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

net/publication/326750149

Identification of sewage markers to indicate sources of contamination: Low

cost options for misconnected non-stormwater source tracking in stormwater
systems

Article  in  Science of The Total Environment · August 2018

DOI: 10.1016/j.scitotenv.2018.07.448

CITATION READS

1 91

9 authors, including:

Hailong Yin Jingshui Huang

Tongji University Tongji University
39 PUBLICATIONS   236 CITATIONS    5 PUBLICATIONS   91 CITATIONS   

SEE PROFILE SEE PROFILE

Rongchang Wang
Tongji University
41 PUBLICATIONS   545 CITATIONS   

SEE PROFILE

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

the National Natural Science Foundation of China (51878466) View project

the National Key Research and Development Program in China (2016YFC0400805) View project

All content following this page was uploaded by Rongchang Wang on 21 October 2018.

The user has requested enhancement of the downloaded file.

 Science of the Total Environment 648 (2019) 125–134

Contents lists available at ScienceDirect

Science of the Total Environment

journal homepage: www.elsevier.com/locate/scitotenv

Identification of sewage markers to indicate sources of contamination:

Low cost options for misconnected non-stormwater source tracking in
stormwater systems
Hailong Yin a,b, Ming Xie c, Lunyuan Zhang d, Jingshui Huang b, Zuxin Xu b,e,⁎, Huaizheng Li a,b, Renjing Jiang a,b,
Rongchang Wang b,e, Xu Zeng b,e
a
Key Laboratory of Yangtze River Water Environment, Ministry of Education, Tongji University, Shanghai 200092, China
b
Shanghai Institute of Pollution Control and Ecological Security, Shanghai, China, 200092
c
Tongji Architectural Design (Group) Co., Ltd., Shanghai 200092, China
d
Six Design Institute, Shanghai Municipal Engineering Design Institute (Group) Co., Ltd, Hefei 230061, China
e
State Key Laboratory of Pollution Control and Resource Reuse, College of Environmental Science and Engineering, Tongji University, Shanghai 200092, China

H I G H L I G H T S G R A P H I C A L A B S T R A C T

• Concentrations of fifty-two markers

through industrial WWTPs were inves-
tigated.
• Eight markers were found stable to indi-
cate industrial and domestic contamina-
tion.
• Benchmark data to distinguish between
industrial and domestic sewage was
provided.
• Tracking of illicit domestic discharges
was manifested using acesulfame and
theanine.

a r t i c l e i n f o a b s t r a c t

Article history: There has been increasing research focusing on the detection and occurrence of wastewater contamination in
Received 3 June 2018 urban water systems. To find suitable markers to indicate industrial and domestic sewage flows inappropriately
Received in revised form 17 July 2018 entering storm drains, this study investigated the occurrence and fate of 52 chemical markers through wastewa-
Accepted 30 July 2018
ter treatment facilities of manufacturers of agricultural and sideline products, beverage products, and pharma-
Available online 1 August 2018
ceutical products, which are also consumed in our daily life. Of the 52 candidate markers, sodium, chloride,
Keywords:
potassium, isomalto-oligosaccharide, acesulfame, theanine, glycerol, and clarithromycin were found to be con-
Chemical markers servative markers, with an average change in concentrations through the wastewater treatment processes of
Wastewater contamination b30%. These markers are useful in identifying industrial and domestic sewage flow contamination in urban
Misconnections sewers. Specially, sodium, chloride, potassium, isomalto-oligosaccharide, acesulfame, and clarithromycin exhib-
Source tracking ited higher concentrations in blackwater than in greywater, with detected average concentrations of 43.8 mg/L,
Stormwater sewers 189 mg/L, 37.3 mg/L, 123 μg/L, 37.2 μg/L, and 0.99 μg/L in blackwater, respectively. In contrast, theanine and glyc-
erol were observed with higher concentrations in greywater than in blackwater (average 10.1 μg/L and 19.5 μg/L
in greywater, respectively). The benchmark concentrations to discriminate between industrial and domestic
sewage were also presented. A study in a storm drainage system of downstream Taihu catchment, China

⁎ Corresponding author.
E-mail address: xzx@tongji.edu.cn (Z. Xu).

https://doi.org/10.1016/j.scitotenv.2018.07.448
0048-9697/© 2018 Elsevier B.V. All rights reserved.
126 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134
demonstrated the usefulness of the markers as low-cost options to trace and quantify misconnected wastewater
entries into storm drains, while denoting priority areas for misconnected entries correction.
1. Introduction
For watershed managers, the location of potential sources of con-
tamination is an important step in addressing urban water quality con-
cerns. In the urban river catchment, where separate sewer and storm
networks have been developed, inappropriate or illicit non-
stormwater discharge from the storm drainage system may account
for significant pollution of the receiving waters, including wastewater,
which is be an important source of dry-weather flow (Pitt et al., 1993;
Field et al., 1994; Brown et al., 2004; Xu et al., 2014, 2016; Revitt and
Ellis, 2016). Additionally, on wet-weather days, contaminated urban
stormwater runoff consists of traditional precipitation that drains from
city surfaces as well as the waters from illicit and/or inappropriate
wastewater flows into the storm drainage system, thus aggravating
the overflow pollution (Zhang et al., 2015, 2016; Yin et al., 2017).
Efficient implementation of substantial efforts to eliminate the illicit
connections would require a thorough assessment of sewer conditions;
furthermore, prioritization of high-risk areas would allow the utilities to
focus their efforts and proactively treat problems. Most of the city-wide
assessments so far have relied upon physical methods, such as CCTV
(closed circuit television), temperature based thin-film cabling exami-
nation, and flow measurements. However, these methods are often
too detail-focused and labor-intensive to achieve system-wide tracking
of contaminated sources in sewer dry-weather flow.
Optionally, a few marker species could be used to identify and quan-
tify sewage source discharge, which is usually less costly to perform
than by thorough physical inspections. Recently various chemical
markers have been utilized to detect wastewater contamination in sur-
face waters and groundwater, and these markers can be generally clas-
sified as pharmaceutical and personal care products (PPCPs), artificial
sweeteners, fluorescent whitening agents (FWAs), sterols/stanols, iso-
topic compositions, and other alternative potential markers (Lim et al.,
2017). Thirty-two chemical markers, including 12 PPCPs, 3 artificial
sweeteners, 4 FWAs, 8 sterols/stanols, 2 isotopic compositions, and 3
other alternative potential markers, have been investigated. For exam-
ple, in recent studies, researchers demonstrated that PPCPs and artificial
sweeteners could serve as promising markers of sewage contamination
(Nakada et al., 2008; Buerge et al., 2009; Kasprzyk-Hordern et al., 2009;
Gasser et al., 2011; Daneshvar et al., 2012; Fono and Sedlak, 2005;
Kuroda et al., 2012; Tran et al., 2014a, Tran et al., 2014b; Sun et al.,
2016; Yang et al., 2017).
Although these reported markers have shown to be useful in indicat-
ing anthropogenic domestic sewage contamination in surface water
bodies (i.e., rivers, canals, lakes, and coastal waters) and groundwater,
it is still challenging to further distinguish raw sewage sources, consid-
ering pollution may result from various domestic and industrial activi-
ties. For example, flow discharge of higher concentrations may be
observed as a result of illicit industrial wastewater entry into the
storm pipes of an urbanized area, where food, beverage, and pharma-
ceutical manufacturing industries are present. Considering these prod-
ucts are consumed in daily life, these potential markers associated
with a variety of industries may also be effective indicators of domestic
sewage discharge. From this perspective, identifying new markers will
facilitate the identification and quantification of industrial wastewater
contamination in the dry-weather flow, and also promote our under-
standings of the “fingerprint” of various domestic activities.
Hence, the objective of our study was to find the chemical markers to
indicate sewage contamination on the basis of three typical industrial
activities in an urbanized area (i.e., food, beverage, and pharmaceutical
processing or manufacturing). To meet this objective, a large number of
© 2018 Elsevier B.V. All rights reserved.
candidate marker species associated with these industries were consid-
ered according to production process survey and through laboratory
testing of onsite collected water samples. We aimed to identify several
suitable indicators that would be useful in tracking and quantifying po-
tential sewage sources. Finally, the findings were applied in an urban
drainage system of downstream Taihu catchment area, China to demon-
strate the use of these new markers in detecting inappropriate sewage
connections to storm drains.
2. Materials and methods
2.1. Sampling sites
To identify new markers indicating sewage contamination in dry-
weather flow, a total of 14 sampling sites were chosen. These sites are
located in the wastewater treatment facilities of industrial enterprises
in Shanghai, China. The facilities cover 3 industrial trades and are asso-
ciated with 12 sub-industrial trade types, which are related to agricul-
tural food processing and manufacturing, alcohol, beverages, and
refined tea manufacturing, and pharmaceutical manufacturing. A de-
scription of the sampling sites is given in the Supporting Information
(Section S1). Additionally, one sampling site in a residential community
was also chosen to establish the data library of screened chemical
markers in sanitary sewage. Specifically, the sampling site is a residen-
tial community close to our university (i.e., No. 8 Anshan Village,
which was built in the late 1970s).
2.2. Candidate markers and samples collection
It is necessary to select candidate marker species before conducting
sample collection and analyses. For each chosen industrial trade type,
candidate markers were determined based on the raw materials used
and their production processes, as explained in the Supporting Informa-
tion (Section S2). In total, 52 candidate markers were selected, including
14 markers for agricultural and sideline food processing and
manufacturing, 33 markers for alcohol/beverage products manufactur-
ing (4 markers overlapped with those for agricultural products
manufacturing), and 13 markers for pharmaceutical products
manufacturing (4 markers overlapped with those for agricultural prod-
ucts manufacturing, as well).
Ideal markers of contaminated dry-weather flow should also exhibit
conservative behavior (i.e., no significant concentration change due to
biological or chemical processes), besides their obvious differences in
concentrations between polluted and relatively clear sources. To deter-
mine if the candidate markers exhibit conservative or reluctant behav-
ior, sewage samples were measured at the influent and effluent of
wastewater treatment plants (WWTP) for each enterprise. Specifically,
two sampling activities were performed. The first sampling activity
was conducted in May, 2015, and the second was conducted in Nov.,
2015. For each sampling activity, the sewage samples were collected
during the period of 9 a.m. to 6 p.m. with a sampling interval of 1 h,
resulting in total 20 samples for each enterprise. The sewage samples
(1.5 L each) were kept at approximately 4 °C and transported back to
the laboratory for immediate processing.
As mentioned above, water samples were also collected in one resi-
dential community. Sanitary sewage from this community flows
through two parallel indoor and outdoor sewer pipes: one is for black-
water from the toilet and another is for greywater from bathing, laun-
dry, dish washing, and cooking. The sampling locations were located
at the sewer network outlets of this residential community. Sample
 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134 127

collection continued for two days, during which greywater and black- Table 1
water were sampled separately. Sewage was sampled every 3 h, Analytical methods for the candidate markers.

resulting in 8 samples per day and a total of 16 samples for greywater Parameter Analytical method
and blackwater, respectively. The sewage samples were maintained at Ammonia Colorimetry
approximately 4 °C using ice packs and transported to the laboratory Oils Infrared spectrophotometry
daily for immediate processing and analysis. Sodium, potassium ICP-AES (inductively-coupled
plasma atomic emission
spectroscopy)
2.3. Chemical analyses
Chloride Ion chromatography
Protein BCA (Bicinchoninic acid) protein
The collected samples were analyzed for the candidate markers assay kit
listed above. A strict regime of quality control was employed before Sugar Anthracenone colorimetry
the onset of sampling and analysis program. Quantification of the target Starch Phenol sulfuric acid colorimetry
Butyric acid, oxalic acid, acetic acid, HPLC (high performance liquid
compounds was performed by the internal standard and the external butanedioic acid, pyruvic acid, fumaric chromatography)
standard method. For every set of 3 samples, a procedure blank and a acid, lactic acid, propanoic acid, malic acid,
standard solution were run to check for interference and cross tartaric acid, vanillic acid, ascorbic acid,
contamination. sorbic acid, hexylic acid
Caffeine, benzoic acid, theanine HPLC
For the candidate markers including organic acids, maltose, fructose,
Fructose, glucose, maltose HPLC
glucose, isomalto-oligosaccharide (ISO), caffeine, benzoic acid, Lemon yellow, sunset yellow HPLC
theanine, lemon yellow, sunset yellow, methanal, aldehyde compounds, Methanal GC–MS (gas chromatography-mass
ethyl acetate, glycerol, alcohols, artificial sweeteners, oxazocilline, ibu- spectrometer)
profen, naproxen, clarithromycin, sulfamethoxazole, trimethoprim, Aldehyde acetal, ethanal, furfural, GC–MS
benzaldehyde
paracetamol, salicylic acid, epigallocatechin gallate (EPI) and citric
Ethyl acetate, ethanol, glycerol, n-propyl GC–MS
acid, their concentrations were quantitatively determined by the inter- alcohol, isobutyl alcohol, n-butyl alcohol
nal standards using peak height. The recoveries were calculated from in- Acesulfame, aspartame LC-MS (Liquid
dustrial WWTP influent, WWTP effluent and domestic sewage samples chromatography-mass
spectrometer)
spiked with the standard solutions at appropriate concentrations con-
Oxazocilline, ibuprofen, naproxen, LC-MS
sidering the background concentrations of the target compounds in clarithromycin, sulfamethoxazole,
unspiked samples. Limit of quantification (LOQ) for each target com- trimethoprim, paracetamol, salicylic acid
pound was determined as the minimum detectable concentration Epigallocatechin gallate LC-MS
which was ten times of the signal to noise ratio (SNR) near the target Isomalto-oligosaccharide LC-MS

peak in different matrix. An instrumental blank, a procedural blank,

and a standard mixture (domestic sewage: 10 ng/L; WWTP influent
and effluent: 100 μg/L) were applied for each batch to check potential
background contamination and instrument performance. All the target
compounds were not detected in all blank samples.
For the other indicators including ammonia, oils, sodium, potassium,
chloride, protein, sugar and starch, their concentrations were quantita-
tively determined by external standard methods. The regression coeffi-
cients for the calibration standards, injected at appropriate
concentrations for ammonia, oils, sodium, potassium, chloride, protein,
sugar and starch, were N0.999.
The analytical method for each parameter is summarized in Table 1.
Detailed information for the analyses of these markers is provided in the
Supporting Information S3.
3. Results and discussions
3.1. Screening of sewage markers as indicators of source flow
contamination
Measured concentrations of the candidate markers at the influents
and effluents of WWTPs associated with agricultural and sideline prod-
ucts processing or manufacturing, alcohol/beverage manufacturing, and
pharmaceutical manufacturing are summarized in Fig. 1. More details
concerning the measured data are listed in the Supporting Information
S4. Thereby, for each marker, the average removal efficiency via
WWTP processes is presented in Fig. 2.
It is necessary to set criteria for the screening of ideal markers based
on the calculated marker removal ratio via WWTP process. A study on
the occurrence of pharmaceuticals in an urban sewer reported the aver-
age removal ranged from −10% to 10%, could be considered negligible
when compared with the overall uncertainty associated with the con-
centration values (Jelic et al., 2015). The compounds arrived unchanged
(−10%–10% removal) or slightly changed (b30% removal), as delivered
via sewer pipe, may be used as stable markers in the sewage epidemiol-
ogy approach (O'Brien et al., 2014; Jelic et al., 2015). According to
Kasprzyk-Hordern et al. (2009) and Yang et al. (2017), the compounds
with low removal rate (b50%) could be selected as conservative indica-
tors of wastewater.
Referring to these previous studies, we selected chemical markers
with average removal efficiencies of b30% as the screened ideal or stable
markers to indicate sewage contamination in dry-weather flow. There-
fore, the screened chemical markers associated with the investigated in-
dustrial trades are sodium, chloride, potassium, isomalto-
oligosaccharide (ISO), acesulfame, tartaric acid, theanine, pyruvic acid,
hexylic acid, glycerol, n-propyl alcohol, isobutyl alcohol, and
clarithromycin (see Fig. 2).
Sodium, chloride, and potassium were detected and screened as
ideal markers in both agricultural/sideline products processing or
manufacturing, and pharmaceutical manufacturing. Barrett et al.
(1999) listed the major cations and anions, including potassium, so-
dium and chloride, as candidate markers to identify urban groundwater
recharge sources. Among the major ions, only chloride was thought to
be reasonably conservative. However, based on our study, we found
that in addition to chloride, sodium and potassium also exhibit conser-
vative behavior through WWTP processes. In this case, potassium was
detected at extreme high concentration levels in WWTP effluents of
pharmaceutical manufacturing, with the average values ranging from
2525–3813 mg/L. Sodium and chloride also exhibited significantly
high concentrations in effluent flows of investigated pharmaceutical
manufacturing, with concentrations of 5040 mg/L and 5897 mg/L, re-
spectively, at the highest level. Additionally, sodium, chloride and po-
tassium were found in concentrations of approximate 500–600 mg/L
for some sub-industrial types related to agricultural and sideline prod-
ucts manufacturing. This indicates that inorganic ions of sodium, chlo-
ride, and potassium are useful as wastewater-specific parameters to
detect illicit industrial flow entries into separate storm drains and
aquatic environments.
Artificial sweeteners are now used worldwide as tabletop sweet-
eners and food additives to sweeten diet beverages. The artificial
128 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134

(a) agricultural and sideline industries

(b) drinks and beverage industries

(c) pharmaceutical industries

Fig. 1. Concentrations of candidate markers detected in influents and effluents of WWTPs. The error bars represent the analytical measurement uncertainty associated with reported
concentrations (confidence interval: 70%).

sweetener, acesulfame, was viewed as an ideal marker for the evalua-
tion of the impact of wastewater contamination on urban water systems
in the last decade, due to its recalcitrance to transformation (Buerge
et al., 2009; Tran et al., 2014a). In this study, concentrations of
acesulfame at WWTP effluents were 72%–84% of that at the influents,
exhibiting conservative behavior. The highest acesulfame concentration
was detected in the wastewater of candy manufacturing, with average
data of 4.57 mg/L and 3.27 mg/L at WWTP influents and effluents re-
spectively. In the treatment plant influent for non carbonated and tea
beverage manufacturing, the measured acesulfame concentration was
approximate 2.3 times that of the acesulfame concentration from car-
bonated beverage manufacturing. Therefore, it can be inferred that
acesulfame consumed via low-calorie beverages and other products, is
excreted through urine or feces and ends up in the environment.
Clarithromycin is a useful conservative marker based on its detec-
tion in pharmaceutical manufacturing wastewater. It was detected in
concentrations ranging from 4.82–8.14 mg/L and 4.21–6.95 mg/L at
WWTP influents and effluents respectively, corresponding to a removal

Fig. 2. Average removal of candidate markers via wastewater treatment processes.

 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134 129

ratio of 7%–15%. There is debate as to whether clarithromycin is readily
biodegradable, due to its reported wide removal range of 0–99% (Tran
et al., 2018). Despite this, a low removal rate for clarithromycin was
found in some WWTPs. For example, elimination rates of clarithromycin
in two conventional activated sludge systems of Zurich, Switzerland
ranged between 5.6%~14% on average, over different seasons (Gobel
et al., 2007). Clarithromycin concentrations remained almost un-
changed through the treatment processes of six WWTPs in Canada,
i.e., 120 μg/L at the influent versus 110 μg/L at the effluent (Guerra
et al., 2014). A study on in-sewer transformation of pharmaceuticals
in a pressurized sewer showed an average removal efficiency of 26%
for antibiotic clarithromycin, which also falls within the range of conser-
vative markers (Jelic et al., 2015). From this perspective, clarithromycin
is a relatively stable marker for identifying contamination of sewer net-
works or storm drains.
For the rest of screened markers, including theanine, tartaric acid,
isomalto-oligosaccharide, pyruvic acid, hexylic acid, glycerol, n-propyl
alcohol, and isobutyl alcohol, no published studies are currently avail-
able regarding their concentrations in wastewater. This is, to our knowl-
edge, the first study to investigate the occurrence and concentrations of
these markers in wastewater.
Pitt et al. (1993) listed the chemicals indicative of manufacturing in-
dustrial activities, which were used to provide guidance for investigat-
ing illicit non-stormwater entries into storm drains. Specifically,
mineral acids and organic acids were thought to be significant chemicals
in industrial wastewater associated with brewing, distilleries and fer-
mentation plants; starch and sugars were thought to be significant
chemicals in industrial wastewater of dairies, food industries, sugar re-
fining, and preserves processing. However, no detailed information
concerning their concentrations in wastewater is available. In our
study, we established a data library of chemical markers associated
with industrial activities and evaluated the suitability of these chemicals
as indicators of sewage contamination. For example, we found that
starch and sugars may not be ideal for indicating wastewater flow of
food or drink manufacturing, as the removal efficiencies for these com-
pounds was N75%, as indicated in Fig. 2. We also screened the organic
acids and alcohols that could be used as conservative markers for de-
tecting industrial wastewater from brewing, distilleries and fermenta-
tion plants among a series of candidate parameters, such as tartaric
acid, pyruvic acid, hexylic acid, n-propyl alcohol, and isobutyl alcohol,
based on this study.
3.2. Data library of screened markers in contaminated source flows
3.2.1. Data library of screened markers in domestic wastewater
For the screened markers, Table 2 presents their concentrations
along with the occurrence rates in domestic greywater and blackwater.
Inorganic substances, including sodium, chloride, and potassium,
were measured in higher concentrations in blackwater than in
greywater. In greywater from our study site, the concentrations of so-
dium, chloride and potassium were 65.3 mg/L, 52 mg/L, and
15.5 mg/L, on average respectively; in black water, their average con-
centrations were 80.6 mg/L, 153 mg/L, and 40.4 mg/L, respectively.
The increase in sodium, chloride and potassium concentration following
domestic water usage further suggests its potential as a tracer
parameter.
Organic substances, including acesulfame, isomalto-oligosaccharide,
and clarithromycin, were also measured in higher concentrations in
blackwater than in the greywater. Artificial low-calorie sweeteners,
such as acesulfame, are regularly consumed in large quantities. After in-
gestion, the sweetener passes through the human metabolism large un-
affected, are quantitatively excreted via feces. A high concentration of
acesulfame was frequently found at the sampling sites located in resi-
dential/commercial areas with high population density. Considering
greywater is estimated to account for 70%–75% of domestic wastewater
(Blanky et al., 2015; Yi et al., 2011), the detected average acesulfame
concentration in blending domestic wastewater was approximately
10.4–12.2 μg/L, based on measured data for blackwater and greywater
in Table 2.
Isomalto-oligosaccharide (IMO) is a mixture of short-chain carbohy-
drates that has a digestion-resistant property. IMO is found naturally in
some foods and is also manufactured commercially. The raw material
used for manufacturing IMO is starch, which is enzymatically converted
into a mixture of IMOs. The measured average IMO concentration in
blackwater was as high as 123 μg/L, with a range of 40–200 μg/L. By con-
trast, no IMO was detected in greywater in our sampling site.
Clarithromycin is an antibiotic used to treat various bacterial infec-
tions, and it can be classified as a PPCP, when discussing the use of
chemical markers. Most PPCP contamination is generated from the use
of medicines and cosmetic products. While a series of PPCPs
(e.g., acetaminophen, diclofenac, caffeine, bisphenol A, propyl paraben,
triclocarbon, metoprolol, propranolol, ibuprofen, atenolol, carbamaze-
pine, and oxypurinol) has been found ubiquitously in environmental
waters and widely reported in wastewater, surface water and ground-
water, relatively less information concerning the clarithromycin con-
centration in sanitary wastewater was found. In our study, the average
clarithromycin concentration in blackwater was 0.99 μg/L, with the
measured range extending from 0.68 to 1.67 μg/L. This corresponds to
blending residential wastewater concentration of 0.30 μg/L approxi-
mately considering no clarithromycin was detected in greywater.
Compared to the aforementioned six markers (sodium, chloride, po-
tassium, isomalto-oligosaccharide, acesulfame, and clarithromycin),
theanine and glycerol exhibited a different trend: higher concentrations
were observed in greywater rather than blackwater. The averaged

Table 2
Summary of measured marker concentrations for sanitary sewage.

Markers Unit n Greywater Blackwater

Min.–Max. Av. ± Std. Occ. Min.–Max. Av. ± Std. Occ.

Sodium mg/L 16 22.2–68.3 38.7 ± 11.6 100% 16.7–63.7 43.8 ± 10.7 100%
Chloride mg/L 16 57.8–82.5 68.2 ± 7.46 100% 159–232 189 ± 21.8 100%
Potassium mg/L 16 12.9–36.7 23.6 ± 6.61 100% 25.0–56.0 37.7 ± 7.5 100%
Isomalto-oligosaccharide μg/L 16 bLOQ / 0 40.0–200 123 ± 67.7 100%
Acesulfame μg/L 16 1.07–1.78 1.47 ± 0.22 100% 27.9–51.2 37.2 ± 6.1 100%
Tartaric acid μg/L 16 0–0.50 0.07 ± 0.16 19% 0–0.30 0.03 ± 0.09 13%
Theanine μg/L 16 7.50–12.4 10.1 ± 1.29 100% 0.04–0.23 0.10 ± 0.05 100%
Pyruvic acid μg/L 16 bLOQ / 0 0–0.02 0 ± 0.01 19%
Hexylic acid μg/L 16 0.02–0.27 0.10 ± 0.06 100% 0–0.05 0.03 ± 0.01 69%
Glycerol μg/L 16 8.17–43.2 19.5 ± 9.86 100% bLOQ / 0
n-Propyl alcohol μg/L 16 0.02–0.17 0.09 ± 0.05 100% bLOQ / 0
Isobutyl alcohol μg/L 16 0–0.03 0.02 ± 0.01 38% bLOQ / 0
Clarithromycin μg/L 16 bLOQ / 0 0.68–1.67 0.99 ± 0.30 100%

n = number of samples; Occ.: occurrence rate of this substance in grey or blackwater, in percentage (%); Min = minimum concentration; Max = maximum concentration; Av = average
concentration; Std = standard concentration of measured samples; and LOQ = limit of quantification.
130 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134

measured theanine concentration in greywater was 10.1 μg/L (10.1 ±
1.29 μg/L, range 7.50–12.4 μg/L), while that in blackwater was only
0.10 μg/L (0.10 ± 0.05 μg/L, range 0.04–0.23 μg/L). For glycerol, the av-
eraged measured concentration in greywater was 19.5 μg/L (19.5 ± 9.86
μg/L, range 8.17–43.2 μg/L); in contrast, no glycerol was detected in
blackwater samples. To the best of our knowledge, this is the first report
on the theanine and glycerol concentrations in sanitary wastewater.
Theanine is a non-protein amino acid present in tea and also used as
an additive in food industry. Unlike the artificial sweetener acesulfame,
theanine is not refractory within the human body; it can be degraded
into glutamic acid and ethylamine by glutaminase in the kidney and ex-
creted through human urine (Unno et al., 1999). This can explain why
theanine concentration in blackwater is low. Considering theanine is
an essential amino acid mainly found in tea and accounts for over 50%
of the total free amino acid in tea (Zheng et al., 2017), high theanine
concentration detected in greywater may be pertinent to the daily tea
consumption. Use of theanine as a food additive also contributes to
the increased theanine concentration in greywater. This finding is inter-
esting, because theanine can be employed as a novel chemical marker to
discriminate greywater from blackwater. For example, a decrease in the
concentration of foul markers (e.g., nitrogen species and artificial sweet-
ener) down the sewer pipe may indicate the entry of relatively clear wa-
ters. In this case, the use of theanine could be helpful in further clarifying
the relatively clean water types, e.g., greywater or infiltrated
groundwater.
Glycerol, also called glycerin, is a simple polyol compound. It is a col-
orless, odorless, viscous liquid that is sweet tasting and non-toxic. It is
widely used in pharmaceutical and personal care product preparations,
mainly as a means of providing lubrication, and as a humectant. It is
found in toothpaste, mouthwashes, skin care products, shaving cream,
hair care products, shampoos, soaps, and water-based personal lubri-
cants. This can explain why detected glycerol concentration is high in
greywater, which is wastewater generated from bathing, showering,
hand washing, laundry, and the kitchen sink. Glycerol is almost not ob-
served in blackwater. Accordingly, like theanine, glycerol is a useful
chemical marker to distinguish greywater from blackwater.
Organic acids and alcohols, including tartaric acid, pyruvic acid,
hexylic acid, n-propyl alcohol, and isobutyl alcohol, were detected at
very low concentrations in sanitary sewage, with average concentra-
tions b0.10 μg/L. Tartaric acid was observed in just 19% of greywater
samples and 13% of blackwater samples; pyruvic acid and n-propyl alco-
hol isobutyl alcohol were not observed in blackwater or greywater sam-
ples. This shows that while these five acids and alcohols are ubiquitous
in industrial wastewater of brewing, distilleries and fermentation
plants, they are not suitable for indicating domestic wastewater con-
tamination, due to their negligible concentrations in collected water
samples from the residential community.
3.2.2. Benchmark concentrations to distinguish between industrial and do-
mestic discharges
Based on above discussions, eight conservative markers can be used
to indicate both industrial and sanitary sewage. For the catchment that
has residential sites as well as industrial generating sites, Table 3 further
summarized typical benchmark concentrations that can distinguish be-
tween unique industrial or sanitary wastewaters, which were deter-
mined based on the measured maximum value in blackwater or
greywater.
Among the eight markers, potassium is proved to be an excellent in-
dicator of a broad range of industrial discharges. In our case, potassium
is able to discriminate six types of industrial sources from domestic sew-
age. Chloride and sodium are capable of discriminating four industrial
sources; acesulfame is useful for capturing three industrial flow types
from the blending of industrial/sanitary wastewater. Extreme high
marker concentration levels were found in pharmaceutical manufactur-
ing sources, differing by 28–3820 times as compared to benchmark con-
centration, especially for the marker of clarithromycin. For the other

Table 3
Benchmark concentrations to discriminate between domestic and industrial sewage.

Industrial Sodium Chloride Potassium ISO Acesulfame Theanine Glycerol Clarithromyci

benchmark (mg/L) (mg/L) (mg/L) (µg/L) (µg/L) (µg/L) (µg/L) n (µg/L)
concentration ุ70 ุ250 ุ60 ุ200 ุ60 ุ15 ุ50 ุ2.0
Typical concentration range in industrial flow types
Aquatic products 254–479 798–1143 130–143
processing
Vegetable oil 27.3–79.4 126–156 352–433
processing
Starch and sugar 12000–13
44.2–47.8 36.3–45.7 23.7–27.2 800
manufacturing
Bakery products 414–514 456–776 476–532
manufacturing
Soybean products 325–487 536–632 524–698
processing
Candies/ chocolate 2240–4220
34.5–42.3 34.1–-45.3 4.4–8.3
manufacturing
Dairy products 357–423
39.5–55.7 39.4–52.3
manufacturing
Beer manufacturing 640–820
Carbonated beverage 260–400
manufacturing
Non-carbonated and 610–820 1070–1510
tea beverage
manufacturing
Bio-pharmaceutical 3013–5864 5087–7006 2065–4013 4010–7640
manufacturing

Note 1: The markers denoted by grey shading, can distinguish industrial sources from domestic sewage in 100% of samples.
Note 2: Concentrations of tartaric acid, pyruvic acid, hexylic acid, n-propyl alcohol and isobutyl alcohol in treated flows of beer
manufacturing were average 1.74 mg/L, 1.06 mg/L, 0.53 mg/L, 0.97 mg/L and 0.35 mg/L respectively; almost no marker data were
detected in sanitary sewage.
Note 3: Concentrations of tartaric acid, pyruvic acid, hexylic acid, n-propyl alcohol and isobutyl alcohol in treated flows of rice wine
manufacturing were average 2.8 mg/L, 0.66 mg/L, 0.38 mg/L, 0.16 mg/L and 0.23 mg/L respectively.
 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134
industrial flow types, their detected marker concentrations were ap-
proximately 1.8–101 times of benchmark data.
3.3. Detection of illicit discharges connected to storm drains using the
screened markers-demonstration study in downstream Taihu catchment,
China
The demonstration study site is located in downstream Taihu catch-
ment area, China and covers an area of 1.92 km2. Characterized as a fi-
nancial and trade zone, this area was designed as separate sewer and
storm drainage systems in the 1990s, with a storm pumping station
installed to pump the collected urban runoff into the nearby river on
wet-weather days (see Fig. 3). The current circumstance is non-
storm water sources also find their way into storm drains on
dry-weather days, resulting in unexpected dry-weather overflow
events through storm pumps operation. Therefore, the goal of the
demonstration project is to find the non-stormwater sources with
inappropriate entry into the storm drains using the screened
chemical markers, thus proving a strategy for correcting or fixing
the inappropriate discharge.
3.3.1. Process for tracing inappropriate flow entries in the field
A 72-h antecedent dry period was observed prior to the fieldwork to
minimize the possibility of sampling stormwater runoff rather than il-
licit connections.
There is only one outfall for the studied storm drainage system,
which is connected to the storm pumping station. At terminal outfall,
water samples were extracted from the terminal wet well using an
ISCO 6712 automatic field sampler (Teledyne, Lincoln, Nebraska, USA).
Water samples were collected every 3 h, resulting in 8 samples per
day. For the within-sewer sample collection or the track-down sam-
pling, manhole covers were lifted, and a bucket that could be laid hori-
zontally into shallow flow was used to collect water for a timed interval
of 3 h, resulting in 8 samples per day.
Marker investigations were conducted in two stages. The first stage
was performed during the period of Mar. 3 to Mar. 5, 2016, when
Fig. 3. Investigated storm drainage system with misconnected dry-weather flow entries in downstream Taihu catchment area, China.
131
terminal wet well and 7 up-pipe manholes along the trunk storm pipe
were investigated (see Fig. 3). The second stage took place from Apr.
14 to Apr. 16, 2016. We further refined the water sampling within the
final segment (i.e., between manhole 7 and 8), where 2 additional sam-
pling manholes (i.e., manhole 7a and 7b) were supplemented so as to
further clarify the types of non-stormwater sources inappropriately
connected to the storm drains, as seen in Fig. 3. Moreover, we measured
the dry-weather flows along the trunk pipe at three manholes
(i.e., manhole 2, 4, and 6) and terminal outfall during the first stage of
the investigation (see Supporting Information S5), so as to quantify
the non-stormwater flows of different source types along the storm
drains, based on the presence of markers.
Additionally, a groundwater well was drilled at the study site (see
Fig. 3) to monitor the marker profiles of shallow groundwater, consider-
ing potential groundwater inflow into the storm drains occurs in this
area, due to normal shallow groundwater levels (i.e., approximate
2.60 m) that are higher than the storm pipe invert, especially for deeper
storm drains that utilize gravity flow. Specifically, the groundwater
samples were collected by monitoring the well at a depth of 2 m
below ground; the samples were drawn from 30 cm and 60 cm below
the groundwater table, which represents the hydrochemistry of shallow
groundwater seeping into the storm drains. The sampling activities
were conducted for one week, and two samples were collected daily,
resulting in a total of 14 samples.
3.3.2. Determination of illicit discharges using the measured markers profile
Based on above analysis, considering this catchment features do-
mestic and commercial activities, the employed markers used to indi-
cate sewage contamination at this demonstration site were
acesulfame and theanine. For groundwater, the total hardness was mea-
sured as a supplementary indicator of storm drain infiltration, consider-
ing Ca and Mg compounds in the aquifer are likely to be dissolved by
percolating rainwater made acidic by carbon dioxide, based on previous
studies (Pitt et al., 1993; Xu et al., 2016).
The results of the measured marker concentrations down the trunk
pipe are shown in Fig. 4. As shown in this figure, the measured acesulfame
132 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134
(a)
(b)
(c)
Fig. 4. Plot showing the measured marker concentrations (a) acesulfame, (b) theanine and
(c) total hardness down the trunk storm pipe (mean ± standard deviation).
concentration profile initially increased from 5.37 μg/L to 8.82–8.88 μg/L at
manholes 3 and 4 and then decreased to 3.32 μg/L at terminal outfall. The
highest acesulfame concentrations are comparable to domestic wastewater
concentrations (10.4–12.2 μg/L) as discussed above, which is indicative of
raw sewage entry into the storm drains. Another concern was acesulfame
concentration fell sharply from 7.02 μg/L to 3.32 μg/L between manholes
7 and 8, decreasing by 52.7%. This indicates large non-stormwater inflows
from relatively clear sources instead of contaminated sources that entered
into the storm drains downstream of manhole 7.
The use of theanine and total hardness can further determine the char-
acteristics of non-stormwater sources connecting to the terminal trunk. Ac-
cording to the observed marker data collected at four manholes
(i.e., manhole 7, 7a, 7b and 8), while acesulfame concentration decreased
through the terminal segment, total hardness inversely increased from
126 mg/L to 191 mg/L. In addition, theanine concentration within this seg-
ment firstly increased from manhole 7 to 7a (i.e., from 7.37 μg/L to 8.42
μg/L) and subsequently decreased to 6.34 μg/L at the terminal outfall. In-
creased concentrations in total hardness and theanine within this segment
indicate that the non-stormwater flows were associated with large
amounts of groundwater and greywater. Furthermore, it is our goal to
quantify the non-stormwater flows of different source types in the storm
drains, based on chemical mass balance approach.
3.3.3. Quantification of illicit discharges using the mass balance of mea-
sured markers
In conjunction with measured markers, entry-exit chemical mass
balance (CMB) was established to determine source flow contributions
to the storm drain discharge on dry-weather days. A flow chart for
source apportionment and results output is presented in Fig. 5. In this
case, source flow components were apportioned by the parallel CMB
equations as a form of 3 × 3 matrix, corresponding to three potential
source types of blackwater, greywater and infiltrated groundwater.
Monte Carlo-based CMB approach was employed to overcome the un-
certainty in source apportionment modeling arisen by the factors such
as errors in the measurement of the input and output parameters, spa-
tial and temporal variability in the source profiles, and possible reac-
tions and mass exchange between phases (Xu et al., 2016). Thereby
the modeling output was presented while satisfying the apportioned
flow components added up to 100%.
Source apportionment results showed that non-stormwater flows
with inappropriate entries into the storm drains within this area mainly
arose from two priority areas. The first priority area was the trunk seg-
ment between manholes 2 and 4 (i.e., segment 2), where 1025 m3/d
blackwater and 3158 m3/d were quantified. The ratio of quantified
blackwater to greywater was about 0.25 to 0.75, coinciding well with
the flow components of residential wastewater as mentioned above
(Blanky et al., 2015; Yi et al., 2011). This indicated that non-
stormwater flows into trunk segment 2, were mainly from nearby resi-
dential communities, according to the land use map. This circumstance
may be explained as follows. Historically, sanitary sewers were not in
existence in this catchment, since early storm drains preceded the de-
velopment of sanitary sewers. Therefore, sewage from these older com-
munities is connected to early municipal storm pipes under the traffic
road and discharged into nearby watercourses. However, after the con-
struction of separate municipal sewage system in the 1990s in this area,
sewage discharge sources from these old residential areas are still not
reconnected to the developed municipal separate sewers, possibly due
to undeveloped separate sewers and storm pipes in these living com-
munities. Often, old combined sewer pipes were connected to the mu-
nicipal storm drains to prevent flooding in case of rainfall.
The second priority area was trunk segment 4 between manhole 6
and terminal outfall, where 7972 m3/d non-stormwater flow including
3164 m3/d greywater and 4808 m3/d infiltrated groundwater was
quantified. This accounted for 53.1% of the total dry-weather flow inap-
propriately entering the storm drains. In contrast, very less blackwater
was quantified. Large amounts of greywater entry were associated
with the wastewater from nearby commercial buildings that utilize
kitchen sink, and other greywater facilities. Another interesting phe-
nomenon was that the vast majority of infiltrated groundwater came
from trunk segment 4, accounting for 90% of the total groundwater dis-
charge within the entire storm drain system. Potential pipe defects as
well as gradually deepened pipe inverts down the trunk pipe are the
reasons for the generation of large amounts of groundwater inflow
within the terminal segment. This indicates that obvious groundwater
infiltration may not occur throughout the entire storm drain system
but instead occurs at several key pipe segments.
Therefore, based on the above discussion, we can narrow our inves-
tigation to one or several key pipe segments where inappropriate en-
tries of residential wastewater, commercial wastewater, or
groundwater are of major concern. In this way, the ability to track
major discharges to sources will be less labor intensive and more effec-
tive. This case study strengthens the importance of identifying markers
to indicate dry-weather flow contamination within storm drains.
4. Conclusions
The present study provides the first quantitative data on the occur-
rence and removal of 52 chemical markers during biological treatment
at WWTPs of industrial enterprises associated with agricultural and
 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134 133

observations of blackwater source 1 source data generation based on normal distribution

data library of acesulfame, theanine and blackwater greywater groundwater
total hardness in blackwater respectively source 2 μ σ μ σ μ σ
acesulfame (μg/L) 37.2 6.1 1.47 0.22 0.02 0.002
theanine (μg/L) 0.10 0.05 10.1 1.29 0.01 0.003
observations of greywater source 3 total hardness (mg/L) 154 21.5 127 12.8 300 50.1
data library of acesulfame, theanine and
total hardness in greywater respectively
outfall manhole 6 manhole 4 manhole 2
μ σ μ σ μ σ μ σ
observations of groundwater chemistry acesulfame (μg/L) 3.32 1.8 6.97 2.16 8.82 3.14 6.63 2.3
data library of acesulfame, theanine and theanine (μg/L) 6.34 2.27 7.13 1.34 6.12 2.18 5.85 1.93
total hardness in groundwater exit total hardness (mg/L) 191 10.3 125 12.5 137 12.2 184 11.8
respectively

source apportionment based on CMB matrix

observations of storm drains outfall and
constraint condition: source flow components adding up to 100%
selected manholes
measured data sets of acesulfame,,
theanine and total hardness at wet well
and manholes apportioned source flow components and general source chemistry

(a) Source apportionment schemes for non-stormwater entries into storm drains

(b) Source apportionment modeling output

Fig. 5. Apportioned percent source flow components and dry-weather flows of three source types within the divided four segments of the storm drain.

sideline products, alcohols/beverage products, and pharmaceutical
products, which are also consumed in our daily life. The purpose of
this work is to identify new chemical markers that can be used to
track sources of contamination in dry-weather flow. This is the first
study to screen chemical markers associated with industrial activities,
and meanwhile the results deepen our understanding of domestic sew-
age markers. The major findings are as follows:
– Sodium, chloride, potassium, isomalto-oligosaccharide, acesulfame,
tartaric acid, theanine, pyruvic acid, glycerol, and clarithromycin
were identified as conservative markers, with changes in concentra-
tion through WWTP processes of b30%. These markers are useful to
indicate potential inappropriate industrial flow entries into storm
drains, arising from agricultural food processing or manufacturing,
drinks manufacturing, and pharmaceutical manufacturing.
– After ingestion and digestion through the human body, sodium,
chloride, potassium, isomalto-oligosaccharide, acesulfame,
theanine, glycerol, and clarithromycin were detected at certain con-
centration levels in domestic raw sewage, whereas tartaric acid and
pyruvic acid were almost not detected in domestic raw sewage sam-
ples. Specifically, sodium, chloride, potassium, isomalto-
oligosaccharide, acesulfame, and clarithromycin were found in
higher concentrations in blackwater than in greywater, and in con-
trast, theanine and glycerol were measured in higher concentrations
in greywater than in blackwater.
– Acesulfame and theanine were implemented as markers, in
combination with CMB-based source apportionment modeling,
to quantify blackwater and greywater entries into the
stormwater system at a domestic and commercial catchment in
downstream Taihu Lake, China. Total hardness was also
measured to discriminate and quantify groundwater discharge
into the storm pipes. At this demonstration site, priority areas
that received inappropriate residential wastewater, commercial
wastewater, and infiltrated groundwater were identified. This
demonstrates that the assessment of suitable chemical markers
to indicate source flow contamination is very important when
applying a low-cost approach to track inappropriate connections
to stormwater systems.
Acknowledgment
This work was supported by China's Major S&T Project on Water Pol-
lution Control and Treatment (Grant nos. 2017ZX07206-001 and
2013ZX07304-002).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.scitotenv.2018.07.448.
134 H. Yin et al. / Science of the Total Environment 648 (2019) 125–134
References
Barrett, M.H., Hiscock, K.M., Pedley, S., Lerner, D.N., Tellam, J.H., French, M.J., 1999. Marker
species for identifying urban groundwater recharge sources: a review and case study
in Nottingham, UK. Water Res. 33 (14), 3083–3097.
Blanky, M., Rodríguez-Martínez, S., Halpern, M., Friedler, E., 2015. Legionella
pneumophila: from potable water to treated greywater, quantification and removal
during treatment. Sci. Total Environ. 533, 557–565.
Brown, E., Caraco, D., Pitt, R., 2004. Illicit Discharge Detection and Elimination, a Guidance
Manual for Program Development and Technical Assessments. 378. Center for Wa-
tershed Protection, Maryland, U.S.
Buerge, I.J., Buser, H.R., Kahle, M., Muller, M., Poiger, H., 2009. Ubiquitous occurrence of
the artificial sweetener acesulfame in the aquatic environment: an ideal chemical
marker of domestic wastewater in groundwater. Environ. Sci. Technol. 43 (12),
4381–4385.
Daneshvar, A., Aboulfadl, K., Viglino, L., Broseus, R., Sauve, S., Madoux-Humery, A.S.,
Weyhenmeyer, G.A., Prevost, M., 2012. Evaluating pharmaceuticals and caffeine as in-
dicators of fecal contamination in drinking water sources of the Greater Montreal re-
gion. Chemosphere 88 (1), 131–139.
Field, R., Pitt, R., Lalor, M., Brown, M., Vikelis, W., Phackston, E., 1994. Investigation of dry
weather pollutants entries into storm drainage systems. J. Environ. Eng. 120 (5),
1044–1066.
Fono, L.J., Sedlak, D.L., 2005. Use of the chiral pharmaceutical propranolol to identify sew-
age discharges into surface waters. Environ. Sci. Technol. 39 (23), 9244–9252.
Gasser, G., Rona, M., Voloshenko, A., Shelkov, R., Lev, O., Elhanany, S., Lange, F.T., Scheurer,
M., Pankratov, I., 2011. Evaluation of micropollutant tracers. II. Carbamazepine tracer
for wastewater contamination from a nearby water recharge system and from non-
specific sources. Desalination 273 (2), 398–404.
Gobel, A., MaArdell, C.S., Joss, A., Siegrist, H., Giger, W., 2007. Fate of sulfonamides,
macrolides, and trimethoprim in different wastewater treatment technologies. Sci.
Total Environ. 372 (2–3), 361–371.
Guerra, P., Kim, M., Shah, A., Alaee, A., Smyth, S.A., 2014. Occurrence and fate of antibiotic,
analgesic/anti-inflammatory, and antifungal compounds in five wastewater treat-
ment processes. Sci. Total Environ. 473, 235–243.
Jelic, A., Rodriguez-Mozaz, S., Barcelo, D., Gutierrez, O., 2015. Impact of in-sewer transfor-
mation on 43 pharmaceuticals in a pressurized sewer under anaerobic conditions.
Water Res. 68, 98–108.
Kasprzyk-Hordern, B., Dinsdale, R.M., Guwy, A.J., 2009. Illicit drugs and pharmaceuticals in
the environment-forensic applications of environmental data, part 2: pharmaceuti-
cals as chemical markers of fecal water contamination. Environ. Pollut. 157 (6),
1778–1786.
Kuroda, K., Murakami, M., Oguma, K., Muramatsu, Y., Takada, H., Takizawa, S., 2012. As-
sessment of groundwater pollution in Tokyo using PPCPs as sewage markers. Envi-
ron. Sci. Technol. 46 (3), 1455–1464.
Lim, F.Y., Ong, S.L., Hu, J., 2017. Recent advances in the use of chemical markers for tracing
wastewater contamination in aquatic environment: a review. Water 9 (7), 143–168.
Nakada, N., Kiri, K., Shinohara, H., Harada, A., Kuroda, K., Takizawa, S., Takada, H., 2008.
Evaluation of pharmaceuticals and personal care products as water-soluble molecular
markers of sewag0065. Environ. Sci. Technol. 42 (17), 6347–6353.
View publication stats
O'Brien, J.W., Thai, P.K., Eaglesham, G., Ort, C., Scheidegger, A., Carter, S., Lai, F.Y., Mueller,
J.F., 2014. A model to estimate the population contributing to the wastewater using
samples collected on census day. Environ. Sci. Technol. 48 (1), 517–525.
Pitt, R., Lalor, M., Field, R., Adrian D.D., Barbel, D., 1993. Investigation of inappropriate pol-
lutant entries into storm drainage systems: a User's Guide. 87. Risk Reduction Engi-
neering Lab., US EPA, Cincinnati.
Revitt, D.M., Ellis, J.B., 2016. Urban surface water pollution problems arising from miscon-
nections. Sci. Total Environ. 551–552, 163–174.
Sun, Q., Li, M., Ma, C., Chen, X., Xie, X., Yu, C.P., 2016. Seasonal and spatial variations of
PPCP occurrence, removal and mass loading in three wastewater treatment plants lo-
cated in different urbanization areas in Xiamen, China. Environ. Pollut. 208 (Pt B),
371–381.
Tran, N.H., Hu, J., Li, J., Ong, S.L., 2014a. Suitability of artificial sweeteners as indicators of
raw wastewater contamination in surface water and groundwater. Water Res. 48,
443–456.
Tran, N.H., Li, J., Hu, J., Ong, S.L., 2014b. Occurrence and suitability of pharmaceuticals and
personal care products as molecular markers for raw wastewater contamination in
surface water and groundwater. Environ. Sci. Pollut. Res. 21 (6), 4727–4740.
Tran, N.H., Reinhard, M., Gin, K.Y.H., 2018. Occurrence and fate of emerging contaminants
in municipal wastewater treatment plants from different geographical regions-a re-
view. Water Res. 133, 182–207.
Unno, T., Suzuki, Y., Kakuda, T., Hayakawa, T., Tsuge, H., 1999. Metabolism of theanine, γ-
glutamylethylamide, in rats. J. Agric. Food Chem. 47 (4), 1593–1596.
Xu, Z.X., Yin, H.L., Li, H.Z., 2014. Quantification of non-stormwater flow entries into storm
drains using a water balance approach. Sci. Total Environ. 487, 381–388.
Xu, Z.X., Wang, L.L., Yin, H.L., Li, H.Z., Schwegler, B.R., 2016. Source apportionment of non-
storm water entries into storm drains using marker species: modeling approach and
verification. Ecol. Indic. 61, 546–557.
Yang, Y.Y., Liu, W.R., Liu, Y.S., Zhao, J.L., Zhang, Q.Q., Zhang, M., Zhang, J.N., Jiang, Y.X.,
Zhang, L.J., Ying, G.G., 2017. Suitability of pharmaceuticals and personal care products
(PPCPs) and artificial sweeteners (ASs) as wastewater indicators in the Pearl River
Delta, South China. Sci. Total Environ. 590, 611–619.
Yi, L.L., Jiao, W.T., Chen, X.N., Chen, W.P., 2011. An overview of reclaimed water reuse in
China. J. Environ. Sci. 23 (10), 1585–1593.
Yin, H.L., Lu, Y., Xu, Z.X., Schwegler, B.R., 2017. Characteristics of the overflow pollution of
storm drains with inappropriate sewage entry. Environ. Sci. Pollut. Res. 24,
4902–4915.
Zhang, J., Wang, J., Hua, P., Krebs, P., 2015. The qualitative and quantitative source appor-
tionments of polycyclic aromatic hydrocarbons in size dependent road deposited
sediment. Sci. Total Environ. 505, 90–101.
Zhang, J., Hua, P., Krebs, P., 2016. The influences of dissolved organic matter and surfac-
tant on the desorption of Cu and Zn from road-deposited sediment. Chemosphere
150, 63–70.
Zheng, H.J., Ni, D.J., Yu, Z., Liang, P., 2017. Preparation of SERS-active substrates based on
graphene oxide/silver nanocomposites for rapid zdetection of L-Theanine. Food
Chem. 217, 511–516.