The reliability of energy performance diagnostics (EPD) again called into question by a new study

The first deadline of the Climate Law took effect this Sunday, January 1, 2023: G-class housing whose consumption exceeds 450 kWh/m²/year (i.e. approximately 90.000 housing units) is now prohibited from being rented. The precision of the calculation of the consumption and the energy label is therefore very important since it will determine the ability of an owner to rent his property, and this also has a significant impact on the value of the latter.

A new method of study that leads to the same results

The reliability of DPEs is often questioned. This is evidenced in particular this study of the UFC Que Choisir which, for certain dwellings, had obtained very different results according to the diagnosticians. The Hello Watt study, carried out on more than 200 dwellings based on consumption data and DPEs, corroborates the results of the UFC Que Choisir study carried out on 7 dwellings.

Hello Watt compared the energy label of the 221 homes for which their teams have the DPE* with their energy consumption per m2**. Result :

In 29% of cases, the energy label corresponds to our calculation of consumption per m2
In 40% of cases, the label does not match, and a neighboring label is assigned (for example: B or D instead of C)
In 31% of cases, the label does not match, and it is not even a neighboring label that is assigned (for example B for D or, more seriously, G instead of E)

In total, therefore, according to these calculations, 71% of the DPEs are not correct, ie they do not correspond to the consumption of the dwelling in kWh/m²/year.

An approximation that poses a problem, since with the new requirements of the Climate Law, a difference - even of a slice - can have significant consequences. For example, between an E classification and an F classification, it is the thermal sieve status that is at stake, and with it a ban on rental and a significant discount in real estate value.

The method is different from the studies carried out by consumer associations (UFC que Choisir in particular) - where different diagnosticians gave different DPEs for the same dwelling - but the result is the same: the DPEs are currently not reliable on the market.

DPEs not much more accurate than if the letter of the DPE was randomly assigned

If we imagined an unscrupulous diagnostician who would assign the labels at random (respecting the national distribution and without even looking at the accommodation), he would be good in 25% of cases (vs 29% above)!

Consequently, there is an urgent need to improve the reliability of DPEs, which requires better training and better supervision of the sector, which is still young.

How was the Hello Watt study carried out?

As a reminder, the DPEs analyze the characteristics of a dwelling to estimate its consumption per m2. From these estimates, a label is assigned.

Among the homes with an energy label and listed in the ADEME database, Hello Watt has identified 221 that use the Hello Watt application. For these dwellings, their actual consumption is known, as well as their surface area.

It is thus possible to match their actual consumption/m2 with their energy label, and therefore with their estimated consumption/m2.

*ADEME Open Data Portal

**From the postal addresses indicated for each DPE within the ADEME dataset, it is possible, thanks to the Hello Watt application, to find the surface area of the accommodation and its energy consumption.

The complete study is available in by clicking here.