b'GREEN ROOF SYSTEMSrelated linearly, and the rainwater reten-tion potential of the green roof is only about 21 per cent, while in the summer, the rainwater retention capacity increases as high as 100 per cent (Fig. 3(b)). The green roof also delays rain runoffs for hours, which helps to shave and shift the peak rainwater manage-ment system load.DURING SUMMER DAYS (FIG. 5(A)), THE ENTIRE HEAT TRANSFER THROUGH THE GR AND GM ROOFS CAN BE CHARACTERIZED AS HEAT GAIN WHILE BOTH HEAT GAINS AND HEAT LOSSES ARE EXPERIENCED THROUGH THE CR ROOF. The GR sheathing board temperature is more stable than the CR sheathing due to the thermal mass effect of the green roof and the evapotranspiration and shading effect of the plants. In summer (Fig. 4(a)), the GR delays the peak temperature and the fluctuation range of the sheathing board temperature in the GR is 6oC, whereas the CR is 28 oC. Since the waterproofing membranes FIG. 2. SENSOR LAYOUT IN GR (A), GM (B) AND CR (C). are directly installed on top of the sheathing boards, the waterproofing space prior to filling the area with theare installed to measure the solar radia- membrane in the GR experiences less growing media. Fertilizer is appliedtion received and reflected, and the raintemperature fluctuation and lower peak every week and irrigation is providedrunoff from the test roofs, respectively.temperature, resulting in longer service every other day for two months untilHeat flux transducers are installed onlife when compared to that of the CR. the plants are fully established.top of the vapour barrier of the threeA green roof keeps the sheathing board warmer in the winter and thereby roof modules. An additional heat fluxreduces the risk of condensation when Fig. 2(a) to (c) present the sensorstransducer is placed at the growingcompared to the CR sheathing (Fig. layout in GR, CR and GM, respec- media bottom of the GR and GM4(b)).tively. Type T Thermocouples and GS1roofs. Two additional thermocouples Moisture Content Sensors are installedare installed in each roof module toDuring summer days (Fig. 5(a)), the at the top, middle and bottom layersmeasure the roof sheathing and entire heat transfer through the GR of the growing media. Additionalceiling temperatures. and GM roofs can be characterized as Thermocouples are installed at theheat gain while both heat gains and bottom of the drainage mat and the rainThe moisture content in the growingheat losses are experienced through runoff collection slit. A set of tempera- media of the green roofs fluctuatesthe CR roof. The peak heat gain in the ture and relative humidity sensors arebetween 10 per cent and 35 per centCR matches the outdoor temperature installed at one inch and 12 inchesfrom September to mid-October andpeak time (4 p.m.), while the peak heat above the respective roof surfaces.mid-May to June (Fig. 3(a)). Duringfluxes in GR and GM lags by five hours Pyranometer type CMP3-L and tippingthe wet period, October to mid-May,(9 p.m.). When compared to CR, the bucket rain gauges type E-802-KIPP004the rain runoff and rainfall amounts areamplitudes of the GM and GR heat 18 BCBEC ELEMENTSA BCBEC PUBLICATION'