File(s) under permanent embargo

Reason: Restricted by author. A copy can be supplied under Section 51(2) of the Australian Copyright Act 1968 by submitting a document delivery request through your library or by emailing

Stormwater biofiltration systems; development, optimization and application

posted on 09.02.2017, 02:36 authored by Zinger, Yaron
Stormwater runoff is a leading cause of water quality degradation in many urban waterways and receiving waters. In addition, rapid urbanisation and climate change effects are elevating the pressure on the use and resourcing of freshwater supplies. Stormwater harvesting has the potential to harness this conventional nuisance into a reliable potable resource if a suitable treatment can be achieved. Excess nutrients and nitrogen in particular are carried by stormwater, potentially leading to eutrophication. Biofilters, also known as bioretention systems, have shown the potential to remove nutrients from stormwater, thus protecting receiving waters as well as providing significant landscape amenity and urban microclimate benefits. In biofilters, nitrogen compounds can be transformed and ultimately converted into nitrogen gas by coupled nitrification and denitrification, providing the sustainable removal of nitrogen. Current biofilter designs, however, have not yet been optimised for efficient nitrogen removal. Additionally, current biofilter systems are considered a "black box" in terms of nitrogen species transformation, with little known about the variations in their performance, particularly in relation to the harsh wetting and drying environment to which they are subjected. The present thesis has examined the processes involved in nitrogen removal (and to a lesser degree phosphorus removal), focusing particularly on nitrate removal dynamics and its optimisation in biofilters. The first step was a large scale base-line study that was designed to quantify the removal performance of conventional biofilter designs. The findings targeted the need to enhance NOx removal, by optimising components of the design, leading to new configurations. The novel design was tested for typically harsh operational conditions, such as prolonged drying periods and system recovery. In order to meet water quality guidelines, laboratory results were validated in the field through a full-scale biofiltration system, which also tested the effectiveness of the optimised designs in removing a range of pollutants from urban runoff. In the first stage of the thesis, a large scale study of 140 columns tested eight different biofilter design and operational factors. Overall, this study revealed that whilst biofilters could readily remove high levels of sediment (averaging 98% removal), phosphorus (85%) and heavy metals (greater than 90% removal for most metals), nitrogen removal was often poor. NOx in particular, leaches from the biofilters after dry weather spells, In addition, NOx removal was strongly dependent on the type of vegetation. It was concluded, therefore, that systems should be carefully designed, paying particular attention to the specification of the soil media and selection of the plants to assure the required nutrient removal. For the conditions testing, a biofilter system of 2% of its impervious area with a minimum filter media depth of 5OO mm was found to be satisfactory. Finally, the biofilter columns demonstrated the facility to achieve and maintain removal capacity even under high concentration inflows. The next study investigated nitrogen transformations and improved removal of NOx through denitrification. In order to achieve this, 18 advanced biofilter columns were constructed and incorporated into different levels of a saturated zone (SAZ), supplemented with a carbon source. Sampling ports enabled measurement of nitrogen transformations throughout the filter depth profile. The SAZ design columns demonstrated removal of NO x, ammonia, organic nitrogen and mean TN removal of up to 74%. The columns, which included carbon substrate in their SAZ, demonstrated more than 99% success in removing NOx, statistically more than the control columns that did not use carbon which removed less than 50% NOx. Moreover, the depth concentration profile exhibited the highest NOx reduction along the SAZ biofilter section, suggesting that the addition of organic carbon as an electron donor in the saturated zone is beneficial to the rate of denitrification; a saturated zone depth of 450 mm was found to be effective. Moreover, a subsequent study investigated the efficiency of the SAZ design during prolonged drying and subsequent rewetting periods, and found that having a saturated zone (SAZ) is critical for efficient nitrogen removal in dry periods of more than two weeks. Without the SAZ, the biofilters behave as a source rather than a sink for nitrogen and NOx in particular. Furthermore, the SAZ design showed much faster recovery of N removal upon rewetting; the SAZ design biofilters were able to recover nitrate removal after only one rewetting event. Without the SAZ, the recovery time may be longer than the antecedent dry period itself, meaning that net leaching will occur during several storm events before net removal is re-established. Finally, the laboratory biofilter results were validated in the field by introducing a large scale biofilter pilot in Israel adopting a dual mode system; 1. A stormwater harvesting operational mode (during the rainy season) and 2. An aquifer recovery mode (during the dry season) for treatment of highly pollutant groundwater with nitrate. The removal performance for sediments and nutrients in the field was similar if not better than predicted in the laboratory; TSS concentrations were reduced by 99.4% (lab; 98.1%), TP by 94% (lab; 70%), and TN by 65% (lab; 64%, SAZ=6OO mm). The field study results confirmed a high removal performance not only for nutrients, but also for heavy metals, pathogenic indicators, and TOC. The biofilter was found to treat the stormwater and met water quality standards for irrigation and stream health, achieving even the most stringent local drinking water guidelines (not for pathogens contamination). For example, it demonstrated high reductions of E-coli and Faecal Coliforms in the range of 2- 3 log reductions, and below the maximal permitted values for the majority of metals and measured nutrients. This does not mean that the outflows are directly drinkable without additional filtration and disinfection, but that the data demonstrates the potential of stormwater to eventually become the first stage in a potable water source or alternatively this can be safely recharged into the aquifer. Aquifer recovery application results show potential for nitrate removal in the remediation of contaminated groundwater, albeit at low flow rates and under batch flow regimes. In these conditions, the biofilter managed to remove up to 73% of the nitrate concentration within the contaminated aquifer and met the drinking water guideline for nitrate. The present research contributed many recommendations for the design of biofilters and operational recommendations that are listed in FAWB adoption guidelines (2009)1. One of the key design recommendations arising from the present research, however, is that, where possible, biofilters should incorporate SAZ and a supplementary carbon source within the filter media, to enhance their robustness and nitrogen removal. The presence of the SAZ design can buffer some inefficiency ineffective traits of conventional biofilters, while at the same time sustaining vegetation growth during dry periods. In fact, retrofitting the SAZ into 'simple' biofilters is recommended if the existing biofilter has inadequate N removal and if N discharges poses a potential threat to the receiving environment. A number of knowledge gaps and research challenges were identified from the current research. For example, the need to enhance the removal of organic nitrogen from stormwater, since it was observed as the primary N form in the biofilter effluent (86% of total N). This study also suggests that biofilters when deployed in practice as a decentralised system may serve several purposes simultaneously. This would require further research and testing to allow the optimisation of stormwater harvesting and the aquifer recovery of nitrate through a constant flow regime. This research has provided comprehensive insights and practical design recommendations to improve biofilter performances, while allowing safer and more versatile use. The practical applications of this research are currently being adopted in Australia, Israel and in other countries.


Campus location


Principal supervisor

Ana Deletic

Additional supervisor 1

Tim Fletcher

Year of Award


Department, School or Centre

Civil Engineering


Doctor of Philosophy

Degree Type



Faculty of Engineering