4 Living roofs in 3 locations: Does configuration affect runoff mitigation?

Abstract

Summary Four extensive living roofs and three conventional (control) roofs in Auckland, New Zealand have been evaluated over periods of 8 months to over 2 yrs for stormwater runoff mitigation. Up to 56% cumulative retention was measured from living roofs with 50–150 mm depth substrates installed over synthetic drainage layers, and with >80% plant coverage. Variation in cumulative %-retention amongst sites is attributed to different durations of monitoring, rather than actual performance. At all sites, runoff rarely occurred at all from storms with less than 25 mm of precipitation, from the combined effects of substrates designed to maximize moisture storage and because >90% of individual events were less than 25 mm. Living roof runoff depth per event is predicted well by a 2nd order polynomial model ( R 2 = 0.81), again demonstrating that small storms are well managed. Peak flow per event from the living roofs was 62–90% less than a corresponding conventional roof’s runoff. Seasonal retention performance decreased slightly in winter, but was nonetheless substantial, maintaining 66% retention at one site compared to 45–93% in spring-autumn at two sites. Peak flow mitigation did not vary seasonally. During a 4-month period of concurrent monitoring at all sites, varied substrate depth did not influence runoff depth (volume), %-retention, or %-peak flow mitigation compared to a control roof at the same site. The magnitude of peak flow was greater from garden shed-scale living roofs compared to the full-scale living roofs. Two design aspects that could be manipulated to increase peak flow mitigation include lengthening the flow path through the drainage layer to vertical gutters and use of flow-retarding drainage layer materials.