Modeling stormwater transport through unsaturated green roof substrates
Licentiate thesis, 2018
In recent decades there has been an increase in research regarding green roofs and similar technologies. This increased interest is driven by the requirements of urban development and its effects both on humans and the environment. Additionally, the predicted increase in weather severity in the future is raising concerns on the capabilities of urban environments and their stormwater management systems to cope with the increase.
Green roofs can be used as a space-conscious solution for improving stormwater management in urban areas as well as contributing to, for example, building protection and pollution and noise reduction. In order to fully utilize them effectively for stormwater runoff reduction it is necessary to quantify their effect and optimize their performance in a given climate. This optimization can take the form of placement on structures or by design within the green roof construction itself.
This work focuses on optimization of design by applying computational fluid dynamics and lattice Boltzmann theory to the soil growth substrate. Computational fluid dynamics is used for modeling the flow through the green roof growth substrate (soil layer) at the macrososcopic scale while a lattice Boltzmann model is applied to the mesoscopic (soil particle) scale. Using these methods, the efficacy at water retention and drainage of given soil particles and full-sized green roofs can be determined. This work covers the framework covering both scales however the methodology is applied only to the mesoscopic scale.
The focus within the mesoscopic scale is primarily on the hydrophilicity of the particles in the soil and its impact on liquid imbibition. Also included is an exploration on the liquid-air interfacial area and liquid penetration depth to aid in the analysis of the results. The findings of the study suggest particle hydrophilicity plays an important role in the imbibition process, particularly under light to medium rainfall conditions. In addition a pore blocking phenomenon is identified which requires further analysis. Finally, plans for future work and the closure of the two-framework methodology proposed in this work is discussed.