A building's DHW demand is not obtained by adding up per-person demands. The reference method, set out in the ADEME and COSTIC technical guide, brings each flat back to a number of standard dwellings, the standard dwelling being a three-room social housing flat occupied by 2.1 people. Average daily demand per standard dwelling is 125 ± 50 litres at 40°C, or roughly 70 litres at 60°C with cold water at 16°C. Expressed per dwelling, demand comes out at around 110 litres at 40°C in private housing and 130 litres in social housing.
Why a building's demand is not the sum of individual demands
The three effects that shift the result
Three effects separate dwelling scale from building scale, and they do not pull the same way. Diversity between dwellings smooths the peaks, so the variation observed at building scale is smaller than at flat scale. Vacant dwellings pull the average down. Communal distribution losses pull it up. A calculation that ignores any one of the three is wrong by tens of percent.
What the measured values contain
Building-scale demand values were established from measurements taken at the communal generation point. They therefore already include part of the distribution losses, communal and individual, down to the outlets. Because the monitored stock mixes new buildings with buildings several decades old, that share varies with the insulation thicknesses, pipe lengths, diameters and distribution temperatures encountered.
What the method does not replace
This calculation gives a demand, not an installation. Choosing the generation architecture and the generator is a separate decision, handled in communal DHW generation. Sizing a solar contribution follows yet other ratios, developed in communal solar DHW. The three come in that order: the demand, then the architecture, then the solar.
Counting the building's standard dwellings
The definition of the standard dwelling
The standard dwelling is a three-room flat, the most common type, taken from social housing where monitoring data is most plentiful, and occupied by 2.1 people, matching the average occupancy rates recorded. The equivalence coefficients account for the differences in occupancy between flat types, since demand depends directly on them. Sanitary fittings, bath or shower, are not taken into account: their influence is smaller, occupants now mostly taking showers.
The table of equivalence coefficients
| Dwelling type | Private stock, average occupancy | Private stock, coefficient | Social stock, average occupancy | Social stock, coefficient |
|---|---|---|---|---|
| One room | 1.2 | 0.6 | 1.2 | 0.6 |
| Two rooms | 1.4 | 0.7 | 1.4 | 0.7 |
| Three rooms | 1.9 | 0.9 | 2.1 | 1 |
| Four rooms | 2.3 | 1.1 | 3 | 1.4 |
| Five rooms | 2.7 | 1.3 | 3.7 | 1.8 |
| Six rooms or more | 2.9 | 1.4 | 3.9 | 1.9 |
The average occupancy rates come from USH-DEEF calculations based on the 2008 population census and the 2006 national housing survey. A four-room social housing flat therefore counts as 1.4 standard dwellings, and a three-room private flat as 0.9.
A worked count
On a 30-unit private building made up of 6 one-room, 9 two-room, 10 three-room and 5 four-room flats, the count gives 6 × 0.6 plus 9 × 0.7 plus 10 × 0.9 plus 5 × 1.1, so 24.4 standard dwellings. Average daily demand comes out at 24.4 × 125, roughly 3,050 litres a day at 40°C. Write the count type by type on the calculation note: that is what a reviewer will check, not the total.
The demand values to use
The table of average daily demands
| Counting basis | Average daily demand at 40°C | Scope |
|---|---|---|
| Per standard dwelling | 125 ± 50 L | all monitored sites |
| Per dwelling, private housing | around 110 L | private stock buildings |
| Per dwelling, social housing | around 130 L | social stock buildings |
| Per one-room flat, social housing | around 75 L | annual meter readings |
| Per five-room flat, social housing | around 190 L | annual meter readings |
Demand per dwelling in the private stock is slightly lower than in the social stock, because of lower occupancy. Uncertainty on the values from remote monitoring is of the order of 5 to 10%, and of the order of 25% on values extrapolated from occupancy statistics.
Converting a volume into energy
The core of the calculation is one line. Useful energy is V × 1.163 × ΔT, where V is the volume in litres and ΔT the difference between the cold water temperature and the temperature the volume is expressed at. The result is in watt-hours. On the 3,050 litres of the previous example, with cold water at 14°C and a volume expressed at 40°C, daily useful demand is 3,050 × 1.163 × 26, roughly 92 kWh.
The unit error that doubles the result
The most common mistake is to take a volume expressed at 40°C and apply the difference between cold water and a 60°C generation setpoint. Blending is then counted twice. On the same example, demand would go from 92 to 163 kWh, and the plant that came out of it would be oversized by 77%. Fix the temperature the volume refers to, write it down, and take the matching difference.
The conventional assessment calculation, which is not yours
The 3CL-DPE 2021 method for blocks of flats
The 3CL-DPE 2021 method does not use the standard dwelling. It works out an average habitable floor area per dwelling, dividing the building's total habitable area by the number of dwellings, then derives a maximum occupancy coefficient. If that average area is below 10 m², Nmax is 1. Between 10 and 50 m², Nmax is 1.75 − 0.01875 × (50 − Savg). From 50 m² up, Nmax is 0.035 × Savg.
From occupancy coefficient to annual demand
The equivalent-adult count follows: Nadeq is Ndwellings × Nmax if Nmax is below 1.75, and Ndwellings × (1.75 + 0.3 × (Nmax − 1.75)) above it. Monthly demand is then 1.163 × Nadeq × 56 × (40 − Tcold) × nd, with Tcold the month's average cold water temperature for the climate zone and nd the number of occupied days, one week of absence being counted in December.
Why not to size with it
The conventional assessment calculation exists to compare buildings while stripping out occupant behaviour. It reduces DHW demand to a function of floor area and climate zone, which is exactly what must not be done to size a plant. Use it for the label and for a funding file, never for a plant room.
From demand to sizing, and how to evidence it
From energy to power
Power follows from the energy over the peak period divided by the duration of that peak. The longer the peak period considered, the lower the average draw rate, which shifts the trade-off between generation power and storage volume. A large store allows moderate power, a small store forces the opposite. The target recharge time, often one to two hours, closes the reasoning.
The loss line, costed separately
Secondary return losses are not a detail. The SOCOL 2021 sheet records a ratio between distribution losses and DHW demand generally between 0.6 and 1.5, and uses working defaults of 0.6 for new buildings and 1 for existing ones. On an existing building, measure rather than estimate: loop flow and return temperatures, circulation rate. The reduction levers are detailed in DHW secondary return.
What makes a calculation note defensible
A defensible calculation note carries the assumptions before the results: the standard dwelling count type by type, the unit demand used and its source, the temperature the volume is expressed at, the cold water temperature, the peak duration considered, the distribution loss assumption. Add the site readings that support it and the temperatures imposed by hygiene requirements, detailed in legionella safety devices. With Argile, those assumptions are entered once and come out dated on every work plan, which lets you compare variants without redoing the calculation and present a file that a decision-making body can actually vote on. The same requirement applies on the heating side, where Argile generates a sizing note compliant with NF EN 12831-1 from the data recorded, assumptions included.
