*Building Description*

The building is a multipurpose social center with 1160 m<sup>2</sup> anda single floor, located in the southeast of the town (Figure 2). Since it was initially designed as a day care center for senior citizens, the building has some common spaces and facilities such as a dining room, dressing rooms, kitchen etc.

**Figure 2.** Building main facade (north).

The building has a flat roof with an "inverted roof" typology, characterized by having the thermal insulation (extruded polystyrene, XPS) over the waterproofing membrane, and over this thermal insulation liesa geotextile filter and a layer of gravels.

In our case, the detailed structure (Figure 3) incorporates the following elements: gravel layer (gravel diameter in the range of 20 to 50 mm and 1700 kg/m<sup>3</sup> bulk density, 50 mm of layer thickness), geotextile filter layer (2 mm), extruded polystyrene (XPS) insulation (40 mm), waterproofing membrane (5 mm) and concrete hollow block (300 mm).

**Figure 3.** Initial roof structure.

The green roof was built over the present "inverted roof". It was decided to remove the layer of gravel and a water retention layer was added below the growing medium (separated with a filter fabric layer). This storage layer increases the capacity of the roof for retaining water after a rain episode and significantly reduces the amount of runoff generated. Figure 4 displays the green roof structure, which includes the following layers: growth medium (80 mm thickness), permeable textile layer (2 mm), drainage layer (water storage layer, 30 mm), geotextile layer/root barrier layer (3 mm), XPS insulation (40 mm), waterproofing membrane (5 mm) and a concrete hollow block (300 mm).

**Figure 4.** Green roof structure.

The growth medium is a mixture of conventional gardening organic substrate (40%), volcanic lava rocks (40%) and silica sand (20%). In the upper part of the green roof, there are plants covering almost the entire area with a height in the range of 50 to 150 mm. These plants are genus sedum (a mixture of sedum album AH, sedum floriferum, sedum sediform, sedum reflexum, sedum spurium, sedum moranense and sedum acre).

Figure 5 displays the plan view of the building, denoting the roof area where the green roof was installed by the dotted line. The building area under controlled conditions with green and conventional rooftop was 280 m2. Figure 6 shows the green roof already installed.

**Figure 5.** General view of building roof (dotted line indicating green roof affected area).

**Figure 6.** General view of greenroof.

The monitoring system used for these experiments is presented in Figure 7.

**Figure 7.** Monitoring system scheme.

A set of six type T thermocouples, whose positions and missions are detailed in Table 1, enables to determine the evolution of the temperature at different layers of the roof. These temperature measurements allow identifying similar ambient temperature conditions in the registered data sets and provide the data required for the simulation tools.


**Table 1.** Thermocouple positions.

Specifications of the thermocouple sensors were: Probe PT100 RS PRO M16, PT100, +100 to +450 ◦C, diameter 6 mm, Connection head, Class B 4 Stainless Steel.

All these thermocouple signals were stored every minute in a data logger, together with the electricity consumption of the two heat pumps and the impulsion unit of the air conditioning system of the area covered by the green roof. In addition, wind velocities and solar radiation were provided by a nearby meteorological station.

The area affected by the green roof is about 1/4 of the total building surface and it has an independent air conditioning system. There are two heat pumps (Mitsubishi Electric PEA-RP250GA, with an input power of 8.455 kW each). Additionally, there is a common air impulsion unit of 3.05 kW of input power. Thus, maximum total input power, for both cooling and heating models, is about 19.96 kW.

This monitored building area is about 280 m<sup>2</sup> and it was closed during the testing periods to guarantee the control on internal loads and other factors that could affect the energy consumption. In this way, the test conditions are the same along the experimental campaigns with the conventional and the green roof. Only the external variables (temperature, solar radiation, wind etc.), which are

experimentally monitored, change. These campaigns cover different indoor comfort temperatures, in arange from 22 ◦C to 28 ◦C.
