The reference figure for domestic hot water consumption per person is 56 litres a day at 40°C. It is written into annex 1 of the French order of 31 March 2021, which carries the 3CL-DPE 2021 method, and confirmed by the ADEME and COSTIC technical guide from 398 monitored dwellings, with a standard deviation of 23 litres. The same method uses 79 litres per person per day for heavy-use behaviour, roughly 40% more. The same volume expressed at 55°C drops to 35 litres, because fewer hot litres deliver the same energy.
Where the per-person DHW benchmark comes from
What the 3CL-DPE 2021 method uses
The DHW demand in an energy assessment does not depend on the actual occupants. The method goes through an equivalent-adult count, derived from the average habitable floor area of the dwelling, then multiplies that count by 56 litres a day at 40°C. The full expression is Becs = 1.163 × Nadeq × 56 × (40 − Tcold) × nd, where Tcold is the month's cold water temperature and nd the number of occupied days. The calculation conventionally counts one empty week in December, so 358 days a year.
The table of available reference values
Five figures circulate, and they do not say the same thing, because they are not expressed at the same temperature nor intended for the same use.
| Reference | Value | Volume temperature | What it is for |
|---|---|---|---|
| 3CL-DPE 2021, conventional behaviour | 56 L/day/person | 40°C | calculating assessed DHW demand |
| 3CL-DPE 2021, heavy use behaviour | 79 L/day/person | 40°C | bounding the top of the range |
| ADEME and COSTIC monitoring, measured average | 56 ± 23 L/day/person | 40°C | setting a quotation assumption |
| ADEME and COSTIC monitoring, generation equivalent | 35 ± 14 L/day/person | 55°C | sizing a store |
| SOCOL 2021 sheet, solar sizing basis | 30 L/day/person | 60°C | sizing a collector array |
A figure expressed at 40°C is a blended volume, as it leaves the shower. A figure expressed at 55 or 60°C is a volume to be generated. Confusing the two oversizes the plant by roughly 60%.
What happens to the 59 litre figure
It appears in no primary source. Neither the 3CL-DPE 2021 method, nor the ADEME and COSTIC guide, nor the SOCOL sheet carries that value. The nearest thing to it is the 56 litre average, to which the 23 litre standard deviation must be added to understand the real spread. On a calculation note, write 56 and cite the source: a figure with no reference behind it cannot be defended in a review.
What moves DHW consumption
Household size, ahead of everything else
The instrumented monitoring shows demand per person falling as the household grows. The type of housing, social or private, house or flat, has no significant influence on the average.
| Situation | Demand per person at 40°C | Dwelling demand at 40°C |
|---|---|---|
| Single occupant | 80 L/day | 80 L/day |
| Average of the 398 monitored dwellings, 2.5 people | 56 L/day | 140 L/day |
| Five-person household | 45 L/day | 225 L/day |
The right-hand column is a product, not a measurement: it takes you from the per-person benchmark to the volume to be sized, provided the assumed occupancy is written on the survey sheet.
Flow rate at the outlet and draw duration
The volume drawn is flow rate times duration. A shower head at 12 l/min over six minutes is 72 litres; the same duration at 7 l/min is 42. Measure the flow rate on site, graduated bucket and stopwatch, rather than repeating the figure printed on the box: the gap between the two is common on an installation with no pressure reducing valve. The measured flow is also what lets you justify a flow restrictor on the quotation.
Configurations where the benchmark no longer applies
A per-person benchmark assumes permanent occupancy. It does not hold for short lets, holiday accommodation, second homes, house shares with turnover, or where a business is run from the dwelling. In those cases start from readings: water bills over 12 to 24 months, hot water meter, flow measurements. The volume genuinely drawable is then worked through in DHW storage volume.
From litres to kWh on a calculation note
The expression and the assumptions to write down
One litre of water raised by 1 kelvin takes 1.163 Wh. Useful demand is therefore V × 1.163 × ΔT, in watt-hours. Two assumptions have to appear explicitly: the temperature the volume refers to, and the cold water temperature used. Assessment methods take a monthly cold water temperature by climate zone; on site, a value measured at the tap in winter beats an annual average.
The annual demand table by household
Calculated from the volumes above, with cold water at 12°C, a volume expressed at 40°C and 358 occupied days.
| Situation | Volume at 40°C | Daily useful demand | Annual useful demand |
|---|---|---|---|
| Single occupant | 80 L/day | 2.6 kWh | 933 kWh |
| Household of 2.5 people | 140 L/day | 4.6 kWh | 1,632 kWh |
| Five-person household | 225 L/day | 7.3 kWh | 2,623 kWh |
These are useful demands at the outlet. They include neither generator efficiency, nor storage losses, nor distribution losses: metered consumption will necessarily be higher.
The unit error that wrecks a quotation
The classic mistake is to take a volume expressed at 40°C and apply the rise between cold water and a 60°C storage setpoint. Blending is then counted twice and demand inflates by 60 to 80%. The generator that comes out of it is oversized, it short cycles, and the gap between the consumption announced and the consumption billed becomes impossible to explain to the client.
What the benchmark does not cover
Distribution and secondary return losses
The 56 litre benchmark is a draw-off demand. Everything lost between the generator and the tap is added on top, and in multi-occupancy buildings those losses are heavy: the SOCOL sheet notes that the ratio between secondary return losses and DHW demand generally sits between 0.6 and 1.5. In other words, the return loop can consume as much as the showers. The subject is handled in DHW secondary return, and the effect of insulation in the distribution efficiency table.
The temperatures imposed in dwellings
Approved Document G caps bath outlets at 48°C in new dwellings, while legionella control under HSE guidance L8 calls for storage at 60°C and 50°C reached at the outlets within a minute. Those two constraints are reconciled by thermostatic mixing at the outlet. They set the generation setpoint whatever the volume consumed.
Scaling up to a whole building
Adding up per-person demands does not give a building's demand. Diversity between dwellings cuts the peak, vacant flats cut the average, and communal distribution losses push it back up. The transposition goes through a different accounting unit, developed in calculating DHW demand for a multi-occupancy building.
Securing the costing and the commissioning
What goes on the quotation
Show the demand assumption used, in litres a day at 40°C, the occupancy taken into account and the source cited. Add the generator's commercial reference with its declared tapping profile, the safety group, the expansion vessel, the thermostatic mixing valve and the insulation of the connecting pipework. The link between the stated demand and the chosen appliance is read on the DHW tapping profiles.
What gets recorded at handover
Record the generation setpoint, the temperature at the least favourable outlet after mixing, the pressure at the reducing valve, the flow rate measured at the main outlets and the programmed thermal treatment cycle. Date it, have it signed, attach the product fiche. A demand disputed six months later is settled on those readings, not on an argument about habits.
Setting the demand at the pre-visit stage with Argile
With Argile, DHW demand is qualified from the office, before you travel: occupants, outlets, existing generator, siting constraints. You arrive with costed work plans and a traceable demand assumption, which leaves you time to take the measurements that matter instead of rebuilding the file from memory. The quotation comes out with the funding already applied and the supporting documents already attached.
