ONLINE BRUSSELS HERITAGE - ONE-DAY SEMINAR - 11/12/2014

We'll now focus on one particular week in winter to take a detailed look at what is happening (fig. 7). The temperature of the house is stable, with a low inertia effect. The external temperature is cool and variable. We can see that the curve corresponding to the measurement displays more significant variations than the simulation curve, even though the set-point temperatures of the simulation are those that were measured (fig. 8). From a consumption point of view, this means that the boiler, in reality, is continually stopping and starting, a phenomenon that is not taken into account by the simulation. This typically illustrates a difference arising from the way in which the calculation engine takes the reality into account. In practical terms, the calculation somewhat smooths out the heating system even though, in reality, an old boiler operates a lot less smoothly. Inversely, when the heating flow is imposed on the model and changes in temperature are observed, we notice that changes are much more marked in the calculation. In this case, the inertia of the heating system is not taken into account in the calculation, which constitutes yet another bias of my mathematical tool. This phenomenon is not very significant over an average. However, it becomes very significant when we look at exactly what happens on an hourly basis.

CASE STUDY 2: THE PARISIAN APARTMENT

The apartment dates from the late 19th/eariy 20th century (fig. 2). With a surface area of 108 m², comprising three bedrooms, it is located on the 5th floor of a handsome apartment building. The façades are made from hard limestone (called Paris stone) with beautiful dressed stone on the external street-facing side and plaster render on the interior walls. Thinner, less elaborate breezeblocks covered in render are found in the courtyard-facing side. Paris stone, which is very hard, offers very poor thermal conductivity for a limestone. All of the windows are single-glazed and original. They are thermally inefficient. A family of two adults and two teenagers occupies the apartment. All family members are occupied during the day. There is no mechanical ventilation apart from a basic extractor in the bathroom, added retrospectively. The family opens the windows every day.

The same method as previously was applied. This gives a graph of annual consumption (fig. 9). Energy consumption is a bit lower than the house in Noisiel: 165 kWh/m²/year, the correct figure for a building with single glazing. The extreme contiguity plays a positive role here. The occupants, who are quite well off, have no hesitation in turning on the heating; we are not dealing with a scenario of energy insecurity here that would require the set-point temperatures to be lowered. On the contrary, the apartment is comfortable but, nevertheless, does not consume a particularly high amount of energy. This example illustrates how certain

Fig. 7

Dynamic analysis over one week (source: Cerema).

Fig. 8

Changes in heating output measured and calculated. The interior temperature is imposed (source: Cerema).

60 | Analysis of uncertainties in dynamic thermal simulations for old housing