In urban areas roofs are a critical part of the building envelopes which are highly susceptible to solar radiation and other environmental changes, thereby, influencing the indoor comfort conditions for the occupants. Roofs account for large amounts of heat gain/loss, especially, in buildings with large roof area and not many floors. Additionally, in urban areas roofs play a very important role on stormwater management and present many environmental benefits.
Some of the green roof (which includes a layer of vegetation, growing media and additional drainage/auxiliary layers) benefits are stormwater management , improved water run-off quality , improved urban air quality . Other benefit is the improvement of thermal behaviour of the building and energy savings as showed in this paper and previous bibliography , especially for summer period where cooling load savings can be around 10-30% (depending on previous insulation of the roof) .
Thermal behaviour of a building and so, the impact of green roof installation on the building energy consumption is not an easy subject. Thermal conductivity of used materials is an important factor, but also other variables as internal loads (lights, computers, people,..) or roof reflectance to solar radiation can play a very important role, especially in summer period. In the frame of the E2STORMED  project it was installed a green roof on a public building and it was monitored the impact on energy consumption of the air conditioning system.
The objective of this work is to show the results on energy saving of the green roof on buildings with Mediterranean climate, based on experimental data.
It has been monitored the energy consumption of the air conditioning system of the building (power metering every minute) and temperature profile (6 thermocouples type T, T1 to T6) of the roof for 15 months (July 2013 to September 2014). The monitored building area is about 280 m2 and during the period of July 2013 to February 2014, the energy consumption associated to Air conditioning system is due to the "conventional" roof. In February 2014 the external gravel layer was substituted by a green roof (see figure 1). Since March 2014 to September 2014 the energy consumption corresponded to the green roof. Monitored building area is not being used during the testing period to guarantee control on internal loads and other factors that could affect the energy consumption. Hence, the test conditions are the same except monitored variables (temperature, solar radiation, wind,...).
First evaluation on energy savings was made by direct comparison of similar days in terms of external temperature and solar radiation. For the identification of similar days for comparison purposes, it was used data from a meteorological station and one thermocouple, T3, located in the roof below the insulation material but in an area where green roof was not installed. This thermocouple has been metering since the beginning, so similar days should have similar values of T3 at the same time during the day.
(FIGURE 1: a picture showing constructive details and layers of both conventional and green roof)
# Results and Discussion
On the one hand, it was analysed the effects of heat storage of the materials. In conventional roof, gravel layer had a significant heat storage effect as it can be observed from the temperatures in the interface between gravel and XPS insulation layer (figure 2). In fact, it was recorded temperatures of approximately 45-55ÂșC in summer. Once green roof was installed, in the interface green roof-XPS, maximum temperatures where about 30-35ÂșC, so heat storage effect was softened and delayed. Taking into account the working schedule of the building, from 8.00 -- 15.00, no impact is observed in the electrical consumption due to this heat storage effect.
(FIGURE 2: a graph showing temperature versus time, under main layers, in summer, for both conventional and green roof)
On the hand, the study was focused on comparing power consumption before and after installing the green roof during a very similar day in summer. The comparison was made in the period from 10.30 to 15.00 pm with very similar external temperatures (differences lower than 1ÂșC and temperature values at T3 with differences lower than 0.5 ÂșC). In this study, an average power reduction of 26.75% was observed (see figure 3).
(FIGURE 3: a graph showing air conditioning power consumpition and external temperature versus time, under main layers, in summer, for both conventional and green roof)
The monitoring period of 15 months has permitted to evaluate the building energy behaviour during summer and winter time. During this period enough experimental data has been collected to analyse a summer time period with conventional and a green roof, and compare them. In this regard, it has been concluded that greenroof can modify notably the thermal behaviour of the roof regarding insulation and solar radiation effects. Energy savings for summer period can reach 25% for one floor building with average insulation in Mediterranean areas.
 J.C. Berndtsson, L. Bengtsson, K. Jinno, Runoff water quality from intensive and extensive vegetated roofs, Ecological Engineering 35 (3) (2009) 369--380.
 J. Mentens, D. Raes, M. Hermy, Green roofs as a tool for solving the rainwater runoff problem in the urbanized 21st century? Landscape and Urban Planning 77 (2006) 217--226.
 M. Santamouris, et al., Investigating and analysing the energy and environmental performance of an experimental green roof system installed in a nursery school building in Athens, Greece, Energy 32 (2007) 1781--1788.
 J. Yang, Q. Yu, P. Gong, Quantifying air pollution removal by green roofs in Chicago, Atmospheric Environment 42 (31) (2008) 7266--7273.
 E2STORMED (Improvement of energy efficiency in the water cycle by the use of innovative storm water management in smart Mediterranean cities) is funded by the MED Programme of the European Union.