Demand for data centre facilities has reached record levels, driven by the explosive acceleration of AI computing and the continued growth of cloud computing. Recent forecasts project that global data centre capacity will need to double by 2030. That was underscored earlier this year when NVIDIA reported $1 trillion in orders for its next-generation GPUs. Hyperscale tech companies who are making these investments in GPUs require AI-ready digital infrastructure to be built on a massive scale and on an aggressive timeline. This will drive data centre demand that is projected to total $3 trillion in global data centre investment over the next few years.
This represents an enormous opportunity for investors. However, for investors to be successful they must successfully mitigate significant risks that exist in the market for digital infrastructure. Data centre developments are highly capital-intensive, creating significant financial exposure to investors. These developments also involve complex site selection, power acquisition, permitting processes and precise technical specifications. Due diligence, risk mitigation, and expertise in each phase of a development are therefore critically important – particularly in markets like London and large EU cities, which face major constraints for digital infrastructure.
HartDixon has extensive experience helping investors navigate these challenges in ways that position their data centre projects for success, whether that’s with greenfield data centre projects built from the ground up, or the redevelopment of legacy facilities into data centres that can support AI computing.
Risks and success factors for greenfield data centre projects:
For investors pursuing greenfield development of data centres, there are a number of complexities that must be navigated for each phase of a project. Due diligence, careful planning and deep expertise is required for these projects.
- Site acquisition challenges – There are significant site selection constraints for greenfield projects, not only in FLAP-D marketing, but also secondary markets in EMEA. Sites must have the right combination of electrical power, water supplies and access to fibre connectivity, and more. Competition for sites that meet the needs of advanced data centres can be fierce, driving up prices and risk. Extensive due diligence is needed to determine that a site can realistically meet the needs of an AI-ready data centre.
- Rapidly changing specifications for AI data centres – One of the most complex challenges facing data centre developers is ensuring that their site and design will be future-proof. This is increasingly challenging due to the pace of change in GPU design, which dictates the structural, power, cooling, and other needs of facilities. Investors must have expertise and insight into evolving data centre requirements to ensure their data centre will not be at risk of obsolescence and re-development before the projected lifecycle of the asset.
- Difficulty accessing adequate power with pricing stability – AI data centres require far more power than traditional data centres, with IT loads rapidly increasing in density. And for data centre facilities to have value, customers want to see that access to power is guaranteed rather than just potential. Customers also want power access that protects them against pricing volatility. Negotiating power agreements can be a rigorous, time and money consuming process. Once approved, connection timescales can be prolonged, requiring developers to consider long term strategies or to explore options for on-site power generation until grid power comes online. Expertise is essential to navigate this process to deliver power on the timeline that is needed to adhere to aggressive customer schedules for occupying a data centre and launching operations.
- Planning and permitting setbacks can lead to delays or threaten projects – One of the riskiest aspects of data centre development is the planning process. Many greenfield sites must successfully go through a laborious planning process before they can move forward. These processes are increasingly difficult to navigate, particularly as public opposition to data centres increases and places pressure on government entities responsible for permitting. The process often includes detailed analyses of noise impact, economic impact, environmental impact and more – all of which require expert support to meet the requirements of local authorities. Extensive due diligence and expertise in data centre permitting is indispensable for avoiding costly delays or cancelled projects that strand investment capital.
- Sustainability and ESG matters more than ever – Because AI computing uses far more resources than traditional IT environments, customers and government entities are shining a light on the sustainability of new data centres. AI customers are asking for more robust reporting that goes far beyond the traditional Power Usage Effectiveness (PUE) metric. Investors need to understand how sustainability metrics are evolving to ensure they can deliver on customer requirements and public reporting requirements for energy efficiency, water efficiency, carbon efficiency and more. Data centres also face a rising number of regulations related to data sovereignty, waste heat, heat re-use, ESG reporting and more. Developers must understand these evolving requirements and consider the plans early to avoid the need for costly re-designs and retrofits later.
Risks and success factors for redevelopment of legacy properties:
Investors pursuing redevelopment of legacy properties face many of the same challenges and risks discussed above, including alignment with rapidly changing customer specifications, sustainability requirements and planning obstacles. However, redevelopment has some unique challenges that come with their own risks.
- Redeveloping industrial sites with adequate power – Given the obstacles to acquiring adequate power for AI computing, investors often look at redeveloping industrial sites in order to take advantage of their built-in access to power. However, investors need the right strategy to mitigate the risks of these redevelopment projects. AI data centres require very specific power equipment and power resilience, requiring upgrades to industrial sites. Sites that have adequate power may not meet customer requirements for diversity of fibre paths. The existing footprints and infrastructure of an industrial site may not align well with facility designs that are optimised for AI computing, and other investments may be required to future-proof a site against the rapid changes happening in designs and requirements for AI data centres.
- Upgrading existing data centres – Existing data centres are attractive candidates for investment as the site selection, power, construction and more are already completed. These facilities are already generating revenue, which makes them financial assets from day one. However, investors need to have the plan and execution considered in order to successfully upgrade these assets to support more dense usage while maximising potential revenue. Existing data centres may have access to adequate power for cloud computing, but not AI-levels of power consumption. These facilities typically are not designed for the advanced power systems or liquid cooling systems required by next-gen GPUs, and other mechanical systems and infrastructure may be reaching the end of its lifecycle, requiring investment that must be considered at the outset. Existing data centres may also need other upgrades in order to meet the technical specifications and SLAs that come from AI customers, and due to contractual obligations with existing IT tenants, redevelopment of these data centres may require a risk managed and phased approach that protects operations in one portion of the site while upgrades are done in the other.
A blueprint for risk mitigation and successful projects:
HartDixon has extensive experience supporting investors through the complexities of both greenfield and redeveloped data centre projects. Our role is to provide the strategic oversight, technical advisory and project leadership needed to help clients accelerate delivery programmes, reduce risk, overcome operational and regulatory challenges, whilst protecting the long-term asset value.
- Technical due diligence at each phase
- Site appraisal, selection and acquisition support
- Master planning co-ordination and consultant team management
- Project Management and Employer’s Agent services during both design and construction
- Utility feasibility and power connection strategies
- Planning and stakeholder co-ordination
- Contract administration and delivery oversight
- Alignment with customer, operational and industry requirements
- Sustainability, compliance and risk management
- Planned Preventative Maintenance (PPM) strategies and ongoing asset support
Contact Charlie Railton today to learn more about how we can mitigate the risk and complexity of your data centre development strategy.
