The influence of dual-substrate-layer extensive green roofs on rainwater runoff quantity and quality

- Dual-substrate-layer (DSL) extensive green roofs were effective at retaining rainfall.
- DSL green roofs were sinks for organics, heavy metals and nitrogen in all cases.
- DSL green roofs were able to mitigate mild acid rain.
- DSL green roofs' runoff did not show any first flush effect.
- DSL green roofs were practical in application for retaining rainwater in urban areas.

This study investigates the ability of dual-substrate-layer extensive green roofs to retain rainwater and reduce pollutant leaching. The substrates in dual-substrate-layer green roofs consist of an upper organic nutrition layer for plant growth and a lower inorganic adsorption layer for water retention and pollutant reduction. One traditional single-substrate-layer extensive green roof was built for comparison with dual-substrate-layer green roofs. During the experimental period, dual-substrate-layer green roofs supported better natural vegetation growth, with coverage exceeding 90%, while the coverage in single-substrate-layer green roof was over 80%. Based on the average retention value of the total rainfall for four types of simulated rains (the total rainfall depth (mm) was 43.2, 54.6, 76.2 and 86.4, respectively), the dual-substrate-layer green roofs, which used the mixture of activated charcoal with perlite and vermiculite as the adsorption substrate, possessed better rainfall retention performance (65.9% and 55.4%) than the single-substrate-layer green roof (52.5%). All of the dual-substrate-layer green roofs appeared to be sinks for organics, heavy metals and all forms of nitrogen in all cases, while acted as sources of phosphorus contaminants in the case of heavy rains. In consideration of the factors of water retention, pollution reduction and service life of the green roof, a mixture of activated charcoal and/or pumice with perlite and vermiculite is recommended as the adsorption substrate. The green roofs were able to mitigate mild acid rain, raising the pH from approximately 5.6 in rainfall to 6.5-7.6 in green roof runoff. No signs of a first flush effect for phosphate, total phosphorus, ammonia nitrogen, nitrate nitrogen, total nitrogen, organics, zinc, lead, chromium, manganese, copper, pH or turbidity were found in the green roof runoff. Cost analysis further proved the practicability of dual-substrate-layer green roofs in retaining rainwater, and their long-term rainwater runoff quantity and quality performance in urban environments merit further investigation.

245