility, proposed rent commencement, pre-let description and stated power tranches. Cover/approval dated 5 August 2026; exact first-publication day not established. Issuer statements, not independent inspection of contracts or works.
- SDC M40 Campus, The planning process, Timeline, as reviewed on 1 October 2026. Conditional application, construction and first-operation targets; exact webpage publication/revision date not established.
- SDC M40 Campus, The data centre campus, Operations, as reviewed on 1 October 2026. First-building ramp-up and phased campus delivery are developer expectations; exact webpage publication/revision date not established.
Report files are not reproduced here. The two SDC pages are accounts from the same developer, not independent corroboration. This article is not engineering, legal or investment advice.