Simon Harris, Head of Critical Infrastructure at BCS (Business Critical Solutions), explores how organisations can improve the performance of their data centres and meet the ever increasing economic, societal and legislative demands.
As the world increasingly turns its attention to the action required to limit the damage being done to the environment by human activity, the data centre sector is getting to grips with the role it must play in reducing carbon emissions. Energy consumed by data centres is attracting more attention as a matter of concern, as global computing capacity continues to rapidly increase. Already, data centres use an estimated 200 terawatt hours (TWh) each year globally – more than the national energy consumption of some countries.
The stakes are high
By January 1 2030, The Climate Neutral Data Centre Pact (CNDCP) requires signatories to make a binding commitment to achieve Power Use Effectiveness (PUE) of between 1.3 and 1.4 in sites built up to 2025, reflecting the increased challenges of interventions within operational environments. Among these the older data centres will have often been constructed with PUEs of 2.0 or more. In their current form they could be significantly distant from the required standard and in breach of their owners’ CNDCP commitments. They will also look increasingly unattractive to tenants pursuing socially responsible energy and carbon agendas.
With ambitious targets to be achieved it also begs the question that if our sector doesn’t get ahead of these targets, will the self-regulatory initiative become legislative and regulated? We believe our sector is at a crossroads with one route being proactive, investing in new technologies, self-generation and looking at innovative storage solutions to reach climate neutral targets. The other route is having legislation and regulation imposed on us and having to react to the imposition of energy, water and emission targets that we have no influence over.
The challenges are most complex for the stock of data centres spread throughout Europe around 60% of which are in excess of 18 years old. Consistent with buildings in other real estate sectors, there is a material and differentiated challenge in dealing with the existing stock of data centres built from the early noughties onwards which are either partially or fully occupied.
So how can organisations improve the performance of their data centres and meet the ever increasing economic, societal and legislative demands?
Rebuilding is not always the best answer
From the outside, the simple response to this problem would be to demolish and reconstruct through a managed vacation and migration of IT processing. However, the original design and financial plans for these sites were often authored with the structural and architectural elements having a life expectancy of 60 years. There is also a significant embodied carbon penalty to pay in demolition and rebuilding. A substantial amount of the construction work involves the use of energy dense concrete and steel to such an extent that refurbishing an existing facility can save in the order of 70%-80% of the carbon cost of a new build.
In many cases, an upgrade and refresh to critical infrastructure could liberate trapped electrical capacity for deployment to serve higher density and growing IT loads, for example through UPS replacement or cooling solution changes. These types of interventions will be more easily accommodated in Tier III facilities having two concurrently maintainable power and cooling paths, although the work will require careful planning and right-first-time execution. Nevertheless, such a solution overcomes the power availability challenges and takes the facility down a path towards better PUE performance.
Other considerations
At BCS our experience shows that there are some key project success factors:
- Clarity of brief
Within existing facilities, it is easy to allow the scope of the project to expand as operational stakeholders have their say and more facts and shortcomings about the facility come to light. The brief points the way in preventing budget and programme pressures being realised by focusing on scope that enhances or protects asset performance and value.
- Surveys and validation of record information
It is often wrongly assumed that record information is accurate and up to date. These should be checked and recommissioned if necessary.
- Budget setting and consideration of the detail of the project
Broad-brush cost estimates based upon historic norms are usually inappropriate and likely to give misleading answers. Consideration needs to be given to the project specifics and prevailing working constraints.
- Recognition of working constraints
Operational procedures at data centre sites often prevent uninterrupted working and change freezes are not uncommon. Projects tend to have extended execution times compared to new construction.
- Risk management processes
Project planning and execution needs to be aligned to operational risk issues and downstream impacts. Well-considered risk registers are required providing clarity about owners and actions, together with strong risk management processes.
- High levels of engagement with operational teams
Ensuring that project work can take place in a way that is congruent with the continuing business-as-usual operation of the facility is essential.
- Back-out plans agreed before changeovers are executed
Many projects will have one or several key changeover activities that require right-first-time execution and a pre-agreed process to ensure service availability in the event of unexpected difficulties
- Capable teams with relevant expertise and experience
A selection of skilled and experienced design and construction teams are essential to successful project delivery. Appointing project participants should therefore be done on the basis of a thorough pre-qualification process or through pre-existing knowledge of the capability of organisation’s key individuals.
- Financial considerations
Full project financial picture must be considered, inclusive of the tax position. Government tax incentives in the form of capital allowances are available to support and encourage businesses to undertake capital investment. These incentives are obtained through savings in Corporation Tax. With expenditure on qualifying plant and machinery likely to be substantial, Corporation Tax paying UK-based data centre owners should ensure that capital allowance benefits are maximised in order to improve their return on investment.
Obtaining the relief is not an automatic process and the tax rules are complex and often misunderstood. As a result, many businesses miss out on the tax relief available to them. Appointing a specialist capital allowances consultant with complex engineering systems experience will deliver tangible benefits.
2. Potential efficiency gains through investment
Investments in next-generation computing, storage and heat removal technologies will be required to avoid potentially steep energy use growth later this decade. This is in addition to parallel investments in renewable power sourcing to minimise the climate implications of unavoidable data centre energy use.
Conclusion
At BCS we have seen significant improvements to sustainability delivered as part of a general refreshment of assets that are beyond their economic life as we look to renew, reuse, refresh and repeat rather than rebuilding.
