The role of living roofs in holistic stormwater management systems - Living Roofs in Integrated Urban Water Systems - page 30

events produce relatively little rainfall. In Auckland, 80 percent of 396 events
were less than 15 mm of rainfall, while 90 percent of events were less than
25 mm over 28 months of monitoring. These events were shown to be satisfac-
torily retained by extensive living roofs with appropriately designed growing
mediums providing water-holding capacities in excess of ~11-20 mm. When
monitored concurrently there was no statistical difference in four-month cumula-
tive retention eficiencies amongst these living roofs ranging ~60 mm-150 mm
growing media depth (Fassman-Beck et al. 2013a).
From these four extensive living roofs in Auckland, storms less than approxi-
mately 10 mm did not produce meaningful runoff (more than a couple of mm).
In comparison, Palla et al. (2011) identiied that storms less than 8 mm did not
produce runoff from a 200 mm intensive living roof in Genoa, Italy. A comparison
of median per event retention for storms greater than 25 mm at these sites shows
that there is no substantial difference in performance, particularly when variabil-
ity is taken into account ( Table 2.1 ). Ultimately, the increased initial and long-term
costs associated with intensive living roofs are not justiied in terms of stormwa-
ter management.
Assembly design characteristics such as the presence/absence of water reten-
tion cells or fabrics, effects of irrigation, and roof slope have only minimally been
investigated, with inconsistent results amongst studies (Getter et al. 2007; Hatha-
way et al. 2008; VanWoert et al. 2005; Villarreal and Bengtsson 2005). Intuitively,
a steeper slope should provide faster drainage and less detention. Qualitative
observation of sloped living roofs indicates less water is retained near the ridgeline
compared to eaves, with clear impacts for plant health (Fassman et al. 2013). Any
single factor alone may not have an inluence on stormwater mitigation, but the
combination of slope and media depth or composition, for example, may be
important. Other assembly inluences may include quality control during construc-
tion and long-term changes in media physical and chemical composition (Durhman
et al. 2007; Fassman et al. 2013; Grifin 2014). Only a single study to date has
compared concurrent monitoring of a modular tray system against a built-in-place
assembly (Carson et al. 2013). The researchers hypothesized that the modular tray
system's constricted drainage design enhanced retention performance by limiting

Table 2.1 Empirical evidence from one intensive and four extensive living roofs: median (standard deviation) large storm
per event retention *

Rainfall depth range (mm) % retention per event according to growing media depth (mm)

200 (intensive) 150 (extensive) 100 (extensive, two sites) ~60 (extensive)

25-69 51 (36) 48 (19) 31 (25), 58 (16) 49

>70 11 (8) 39 (29) 22 (27), 32 n/a

Note
* Performance statistics for the intensive roof are in reference to rainfall, whereas the extensive roofs are in reference to runoff from a
conventional roof surface. The latter case would provide slightly lower percentage retention than compared to rainfall. Data from
Fassman-Beck et al. (2013a) and Palla et al. (2011).