CPD PROGRAMME | HEATING AND COOLING Screw compressor Evaporator Economiser Filter dryer Controller with local interface Condenser Figure 1: An example of a commercially available heat pump with variable speed twin-rotor screw compressor employing R-1234ze and flooded heat exchangers (Source: Carrier) appropriate concurrent heat demands or thermal storage facilities, although, in many buildings, much of the heat cannot be usefully used immediately on site. This presents opportunities to redistribute this heat while, at the same time, being able to integrate large-scale renewable heat sources, using heat networks. However, the EUs most recent assessment7 on the uptake of renewable energy sources notes that for heating and cooling the barriers are mainly due to shortcomings related to the capacities of the district heating networks. Currently, district heating provides 9% of the EUs heating; in 2012, the main fuel was gas (40%), followed by coal (29%) and biomass (16%).6 District heating can integrate renewable electricity (through heat pumps) by exploiting natural heat resources, such as geothermal, ground and water sources. Also, as stated in the European Commission document,6 synergies between wasteto-energy processes and district heating/ cooling could provide a secure, renewable and, in some cases, more affordable energy in displacing fossil fuels. Driven by F-Gas legislation for new lowGWP refrigerants and the requirements of Ecodesign and standards such as BS-EN 148258 to meet increasingly robust efficiency requirements (as measured by seasonal coefficient of performance (SCOP) and seasonal energy efficiency ratio (SEER)) heat pumps have been designed that are capable of delivering relatively high hot-water temperatures at high system efficiencies. This has increased the opportunity to employ such systems so that large heat producers for example, data centres can be used to provide heat for heat pump-powered district heating systems. The prime objective in controlling a data centre environment is to provide the correct amount of cooling, as efficiently and as reliably as possible, so as to allow the computing equipment to operate most effectively. The operational effectiveness, in terms of energy, is typically measured in terms of its power usage effectiveness (PUE), although there is increasing interest in the additional evaluation of energy reuse effectiveness (ERE), which indicates how otherwise wasted energy is used (see boxout PUE and ERE). Traditionally, cooling has been provided using mechanical vapour compression cooling, with the heat being rejected to the outside atmosphere through an air, or water, heat-rejection system. As data centre cooling represents a significant proportion of the buildings energy consumption, and is typically considered to be around 40% of total energy use, there is an increasing focus on improving the efficiency of cooling solutions to reduce the real PUE and provide a smaller ERE. Whether using mechanical cooling equipment alone or in combination with evaporative or air-based cooling systems, this still involves rejecting heat to the local environment. A common design approach is to recover the heat directly from the cooling system at the relatively low temperatures associated with vapour compression refrigeration, using heat recovery condensers or desuperheaters added between the compressor and the condenser. Using heat-recovery condensers, such systems can achieve water temperatures up to 50C or, when using desuperheaters, up to 60C (but with significantly less heat available in the latter case). Depending upon the application, such water temperatures, combined with the significant amounts of available heat, can offer economic advantages and reduce the ERE. However, as the rejected heat is typically not at temperatures required for many district heat networks, heat pumps can be applied to deliver the heat at a more usable temperature. Modern refrigeration technology and low-GWP refrigerants such as the hydrofluoroolefins (HFOs) can provide heat pumps capable of delivering hot water up to 85C. The temperature that a heat pump, such as that shown in Figure 1, is able to produce economically will be dependent on the temperature of the source heat, as well as the design and operational characteristics of the heat pump. Manufacturers data12 indicates that, for example, the heat pump in Figure 1 can economically deliver temperatures of 85C (COP 2.26) at a modest source water temperature of 20C (leaving the evaporator at 15C). The COP of the heat pump system can be significantly improved by employing two modular heat pumps in series, as illustrated in Figure 2. Series-connected heat pumps can economically deliver temperatures that are associated with standard low-temperature hot water heating from a relatively low temperature source. These can provide an efficient means of heat recovery, either as stand-alone installations or to supplement traditional boilers in applications such as district heating or industrial processes. By using ultra-low GWP refrigerants, such as HFO R-1234ze, in well-designed and installed systems, the potential environmental impact is reduced significantly compared with previousgeneration hydrofluorocarbon (HFC) refrigerants. Such systems can simultaneously produce chilled and hot water to supplement boilers as well as replace chillers, and attain an operating COP of 3.0 or more, so benefiting the ERE. PUE AND ERE The power usage effectiveness (PUE) is defined as the total energy needed for everything used in the bounds of the data centre facility divided by the energy used for computing PUE = (Cooling energy + small power energy + lighting energy + IT energy ) (IT energy) Conventional data centres typically have a PUE of about 2.0; many current hyperscale facilities have achieved PUEs better than 1.2.9,10 Energy reuse effectiveness (ERE) provides an indication of the useful energy recovery from a data centre that is used, for example, to heat buildings or drive an industrial process (aside that this is used directly in heating and cooling for the data centre itself). ERE = (Cooling energy + small power energy + lighting energy + IT energy reused energy) (IT energy) On its own, an ERE of 1.0 does not imply an efficient data centre infrastructure PUE and ERE together provide a better impression of the performance of data centres that reuse energy. See ERE: A metric for measuring the benefit of reuse energy from a data center11 for a more detailed explanation of the evaluation of ERE. 84 September 2019 www.cibsejournal.com CIBSE Sep19 pp83-86 CPD.indd 84 23/08/2019 16:56