| HEAT PUMPS world. As somewhat controversially illustrated3 by Iman et al, the reliability of the modelled solution will also be dependent on the knowledge and skill of the modeller. The peak heating, cooling and hotwater demand for a development may be obtained by calculation and modelling, or by monitoring demands of existing buildings. Actual data derived from similar operational buildings are extremely useful as a check on proposed designs. As discussed in CP1,1 when employing communal networks for multi-residential systems, the peak heating demand for the building is always found to be significantly less than the sum of the peak design demands calculated for the individual dwellings. For example, many dwelling space heating demand calculations assume that there is heat loss to adjacent unoccupied properties. If these heat demands are then multiplied by the number of dwellings to give a block heat demand, there will be significant oversizing. A space heating diversity factor may be applied to the individual dwelling space heating demands, calculated to include heat losses to adjacent properties. CP11 recommends the rule of thumb empirical formula from the Danish guidebook Varme Stbi,4 illustrated in Figure 1. For a small number of apartments, it would not be appropriate to apply the diversity factor but, as can be seen, groups of more than 10 apartments are likely to have a coincident heating demand of less than twothirds the sum of individual peak demands. CIBSE AM16 highlights that accurate calculation of DHW loads is particularly important for the design of heat pump systems in multi-residential buildings, as DHW forms a large proportion of the overall heating load, in terms of both peak load and annual energy consumption. The flow temperature in the network is set to meet the specific needs of the residential systems, and this is often determined by the temperature required to meet domestic DHW needs. (The actual maximum temperature required for the DHW in the dwelling will depend on whether instantaneous or storage hot water is used see the recent CIBSE note, Domestic hot water temperatures from instantaneous heat interface units, for a clear explanation of current thinking.) Realistic assessment of diversity (of use) is particularly important. There are continuing efforts to produce metrics that can be applied by the designer, including the CIBSE-supported Loading Unit Normalisation Assessment (LUNA) project,5 as it is thought that current practice tends to overestimate demand1. CIBSE CP11 recommends that the method of Danish 1.0 0.9 0.8 Diversity factor CPD PROGRAMME Space heating diversity factor = 0.62 + (0.38/N) 0.7 0.63 0.6 0.5 0.4 0.3 Example hot water diversity factor based on CP1 and DS 439 0.2 0.1 0.0 0 5 10 15 20 25 30 35 40 45 50 Number of dwellings, N Figure 1: Example of space heating diversity factors noted as rule of thumb in CP1 for UK applications, and diversity factors for instantaneous domestic hot-water systems for dwellings, based on equation included in Danish standard DS 439 and employing default values (Source: CIBSE CP1) standard DS 4396 may be applied, which employs a (relatively simple) equation that is clearly explained in CP1 with an excellent example and produces a demand in terms of a nominal cumulative design DHW flow. This was applied to provide the example diversity graph in Figure 1. The resulting peak load on the heat network effectively relates to the demand from instantaneous hot-water heating for example, as would be realised by employing heat interface units (HIUs) with a higher-temperature network. When applying ambient loops in conjunction with in-home heat pumps, it is highly likely that there will be some form of DHW storage, such as individual hot-water cylinders in each apartment. This can significantly reduce the peak load on the heat network, as heating of the water store may be scheduled to minimise the coincidence of multiple residences drawing heat at the same time. (Maximising this benefit requires considered design, building-wide control, and the compliance of the residents.) The building network distribution losses will be dependent on the location and temperature of the pipework, and the insulation. So, for example, when considering Flow o C 80 80 80 80 70 70 70 70 60 60 60 60 30 30 20 20 Return C (Flow Return) K Network heat loss Average loss per kW metre, W.m-1 80 70 60 50 70 60 50 40 60 50 40 30 30 20 20 10 0 10 20 30 0 10 20 30 0 10 20 30 0 10 0 10 8.72 8.05 7.38 6.71 7.38 6.71 6.04 5.38 6.04 5.37 4.71 4.05 2.06 1.41 0.75 0.38 o 9.88 9.12 8 37 7.61 8.36 7.61 6.85 6.10 6.85 6.09 5.34 4.59 2.33 1.59 0.85 0.43 Table 2: Illustration of heat losses from insulated 880m communal heat network (Source: Glen Dimplex Heating & Ventilation8) 46 January 2022 www.cibsejournal.com CIBSE Jan 22 pp45-48 CPD 190.indd 46 23/12/2021 14:08