Water quality impacts of young green roofs in a tropical city: a case study from Singapore

This study examined the effects of two substrates (SOIL and COMMERCIAL) and grass on the green roof runoff quality in Singapore. Ten events were sampled over a 9-month period. Rainfall and green roof runoff from grass and bare experimental configurations were tested for total organic carbon (TOC), nitrogen and phosphorus nutrients (N, NH4+-N and PO43−-P), cations/anions and trace metals (Fe, Cu, Zn, Cd and Pb). All configuration units neutralised acid rainfall and removed metals except Fe despite their proximity to an industrial area. Concentrations decrease over the monitoring period for most water quality variables. The COMMERCIAL (COM) configurations elevated Cl− (3.8–10.8 ppm), SO42− (1.5–32.4 ppm), NO3−-N (7.8–75.6 ppm) and NH4+-N (22.0–53.1 ppm) concentrations in the runoff. Concentrations of NO3−-N (4.5–67.7 ppm) and NH4+-N (14.7–53.0 ppm) remained high at the end of the monitoring period for the COMgrass configuration, even with dilution from monsoon rainfall, making it suitable as an irrigation water source and a fertiliser substitute. The SOIL substrate retained N-nutrients, TOC and trace metals with concentrations comparable or below rainfall inputs. This substrate is suitable for widespread green roof applications in Singapore and other tropical cities. We recommend substrate testing before their approval for use on green roofs and encourage the long-term monitoring of these systems.

Copper, Zinc and Iron Contamination in Wellington Streams After Rainfall Events

<p>This work set out to determine the concentrations of dissolved copper (Cu), zinc (Zn) and iron (Fe) during base and wet weather flow at streams throughout the Wellington region. The secondary objective was to investigate possible sources of heavy metals during rainfall events. The concentrations of the three dissolved trace metals Cu, Zn and Fe were measured at 13 sites on five streams during base flow conditions and during runoff events (wet weather flow) in the Wellington region between January and July 2011. More than 240 base flow and 100 wet weather flow samples have been analysed for the three dissolved metals. Additionally, rainfall, roof runoff and paved surface runoff samples have been collected and analysed. The analysis was performed by Flame Atomic Absorption Spectroscopy (FAAS). A pre-concentration procedure using Chelex-100, a chelating polymeric resin bead, was developed and successfully used to enhance the concentrations of dissolved Cu and dissolved Zn. The recorded data were compared to the recommended long-term (chronic) toxicity triggers; the Australian and New Zealand Environment and Conservation Council freshwater toxicity trigger values (ANZECC (2000) TV) for dissolved Cu and Zn, and the Canadian trigger value (CTV) for dissolved Fe. Additionally, the concentrations of dissolved Cu and Zn in storm water samples were compared against the recommended short-term (acute) toxicity triggers, the United States Environmental Protection Agency (USEPA 2006) Criteria Maximum Concentrations (CMC). The medians of dissolved heavy metals concentrations; Cu, Zn, and Fe, all of which are potentially toxic to aquatic life, exceeded the long-term (chronic) toxicity guidelines at one of the studied sites for Fe, nine sites (69%) for Cu and 10 sites (77%) for Zn in base flow conditions. Comparison of base flow monitoring data to previous reports showed that the concentrations of the studied metals have increased over the last five years. Storm water (wet weather flow conditions) contained elevated levels of dissolved heavy metals in comparison to base flow concentrations which is consistent with what has been reported previously. Dissolved Cu and Zn exceeded the acute toxicity criteria at sites of suburban residential areas. The median of dissolved Fe concentration exceeded the sustained toxicity exposure trigger at eight of the studied sites (61%). Distinct catchment type contaminant concentrations (dissolved Cu and Zn) were observed during storm runoff events with a concentration pattern of suburban residential > commercial > light residential > rural catchment. Dissolved Fe exhibited a similar pattern but in this case the concentration in rural catchments was higher than in light residential catchments. These observations were attributed to the high traffic of vehicles passing nearby the area; accumulated particulate materials; and corrosion of materials containing heavy metals, such as galvanised and copper roofs, gutter systems and building construction materials. The strongest and most obvious first flush effect was observed with dissolved Cu followed by Zn where the phenomenon was observed in six storm runoff events for Cu and five events for Zn. The first flush effect of dissolved Fe was present in three out of eight storm runoff events. The concentrations of dissolved metals were of the same order of magnitude as those previously reported for the Wellington region, but this work recorded the highest expected concentrations, particularly, for dissolved Fe. The reported data are consistent with data sets from other New Zealand regions. The investigation of possible sources of dissolved heavy metals in storm runoff samples showed that rainfall water contained markedly elevated concentrations of dissolved Zn and smaller Cu concentrations, 0.04-0.075 and 0.0018-0.01 mg/L respectively, in comparison to the ANZECC (2000) TVs, 0.008 and 0.0014 mg/L respectively. The concentrations of dissolved Fe were below the CTV level of 0.3 mg/L. Most studies conducted in New Zealand did not take into account the atmospheric precipitation contribution to the elevated concentrations of dissolved metals during runoff events. Roof runoff samples had similar dissolved Cu and Fe concentration to those recorded in atmospheric rainfall water, but Zn was found to be higher in galvanised roof runoff. First flush samples from roof runoff had higher concentrations of all three metals than the delayed runoff samples, indicating the presence of accumulated particles containing metals. Paved surface runoff samples had concentrations of dissolved Cu and Zn higher than the chronic toxicity triggers, but the medians did not exceed the acute toxicity guidelines. The value of the median for the concentrations of dissolved Fe was below the CTV criteria. Similar results have been published for surface runoff in New Zealand and the international literature related to this field.</p>

Copper, Zinc and Iron Contamination in Wellington Streams After Rainfall Events

<p>This work set out to determine the concentrations of dissolved copper (Cu), zinc (Zn) and iron (Fe) during base and wet weather flow at streams throughout the Wellington region. The secondary objective was to investigate possible sources of heavy metals during rainfall events. The concentrations of the three dissolved trace metals Cu, Zn and Fe were measured at 13 sites on five streams during base flow conditions and during runoff events (wet weather flow) in the Wellington region between January and July 2011. More than 240 base flow and 100 wet weather flow samples have been analysed for the three dissolved metals. Additionally, rainfall, roof runoff and paved surface runoff samples have been collected and analysed. The analysis was performed by Flame Atomic Absorption Spectroscopy (FAAS). A pre-concentration procedure using Chelex-100, a chelating polymeric resin bead, was developed and successfully used to enhance the concentrations of dissolved Cu and dissolved Zn. The recorded data were compared to the recommended long-term (chronic) toxicity triggers; the Australian and New Zealand Environment and Conservation Council freshwater toxicity trigger values (ANZECC (2000) TV) for dissolved Cu and Zn, and the Canadian trigger value (CTV) for dissolved Fe. Additionally, the concentrations of dissolved Cu and Zn in storm water samples were compared against the recommended short-term (acute) toxicity triggers, the United States Environmental Protection Agency (USEPA 2006) Criteria Maximum Concentrations (CMC). The medians of dissolved heavy metals concentrations; Cu, Zn, and Fe, all of which are potentially toxic to aquatic life, exceeded the long-term (chronic) toxicity guidelines at one of the studied sites for Fe, nine sites (69%) for Cu and 10 sites (77%) for Zn in base flow conditions. Comparison of base flow monitoring data to previous reports showed that the concentrations of the studied metals have increased over the last five years. Storm water (wet weather flow conditions) contained elevated levels of dissolved heavy metals in comparison to base flow concentrations which is consistent with what has been reported previously. Dissolved Cu and Zn exceeded the acute toxicity criteria at sites of suburban residential areas. The median of dissolved Fe concentration exceeded the sustained toxicity exposure trigger at eight of the studied sites (61%). Distinct catchment type contaminant concentrations (dissolved Cu and Zn) were observed during storm runoff events with a concentration pattern of suburban residential > commercial > light residential > rural catchment. Dissolved Fe exhibited a similar pattern but in this case the concentration in rural catchments was higher than in light residential catchments. These observations were attributed to the high traffic of vehicles passing nearby the area; accumulated particulate materials; and corrosion of materials containing heavy metals, such as galvanised and copper roofs, gutter systems and building construction materials. The strongest and most obvious first flush effect was observed with dissolved Cu followed by Zn where the phenomenon was observed in six storm runoff events for Cu and five events for Zn. The first flush effect of dissolved Fe was present in three out of eight storm runoff events. The concentrations of dissolved metals were of the same order of magnitude as those previously reported for the Wellington region, but this work recorded the highest expected concentrations, particularly, for dissolved Fe. The reported data are consistent with data sets from other New Zealand regions. The investigation of possible sources of dissolved heavy metals in storm runoff samples showed that rainfall water contained markedly elevated concentrations of dissolved Zn and smaller Cu concentrations, 0.04-0.075 and 0.0018-0.01 mg/L respectively, in comparison to the ANZECC (2000) TVs, 0.008 and 0.0014 mg/L respectively. The concentrations of dissolved Fe were below the CTV level of 0.3 mg/L. Most studies conducted in New Zealand did not take into account the atmospheric precipitation contribution to the elevated concentrations of dissolved metals during runoff events. Roof runoff samples had similar dissolved Cu and Fe concentration to those recorded in atmospheric rainfall water, but Zn was found to be higher in galvanised roof runoff. First flush samples from roof runoff had higher concentrations of all three metals than the delayed runoff samples, indicating the presence of accumulated particles containing metals. Paved surface runoff samples had concentrations of dissolved Cu and Zn higher than the chronic toxicity triggers, but the medians did not exceed the acute toxicity guidelines. The value of the median for the concentrations of dissolved Fe was below the CTV criteria. Similar results have been published for surface runoff in New Zealand and the international literature related to this field.</p>

Evaluation of Harvested Rainwater Quality As a Drinking Water Source in Urban Land Uses in Western Province Sri Lanka

Increasing of population around the world has imposed considerable strains on the water resources. Hence, regulatory authorities have faced significant challenges to expand their water supply schemes not only due to financial constraints but also due to limited water resources. Consequently, there has been a growing interest, especially in developing countries, in harvested roof run off as an alternative source of drinking water. However, in determining the end use and the potential success of such an option, the possible problems associated with water quality need to be analyzed and the feasibility of using rainwater as a source of water for household use should be determined. Therefore, this research study was focus on evaluating roof runoff as a drinking water source in different urban land uses where different urban activities are present. For this purpose, a roof runoff samples were collected from three selected land uses namely industrial, commercial and residential. The samples have collected from each land use with three selected roofing materials which are common to the area. This is to evaluate the roof runoff quality based on the variability of land use pattern as well as the variability on roofing materials. All the collected samples have tested for a range of water quality parameters namely, pH, alkalinity, Hardness, Turbidity, TS, COD, Nitrogen (Ammonia), chloride and biological contaminations. Both uni-variate and stacked area analysis techniques were used in the analysis of test results. Based on the outcomes, recommendations are provided to use harvested roof runoff as a drinking water source in urban land uses in Sri Lanka. Keywords urban land uses; drinking water quality parameters; rainwater harvesting

Hydrological Modeling of Green Infrastructure to Quantify Its Effect on Flood Mitigation and Water Availability in the High School Watershed in Tucson, AZ

Green Infrastructure (GI) practices are being implemented in numerous cities to tackle stormwater management issues and achieve co-benefits such as mitigating heat island effects and air pollution, as well as water augmentation, health, and economic benefits. Tucson, Arizona is a fast-growing city in the semiarid region of the southwest United States and provides a unique landscape in terms of urban hydrology and stormwater management, where stormwater is routed along the streets to the nearest ephemeral washes. Local organizations have implemented various GI practices, such as curb cuts, traffic chicanes, roof runoff harvesting, and retention basins, to capture the excess runoff and utilize it on-site. This study models the 3.31 km2 High School watershed in central Tucson using the Automated Geospatial Watershed Assessment (AGWA) tool and the Kinematic Runoff and Erosion (KINEROS2) model. Each parcel in the watershed was individually represented using the KINEROS2 Urban element to simulate small-scale flow-on/flow-off processes. Seven different configurations of GI implementation were simulated using design storms, and we stochastically generated 20 years of precipitation data to understand the effects of GI implementation on flood mitigation and long-term water availability, respectively. The design storm analysis indicates that the configuration designed to mimic the current level of GI implementation, which includes 175 on-street basins and 37 roof runoff harvesting cisterns, has minimum (<2%) influence on runoff volume. Furthermore, the analysis showed that the current level of GI implementation caused an increase (<1%) in peak flows at the watershed outlet but predicted reduced on-street accumulated volumes (>25%) and increased water availability via GI capture and infiltration. When the GI implementation was increased by a factor of two and five, a larger reduction of peak flow (<8% and <22%, respectively) and volume (<3% and <8%, respectively) was simulated at the watershed outlet. The 20-year analysis showed that parcels with roof runoff harvesting cisterns were able to meet their landscape irrigation demands throughout the year, except for the dry months of May and June. Additionally, stormwater captured and infiltrated by the on-street basins could support xeric vegetation for most of the year, except June, where the water demand exceeded volume of water infiltrated in the basins. The current level of GI implementation in the High School watershed may not have significant large-scale impacts, but it provides numerous benefits at the parcel, street, and small neighborhood scales.

Using Multiple Isotopes to Identify Sources and Transports of Nitrate in Urban Residential Stormwater Runoff

Increased nitrogen (N) from urban stormwater runoff aggravates the deterioration of aquatic ecosystems as urbanisation develops. In this study, the sources and transport of nitrate (NO3−) in urban stormwater runoff were investigated by analysing different forms of N, water isotopes (δD-H2O and δ18O-H2O), and NO3− isotopes (δ15N-NO3− and δ18O-NO3−) in urban stormwater runoff in a residential area in Hangzhou, China. The results showed that the concentrations of total N and nitrate N in road runoff were higher than those in roof runoff. Moreover, high concentrations of dissolved organic N and particulate N in road runoff led to significantly different TN concentrations in road runoff (mean: 3.76 mg/L) and roof runoff (mean: 1.23 mg/L). The high δ18O-NO3− values (mean: 60 ± 13.1‰) indicated that atmospheric deposition was the predominant NO3− source in roof runoff, as confirmed by the Bayesian isotope mixing model (SIAR model), contributing 83.6–97.8% to NO3−. The SIAR model results demonstrated that atmospheric deposition (34.2–91.9%) and chemical fertilisers (6.27–54.3%) were the main NO3− sources for the road runoff. The proportional contributions from soil and organic N were smaller than other sources in both the road runoff and roof runoff. For the initial period, the NO3− contributions from atmospheric deposition and chemical fertilisers were higher and lower, respectively, than those in the middle and late periods in road runoff during storm events 3 and 4, while an opposite trend of road runoff in storm event 7 highlighted the influence of short antecedent dry weather period. It was suggested that reducing impervious areas and more effective management of fertiliser application in urban green land areas were essential to minimize the presence of N in urban aquatic ecosystems.

EVALUATING THE EFFICIENCY OF HOUSEHOLD STORMWATER DETENTION SYSTEM

This paper describes the evaluation of water storing capacity of a household stormwater detention system based on field data. Collection of field data is often sidelined due to the cost and human capital incurred. However, the true value of field data is demonstrated here by comparing the observed and design data. A field test is completed in a real-life terrace house, utilizing the house’s 95m2 side canopy as roof catchment and 4.40m x 4.70m car porch area to station a detention tank. Precast concrete modular units with 3.9m3 effective storage volume are assembled within the tank. Downpipe with 0.1m diameter is installed to connect the roof gutter to the detention tank; while pipeline with 0.05m diameter is installed as the outlet from tank to the house perimeter drain. The mentioned setup is subjected to actual rainfalls from December 2019 till February 2020 that corresponded with the peak of Northeast Monsoon season. Ten observed storm events with peak hourly total rainfall readings ranging from 22 to 48mm are selected for analysis. Rainfall and water level readings from the field test allow the derivation of roof runoff volume and detained water volume in the tank. It is found that the household stormwater detention system is able to capture about 50% of the roof runoff. However, the current setup is found to cause flooding for rainfall over 40mm. The flooding issue, however, is undetected by the design data that underestimated the water storing capacity. This is due to the use of uncommon precast concrete modular units that may not have its flow characteristics represented by existing formula and model. No matter how uncommon the modular units be, various types and forms of stormwater detention system are becoming the new normal in the industry and field test is the best tool to validate their performances.