PERFORMANCE EVALUATION OF DIFFERENT GREEN ROOF DESIGNS CHENGDU, CHINA Richard L. Stanford, P.E., Dr. Shaw L. Yu, Liu Ruifen, Richard Field, PE, D. WRE, DEE, Anthony N. Tafuri, Lin Luo and Yun Deng
Layout of Green Roofs Two green roof experimental segments were constructed on a single rooftop at Sichuan University Each green roof segment was 20 m X 4.5 m (90 m²)
Green Roofs Cross Section Garden soil Green Roof A Green Roof B Source: Green roof and vertical greening, Chengdu Technical Guidelines (on trial, issued in 2005 by local government) Source: Greenlink Kusters Company (one of the Germany s biggest landscaping companies started to design Green Roofs)
Construction
Carbon Residue
Monitoring Water Quantity Water Quality Temperature
Water Quantity Monitoring Rainfall Gauge Water Level logger Runoff Barrels for Green Roofs
V-Notched Weir 8
Water Quality Monitoring Roof A Roof C Roof B Sampling Spots of Green Roofs Sampling Spot of C Roof
Minimum Water Quality Storm Sampling Criteria An effective rainfall event: duration for at least one hour; precipitation at least 2 mm; interval between two events at least six hours. Sample type: composite Sample technique: manual Flow measurement method: volume method 10
Water Quality Analyses Parameter Method Implementation Reporting Testing of Standards Limit Place TSS Gravimetric method SEPA method 3.1.7-SS (A) 5mg/L Our Lab PH Method with Portable PH Meter SEPA method 3.1.6-PH (B) Our Lab SM 5220B COD Open Reflux Method (Similar as GB11914-1989) 5mg/L Our Lab SM 5210B BOD 5 Dilution and seeding method (Similar as GB7488-1987) 2.0 mg/l Our Lab TOC NO 2 -N Non-dispersive infrared absorption method N- (1-naphthyl)-ethylenediamine spectrophotometric method SEPA method 3.3.5-TOC(A) SEPA method 3.3.11-NO 2 (A) 0.5 mg/l 3µg/L Testing Center out of University Testing Center in University NO 3 -N UV spectrophotometric method SEPA method 3.3.10-NO 3- (B) 0.08mg/L NH 3 -N Titrimetric method SEPA method 3.3.12- NH 3 (A) 0.2mg/L Our Lab TN TP Cu Pb Zn Cd Alkaline potassium persulfate digestion- UV spectrophotometric method Ascorbic acid method Inductively coupled plasma-atomic emission spectrometry SEPA method 3.3.9-N(A) 0.05mg/L Our Lab SM 4500-P.E (Similar as GB11893-1989) EPA method 200.7 0.05mg/L 10µg/L 10µg/L 10µg/L 3µg/L Our Lab Testing Center in University 11
Temperature Monitoring Thermometer Test Point Sensor Location Temperature Collection from Different Layers of B Green Roof
Thermistor Locations Room under Green Roof B Room under Green Roof A Room under Non-green Area Top Floor Position of Thermocouple 13
RESULTS
First, the storms Table 3.14-5.30 rainfall events monitoring Hydrologic characteristics of the thirteen monitored rainfall events Event Date 1 Precipitation depth (mm) Rainfall duration (hours) Antecedent dry days 1 4/06 9.2 17 0 2 5/08 22.8 14.6 1 3 5/10 1.0 5.2 0 4 5/26 7.2 9.2 0 5 5/30 >11.4 7.4 0 6 6/28 2.4 5.5 0 7 7/07 61.2 3.2 8 8 7/13 16.6 5.1 2 9 8/11 >16.4 <2.5 5 10 8/13 >27 7 1 11 8/18 24.0 7.6 4 12 8/19 51.8 20 0 13 9/05 11.2 8.4 0 1 All dates are in 2010. Note: Red color means runoff occurance from green roofs in the rainfall event 15
Peak Shaving and Delay Event Date Start Antecedent Dry Period (day) Rain Duration (hour) Rainfall Total (mm) Runoff Total (mm) Peak Rainfall Intensity (mm/5mins) Peak Runoff Intensity (mm/5mins) Delay to Start of Runoff (min) A B C A B C A B C 7/7 8 3.1 61.2 33.8 36.8 52.6 7 3.14 2.62 4.59 87 107 7 7/13 2 4.1 16.6 6.2 5.7 16.1 0.8 0.34 0.22 0.7 225 265 185 7/24 1 32.5 71.2 55.1 55.2 65.8 4 0.91 0.73 2.42 165 175 90 8/11 5 <2.5 >16.4 4 10 16.4 unknown 0.19 0.39 2.76 8/13 1 <7 >27 12.7 9.3 27 unknown 0.75 0.58 2.59 8/18 4 7.6 24 13.3 8.8 18.4 4.6 0.79 1.32 3.79 306 286 146 8/19 0 20 51.8 36.2 37.7 48.8 2.2 1.17 1.24 1.88 475 480 246 9/4 2 16.4 18.6 10.3 7.9 18.5 0.6 0.19 0.18 0.43 575 670 135 9/5 0 8.4 11.2 10.2 7.3 12.9 0.6 0.33 0.2 0.65 90 90 55 9/8 2 10.6 16.4 5.3 3.8 15.2 1 0.39 0.15 1 470 500 375
Retention and Peak Intensity Event Date Retention (mm) Retention (%) Peak Runoff Intensity (mm/5mins) Peak Runoff Intensity (%) Delay to Start of Runoff (hr) Period Between Peak and Peak Runoff (hr) A B C A B C A B C A B C A B C A B C 7/7 27.4 24.4 8.6 44.8 39.9 14.1 3.14 2.62 3.14 44.9 37.4 65.6 1.5 1.8 0.1 0.3 0.5 0.0 7/13 10.4 10.9 0.5 62.7 65.7 3.0 0.34 0.22 0.34 42.5 27.5 87.5 3.8 4.4 3.1 3.2 3.3 2.8 7/24 16.1 16 5.4 22.6 22.5 7.6 0.91 0.73 0.91 22.8 18.3 60.5 2.8 2.9 1.5 11.3 11.3 0.0 8/18 10.7 15.2 5.6 44.6 63.3 23.3 0.79 1.32 3.79 17.2 28.7 82.4 5.1 4.8 2.4 0.3 0.2 0.1 8/19 15.6 14.1 3 30.1 27.2 5.8 1.17 1.24 1.88 53.2 56.4 85.5 7.9 8.0 4.1 5.1 5.2 3.6 9/4 8.3 10.7 0.1 44.6 57.5 0.5 0.19 0.18 0.43 31.7 30.0 71.7 9.6 11.2 2.3 12.6 13.2 0.1 9/5 1 3.9-1.7 8.9 34.8-15.2 0.33 0.2 0.65 55.0 33.3 108.3 1.5 1.5 0.9 1.1 0.8 0.4 9/8 11.1 12.6 1.2 0.7 0.8 0.1 0.39 0.15 1 39.0 15.0 100.0 7.8 8.3 6.3 0.4 0.7 0.2 Average= 12.6 13.5 2.8 32.4 39.0 4.9 0.91 0.83 1.52 38.3 30.8 82.7 5.0 5.4 2.6 4.3 4.4 0.9 17
Typical Runoff Hydrograph For low intensity rainfall, two green roofs runoff generated under saturated condition. Since A green roof had a higher field capacity, it can effectively delay runoff occurance. 18
Conclusions Water Quantity An average of 12.6 mm or 32.4% of the total precipitation of the sampled 11 events was retained on Green Roof A and 13.5 mm or 39% on Green Roof B. The maximum runoff rate from Green Roof A averaged 0.91 mm/5mins or 38.3% of the peak rainfall intensity and from Green Roof B averaged 0.83 mm/5mins or 30.8 of the peak rainfall intensity. An average of 5.0 hours was delayed from the start of the rainfall event by Green Roof A and 5.4 hours by Green Roof B. Both green roofs can change stormwater runoff dynamics through lowering (attenuation) An average and delaying of 4.3 hours the peak between runoff. the Compared peak intensity to A green of the roof, rainfall B green to roof advanced the in peak reducing runoff runoff rate was peak observed flow by extending from Green the Roof whole A and runoff 4.4 hydrography. hours Inflection from point Green was shown Roof B. on each curve at 21:10. 19
Results of Water Quality Monitoring Parameter Table Statistical Mean summary of concentrations of of the major common pollutants pollutants in rainfall, green from roof control and control roof roof in runoff each event No. of Observations Raw Data Mean (EMC) Standard Deviation Normallydistributed? Normallydistributed? Lognormally Transformed Data Mean* (EMC) Standard Deviation* Data set used for comparisons Unit:mg/L Green Roof A TSS 12 Y** 8.4 5.9 NC NC NC Raw ph 11 Y 7.8 0.4 NC NC NC Raw COD 12 Y 17.0 11.9 Y 18.5 19.5 Transformed TP 12 N 0.09 0.06 Y 0.087 0.005 Transformed TN 12 N 2.55 2.3 Y 2.60 1.80 Transformed Green Roof B TSS 13 Y** 9.7 10.4 NC NC NC Raw ph 11 Y 7.6 0.5 NC NC NC Raw COD 13 N 27.5 19.9 Y 28.8 25.1 Transformed TP 13 Y 0.16 0.06 Y 0.159 0.009 Transformed TN 13 N 15.69 17.5 Y 18.87 45.93 Transformed Control Roof ( C ) TSS 13 Y 13.1 9.7 NC NC NC Raw ph 11 Y 7.6 0.3 NC NC NC Raw COD 13 N 22.6 24.4 Y 26.8 53.4 Transformed TP 13 Y 0.04 0.03 Y 0.046 0.001 Transformed TN 13 N 4.10 2.9 Y 4.12 2.30 Transformed Rainfall TSS 6 Y 0.0 0.0 NC NC NC Raw ph 10 Y 6.3 0.9 NC NC NC Raw COD 10 N 2.2 1.9 Y 2.2 1.3 Transformed TP 12 Y 0.04 0.02 Y 0.037 0.001 Transformed TN 11 Y 2.36 1.65 Y 2.39 1.19 Transformed NC Not calculated. * Mean and standard deviation back-transformed to arithmetic values ** One-half quantitation level substituted for data reported below quantitation level. Kaplan-Meyer (1958) procedure used to calculate mean and standard deviation for censored data set. 20
Water Quality Summary Statistics Summary Comparison Statistics for Water Quality Analyses Rainfall Control Roof Green Roof A Green Roof B TSS EMC 0.0 13.1 8.4 10.4 Comparison Rainfall Control Roof Green Roof A Green Roof B ph EMC 6.3 7.6 7.8 7.6 Comparison Rainfall Control Roof Green Roof A Green Roof B COD EMC 2.2 22.6 27.5 17.0 Comparison Rainfall Control Roof Green Roof A Green Roof B TP EMC 0.04 0.04 0.09 0.16 Comparison Rainfall Control Roof Green Roof A Green Roof B TN EMC 2.36 4.10 2.55 15.69 Comparison Rainfall Green Roof A Control Roof Green Roof B
Where do the Nutrients come from? Part 1 Changes in Nitrogen and Phosphorus Concentrations in Sedum mat soil
Where do the Nutrients come from? Part 2 Changes in Nitrogen and Phosphorus Concentrations in Growth Media
Where do the Nutrients go? Material Analyses of total nitrogen and total phosphorus in green roof construction materials Sedum mat soil Total Nitrogen Total Phosphorus Growth media Total Nitrogen Total Phosphorus Carbon residue Total Nitrogen Total Phosphorus Sedum mat soil Total Nitrogen Total Phosphorus Growth media Total Nitrogen Total Phosphorus Initial concentrations Green Roof A 0.156 0.204 0.063 0.088 ND ND Green Roof B 0.156 0.204 0.119 0.082 Concentrations 7 months after construction 0.107 0.112 0.089 0.086 0.082 0.056 0.130 0.104 0.051 0.083 Values are in percent. ND Not determined, assumed 0.0% because of the thermal method of producing carbon residue
Conclusions Water Quality Initial concentrations of nutrients in runoff were HIGHER than the concentrations of nutrients in the rainfall. Green Roof A showed an increase in EMC of 225% TP and 174% TN. Green Roof B showed an increase in EMC of 400% TP and 665% TN. The increase in nutrients appeared to come from leaching of the sedum mat and from leaching of the manufactured growth media (soil). Leaching was less from the local garden soils. The carbon residual layer of Green Roof A seems to absorb some of the nutrients, which resulted in a lower increase in nutrient concentrations in the effluent. Both green roofs appeared to reduce TSS, but neither had a TSS effluent concentration that differed statistically from the control roof.
Temperature Monitoring Results 26
Temperature Data from Different Layers The temperature fluctuations in the structural layers of Green Roof A were similar to the air temperature fluctuation. The plant mat had the largest temperature fluctuation and the inner ceiling surface had the smallest.
Temperature Monitoring Results
Temperature Data from Different Rooms While the interior ceiling surface temperature below Green Roof B changed with the air temperature, the interior ceiling surface temperature fluctuation below Green Roof A was similar to the interior surface temperature fluctuation of the room without an overlying green roof. 29
Temperature Monitoring Results 30
Temperature Monitoring Results Statistical index Average diurnal temperature Average diurnal maximum temperature Average diurnal minimum temperature Average diurnal temperature fluctuation temperatures on different interior ceilings surfaces B green roof VS Non-green A green roof VS B green roof A green roof VS Non-green no-significant difference no-significant difference no-significant difference no-significant difference no-significant difference significant difference no-significant difference no-significant difference no-significant difference no-significant difference no-significant difference significant difference The green roofs reduced the average maximum diurnal temperature by 1.5 C, and the average diurnal temperature fluctuation of the room beneath the green roof by 0.8 C. 31
Next steps Continue monitoring green roofs for water quantity and quality to see if the mat and media stop leaching nutrients. Evaluate mass balances of nutrients to see if the total amount of nutrients released is lower or higher than the control roof after considering the reduction in runoff quantity from the green roofs. Perform an energy analysis of the temperature differences to see if there is a significant energy savings resulting from the green roofs. Combine the two designs (i.e., add carbon residue layer to Green Roof B) to see if the effects of the layer are additive.