on this page
Demand for data centre capacity is accelerating at a pace few sectors have experienced. Global demand for data storage and processing capability is expected to rise sharply over the coming decade, with many forecasts seeing current capacity doubling by 2030.
At the delivery level, that growth is colliding with a complex set of interlinked constraints including location and size of land, power availability, planning risks, sustainability and supply chain resilience.
With speed of delivery a key factor, demand for major contractors with proven track records continues to grow. At the same time, the pool willing and able to carry programme, performance and price risk at scale is narrowing as commercial pressures intensify. This concentration of demand and narrowing of effective delivery capacity is contributing to cost escalation, alongside rising materials and energy prices and the scarcity of specialist skills.
Even if site, power and planning constraints were resolved tomorrow this market structure would remain, placing upward pressure on delivery costs. The key question is how can organisations engage the construction market to deliver at scale, sustainably and at a market-appropriate cost.
The strain shows most clearly in the way projects are procured. Design and Build (D&B) remains the default delivery model for large data centre projects, offering a familiar single point of responsibility and a risk-transfer-based commercial framework. This approach is seen as particularly attractive for large, complex assets with high capital outlays, especially where projects are investor-funded and risk profile influences funder confidence.
However, when applied repeatedly at scale, the model shows its limits. Two-stage D&B procurement often narrows competition as designs mature and specialist, long-lead packages are defined. In response to heightened delivery risk, organisations gravitate towards a diminishing group of contractors able to carry programme and performance responsibility, reinforcing concentration rather than expanding capacity.
Owner Furnished Contractor Installed (OFCI) arrangements are also introduced alongside traditional D&B models to shorten procurement timescales, secure estate-wide systems and reduce price uncertainty. However, they often dilute D&B’s risk transfer. Responsibility for critical systems and interfaces becomes fragmented, leaving greater responsibility for time, cost and quality with the client, without expanding market capacity or sustaining the competitive tension needed to manage cost and programme escalation.
What this reveals is a mismatch between procurement models designed for individual, complex assets and a sector now seeking industrial-scale, repeatable delivery.
Rethinking risk, standardisation, and control
The core challenge of scale and scarcity requires a rethink of how risk, design and delivery are structured. For organisations delivering portfolios of data centres facilities can increasingly be reframed as functional, high-tech industrial or manufacturing assets rather than singular flagship projects creating a clear opportunity for greater standardisation through organisation-held reference designs and open technical protocols.
By standardising more of what is currently bespoke, organisations can retain greater control over design intent and performance, widen the pool of delivery partners and restore competition at scale, enabling repeatability across portfolios while bringing decision-making forward to be managed strategically and at lower cost. This is not about eliminating risk, but distributing it more intelligently to meet demand.
Unlocking Capacity
A more considered approach to risk, design and procurement also creates the conditions for construction to industrialise. Modern Methods of Construction (MMC) are often presented as a route to faster, cheaper and more certain delivery. In practice, however, without confidence in repeat demand and clearly defined interfaces, supply chains in the data centre sector cannot justify the investment required to industrialise.
Where organisations define common reference designs, technical interfaces and performance criteria at portfolio level, MMC can genuinely take hold. Repeatable pipelines allow manufacturers to invest, processes to stabilise, learning to be embedded and productivity to improve.
The potential benefits are substantial: shorter programmes, manufacturing-level quality and reduced pressure on site labour. However, MMC is not an innovation that can be applied in isolation; it is a consequence of how projects are procured.
An industry-level perspective
As data centre development accelerates, the question is whether the market is structured to deliver it efficiently at scale. Delivery costs are rising, driven by a scarcity of skills and a concentration of work in a narrow part of the supply chain. That scarcity is reinforced by procurement models ill-suited to industrial-scale, repeatable delivery. The result is a cost curve rising faster than many business cases can sustain.
This is why the case for greater standardisation is economically compelling. By holding more design intent, performance criteria and technical interfaces at portfolio level, organisations can widen the pool of delivery partners, restore competition, and plan, control and manage outcomes with far greater certainty across delivery programmes. In doing so, they create the conditions for repeatable, investable delivery models - including the effective use of Modern Methods of Construction and appropriate traditional procurement approaches, rather than relying on one-off solutions.
Meeting this challenge will require organisations to rethink how much control and responsibility they are willing to hold, and how they work with delivery partners over the long term. Taking greater ownership of design, interfaces and delivery strategy involves earlier commitment, The capabilities already exist within the market and are routinely deployed through long-term, collaborative partnerships, enabling organisations to realign supply and demand and move away from the economics of scarcity that currently constrain delivery.
The next phase of data centre expansion will be defined by who can organise delivery at scale. Those that do will be best placed not only to meet demand, but to strengthen the UK’s position as a competitive and investable home for the next generation of data infrastructure.
Please note that you are now leaving the AtkinsRéalis website (legal name: AtkinsRéalis Group inc.) and entering a website maintained by a third party (the "External Website") and that you do so at your own risk.
AtkinsRéalis has no control over the External Website, any data or other content contained therein or any additional linked websites. The link to the External Website is provided for convenience purposes only. By clicking "Accept" you acknowledge and agree that AtkinsRéalis is not responsible, and does not accept or assume any responsibility or liability whatsoever for the data protection policy, the content, the data or the technical operation of the External Website and/or any linked websites and that AtkinsRéalis is not liable for the terms and conditions (or terms of use) of the External Website. Further, you acknowledge and agree that you assume all risks resulting from entering and/or using the External Website and/or any linked websites.
BY ENTERING THE EXTERNAL WEBSITE, YOU ALSO ACKNOWLEDGE AND AGREE THAT YOU COMPLETELY AND IRREVOCABLY WAIVE ANY AND ALL RIGHTS AND CLAIMS AGAINST ATKINSRÉALIS, AND RELEASE, DISCHARGE, INDEMNIFY AND HOLD HARMLESS ATKINSRÉALIS, ITS OFFICERS, EMPLOYEES, DIRECTORS AND AGENTS FROM ANY AND ALL LIABILITY INCLUDING BUT NOT LIMITED TO LIABILITY FOR LOSS, DAMAGES, EXPENSES AND COSTS ARISING OUT OF OR IN CONNECTION WITH ENTERING AND/OR USING THE EXTERNAL WEBSITE AND/OR ANY LINKED WEBSITES AND ANY DATA AND/OR CONTENT CONTAINED THEREIN.
Such waiver and release specifically includes, without limitation, any and all rights and claims pertaining to reliance on the data or content of the External Website, or claims pertaining to the processing of personal data, including but not limited to any rights under any applicable data protection statute. You also recognize by clicking “Accept” that the terms of this disclaimer are reasonable.
The information provided by Virtua Research cited herein is provided “as is” and “as available” without warranty of any kind. Use of any Virtua Research data is at a user’s own risk and Virtua Research disclaims any liability for use of the Virtua Research data. Although the information is obtained or compiled from reliable sources Virtua Research neither can nor does guarantee or make any representation or warranty, either express or implied, as to the accuracy, validity, sequence, timeliness, completeness or continued availability of any information or data, including third-party content, made available herein. In no event shall Virtua Research be liable for any decision made or action or inaction taken in reliance on any information or data, including third-party content. Virtua Research further explicitly disclaims, to the fullest extent permitted by applicable law, any warranty of any kind, whether express or implied, including warranties of merchantability, fitness for a particular purpose and non-infringement.
The consensus estimate provided by Virtua Research is based on estimates, forecasts and predictions made by third party financial analysts, as described above. It is not prepared based on information provided by AtkinsRéalis and can only be seen as a consensus view on AtkinsRéalis' possible future results from an outside perspective. AtkinsRéalis has not provided input on these forecasts, except by referring to past publicly disclosed information. AtkinsRéalis does not accept any responsibility for the quality or accuracy of any individual or average of forecasts or estimates. This web page contains forward-looking statements based on current assumptions and forecasts made by third parties. Various known and unknown risks, uncertainties and other factors could lead to material differences between AtkinsRéalis' actual future results, financial situation, development or performance, and the estimates given here.