Data Centre MEP Cost Management: Why M&E Is the Project

On a hyperscale or colocation data centre, power, cooling, containment, fire suppression and communications infrastructure can account for 55–70% of construction cost. Within a services-intensive fit-out, the proportion can approach 80%. These are indicative ranges, not a universal benchmark: land, shell, utility connections, IT equipment, redundancy and the definition of construction cost can change the comparison materially. The central point remains: M&E is not a subcontract to manage at arm's length. It is the principal commercial exposure.

A small percentage movement in the electrical package can outweigh an apparently substantial saving elsewhere in the building. Cost control therefore needs to follow the engineering decisions, not merely report the financial result after equipment has been ordered. Capacity, resilience, phasing and operating requirements must be visible in the cost plan from the start.

The procurement challenge — speed vs competition

Data centre programmes are compressed by customer commitments, power availability and the commercial value of bringing capacity online. Main contractors may be sole-sourced or selected from a shortlist of two. The specialist market for critical power and precision cooling is narrower still, particularly when several large projects compete for the same production slots and installation teams.

Limited competition reduces price tension, while programme pressure encourages award before the Scope is complete. A headline package price can therefore be less reliable than it looks. Review exclusions, currency assumptions, escalation, factory testing, delivery terms and commissioning obligations alongside the tender total. Record which design information underpins each commitment and what changes require a price adjustment.

Early engagement can protect the programme without abandoning commercial discipline. Use open-book reconciliation, targeted specialist enquiries and a clear process for converting early orders into the final package. Separate the price of securing manufacturing capacity from the price of the fully coordinated installation.

Critical power — where the money is

Generator sets, UPS equipment, HV/LV distribution, PDUs and busbar trunking are among the largest MEP cost items. Large transformers and generators may require 40–52 weeks or longer in constrained markets, although current supplier quotations must determine the actual programme. Waiting for every detail to be resolved before procurement can make the completion date impossible.

The commercial consequence is that equipment is often purchased against outline specifications. Later changes to redundancy, fault levels, protection, harmonic performance or maintainability then affect equipment already in manufacture. Variations can include redesign, abortive work, factory retesting and delayed delivery, not just the difference in catalogue price.

Before releasing an order, establish the design freeze points and the decisions still open. Track equipment quantities against usable IT load and resilience assumptions, not simply gross floor area. Check what factory acceptance testing, transport, lifting and energisation support the supplier includes. Those interfaces can carry significant cost even where the equipment price itself is fixed.

Cooling — the value engineering trap

Precision air, chilled water and adiabatic cooling systems are frequent value engineering targets because their capital costs are highly visible. But a cheaper cooling arrangement can alter power demand and PUE, water consumption, structural loads, space requirements and the commissioning sequence. A saving in one package can produce a larger cost in another.

Changing plant selection after structural or electrical design has progressed may require additional supports, revised distribution and new controls logic. Evaluate the total installed and operational effect, not just the plant tender. Account for resilience, part-load performance, maintenance access and the design conditions under which the proposed solution must work. Where future rack densities are uncertain, make the capacity assumptions explicit rather than hiding them in contingency.

A credible value engineering proposal includes the engineering assessment, cross-package cost reconciliation and programme impact. It should also identify who accepts any operational compromise. Otherwise the team may book a capital saving that disappears during integration or reappears as an owner-funded upgrade.

What effective MEP cost control looks like

• Specialist MEP QS involvement from RIBA Stage 2, aligned with capacity, resilience and phasing decisions.
• Independent cost plans benchmarked against genuinely comparable schemes, with location, date, scope and IT-load assumptions normalised.
• An equipment procurement strategy that identifies long-lead commitments, remaining design risk and supplier interfaces.
• Compensation Event or variation assessment throughout design development, with a live forecast of committed and unresolved change.
• Post-contract commercial management through integrated systems testing, commissioning and handover, not just mechanical completion.

The most useful report distinguishes the original budget, committed cost, assessed change, unresolved exposure and forecast final cost. It connects each movement to a decision the team can act on. That is how commercial management supports delivery rather than becoming a retrospective explanation of overspend.