State-of-the-Art Green Roofs: Technical Performance and Certiﬁcations for Sustainable Construction

: Green roof systems, a technology which was used in major ancient buildings, are currently becoming an interesting strategy to reduce the negative impact of traditional urban development caused by ground impermeabilization. Only regarding the environmental impact, the application of these biological coatings on buildings has the potential of acting as a thermal, moisture, noise, and electromagnetic barrier. At the urban scale, they might reduce the heat island e ﬀ ect and sewage system load, improve runo ﬀ water and air quality, and reconstruct natural landscapes including wildlife. In spite of these signiﬁcant beneﬁts, the current design and construction methods are not completely regulated by law because there is a lack of knowledge of their technical performance. Hence, this review of the current state of the art presents a proper green roof classiﬁcation based on their components and vegetation layer. Similarly, a detailed description from the key factors that control the hydraulic and thermal performance of green roofs is given. Based on these factors, an estimation of the impact of green roof systems on sustainable construction certiﬁcations is included (i.e., LEED—Leadership in Energy and Environment Design, BREEAM—Building Research Establishment Environmental Assessment Method, CASBEE—Comprehensive Assessment System for Built Environment E ﬃ ciency, BEAM—Building Environmental Assessment Method, ESGB—Evaluation Standard for Green Building). Finally, conclusions and future research challenges for the correct implementation of green roofs are addressed. Cooling potential of green roofs can be around 3.02 kWh per day for an LAI (leaf area index) of 4.5. This is enough to maintain an average room temperature of 25.7 ◦ C. [46] Experimental study for measuring energy savings in cooling for Mediterraneancontinentalclimate. Energy savings (16.7%) for cooling and an increase in energy consumption for heating (11.1%) were observed compared to conventional ﬂat roofs. These results correspond to a 20% to > 85% area covered by vegetation.


Introduction
Due to its multiple benefits, biological coatings on buildings (i.e., green roofs and walls) were used by different human civilizations [1,2]. According to the scientific community, there is testimonial evidence of the use of biological coatings on Babylon's Hanging Gardens and Babel Tower [2][3][4]. Similarly, religious buildings called "Ziggurats" were constructed in Mesopotamia using plant growth on their building surfaces most probably to reduce heat interchange with the environment [3]. Biological coatings were also part of vernacular architecture in Scandinavia. For instance, in Norway, green roof systems are still used as a thermal isolation system [5]. (a) Vegetation layer: This can be composed of different biological species, usually plants or even trees. The respective species selection is based on availability, weather, maintenance conditions, and the substratum's depth, which depends on the structural building capacity [9]. Experience shows that green roof plants need to be chosen with care, as not all plants are suitable for growing in this way. For example, when choosing plants for a green roof, they need to be able to withstand wind and frost, be drought-resistant, tolerate living in poor soil, and require low maintenance. Also, green roof plants should also be attractive and offer food and shelter for wildlife. Although, many plants were proposed for green roofs in the literature, there is a need for comparisons among various types of plants to provide design guidelines for selecting the most appropriate vegetative layer for a given green roof [10]. Interesting research is now being conducted to analyze CO2 reduction and O2 production from different plant species; hence, this factor should also be included in the plant selection process for increasing the positive environmental impact of the biological coatings in lifecycle analysis.
(b) Substratum: The function of this layer is to physically support the vegetation and to supply the required nutrients for their development. Also, this layer should have the ability to store and gradually release excess rainwater, keeping enough moisture to later reduce maintenance activities. This layer is usually based on a combination of minerals with organic matter and other nutrients such as nitrogen, phosphorus, potassium, and magnesium [11,12]. Physical properties of a typical substratum are presented in Table 1. Table 1. Physical properties of a typical substratum used in green roofs [13].

Physical Property
One-Layer System Multi-Layered System Water retention Minimum 20% Minimum 35% Water permeability Minimum 60 mm/min Minimum 0.6 mm/min Air content (fully saturated) Minimum 10% Minimum 10% pH 6.5 to 9.5 6.5 to 8.0 (c) Filter: This building protective layer is usually a geotextile; its main function is to block the substratum's material flow to the drainage layer and to keep it in the right position. This geotextile, chemically neutral, has to be resistant to acid and alkaline attacks [3].
(d) Drainage layer: Also known as a drainage system, this layer is composed of granular-based material layers and pipelines. This is the key to appropriate vegetation growth and control of building water-associated problems such as filtrations. In particular, the drainage layer is responsible for the equilibrium between water excess and scarcity. It is normally formed by gravel, which is a natural  Ovacen (2014) [8]. Table 1. Physical properties of a typical substratum used in green roofs [13].

Physical Property One-Layer System Multi-Layered System
Water retention Minimum 20% Minimum 35% Water permeability Minimum 60 mm/min Minimum 0.6 mm/min Air content (fully saturated) Minimum 10% Minimum 10% pH 6.5 to 9.5 6.5 to 8.0 (c) Filter: This building protective layer is usually a geotextile; its main function is to block the substratum's material flow to the drainage layer and to keep it in the right position. This geotextile, chemically neutral, has to be resistant to acid and alkaline attacks [3].
(d) Drainage layer: Also known as a drainage system, this layer is composed of granular-based material layers and pipelines. This is the key to appropriate vegetation growth and control of building water-associated problems such as filtrations. In particular, the drainage layer is responsible for the equilibrium between water excess and scarcity. It is normally formed by gravel, which is a natural crushed rock [14]. Research is now being conducted to reduce the use of natural aggregates to minimize the negative environmental impact of using nonrenewable resources in green roofs [15].
(e) Anti-root barrier: This high-density polyethylene layer protects the waterproof membrane from possible tearing caused by roots. It also functions as a second protection from the substratum to the building [3].
(f) Waterproof membrane: This layer is used to block the water flow to the building slab. Some of the most used waterproof membranes are thermo-polymer elastomers such as EPDM (ethylene propylene diene) or thermoplastic polyolefin (TPO), which are also good as root barriers [14].

Classification
Although there are different criteria to classify green roofs, the most common classification at the international level is based on the characteristics of the vegetation layer, as presented below [16].
(a) Extensive: This green roof type uses plants with low-moisture needs ( Figure 2). As a consequence, it uses a substratum with thickness between 5 and 15 cm. This biological coating is usually implemented on places with difficult access for pedestrians, can be moistened with rainwater, Coatings 2020, 10, 69 4 of 14 and can be placed on existing structures when the building has a proper design (i.e., roof live load considered as indicated in design codes). Its weight varies from 60 to 140 kg/m 2 [17].

Classification
Although there are different criteria to classify green roofs, the most common classification at the international level is based on the characteristics of the vegetation layer, as presented below [16].
(a) Extensive: This green roof type uses plants with low-moisture needs ( Figure 2). As a consequence, it uses a substratum with thickness between 5 and 15 cm. This biological coating is usually implemented on places with difficult access for pedestrians, can be moistened with rainwater, and can be placed on existing structures when the building has a proper design (i.e., roof live load considered as indicated in design codes). Its weight varies from 60 to 140 kg/m 2 [17].
(b) Intensive: The green roofs belonging to this type have the possibility of using a great variety of plants and even tree species ( Figure 3). Hence, they generally require substratum of great depth, usually superior to 15 cm. Based on the last aspect, these coatings can allow pedestrian access, require artificial water irrigation systems, and are placed on structures specifically designed to bear these additional loads. Their weight varies from 250 to 400 kg/m 2 [20]. (c) Semi-intensive: This type of green roof considers an intermediate system between intensive and extensive systems with species that grow over a 10 to 30-cm-deep substratum ( Figure 4). It allows partial pedestrian access and requires artificial irrigation systems. Its weight varies from 60 to 140 kg/m 2 [23].  [18]; (b) Nanyang Technological University, Singapore [19].
(b) Intensive: The green roofs belonging to this type have the possibility of using a great variety of plants and even tree species ( Figure 3). Hence, they generally require substratum of great depth, usually superior to 15 cm. Based on the last aspect, these coatings can allow pedestrian access, require artificial water irrigation systems, and are placed on structures specifically designed to bear these additional loads. Their weight varies from 250 to 400 kg/m 2 [20].

Classification
Although there are different criteria to classify green roofs, the most common classification at the international level is based on the characteristics of the vegetation layer, as presented below [16].
(a) Extensive: This green roof type uses plants with low-moisture needs ( Figure 2). As a consequence, it uses a substratum with thickness between 5 and 15 cm. This biological coating is usually implemented on places with difficult access for pedestrians, can be moistened with rainwater, and can be placed on existing structures when the building has a proper design (i.e., roof live load considered as indicated in design codes). Its weight varies from 60 to 140 kg/m 2 [17].
(b) Intensive: The green roofs belonging to this type have the possibility of using a great variety of plants and even tree species ( Figure 3). Hence, they generally require substratum of great depth, usually superior to 15 cm. Based on the last aspect, these coatings can allow pedestrian access, require artificial water irrigation systems, and are placed on structures specifically designed to bear these additional loads. Their weight varies from 250 to 400 kg/m 2 [20].  (c) Semi-intensive: This type of green roof considers an intermediate system between intensive and extensive systems with species that grow over a 10 to 30-cm-deep substratum ( Figure 4). It allows partial pedestrian access and requires artificial irrigation systems. Its weight varies from 60 to 140 kg/m 2 [23]. (c) Semi-intensive: This type of green roof considers an intermediate system between intensive and extensive systems with species that grow over a 10 to 30-cm-deep substratum ( Figure 4). It allows partial pedestrian access and requires artificial irrigation systems. Its weight varies from 60 to 140 kg/m 2 [23].

Hydraulic Performance
Based on the non-permeable large areas generated by conventional urban development, green roofs might reduce the negative environmental impact, particularly by partially recovering the natural water cycle in a basin. These biological coatings on buildings have the ability to store water

Hydraulic Performance
Based on the non-permeable large areas generated by conventional urban development, green roofs might reduce the negative environmental impact, particularly by partially recovering the natural water cycle in a basin. These biological coatings on buildings have the ability to store water in the substratum, in which a fraction is absorbed by plants and returned to the atmosphere via evaporation and transpiration processes [26,27]. In addition to the efficient management of the evapotranspiration water fraction, another water fraction is absorbed, infiltrated, and stored inside the system layers. Depending on the precipitation intensity, substratum type, and depth, studies showed that green roofs have the capacity to store between 40% and 80% of annual precipitation that falls on them. Additionally, it was reported that a 12-cm layer takes 12 h to start releasing the stored water during a rain event, and it continues releasing it during the next 21 h, approximately [28]. Based on Kok et al. (2016) [29] and Masseroni and Cislagi (2016) [30], the discharge peak from a storm can be reduced by 26% to 35%. Hence, it is deduced that green roofs, in addition to reducing the magnitude of water volume captured by rain drainage systems, also delay the water discharge time, because the substratum and other layers require time to saturate and discharge the stored water, reducing potential flood impacts when used intensively at the urban scale [26]. In Germany, Spain, Holland, England, and the United States, the performance of different green roof types was evaluated toward the quantity and quality of urban runoff water. For example, van Woert et al. (2005) [14] indicated that the implementation of green roofs reduces the excess of runoff water, and Stovin (2009) [17] found 34% rainwater retention in the 11 precipitation events monitored. In this article, the retention capacity, measured as the water percentage stored in different green roofs compared with the water that passes through them, is presented in Table 2. From the results, it can be observed that retention capacities can range from 45% to 78% for different green roof systems. The variation in the results is associated with the substratum depth, initial water content, vegetation age and type, and precipitation intensity and distribution. Vegetated roofs consisted of the following layers: drainage (15 mm), water retention fabric (7.5 mm), additional retention fabric as vegetation carrier (7.5 mm), and 25 mm of growing media. Twelve additional roof platforms were used to examine roof slope (6 with slope 2% and 6 with slope 6.5%) and media depth (for 2% slope 25 mm and 40 mm; for 6% 40 mm and 60 mm).  On the other hand, as mentioned previously for the drainage layer, recent research is evaluating new material sources to replace natural aggregates traditionally used on green roofs. In this case, drainage systems composed of PEAD plates (recycled high-density polyethylene), reused PET bottles (polyethylene terephthalate), and vulcanized recycled rubber particles were evaluated and compared with basal gravel, the conventional system used as a drainage system ( Figure 5). The results indicated that, for low-intensity precipitations (simulations of 1 and 4 mm) and duration of a couple of hours, granular drainage systems (rubber and gravel) were very efficient, as they kept almost all precipitation water, and retention coefficients were about 100%. For the same conditions, reused PET bottles and PEAD plates retained half of the simulated precipitation, while retention coefficients were about 50%. For higher-intensity precipitations (simulations around 50 mm), the retention coefficients were near 30% for all green roof systems, except for the system composed of PET bottles, which had a retention capacity near to 0% [15].
Using the same experimental methodology but including a simulation for the Singapore precipitation regime, extensive green roofs of 12 cm were not enough to obtain a significant retention coefficient. The authors concluded that a proper green roof design has to be done for each precipitation condition [38]. Hence, more research has to be conducted to improve the retention coefficients in green roofs. On the other hand, as mentioned previously for the drainage layer, recent research is evaluating new material sources to replace natural aggregates traditionally used on green roofs. In this case, drainage systems composed of PEAD plates (recycled high-density polyethylene), reused PET bottles (polyethylene terephthalate), and vulcanized recycled rubber particles were evaluated and compared with basal gravel, the conventional system used as a drainage system ( Figure 5). The results indicated that, for low-intensity precipitations (simulations of 1 and 4 mm) and duration of a couple of hours, granular drainage systems (rubber and gravel) were very efficient, as they kept almost all precipitation water, and retention coefficients were about 100%. For the same conditions, reused PET bottles and PEAD plates retained half of the simulated precipitation, while retention coefficients were about 50%. For higher-intensity precipitations (simulations around 50 mm), the retention coefficients were near 30% for all green roof systems, except for the system composed of PET bottles, which had a retention capacity near to 0% [15].
Using the same experimental methodology but including a simulation for the Singapore precipitation regime, extensive green roofs of 12 cm were not enough to obtain a significant retention coefficient. The authors concluded that a proper green roof design has to be done for each precipitation condition [38]. Hence, more research has to be conducted to improve the retention coefficients in green roofs.

Thermal Performance
In general, green roof systems produce a high thermal isolation effect in buildings by increasing thermal inertia. This interesting property is mainly due to the green roof components which work as thermal isolation chambers, preventing and delaying the temperature amplitudes due to their high thermal capacities in hot regions, as well as heat winter loss reduction in cold regions. Indeed, a green roof can reduce energy consumption from air conditioning in buildings to 50% [39]. In addition to the good thermal performance of green roofs, they also reduce the surrounding environment's temperature via vegetative physiological processes such as evapotranspiration, photosynthesis, and ability to store water [39]. Scale experiments in outdoor conditions on different green roof systems developed in Cali (Colombia) showed that, during days where the environmental temperature was high (over 35° C), there was a temperature reduction between 10.6 and 11.7 °C below the green roof prototypes [15]. These results are similar to those reported in the literature for tropical climates. More details about the prototype evaluation can be seen in Figure 6.  Table 3 summarizes results from several researchers about the thermal performance of different green roof systems. From these results, a trend to evaluate the thermal behavior of green roof systems during warm and cold seasons can be identified. Although some researchers showed a more efficient

Thermal Performance
In general, green roof systems produce a high thermal isolation effect in buildings by increasing thermal inertia. This interesting property is mainly due to the green roof components which work as thermal isolation chambers, preventing and delaying the temperature amplitudes due to their high thermal capacities in hot regions, as well as heat winter loss reduction in cold regions. Indeed, a green roof can reduce energy consumption from air conditioning in buildings to 50% [39]. In addition to the good thermal performance of green roofs, they also reduce the surrounding environment's temperature via vegetative physiological processes such as evapotranspiration, photosynthesis, and ability to store water [39]. Scale experiments in outdoor conditions on different green roof systems developed in Cali (Colombia) showed that, during days where the environmental temperature was high (over 35 • C), there was a temperature reduction between 10.6 and 11.7 • C below the green roof prototypes [15]. These results are similar to those reported in the literature for tropical climates. More details about the prototype evaluation can be seen in Figure 6.

Thermal Performance
In general, green roof systems produce a high thermal isolation effect in buildings by increasing thermal inertia. This interesting property is mainly due to the green roof components which work as thermal isolation chambers, preventing and delaying the temperature amplitudes due to their high thermal capacities in hot regions, as well as heat winter loss reduction in cold regions. Indeed, a green roof can reduce energy consumption from air conditioning in buildings to 50% [39]. In addition to the good thermal performance of green roofs, they also reduce the surrounding environment's temperature via vegetative physiological processes such as evapotranspiration, photosynthesis, and ability to store water [39]. Scale experiments in outdoor conditions on different green roof systems developed in Cali (Colombia) showed that, during days where the environmental temperature was high (over 35° C), there was a temperature reduction between 10.6 and 11.7 °C below the green roof prototypes [15]. These results are similar to those reported in the literature for tropical climates. More details about the prototype evaluation can be seen in Figure 6.  Table 3 summarizes results from several researchers about the thermal performance of different green roof systems. From these results, a trend to evaluate the thermal behavior of green roof systems during warm and cold seasons can be identified. Although some researchers showed a more efficient Figure 6. Green roof prototypes evaluated using reused and recycled materials in their drainage systems [15]. Table 3 summarizes results from several researchers about the thermal performance of different green roof systems. From these results, a trend to evaluate the thermal behavior of green roof systems during warm and cold seasons can be identified. Although some researchers showed a more efficient behavior in hot weather compared to cold weather, some others showed almost the same behavior. Based on the review, more research should be conducted when the substratum is in saturated conditions during cold weather; then, the isolating effect of green roofs seems to be significantly reduced [40]. Table 3. Relevant work related to the thermal performance of green roofs.

Approach Proposed Major Findings Reference
Long-term experimental analysis to compare thermal performance from a conventional roof respect to a green roof.
Under typical Mediterranean climate conditions, the green roof system provides different behavior according to the season. The green roof performance is meaningful during warm seasons, while this technology does not show a significant difference with a conventional roof during cold seasons. [41] Model based on energy balanced equations expressed for foliage soil media and simulations.
During the exposure to warm environmental conditions, the evapotranspiration provides evaporative cooling that increases the thermal resistance of a green roof system. [42] Analysis of transmittance and heat flux.
During the winter season, the green roof provides further isolation even in saturated conditions. In the summer, the green roof decreases incoming heat fluxes and ceiling temperatures. [43] Study during warm and cold periods with 3 different roof conditions.
During warm periods, the evaluated green roof reduced heat gain over 90%. During the cold period, the evaluated green roof system reduced heat loss between 70% and 84%. These results are comparable to those obtained with conventional ceramic and metallic roofs. [44] Numerical model and experimental validation for energy savings (comparison approach).
After evaluating extensive, semi-intensive, and intensive green roofs, the extensive one involved higher cooling energy demand than semi-intensive (2.8-fold) and intensive ones (5.9-fold). [45] Mathematical model and experimental validation.
Cooling potential of green roofs can be around 3.02 kWh per day for an LAI (leaf area index) of 4.5. This is enough to maintain an average room temperature of 25.7 • C. [46] Experimental study for measuring energy savings in cooling for Mediterranean continental climate.
Energy savings (16.7%) for cooling and an increase in energy consumption for heating (11.1%) were observed compared to conventional flat roofs. These results correspond to a 20% to >85% area covered by vegetation. [47,48]

Certifications for Sustainable Construction
In order to evaluate the impact of green roof system implementation on major sustainable construction certifications, a preliminary qualitative impact analysis must be developed to assess the decision process. A suggested impact of the green roof system that varies from not related to high is presented in Tables 4 and 5, which summarize the impact of green roofs on the overall certification systems based on the degree of relevance of green roof benefits concerning the main criteria for each system. Thus, criteria regarding energy and water involve a strong (high impact) relationship (efficiency, cost reduction), while indoor and outdoor environment improvements involve indirect benefits (medium impact). Finally, innovation is considered as a less relevant criterion since its measurement in buildings is still subjective in terms of green roofs, and it can be interpreted as an indirect benefit or consequence. Table 4. Suggested impact of green roof systems in sustainable construction certifications.

High
In this level, a green roof has a direct influence on the evaluation criteria providing points in relation to conventional roofs. There is a strong relation to energy and water efficient use criteria.

Medium
Although, in this level, the implementation of green roof systems has an impact on the criteria, complementary technologies are required to be explicitly valued.

Low
At this level, the green roof implementation is related to the criteria, but it does not necessarily mean a performance improvement.

Not related
The implementation of green roof systems has no impact on the evaluation criteria in this level. interpreted as an indirect benefit or consequence.

High
In this level, a green roof has a direct influence on the evaluation criteria providing points in relation to conventional roofs. There is a strong relation to energy and water efficient use criteria.

Medium
Although, in this level, the implementation of green roof systems has an impact on the criteria, complementary technologies are required to be explicitly valued.

Low
At this level, the green roof implementation is related to the criteria, but it does not necessarily mean a performance improvement. Not related The implementation of green roof systems has no impact on the evaluation criteria in this level. From the analysis of the impact of green roofs on construction certifications, it is clear that the British Accreditation, BREEAM (Building Research Establishment Environmental Assessment Method), considers the use of this technology more relevant. BREEAM assumes a high impact of green roofs in energy, water, pollution, land use, and ecology. Contrary to BREEAM, the Japanese Accreditation, CASBEE (Comprehensive Assessment System for Built Environment Efficiency), does not consider the use of green roofs crucial for sustainable construction. However, this accreditation system gives a reasonable importance to green roofs regarding energy resources and materials and in outdoor environments. On the other hand, LEED (Leadership in Energy and Environment Design), ; medium impact: interpreted as an indirect benefit or consequence. Table 4. Suggested impact of green roof systems in sustainable construction certifications.

High
In this level, a green roof has a direct influence on the evaluation criteria providing points in relation to conventional roofs. There is a strong relation to energy and water efficient use criteria.

Medium
Although, in this level, the implementation of green roof systems has an impact on the criteria, complementary technologies are required to be explicitly valued.

Low
At this level, the green roof implementation is related to the criteria, but it does not necessarily mean a performance improvement. Not related The implementation of green roof systems has no impact on the evaluation criteria in this level. From the analysis of the impact of green roofs on construction certifications, it is clear that the British Accreditation, BREEAM (Building Research Establishment Environmental Assessment Method), considers the use of this technology more relevant. BREEAM assumes a high impact of green roofs in energy, water, pollution, land use, and ecology. Contrary to BREEAM, the Japanese Accreditation, CASBEE (Comprehensive Assessment System for Built Environment Efficiency), does not consider the use of green roofs crucial for sustainable construction. However, this accreditation system gives a reasonable importance to green roofs regarding energy resources and materials and in outdoor environments. On the other hand, LEED (Leadership in Energy and Environment Design), ; low impact: interpreted as an indirect benefit or consequence.

High
In this level, a green roof has a direct influence on the evaluation criteria providing points in relation to conventional roofs. There is a strong relation to energy and water efficient use criteria.

Medium
Although, in this level, the implementation of green roof systems has an impact on the criteria, complementary technologies are required to be explicitly valued.

Low
At this level, the green roof implementation is related to the criteria, but it does not necessarily mean a performance improvement. Not related The implementation of green roof systems has no impact on the evaluation criteria in this level. From the analysis of the impact of green roofs on construction certifications, it is clear that the British Accreditation, BREEAM (Building Research Establishment Environmental Assessment Method), considers the use of this technology more relevant. BREEAM assumes a high impact of green roofs in energy, water, pollution, land use, and ecology. Contrary to BREEAM, the Japanese Accreditation, CASBEE (Comprehensive Assessment System for Built Environment Efficiency), does not consider the use of green roofs crucial for sustainable construction. However, this accreditation system gives a reasonable importance to green roofs regarding energy resources and materials and in outdoor environments. On the other hand, LEED (Leadership in Energy and Environment Design), ; not related: interpreted as an indirect benefit or consequence.

High
In this level, a green roof has a direct influence on the evaluation criteria providing points in relation to conventional roofs. There is a strong relation to energy and water efficient use criteria.

Medium
Although, in this level, the implementation of green roof systems has an impact on the criteria, complementary technologies are required to be explicitly valued.

Low
At this level, the green roof implementation is related to the criteria, but it does not necessarily mean a performance improvement. Not related The implementation of green roof systems has no impact on the evaluation criteria in this level. From the analysis of the impact of green roofs on construction certifications, it is clear that the British Accreditation, BREEAM (Building Research Establishment Environmental Assessment Method), considers the use of this technology more relevant. BREEAM assumes a high impact of green roofs in energy, water, pollution, land use, and ecology. Contrary to BREEAM, the Japanese Accreditation, CASBEE (Comprehensive Assessment System for Built Environment Efficiency), does not consider the use of green roofs crucial for sustainable construction. However, this accreditation system gives a reasonable importance to green roofs regarding energy resources and materials and in outdoor environments. On the other hand, LEED (Leadership in Energy and Environment Design), . From the analysis of the impact of green roofs on construction certifications, it is clear that the British Accreditation, BREEAM (Building Research Establishment Environmental Assessment Method), considers the use of this technology more relevant. BREEAM assumes a high impact of green roofs in energy, water, pollution, land use, and ecology. Contrary to BREEAM, the Japanese Accreditation, CASBEE (Comprehensive Assessment System for Built Environment Efficiency), does not consider the use of green roofs crucial for sustainable construction. However, this accreditation system gives a reasonable importance to green roofs regarding energy resources and materials and in outdoor environments. On the other hand, LEED (Leadership in Energy and Environment Design), BEAM Plus (Building Environmental Assessment Method), and ESGB (Evaluation Standard for Green Building) give a medium impact of using green roofs in their accreditation systems.
In general, for the accreditation systems, green roofs are considered as a low-level innovation, unless they are integrated with other technologies such as renewable energies, cogeneration, and so on. That is due to the fact that green roof systems were implemented in many projects worldwide in the past. However, the real fact is that the impact of the implementation was never properly quantified for the hydraulic and thermal performances. Thus, other environmental benefits such as reduction of the heat island effect and sewage system load, improvement of runoff water and air quality, and reconstruction of natural landscapes remain to be estimated for each green roof case.
As a reference for the reader, a description of major construction certifications is included in Table 6. This certification has 3 levels: 1 start: over 33% marks 2 starts: over 67% marks 3 starts: over 80% marks

Conclusions and Research Challenges
Although green roofs were used in major ancient buildings, their implementation in modern infrastructure was restricted until Le Corbusier included them among the main building conceptual design points. However, green roofs became more a landscaping action rather than a technical solution with significant environmental, technical, economic, and social benefits. This is mainly due to the fact that there was no proper design, construction, and maintenance of these biological coatings. Intensive, semi-intensive, and extensive green roofs require a proper conceptualization that allows identification of each component and their functions for the overall performance.
Currently, the economic benefits from reducing building energy consumption and sewage system load attracted the attention of urban developers for implementing green roofs on buildings. This review showed that, only regarding the hydraulic performance, retention capacities vary from 45% to 78% for different green roof systems reported in the literature. Thus, more research has to be conducted to estimate the real impact of substratum depth, initial water content, vegetation age and type, and precipitation regime on the hydraulic performance. Similarly, although thermal gradients up to 10 • were reported using green roofs in tropical climates, more research on the application of green roofs in cold climates has to be done. In this case, when the substratum is saturated, the isolating effect is significantly reduced. Therefore, combined models that integrate hydraulic and thermal performances should be developed. Thermal and hydraulic performances mostly control the complete green roof system.
Even though most accreditation systems for sustainable constructions do not give significant importance to green roof implementation, an accurate quantification of the environmental, technical, economic, and social benefits of green roofs might help to improve this situation, as well as address the lack of legislation for this technology in most countries.
Finally, from this review of the state of the art, research challenges that can be undertaken in the short or medium term toward an effective implementation of green roofs are presented in Table 7. Table 7. Research challenges facing green roofs reported in the literature.

Component Research Topics
Vegetation layer