See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/281930800

Human and environmental health risks and benefits associated with use of
urban stormwater

Article  in  Wiley Interdisciplinary Reviews: Water · September 2015

DOI: 10.1002/wat2.1107

CITATIONS READS

27 4,221

5 authors, including:

Sunny Jiang Keah Ying Lim

University of California, Irvine University of California, Irvine
111 PUBLICATIONS   7,005 CITATIONS    13 PUBLICATIONS   414 CITATIONS   

SEE PROFILE SEE PROFILE

Eric (Xiao) Huang David Mccarthy

California Institute of Technology Monash University (Australia)
28 PUBLICATIONS   709 CITATIONS    229 PUBLICATIONS   6,086 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Rural mental health. View project

From stormwater to potable water via Water Sensitive Urban Design? View project

All content following this page was uploaded by Eric (Xiao) Huang on 29 December 2017.

The user has requested enhancement of the downloaded file.

 Advanced Review

Human and environmental health

risks and benefits associated with
use of urban stormwater
Sunny C. Jiang,1* Keah-Ying Lim,1 Xiao Huang,1
David McCarthy2 and Andrew J. Hamilton3

For stormwater harvesting to achieve its full potential in mitigating water scarcity
problems and restoring stream health, it is necessary to evaluate the human and
environmental health risks and benefits associated with it. Stormwater harbors
large amounts of pollutants and has traditionally been viewed as a leading cause
of water-quality degradation of receiving waters. Harvesting stormwater for house-
hold use raises questions of human exposure to pollutants, especially human patho-
gens, which have the potential to cause large-scale disease outbreaks. These issues
are compounded by uncertainties relating to the performance of stormwater treat-
ment technologies in pathogen removal. Quantitative microbial risk assessment
provides an objective risk estimate based on scientific data and the best assump-
tions, which can be used to educate and instil confidence in stakeholders of the prac-
tice. Although limited, human health risk studies have positively supported the use
of minimally treated rainwater and stormwater for some non-potable applications.
In addition to the well-known benefit of preserving the stream hydrology and ecol-
ogy, wetlands used for harvesting stormwater can also provide new habitats
for wildlife that benefit environmental health. A fundamental change from viewing
stormwater as waste to resource requires the coordinated efforts in research, edu-
cation, and effective communication. © 2015 Wiley Periodicals, Inc.
How to cite this article:
WIREs Water 2015, 2:683–699. doi: 10.1002/wat2.1107

INTRODUCTION coverage and stream flow.1–3 Efforts to prevent and

reverse urban stream syndrome require harvesting a

P opulation growth, rapid urbanization, and climate

change have been straining our traditional water
resources and degrading the environment. Several
major portion of stormwater running off impervious
surfaces and using the harvested water for local irriga-
tion of landscape to improve evapotranspiration, or
recent studies and reviews have analyzed the impact for other non-potable uses that involve exporting the
of urban development on natural hydrological system harvested water through sewage systems.4 This review
and coined the term ‘urban stream syndrome’ to paper uniquely focuses on the validity of use of har-
describe the complex issues related to change in land vested stormwater for local applications. Harvesting
is hereby defined as collection of surface runoff
from impervious surfaces, which is consistent with that
*Correspondence to: sjiang@uci.edu
1
used in the previous reviews on stream hydrology. As
Civil and Environmental Engineering, University of California,
Irvine, CA, USA such, groundwater infiltration with the potential for
2
Department of Civil Engineering, Monash University, Melbourne,
groundwater recharge is not included in this review
Australia as harvesting. Furthermore, this paper will not dupli-
3
The University of Melbourne, Federation University Australia, Mt cate the previous reviews4 on quantity of the storm-
Helen Campus, Australia water that should be partitioned between infiltration
Conflict of interest: The authors have declared no conflicts of interest to recharge groundwater aquifer and harvesting for
for this article. local use.

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 683

 Advanced Review wires.wiley.com/water

Extending from the benefit of reversing the urban is an important component of stormwater harvesting, a
stream syndrome, an additional benefit of stormwater clear distinction is made between the quality of rainwa-
harvesting lies in applying the minimally treated storm- ter harvested directly from rooftops in the rain tanks
water or untreated rainwater for non-potable household and precipitation collected in stormwater channels.
applications to alleviate the pressure on drinking water Rooftop-harvested rainwater is less polluted with
resource. Stormwater harvesting presents the opportu- chemical and microbial contaminants than storm-
nity to shift the urban water management paradigm water, which mobilizes accumulated pollutants from
from the traditional view of stormwater as a pollutant the ground (i.e., from motor vehicles, animal wastes,
and flood risk to the view of new water resource.5 How- and lawn maintenance). In addition to precipitation,
ever, the quality of stormwater and potential health con- urban runoff from landscape and agriculture irriga-
cern associated with household use should be evaluated. tion, car washes and wash down of grounds also con-
Treatment systems commonly used for improving tribute to stormwater in metropolitan areas (Figure 1).
stormwater quality (e.g., constructed wetlands, biofil- During dry-weather conditions, urban runoff is the
ters) should be examined if stormwater is used in direct main source of stormwater in storm drain system.
human contact activities (showering, laundry, home In many cities, the underground storm channels have
crop gardens irrigation, etc.). Among the array of health connectivity with groundwater through seepage.
hazards that may be carried by stormwater, microbial However, the contribution of the groundwater to the
pathogens are the most important public health concern stormwater flow is highly variable and poorly quanti-
due to their low dose of infection, the acute nature, the fied. Cross contamination of stormwater and sewage
secondary transmission, and the potential for large-scale due to aging infrastructure, poor design, and poor
disease outbreaks. There is also a lack of understanding implementation has also been reported,11,12 where
of microbial removal efficiency because the earlier the degree of seepage and exchange vary significantly
stormwater harvesting systems have focused on removal from city to city (Figure 1). The diversity of stormwater
of total nitrogen, phosphorous, and other chemical con- sources leads to considerable variability in both water
stituents. This paper focuses on the potential human quality and quantity.
health risks posed by microbial contaminants during
household uses of stormwater, and human and environ-
mental health benefits beyond the well-studied stream Microbial Pathogens
ecology and drinking water substitution.6–8 Stormwater quality studies have traditionally focused
on physiochemical parameters, nutrients, heavy
metals, and micropollutants due to their accumulative
MICROBIAL QUALITY OF environmental damaging effects on the receiving
STORMWATER waters. The National Stormwater Quality Database
In cities of developed countries, stormwater are col- in the U.S. collected more than 10 years of stormwater
lected from impervious surfaces by extensive networks monitoring data from the National Pollutant Discharge
of engineered underground storm drainage systems9 Elimination System MS4 (municipal separate storm
(Figure 1). In most cities, urban stormwater empties sewer system) stormwater permit program, which
directly to rivers, creeks, or coastal waters without includes about 9000 rain events across the U.S. with
any treatment. In other cities where storm drain and information of more than 125 different stormwater
sewer collection systems are combined, stormwater is quality constituents. Fecal indicator bacteria (FIB,
piped to sewage treatment plants together with munic- i.e., Escherichia coli and fecal coliforms also known
ipal sewage for treatment. These combined systems can as thermotolerant coliforms) are the only microbial
present serious problems during wet weather, when the data collected in the database.
stormwater overwhelms the design volume of sewage MS4 stormwater permit program data from the
treatment plants and has to be released (with raw sew- U.S. and a large volume of international literature con-
age) to the receiving water without any treatment.10 sistently showed very high concentrations (i.e.,
During this combined sewer and stormwater overflow >10,000 CFU/100 mL) of FIB in stormwater and in sur-
(CSO) event, the receiving water is severely contami- face water receiving storm runoff. FIB concentration is
nated by raw sewage. influenced by the intensity of watershed development,
stream flow, and antecedent precipitation.13 But the
first-flush loading of FIB is not always seen, indicating
Source Matters that the FIB may have an ecological origin rather than a
Rainwater and snowmelt are the main sources of direct fecal source.13 The adequacy of using FIB to indi-
stormwater (Figure 1). Although rainwater harvesting cate human pathogens in stormwater has been

684 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

(b)
(a)
(a)
(b)

(a)

(c)
(b) (b)

(c) (d) (d) Retention/treatment

(c)
(e)
Storm channel
(d)
Sanitary sewer system

FI GU RE 1 | Schematic illustration of urban stormwater generation and harvesting scenarios. Precipitation or irrigation runoff from impervious
surface is collected either directly to underground storm channels through street gutters (a) or infiltrated through sandfilters/biofilters (b) and entering the
channels through the perforated underdrains (c). The close approximation between sanitary sewer lines and underground storm channels may cause cross
contamination of stormwater with sewage due to aging infrastructure, poor design and poor implementation (d). Stormwater harvesting is achieved by
piping the stormwater from main channel (e) to retention/treatment systems such as artificial wetlands and recharge basin.

questioned.14,15 Numerous studies have been carried technical challenges in recovering pathogens from
out to identify the sources of fecal contamination in stormwater directly (Table 1).
waters with high concentrations of FIB using microbial In spite of the lack of consistent and quantitative
source-tracking technology in order to better under- data, numerous studies have shown the presence of
stand the microbiological health risk associated with microbial pathogens, including Giardia, Cryptospori-
stormwater.15 dium, toxic E. coli, Salmonella, Campylobacter,
The quantitative data on human pathogens in and human viruses, in stormwater and surface water
stormwater are sparse, which is largely due to the dif- receiving storm runoff (Table 1). Although rooftop-
ficulties of detecting and quantifying pathogens. Path- harvested rainwater is well recognized to have
ogen detection requires concentration of large volumes better water quality than harvested stormwater in
of water. The presence of relatively high concentrations many studies, it is not free of microbial pathogens.
of suspended solids (>2000 mg/mL)16 and grease Pathogens including Giardia, Cryptosporidium,
(>1000 mg/mL)17 in stormwater significantly chal- Salmonella, Camplyobacter and Legionella pneumo-
lenges the technology that can be applied for concentra- phila were found in rainwater tanks tested in
tion and recovery of pathogens. There are also large Australia, Denmark, France, New Zealand, UK, and
amounts of hydrocarbons and heavy metals (>5 mg/ USA (Table 1).
L) that inhibit the molecular detection method used To embrace the concept of stormwater harvesting
to identify and quantify a specific pathogen.16 Several for human uses, it is of paramount importance to
reports have indicated that microbial pathogens are remove human pathogens in stormwater during the
more frequently detected in the receiving water near practice, which will prevent the potential outbreaks
storm drains than in the stormwater itself due to the of infectious diseases (Figure 2).

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 685

 Advanced Review wires.wiley.com/water

TABLE 1 | Microbial Pathogens in Rain Tank Water, Stormwater, and Receiving Water Near Storm Drains
% positive (total samples) or concentration/L
Pathogens Rainwatera Stormwatera Receiving watera References
Cryptosporidium spp. 0% (20) 18
0% (214) 19
37% (59) 20
77% (120), 21
0.04–1.5 oocysts
35% (17) 22
4% (125) 23
40% (20), 24
0.07–0.31 oocysts
Giardia spp. 19% (21) 25
8% (214), 19
0.6–3.6 cysts
13% (24), 26
120–580 cysts
0% (20) 18
19% (59) 20
0% (17) 22
7% (14) 27
0% (125) 23
40% (20), 24
0.05–3.77 cysts
50% (2) 28
Camplyobacter spp. 4% (27) 25
0% (214) 19
17% (23), 26
5–110 cells
1.5% (67) 29
13% (42) 18
3% (59) 20
94% (54), 30
<1–43 MPNIUb
96% (23) 12
12% (17) 22
0% (142) 31
37% (24) 32
0% (125) 23
Legionella spp. 27% (27) 25
4% (214), 19
60–170 cells
15% (67) 29
2% (54), 30
10,000 cells
71% (7) 22
7% (14) 27
0% (418) 31
Salmonella spp. 11% (27) 25

686 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

TABLE 1 | Continued
% positive (total samples) or concentration/L
Pathogens Rainwatera Stormwatera Receiving watera References
7% (214), 19
65–380 cells
4% (24), 26
7300 cells
3% (67) 29
0% 18
(n > 60)
32% (22) 12
0.1% (798) 31
0.9% (125) 23
Adenovirus 65% (23) 12
3% (59) 20
91% (23) 33
61% (18), 34
4100–38,000 gcc
42% (52), 35
250–7000 gc
29% (14), 36
1500–6500 gc
290–400 gc 37
9240–12,400 gc 38
33% (12), 14
880–7500 gc
52% (21) 39
16% (114), 40
220–60,000 gc
2% (61), 41
30–8430d gc
Norovirus 75% (52), 35
135–19,000 gc
1270–147,000 gc 42
a
Rainwater represents rooftop-harvested rainwater stored in rainwater tank. Stormwater represents stormwater runoff collected from storm drains, ditches, or
stormwater outfall. Receiving water represents water samples collected from surface water that are affected (or potentially affected) by stormwater discharges.
b
MPNIU is most probable number of infectious units. The range is reported as lower detection limit value and maximum observed value.
c
gc is genome copy.
d
Represents range of detection limits for the 60 non-detected samples. The only positive sample is within this range.

MICROBIAL REMOVAL DURING The concept of utilizing stormwater as non-

STORMWATER HARVESTING potable water supply is not new but has only recently
received appreciable attention.43 For example, the
Treated stormwater is suitable for various purposes Santa Monica Urban Run-off Facility (SMURF) in Cal-
depending on the treatment technology applied. ifornia, USA, has been harvesting/treating urban runoff
Non-potable applications such as landscape irrigation, from its main stormwater drains for landscape irriga-
car washing, and toilet flushing (Figure 2) are the most tion and toilet flushing since 2001. Rainwater and
common type of end-uses practiced in different coun- stormwater have been used in various places in
tries.20,43 Applications that involve much closer contact Australia for various non-potable purposes for decades
with the water, such as showering and swimming pool but are not well documented or studied. Chlorination
filling, are theoretically possible if the stormwater is trea- and UV disinfection are often used for disinfection to
ted adequately to ensure human safety. ensure the safety of stormwater applications.20 The

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 687

 Advanced Review wires.wiley.com/water

Showering and toilet-flushing:

Both actions can produce respirable
micron-sized respirable aerosols

Laundry:
Human exposure to pollutants in
laundry water may occur either
through inhalation of aerosols during
laundry or transmitted to human
through handling of wet laundry

Secondary transmission of Consumption of fresh produce

disease can occur through can expose consumers to
direct/indirect contact pollutants carried by the
between individuals stormwater-irrigated produce

Children playing outdoor may

come into contact with lawn
irrigation water and/or soil
dampened by the water

Food crop irrigation:

Car washing: Aerosols generated during crop
Lawn irrigation: irrigation can expose individual to
Pressurized water splashed against car’s Sprinkling action can produce pollutants carried in irrigation water
surface is known to produce micron-sized micron-sized respirable aerosols through inhalation. Pollutants can also
respirable aerosols
attach onto the crop’s surface

F I G U R E 2 | Schematic illustration of route of human exposure to stormwater pollutants during different stormwater use. Inhalation of aerosols
during laundering, showering, toilet flushing, car washing, lawn and home garden irrigation represents the major route of human exposure to pollutants
carried in the stormwater, while consumption of fresh produce irrigated by stormwater exposes human to pollutants through ingestion. Secondary
transmission during person-to-person contact also contributes to an important portion of human exposure.

modern stormwater harvesting practice focuses Rainwater Harvesting System

on using low-impact development (LID), natural sys- Rain tank is the earliest LID system used to capture
tems, or green engineering for treatment. With subtle rainwater for later use and has been reviewed exten-
differences, all three terms describe similar approaches sively elsewhere.29,47,48 While many variations of rain-
for managing stormwater. The U.S. EPA defines LID water harvesting systems have been developed, all are
as ‘an approach to land development (or re- made up of the same basic components, which include
development) that works with nature to manage storm- a collection area, a conveyance system, a first-flush
water as close to its source as possible’.44 This differs diversion/screening system, and a storage system. Rain-
from the traditional engineering practice of harvesting water is most commonly collected from the rooftop of
stormwater using constructed pipes, tanks and con- buildings before it reaches the ground. While rooftops
crete structures because LID adds the direct benefits and rain gutters are less contaminated (at least for most
of environmental preservation. Rain tanks, artificial cases where atmosphere deposition is not the major
wetlands, biofilters/bioswales are some examples of contributor of pollutants), they are subjected to the
the LID systems. accumulation of tree litters, atmospheric deposits, ani-
There have been a number of reviews focusing mal feces (e.g., birds, squirrels), and traces of weathered
on stormwater treatment technologies, costs, and eco- roofing materials (e.g., heavy metals).47 Immediately
nomical trade-offs.45,46 These previous reviews sum- after a rain event, these contaminants are rinsed off
marized the state of technology at the time for in large amount, producing a rather polluted initial
removal of nutrients and chemical contaminants in runoff from the first 0.25 cm of rain. These ‘first
stormwater treatment systems and pointed out that flushes’ are usually discarded in the modern design of
there was a lack of information on removal of microbial rain tanks using diverters to reduce pollutants in the
pathogens. collection tank. In addition, leaf screens and mesh

688 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

filters are usually employed before the storage tanks to governing the removal of microbial pathogens in wet-
prevent introduction of debris and pests (e.g., mosqui- lands. Bavor et al.56 indicated that the establishment of
toes) into the tank. Intuitively, the water quality of har- vegetation could improve the removal of FIB through
vested rainwater is highly variable. There is essentially the enhanced sedimentation. In contrast, Hathaway
no direct engineering measure for the reduction of et al.57 concluded a lower plant coverage in wetland
microbial pathogens in the rain tank. The fate of the improved FIB removal due to high exposure to sun-
microbial pathogens and FIB in rain tank involves both light. The variability observed is likely caused by the
natural decay and regrowth due to complicated envi- size of the wetland, the residence time, the quality of
ronmental conditions. The biofilm growth in the rain influent, and local conditions. Residence time is per-
tank is controversial; some suggested the benefit of bio- haps the most important factor in controlling the
film at adsorbing trace metals and other pollutants,49 removal rates of microbial pathogens in the wetland,
while others indicated the risk of supporting the sur- but is ignored by many studies. Struck et al.58 found
vival and growth of human pathogens.50 There are sev- the removal of FIB generally followed the first-order
eral reports of presence of pathogens and FIB in the rain decay model as a function of time. The decay rate is sig-
tank water (see Section Microbial Removal During nificantly faster in the first 50 h over the 100-h study
Stormwater Harvesting) but no report on pathogen fate period. These results imply that neither sedimentation
in the rain tank. The Australian Guidelines 23 for nor sunlight exposure will be sufficient to treat the
stormwater harvesting and reuse relied on the ratio rapid flow of stormwater through the wetland when
of FIB–to-pathogen in human sewage to determine the residence time is less than 50 h, especially during
the pathogen reduction for use of rainwater.20 The a heavy rain event.
new field data that are being collected throughout the According to the International Stormwater BMP
world will likely improve our current understanding Database (ISBD) 2014,59 the average removal rate of
of the fate of pathogen in rain tank and the health risk fecal coliform, E. coli, and Enterococcus in the five wet-
associated with uses of rainwater. lands investigated were 91, 53, and 61% (Table 2).
Similar or slightly higher rates of removal are also
reported in two Australian studies (Table 2). Based
Retention Ponds and Wetlands on the current data, none of the wetland effluent can
Retention ponds are constructed basins that are mainly meet the microbial water quality criteria for primary
used to mitigate the peak flow during rain events. Con- contact recreation (30-day geometric mean for E. coli
structed wetlands are similar to wet retention ponds <126 CFU/100 mL, Enterococcus < 35 CFU/100 mL),
that incorporate plants in shallow pools. Besides peak which indicates potential health risks are associated
flow mitigation, wetlands are also used for stormwater with the wetland-treated water. However, others argue
quality control. Typical wetland designs include a deep that the low removal rate of FIB is due to its regrowth in
pond at the inlet (sediment forebay) to decrease the wetlands. The behavior of pathogen in wetland can be
water velocity and sediment load, followed by shallow significantly different than that of the FIB. So far, there
water areas with wetland plants, and outlet structures is no reported study on specific pathogen removal in the
to control the hydraulic regime of the wetland. For the stormwater wetland, which highlights the need for
proper functioning of the system, a permanent flow future field scale studies to address this gap.20
condition is required to support the growth of wetland
plants.51,52
The transport and fate of microbial pathogens Biofilters
have been extensively studied in constructed waste- Biofilters are primarily designed to treat and attenuate
water wetlands.53,54 However, very few (if any) similar stormwater runoff for a specified water volume.44
studies have been carried out in stormwater wetlands, While they provide limited flood retention capacity
which is likely due to the lack of tools for detecting low compared to wetlands due to their relatively smaller
concentration of human microbial pathogens in storm- footprints, biofilters easily enable stormwater treat-
water. Researchers have relied on FIB or indictor ment in urban settings and established communities.
viruses to understand the fate of microbial pathogen The typical designs of biofilters (e.g., grassed channel,
in stormwater wetlands, in spite of all the well-recog- dry swale, wet swale, or bio-swale) allow stormwater
nized disadvantages of this approach, such as the to flow through the systems horizontally or vertically.
regrowth of indicators under favorable conditions Pollutants in the stormwater are removed through bio-
and the contribution of nonhuman sources.55 Sedimen- logical uptake by plants and biofilm, filtration through
tation, sunlight exposure, water temperature, and the a subsoil matrix, and/or infiltration into the underlying
adsorption to biofilms are considered as main factors soils. In recent years, several studies have been

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 689

 Advanced Review wires.wiley.com/water

TABLE 2 | Microbial Removal Efficiency of Stormwater Water Treatment Systems

Treatment Microorganisms Removal Efficiencies Country of Study Sites References
Wetland E. coli 33–96% U.S. 57
53% U.S. 59
Enterococci 1 log10 Australia 60
61% U.S. 59
Fecal coliform 56–98% U.S. 57
91% U.S. 59
0–2 log10 Australia 56
Wet retention pond E. coli 46% U.S. 61
Fecal coliform 54–99.8% Lab study 62
84% U.S. 59
70% U.S. 61
0–0.5 log10 Australia 56
Biofilter E. coli 79–93% Australia 63
3 log10 Israel 64
1–2 log10 Lab study 65
97% Lab study 66
Enterococci 79–92% Australia 63
Fecal coliform 2 log10 Israel 64
C. perfringens >97% Australia 63
F-RNA coliphage 1–5 log10 Lab study 65
82% Lab study 66
1–3 log10 Australia 63
Adenovirus <1 log10 Australia 63

conducted to understand the fate of microbial patho- the system.69 It should be cautioned that most of these
gens in the biofilters. Using laboratory-scale biofilters, conclusions are based on laboratory-scale studies due
these studies showed that plants play a crucial role in to the lack of field studies.
the removal of microbial pathogens, yet the mechan-
isms of interaction between plants and the seeded indi-
cator bacteria and viruses are still not clear.63,65 Cost and Water Quality
Traditional filter media, such as sand, soil, zeolite, The cost of water treatment often increases with
and anthracite, are often used for biofilter construction; improvement in microbial quality of the finished water.
improved microbial pathogen removal efficiency were Additional treatments including microfiltration, UV
observed in media incorporated with Cu and Zn com- radiation, and chlorine disinfection can be included
pounds.67 Studies also showed that the microbial path- to further reduce pathogen loads in finishing water.
ogen removal efficiency is significantly affected by the Most of the current cost estimates for stormwater treat-
dry and wet weather conditions. Adding a submerged ment, however, do not include the cost of disinfection.
zone at the bottom of the biofilter improves the micro- Among 12 stormwater harvesting cases studied in New
bial removal efficiency of biofilters because it prevents South Wales (NSW), Australia, only two included dis-
the formation of fine fissures and macro pores during infection processes into the cost estimation. The cost
extended dry period.65,68 Overall, these myriad designs structure of stormwater harvesting is highly complex
and operational characteristics of biofilters have led to and variable, which includes capital costs, recurrent
inconsistent pathogen removal efficiencies as reported costs, and water quality benefits unit costs. Although
across the literature (Table 2). Human viral pathogen it is difficult to compare between countries and regions,
was only tested once in a biofiltration laboratory-scale recent studies in NSW have indicated that the average
study and showed less than one log removal through levelized cost for treated stormwater is higher than the

690 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

mains water prices in the Sydney Greater Metropolitan applications that are familiar to and more acceptable
Area in 2005–2006.70 An evaluation of the water sup- to the public (e.g., irrigation). This also compares with
ply options for Melbourne that compares traditional the public’s perception of wastewater reuse, where indi-
supply sources and alternative sources indicates rect potable reuse of recycled wastewater is usually fer-
stormwater harvesting as the lowest cost option for vently opposed by the public when essential
greenfield development among options including was- information about the safety of the water is not com-
tewater recycling and rainwater harvesting.71 How- municated effectively.75 However, the situation can
ever, the finishing stormwater quality is not intended be the opposite if the risks are communicated transpar-
for direct human contact and did not include the ently with all the stakeholders involved, which was the
cost of additional treatment beyond wetland case for the renowned Orange County Water District’s
treatment. Dandy et al.72 used City of Salisbury, South indirect potable use of treated wastewater.76
Australia, as a case study to demonstrate the frame-
work and tools needed for estimating a number of Estimated Risk
stormwater harvesting options. The report emphasized Most often, the variation of perceived risks among indi-
the importance of incorporating a broader multi- viduals can be reduced through estimating and commu-
criteria analysis to economic, environmental, and nicating the risk objectively. Estimated risk is a
social criteria. technical assessment of risk through the collection
and application of scientific facts that are relevant
to the risk. For example, the potential risk of storm-
RISKS OF USING STORMWATER water use is first identified by an understanding of
the hazards in stormwater. The degree of exposure to
Human Health Risk
these hazards can differ from one type of water appli-
It is important to recognize that all human activities
cation to another, marked by the different levels
involve risk. It is appropriate to compare the risk of a
of water contact (i.e., toilet-flushing vs. showering), fre-
new water practice with existing standards or water
quency of the water application (i.e., daily vs. weekly),
supplies. In terms of the risk associated with human
and also disease transmission routes (i.e., breathing
use of harvested stormwater, health risk can be defined
contaminated aerosols vs. ingesting contaminated
as actual risk, perceived risk, and estimated risk.
water/food, see Figure 2). Finally, the risk is determined
Through analyses of these risks, decisions on water uses
by the pathogen’s ability to reach infection site in
can be made to advance and refine the practice.
human body, the pathogen’s potency to induce an
infection, and an individual’s immunity against the
Actual Risk pathogen.
The actual risk of stormwater use can be expressed The quantitative microbial risk assessment
as the number of people whose health statuses are com- (QMRA) has been adopted to provide a scientific basis
promised (i.e., hospitalized) through using treated to model these risks.73 Estimated risks of water appli-
stormwater. While epidemiological studies may offer cations are generally compared to the acceptable drink-
a glimpse into the actual risk of such practice, ing water risk benchmarks recommended by U.S. EPA
the results can be compromised by manifold of uncer- and WHO for safety assurance. These benchmarks are
tainties. For example, people who use treated storm- set at threshold of ≤1 infection case/10,000 persons-
water can get ill from many different sources of year by U.S. EPA and ≤1 DALY/1,000,000 persons-
contamination, such as eating contaminated food in year by WHO.77,78 Table 3 shows a summary of risk
a restaurant or through secondary infection from estimates associated with the applications of untreated
another person. There are also challenges in conducting rooftop-harvested rainwater and treated urban storm-
ethical experiments that can validate the actual risk. water from four QMRA studies. Based on these risk
Moreover, stormwater harvesting practice is still estimates, harvested stormwater is only suitable for toi-
in its early development and collection of the relevant let flushing, whereas rooftop-harvested rainwater is
epidemiological data is inevitably difficult, if not suitable for various applications including toilet flush-
impossible. ing, showering, and garden hosing. While the absolute
risk of using rainwater for food-crop irrigation is ques-
Perceived Risk tionable, a comparative risk analysis has shown that
Perceived risk of stormwater varies widely from person rainwater-irrigated food-crops present risks that are
to person due to the lack of understanding of storm- at least 10-fold lower than food-crops irrigated using
water harvesting and effective public education.73,74 reclaimed wastewater, which is a common agricultural
This has resulted in conservative stormwater practice.80,82 Overall, these risk assessments indicate

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 691

 TABLE 3 | Quantitative Microbial Risk Assessment of Human Health Risks Associated with Use of Rooftop-Harvested Rainwater and Stormwater

692
Risk Estimate
Infection cases/ DALYs/
Type of Target 10,000 persons- 1,000,000 Frequency of Applications
Water Applications Pathogen Exposure Route yeara persons-yeara Considered References
Advanced Review

Rainwater Toilet flushing Camploybacter Aerosol ingestion 0.00004–1 0.01–400 Applications are engaged six 79
spp. (0.2) (70) times per day, but exposure
events per year varies from
0–219 times. (based on
0–30% pathogen detection).
Drinking Salmonella spp. Liquid ingestion 10–53 Applications are engaged daily, 19
Giardia lamblia 20–130 but only 18 exposure events
per year (based on 5%
Showering Salmonella spp. Aerosol ingestion 0.02–0.1
pathogen detection).
Giardia lamblia 0.04–0.2
L. pneumophila Aerosol inhalation 0.003–0.007
Garden hosing Salmonella spp. Liquid ingestion 0.01–0.08 Applications are engaged
Giardia lamblia 0.03–0.2 biweekly, but only five
exposure events per year
L. pneumophila Aerosol inhalation 0.002–0.006
(based on 5% pathogen
detection).
Foodcrop irrigation (including Salmonella spp. Ingestion of 0.3–27 0.2–16 Food crops are consumed daily. 80
tomatoes, lettuce, and contaminated (1–11) (0.9–7) Every infection case is an
cucumber) Giardia lamblia foodcrop 3–77 0.8–21 illness case.

© 2015 Wiley Periodicals, Inc.

(6–55) (1–4)
Stormwaterb Toilet flushing Adenovirusa Aerosol inhalation 0.0002–0.01 0.001–0.03 Applications are engaged four 81
(0.001–0.002) (0.003–0.005) times daily, totaling up to
Norovirusc Aerosol ingestion 0.003–2 0–5 × 10−9 1460 exposure events per
(0.005–0.3) (0–1 × 10−16) year.
Showering Adenovirusa Aerosol inhalation 0.0006–21 0.002–56 Applications are engaged daily.
(0.004–0.6) (0.01–2) Each event lasted for 20 min.
Norovirusc Aerosol ingestion 1–6,954 (4 × 10−9–6 × Aerosol distributions of hot
(3–430) 10−5) and cold shower were
considered.
Foodcrop irrigation (lettuce only) Norovirusc Ingestion of 1961–9998 0.002–140 Food crops are consumed
contaminated (6810–9730) (0.1–51) 90, 180, or 270 times per year.
foodcrop
a
Unbracketed values represent the minimum and maximum of each risk estimate. Bracketed values represent the mean or median values (a range of mean/median is given for risk estimates that covered different scenarios).
b
Virus concentration for stormwater were inferred from data for receiving waters affected by stormwater discharge due to the lack of reliable virus quantification method for stormwater.
c
Norovirus is assumed to be in monodispersed form.

Volume 2, November/December 2015

wires.wiley.com/water
 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

that potable uses of rainwater, and more so for storm- chemicals and toxins in the wetland can pose hazards
water, are not advisable unless adequate risk manage- to wildlife.85 There is also anecdotal evidence suggest-
ment strategies (e.g., additional water treatment) are ing that stormwater harvesting wetlands can become a
undertaken. The outcome of QMRA can also be pre- breeding ground for pathogenic microbes that infect
sented in a more practical format, such as ‘tolerable waterfowls and other wetland residents.86 The highly
pathogen concentration in stormwater’ and/or ‘treat- eutrophic stormwater wetland may also encourage
ment technologies/log removal required’ for each spe- the growth of weeds, pests and invasive species, nega-
cific stormwater application (i.e., as used in the tively impact local biodiversity.87 Little data have been
Australian Guideline 23).20 These formats are calculated collected in this aspect of research since stormwater
based on a desired health baseline, such as the WHO’s harvesting wetlands are still at their infancy. The risk
≤10−6 DALYs threshold. It should also be noted that of creating an ecological trap, where wildlife is
these risk estimates contain caveats and data gaps attracted to a potentially dangerous situation, may also
that are intrinsic to most QMRA models. Furthermore, have legal implications. Many countries have laws
water systems fed by stormwater are also prone to which confer liability on individuals or organizations
harbor opportunistic pathogens such as Legionella who act to the detriment of waterbirds, whether
spp. (as found in rainwater tanks). The health risks intended or not (e.g., the USA’s Migratory Bird Treaty
of opportunistic pathogens have not been evaluated. Act of 1918 [amended 1974]).
The QMRA process thus helps to guide additional
research and data collection to further improve the risk
estimates. BENEFITS OF USING STORMWATER
Ultimately, perhaps the most significant contri- Human Health Benefits
bution of QMRA is not the quantified risk value but The most direct human health benefit from stormwater
the linking of relevant sciences together during the risk harvesting for local use is to reduce the pollutant loads
modeling process to gain a better conceptual under- to the receiving water that is used for human recreation.
standing of the risk.83 This also empowers the stake- The impact of stormwater runoff to coastal water qual-
holders with confidence in using stormwater and the ity degradation is well documented.88 In fact, the Cal-
knowledge to troubleshoot issues related to storm- ifornia Department of Public Health advises beach
water harvesting systems. users to avoid contact with ocean and bay waters for
a period of 3 days (72 h) after rainfall ends due to con-
tamination of ocean and bay water by urban runoff.89
Environmental Health Risk Elevated health risks were found among beachgoers
Many aspects of environmental health risks have been recreating at sites polluted by stormwater runoff in
addressed in the previous reviews. The negative impact comparison with those away from the storm drains.90
of over-harvesting of stormwater that results in reduc- QMRA also indicates elevated human health risks from
tion of stream baseflow84 is well documented. For that exposure to stormwater-affected receiving waters.91–93
matter, the Australian guidelines for stormwater har- Stormwater harvesting will reduce the loading of pollu-
vesting have suggested for a tiered approach to investi- tants to the recreation water and effectively avert these
gate the risk of overextracting stormwater.20 Based on hazards.
the guidelines, any stormwater harvesting schemes that Stormwater harvesting infrastructures, i.e.,
extract more than 10% of the annual runoff within rain garden, wetlands, also provide indirect human
their catchment area are required to conduct detailed health benefit by providing new relaxation and recrea-
investigation for reducing any potential environmental tional sites, which have well-documented benefit for
risks. Urban landscape, golf course and home garden both human body and minds.94 The third human
irrigation using stormwater may also negatively impact health benefit of using harvested stormwater to supple-
soil salinity, chemical composition, and plant health ment the traditional water supplies in household is
(i.e., chlorine). These impacts tend to be chronic and the reduction of human stress on water shortage
accumulative and will require additional years of expe- and the improvement of livability of the home
rience to collect additional data.70 environment.95,96
Another potential environmental health risk that
has not been well recognized may be caused by the har-
vesting infrastructure (i.e., stormwater harvesting wet- Environmental Health Benefits
lands). These wetlands were constructed to trap toxic The best-known environmental benefits of stormwater
chemicals, heavy metals, and pharmaceuticals in the harvesting are for stream hydrology and ecology resto-
stormwater. The long-term accumulation of these ration. Since both subjects have been discussed

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 693

 Advanced Review wires.wiley.com/water

extensively in the literature, they are not reiterated in drained. There is more to wetland biodiversity than avi-
this paper. Another benefit that has been largely fauna, Jenkins et al.111 have shown that stormwater
ignored but equally important is the direct benefit from wetlands can support diverse aquatic macroinverte-
construction of new wetlands for stormwater treat- brate and vegetative communities. Herbaceous vegeta-
ment. This is non-trivial, particularly with the back- tive communities harvested from stormwater wetlands
drop of global wetland decimation. Roughly half of also hold potential as sources of energy, fiber, and other
the world’s wetlands have disappeared since commodities.112 The ecological roles and benefits of
1900.97–99 stormwater wetland will always be regionally specific,
Intuitively, it may seem unlikely that stormwater but the existing examples around the world suggest
treatment wetlands could make an impact on a global that these systems at least have the potential to be of
scale. Nonetheless, their potential contribution may be direct worth from a biodiversity conservation
surprisingly significant in certain areas. This is perhaps perspective.
best illustrated by the situation in Melbourne,
Australia. By 2013 there were about 435 constructed
wetland systems used in Melbourne for treating storm-
CONCLUSION
water and many more have been slated for construc-
tion.100,101 The number of individual ponds Stormwater has the potential to be used as a new
comprising a system varies but is generally in the order water resource to meet human demand. The suitability
of 3–10, with the average pond covering about half a of its application depends on the treatment technologies
hectare.102 This clearly represents a large amount of employed, the concentration and type of contaminants
wetland reconstruction for a city that has drained most carried in the stormwater, and the designated use.
of its natural wetlands for development. Microbial pathogens are of the most concern
These stormwater wetlands constructed for when water comes in direct contact with humans
water treatment are not substitutes for natural wet- through showering, toilet flushing, and consumption
lands, although habitat considerations are often incor- of stormwater-irrigated food-crops. The disease risks
porated into the design.103–105 The location of the of using treated stormwater can be assessed through
wetlands in a landscape context—for example, with QMRA. The current risk estimates suggest that the pas-
respect to the movement of migratory species—is sive stormwater treatment provided by LIDs is not ade-
also likely to be a relevant factor in assessing their bio- quate to support applications of stormwater beyond
diversity value. In sum, despite the fact that wildlife lawn irrigation and toilet-flushing. Adoption of the
habitat is often promoted as a benefit of stormwater QMRA approach can assist with decision making and
treatment wetlands,85 there have been very few risk management. This may offer degrees of confidence
attempts to date to determine their broader involve- for stakeholders to adopt the stormwater harvesting
ment in biodiversity conservation even at reasonably practice.
local spatial scales. Pollution mitigation through harvesting storm-
A recent study on waterbird use of Melbourne’s water can benefit human health by reducing human
stormwater treatment wetlands102 found that on a exposure to pollutants during water recreation, and
per-area basis these wetlands tend to support more benefit environmental health through preventing
individual birds as well as species than natural wetlands urban stream syndrome. Stormwater harvesting wet-
in south-eastern Australia.106 In addition, they are lands also offer new habitats for birds and other wild-
located in areas where wetlands are critically needed life. Poor management of stormwater harvesting,
for many species. Numerous Australian duck species, however, can result in overdrawing of natural
for example, breed on ephemeral inland wetlands baseflow and may also endanger wildlife that gather
and use permanent coastal wetlands as non-breeding at regions of treatment wetlands with concentrated
refuges,107–110 but many of these wetlands have been pollutants.

ACKNOWLEDGMENTS
Funding for this work was provided by U.S. National Science Foundation Partnerships for International Research
and Education (OISE-1243543). The discussions and feedback from Professors Stanley Grant and Jean-Daniel
Saphores during the project development are acknowledged.

694 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

REFERENCES
1. Meyer JL, Paul MJ, Taulbee WK. Stream ecosystem 14. Jiang S, Noble R, Chui WP. Human adenoviruses and
function in urbanizing landscapes. J North Am Benthol coliphages in urban runoff-impacted coastal waters of
Soc 2005, 24:602–612. Southern California. Appl Environ Microbiol 2001,
2. Walsh CJ, Roy AH, Feminella JW, Cottingham PD, 67:179–184.
Groffman PM, Morgan RP. The urban stream syn- 15. Robertson BH, Nicholson JKA. New microbiology
drome: current knowledge and the search for a cure. tools for public health and their implications. Annu
J North Am Benthol Soc 2005, 24:706–723. Rev Public Health 2005, 26:281–302.
3. Wenger SJ, Roy AH, Jackson CR, Bernhardt ES, Carter 16. U.S. EPA. Preliminary Data Summary of Urban Storm
TL, Filoso S, Gibson CA, Hession WC, Kaushal SS, Water Best Management Practices. EPA 821-R-99-
Marti E, et al. Twenty-six key research questions in 012, Washington, DC, Office of Water, United States
urban stream ecology: an assessment of the state of Environmental Protection Agency, 1999.
the science. J North Am Benthol Soc 2009,
17. Stenstrom MK, Silverman GS, Bursztynsky TA. Oil and
28:1080–1098.
grease in urban stormwaters. J Environ Eng 1984,
4. Askarizadeh A, Rippy MA, Fletcher TD, Feldman D, 110:58–72.
Peng J, Bowler P, Mehring A, Winfrey B, Vrugt J,
AghaKouchak A, et al. From rain tanks to catchments: 18. Bannister R, Westwood J, McNeill A. Investigation
Use of low-impact development to address hydrologic of microbiological and chemical water quality in
symptoms of the urban stream syndrome. Environ Sci rainwater tanks in Victoria, Report No. 139/97.
Technol 2015. doi:10.1021/acs.est.5b01635. Melbourne,Victorian Department of Natural
Resources and Environment, 1997.
5. Marlow DR, Moglia M, Cook S, Beale DJ. Towards
sustainable urban water management: a critical reas- 19. Ahmed W, Vieritz A, Goonetilleke A, Gardner T.
sessment. Water Res 2013, 47:7150–7161. Health risk from the use of roof-harvested rainwater
in Southeast Queensland, Australia, as potable or non-
6. Burns MJ, Fletcher TD, Walsh CJ, Ladson AR, Hatt BE.
potable water, determined using quantitative microbial
Hydrologic shortcomings of conventional urban storm-
risk assessment. Appl Environ Microbiol 2010,
water management and opportunities for reform.
76:7382–7391.
Landsc Urban Plan 2012, 105:230–240.
20. NRMMC–EPHC–AHMC. Australian Guidelines for
7. Grant SB, Saphores JD, Feldman DL, Hamilton AJ,
Water Recycling: Stormwater Harvesting and Reuse.
Fletcher TD, Cook PLM, Stewardson M, Sanders BF,
Canberra, Natural Resource Management Ministerial
Levin LA, Ambrose RF, et al. Taking the “Waste”
Council, the Environment Protection and Heritage
Out of “Wastewater” for human water security and
Council and the National Health and Medical Research
ecosystem sustainability. Science 2012, 337:681–686.
Council, 2009.
8. Violin CR, Cada P, Sudduth EB, Hassett BA, Penrose
21. Ruecker NJ, Braithwaite SL, Topp E, Edge T, Lapen
DL, Bernhardt ES. Effects of urbanization and urban
DR, Wilkes G, Robertson W, Medeiros D, Sensen
stream restoration on the physical and biological struc-
CW, Neumann NF. Tracking host sources of Crypto-
ture of stream ecosystems. Ecol Appl 2011,
sporidium spp. in raw water for improved health risk
21:1932–1949.
assessment. Appl Environ Microbiol 2007,
9. U.S. EPA. Stormwater homepage. Available at: http:// 73:3945–3957.
water.epa.gov/polwaste/npdes/stormwater/. (Accessed
September 15, 2014). 22. Albrechtsen H. Microbiological investigations of rain-
water and graywater collected for toilet flushing. Water
10. U.S. EPA. Combined sewer overflows (CSO) home. Sci Technol 2002, 46:311–316.
Available at: http://water.epa.gov/polwaste/npdes/cso/.
(Accessed September 15, 2014). 23. Simmons G, Hope V, Lewis G, Whitmore J, Gao W.
Contamination of potable roof-collected rainwater in
11. McCarthy DT, Hathaway JM, Hunt WF, Deletic A.
Auckland, New Zealand. Water Res 2001,
Intra-event variability of Escherichia coli and total sus-
35:1518–1524.
pended solids in urban stormwater runoff. Water Res
2012, 46:6661–6670. 24. Arnone R, Walling J. Waterborne pathogens in urban
watersheds. J Water Health 2007, 5:149–162.
12. Sidhu JPS, Hodgers L, Ahmed W, Chong MN, Toze S.
Prevalence of human pathogens and indicators in 25. Ahmed W, Huygens F, Goonetilleke A, Gardner T.
stormwater runoff in Brisbane, Australia. Water Res Real-time PCR detection of pathogenic microorganisms
2012, 46:6652–6660. in roof-harvested rainwater in Southeast Queensland,
13. Rowny JG, Stewart JR. Characterization of nonpoint Australia. Appl Environ Microbiol 2008,
source microbial contamination in an urbanizing water- 74:5490–5496.
shed serving as a municipal water supply. Water Res 26. Ahmed W, Hodgers L, Sidhu J, Toze S. Fecal indicators
2012, 46:6143–6153. and zoonotic pathogens in household drinking water

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 695

 Advanced Review wires.wiley.com/water

taps fed from rainwater tanks in Southeast Queensland, polyomavirus JC and noroviruses in source waters
Australia. Appl Environ Microbiol 2012, 78:219–226. and drinking water using quantitative PCR. J Virol
27. Vialle C, Sablayrolles C, Lovera M, Huau M-C, Jacob Methods 2009, 158:104–109.
S, Montréjaud-Vignoles M. Water quality monitoring 39. Jiang SC, Chu W. PCR detection of pathogenic viruses
and hydraulic evaluation of a household roof runoff in southern California urban rivers. J Appl Microbiol
harvesting system in France. Water Resour Manag 2004, 97:17–28.
2012, 26:2233–2241.
40. Choi S, Jiang SC. Real-time PCR quantification of
28. Birks R, Colbourne J, Hills S, Hobson R. Microbiolog- human adenoviruses in urban rivers indicates genome
ical water quality in a large in-building, water recycling prevalence but low infectivity. Appl Environ Microbiol
facility. Water Sci Technol 2004, 50:165–172. 2005, 71:7426–7433.
29. Chapman H, Cartwright T, Huston R, O’Toole J. 41. Rajal VB, McSwain BS, Thompson DE, Leutenegger
Water Quality and Health Risks from Urban Rainwa- CM, Wuertz S. Molecular quantitative analysis of
ter Tanks. Cooperative Research Centre for Water human viruses in California stormwater. Water Res
Quality and Treatment: Salisbury, SA; 2008. 2007, 41:4287–4298.
30. Lampard J, Chapman H, Stratton H, Roiko A,
42. Calgua B, Fumian T, Rusinol M, Rodriguez-Manzano
McCarthy D. Pathogenic bacteria in urban stormwater
J, Mbayed VA, Bofill-Mas S, Miagostovich M, Girones
drains from inner-city precincts. In: WSUD 2012:
R. Detection and quantification of classic and emerging
Water Sensitve Urban Design: Building the Water Sen-
viruses by skimmed-milk flocculation and PCR in river
sitve Community: 7th International Conference on
water from two geographical areas. Water Res 2013,
Water Sensitive Urban Design, Melbourne Cricket
47:2797–2810.
Ground, Engineers Australia, 21–23 February, 2012.
31. Holländer R, Bullermann M, Grob C, Hartung H, 43. Philp M, McMahon J, Heyenga S, Marioni O,
König K, Lücke F, Nolde E. Microbiological and hygi- Jenkins G, Maheepala S, Greenway M. Review of
enic aspects of the use of rainwater as process water for stormwater harvesting practices. Urban Water
toilet flushing, garden irrigation and laundering. Security Research Alliance Technical Report No. 9,
Gesundheitswesen 1996, 58:288–293. The Urban Water Security Research Alliance, City East,
QLD, 2008.
32. Savill M, Hudson J, Ball A, Klena J, Scholes P, Whyte R,
McCormick R, Jankovic D. Enumeration of Campylo- 44. U.S. EPA. Low impact development. Available
bacter in New Zealand recreational and drinking at: http://water.epa.gov/polwaste/green/. (Accessed
waters. J Appl Microbiol 2001, 91:38–46. September 15, 2014).
33. Sidhu JPS, Ahmed W, Gernjak W, Aryal R, McCarthy 45. Fletcher TD, Deletic A, Mitchell VG, Hatt BE. Reuse of
D, Palmer A, Kolotelo P, Toze S. Sewage pollution in urban runoff in Australia: a review of recent advances
urban stormwater runoff as evident from the wide- and remaining challenges. J Environ Qual 2008, 37:
spread presence of multiple microbial and chemical S116–S127.
source tracking markers. Sci Total Environ 2013, 46. Hatt BE, Deletic A, Fletcher TD. Integrated treatment
463:488–496. and recycling of stormwater: a review of Australian
34. Haramoto E, Kitajima M, Katayama H, Ohgaki S. practice. J Environ Manage 2006, 79:102–113.
Real-time PCR detection of adenoviruses, polyoma-
47. Abbasi T, Abbasi S. Sources of pollution in rooftop
viruses, and torque teno viruses in river water in Japan.
rainwater harvesting systems and their control. Crit
Water Res 2010, 44:1747–1752.
Rev Environ Sci Technol 2011, 41:2097–2167.
35. Kishida N, Morita H, Haramoto E, Asami M, Akiba M.
One-year weekly survey of noroviruses and enteric ade- 48. Bitting J, Kloss C. Managing Wet Weather with Green
noviruses in the Tone River water in Tokyo metropol- Infrastructure. Municipal Handbook: Green Infrastruc-
itan area, Japan. Water Res 2012, 46:2905–2910. ture Retrofit Policies, EPA 833-F-08-008, Washington,
DC, US Environmental Protection Agency, 2008.
36. Dong Y, Kim J, Lewis GD. Evaluation of methodology
for detection of human adenoviruses in wastewater, 49. Rasid RA, Rahman RA, Rasid RA. Biofilm and multi-
drinking water, stream water and recreational waters. media filtration for rainwater treatment. J Sustain
J Appl Microbiol 2010, 108:800–809. Dev 2009, 2:196–199.
37. Albinana-Gimenez N, Clemente-Casares P, Bofill-Mas 50. Schets F, Italiaander R, Van Den Berg H, de Roda HA.
S, Hundesa A, Ribas F, Girones R. Distribution of Rainwater harvesting: quality assessment and utiliza-
human polyomaviruses, adenoviruses, and hepatitis E tion in The Netherlands. J Water Health 2010,
virus in the environment and in a drinking-water treat- 8:224–235.
ment plant. Environ Sci Technol 2006, 40:7416–7422. 51. IRWD. Natural treatment system guidelines. 2012.
38. Albinana-Gimenez N, Clemente-Casares P, Cagua B, Available at: http://www.irwd.com/images/pdf/water-
Huguet JM, Courtois S, Girones R. Comparison of sewer/NTS%20Design%20Guidelines%20May%
methods for concentrating human adenoviruses, 202012%20final_2.pdf. (Accessed August 16, 2015)

696 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

52. U.S. EPA. Storm Water Technology Fact Sheet: Storm Israel. In: 12th International Conference on Urban
Water Wetlands. EPA 832-F-99-025. Washington, Drainage, Porto Alegre, Brazil, 11–16 September,
DC, Office of Water, United States Environmental Pro- 2011.
tection Agency, 1999. 65. Li YL, Deletic A, Alcazar L, Bratieres K, Fletcher TD,
53. Hench KR, Bissonnette GK, Sexstone AJ, Coleman JG, McCarthy DT. Removal of Clostridium perfringens,
Garbutt K, Skousen JG. Fate of physical, chemical, and Escherichia coli and F-RNA coliphages by stormwater
microbial contaminants in domestic wastewater follow- biofilters. Ecol Eng 2012, 49:137–145.
ing treatment by small constructed wetlands. Water Res 66. Bratieres K, Fletcher T, Deletic A, Alcazar L, Le Cous-
2003, 37:921–927. tumer S, McCarthy D. Removal of nutrients, heavy
54. Karim MR, Manshadi FD, Karpiscak MM, Gerba CP. metals and pathogens by stormwater biofilters. In:
The persistance and removal of enteric pathogens in 11th International Conference on Urban Drainage,
constructed wetlands. Water Res 2004, 38:1831–1837. Edinburgh, Scotland, UK, 2008.
55. Leeming R, Bate N, Hewlett R, Nichols PD. Discrimi- 67. Li LQ, Davis AP. Urban stormwater runoff nitrogen
nating faecal pollution: a case study of stormwater composition and fate in bioretention systems. Environ
entering Port Phillip Bay, Australia. Water Sci Technol Sci Technol 2014, 48:3403–3410.
1998, 38:15–22.
68. Chandrasena GI, Pham T, Payne EG, Deletic A,
56. Bavor H, Davies C, Sakadevan K. Stormwater treat- McCarthy DT. E. coli removal in laboratory scale
ment: do constructed wetlands yield improved pollu- stormwater biofilters: influence of vegetation and sub-
tant management performance over a detention pond merged zone. J Hydrol 2014, 519:814–822.
system? Water Sci Technol 2001, 44:565–570.
69. Chandrasena GI, Deletic A, Ellerton J, McCarthy DT.
57. Hathaway J, Hunt W, Jadlocki S. Indicator bacteria
Evaluating Escherichia coli removal performance in
removal in storm-water best management practices in
stormwater biofilters: a laboratory-scale study. Water
Charlotte, North Carolina. J Environ Eng 2009,
Sci Technol 2012, 66:1132–1138.
135:1275–1285.
70. Department of Environment and Conservation
58. Struck SD, Selvakumar A, Borst M. Prediction of efflu-
NSW. Managing urban stormwater: harvesting and
ent quality from retention ponds and constructed wet-
reuse. 2006. Available at: http://www.environment.
lands for managing bacterial stressors in storm-water
nsw.gov.au/resources/stormwater/managestormwa-
runoff. J Irrigation Drainage Eng 2008, 134:567–578.
terb06137.pdf. (Accessed August 16, 2015)
59. Leisenring M, Clary J, Hobson P. International
71. Moran A. Water supply options for Melbourne.
stormwater best management practices (BMP)
Institute of Public Affairs Occasional Paper, 2008.
database pollutant category statistical summary
Available at: https://www.ipa.org.au/library/publica-
report: solids, bacteria, nutrients, and metals.
tion/1222147673_document_moran_watersupply-
2014. Available at: http://www.bmpdatabase.org/
melbourne.pdf. (Accessed on August 16, 2015)
Docs/2014%20Water%20Quality%20Analysis%
20Addendum/BMP%20Database%20Categorical_ 72. Dandy G, Ganji A, Kandulu J, Hatton MacDonald D,
StatisticalSummaryReport_December2014.pdf. Marchi A, Maier H, Mankad A, Schmidt CE.
(Accessed August 16, 2015) Managed Aquifer Recharge and Stormwater Use
60. Davies C, Bavor H. The fate of stormwater-associated Options: Net Benefits Report. 2013. Available at:
bacteria in constructed wetland and water pollution http://www.goyderinstitute.org/uploads/U.2.1.%
control pond systems. J Appl Microbiol 2000, 20MARSUO%20Net%20Benefits%20Report%20
89:349–360. (T5)_for%20RAC.pdf. (Accessed August 16, 2015)
61. Hathaway J, Hunt W. Indicator bacteria performance 73. Dobbie MF, Brown RR. Risk perceptions and receptiv-
of storm water control measures in Wilmington, ity of Australian urban water practitioners to storm-
North Carolina. J Irrigation Drainage Eng 2011, water harvesting and treatment systems. Water Sci
138:185–197. Technol 2012, 12:888–894.
62. Rusciano G, Obropta C. Bioretention column study: 74. Wu ZF, McKay J, Keremane G. Issues affecting commu-
fecal coliform and total suspended solids reductions. nity attitudes and intended behaviours in stormwater
Trans ASABE 2007, 50:1261–1269. reuse: a case study of Salisbury, South Australia. Water
63. Chandrasena G, Filip S, Zhang K, Osborne C, Deletic 2012, 4:835–847.
A, McCarthy D. Pathogen and indicator microorgan- 75. Hurlimann A, Dolnicar S. When public opposition
ism removal in field scale stormwater biofilters. In: Pro- defeats alternative water projects–the case of Too-
ceedings of the 7th International Conference on Water woomba Australia. Water Res 2010, 44:287–297.
Sensitive Urban Design. Melbourne, Australia, 21–- 76. OCWD-OCSD. About GWRS – World’s largest
23 February 2012. water purification system for potable reuse. 2003–04.
64. Zinger Y, Deletic A, Fletcher T, Breen P, Wong T. A Available at: http://www.gwrsystem.com/. (Accessed
dual-mode biofilter system: case study in Kfar Sava, August 16, 2015)

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 697

 Advanced Review wires.wiley.com/water

77. U.S. EPA. Occurrence and exposure assessment for the of quantitative microbial risk assessment (QMRA).
final long term 2 enhanced surface water treatment rule. Water Res 2010, 44:4692–4703.
EPA 815-R-06-002. Washington, DC, Office of Water, 92. McBride GB, Stott R, Miller W, Bambic D, Wuertz S.
United States Environmental Protection Agency, 2005. Discharge-based QMRA for estimation of public health
78. World Health Organization. Guidelines for Drinking- risks from exposure to stormwater-borne pathogens in
water Quality: Incorporating 1st and 2nd Addenda, recreational waters in the United States. Water Res
vol. 1. 3rd ed. Geneva: WHO Press; 2008. 2013, 47:5282–5297.
79. Fewtrell L, Kay D. Quantitative microbial risk assess- 93. Tseng LY, Jiang SC. Comparison of recreational health
ment with respect to Campylobacter spp. in toilets risks associated with surfing and swimming in dry
flushed with harvested rainwater. Water Environ J weather and post-storm conditions at Southern Califor-
2007, 21:275–280. nia beaches using quantitative microbial risk assess-
80. Lim KY, Jiang SC. Reevaluation of health risk bench- ment (QMRA). Mar Pollut Bull 2012, 64:912–918.
mark for sustainable water practice through risk analy- 94. City of Melbourne. Royal Park stormwater harvesting
sis of rooftop-harvested rainwater. Water Res 2013, project. 2013. Available at: http://www.clearwater.asn.
47:7273–7286. au/user-data/case-studies/plans-designs/Royal-Park-
81. Lim KY, Hamilton AJ, Jiang SC. Assessment of public Case-Study_FINAL.pdf. (Accessed August 16, 2015)
health risk associated with viral contamination in har- 95. Corbett D. Linking Urban Planning and Water Pla-
vested urban stormwater for domestic applications. nning to Achieve Sustainable Development and Live-
Sci Total Environ 2015, 523:95–108. ability Outcomes in the New Growth Areas of
82. Hamilton AJ, Stagnitti F, Premier R, Boland AM, Hale Melbourne, Australia. Int J Soc Behav Educ Econ
G. Quantitative microbial risk assessment models for Manag Eng 2012, 6:220–228.
consumption of raw vegetables irrigated with reclaimed 96. Melbourne Water. Melbourne water’s submission:
water. Appl Environ Microbiol 2006, 72:3284–3290. Melbourne’s water future. Undated. Available at:
83. Bichai F, Smeets PWMH. Using QMRA-based regula- http://melbournewater.com.au/aboutus/news/Docu-
tion as a water quality management tool in the water ments/Melbourne_s_Water_Future_submission.pdf.
security challenge: experience from the Netherlands (Accessed August 16, 2015)
and Australia. Water Res 2013, 47:7315–7326. 97. OECD. Guidelines for the Aid Agencies for improved
84. Walsh CJ, Fletcher TD, Burns MJ. Urban stormwater Conservation and Sustainable Use of Tropical and Sub-
runoff: a new class of environmental flow problem. tropical Wetland. Paris: Wiley; 1996.
PLoS One 2012, 7:e45814. 98. Pineau O. Conservation of wintering and migratory
85. Sparling DW, Eisemann JD, Kuenzel W. Contaminant habitats. In: Hafner H, ed. Heron Conservation.
exposure and effects in red-winged blackbirds inhabit- London: Academic Press; 2000, 237–250.
ing stormwater retention ponds. Environ Manage
99. Tiner RW. Assessing cumulative loss of wetland func-
2004, 33:719–729.
tions in the Nanticoke River watershed using enhanced
86. Murray CG, Hamilton AJ. REVIEW: perspectives on National Wetlands Inventory data. Wetlands 2005,
wastewater treatment wetlands and waterbird conser- 25:405–419.
vation. J Appl Ecol 2010, 47:976–985.
100. Melbourne Water. Melbourne water: annual report.
87. Moore TL, Hunt WF. Urban Waterways: Stormwater 2009. Available at: http://www.melbournewater.com.
Wetlands and Ecosystem Services. AGW-588-22, au/aboutus/reportsandpublications/Annual-Report/
North Carolina Cooperative Extension, 2011. Documents/2008-09%20Sustainability%20Report.
88. Burton GAPRE. Stormwater Effects Handbook : A pdf. (Accessed August 16, 2015)
Toolbox for Watershed Managers, Scientists, and Engi- 101. Melbourne Water. Melbourne water: annual report.
neers. Boca Raton, FL: Lewis; 2002. 2013. Available at: http://www.melbournewater.com.
89. County of Los Angeles. Environmental health- beach au/aboutus/reportsandpublications/Annual-Report/
advisories. 2015. Available at: http://publichealth. Documents/2013_Annual_Report.pdf. (Accessed
lacounty.gov/phcommon/public/eh/water_quality/ August 16, 2015)
beach_grades.cfm. (Accessed August 16, 2015) 102. Murray CG, Kasel S, Loyn RH, Hepworth G, Hamilton
90. Haile RW, Witte JS, Gold M, Cressey R, McGee C, AJ. Waterbird use of artificial wetlands in an Australian
Millikan RC, Glasser A, Harawa N, Ervin C, Harmon urban landscape. Hydrobiologia 2013, 716:131–146.
P. The health effects of swimming in ocean water con- 103. Melbourne Water. Constructed wetlands
taminated by storm drain runoff. Epidemiology 1999, guidelines. nd. Available at: http://www.melbourne-
10:355–363. water.com.au/planning-and-building/standards-and-
91. Ashbolt NJ, Schoen ME, Soller JA, Roser DJ. Predicting specifications/design-general/pages/constructed-wet-
pathogen risks to aid beach management: the real value lands-guidelines.aspx. (Accessed August 16, 2015)

698 © 2015 Wiley Periodicals, Inc. Volume 2, November/December 2015

 WIREs Water Human and environmental health risks and benefits associated with use of urban stormwater

104. U.S. EPA. Constructed wetlands for wastewater treat- 108. Hamilton A, Taylor I. Seasonal patterns in abundance
ment and wildlife habitat: 17 case studies. 1993. Avail- of waterfowl (Anatidae) at a waste stabilization pond in
able at: http://water.epa.gov/type/wetlands/upload/ Victoria. Corella 2004, 28:61–67.
constructed-wetlands.pdf. (Accessed August 16, 2015)
109. Hamilton A, Taylor I, Rogers P. Seasonal and diurnal
105. Wong T, Breen P, Somes T, Lloyd S. Managing Urban patterns in abundance of waterbirds at a waste stabili-
Stormwater Using Constructed Wetlands. 2nd ed. zation pond, Victoria. Corella 2004, 28:43–54.
Clayton: Cooperative Research Centre for Catchment
Hydrology and Cooperative Research Centre for Fresh- 110. Pert P. Changes in abundance of grey teal (Anas gracilis)
water Ecology; 1999. in two saline habitats in southern Victoria. Master’s
Thesis, Charles Sturt University, Thurgoona, New
106. Murray CG, Loyn RH, Kasel S, Hepworth G, Stama- South Wales, 1997.
tion K, Hamilton AJ. What can a database compiled
over 22 years tell us about the use of different types 111. Jenkins G, Greenway M. Restoration of a constructed
of wetlands by waterfowl in south-eastern Australian stormwater wetland to improve its ecological and
summers? Emu 2012, 112:209–217. hydrological performance. Water Sci Technol 2007,
56:109–116.
107. Frith H. Breeding and movements of wild ducks in
inland New South Wales. CSIRO Wildlife Res 1957, 112. Mitsch WJ, Gosselink JG. Wetlands. Hoboken, NJ:
2:19–31. John Wiley and Sons; 2007.

Volume 2, November/December 2015 © 2015 Wiley Periodicals, Inc. 699

View publication stats