Science of the Total Environment 499 (2014) 265–275

Contents lists available at ScienceDirect

Science of the Total Environment

j ourna l homepage: www.e lsev ie r .com/ locate /sc i totenv

Substance flow analysis and assessment of environmental exposurepotential for triclosan in mainland China

Chu-Long Huang a,b, Hwong-Wen Ma c, Chang-Ping Yu a,⁎a Institute of Urban Environment, Chinese Academy of Sciences, 1799, Jimei Road, Xiamen 361021, Chinab Department of Resources and Environmental Sciences, Quanzhou Normal University, 398, Donghai Street, Quanzhou 362000, Chinac Graduate Institute of Environmental Engineering, National Taiwan University, 71, Chou-Shan Road, Taipei 10660, Taiwan

H I G H L I G H T S

• 78% of TCS consumed in 2008 unsustainably discharged into the natural environment.• Most TCS in China is discharged to surface water sediment, ocean, and soil.• Higher consumption levels increased TCS accumulation in the environment.• Environmental exposure potential to TCS in China is increasing.• SFA is a well-established tool to provide information for exposure assessment of TCS.

Abbreviations: TCS, triclosan; SFA, substance flow analypersonal care products; WWTPs, wastewater treatment pHCE-TCS, per capita annual urban household consumptiproducts.⁎ Corresponding author. Tel.: +86 592 6190768.

E-mail address: cpyu@iue.ac.cn (C.-P. Yu).

http://dx.doi.org/10.1016/j.scitotenv.2014.08.0320048-9697/© 2014 Elsevier B.V. All rights reserved.

a b s t r a c t

a r t i c l e i n f o

Article history:Received 1 May 2014Received in revised form 9 August 2014Accepted 10 August 2014Available online 3 September 2014

Editor: E. Capri

Keywords:TriclosanSubstance flow analysisEnvironmental exposure potentialChina

Triclosan (TCS) is a widely-used antimicrobial agent in many consumer products around theworld, and China is amajor producer and consumer of TCS. In this study substance flow analysis (SFA) was used to construct a staticmodel of anthropogenic TCSmetabolism in China in 2008. The systematic SFA resultswere used to determine pos-sible exposure pathways and trends in environmental exposure potential through different pathways. TCSdischarged in wastewater mainly flowed into surface water sediment, ocean, and soil, where it accumulates inaquatic and agricultural products that may pose a higher risk to human health than brief exposure during con-sumption. Only 22% of TCS discharged was removed in the built environment with the remainder dischargedinto the natural environment, indicating that anthropogenic TCS metabolism in China is unsustainable. Per capitaTCS consumption increased 209% from 2003 to 2012, resulting in increased discharge and accumulation in the en-vironment. If current trends continue, it will increase to 713mg capita−1 yr−1 in 2015 and 957mg capita−1 yr−1

in 2020. Accordingly, annual environmental exposure potential will increase from 388 mg capita−1 in 2008 to557mg capita−1 in 2015 and 747mg capita−1 in 2020, indicating an increasing trend of exposure to environmen-tal TCS. Results of Pearson correlation analysis suggested that feasible countermeasures to reduce environmentalexposure potential for triclosan would include encouraging the development of small cities, raising awareness ofhealth risks, nurturing environmentally-friendly consumer values, and improving the environmental performanceof TCS-containing products.

© 2014 Elsevier B.V. All rights reserved.

1. Introduction

Triclosan (TCS) has been marketed for over 40 years (Glaser, 2004)and is produced by manufacturers located in the U.S.A., Switzerland,the Netherlands, China, India, South Korea, and elsewhere (APUA,

sis; PPCPs, pharmaceuticals andlants; MeTCS, methyl-triclosan;on expenditure on TCS-related

2011). It has been suggested that TCS inhibits lipid biosynthesis by act-ing upon enzymes (Levy et al., 1999;McMurry et al., 1998); as a result, ithas become a widely-used antimicrobial agent in pharmaceuticals andpersonal care products (PPCPs).

However, because TCS is an artificial chemical with no known natu-ral sources (HCEC, 2012), the continuous consumption of TCS createsthe risk of “environmental allelopathy”, meaning that humans maynot be acclimatized to the changing human–microorganisms relation-ship caused by the continuing environmental occurrence or accumula-tion of TCS (Gautam et al., 2014; Halden and Paull, 2005; Lindstromet al., 2002; Miller et al., 2008). Environmental accumulation is a resultof TCS migrating from consumer products to the natural environment

266 C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

(Davis et al., 2012; Hua et al., 2005; Ying and Kookana, 2007) where itpotentially disrupts critical ecological processes performedby beneficialmicroorganisms (Dokianakis et al., 2004; Neumegen et al., 2005; Priceet al., 2010). TCS is acutely toxic to some aquatic organisms (particularlycertain algae species) at low μg L−1 levels, and has been shown tomod-ulate thyroid functions in amphibians at concentrations as low as0.15 μg L−1 (Chu and Metcalfe, 2007). Environmental accumulationcould also result in the appearance of bacteria strains with TCS-insensitive enzymes, including so-called ‘super-bugs’ (bacteria thathave developed resistance to the effects of antibiotic drugs) such asthe new strains of tuberculosis (Phillips, 1998). Thus, a microbial envi-ronmentwhich is continually changed by TCSmay result in disease out-breaks among vulnerable people. TCS has been found to be omnipresentin human tissues (plasma, milk, liver, adipose and brain) around theworld, including in Australia, the U.S.A., Germany, Sweden, Belgium,South Korea, and China (Allmyr et al., 2006; Dann and Hontela, 2011;Geens et al., 2012; Li et al., 2013; NICNAS, 2009; SCCP, 2009); forexample, TCS was detected in 74.6% of the urine samples of the generalBelgian population (Pirard et al., 2012). TCS has been classified as a po-tential carcinogen and toxicant (Fang et al., 2010) and can be metabo-lized to dichlorodibenzo-p-dioxin (Lores et al., 2005). Consumers whouse TCS-containing products on a daily basis risk reduced levels ofDNA methylation in human hepatocytes and down-regulated MBD2,MBD3, and MeCP2 gene expression (Ma et al., 2013).

TCS has been identified as a chemical which should be a monitoringand research priority due to human toxicity and evidence of adverse en-vironmental effects and endocrine disruption (Clarke and Smith, 2011).However, studies usually focus on independent assessment of individu-al risk sources, mainly by microexamination of molecules or cells(Binelli et al., 2009; Canesi et al., 2007; Crofton et al., 2007; HCEC,2012; James et al., 2010; Kawanai, 2011; Lin et al., 2010; Marlatt et al.,2013; Matozzo et al., 2012; Tamura et al., 2012). At present, existingknowledge of the potential environmental and human risks from TCSuse is insufficient to carry out a comprehensive risk assessment due toa general lack of data related to TCS distribution in the environmentand the amounts of TCS transmitted from the environment to humans,and the most important transmission pathways.

Substance flow analysis (SFA) is usually used to assess the sustain-ability of socioeconomic development and environmental change, par-ticularly from the perspective of improving substance flow efficiency(Huang et al., 2012; Zhang et al., 2014), but more recently it has alsobeen used to investigate hazardous material for environmental risk as-sessment and management (Asari et al., 2008; Chèvre et al., 2013;Eriksson et al., 2008; Earnshaw et al., 2013; Herva et al., 2012; Longet al., 2013; Oguchi et al., 2013; Ono, 2013). The integration of SFAand risk assessment can convey amore comprehensive picture than tra-ditional risk assessment; this also facilitates examination of risk distri-bution in a systemic way, bringing out additional information that canbe used to formulate efficient management policies (Ma et al., 2007).In this study, we used a mass balance approach to comprehensivelyquantify TCS flows and distribution in China to determine possible ex-posure pathways for TCS. Then, trends in TCS consumption and environ-mental exposure potential for TCS were analyzed in order to identifyappropriate measures to reduce the risks resulting from TCS discharge.

2. Methodology

2.1. Analytical framework

SFA is an analytical method used to systematically assess the flowand stock of a substance through a given system defined in space andtime (Brunner and Rechberger, 2004; Huang et al., 2014). It can beused to help construct a systematic database and determine criticallinks or pathways, and can also facilitate the quantifying of substanceflows within the socio-economic and environmental systems (Brunner

and Rechberger, 2004; European Communities, 2001; Huang et al.,2012, 2013; Sendra et al., 2007).

To understand TCS flowdistribution in the anthroposphere in China,a coupled human and environmental system approach was adopted(Srinivasan et al., 2013; Stannard and Aspinall, 2011). The systemboundary is the national border of mainland China, without HongKong, Macau and Taiwan (Fig. 1). The system was divided into threegroups: 1) TCS flows in the built environment which include importsand exports, production, consumption, discharge and removal (includ-ing degradation, landfill, and incineration by built facilities); 2) TCSflows in environment media; and 3) human exposure.

2.2. Data collection

Because it is not known howmuch TCS is released from domestic orindustrial sites (e.g. where TCS is incorporated into plastic and textileitems) into the environment, TCS discharge mass balance was basedon monitoring of TCS concentration in waterways and effluents. Wefound that the monitoring data for China in the published literaturewas more complete in 2008 than in other years, so 2008 was selectedfor this study. TCS and other related data in China for 2008were derivedfrom 1) peer-reviewed literature; 2) Chinese government statisticalyearbooks or bulletins, which represent the best available data for quan-tification and forecast of anthropogenic TCS consumption and dischargeflows in China (this includes but is not limited to the National Bureau ofStatistics, Ministry of Housing and Urban–Rural Development, Ministryof Water Resources, Ministry of Environmental Protection, General Ad-ministration of Customs, Ministry of Commerce, National Developmentand ReformCommission, andDevelopment Research Center of the StateCouncil); 3) industrial reports from chemical associations and researchunits, especially ‘Report of TCS technology andmarket in China’ (S6CCN,2013), which provides data on national production, consumption,import, and export of TCS; 4) personal communication with somecompanies (e.g., Weifang Aoyou Chemical Technology Co., Ltd); 5)survey of several major producers and users of TCS; 6) mass balanceacross the socio–economic–environmental system; 7) proxy dataand known scientific relationships; and 8) estimation by industrialassociation studies.

2.3. Accounting approach

2.3.1. TCS flow distributionWe assumed that 1) the stock of TCS in production or consumption

was the same in different years, so TCS stocks need not be considered in2008, and 2) TCS spreads out in wastes or each environmental mediumuntil an even concentration is achieved in a basin. This allowed us to de-velop a static physical flow model with detailed TCS balance calcula-tions as indicated by Eqs. (1)–(14) in Table 1.

2.3.2. TCS consumption trendsIn order to select a reference indicator for TCS consumption forecast-

ing, Pearson correlation analysis was conducted between TCS consump-tion and its most likely factors based on the literature (Katz et al., 2013;NICNAS, 2009; Zhao et al., 2013). According to results of Pearson corre-lation analysis (Table S1, Supplementary data), the highest correlationcoefficient was between population and apparent consumption of TCS,so annual increase in per capita TCS consumption (Ra) is taken as an in-dicator for forecasting; this was calculated by Eq. (15) and validated byEq. (16).

Ra ¼ Vr=Vsð Þ1= yr−ysð Þ−1� �

� 100% ð15Þ

Ry ¼ Vyt−Vyt−1

� �=Vyt−1

� �� 100% ð16Þ

Fig. 1.Key TCS flows and possible exposure pathswithin a country. TCS flows in the built environment are shown in square boxes and human and environmental media in the ovals (‘builtenvironment’ refers to artificial facilities during imports and exports, production, consumption, discharge and removal, including factories, storehouse, WWTPs and waste treatmentplants; ‘environmental media’ refers to air, sediment, freshwater, sea and soil). Dashed arrows mean the pathways of TCS removal from the system. Big dashed box indicates the nationalboundary.

267C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

where Ra is average annual increase rate of per capita TCS consumption(%); Vr is per capita TCS consumption in reference year (kg person−1);Vs is per capita TCS consumption in starting year (kg person−1); yr isreference year (i.e. 2012) and ys is starting year (i.e. 2012minus numberof years for average annual increase rate calculation); Ry is year-on-yearincrease rate in a given year (%); Vyt is per capita TCS consumption intarget year (kg person−1); Vyt − 1 is per capita TCS consumption inthe year before target year (kg person−1).

Cp1 ¼ Pt � Cpc ð17Þ

Cp2 ¼ Vra � 1þ Ra=100ð Þn ð18Þ

where Cp1 is apparent TCS consumption for predicted years (kg), whichcan be validated by Eq. (18); Pt is total population in predicted year(person); Cpc is per capita TCS consumption for predicted year(kg person−1), based on per capita TCS consumption in reference yearand Ra or Ry; Cp2 is apparent TCS consumption for predicted years(kg); Vra is apparent TCS consumption in reference year (2012) (kg);n is the number of years between the reference year and predictedyear (yr).

2.3.3. Environmental exposure potentialAssuming that 1) the capacity of WWTPs to remove TCS and 2) the

proportional structure of TCS mass variables from consumption willnot change from 2008 to the predicted year, TCS environmental expo-sure potential can be estimated by Eq. (19).

EEPx ¼ Cx−Dsr−Dtd−Dta � 22:1%ð Þ � 106h i

=Pt ð19Þ

where EEPx is TCS environmental exposure potential (mg capita−1); Dsr

is removed TCS by incineration and landfill from total TCS dischargedwith solid waste (‘Assumption for Eq. (2)’, Supplementary data) (kg).The sum of Dtd and Dta is Dt (see Eq. (3)); Dtd is TCS mass removed bybiodegradation or other processes in WWTPs (kg) (this was 67.71% ofDt, as 32.29% of Dt was adsorbed onto sludge after treatment byWWTPs (Heidler and Halden, 2008)); Dta is TCS mass absorbed in thesludge from WWTPs (kg). It can be assumed that 22.1% of Dta is

removed by landfill with leachate collection and treatment systems(MOC, 2006), for details see ‘Assumption for Eq. (2)’ in Supplementarydata.

Ypy ¼ Yr=C2008 � Cpy ð20Þ

where Ypy is variable in predicted year (kg), i.e. Dsr, Dtd, and Dta in 2015or 2020 (see Eq. (19)); Yr is variable in reference year (kg), i.e. Dsr, Dtd,and Dta in 2008; C2008 is TCS apparent consumption in 2008 (kg); Cpyis TCS apparent consumption in predicted years (kg), i.e. Cp1 and Cp2,based on Eqs. (17) and (18), respectively.

3. Results and discussion

3.1. Production and consumption of TCS

Total use of TCS in China before 2000 was only a few tons, most ofwhichwas imported. However, in the past decademarket demand com-bined with technological advances strongly promoted the progress ofthe TCS-related industry, transforming the country from a net importerto a net exporter. TCS production increased from 500 t in 2003 to 2900 tin 2012, most of which was produced by a few major manufacturers(Table S3, Supplementary data); since 2006 over 70% of the productionhas been exported and import of TCS remains very low (Table S4, Sup-plementary data). TCS consumption increased from 250 t in 2003 to810 t in 2012 (Table S4, Supplementary data); based on total population(NBSC, 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012), percapita consumption of TCS in China increased from 193 mg yr−1 in2003 to 598mg yr−1 in 2012, rapidly approaching levels seen in devel-oped countries in recent years (Table S5, Supplementary data). This hasinevitably resulted in increasing discharge and environmental accumu-lation of TCS.

3.2. TCS discharge and environmental fate

3.2.1. TCS mass with wastewater discharge

3.2.1.1. TCS with wastewater discharged into the environment. UsingEqs. (11) and (12), the results estimated by population show that TCSmass inventories in air, soil, water, and sediment from wastewater

Table 1Equations used for the calculations of triclosan flows in China.

Equation No. Note Source

∑m

i¼1Pi þ∑

n

j¼1I j ¼ ∑

a

x¼1Cx þ ∑

b

y¼1Ey

(1) Pi, Cx, Ij, and Ey are amounts of TCS production, consumption, import, and export,respectively (kg). Cx is apparent consumption.

S6CCN (2013)

∑a

x¼1Cx ¼ ∑

c

w¼1Dw þ∑

d

s¼1Ds þ ∑

e

v¼1Dv

(2) Dw, Ds, and Dv are TCS mass discharged by wastewater, solid waste and volatilization,respectively (kg).[1]

Dw: Eq. (3); Ds: Eq. (2)

∑c

w¼1Dw ¼ ∑

h

o¼1Do þ∑

f

t¼1Dt

(3) Do includes both TCS mass being treated but not removed by wastewater treatment plants(WWTPs) and not being treated by WWTPs (kg); Dt is TCS mass removed by WWTPs(kg).[2]

Do: Eq. (7); Dt: Eq. (6)

Dt = Rr × Mau (4) Rr is TCS removal rate of WWTPs, i.e. the median of TCS removal efficiencies of WWTPsaround the world (%); Mau is total TCS mass in the urban wastewater discharge whichwould be treated by WWTPs (kg).

Rr: Table S7; Mau: Eq. (5)

Mau = (Do + Dt) × Wtu/W (5) Wtu is the treated urban wastewater discharge (m3); W is total wastewater discharge inChina (m3).

Wtu: MOHURD (2012);W: WRC (2008)

Dt = Do × (Wtu/W) ÷ (1/Rr − Wtu/W) (6) This is derived from Eqs. (4) and (5).Do = Mt + Mo (7) Mt is TCS flowing into continental environmental media (kg); Mo is amount of TCS flowing

into the ocean (kg).Mt: Eq. (8); Mo: Eq. (13)

Mt = Mair + Msoil + Mwater + Msediment (8) Mair, Msoil, Mwater, and Msediment are parts of TCS mass sourced from wastewater dischargeswhile flowing into the air, soil, water, and sediment, respectively (kg).[3]

Mwater, Msediment: Eq. (9); Mair,Msoil: Eq. (12)

Mti = Mm × Pb/Ps (9) Mti is the mass inventory of TCS in a given basin or area, including Mwater and Msediment[4] ; Mm

is the measured mass inventory in the simulated area of a given basin (kg); Pb is the totalpopulation in a given basin, Ps is the population in the simulated area of the given basin (person).

Mm: Eq. (10); Pb: Eq. (11)[5];Ps: Zhao et al. (2013)

Mm = Ms × Cm/Cs (10) Ms is the simulated mass inventory in the simulated area of a given basin (kg); Cm is themeasured TCS concentration (ng L−1 or ng g−1); Cs is the simulated TCS concentration(ng L−1 or ng g−1).

Ms: Table S6; Cm, Cs: Zhao et al.(2013)[6]

Pb = Wtwu/Wpcwu (11) Wtwu is the total water use in a given basin (m3); Wpcwu is per capita water use in a givenbasin (m3 person−1).

Wtwu, Wpcwu: local waterbulletins

Mtj = Ms × Pb/Ps (12) Mtj is the simulated mass inventory of TCS in the air and soil in a given basin, including Mair

and Msoil.[7]Ms: Table S6; Pb, Ps: see Eq. (9)

Mo = Mww + Mwsp + Mws (13) Mo is total amount of TCS flowing into the ocean (kg); Mww, Mwsp, and Mws are the amountsof TCS in water, suspended particulate matter, and sediment, respectively (kg).

Mww, Mwsp, Mws: Eq. (14)

Moi = Wsea × Ci (14) Moi is Mww, Mwsp, or Mws; Wsea is water volume of runoff into the ocean (m3); Ci is themeasured TCS concentration in water (ng L−1), suspended particulate matter (ng L−1) orsediment (ng g−1).[8]

Wsea: local water bulletins; Ci:Zhao et al. (2013) and Table 3.

Notes: [1]. Dv is assumed to be 0 kg, 24.7% of Ds was removed by incineration and landfill, see ‘Assumption for Eq. (2)’ in Supplementary data.[2]. Dt is calculated by Eqs. (4)–(13) after assuming that 1) TCS concentration of untreated rural domesticwastewater is the same as untreated urban domesticwastewater, and 2) all ruraldomestic wastewater has not been treated by WWTPs.[3]. Mair, Msoil, Mwater, and Msediment were estimated by population in accordance with the following three facts (see Table S6 and Table 3): (1) Based on TCS-relevant literature, a Pearsoncorrelation analysis showed the correlation coefficient of total population was highest among selected factors affecting TCS consumption, the secondwas urbanization level, and the thirdwas household consumption index at constant prices (Table S1, Supplementary data); (2) Experimental evidence shows that the mass inventories of TCS in riverine environments at theriver basin scale are significantly influenced by population (Zhao et al., 2013); (3) TCS is mainly used in the manufacture of PPCPs owing to its distinctive physicochemical properties(Table S2, Supplementary data). For example, in theU.S.A., over 95% of TCS is used in consumer products that are disposed of in residential drains (APUA, 2011; Reiss et al., 2009),meaningthat TCS is mainly discharged by domestic wastewater (HCEC, 2012).[4]. The calculation ofMwater andMsedimentwas estimated by themeasured TCS concentration inwater and sediment by Zhao et al. (2013) and other literature (see Eqs. (9)–(11)), becauseTCS is sufficiently persistent in the environment (Katz et al., 2013).[5]. Pb for the Hai and Huang Rivers was adapted from HRCMWR (2012) and HRCMWR (2013), respectively.[6]. Both are sourced from Zhao et al. (2013), but Cm of water and sediment for the Yangtze River is the median between the simulated value from Zhao, et al. (2013) and the measuredvalue from Zheng et al. (2011); Cm of water and sediment for the whole country is calculated on MOWR (2009) and MOWR (2008).[7]. Mair and Msoil were estimated according to the simulatedmass inventory for China's major rivers at the watershed scale by fugacity Level III model simulation (Zhao et al., 2013), be-cause there was insufficient information on measured TCS concentrations in air and soil; mass inventories of TCS in air and soil are extremely limited compared with that in water andsediment.[8]. Ci in suspended particulate matter in the Yangtze River (Cspm-yangtze) was calculated using Eq. (S1).

268 C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

discharge were 1.2 kg, 68.5 kg, 5893 kg, and 144,903 kg in 2008, re-spectively (Table S6, Supplementary data). In contrast, based onbasin area, the simulated TCS mass inventories would be more thantotal apparent consumption (Table S6, Supplementary data), sug-gesting that estimation based on population level yields more accu-rate results than simulation by area.

TCS discharged into waters will rapidly degrade or be absorbed intothe sediment. Thus, it would be better to simulate mass inventories inwater and sediment using measured TCS concentrations in five largerivers in China (Zhao et al., 2013). TCS mass inventories in air and soilsimulated by the fugacity Level III model can be accepted because 1)the simulated values were generally within the same order of magni-tude as the measured values (Zhao et al., 2013), and 2) TCS is more dif-ficult to measure in air and soil than in water and sediment, because①TCS is more likely to receive solar light for photolysis and has muchshorter half-life in air than that in water and sediment (SCCS, 2010);② the distribution of TCS is more uneven in soil than in water and sed-iment due to its sources and diffusion modes, and TCS concentrationand its half-life vary between different soil types (Ying et al., 2007;

Langdon et al., 2011; Waria et al., 2011; Lozano et al., 2012); and 3)given the very low volatility of TCS, partitioning to air is not expectedto be a significant migration pathway (NICNAS, 2009).

Applying the simulated values of 1.2 kg in air and 68.5 kg in soilgiven by the fugacity Level III model (Table S6, Supplementary data)and the estimated values of 14,958 kg in water and 188,807 kg in sedi-ment (based on the measured TCS mean concentration; see Table 2),then, total mass inventory in all continental environmental media inChina was 203,834 kg. Because 168,951 kg TCS flowed into the ocean(Table 3), there should be 372,785 kg TCS directly discharged into theaquatic environment by wastewater discharge (Table 4). This meansthat 56.48% of the total TCS consumed in China in the year 2008 directlyflowed into the environment, while the other 43.52% flowed intoWWTPs. After the discharge of TCS into the aquatic environment withwastewater (regardless of whether or not it was treated by WWTPs),the TCS mass distribution in different environmental media showsthat more than 50% was in freshwater sediment, 45% in the ocean, 4%in the freshwater, and less than 1% in soil and air (Fig. 2). Thus, the de-velopment of environmental TCS control measures should pay more

Table 2Mass inventory of TCS in water and sediment in mainland China in 2008.

River basin Medium TCS concentrationng L−1 in water,ng g−1 in sediment

Population×106 person

Mass inventorykg

Simulated Measured Simulated area Entire basin Simulated1 Measured2 Entire basin3

Liao River Water 30.20 27.40 22.96 119.32 136 123 641Sediment 98.50 5.46 2234 124 644

Hai River Water 2.36 6.89 11.76 138.00 30 88 1028Sediment 13.50 2.31 862 148 1731

Yellow River Water 12.10 8.70 31.28 110.00 73 52 184Sediment 4.76 1.37 1522 438 1540

Zhujiang River Water 36.20 95.70 10.18 119.92 22 57 676Sediment 131.00 248.00 754 1427 16,815

Yangtze River Water 7.14 16.57a 52.63 453.86 266 617 5323Sediment 67.70 67.70b 7572 7572 65,298

China Water 18.30 46.45c 118.63 1328.02 526 1336 14,958Sediment 50.30 65.54d 12,944 16,866 188,807

Notes: Population in the simulated area of each river basin was adapted from Zhao et al. (2013); population for the entire basin was fromNBSC (2012); TCS concentrations are from Zhaoet al. (2013), except for “a, c, d”which are calculated as shown in “a, d, f” of notes in Table 3, respectively. “b” is simulated value from Zhao et al. (2013). Themeasured TCS concentrationswere noted in note 1 of Table 3.

1 TCS mass inventory by fugacity Level III model simulation (Zhao et al., 2013).2 TCS mass inventory calculated by Eq. (10).3 TCS mass inventory calculated by Eq. (9).

269C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

attention to reducing the amount of TCSdischarged into freshwater sed-iment and coastal waters, due to the much lower proportion of TCSmass present in other environmental media.

3.2.1.2. TCS with wastewater removed by WWTPs. Although TCS has anti-microbial properties, some TCS is treated and degraded by microbialbiodegradation in WWTPs (NICNAS, 2009). The activated sludge treat-ment method has been estimated to remove approximately 95% of theTCS present (Wilson et al., 2009), and this method is currently used in80% of Chinese urbanWWTPs (EPC, 2008). However, the factors affect-ing degradation and sorption of TCS are complex and variable, and in-clude water pH, solar light intensity, river depth, suspended solidsconcentration as well as characteristics such as organic carbon content,particle size distribution and water convection (turbulence) rate, watertemperature (which varies considerably throughout the year and at dif-ferent water depths), and the amount of surface biofilm (proportionalto water surface area to volume ratio) (Sabaliunas et al., 2003). The

Table 3Amount of TCS flowing into the ocean from mainland China in 2008.

River Mean concentration of TCS1 Runoff into the o×108 m3

In waterng L−1

In suspendedparticulate matterng L−1

In sedimentng g−1

Liao River 27.40 7.03 5.46 103Hai River 6.89 1.66 2.31 25Yellow River 8.70 1.24 1.37 141.6Zhujiang River 95.70 18.00 248.00 5342Yangtze River 16.57a 3.31b 67.70c 8530China 46.45d 8.87e 65.54f 22,805g

1 From Zhao et al. (2013), measured in both low-flow and high-flow seasons from Decembesediment discharge, etc.) for 2008 were used.

2 Water Sources Bulletin for river basins in China in 2008.3 MOWR (2009).4 The product of ‘mean concentration of TCS in water’ and ‘runoff into the ocean’.5 The product of ‘mean concentration of TCS in suspended particulate matters’ and ‘runoff in6 The product of ‘mean concentration of TCS in bed sediment’ and ‘sediment discharge’.a Median between the simulated value from Zhao, et al. (2013) and the measured value fromb Calculated value using Eq. (S1) in Supplementary data.c Simulated value from Zhao et al. (2013).d The value of total amount of TCS flowing into the ocean by water divided by total runoff ine The value of total amount of TCS flowing into the ocean by suspended particulate matter df The value of total amount of TCS flowing into the ocean by bed sediment divided by sedimg The sum of runoff into the ocean, water volume flowing into international boundary rivers,h Assuming that sediment discharge is the sediment flowing into the ocean, national sedime

stations of the main rivers in China’ and ‘the ratio of national runoff to runoff of representative

rate of biodegradation of TCS is generally proportional to exposuretime and inversely proportional to TCS concentration; the degradationrate may also depend on the microbial community, whether the com-munity has been acclimatized to TCS and the environmental conditions(NICNAS, 2009). In view of the range of removal efficiencies under dif-ferent treatment configurations (see Table S7, Supplementary data),the average removal efficiency of Chinese WWTPs was assumed to be78% (Table 4). Then, based on Eqs. (4)–(14), TCS removed by WWTPscan be estimated as 147,125 kg (Table 4).

We assumed that 1) all TCS in the untreated wastewater flows intoenvironmental media, 2) the average removal efficiency of ChineseWWTPs is 78%, 3) TCS is evenly distributed in all kinds of wastewater,and 4) the ratio of total treated domestic wastewater to total domesticwastewater discharge in 2008 is the ratio of treated wastewater tototal wastewater, i.e. 36% (Table 4). Assumption 4) is supported by thefact that there were only 1440 WWTPs in China in 2008 (this total in-creased to 2496 WWTPs in 2010); these were located in 93% of cities,

cean2 Sediment discharge3

×107 kgTCS mass flowing into the ocean kg

By water4 By suspendedparticulate matter5

By bedsediment6

109 282 72 615 17 4 0.4

13,000 123 18 1784460 51,123 9616 11,061

13,000 14,134 2827 880165,294h 105,929 20,228 42,794

r 2007 to December 2009; as most were sampled in 2008, statistical data (runoff volume,

to the ocean’.

Zheng et al. (2011).

to the ocean.ivided by total runoff into the ocean.ent discharge.andwater volume flowing across national borders. Data are sourced fromMOWR (2008).nt discharge should be the product of ‘sediment discharge in representative hydrologicalhydrological stations in China’.

Table 4Total TCS discharged by wastewater from the Chinese anthroposphere.

Items Untreated TreatedUrban

Total

Rural Urban

Wastewater discharge, ×108 m3 332 151a 275a 758b

TCS removal rate of WWTPs, % 78c

TCS flowing into aquaticenvironment

Removed

TCS mass, kg 372,785d 147,125e 519,910

Notes:a MOHURD (2012).b WRC (2008).c Median of TCS removal efficiencies of WWTPs around the world (Table S7, Sup-

plementary data).d Based on Tables 2 and 3.e Based on Eqs. (4)–(14).

270 C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

63% of counties, but in only a few villages where nearly half the totalpopulation lives (MOHURD, 2012). Then, TCS removed by WWTPs canbe estimated to be 145,991 kg (36% × 78% × 519,910 kg = 145,991 kg),very close to the total of 147,125 kg given in Table 4. Thus, the estimatedvalue of TCS removed by WWTPs calculated using Eqs. (4)–(14) wouldseem to be acceptable.

After adsorption, biodegradation, and photo-degradation inWWTPs,TCS concentrations decline rapidly (Boyd et al., 2003; Morrall et al.,2004; Sabaliunas et al., 2003; Singer et al., 2002). As 32.29% of Dt isadsorbed onto biosolids (Heidler and Halden, 2008), 47,507 kg TCSshould be adsorbed onto biosolids (Dta), and the remaining 99,618 kgdegraded. As 22.1% of Dta is removed by landfill with leachate collectionand treatment systems (Eq. (19)), there should be 37,008 kg TCSdischarged to the soil environment (i.e. 47,507 kg–10,499 kg) (Fig. 3).

3.2.2. TCS mass with solid waste dischargeIf 372,785 kg of TCS flowed into the aquatic environment, total dis-

charge of TCS by wastewater from the Chinese anthroposphere shouldbe 519,910 kg (Table 4), calculated by the measured TCS concentration(Zhao et al., 2013). Thus, as apparent consumption in China in 2008was660,000 kg, 140,090 kg TCS should have been discharged with solidwaste and volatilization. Because volatilization of TCS is assumed as 0and the TCS removed by incineration and landfill was 24.7% of totalTCS discharged with solid waste (‘Assumption for Eq. (2)’, Supplemen-tary data), the TCS removed by incineration and landfill should be34,602 kg,with the remaining 105,488 kg entering the soil environment(Fig. 3).

3.3. Substance flow analysis for TCS in China in 2008

Based on the above SFA for TCS, the distribution and flux of TCS inthe Chinese anthroposphere from production to discharge and its final

In soil, 68.50 , 0%

In sea, 168951, 45%

In air, 1.20 , 0%

Fig. 2. Proportions of annual TCS discharge in different environmental media i

environmental fate are illustrated in Fig. 3, fromwhich the following re-sults can be deduced.

(1) About 73% of total TCS production in Chinawas exported in 2008,while the imported TCS mass accounted for less than 1% of totalproduction (Fig. 3). This indicates that the proportion of theexported environmental TCS consumption burden (or theimported environmental TCS production burden) was veryhigh, and will probably remain high in the coming years becausethe proportion of total TCS production for export increased from52% in 2003 to 77% in 2006, and has remained above 70% since2006 (Table S4, Supplementary data).

(2) A total of 660,000 kg TCS was consumed in China in 2008(S6CCN, 2013), with about 79% of total consumption beingdischarged in wastewater and 21% in solid waste (Fig. 3). Thisimplies that at most 79% of the total consumption was actuallyused, with up to 21% being wasted during consumption. It is ur-gent to reduce TCS discharge in solid waste, on account of thelonger half-life of TCS in soil than in water and the lower treat-ment rate of solid waste than that of wastewater (The treatmentrate of urban harmless garbage in China increased from 52% in2005 to 80% in 2011 (NBSC, 2006, 2012), but was always lessthan the urban sewage treatment rate, which increased from56% in 2006 to 82% in 2010 (MOHURD, 2012).). To do this, thediscarding of TCS containing products should be reduced and thepackaging of TCS containing products should be improved.

(3) Only 22% of total TCS discharge was removed in the built environ-ment, including degradation and landfill/incineration,while 78%oftotal dischargeflowed into the environment (Fig. 3). This also indi-cates the unsustainability of societal TCS metabolism.

(4) Removal efficiency was about 21% by wastewater and about 25%by solid waste (excluding sludge from WWTPs) (Fig. 3). The lowlevel of TCS removal efficiency in the built environment can be ex-plained as follows. WWTPs are quite efficient in removing TCSfrom wastewater, but mainly due to secondary treatment(HCEC, 2012). Only about 29% of total treating capacity wasprovided with secondary wastewater treatment in 2008 (seeSection 3.2.1.2 above) and very little water treatment occurs inrural areas where half of the population lives.

(5) Although a total of 515,281 kg TCS flowed into the environmentin 2008, its distribution in different environmental media wasvery uneven (very close to 0.0% in the air, 27.7% in the soil, 2.9%in freshwaters, 36.6% in the sediment of freshwaters, and 32.8%in the ocean) (Fig. 3). The half-life for TCS dissolved in deionizedwater when photo-degraded by natural sunlight is only 5 h(Latch et al., 2005), but is commonly more than 40 days in waterand aerobic sediment, 70 days in anaerobic sediment (HCEC,2012) and over 100 days in soil (Lozano et al., 2012; Waria et al.,2011); in addition, the solubility of TCS in water is low (Table S2,Supplementary data). Accordingly, indirect exposure to humans

In sediment, 188807,51%

In continentalwater, 14958, 4%

n mainland China in 2008, excluding TCS discharged by solid waste (kg).

Fig. 3. Substance flow analysis for TCS inmainland China in 2008 (kg). Production, consumption, and export are shown in cube boxes; degradation, landfill and incineration inwide arrowwith dots; others are shown in cylinders. Linear arrows indicate TCS flow direction in the system.Mass data is represented for the 31 provinces in China, not including Taiwan, Hong Kong,and Macao. We assume that process loss rate in production and consumption is 0, and all storage volume of TCS is consumed in the same year.

271C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

via the environment would mainly come from the food chain, in-cluding aquatic and agricultural products (e.g. beans grown onland which had received PPCPs from biosolids or wastewater;Wu et al., 2010). TCS metabolites (methyl-triclosan, MeTCS) existfor amuch longer period in the environment, increasing in concen-tration over time in some sediment (HCEC, 2012) to amuchhigherconcentration than its parent compound at the same sample sites.This results in a higher concentration of MeTCS than TCS in biota,including aquatic organisms (NICNAS, 2009). TCS metabolitesprobably pose a more serious risk than the parent chemical tothe human diet (Dann and Hontela, 2011). More attention shouldbe paid to reducing the proportion of TCSflowing into the environ-ment by recycling and removing in the anthroposphere, ratherthan focusing on removing TCS which has accumulated in theenvironment.

3.4. Exposure potential analysis

3.4.1. Distribution of exposure potential through TCS flow system in 2008According to the three routes of human exposure to TCS (Fig. 1) and

the results of TCS SFA (Fig. 3), the annual exposure potential of theworkers processing products containing TCS was 1,113,689 mg percapita (total TCS production divided by the number of employees inchemical raw material and products manufacturing in 2008; the num-ber of employees sourced from NBSC, 2009). Annual exposure potentialfor the consumerwas 497mgper capita (total TCS consumption dividedby total population in China). In the environment, annual exposure po-tential for residents living in environments where TCS-accumulationoccurs was 388 mg per capita (annual cumulative TCS amount in envi-ronment from the TCS flow systemdivided by total population, excludingthe part of TCS mass degraded by WWTPs and removed by municipalsolid waste management facilities).

Obviously, exposure potential for workers in the production processis much higher than for consumers and residents, and exposure occursmainly during production of TCS for export (see Section 3.3-(1)), indi-cating that the majority of exposure burden during production isimported. Nevertheless, the smaller size of the exposed population inthe production process means that appropriate measures should be ef-fective in reducing the risk to workers (Shih and Ma, 2011). Althoughexposure potential for residents is slightly lower than for consumers,the risk to residents may be higher because TCS can persist in the envi-ronment (Chalew and Halden, 2009; Miller et al., 2008; Ying et al.,2007) and be stored in food and beverages. For example, in one study

most of the bottled water in Guangzhou was found to contain TCS,with concentrations ranging from 0.6 to 9.7 ng L−1 (Li et al., 2010). Lon-ger exposure to TCS is usuallymore harmful to humanhealth. For exam-ple, for human natural killer cells which defend against tumor cells andvirally infected cells by binding to these targets, the binding function de-creased 37 ± 19% for a 24 h exposure of 5 μM TCS, but 67 ± 18% for a48 h exposure (Hurd-Brown et al., 2013). In addition, the exposurepathways for consumers and residents are mainly via dermal contactand oral ingestion, but inhalation is also an important environmentalexposure pathway for residents, for example inhalation of TCS-contaminated household dust (Fan et al., 2010; HCEC, 2012). Basedon the available animal data, TCS does not meet the Approved Criteria(NOHSC, 2004) for classification for acute oral and dermal toxicity, butbased on effects seen after a single exposure in a repeat dose inhalationtoxicity study, TCS does meet the Approved Criteria for classification as‘toxic by inhalation (R23)’.

3.4.2. Trends in environmental exposure potentialBecause TCS is a human-made chemical, changes in its environmen-

tal exposure potential are mainly determined by consumption.According to the Pearson correlation coefficient rank of different fac-

tors affecting TCS consumption (Table S1, Supplementary data), totalpopulation is the strongest factor affecting TCS consumption. Thus, pre-diction of TCS consumption change can be based on population, i.e.annual rate of increase of TCS consumption per capita, which can be cal-culated by Eq. (15) and validated by Eq. (16).

Results show that per capita TCS consumption in China increased209% between 2003 and 2012, from 193 to 598 mg TCS yr−1 (Table S8,Supplementary data). The fluctuation range of the average annual in-crease rate for years (Ra) based on Eq. (15) was lower than the year-on-year increase rate (Ry) based on Eq. (16), but both Ra andRy are consistentin the changing trendwhich decreased from nearly 16% in 2003 to nearly6% in 2012, and both Ra and Ry from 2003 to 2006 were more than twicethat from2007 to 2012 (Table S8, Supplementary data). Thus, performinga cross-check calculation of per capita consumption increase rates sug-gests that the results are reasonable; it can be assumed that the averageannual increase rate of per capita TCS consumption is 6% from any refer-ence year to the predicted year. In that case, taking 2012 as the base year,per capita TCS consumption will increase from 598 mg TCS yr−1 in 2012to 713mg TCS yr−1 in 2015 and 957mg TCS yr−1 in 2020 (Table S8, Sup-plementary data). Based on Eq. (17) and population in predicted years(Men and Zeng, 2004), total TCS consumption would be 962 t in 2015and 1301 t in 2020; the results are acceptable according to the goodness

Fig. 4. Projection of TCS consumption in 2015 and 2020. (Data sources and the calculationcan be seen in Table S8, Supplementary data.)

Table 6Annual environmental exposure potential of TCS in environmental media.

Air Freshwater Sediment Soil Ocean

A 2008 1.20 14,958 188,807 142,565 168,9512015 1.75 21,801 275,181 207,784 246,2422020 2.37 29,489 372,219 281,055 333,074

B 2008 0.125 0.007 0.289 6.658 0.0742015 0.182 0.010 0.421 9.704 0.1082020 0.246 0.013 0.570 13.126 0.146

Area Runoff Sediment Area RunoffC 9,600,000a 22,805b 65,294c 21,412.36d 22,805b

Notes: A is annual cumulative TCSmass in environmental media, kg. A for 2008 is sourcedfrom Fig. 3; A for 2015 and 2020 was calculated as Eq. (S2); B is annual cumulative TCSconcentration in environmental media, mg km−2 in air, mg m−2 in soil, mg m−3 inwater and ocean, mg kg−1 in sediment; B = A/C, excluding unit conversion in thisequation.

a China's land area (NBSC, 2012), km2.b Runoff into the ocean (see Table 3), ×108 m3.c Sediment discharge (see Table 3), ×107 kg.d Sewage irrigation area, km2, calculated based on effective irrigation area (NBSC, 2009)

and the rate of sewage irrigation area to effective irrigation area, 7.3% (Zhao, 2010).

272 C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

offit of themodel (R2= 0.97) (Fig. 4). This predicted increase in TCS con-sumptionwould result not only in increasing pressure on the biotic envi-ronment, especially susceptible aquatic organisms and algae, but also inincreased potential human exposure.

As per capita TCS consumption continues to increase, annualenvironmental exposure potential will rapidly increase from388 mg capita−1 yr−1 in 2008 to 557 mg capita−1 yr−1 (i.e.1526 μg capita−1 d−1) in 2015, and 747 mg capita−1 yr−1

(2047 μg capita−1 d−1) in 2020 (Table 5).

y = 0.0019x2 - 0.9139x + 755.53

R2 = 0.9478

y = 8E-05x2 + 0.017x + 1289

R2 = 0.9534

y = 2E-05x2 - 0.0036x + 41.087

R2 = 0.9585

600

800

1000

1200

1400

1600

opul

atio

n ×

106

pers

onol

d co

nsum

ptio

n in

dex

-TC

S U

SD c

apita

-1

20.0

30.0

40.0

50.0

60.0

aniz

atio

n le

vel %

3.4.3. Distribution of environmental exposure potential in differentenvironmental media

Discharged TCS is not evenly distributed across different media.Based on the results presented in Section 3.3-(5) and Fig. 3, annual en-vironmental exposure potential was from freshwater sediment, ocean,soil, freshwaters and air in descending order of size (Table 6). The possibleexposure pathway may be agricultural products, especially aquatic prod-ucts which may accumulate TCS from freshwater sediment and freshwa-ter, the ocean, and or even directly fromdrinkingwater. For example, TCSwas detected infive out of six drinkingwater plants in Guangzhou,wherethe concentration in tap water was up to 14.5 ng L−1, and daily adultintake from drinking 2 L of tap water was estimated to be 10 ng day−1

(Li et al., 2010). In summary, the uneven distribution of TCS dischargeacross environmental media around the country worsens the TCS risk.

Associated with the increase in annual cumulative mass of TCS inthe environment from 2008 to 2020, the annual environmental ex-posure potential will accordingly increase to 0.246 mg km−2 in theair, 0.013 mg m−3 in freshwaters, 0.570 mg kg−1 in the sedimentof freshwaters, 13.126 mg m−2 in sewage irrigation areas, and0.146 mg m−3 with runoff flowing into the ocean (B, Table 6). Thismeans that TCS environmental exposure potential will rapidly increasein each media, especially in soil and the aquatic environments.

Table 5Estimation of annual TCS environmental exposure potential in China.

Cx, kg Dsr, kg Dtd, kg Dta, kg Dtar, kg Pt, ×104

personsEEPx, mgcapita−1

2008 660,000 34,602 99,618 47,507 10,499 132,802a 3882015 962,000 50,435 145,201 69,245 15,303 134,831b 5572020 1,301,000 68,208 196,369 93,646 20,696 135,989b 747

Notes: Cx is apparent consumption from Table S8 (Supplementary data); Dsr, Dtd, Dta canbe seen in Eq. (19); Dtar is the degraded part of Dta (see Section 3.2.1.2); Pt is the total pop-ulation in a given year; EEPx is annual TCS environmental exposure potential based onEqs. (19) and (20).

a From NBSC (2009).b Based on Men and Zeng (2004).

3.4.4. CountermeasuresBased on the literature (Katz et al., 2013; NICNAS, 2009; Zhao et al.,

2013), results of Pearson correlation analysis (Table S1, Supplementarydata) and the goodness of fit (Fig. 5), the most relevant factors affectingTCS consumption were urbanization level, total population, per capitaannual urban household consumption expenditure on TCS-relatedproducts (HCE-TCS), and the household consumption index. Thus,these factors should be primarily responsible for the rising magnitudesof TCS flows related to consumption and the rapidly increasing dis-charge of TCS into the environment, and are themost important factorsfor policy making to reduce TCS inventories in the environment.

(1) In order to take full advantage of the capacity to breakdown TCSin both the built and natural environmental systems, it would bebetter that TCS consumption and discharge scatter in small cities.In other words, the development of small cities but not mega-cities should be encouraged by Chinese urbanization policy(CCCPC and SCPRC, 2014). Pollution treatment facilities withhigh TCS removal efficiencies (including both WWTPs and solidwaste incinerators) can be built in small cities or mega-citiesbut are less practical in rural areas due to the high costs of buildingandmanagement, However, compared with small cities, exposure

y = 0.001x2 - 0.5586x + 300.02

R2 = 0.94820

200

400

0 200 400 600 800 1000

Apparent TCS consumption (×103 kg)

Tot

al p

Hou

seh

HC

E

0.0

10.0

Urb

Total population Household consumption indexHCE-TCS Urbanization level

Fig. 5. Most important factors affecting TCS consumption. Urbanization level means therate of urban population to total population; household consumption index is calculatedat constant prices (1978 = 100); HCE-TCS is per capita annual urban household con-sumption expenditure of TCS-related products including clothing (headgear and foot-wear), household daily care products, health care products, and domestic services, USDcapita−1, calculated at 6.3 RMB = US$1.

273C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

potential of people are higher in mega-cities where people live inmore densely populated areas, TCS's consumption and dischargeshould be higher, and its distribution in the environment may bemore uneven, and therefore, environmental TCS will need moretime to degrade due to the persistence of TCS in the environment(Bester, 2005; Morrall et al., 2004; see Table S9 in Supplementarydata).

(2) The population growth rate has already slowed in China (NBSC,2012), but the quality of the population has to be improved to re-duce the use of TCS containing products, e.g. by improving publicaccess to health-related information. Government funded infor-mation campaigns should raise awareness of health risks associat-ed with TCS containing products via newsletters and brochures,mass media and schools.

(3) HCE-TCS depends on population, consumer attitudes and behav-ior, the quality of TCS-related products, investment in TCS indus-try, and so on. To improve the environmental performance ofHCE-TCS, it is necessary to promote positive changes in consumerawareness, beliefs, attitudes, and finally behavior by emphasizingthe value of environmentally friendly consumption, which is pri-marily influenced by personal norms (Minton and Rose, 1997).This can be achieved by local community and general education,while improved quality of TCS-related products and reduced in-vestment in TCS industry could be achieved by suitable economicmeasures, e.g. a sales tax.

(4) The household consumption index ismainly affected by consumerintention and per capita disposable income. To improve the envi-ronmental benefit of the household consumption index, environ-mentally friendly consumer intentions should be established byeducation and mass media encouraging consumers to pay forquality rather than quantity of TCS containing products. In addi-tion, consumer spending on environmentally unfriendly productsshould be discouraged.

3.5. Qualitative uncertainty evaluation

Due topossible statistical inaccuracy and limited information availablefor TCS flows in China, some data used in themass balance calculation arebased on assumptions and/or the collective judgment of the authors; theassessment results should be regarded as an approximate image based onthe best data currently available. For example, the measured TCS meanconcentrations used in Section 3.2 were recorded from December 2007to December 2009 (see Supplementary materials in Zhao et al., 2013),but most were sampled in 2008. This means that other statistical dataused in the mass balance calculation were taken from 2008, and maynot completely match TCS concentrations measured in other years. Also,although sampling sites for the mean concentrations included locationsin five of the largest rivers in China, densely populated areas of China,both sewage impacted streams and larger rivers and both the down-stream and the upstream of effluent discharge points, and although thesampling months (November, December, May, July, and September)covered both low- and high-flow conditions for the sampled rivers, thesampling could not consider all TCS distribution factors and did notcover all TCS discharge points or discharge months. Because the half-lifeof TCS is 40–56 days in water, 40–56 days in aerobic sediment andmore than 70 days in anaerobic sediment (HCEC, 2012), we assumedthat all the measured TCS mass in the five rivers was discharged in2008, but a small part of themeasuredmass could be from the accumula-tion of trace amounts of undegraded TCS discharged in previous years.

All data in Section 3 were validated by mass balancing (Fig. 3), andtend to confirm our calculations for some missing data.

4. Conclusion

China has experienced a rapid increase in TCS production and con-sumption during the past decade. Results of TCS SFA in 2008 showed

that 1) there was high environmental burden shift among China andother countries; 2) there was great potential to reduce TCS dischargein solid waste; 3) the majority of consumed TCS was discharged inwastewater, then mainly flowed into freshwater sediment, ocean, andsoil, where TCS can accumulate in aquatic and agricultural productsand pose a potentially higher dietary risk than the risk resulting frombrief exposure using TCS containing products; and 4) societal TCS me-tabolism in China is unsustainable, as shown by the fact that 78% ofTCS consumed in 2008 was discharged into the natural environment.

Due to continuing socio-economic development, TCS consumptionwill increase from 660,000 kg in 2008 to about 1,301,000 kg in 2020.Accordingly, the annual environmental exposure potential will increase93% by 2020, compared to 2008. To curb growing environmental expo-sure potential from TCS, appropriate countermeasureswould include 1)encouraging the development of small cities; 2) enhanced public recog-nition that TCS containing products are not essential for a healthy life; 3)promoting positive changes in consumer awareness, beliefs, attitudes,and then behavior; 4) improving the environmental performances ofTCS-related products and industry by effective socio-economicmeasures;and 5) increasing the environmental benefit of the household consump-tion index.

This study supports the view that SFA integrated with exposure po-tential analysis is a useful approach to 1) illustrate the distribution andflux of a pollutant in a system, 2) discover the main pathways of envi-ronmental exposure potential by determining the percentages of thepollutant distributed in environmental media, and 3) shed light on op-tions tomitigate negative environmental consequences of the pollutant.However, detailed polices to reduce health riskswill requiremore infor-mation about the bioaccumulation efficiencies of the pollutant betweendifferent trophic levels in the food chain, and this will require more sys-tematic research on losses, stocks and fluxes from production to expo-sure. In addition, while mass balancing can be used to estimatemissing data, more accurate results will partly depend on improved ac-curacy of available data, especially in developing countries where sys-tematic scientific data is often unavailable.

Acknowledgments

This research was financially supported by the “CAS/SAFEA Interna-tional Partnership Program for Creative Research Teams” (KZCX2-YW-T08), Science and Technology Innovation and Collaboration TeamProject of the Chinese Academy of Sciences, Science and TechnologyPlanning Project of Xiamen, China (3502Z20120012), and “Educationor research projects for young and middle-aged teachers” funded bythe Fujian Provincial Department of Education (JA13272). We wouldlike to thankProf. YongguanZhu for his valuable suggestions, Dr. JonathanVause for improving English of themanuscript, and the editors and anon-ymous reviewers for their comments and suggested improvements to thepaper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at http://dx.doi.org/10.1016/j.scitotenv.2014.08.032.

References

Allmyr M, Adolfsson-Erici M, McLachlan MS, Sandborgh-Englund G. Triclosan in plasmaand milk from Swedish nursing mothers and their exposure via personal care prod-ucts. Sci Total Environ 2006;372(1):87–93.

APUA (The Alliance for the Prudent Use of Antibiotics). TCS — white paper. 75 KneelandStreet, 2nd Floor, Boston, MA 02111; 2011 [Available at: www.apua.org].

Asari Misuzu, Fukui Kazuki, Sakai Shin-ichi. Life-cycle flow of mercury and recycling sce-nario of fluorescent lamps in Japan. Sci Total Environ 2008;393:1–10.

Bester K. Fate of triclosan and triclosan-methyl in sewage treatment plants and surfacewaters. Arch Environ Contam Toxicol 2005;49:9–17.

Binelli A, Cogni D, Parolini M, Riva C, Provini A. In vivo experiments for the evaluation ofgenotoxic and cytotoxic effects of triclosan in Zebra mussel hemocytes. Aquat Toxicol2009;91(3):238–44.

274 C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

Boyd GR, Reemtsma H, Grimm DA, Mitra S. Pharmaceuticals and personal care products(PPCPs) in surface and treated waters of Louisiana, USA and Ontario, Canada. SciTotal Environ 2003;311(1–3):135–49.

Brunner PH, Rechberger H. Practical handbook of material flow analysis. Florida: LewisPublishers; 2004.

Canesi L, Ciacci C, Lorusso LC, Betti M, Gallo G, Pojana G, et al. Effects of triclosan onMytilus galloprovincialis hemocyte function and digestive gland enzyme activities:possible modes of action on non target organisms. Comp Biochem Physiol C ToxicolPharmacol 2007;145(3):464–72.

CCCPC (The 18th Central Committee of the Communist Party of China) and SCPRC(The State Council of the People's Republic of China). National new urbanizationplan (2014–2020); March 16, 2014 [Available at: www.GOV.cn].

Chalew TEA, Halden RU. Environmental exposure of aquatic and terrestrial biota to triclo-san and triclocarban. J Am Water Works Assoc 2009;45(1):4–13.

Chèvre N, Coutu S, Margot J, Wynn HK, Bader HP, Scheidegger R, et al. Substance flowanalysis as a tool for mitigating the impact of pharmaceuticals on the aquatic system.Water Res 2013;47(9):2995–3005.

Chu SG, Metcalfe CD. Simultaneous determination of triclocarban and triclosan in munic-ipal biosolids by liquid chromatography tandem mass spectrometry. J Chromatogr A2007;1164:212–8.

Clarke BO, Smith SR. Review of ‘emerging’ organic contaminants in biosolids and assess-ment of international research priorities for the agricultural use of biosolids. EnvironInt 2011;37:226–47.

Crofton KM, Paul KB, Devito MJ, Hedge JM. Short-term in vivo exposure to the water con-taminant triclosan: evidence for disruption of thyroxine. Environ Toxicol Pharmacol2007;24(2):194–7.

Dann AB, Hontela A. Triclosan: environmental exposure, toxicity and mechanisms of ac-tion. J Appl Toxicol 2011;31(4):285–311.

Davis EF, Klosterhaus SL, Stapleton HM.Measurement of flame retardants and triclosan inmunicipal sewage sludge and biosolids. Environ Int 2012;40:1–7.

Dokianakis SN, Kornaros ME, Lyberatos G. On the effect of pharmaceuticals on bacterialnitrite oxidation. Water Sci Technol 2004;50(5):341–6.

Earnshaw MR, Jones KC, Sweetman AJ. Estimating European historical production, con-sumption and atmospheric emissions of decabromodiphenyl ether. Sci Total Environ2013;447:133–42.

EPC (Ministry of Environmental Protection of the People's Republic of China). Preparationinstructions for waste water treatment project technical specification of anaerobic/anoxic/oxic activated sludge process (exposure draft); 2008. p. 1–19 [Available at:http://www.zhb.gov.cn/info/bgw/bbgth/200808/W020080818469754345998.pdf].

Eriksson E, Andersen HR, Ledin A. Substance flow analysis of parabens in Denmarkcomplemented with a survey of presence and frequency in various commodities. JHazard Mater 2008;156:240–59.

European Communities. Economy-wide material flow accounts and derived indicators: amethodology guide. Luxembourg: Office for Official Publications of the EuropeanCommunities; 2001.

Fan X, Kubwabo C, Rasmussen P, Jones-Otazo H. Simultaneous quantitation of parabens,triclosan, and methyl triclosan in indoor house dust using solid phase extractionand gas chromatography–mass spectrometry. J Environ Monit 2010;12:1891–7.

Fang JL, Stingley RL, Beland FA, HarroukW, Lumpkins DL, Howard P. Occurrence, efficacy,metabolism, and toxicity of triclosan. J Environ Sci Health C Environ CarcinogEcotoxicol Rev 2010;28:147–71.

Gautam P, Carsella JS, Kinney CA. Presence and transport of the antimicrobialstriclocarban and triclosan in a wastewater-dominated stream and freshwaterenvironment. Water Res 2014;48:247–56.

Geens T, Neels H, Covaci A. Distribution of bisphenol-A, triclosan and n-nonylphenol inhuman adipose tissue, liver and brain. Chemosphere 2012;87(7):796–802.

Glaser A. The ubiquitous triclosan: a common antibacterial agent exposed. Pestic You2004;24(3):12–7.

Halden RU, Paull DH. Co-occurrence of triclocarban and triclosan in US water resources.Environ Sci Technol 2005;39(6):1420–6.

HCEC (Health Canada, Environment Canada). Preliminary assessment of triclosan (chem-ical abstracts service registry number 3380-34-5). Date modified: 2013-06-20, 10Wellington, 23rd Floor, Gatineau QC, K1A 0H3, Canada: Environment Canada, InquiryCentre; 2012 [Available at: http://www.ec.gc.ca/ese-ees/default.asp?lang=En&n=6EF68BEC-1#archived].

Heidler J, Halden RU. Meta-analysis of mass balances examining chemical fate duringwastewater treatment. Environ Sci Technol 2008;42(17):6324–32.

Herva M, Álvarez A, Roca E. Combined application of energy and material flow analysisand ecological footprint for the environmental evaluation of a tailoring factory. J Haz-ard Mater 2012;237–238:231–9.

HRCMWR (Hai River Commission of Ministry of Water Resources). Introduction of inte-grated plan of water resources for Hai river.Water information network; 2012 [Avail-able at: http://www.hwcc.gov.cn/pub2011/hwcc/wwgj/HWCCzwgk/zfxxgk/hwslgh/lygh/201203/t20120312_341372.htm].

HRCMWR (Huang River Commission ofMinistry ofWater Resources). Outline of integrat-ed plan for Huang river (2012–2030). 2013. [Available at: http://www.yellowriver.gov.cn/zwzc/lygh/zhgh/201303/t20130321_129411.html].

HuaWY, Bennett ER, Letcher RJ. Triclosan inwaste and surfacewaters from the upper De-troit River by liquid chromatography-electrospray-tandem quadrupole mass spec-trometry. Environ Int 2005;31:621–30.

Huang CL, Vause J, Ma HW, Yu CP. Using material/substance flow analysis to support sus-tainable development assessment: a literature review and outlook. Resour ConservRecycl 2012;68:104–16.

Huang CL, Vause J, Ma HW, Yu CP. Urban water metabolism efficiency assessment:integrated analysis of available and virtual water. Sci Total Environ 2013;452–453:19–27.

Huang CL, Vause J, Ma HW, Li Y, Yu CP. Substance flow analysis for nickel in mainland Chinain 2009. J Clean Prod 2014. http://dx.doi.org/10.1016/j.jclepro.2013.12.079. [Availableat: http://www.sciencedirect.com/science/article/pii/S0959652614000043].

Hurd-Brown T, Udoji F, Martin T, Whalen MM. Effects of DDT and triclosan on tumor-cellbinding capacity and cell-surface protein expression of human natural killer cells. JAppl Toxicol 2013;33:495–502.

JamesMO, Li WJ, Summerlot DP, Rowland-Faux L,Wood CE. Triclosan is a potent inhibitorof estradiol and estrone sulfonation in sheep placenta. Environ Int 2010;36:942–9.

Katz DR, Cantwell MG, Sullivan JC, Perron MM, Burgess RM, Ho KT, et al. Factors regulat-ing the accumulation and spatial distribution of the emerging contaminant triclosanin the sediments of an urbanized estuary: Greenwich Bay, Rhode Island, USA. SciTotal Environ 2013;443:123–33.

Kawanai T. Triclosan, an environmental pollutant from health care products, evokescharybdotoxin-sensitive hyperpolarization in rat thymocytes. Environ ToxicolPharmacol 2011;32:417–22.

Langdon KA, Warne MSJ, Smernik RJ, Shareef A, Kookana RS. Degradation of 4-nonylphenol,4-t-octylphenol, bisphenol A and triclosan following biosolids addition to soil under lab-oratory conditions. Chemosphere 2011;84:1556–62.

Latch D, Packer J, Stender B, Van Overbeke J, Arnold W, McNeill K. Aqueous photochem-istry of triclosan: formation of 2,4-dichlorophenol, 2,8-dichlorodibenzo p-dioxin, andoligomerisation products. Environ Toxicol Chem 2005;24(3):517–25.

Levy CW, Roujeinikova A, Sedelnikova S, Baker PJ, Stuitje AR, Slabas AR, et al. Molecularbasis of triclosan activity. Nature 1999;398:383–4.

Li X, Ying GG, Su HC, Yang XB, Wang L. Simultaneous determination and assessment of 4-nonylphenol, bisphenol A and triclosan in tap water, bottled water and baby bottles.Environ Int 2010;36(6):557–62.

Li X, Ying GG, Zhao JL, Chen ZF, Lai HJ, Su HC. 4-Nonylphenol, bisphenol-A and triclosanlevels in human urine of children and students in China, and the effects of drinkingthese bottled materials on the levels. Environ Int 2013;52:81–6.

Lin DS, Zhou QX, Xie XJ, Liu Y. Potential biochemical and genetic toxicity of triclosan as anemerging pollutant on earthworms (E. fetida). Chemosphere 2010;81:1328–33.

Lindstrom A, Buerge IJ, Poiger T, Bergqvist PA, Muller MD, Buser HR. Occurrence and en-vironmental behavior of the bactericide triclosan and its methyl derivative in surfacewaters and in wastewater. Environ Sci Technol 2002;36(11):2322–9.

Long YY, Feng YJ, Cai SS, DingWX, Shen DS. Flow analysis of heavy metals in a pilot-scaleincinerator for residues fromwaste electrical and electronic equipment dismantling. JHazard Mater 2013;261:427–34.

Lores M, Llompart M, Sanchez-Prado L, Garcia-Jares C, Cela R. Confirmation of the forma-tion of dichlorodibenzo-p-dioxin in the photodegradation of triclosan by photo-SPME. Anal Bioanal Chem 2005;381:1294–8.

Lozano N, Rice CP, Ramirez M, Torrents A. Fate of triclosan and methyltriclosan in soilfrom biosolids application. Environ Pollut 2012;160:103–8.

Ma HW, Hung ML, Chen PC. A systemic health risk assessment for the chromium cycle inTaiwan. Environ Int 2007;33:206–18.

Ma HM, Zheng LJ, Li YH, Pan SX, Hu JJ, Yu ZQ, et al. Triclosan reduces the levels of globalDNA methylation in HepG2 cells. Chemosphere 2013;90(3):1023–9.

Marlatt VL, Veldhoen N, Lo BP, Bakker D, Rehaume V, Vallée K, et al. Triclosan exposurealters postembryonic development in a Pacific tree frog (Pseudacris regilla) Amphib-ian Metamorphosis Assay (TREEMA). Aquat Toxicol 2013;126:85–94.

Matozzo V, Formenti A, Donadello G, MarinMG. Amulti-biomarker approach to assess ef-fects of triclosan in the clam Ruditapes philippinarum. Mar Environ Res 2012;74:40–6.

McMurry LM, Oethinger M, Levy SB. Triclosan targets lipid synthesis. Nature 1998;394:531–2.

Men KP, Zeng W. Projection of China's population growth in the next 50 years. J QuantTechnol Econ 2004;3:12–7. [In Chinese].

Miller TR, Heidler J, Chillrud SN, Delaquil A, Ritchie JC, Mihalic JN, et al. Fate of triclosanand evidence for reductive dechlorination of triclocarban in estuarine sediments. En-viron Sci Technol 2008;42(12):4570–6.

Minton AP, Rose RL. The effects of environmental concern on environmentally friendlyconsumer behavior: an exploratory study. J Bus Res 1997;40(1):37–48.

MOC (Ministry of Construction of the People's Republic of China). China urban construc-tion statistics yearbook. Beijing, China: China Architecture and Building Press; 2006[in Chinese].

MOHURD (Ministry of Housing and Urban–Rural Development of the People's Republicof China). Bulletin of discharge and treatment of urban wastewater in Chinafrom 2006 to 2010. 2012. [Available at: www.mohurd.gov.cn/xytj/tjzljsxytjgb/w02012051519761830152701250.docx].

Morrall D, McAvoy D, Schatowitz B, Inauen J, Jacob M, Hauk A, et al. A field study of TCSloss rates in river water (Cibolo Creek, TX). Chemosphere 2004;54:653–60.

MOWR (TheMinistry ofWater Resources of the People's Republic of China). Chinawater re-sources bulletin; 2008 [Available at: http://www.mwr.gov.cn/zwzc/hygb/szygb/qgszygb/201001/t20100119_171051.html].

MOWR (TheMinistry of Water Resources of the People's Republic of China). China's riversediment communique 2008. Beijing, China: China Water & Power Press; 2009.

NBSC (National Bureau of Statistics of China). China statistical yearbook (2004–2012),2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, and 2012 Beijing: China StatisticsPress; 2004 [Available at: http://www.stats.gov.cn/tjsj/ndsj/].

Neumegen RA, Fernández-Alba AR, Chisti Y. Toxicities of triclosan, phenol, and coppersulfate in activated sludge. Environ Toxicol 2005;20(2):160–4.

NICNAS (National Industrial Chemicals Notification and Assessment Scheme), Departmentof Health and Ageing, Australian Government. Triclosan (priority existing chemical as-sessment report no. 30). GPO Box 58, Sydney NSW 2001, Australia: National IndustrialChemicals Notification and Assessment Scheme; 2009. p. 1–485 [January].

NOHSC (National Occupational Health and Safety Commission). Approved criteria forclassifying hazardous substances [NOHSC:1008 (2004)] 3rd ed. ; 2004 [Available at:http://www.safeworkaustralia.gov.au/sites/SWA/about/Publications/Documents/

275C.-L. Huang et al. / Science of the Total Environment 499 (2014) 265–275

258/ApprovedCriteria_Classifying_Hazardous_Substances_NOHSC1008-2004_PDF.pdf].

Oguchi M, Sakanakura H, Terazono A. Toxic metals in WEEE: characterization and sub-stance flow analysis in waste treatment process. Sci Total Environ 2013;463–464:1124–32.

Ono K. Past and future cadmium emissions from municipal solid-waste incinerators inJapan for the assessment of cadmium control policy. J HazardMater 2013;262:741–7.

Phillips H. Too much triclosan? Nature News; 1998. http://dx.doi.org/10.1038/news980813-1 [Available at: http://www.nature.com/news/1998/980813/full/news980813-1.html].

Pirard C, Sagot C, Deville M, Dubois N, Charlier C. Urinary levels of bisphenol A, triclosanand 4-nonylphenol in a general Belgian population. Environ Int 2012;48:78–83.

Price OR, Williams RJ, van Egmond R, Wilkinson MJ, Whelan MJ. Predicting accurate andecologically relevant regional scale concentrations of triclosan in rivers for use inhigher-tier aquatic risk assessments. Environ Int 2010;36:521–6.

Reiss R, Lewis G, Griffin J. An ecological risk assessment for triclosan in the terrestrial en-vironment. Environ Toxicol Chem 2009;28:1546–56.

S6CCN (Shanghai 6CHem Chemical Consultation Network). Report of triclosan tech-nology and market in China; 2013 [Available at: http://www.6chem.com.cn/En-glish/default.asp].

Sabaliunas D, Webb SF, Hauk A, Jacob M, Eckhoff WS. Environmental fate of triclosan inthe River Aire Basin, UK. Water Res 2003;37(13):3145–54.

SCCP (Scientific Committee on Consumer Products). Opinion on triclosan (COLIPA no.P32). European Commission, Health & Consumer Protection Directorate-General, Sci-entific Committee on Consumer Products; 2009.

SCCS (Scientific Committee on Consumer Safety). Opinion on triclosan antimicrobialresistance; 2010 [Available at: http://ec.europa.eu/health/scientific_committees/con-sumer_safety/docs/sccs_o_023.pdf, accessed 15. 12. 2013].

Sendra C, Gabarrell X, Vicent T. Material flow analysis adapted to an industrial area. JClean Prod 2007;15:1706–15.

Shih HC, Ma HW. Life cycle risk assessment of bottom ash reuse. J Hazard Mater 2011;190:308–16.

Singer H, Muller S, Tixier C, Pillonel L. Triclosan: occurrence and fate of a widely used biocidein the aquatic environment: field measurements in wastewater treatment plants, sur-face waters, and lake sediments. Environ Sci Technol 2002;36(23):4998–5004.

Srinivasan V, Seto KC, Emerson R, Gorelick SM. The impact of urbanization on watervulnerability: a coupled human–environment system approach for Chennai, India.Global Environ Chang 2013;23(1):229–39.

Stannard CA, Aspinall RJ. Meeting the challenges of modelling coupled human–environ-mental systems — GLP Nodal Office of Integration and Modelling. Procedia EnvironSci 2011;6:194–8.

Tamura I, Kanbara Y, Saito M, Horimoto K, Satoh M, Yamamoto H, et al. Triclosan, an an-tibacterial agent, increases intracellular Zn2+ concentration in rat thymocytes: its re-lation to oxidative stress. Chemosphere 2012;86:70–5.

Waria M, O'Connor GA, Toor GS. Biodegradation of triclosan in biosolids-amended soils.Environ Toxicol Chem 2011;30:2488–96.

Wilson B, Chen RF, Cantwell M, Gontz A, Zhu J, Olsen CR. The partitioning of triclosan be-tween aqueous and particulate bound phases in the Hudson River Estuary. Mar PollutBull 2009;59:207–12.

WRC (The Ministry of Water Resources of the People's Republic of China). China waterresources bulletin; 2008 [Available at: http://www.mwr.gov.cn/zwzc/hygb/szygb/qgszygb/201001/t20100119_171051.html].

Wu C, Spongberg AL, Witter JD, Fang M, Czajkowski KP. Uptake of pharmaceuticalsand personal care products by soybean plants from soils applied with bio-solids and irrigated with contaminated water. Environ Sci Technol 2010;44:6157–61.

Ying GG, Kookana RS. Triclosan in wastewaters and biosolids from Australian wastewatertreatment plants. Environ Int 2007;33:199–205.

Ying GG, Yu XY, Kookana RS. Biological degradation of triclocarban and triclosan in a soilunder aerobic and anaerobic conditions and comparison with environmental fatemodelling. Environ Pollut 2007;150(3):300–5.

Zhang L, Yang JM, Cai ZJ, Yuan ZW. Analysis of copper flows in China from 1975 to 2010.Sci Total Environ 2014;478:80–9.

Zhao QG. The significance of developing high ecological value of agriculture in China. Sci-ence Times, 2010-09-07; 2010 [Available at: http://www.moa.gov.cn/fwllm/jrsn/201009/t20100907_1645244.htm].

Zhao JL, Zhang QQ, Chen F, Wang L, Ying GG, Liu YS, et al. Evaluation of triclosan andtriclocarban at river basin scale using monitoring and modeling tools: implica-tions for controlling of urban domestic sewage discharge. Water Res 2013;47:395–405.

Zheng C, Zhao J, Bao P, Gao J, He J. Dispersive liquid–liquid microextraction based on so-lidification of floating organic droplet followed by high-performance liquid chroma-tography with ultraviolet detection and liquid chromatography–tandem massspectrometry for the determination of TCS and 2,4-dichlorophenol in water samples.J Chromatogr A 2011;1218(25):3830–6.