By Dominic Cutts, director – System Solutions, Armstrong Fluid Technology

Generation of heat is an unavoidable outcome of electrical resistance within IT processors and, as IT processors become ever more powerful, the demand for cooling is rising incrementally. Chilled water cooling systems have a crucial role to play in dissipating heat to ensure server reliability and prevent downtime, but there is an environmental price to be paid. Overall energy demand of data centres is escalating by between 5% and 10% a year, and they are thought to be responsible for 3% of the total electrical energy consumption of the UK. As a result, the environmental impact of data centres has come under international scrutiny, with a number of organisations developing guidelines and regulatory frameworks within which data centres need to operate.

One such development is the PUE (Power usage effectiveness) metric, which provides a universal benchmark for data centre environmental performance. It has been developed by The Green Grid Association (a non-profit, open industry consortium of end users, policy makers, technology providers, facility architects, and utility companies) and carries significant weight. It provides an at-a-glance picture of the efficiency of the energy consumed in the building both for processing and for cooling. A PUE of 1.5, for example, means that for every unit of power used on core IT equipment, an additional 0.5 units is spent on the supporting infrastructure.

As the power relating to processing is difficult to reduce without impacting on IT capability, it is the infrastructure/HVAC part of the metric which is the focus for increased efficiency. So how do you design HVAC systems for these critical cooling applications?

Three important priorities are:

• Driving down energy wastage by migrating away from capacity based control.
• Developing an incremental upgrade strategy that will alleviate energy wastage related to over-sized plant.
• Removing variables to improve reliability and performance repeatability.

Firstly, traditional capacity-based approaches to control of chilled water systems, designed around independent control loops, are not capable of delivering the levels of PUE demanded by today’s data centres. Demand based control methodologies such as Hartman LOOP digital relational control, however, can achieve the requirements by treating the entire system ‘holistically’, rather than as individual sub-systems.

Key energy consuming components, such as Armstrong Design Envelope pumps, can be sequenced to operate along their peak efficiency natural curve for all load conditions). In addition, Hartman LOOP can calculate and determine the best power relationships between the chiller, condenser pump and tower fan. As well as ensuring the efficiency of individual components, this optimises power relationships across the system, with equipment loading in one device traded off to pick up more load on another. This achieves the same net kWc for a lower kW input (a COP of greater than 7.0 compared to previous averages of between 3.2 and 2.2), achieving significant improvements in PUE.

Secondly, the business model employed by data centre operators involves expanding to full processing capacity over time. So cooling capacity needs to be ‘bolted-on’ in line with these incremental increases in building load. Ability to do this assists the data centre operator’s profitability by preventing the need to front-load capital investment for the facility, and avoids the significant energy wastage of over-sized plant. Modular packaged pump solutions are ideal, as they are pre-designed, off-site manufactured pump packages that are fully-configurable to suit the application. Avoiding bespoke design, they deliver predictability and reproducibility of environmental performance which is so important for critical cooling applications with transparent reporting.

Thirdly, data centres are benchmarked internationally using a reliability metric referred to as ‘The Nines’, which records a range of data such as percentages of IT equipment downtime. The supporting cooling systems play a key role in achieving the best possible IT integrity and must themselves offer outstanding reliability and robustness. Off-site manufacture has proved extremely effective in meeting these requirements. Assembling and integrating a complete chilled water plant room, or pump package, in the controlled environment of a factory, for example, assists the best possible quality of manufacture and component compatibility. Testing can be carried out in the factory under controlled conditions prior to delivery, and the best solutions incorporate fail-safe features. Armstrong’s self-contained IPP-CHW fully-integrated chilled water plant package, for example, is a popular solution.

Whilst the guidelines issued by Green Grid and others (such as the EU Code of Conduct on Data Centres' Energy Efficiency) are currently voluntary rather than mandatory, they provide powerful commercial advantages for those data centre operators able to differentiate themselves through environmental performance. By adopting best practice approaches such as demand-based control, modular system design and off-site manufacture, HVAC partners can be the key to wider commercial success for the data centre operator, as well as enhancing carbon footprint.