Green Roof Opportunities Bruce Gregoire M.S. University of - PowerPoint PPT Presentation

Green Roof Opportunities Bruce Gregoire M.S. University of Connecticut Department of Natural Resources and the Environment

Outline  Background  Green roofs  Objectives  Methods  Results  Conclusions  Future research needs

Background  Urbanized areas - High percentage of impervious surfaces - increased stormwater runoff - nonpoint source pollution - 10% increase in effective impervious area decreases water quality (Booth and Jackson, 1997; Lombardo et al., 2000; Makepeace et al., 1995; Novotny and Olem, 1994; USEPA, 2009)

Background  Roof surfaces - 12% to 21% of imperviousness in urban areas - Accounts for ~ 50% of total stormwater runoff - Contributes to NPS pollution (Bannermann et al., 1993; Boulanger and Nikolaidis, 2003; Forster, 1996)

N Bishop Center

Eagleville Brook Watershed Area: ~80 ha ~ 47% impervious area ~18% roof tops Green roof Eagleville Brook Daylight point

Background  Green roofs offer potential for reducing stormwater runoff  Recommended by EPA as a BMP to control nonpoint source pollution in urban areas.

Green Roofs  Intensive green roofs  Extensive green roofs - Modular green roof systems

Chicago City Hall

Stuttgart-Weilimdorf, Germany ZinCo Int, 2007

Meta-analysis of extensive green roof precipitation retention Figure 1. Meta-analysis of green roof precipitation retention. The solid vertical line represents an average retention of 56%.

Green Roof Research  Nutrients - type of organic matter and fertilizer may act as source of nutrients in green roof runoff  Metals - growing media and vegetation may influence retention of pollutants

Green Roof Research  No modular green roof studies  No green roof water quantity/quality studies in northeastern U.S.

Objectives 1. Evaluate the effect of a modular green roof system in the northeastern United States on stormwater runoff. 2. Evaluate the effect of the green roof on runoff water quality for nutrients, and total and dissolved metals.

Methods  Paired watershed study design - Calibration period - January 25, 2009 to September 1, 2009 - Paired water quantity data collected - precipitation, treatment and control watershed runoff (Clausen and Spooner, 1993)

Treatment watershed (Green roof) Drain Control watershed

Methods  Green roof growing media - 75% lightweight expanded shale - 15% composted biosolids - 10% perlite  Modules planted on Earth Day 2009

Methods  Green roof vegetation - 10 Sedum species, 12 plugs in each module: S. album (Murale), S. foresterianum subsp. elegans (Silver Stone), S. kamtschaticum, S. kamtschaticum var. floriferum (Weihenstephaner Gold), S. reflexum, S. selskianum, S. sexangulare , S. spurium (Dragons Blood), S. spurium (Fuldaglut), and S. spurium (John Creech)

Methods  Treatment period - September 2, 2009 to September 14, 2010 - 248 m ² green roof installed - Paired water quantity monitoring continued - Evapotranspiration measured with weighing lysimeter

Gant Plaza Green Roof  334 modules  81% coverage  2.6% organic matter in growing media  Slow release fertilizer

Schematic cross-section of green roof in Storrs, CT (Modified from Weston Solutions, Inc.). Module Growth Roof media structure Weed block Concrete pavers Root barrier/ Drainage Filter fabric holes

Methods  Treatment period - Paired water quality samples collected - Precipitation, green roof and control runoff - Water quality analysis - TN, TKN, NO 3 +NO 2 -N, NH 3 -N, TP, and PO 4 -P - Total and dissolved Cu, Pb, Zn, Cd, Cr, and Hg

Statistics  Shapiro-Wilk Test for Normality  Water quantity -ANOVA - regression significance during calibration and treatment period -ANCOVA - used to determine changes in slopes and intercepts of treatment and control watershed regressions between the calibration and treatment periods

Statistics  Water quality - No paired water quality data during calibration period - Oneway ANOVA and Tukey means comparison for water quality data - Trimmed mean if 15% - 50% non- detects (USEPA, 2000)

Results  Water Quantity

Calibration Period Average weekly runoff coefficient Watershed Period Control Treatment Calibration 0.70 0.71 Ratio of runoff to precipitation

Treatment Period  Average weekly runoff coefficient Watershed Period Control Treatment Calibration 0.70 0.71 Treatment 0.70 0.56 Ratio of runoff to precipitation

Results Calibration Period Treatment Period (n=29) (n=18) ANCOVA Calibration Treatment Treatment % Characteristic Control Treatment Control Equation Predicted Observed Change F p Adjusted runoff (cm wk -1 ) 1.07 0.55 0.67 T=1.23C 0.78 1.44 1.63 -34 63.81 <0.001 C = Control T = Treatment % change = Observed – Predicted x 100 Predicted  34% reduction in runoff compared to that predicted by the calibration regression

Green Roof Water Balance Input 1 cm % Precipitation 104.9 100 Output 1 cm % Runoff 59.2 56.4 Evapotranspiration 45.4 43.3 Residual -0.3 -0.3 1 September 2009 to June 2010

Precipitation Retention Compared to Other Studies  81% coverage compared to 100% - 100% green roof coverage should increase precipitation retention from 44% to 54%  Plaza roof watershed displayed characteristics of a natural watershed

Meta-analysis of green roof precipitation retention Figure 1. Meta-analysis of green roof precipitation retention. The solid vertical line represents an average retention of 56%.

Results  Water quality

Total Nitrogen b TN (mg L -1 ) a a 1 n = 19 0.1 Control Green roof Precipitation

Ammonia nitrogen 1 b NH 3 -N (mg L -1 ) 0.1 a a 0.01 0.001 n = 19 0.0001 Control Green roof Precipitation

Total Phosphorus TP (mg L -1 ) 1 c b 0.1 a 0.01 0.001 n = 19 0.0001 Control Green roof Precipitation

Orthophosphate PO 4 -P (mg L -1 ) 1 c b 0.1 a 0.01 0.001 n = 19 0.0001 Control Green roof Precipitation

Dissolved Zinc Zn ( µ g L -1 ) Dissolved c 100 b a 10 n = 14 1 Control Green roof Precipitation

Mass Input/Export - Nitrogen NH 3 -N NO 3 +NO 2 -N

Mass Input/Export - Phosphorus PO 4 -P

Mass Import/Export – Total Metals

Conclusions  Water quantity - Green roof precipitation retention 44% - Average runoff coefficient decreased from 0.71 to 0.56

Conclusions  Water quality - Sink for NH 3 -N, Zn, and Pb - Source of TP, PO 4 -P, and total Cu - Reduced mass export of TN, TKN, NO 3 +NO 2 -N, Hg, and dissolved Cu - primarily through a reduction in stormwater runoff

Conclusions  Overall the green roof was effective in reducing stormwater runoff and overall pollutant loading for most water quality contaminants

Future Research Needs  Study effects of growing media using expanded shale and composted biosolids in pollutant retention.  Metal pollutant uptake and stabilization in the growing media by various Sedum species is unknown  No standard exists for the composition of the green roof media and vegetation to reduce NPS pollution

Acknowledgements This project was funded in part by the CT DEP through a US EPA nonpoint source grant § 319 Clean Water Act  Jack Clausen  John Alexopuolos  CT Dept of Environmental Protection  UConn Student Chapter of SWCS  Center for Environmental Sciences and Engineering

Questions