Download as pdf or txt
Download as pdf or txt
You are on page 1of 19

Advancing the Design of Stormwater Biofiltration

September 2007
Contributors

FAWB team

Lucie Alcazar Honours student


Katia Bratières Honours student
Peter Breen Researcher and FAWB Board Member
Sébastien Le Coustumer PhD Student
Ana Deletic Research Manager
Tim Fletcher Project 1 Leader and FAWB Board Member
Belinda Hatt Project 3 Leader (and PhD student)
Justin Lewis Project 4 Leader
Peter Poelsma Researcher
Jennifer Read Researcher
Tricia Wevill Researcher
Tony Wong CEO and Researcher
Yaron Zinger PhD Student

International Collaborators

Sylvie Barraud Research Collaborator from INSA Lyon, France


Godecke Blecken PhD Student from Lulea University, Sweden
Maria Viklander Research Collaborator from Lulea University, Sweden

2
Table of Contents
1. Rationale ....................................................................................................................................4
2. Concept of Biofilter Design .......................................................................................................4
3. Research Methodology ..............................................................................................................6
3.1 Structure of Research Program ............................................................................................6
3.2 Project 1: Technology ..........................................................................................................7
3.3 Project 4: Demonstration and Testing..................................................................................9
4. Key Findings ............................................................................................................................12
4.1 Outline of the Design .........................................................................................................12
4.2 Filter Soil Media ................................................................................................................14
4.3 Vegetation ..........................................................................................................................15
4.4 Anoxic Zone and Carbon Source .......................................................................................16
4.5 Hydraulic Conductivity......................................................................................................16
4.6 Treatment Performance......................................................................................................17
4.7 Construction and Maintenance...........................................................................................17
4.8 Lessons learned from practice so far..................................................................................18
4.9 Other Issues........................................................................................................................18
5. Future research.........................................................................................................................18
5.1 Vegetation ..........................................................................................................................18
5.2 Soil Media ..........................................................................................................................19
5.3 Hydraulic Conductivity......................................................................................................19
5.4 Anoxic Zone and Carbon Source .......................................................................................19
5.5 Treatment Performance......................................................................................................19
Appendix A: Filter Media ..................................................................................................................20
Appendix B: Vegetation.....................................................................................................................39
Appendix C: Anoxic Zone and Carbon Source .................................................................................62
Appendix D: Field Studies.................................................................................................................79
Appendix E: Other Issues ................................................................................................................104

3
1. Rationale
The Facility for Advancing Water Biofiltration (FAWB) was formed in mid-2005 as an
unincorporated joint venture between Ecological Engineering and the Institute for Sustainable
Water Resources (ISWR), Monash University. The following industry collaborators are also
involved:
• Manningham City Council (Vic)
• Melbourne Water (Vic)
• Vic Roads (Vic)
• Landcom (NSW)
• Brisbane City Council (Qld)
• Adelaide and Mount Lofty Ranges Natural Resources Management Board (succeeding The
Torrens and Patawalonga Catchment Water Management Boards) (SA)
• Auckland Regional Council (NZ) (to 30 June 2006)

FAWB is primarily funded through the Victorian State Government’s Science, Technology and
Innovation (STI) grant, industry cash contributions and a direct cash contribution from Monash
University.

The mission of the Facility for Advancing Water Biofiltration (FAWB) is to provide proof of
concept by developing and field-testing a range of biofilter systems that can be applied to specific
market-based needs. This includes the needs of catchment managers, environmental regulators,
public utilities, local governments, land developers, and design engineers.

Water biofiltration is the process of improving water (stormwater and wastewater) quality through
the processes of filtration through biologically influenced media. Stormwater biofiltration systems
include bioretention systems, constructed surface flow wetlands and constructed sub-surface flow
wetlands, how the focus of this document is on bioretention systems.

The purpose of this document is to provide a summary of FAWB’s findings to date on biofiltration
technology 1 . The document starts with a brief outline of the typical design of biofilters, then
explains very briefly the program research methodology, and gives a summary of key findings.
Also included are published research papers from the work so far (12 in total) grouped into five
appendices.

2. Concept of Biofilter Design


A typical biofiltration system consists of a vegetated swale or basin, overlaying a filter medium
(usually soil-based) with a drainage pipe at the bottom (Figure 1). Small, bioretention pods are
often referred to as rain gardens, while linear systems are commonly referred to as bioretention
swales. The design configuration of biofilters is flexible, and possible variations include removal of

1
The research outputs on institutional issues are published in:
(1) Brown & Clark (2007) The Transition to Water Sensitive Urban Design, The Story of Melbourne, Joint Report
of FAWB and National Urban Water Governance Program
(2) Brown & Clark (2007) The transition towards Water Sensitive Urban Design: a socio-technical
analysis of Melbourne, Australia, In the Proc. of NOVATECH 2007, Lyon, France, June 2007.

4
the underdrain (to promote exfiltration into the surrounding soil, and the inclusion of a permanently
wet, anoxic zone at the bottom (to further enhance nitrogen removal).

To refine the design of biofilters and facilitate widespread adoption of these systems, the following
research questions should be answered:
1. Technology questions:
• How do biofilters work?
• How should we design biofilters to work efficiently in a wide range of applications (e.g.
pollution control, stormwater harvesting) and a range of site characteristics (e.g. different
climate, pollutant loads)?
2. Adoption questions:
• What are the factors (policy, regulation, risk, etc.) that advance their widespread
implementation?
• How do we quantify these factors and their relative significance?

Figure 1. Schematics of a typical biofilter (bioretention system)

To test the technology and enable its uptake, FAWB is also committed to:
• Develop adoption tools, such as design methods and adoption guidelines; and
• Demonstrate and test the technology, by supporting construction of a number of full scale
systems.

5
3. Research Methodology
3.1 Structure of Research Program
The entire Research Program is divided into four highly interlinked Projects, as presented in Figure
2.

Project 1: Project 2: Policy


Technology and Risk

Project 3: Adoption Tools

Project 4: Demonstration and Testing

Figure 2. FAWB Projects

The broad aims of each Project are listed below, while a more detailed explanation is given in the
Project reports.

Project 1: Technology aims to overcome technical barriers to wide adoption of the technologies, in
particular to:
• Develop new biofilter designs to optimise performance and ensure long-term sustainability;
• Determine design configurations that optimise treatment performance, and reduce the risk of
soil media clogging;
• Develop new filter media types for targeted pollutants (such as heavy metals, nutrients and
pathogens);
• Determine sustainable pollution loadings, in order to make predictions about effective
lifespan; and
• Determine the performance and risk of using stormwater biofilters as a treatment device for
stormwater harvesting.

Project 2: Policy and Risk aims to develop methodologies/strategies to overcome institutional and
social barriers to widespread adoption of the technologies, in particular to:
• Develop regulatory, policy and strategic guidance; and
• Define risk perception, liability and opportunities.

Project 3: Adoption Tools aims to develop design tools for practitioners, in particular to:
• Develop design methods and algorithms; and
• Develop adoption guidelines.

Project 4: Demonstration and Testing aims to demonstrate the wide capability of novel, multi-
functional designs, by:

6
• Validate laboratory studies and address site specific issues;
• Provide the basis for monitoring of long term robustness under real operating conditions;
• Provide demonstrations of biofiltration systems in a range of urban environments
(streetscapes, greenfield, inner-city retrofits, etc.); and
• Document construction procedures, for use in guidelines and standard drawings.

This document focuses on technology and reports on findings from Project 1 and Project 4 only!

3.2 Project 1: Technology


Based on the aims listed above, three Project Activities have been developed, within Project 1:
• Project Activity 1.01: Vegetation trials;
• Project Activity 1.02: Laboratory biofilter column experiments; and
• Project Activity 1.03: Biofilter optimisation for stormwater reuse.

Activity 1.01: Vegetation trials


20 species commonly used in rain garden design have been tested for removal of the key
stormwater pollutants (Figure 3), including total suspended solids (TSS), key heavy metals, and
total phosphorus (TP) and total nitrogen (TN) and their species (see Read et. al. (in press) in
Appendix B for full details. The list of plants is in Table 1 of this paper). The plants were dosed
with semi-synthetic stormwater for three months and their treatment performance assessed by
analysing the treated stormwater. At the end of the dosing period, plant biomass was measured and
relate to plant performance. Plant stress was also monitored (including the impact of drought and
shade) in a separate trial.

Figure 3. Plant species trials

7
Activity 1.02: Laboratory biofilter column experiments
This is a large activity that involved several independent studies:
(a) Laboratory study of non-vegetated filters. For 42 weeks, the soil-based filters were dosed
with semi-synthetic stormwater using different drying and wetting regimes, and their
treatment (removal of TSS, TP, TN, nutrient species, and heavy metals) and hydraulic
performance (clogging rate) were monitored (Figure 4(a), see Hatt et. al. (2007a) and Hatt
et. al. (2007b) in Appendix A).

(a) (b)
Figure 4: Laboratory biofilter column experiments (a) non-vegetated soil filter media columns (b) standard
columns

(b) Optimisation of standard biofilter design (Figure 4(b)) was carried out using columns filled
with four filter media types, non-vegetated and vegetated with five different plant spices,
and three different filter depths (300, 500 and 700 mm). The impact of different inflow
concentrations and climate (Brisbane and Melbourne) was also studied. In total, 140
columns have been watered over ten months with semi-synthetic stormwater, and their
treatment (removal efficiency of key pollutants) and hydraulic performance (change in
hydraulic conductivity) monitored. Details of the experimental methods are reported in
Fletcher et. al. (2007); see Appendix B. Details of the monitoring of hydraulic performance
were reported in a joint lab/field study by Le Coustumer et. al. (2007); see Appendix D.

(c) The impact of a permanently wet anoxic zone (containing a carbon source) on biofilter
performance was assessed using 18 advanced columns (Figure 5).

8
Figure 5: Testing impact of anoxic zone and carbon source in advanced columns

(d) The long term sustainability of soil media has been investigated in a separate laboratory
study, where three selected soils have been exposed to 15-20 years of continuos loading by
stormwater, the aim being to assess break-through of pollutants. This study has recently
begun, and so will not be reported on here.
(e) The impact of temperature on the performance of biofiltration systems has been studied in
conjunction with Luleå University, Sweden. This work was performed in constant
temperature rooms at Luleå University, using the FAWB standard column design (Fig 4(b))
and experimental procedures. The results of this study, which focuses primarily on cold
climate issues and uses non-Australian plants, are reported in Blecken et. al. (2007) in
Appendix E.

Activity 1.03: Biofilter optimisation for stormwater reuse


Pathogen removal by biofilters was tested using 30 standard columns (Figure 4(b)). Over three
months, the columns were watered with real stormwater spiked by pathogens, and the removal of
three common pathogen indicators (indicators of viruses, protozoe and bacteria) was monitored.
The influence of soil type, plant species, anoxic zone, carbon source, and variable wetting and dry
on pathogen removal was observed. The results for this study are currently being analysed.

3.3 Project 4: Demonstration and Testing

The focus of Project 4 is on testing the novel biofiltration systems constructed in consultation with
FAWB, testing a number of existing systems, and engaging industry on the design, construction,
operation and maintenance of biofilters. The current activities include:
• Activity 4.01 Biofiltration System in Western Sydney
• Activity 4.02 Monash University Carpark Biofiltration System, Melbourne
• Activity 4.03 Wakerley Biofiltration System, Brisbane
• Activity 4.04 Testing existing biofiltration systems
• Activity 4.05 Saturn Crescent Bio Pods, Brisbane
This work was carried out in conjunction with our key industry partners, as outlined below.

9
Activity 4.01: Biofiltration System in Western Sydney
Two trial systems have been built in the saline soils of Western Sydney (Figure 6(a)). The filter
media in the systems has had to be replaced twice due to rapid failure of the hydraulic performance.
Valuable lessons have been learned on how to avoid structurally unsound soil media. A detailed
laboratory study on the performance of unlined systems built in highly saline soils (Fig 6(b)) has
been conducted to study whether these will be impacted upon by saline incursion from the
surrounding soils

(a) (b)
Figure 6: Second Ponds Creek systems: (a) construction of two bioretention systems, (b) Lab study of
performance of bioretention systems built in sodic soils.

Activity 4.02: Monash University Carpark Biofiltration System


This system was designed by FAWB researchers as an experimental facility (Figure 7). Its purpose
is to treat carpark runoff for water use (irrigation of a sports oval). The biofilter contains three
parallel cells, each containing a different soil-based filter media. The system is fully equipped for
accurate field monitoring of hydraulic and treatment performance. The description of the system
and the monitoring program and preliminary results are presented in Hatt et al (2007), included in
Appendix D.

Figure 7: Monash carpark stormwater harvesting system

10
Activity 4.03: Wakerley Bioretention System, Brisbane
The Wakerley biofilter is a large system (servicing an 87 ha catchment) that has been recently built
in Brisbane (Figure 8). To date, FAWB has been involved in helping Brisbane City Council (BCC)
to establish a comprehensive monitoring program. The system is very complex and thus a
challenge to monitor.

Figure 8: Wakerley Bioretention System, Brisbane

Activity 4.04: Testing existing biofiltration systems


The infiltration capacities of thirty-seven biofilters have been tested in situ. Filter media samples
were collected for laboratory measurements of hydraulic conductivity (using standard tests) and
heavy metal concentrations (Figure 9). Over 18 sites in Melbourne, Sydney and Brisbane were
included, and at least three measurements were usually taken at each biofilter, using two different
field tests. A small part of this study has been reported on in Le Coustumer et. al. (2007), included
in Appendix D. The detailed study on hydraulic performance is reported in a joint
FAWB/Melbourne Water report (Le Coustumer et al (in review)).

Figure 9: Testing of existing biofilters

Activity 4.05 Testing Brisbane Bio-Pods


FAWB has been involved in helping BCC to build and test bio-pods, small and functional
biofiltration systems that have been retro-fritted into the urban landscape. The design (which used

11
FAWB filter media guidelines) and implementation of these systems is fully explained in Smith et.
al. (2007), included in Appendix D. The hydraulic and treatment performance of one of the systems
has been tested in two separate experiments (Figure 10). The system was replanted with FAWB
recommended plants between the two experiments. Results from the first sampling event are
presented in Smith et. al., while results from the second sampling event are still to be published,
however preliminary assessment of the results indicates that the infiltration capacity of the system is
unchanged, while the pollutant removal performance, in particularly nutrient removal, has improved
as the vegetation has matured.

Figure 10: BCC and FAWB working together on testing BCC bio-pods.

4. Key Findings
4.1 Outline of the Design

Two main configurations of biofilters are recommended, depending on the objective of the system
(e.g. target pollutants, site opportunities and constraints, etc.). They are:
- Standard biofilter design (Figure 11)
- Biofilter with a submerged anoxic zone (Figure 12)

12
AEROBIC SOIL
(unsaturated)

400- 700 FILTER


MEDIA

TREATED
STORMWATER
Transition layer 150 RIVER SAND
Drainage layer 150 GRAVEL
Drain

100mm Sub-surface Drainage Pipe on 5% grade

Figure 11: Conceptual outline of the design of a standard biofilter

AEROBIC SOIL
(unsaturated)

300- 600 FILTER


MEDIA

TREATED
STORMWATER

ANAEROBIC ZONE
(saturated) 200 SAND &
CARBON
SOURCE
Transition layer 100 RIVER SAND
Drainage layer
150 GRAVEL
Drain

100mm Sub-surface Drainage Pipe on 0% grade

Figure 12: Conceptual outline of the design of a biofilter with submerged anoxic zone

The key findings on each component of these designs are summarised in the sections below, while
full details may be found in the FAWB papers and reports, most of which are included in the
following appendices:
- Appendix A: Filter Soil Media
- Appendix B: Vegetation

13
- Appendix C: Anoxic Zone and Carbon Source
- Appendix D: Field Studies
- Appendix E: Other Issues

4.2 Soil Filter Media

To ensure reliable operation of biofiltration systems, filter media specifications must be adhered to,
in terms of both composition and hydraulic conductivity. FAWB has produced such guidelines,
which are updated as required to reflect new and relevant research insights. These guidelines are
currently under review (as discussed below), and it is anticipated that the a revised version will be
available at the end of September 2007.

The key findings on soil filter media are:


- The hydraulic conductivity should be targeted according to the specific site conditions (e.g.
the size of the system). The hydraulic conductivity of the maximum compacted media
should never be below 50 mm/hour, while 100-400 mm/hour is optimal;
- Particle Size Distribution (PSD): the clay and silt fractions (< 6 microns) should be no more
than 6 % in total, and the distribution of other fractions should be continuos;
- Organic matter should be kept on minimum;
- The total phosphorus content should be minimised, and be at least less than 100 mg/kg;
- Soils used in the filter should be structurally stable, particularly in wet conditions.

Filter media that is placed ‘uncompacted’ will initially show a very high hydraulic conductivity,
which will then rapidly decrease to the design value. It is therefore ESSENTIAL that testing of
hydraulic conductivity be conducted on compacted filter media prior to installation. FAWB has
previously recommended that hydraulic conductivity be determined using the McIntyre and
Jakobsen method (published in Practical Drainage for Golf, Sportsturf and Horticulture (2000),
Ann Arbor Press), which takes into account the effect of compaction of the soil media. However,
this test has proven to be difficult to perform, therefore FAWB is currently revising the procedure
(and is aiming to devise a simple and cost effective test).

While it is tempting to use media with a very high initial porosity (i.e. sand based filters), our study
of the hydraulic performance of non-vegetated filter media showed that such soils are very prone to
surface clogging. This is less prominent in loamy soils, however they do experience reduced
hydraulic capacity due to compaction. Therefore, sourcing sandy loam-type soils and testing them
under compaction should give a good indication of their long-term hydraulic conductivity.

It has been shown that the addition of vermiculite and perlite (around 10% each by volume) to the
soil media helps to maintain hydraulic conductivity, making the biofilter more robust to slight
deviations from the specified filter media characteristics. It may also enhance the long-term
adsorption capacity of the filter media, which is important for heavy metal removal.

Dispersive clays and silts (for example, sandy loam soils from the Western Sydney area) are
unsuitable filter media materials, owing to their unreliability in maintaining a suitable hydraulic
conductivity. The hydraulic testing of soil media under wet conditions should be able to detect
such structural instability, and they should be avoided.

14
Soil media is a key factor for removal of heavy metals. The good news is that all the sand, soil and
soil-based filter media tested (in both laboratory and field studies) demonstrated high removal (over
90%) of heavy metals. In a similar way, media is the key for removal of pathogens.

To achieve a high removal rate of phosphorus, soils should have a low phosphorus content
(FAWB guidelines recommend <100 mg/kg). This is clearly demonstrated by comparing TP
removal in the non-vegetated small diameter columns (Fig 4(a)), which leached phosphorus, to the
non-vegetated standard columns (that are part of 140 large columns-Figure 4(b)), which
demonstrated effective phosphorus removal. The soil used in the small diameter columns contained
substantially high levels of phosphorus than that used in the larger columns. Similar findings are
drawn when results from field tests of Brisbane bio-pods (designed using FAWB filter media
guidelines) are compared with the results from the Monash University carpark biofilter (where the
filter media contains over 300 mg/kg of P).

Without vegetation, most soils will naturally leach some nitrogen. The extent of leaching is
influenced by the presence organic matter, but even more so by soil moisture content. It was found
that, during dry spells, soluble nitrogen accumulates in the soil and is then washed out upon re-
wetting (huge spikes of TN have been recorded in non-vegetated soils).

4.3 Vegetation

Biofilters rely strongly on vegetation and its symbiotic relationships with bacteria and fungi for the
removal of nutrients from stormwater. However, there is marked variation in pollutant removal
(including heavy metals) among plant species. For nitrogen and phosphorus (but not metals,
which are generally effectively removed by any soil-based filter media), some of this variation (20-
37%) could be explained by plant size. However, there was still marked variation among plant
species in pollutant removal per unit plant mass. We expect that some of this variation in pollutant
removal will be due to differences among species in root architecture and physiology, leading to
variation in uptake of pollutants as well as varying effects on soil physicochemistry and the
associated microbial community

Of the species tested extensively so far (Carex appressa, Dianella revoluta, Microleana stipoides,
Leucophyta brownii and Melaleuca ericifolia), C. appressa is the best for nutrient removal (it is
suggested that this is due to rapid spreading of roots throughout the soil media, and the role of
symbiotic fungi around the root rhizosphere). In general, we are not in a position to definitively
say which plants are “good” and “bad” for pollutant removal, but it is clear that plants which are
well adapted to growing in the ephemeral wet/dry conditions of biofilters, and which have
extensive root systems, are likely to be effective. Species tested to date which proven to be
particularly effective for nutrient removal include Carex appressa, Melaleuca ericifolia, Juncus
amabilis and Juncus flavidis.

Biofilters planted with shallow-rooted plants (e.g. Microlaena. stipoides) appear to be ineffective
for nutrient removal, particularly for nitrogen, as do those which have symbiotic relationships with
nitrogen-fixing microbes (e.g. Acacia suaveolens), or those which are adapted to very dry
conditions (e.g. Lomandra longifolia, Banksia marginata). Leucophyta brownii was also
consistently poor in nutrient removal, although this may be due to its poor growth in the wet
conditions that were used in the study so far (see future research section).

15
Interestingly, the other studied species (see Read et. al. in Appendix B) all performed very similarly
and did not appear to significantly influence nitrogen removal.

There was no evidence that plants were physiologically stressed by application of stormwater in this
experiment. However, measures of whole-plant performance may show effects that were not
detected by fluorometry.

4.4 Anoxic Zone and Carbon Source

The presence of an approximately 450 mm deep, permanently submerged anoxic zone (consisting
of sand or gravel) with a carbon source such as hardwood chips (around 5% by volume) will
largely improve nitrate/nitrite (NOx) removal, by promoting denitrification. The presence of an
anoxic zone without a carbon source did not achieve improved NOx removal.

An anoxic zone with carbon is also beneficial for heavy metal removal. This is particularly the
case for copper (Cu), where only systems with this design feature were able to meet the ANZECC
water quality targets for aquatic water health.

However, one of the most important benefits of these permanently wet zones is their ability to
support plant survival during dry periods. This has a huge implication on the treatment
performance; the vegetated systems without an anoxic zone and carbon, that remove over 60% of
TN during regular wetting, begin leaching TN after only three weeks of dry weather and take
longer to recover upon re-wetting. The same systems with an anoxic zone and carbon will ‘fall-
apart’ only after seven weeks of dry weather. Even then, they recover relatively quickly.

However, there are some side effects of anoxic zones. Low levels of ammonium production were
observed, and they are a source of pathogens. If biofilters are to be used for stormwater
harvesting, anoxic zones should be avoided!

4.5 Hydraulic Conductivity

The use of appropriate soil media is an important factor in achieving reliable hydraulic functioning
of the biofilters. If sized accordingly, systems with a hydraulic conductivity of around 100-300
mm/hour (under compaction) should be operational over a considerable period of time.

Biofilters will experience a sharp drop in hydraulic conductivity immediately following


construction (due to compaction and, to a small extent, surface clogging). However, one of the key
findings from the field systems is that hydraulic conductivity will recover over time (at the Monash
carpark system, the hydraulic conductivity was initially 300 mm/hour and decreased to less than 50
mm/hour in the first six months, but then within a year recovered to 200+ mm/hour). We believe
that plant growth is the cause of this recovery (through the creation of macropores by plant roots).

In systems that are under-designed (too small for their catchment) or service catchments with high
silt loads, surface clogging is an existing problem (this is a key finding from our lab study).

16
4.6 Treatment Performance
Vegetated, soil-based biofilters are shown to be effective for the treatment of stormwater, reducing
inflow TSS and TP concentrations by approximately 98 and 85%, respectively. To achieve a high
removal rate of TP, the filter media should have a low phosphorus content (see soil media section).

At least 50% removal of nitrogen is also achievable, but is dependent on vegetation selection. For
both nitrogen and phosphorus removal, C. appressa is very effective. M. ericifolia has taken longer
to mature, however it is also showing signs of facilitating high levels of nitrogen and phosphorus
removal. Selecting filter media which do not have excessive levels of organic matter will help to
prevent leaching of phosphorus, while the biggest variations in TN removal are caused by variable
wetting and drying (with treatment efficiency decreasing substantially after long dry spells).

Importantly, treatment performance will also be reduced if the biofilter is small relative to its
catchment. Whilst no definitive guidance is yet available, a sizing of around 2% of the catchment
area appears to give satisfactory performance for removal of TSS, TP and TN. Further research is
being undertaken to increase nutrient removal, in particular through the retrofitting of an anoxic
zone to ‘poorly-performing’ biofilters.

All biofilters configurations tested (both vegetated and non-vegetated) remove more than 90% of
heavy metals (both particulate and dissolved).

The presence of an anoxic zone with carbon will boost the removal of NOx to over 90%, and to
some extent TN (to 70%). It will also improve Cu removal to meet the ANZECC guidelines for
stream health. However, the most beneficial effect of anoxic zones is buffering the negative impact
of extended dry periods on nitrogen removal.

Preliminary results also show that biofilters may be very effective for the removal of pathogens
(bacteria, viruses and protozoa); this work will be published by FAWB within the next 12 months.

4.7 Construction and Maintenance

Some degree of leaching of fine sediment and nutrients from the soil media will usually occur
during the establishment phase, until the soil has stabilised and plant roots have occupied the soil
volume (this will typically take 2-6 months).

Effective communication between designers and construction contractors is essential, throughout


all stages of the project. It is imperative that quality control issues are addressed in planning and
design, and construction and maintenance throughout the life of the biofiltration system, and that
the design intent is communicated to the contractors, at a pre-construction briefing.

Maintenance requirements could be high during the establishment phase; frequent weed removal is
required and the juvenile vegetation should be watered during extended dry periods. However, the
need for this level of maintenance reduces significantly as the vegetation matures. The
development of mosses on the surface should be discouraged, as these can reduce the hydraulic
capacity of the system. Dense planting of the preferred plants at the time of construction will help
to minimise the extent of weed invasion, and minimise any moss growth.

17
4.8 Lessons learned from practice so far

A study of 37 biofilters, constructed on the east coast of Australia during the last seven years,
shows that 40% of constructed systems have a hydraulic conductivity (K) below 50 mm/hr (the
currently recommended minimum hydraulic conductivity). It appears that the media used varied
greatly between systems, perhaps because of a lack of available guidelines, or because of
inadequate specification and quality control. However, this does not affect the treatment efficiency
of the systems, since most systems are over-sized and their detention storage volume compensates
for reduced media conductivity.

The study broadly revealed two types of systems: some with a high initial K (>200 mm/h) and
some with a low initial K (<20 mm/h). Significant reductions in K are evident for biofilters in the
former group, although most are shown to maintain an acceptably high conductivity. For the
second type of systems (with low initial K), little change occurs over time. Two hypotheses could
explain this phenomenon: on one hand, sediment depositions could be leading to the clogging of
the surface of the system, on the other hand, the creation of macropores through root growth and
dieback may help to minimise the reduction in K. The impact of surface clogging is proportionally
greater in systems which started with a high initial K, most likely because the difference in particle
size distribution between the original filter media and deposited sediments will be greater where the
original media was coarse. In the systems with low initial K, the finer particle size distribution will
be more similar to that of the inflow sediments (although still considerably larger), thus reducing
the proportional impact of any surface clogging effect.

Site characteristics such as filter area as a proportion of catchment area, age of the system and
inflow volume were not found to be useful predictors of media conductivity, with initial
conductivity of the original media explaining the vast majority of variance. It is clear therefore,
that strict attention must be paid to the specification of original filter media, to ensure that it
satisfies current design requirements.

4.9 Other Issues

Biofiltration systems constructed in sodic soils without impermeable lining are not at risk of
exporting salt from in situ soil into local streams. Even after six months of intensive flushing under
controlled, laboratory conditions, they did not leach salt from the surrounding soils.

5. Future research
5.1 Vegetation
While we have shown that some species are more effective than others in pollutant removal
(primarily for nitrogen), it is not certain that the same trends among species will occur in differing
environments, or when plants are grown in competition with other species. We are currently testing
the same species under varied wetting and drying regimes, to determine whether certain species,
which are not effective in nitrogen removal in frequently-wet environments, perform better under
drier conditions.

18
5.2 Soil Media
We are still to find an accurate test for assessment of initial hydraulic conductivity in ‘as built’
systems that could be easily used during construction.

Engineered media (instead of naturally sourced media) should be tested prior to widespread
adoption. They may provide a solution to the problem of ensuring media stability and satisfactory
initial hydraulic performance, but they still need to be tested for leaching and clogging.

We are also awaiting results on break-through of pollutants to assess the long-term sustainability of
the soil media. We hope to be able to provide guidance on how many years a system may operation
for, before the media becomes saturated and starts leaching.

5.3 Hydraulic Conductivity


Clogging issues, as well as soil structure changes and stability, are long-term processes that will
require monitoring over a number of years.

5.4 Anoxic Zone and Carbon Source


The long-term impact of this design feature should be assessed. The length of time for which the
specified carbon source will last is hypothesized, but ongoing tested is needed to verify these
predictions.

5.5 Treatment Performance


Further testing of biofilters in field conditions is essential. It is important to monitor new
systems (hopefully built according to the findings presented above) in order to gain verify the
specified design. These systems, if not designed correctly, may act as sources of pollution
and therefore should be carefully designed, constructed and monitored.

19

You might also like