Year

2013

Degree Name

Master of Environmental Science - Research

Abstract

Urban development is known to significantly impact on catchment hydrology and stormwater runoff quality. Constructed wetlands have been in use since at least the 1960s to ameliorate these impacts and their use in New South Wales (NSW) Australia has been popular since the 1990s. Horsley Estate is a small satellite urban development located within the Illawarra region of NSW, between Dapto and the Illawarra Escarpment. This estate is currently predominately surrounded by rural landuse, however over the next 40 years, urban development is set to expand significantly in the form of the West Dapto Land Release. Four water quality control ponds (WQCP) are currently in service within Horsley Estate, constructed for the purpose of improving urban runoff quality associated with the Estate. In this study, one of these WQCP, “ROB1”, is investigated in terms of its design relative to current best practice and its ability to reduce the Priority Pollutants total nitrogen (TN), total phosphorus (TP) and total suspended solids (TSS) in stormwater runoff in order to inform the design of new urban runoff water quality infrastructure for the Land Release. A review of the literature identified two primary issues concerning determining the pollutant reduction capacity of constructed wetlands, i.e. a lack of technology able to measure TN, TP and TSS at sufficient temporal resolution to accurately estimate pollutant load reductions, and a lack of consistency in methodologies able to quantify the reliability of their results. To address these issues, five time synchronised water quality monitoring stations (WQMS) were set up at each of four inlets and the outlet of ROB1, each containing an auto-sampler, turbidity sensor, conductivity sensor and flow sensor integrated with a programmable data logger. Each WQMS was programmed to record turbidity, conductivity, temperature and flow every five minutes (“high temporal resolution monitoring data”) and to automatically take samples throughout each rain event according to a pre-programmed regime. Standard least squares procedures were then used to fit the data to two multivariate statistical models, termed the “Inlets Predictive Model” (IPM) and the “Outlet Predictive Model” (OPM). Model diagnostics (R2 correlations, summary of fit and analysis of variance indicators) were strong for both TN and TP, for both the IPM and OPM. However, while summary of fit and analysis of variance indicators were relatively strong for TSS for both models, the R2 correlations were considered too weak to be of use (a review of the diagnostic outputs for both models revealed the most like cause was the limits of reporting associated with the laboratory analysis for TSS, rather than the models themselves). Using these models and the high resolution monitoring data, mass balances were then determined for four of the six events for TN and TP (technical problems meant that two events could not be used for this purpose). Results revealed that ROB1 was achieving a level of pollutant reduction akin to current benchmarks. This study also compared the design of ROB1 to current best practice for constructed wetland design, taking into consideration such issues as impacts on lot yield (i.e. size required for effective treatment), ease of maintenance, susceptibility to invasion by weeds, and public safety. This comparison revealed that whilst ROB1 may be achieving a reasonable level of TN and TP reduction, its design is out-dated, and consideration should be given to alternative water quality improvement infrastructure, including the utilization of smaller integrated stormwater management measures to improve water quality outcomes, whilst reducing maintenance costs and public safety risks.

FoR codes (2008)

0502 ENVIRONMENTAL SCIENCE AND MANAGEMENT, 0104 STATISTICS, 0599 OTHER ENVIRONMENTAL SCIENCES

Download

Share

COinS
 

Unless otherwise indicated, the views expressed in this thesis are those of the author and do not necessarily represent the views of the University of Wollongong.