**5. Conclusions**

In this article, scale prototypes of three semi-intensive green roof systems with different types of drainage systems made out of recycled and reused materials (rubber, trays, and bottles) were successfully evaluated and compared to a traditional green roof system that used natural aggregates of (gravel) as its drainage layer.

In terms of hydraulic performance, the behavior of the systems using trays, bottles, rubber, and gravel (reference) was analyzed in different precipitation regimes (typical and intense). The results showed that for "typical precipitations," the granular drainage systems (gravel and rubber) were very efficient because they retained all the precipitation (retention coefficients close to 1.0), while the systems composed of module containers (bottles and trays) retained approximately half of the water supplied at that level of precipitation (retention coefficients close to 0.5). For the "intense precipitation," the coefficients reached values close to 0.3 for all roof systems, except for the gravel system, which reduced its water storage capacity to almost zero (0.08). In general, this study demonstrates the enormous potential of all the green roofs analyzed in this study to reduce the maximum flow of runoff water volumes, as they could increase retention time when they are implemented in large areas at the urban level.

With respect to the thermal behavior, it was possible to verify the effect of temperature reduction of all the roofing systems. During days when the ambient temperature was very high (approximately 50 ◦C), a reduction in temperature that ranged from 10.6 to 11.7 ◦C was found for the investigated green roof systems. This makes it evident that the use of green roof systems with drainage layers made out of recycled and reused materials have, like gravel roofs, the potential to reduce the consumption of electrical energy in buildings derived from artificial cooling.

Finally, a reduction from 33% to 72% in weight per area (dead load) of the green roof was observed when using recycled and reused materials compared to natural materials in the drainage layers. This is significantly important because the ease and costs of implementation of green roofs depends on the structural condition of the building. Therefore, for the load capacity of an existing building, the dead and live loads must be assessed in order to verify whether, with the increase in the dead load generated by the weight of the green roof, the building can withstand the loads added to it without affecting its resistance, as well as if it complies with the specifications of the building construction code. Otherwise, a structural reinforcement must be designed to guarantee safety.

For new buildings, a structural calculation must be made according to all the loads that act on the structural system and that come from the weight of all the permanent elements in the construction (dead loads), the occupants and their belongings (live loads), environmental effects, differential settlements, and dimensional change restriction following current regulations. The last indicates that due to the relatively lower density and lower absorption capacity of the recycled and reused materials evaluated in this research, the implementation of green roof systems, both in existing and new buildings, would be easier and cheaper.

Future research works will be oriented to evaluate the sustainability and economic performance of green roofs based on recycled or reused materials. Considering environmental impacts in terms of material lifecycle and lifecycle costs derived from installing, maintaining, upgrading, and disposing constructive components. All those parameters need to be included to generate holistic evaluations to facilitate the decision-making during green roof design and construction.

**Author Contributions:** Conceptualization, methodology, and editing, J.M. and A.M.-R.; components, classification, and technical performance, A.N. and A.C. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors thank Dos Mundos (Cali, Colombia) and Pontificia Universidad Javeriana Cali (Colombia) for the technical support given during the experiments reported in this article.

**Conflicts of Interest:** The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
