DATA CENTRE MANAGEMENT


Strategies for efficient, effective cooling


Cliff Jones looks at the technologies and key considerations for the specification and design of cooling systems, which are critical to the operation of data centres of any size.


Cliff Jones www.daikin.co.uk


A


Head of applied at Daikin


ll data centres, whatever their size, face the same challenges of continuous operation and management


of very high heat loads. Adequate cooling is essential and must be able to always operate reliably and efficiently.


With the right approach there are


many legacy data centres that can be run more efficiently, with significant savings, which should be actively explored as part of any retrofit or upgrade of existing plant, rather than the like-for-like replacement approach.


Size diff erences For small server rooms, direct expansion (DX) systems are often ideal. They offer cost-effective cooling with built-in settings and require no extra hardware or complex controls, allowing for easy configuration during setup. Various indoor units are available,


including wall-mounted and ceiling- mounted options, to fit different spaces. For larger facilities, chillers paired


with computer room air handling (CRAH) units are typically best. CRAH units cool warm air from the data hall using chilled water coils. However, as computing power increases, especially for AI services, additional cooling methods like direct-to- chip liquid cooling with a coolant distribution unit (CDU) are becoming necessary. The CDU connects to the main chiller, providing targeted cooling for high-performance equipment.


EIBI | NOVEMBER � DECEMBER 2024


Design considerations There are several factors that must be evaluated when considering the design and specification of the cooling system to ensure it meets the needs of the site. Given the high costs of energy and the focus on reducing the environmental impact of our buildings and technology, it is essential that the cooling system operates as efficiently as possible. There are several ways that energy reductions can be achieved through the design and capabilities of the system. To enhance energy efficiency


in cooling plants, incorporating components like variable frequency drive (VFD) compressors is essential, but advanced system controls are crucial. For instance, Daikin’s intelligent Data Centre Manager (iDCM) utilises machine learning to optimise plant operations. Operating under partial load conditions by distributing the cooling load across all units—rather than keeping backups on standby— maximises efficiency. This strategy ensures redundancy; if a unit fails, the remaining units can increase their load to maintain adequate cooling while still operating efficiently. Furthermore, it is valuable to look


for a system that has free cooling capability. This allows the system to use the lower temperature outdoor air during cooler periods to provide the cooling of the water circuit. This means mechanical cooling using the compressor can be reduced significantly, in turn minimising the energy required. For maximum


benefit of free cooling, chilled water temperatures need to be raised as high as practically possible. In the changing decarbonisation


environment, the opportunity to reclaim heat and supply this into district heat networks, or as a minimum make provision for future connection, will benefit the communities’ heating requirements who are reliant on data.


Future needs Data centres are highly specialised environments and as such it is important to select a cooling system that has been designed to meet these requirements. This includes active harmonic filters integrated into the units to prevent the harmonic distortions created by components, such as VFDs, from impacting the server equipment. Additionally, due to the pace of


development in computing technology, data centre requirements will evolve rapidly. It is therefore important to ensure the cooling system has the flexibility to meet future needs. Even for sites under construction that have been designed with both air cooling using CRAH units and liquid cooling using CDUs, there is likely to be a need for adaptation to new requirements. Ideally the cooling system specified should have the flexibility to allow a shift to greater liquid cooling due to an increasing AI load in the coming years with minimal additional investment. A crucial consideration for cooling


systems is the sustainability of the refrigerant supply. For refurbishment projects especially there is an opportunity to achieve circularity.


Designing an effective cooling system requires careful evaluation of various factors to meet the specific needs of the data centre


Through schemes such as Loop by Daikin, refrigerant can be recovered from the units removed from a site and reclaimed to the standard of virgin product ready to be reused. The newly installed units are administratively allocated previously reclaimed refrigerant, closing the refrigerant supply loop.


Choosing the right partner The complexity of data centre cooling projects means that it is critical to have the best possible technical and design support throughout the project. It is often beneficial to look for a manufacturer that has a specialist team dedicated to data centre projects. For example, at Daikin we have a data centre team that spans the EMEA region and is supported by the local Daikin teams. This type of dedicated expert team will have in-depth knowledge of the challenges and requirements specific to data centres and be able to draw on best practice from across the region. It also provides a single point of contact for this element of the project, which is often essential on large scale developments. Data centres have never been


more important but the criticality of cooling these demanding spaces means specialised solutions and in- depth expertise from manufacturers is essential to ensure the required levels of reliability, performance and sustainability. ■


25


Page 1  |  Page 2  |  Page 3  |  Page 4  |  Page 5  |  Page 6  |  Page 7  |  Page 8  |  Page 9  |  Page 10  |  Page 11  |  Page 12  |  Page 13  |  Page 14  |  Page 15  |  Page 16  |  Page 17  |  Page 18  |  Page 19  |  Page 20  |  Page 21  |  Page 22  |  Page 23  |  Page 24  |  Page 25  |  Page 26  |  Page 27  |  Page 28  |  Page 29  |  Page 30  |  Page 31  |  Page 32  |  Page 33  |  Page 34  |  Page 35  |  Page 36  |  Page 37  |  Page 38  |  Page 39  |  Page 40  |  Page 41  |  Page 42  |  Page 43  |  Page 44