Showing papers by "Tim D. Fletcher published in 2011"

Performance and sensitivity analysis of stormwater models using a Bayesian approach and long-term high resolution data

Abstract: Stormwater models are important tools in the design and management of urban drainage systems. Understanding the sources of uncertainty in these models and their consequences on the model outputs is essential so that subsequent decisions are based on reliable information. Model calibration and sensitivity analysis of such models are critical to evaluate model performance. The aim of this paper is to present the performance and parameter sensitivity of stormwater models with different levels of complexities, using the formal Bayesian approach. The rather complex MUSIC and simple KAREN models were compared in terms of predicting catchment runoff, while an empirical regression model was compared to a process-based build-up/wash-off model for stormwater pollutant prediction. A large dataset was collected at five catchments of different land-uses in Melbourne, Australia. In general, results suggested that, once calibrated, the rainfall/runoff models performed similarly and were both able to reproduce the measured data. It was found that the effective impervious fraction is the most important parameter in both models while both were insensitive to dry weather related parameters. The tested water quality models poorly represented the observed data, and both resulted in high levels of parameter uncertainty.

Calibration and sensitivity analysis of urban drainage models: MUSIC rainfall/runoff module and a simple stormwater quality model

Abstract: Model calibration and sensitivity analysis of stormwater models are required to assess model performance; it is very unlikely that non-calibrated models will lead to reasonable results. The aim of this paper is to present results of the calibration and sensitivity analysis of the key parameters used in flow modelling by MUSIC and parameters of a simple stormwater quality model. The assessment of the models is undertaken using a Monte Carlo Markov Chain approach. We describe the models' performance, provide information on their sensitivity to parameters and also discuss the correlation between these parameters. This work will help practitioners to understand importance of the MUSIC parameters that they usually use without calibration. The information reported in the results will also help to guide future development of stormwater quality models and the data needed to support it.

Quantifying the restoration of evapotranspiration and groundwater recharge by vegetated infiltration systems

Abstract: Infiltration-based source-control technologies are increasingly used in urban stormwater management in an attempt to restore pre-development catchment hydrology. To better assess the potential of infiltration systems to restore both the subsurface flow and the evapotranspiration in a catchment, information is needed about the fate of infiltrated water. To investigate this aspect, an experiment was set up to monitor the soil moisture surrounding a vegetated infiltration system (‘raingarden’). The proposed methodology allows partitioning the volume of exfiltrated water between the deep and shallow soil layers of surrounding soils, with the latter volume being hypothesised as available for evapotranspiration. Preliminary results show that: i) less than 3% of the influent water is evapotranspired from the raingarden surface area itself; ii) the heavy clay soil within the lower part of the soil profile significantly limited exfiltration, meaning that most of the water (at least 70%) is exfiltrated into shallow layers. This provides opportunity for water to be later evapotranspired by surrounding vegetation. However, due to the absence of water stress during the studied period, the evapotranspiration from the surrounding soils was not significantly altered by the raingarden, such that the infiltrated water effectively recharged groundwater. The study has implications for the design of stormwater management strategies which aim to reduce the stormwater runoff volume and restore baseflows lost as a result of urbanization.

A Dual-mode Biofilter System: Case study in Kfar Sava, Israel

Abstract: Despite experiencing recent drought conditions, Israel is expected to discharge more than 150 GL of urban runoff to sea due to extensive urbanization along the coastal plains. The underlying aquifer, a vital water resource, has become contaminated mainly by nitrate. Stormwater biofilters, have been demonstrated to be effective for stormwater treatment. A dual-mode biofiltration system has been constructed in Kfar-Sava to combine stormwater harvesting and treatment during the wet season, while being used to treat polluted aquifer groundwater (aquifer recovery) during the dry season. In addition to demonstrating treatment effectiveness, direct and infiltration recharge options of the treated water were tested to determine their relative efficiency. The preliminary results show that the system was able to effectively treat a range of pollutants in urban runoff (heavy metals, nutrients and pathogens) and meet Israeli and Australian guidelines for irrigation, aquifer recharge and streams health. Initial aquifer recovery tests show up to 73% nitrate removal of aquifer polluted water at low biofiltration rates. The Kfar-Sava biofilter marked an important milestone for implementing Water Sensitive Urban Design (WSUD) principles in the Israel and in the next two years Israel will gain at least two pilot systems across the country, with the aim being to establish policies and process to underpin widespread adoption.

Testing and Sensitivity of a Simple Method for Predicting Urban Pollutant Loads

Abstract: This paper tests a simple two-parameter regression model, based on rainfall intensity, for calculating event loads of total suspended solids, total phosphorus, and total nitrogen from urban catchments. It also examines the sensitivity of the model to its two parameters and to the rainfall time step. This was done by using large data sets collected at six urban catchments in temperate Melbourne. It was found that the two-parameter model typically explains approximately 90% of the variation in event loads at a site. The model also predicts the within-event behavior of pollutants when the flow lag time is taken into account, with R2 correlations greater than 0.6 in most cases for both loads and concentrations at a six minute time step. Despite its acknowledged correlation with flow, rainfall intensity over short time steps is shown to be the primary driver of pollutant mobilization, and provides a practical means of predicting pollutant loads by using readily available data.

Modeling the development and consequences of clogging for stormwater infiltration trenches

Abstract: Stormwater infiltration systems are a means of rest oring the natural hydrologic cycle where it has been disturbed by urbanisation. These systems h elp to increase soil infiltration, groundwater recharge and baseflows and reduce direct surface runoff. There is a strong correlation between the directly connected impervio us area of a catchment and corresponding declining health of urban waterways due to changes in the hydrology and water quality. Stormwater infiltration systems help to disconnect impervious areas from waterways and protect the waterway ecosystem. Stormwater infiltration basins and trenches are use d around the world in the UK, Europe, Japan, Australia and the USA. Despite the widesprea d usage, most existing design methods make a number of simplifying assumptions, such as assuming one dimensional (1D) saturated flow. Many existing models have limited capacity to represent antecedent conditions and the effects of lateral flows, particularly from the und erlying soils which may be significant for infiltration trench systems. To assess the benefits of stormwater infiltration systems in terms of modifying the hydrology of a catchment, it is pr eferable to simulate the system under a range of different storm events and intervening dry periods to evaluate measures such as the hydrologic effectiveness of the system (the proport ion of the runoff volume captured) and the frequency of downstream overflows. While stormwater infiltration systems can be very e ffective in reducing the volume and frequency of runoff into receiving waters, they are also potentially subject to sediment clogging, which can lead to reduced performance and eventually system failure. Clogging risks are typically managed with requirements for p re-treatment (a wise precaution) and the use of a ‘clogging factor’ or ‘safety factor. These factors are usually arbitrary and may not have a sound basis. They also do not allow for the progressive development of clogging with time or consider the expected lifespan of a system. In this paper, an overview of a recently developed dynamic two dimensional (2D) variably saturated flow model with representation of the sto rage and development of clogging is provided along with results from application of the model. The model allows a stormwater infiltration trench with a porous media such as gra vel to be simulated. It models soil moistures and flows through the surrounding soils to better r epresent antecedent conditions and their influence on system behavior. The model also simula tes filtration of sediment particles by the storage media and the development of a clogging layer as sediment accumulates at the interface with the soil. The clogging module provid es feedback to the flow modules to progressively influence the flow behavior over time . This allows the long term hydrologic effectiveness of an infiltration system throughout its lifecycle to be evaluated.

Hydraulic modelling of constructed reed-bed wetlands for stormwater treatment

Abstract: The reed-bed filter, widely used to treat wastewater, is adapted here to treat stormwater. Modelling such systems poses problems linked tothe unsteady-state behaviour, itself a consequence of the flow variability during storm events. Under those conditions, hydraulic head variations play an important role in the infiltration processes. The goal of this study is thus to develop simplified models which can be used for the sizing and design configuration of constructed reed-bed wetlands for CSO treatment. The paper proposes five models based on Darcy and Richards approaches, but also accounting for evapotranspiration and capillarity. Those models have been validated against data for a field biofiltration systemslocated at Monash University Car park. Two models based on the concept of barrier filters and simplified Richards equations seem to perform well with good experimental agreement obtained.These conceptual models were then applied to catchments in Lyon and Nantes (France), in order to test whether such simplified models can be feasibly applied to optimise the key design and sizing parameters for constructed reed-bed wetlands. Real inflow time series data collected on these sites were used to develop and test the models. Models have shown to be sensitive to drainage layer and diverted weir level.