Chemosphere 83 (2011) 1320–1325

Contents lists available at ScienceDirect

Chemosphere

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

Concentrations of PFOS, PFOA and other perfluorinated alkyl acids inAustralian drinking water

Jack Thompson a,⇑, Geoff Eaglesham b, Jochen Mueller a

a The University of Queensland, National Research Center for Environmental Toxicology (Entox) 39 Kessels Rd., Coopers Plains, QLD. 4108, Australiab Queensland Health and Forensic Scientific Services (QHFSS), Organics Section 39 Kessels Rd., Coopers Plains, QLD. 4108, Australia

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

Article history:Received 10 December 2010Received in revised form 16 March 2011Accepted 1 April 2011Available online 30 April 2011

Keywords:PFOSPFOADrinking waterExposureIntake

0045-6535/$ - see front matter � 2011 Elsevier Ltd. Adoi:10.1016/j.chemosphere.2011.04.017

⇑ Corresponding author.E-mail address: jack.thompson@uq.edu.au (J. Thom

Perfluorinated alkyl acids (PFAAs) are persistent environmental pollutants, found in the serum of humanpopulations internationally. Due to concerns regarding their bioaccumulation, and possible health effects,an understanding of routes of human exposure is necessary. PFAAs are recalcitrant in many water treat-ment processes, making drinking water a potential source of human exposure. This study was conductedwith the aim of assessing the exposure to PFAAs via potable water in Australia. Sixty-two samples ofpotable water, collected from 34 locations across Australia, including capital cities and regional centers.The samples were extracted by solid phase extraction and analyzed via liquid chromatography/tandemmass spectrometry for a range of perfluoroalkyl carboxylates and sulfonates. PFOS and PFOA were themost commonly detected PFAAs, quantifiable in 49% and 44% of all samples respectively. The maximumconcentration in any sample was seen for PFOS with a concentration of 16 ng L�1, second highest maxi-mums were for PFHxS and PFOA at 13 and 9.7 ng L�1. The contribution of drinking water to daily PFOSand PFOA intakes in Australia was estimated. Assuming a daily intake of 1.4 and 0.8 ng kg�1 bw for PFOSand PFOA the average contribution from drinking water was 2–3% with a maximum of 22% and 24%respectively.

� 2011 Elsevier Ltd. All rights reserved.

1. Introduction

Perfluorinated alkyl acids (PFAAs) and their anions such as per-fluorooctanoic acid (PFOA) and perfluorooctane sulfonate (PFOS)are highly persistent anthropogenic chemicals of some scientificconcern. Due to their ubiquity in the environment, adverse effectsin toxicological studies, and currently uncertain human epidemiol-ogy, efforts have been made to limit their production and releaseinto the environment (Giesy and Kannan, 2002; OECD, 2002; Lauet al., 2007; US EPA, 2009a). In Australia no record of PFAA manu-facture exists, but records of import and use are available for recentyears (2005–2008). In 2007, 1350 kg of PFOS was imported intoAustralia, and products containing PFOA at mg L�1 concentrationswere imported in quantities in the order of tens of kilograms (NIC-NAS, 2007, 2008). Despite the lack of historical direct emissionsand the relatively small inventories suggested by the availableimportation data, measurements of human serum in Australia sug-gests a background contamination similar to that observed inter-nationally (Toms et al., 2009). This includes comparison withmore populous countries, and those with a history of PFAA manu-facture (Fromme et al., 2009).

ll rights reserved.

pson).

Both PFOS and PFOA are well absorbed after oral dosing in ani-mals, with studies showing a subsequent rise in serum concentra-tions proportional to cumulative doses (Seacat et al., 2002;Hundley et al., 2006). In studies of communities with drinkingwater impacted by PFOA, drinking water concentrations and con-sumption rate have been shown to be the biggest determinantsof serum concentrations (Emmett et al., 2006). Analysis of drinkingwater internationally has shown low level contamination in anumber of instances, with measurements of PFAAs typically inthe low parts per trillion range reported (Mak et al., 2009;Quinones and Snyder, 2009). Multiple studies have also measuredthe presence of PFAAs in surface waters such as lakes and rivers(Sinclair et al., 2006; Jin et al., 2009; Naille et al., 2010), and alsoin WWTP effluents (Sinclair and Kannan, 2006; Heidler andHalden, 2008). Given potable water treatment often comprises asimilar suite of techniques as used in a WWTP it is expected thatdrinking water may contain appreciable PFAA concentrations,reflective of those in the source waters. For this reason, a numberof regulatory bodies have suggested guideline values for PFAAs indrinking water. The US EPA have set provisional guidelines of500 ng L�1 and 200 ng L�1 for PFOA and PFOS respectively (USEPA, 2009b). These values have been based on animal toxicologicalstudies, assuming a lifetime exposure, and applying various uncer-tainty factors to extrapolate between species. The German

J. Thompson et al. / Chemosphere 83 (2011) 1320–1325 1321

Drinking Water Commission (TWK) have suggested a guideline va-lue of 300 ng L�1 for combined concentrations of both PFOS andPFOA (Trinkwasserkommission, 2007).

In this study we aimed to evaluate the concentrations of PFAAsin drinking water in Australia, and use this data to estimate thecontribution made by drinking water to previously modeled dailyintakes of PFOS and PFOA in the Australian population (Thompsonet al., 2010).

2. Methods materials

2.1. Sample collection

Samples were collected directly from the drinking water taps atthe 34 sampling locations. One liter HDPE bottles were rinsedwith methanol and milliQ water (Millipore, 0.22 lm filtered,18.2 mX cm�1) and sent overnight via courier to volunteers acrossthe country, with an accompanying return consignment note andletter of instruction. The samples were returned from the samplingpoints via overnight courier. The samples were collected in severalbatches between August and November 2010 with extraction nomore than three weeks after receipt. To save on transportationcosts, samples were shipped at room temperature, but between ar-rival and extraction were kept refrigerated at 4 �C. The samplinglocations are listed along with results in Table 1 and marked witha corresponding code in Fig. 1.

2.2. Extraction and analysis

Prior to extraction samples were fortified with 4 ng (50 lL of0.08 ng lL�1) of mass labeled PFCs (obtained from Wellington lab-oratories, all >98%. 13C and/or 18O PFBA, PFHxA, PFOA, PFNA, PFDA,PFUnDA, PFDoDA, PFHxS, PFOS). These were used to quantify thecorresponding native compounds in the sample, and provide de-tails of method recovery. PFAAs without a corresponding mass la-beled standard were quantified using the closest standard in termsof weight and functional group (i.e. PFBS was quantified using thePFHxS internal standard). One liter samples were extracted usingthe solid phase extraction method, developed by Taniyasu et al.,(2005).

No significant alterations were made to the method. The barrelsof 50 mL polypropylene syringes were used as sample reservoirs,rather than using tubing to transport sample between bottle andSPE cartridge. This was done to avoid large wall surfaces foradsorptive losses and possible PFCA contamination from residualsin polytetrafluoroethylene (PTFE) tubing. The sodium acetate buf-fer wash fraction typically part of the method was omitted withthe aim of achieving maximum recovery and due to the high qual-ity of the water sampled. Target analytes were eluted in 4 mL of0.1% NH3 in methanol, concentrated under a gentle stream of highpurity nitrogen to a volume of 500 lL. Samples were made up with500 lL of water to 1 mL to match the composition of the mobilephase used in the liquid chromatography.

Samples were analyzed using a liquid chromatograph (Shima-dzu Prominence, Shimadzu Corp., Kyoto, Japan) coupled to a triplequadrupole mass spectrometer (QTrap 4000, AB/MDS Sciex, Con-cord, Ontario, Canada). The mass spectrum data were acquired inscheduled MRM mode, with negative electrospray ionization.Chromatographic separation was achieved using a gradient elutionand a mobile phase of 90% and 10% methanol in water, both with5 mM of ammonium acetate. The analytical column was a GeminiC18 5 � 2 mm i.d with 3 lm packing (Phenomenex, Torrance, CA),a second column (Altima C18, 150 � 4.6 mm, 5 lm packing) in-stalled between the injector and solvent reservoirs was used as atrap to retard signals for some PFCs intrinsic in the LC system.

The branched isomers were not separated and shoulder peaksappearing in both MRM transitions were integrated together withthe main peak.

3. Results and discussion

3.1. Qa/Qc

Nine blanks consisting of 1 L of milliQ water were analyzedalong with the samples. PFOA was detected in four of the nineblanks at 0.03–0.27 ng L�1, with an average of 0.09 ng L�1. Thisbackground contamination was taken into account when assigningquantitation limits (average blank + 3 ⁄ SD). Recoveries of internalstandards were estimated by comparing peak areas in the calibra-tion standards, having undergone no extraction or concentrationprocesses, with the corresponding peaks in samples. The averagerecoveries ranged from 65% for 13CPFBA to 88% for 13CPFHxS, withsimilar median recoveries ranging from 62% to 98% for the samecompounds (Table 2). Spike tests of native PFAAs at resulted inrecoveries ranging from 62% (PFDS) to 145% (PFOA), with similarrecoveries across the four different concentration levels (Table 2).

Representative chromatograms of those produced from analysisof blanks, low standards, matrix spikes, and samples are given inthe supporting information.

3.2. Concentrations in drinking water

Table 1 gives the concentrations of the various PFAAs detectedin water samples from around Australia. PFOS and PFOA werethe most commonly detected PFAAs. They were quantifiable in49% and 40% of samples respectively, and were typically found atthe highest concentration of the PFAAs. PFHxS was also detectedrelatively frequently, quantified in 27% of samples, and at concen-trations generally less than PFOS but higher than PFOA. Although asuite of PFAAs were analyzed for, perfluoroundecanoate (PFUnDA)and perfluorodecane sulfonate (PFDS) were not detected in anysample. Perfluorononanoate and decanoate (PFNA, PFDA), whilstdetectable in 40% and 15% of samples respectively, had concentra-tions consistently below their respective limits of quantitation.These four compounds are not included in Table 1. All samplesshowed low concentrations of PFAAs, with a greater percentageof non-detects relative to detection. For site comparisons PFAAconcentrations, including those above instrumental detection limit(0.13–0.18 ng L�1) but below quantitation limits, were summed toprovide a picture of the overall contamination of the site. Themajority of locations had

PPFAAs between <1 and 5 ng L�1, but

several sites stood out as exceptions. Three sites around Sydney;Blacktown, Quakers Hill and North Richmond all had relativelyhigh

PPFAAs concentrations, up to 36 ng L�1 in the case of North

Richmond and 12 ng L�1 in Quakers Hill. The Quakers Hill sampleproviding this concentration was one of four collected in the area,with the other three samples having lower

PPFAAs concentrations

around 5 ng L�1. This suggests a degree of temporal variability inthe concentrations present in drinking water, and as such those re-ported here based on discrete grab samples can only be considereda snapshot of the possible exposures. A plausible explanation forthe higher concentrations in the North Richmond and Quakers Hillregions is not available at present, requiring a more comprehensivesurvey of the area. There are both potable- and waste-water treat-ment plants in the North Richmond area, drawing water from theHawkesbury River and outfalling in a smaller tributary. Whetherthese are impacting each other requires details of the locale notknown to the authors at this time. Two samples from regionalNSW, Gundagai and Yass also had relatively high concentrations,with

PPFAAs around 12 ng L�1. There is a temptation to

Table 1Concentrations (ng L�1) of selected PFAAs in drinking water from around Australia (approximate location).

Location (state)a Map code(Fig. 1)

No.samples(n)b

No. propertiessampledc

PFPeA PFHxA PFHpA PFOA PFBS PFHxS PFOS

Alice Springs (NT) A 1 1 nd nd nd <0.50 nd nd ndBathurst (NSW) BH 1 1 nd nd nd 1.02 <0.71 1.24 1.59Blacktown (NSW) BT 1 1 1.82 0.77 <0.73 1.64 0.74 3.47 3.92Bottled mineral water BW 2 1 nd–

<0.66nd–<0.64

–<0.73 <0.5 <0.71 <0.92 0.75–1.45

Cairns (QLD) CN 1 1 nd nd nd <0.50 nd nd 0.77Canberra (ACT) CB 3 2 nd nd–

<0.64nd <0.5–

0.88nd nd–<0.92 nd–1.84

Campbelltown (NSW) CT 2 1 nd nd nd <0.5–3.76

<0.71 <0.92 <0.66

Coopers Plains (QLD) CP 2 1 nd nd nd nd nd nd ndEmu Plains (NSW) E 2 1 nd <0.64–

0.64<0.73 0.68–

0.80<0.71 1.61–1.98 2.37–2.59

Footscray (VIC) F 2 1 nd <0.64 <0.73 0.64–0.65

<0.71 <0.92 1.18–1.21

Gladstone (QLD) GS 3 2 nd nd nd <0.50 nd nd nd-0.86Glen Helen (NT) GH 1 1 <0.66 nd nd <0.50 nd nd ndGlununga (SA)d GU 4 1 nd nd nd nd nd 11.2–14.4 (w/o filter)

nd (w/filter)15.1–15.6 (w/o filter)nd (w/filter)

Goulburn (NSW) GB 1 1 <0.66 nd nd <0.50 nd nd ndGundagai (NSW) GG 1 1 nd 1.49 <0.73 1.62 <0.71 2.67 4.68Kingston (TAS) KS 2 1 nd–

<0.66nd nd <0.50 nd nd–<0.92 nd–<0.66

Kingsborough (TAS) KB 1 1 nd nd nd 0.73 nd nd 1.76Larrakeyah (NT) LK 1 1 nd nd nd 1.16 nd nd ndLithgow (NSW) LG 1 1 <0.66 <0.64 <0.73 <0.50 nd <0.92 0.76Liverpool (NSW) LP 1 1 1.17 <0.64 nd <0.50 <0.71 1.20 1.68Mundaring (WA) MD 2 1 nd nd nd <0.50 nd–

1.01<0.92 <0.66

Menora (WA) MR 2 1 nd <0.64 nd <0.50 <0.71 <0.92 <0.66Marrickville (NSW) MV 2 1 nd nd nd nd–

<0.50nd nd–<0.92 nd

Marruben (WA) MB 1 1 nd 1.79 0.90 2.1 2.52 9.98 8.07North Richmond

(NSW)NR 3 2 2.58–

3.982.85–5.53

1.25–2.54

5.17–9.66

1.48–2.41

4.21–8.24 1.46–3.32

Nightcliff (NT) NC 3 2 nd nd nd–<0.73

<0.50–1.57

nd–<0.71

nd nd

Parkdale (VIC) P 2 1 nd nd nd <0.50 nd nd <0.66Quakers Hill (NSW) Q 4 2 <0.66–

2.24<0.64–0.77

nd–<0.73

0.58–1.00

<0.71–0.84

1.57–3.86 1.94–4.09

Redland Bay (QLD) RB 2 1 nd–<0.66

nd nd nd–<0.50

nd nd nd

Riddells Creek (VIC)e RC 2 2 nd–<0.66

nd–<0.64

nd–<0.73

<0.50 nd–<0.71

nd–<0.92 <0.66–0.86

Rockhampton (QLD) RH 2 1 nd nd nd <0.50 nd <0.92 <0.66Tumut (NSW) T 1 1 nd nd nd nd nd nd ndWagga Wagga (NSW) W 1 1 nd nd nd 0.54 <0.71 <0.92 <0.66Yass (NSW) Y 1 1 4.23 0.93 nd 2.29 0.72 3.34 2.95% of samples with

detection > IDL24 40 27 96 49 62 75

% of samples withdetection > LOR

13 18 7 44 13 27 49

a TAS = Tasmania, VIC = Victoria, NSW = New South Wales, QLD = Queensland, NT = Northern Territory, ACT = Australian Capital Territory, WA = Western Australia,SA = South Australia.

b Total number of samples taken from geographical location.c Number of houses/properties sampled from at a given geographical location.d Four samples taken from one house, two without filter attached, two with filter attached to tap.e One sample taken at residence with water from rainwater tank.

1322 J. Thompson et al. / Chemosphere 83 (2011) 1320–1325

hypothesise a contributor to these concentrations maybe the pres-ence of an airport located in nearby Wagga Wagga, however a sam-ple taken in Wagga Wagga itself showed very low concentrationsof the few PFAAs detected. As with the North Richmond andQuakers Hill samples, interpretation of these concentrations andany attempt at source apportionment requires a greater knowledgeof the specific area than was available at the time of writing. Thesamples collected from Glenunga SA had an average

PPFAAs of

28 ng L�1 due to relatively high concentrations of PFHxS(13 ng L�1) and PFOS (15 ng L�1). However in samples taken from

the same home after fitting a commercially available home carbonfilter to the tap, no PFAAs were detected. One sample collectedfrom Riddell’s Creek, regional Victoria was from a residence obtain-ing their water from a rainwater tank. No PFAAs were detectedabove quantitation limits in this sample, although small peaksfor PFOA, PFDA, PFBS and PFOS were present in the chromato-grams. This suggests the possibility of the presence of thesecompounds in precipitation in this area, and their atmospherictransport, but direct measurement of rainwater samples wouldbe necessary to confirm this.

w

NR

GB

GG

Y

CBT

E

Q, LP, MV, BT

CT

BHLG

GH, A

CN

KS, KB

MR, MD

GURB, CP

GS

F, P

RC

RH

LK, NC

WA

QLDNT

SA

NSW

TAS

VIC

w

NR

GB

GG

Y

CBT

E

Q, LP, MV, BT

CT

BHLG

w

NR

GB

GG

Y

CBT

E

Q, LP, MV, BT

CT

BHLG

GH, A

CN

KS, KB

MR, MD

GURB, CP

GS

F, P

RC

RH

LK, NC

WA

QLDNT

SA

NSW

TAS

VIC

Fig. 1. Map of Australia including expanded insert of New South Wales, with approximate locations of sampling points marked using codes from Table 1.

Table 2Summary of detection limit and recovery data of selected PFAAs.

Analyte PFPeA PFHxA PFHpA PFOA PFNA PFDA PFUnDA PFBS PFHxS PFOS

IDLa (ng L�1) 0.13 0.13 0.15 0.13 0.15 0.12 0.13 0.14 0.18 0.13LOQb (ng L�1) 0.66 0.64 0.73 0.5 0.74 0.61 0.66 0.71 0.92 0.66Recovery (%) 0.2 ng L�1 120 79 108 125 81 92 97 114 88 93Recovery (%) 0.5 ng L�1 101 108 105 113 114 112 106 112 105 140Recovery (%) 2 ng L�1 107 105 103 145 139 98 101 146 148 150Recovery (%) 5 ng L�1 109 113 107 106 92 97 102 120 118 135Corresponding internal

standard

13C4PFBA 13C2PFHxA 13C2PFHxA 13C4PFOA 13C5PFNA 13C4PFDA 13C2PFUnDA 18O2PFHxS 18O2PFHxS 13C4PFOS

Recovery of internalstandardsc

65 ± 30(62)

73 ± 27 (78) 81 ± 24(83)

80 ± 24(81)

79 ± 26(79)

68 ± 24 (67) 88 ± 28 (98) 81 ± 22(87)

a IDL = instrumental detection limit, calculated based on the standard deviation from eight injections of the lowest standard with a signal to noise > 3, 0.1 ppb in vial,equivalent to 0.1 ppt in samples.

b LOQ = limits of quantitation based on the average concentration in the blanks, plus three times the standard deviation (PFOA), or estimated by multiplying IDL by a factorof 5.

c Average from all samples ± standard deviation (median), n = 64. Recovery spikes of natives performed in duplicate, values given are average.

Table 3Summary of international data on PFOA and PFOS in drinking water (ng L�1).

Country PFOS PFOA References

China <0.1–14.8 <0.1–45.9 Jin et al. (2009)USA <1–57 <5–30 Quinones and Snyder (2009)Brazil 0.58–6.7 0.81–2.8 Quinete et al. (2009)Japan <0.1–6.9 2.3–84 Takagi et al. (2008)Germany <10 <10–68 Wilhelm et al. (2010)India <0.03–8.4 <0.005–2 Mak et al. (2009)Australia 0–16 0–9.7 This study

J. Thompson et al. / Chemosphere 83 (2011) 1320–1325 1323

3.3. Comparison with international data and provisional guidelines

The samples analyzed in this study showed PFOS and PFOA atranges of 0–16 ng L�1 and 0–9.7 ng L�1 respectively. A number ofstudies have been conducted on drinking water in other coun-tries, which can provide a comparison for these concentrationsTable 3. In drinking water from 21 Chinese cities, Jin et al.,2009 reported similarly low concentrations for both PFOS andPFOA, with the exception of samples collected from Guangzhou(10.3 ng L�1 PFOS, 11.5 ng L�1 PFOA) and Shenzen (14.8 ng L�1

PFOS, 45.9 ng L�1 PFOA). While those from Guangzhou were

1324 J. Thompson et al. / Chemosphere 83 (2011) 1320–1325

comparable to the higher concentrations measured here, thePFOA concentration in the Shenzen samples was far in excessof any of the Australian samples. Quinones and Snyder (2009)measured concentrations in finished water from seven drinkingwater facilities in the USA. These ranged from <1 to 57 ng L�1

PFOS and <5 to 30 ng L�1 PFOA, again the maximum concentra-tion far exceeding that measured here. However, as with theChinese study, the concentrations at the lower end of the rangewere comparable with the Australian data. Concentrations ofPFOS and PFOA measured in Brazillian drinking water showedranges closer to those reported here, with mean PFOS and PFOAconcentration of 0.58–6.70 ng L�1 and 0.81–2.82 ng L�1 respec-tively (Quinete et al., 2009). In Osaka Japan, analysis of tap watersamples collected at 14 water purification plants gave PFOS andPFOA concentrations in the ranges of <0.1–6.9 ng L�1 and 2.3–84 ng L�1 respectively (Takagi et al., 2008). The PFOS concentra-tions were similar to those in the Australian drinking water, ex-cept there was a much higher frequency of detection in theJapanese samples. The PFOA concentrations were again muchhigher than those observed in this study. One of the most com-prehensive surveys to date has been that done by Mak et al.,2009 which involved analysis of drinking water from China, In-dia, Japan, USA and Canada. The concentrations of PFAAs re-ported varied greatly between locations, with a mean PFOAconcentration in drinking water from Chinese cities of 10 ng L�1,but varying up to 78 ng L�1 in one city. The concentrations insamples from the remaining countries showed a similar predom-inance of PFOS and PFOA in the PFAAs detected, and were gen-erally lower in terms of

Ptotal PFAAs detected than those in

China. Some exceptions were seen though in several samplesfrom India (

Ptotal PFAAs 100 ng L�1) and Japan (

Ptotal PFAAs

40 ng L�1).In some areas, instances of relatively high contamination have

been reported. In the Ruhr area, Germany, concentrations of PFOAin drinking water up to 519 ng L�1 were reported. These werefound to have resulted from the use of a WWTP sludge as a soilamendment in the area (Skutlarek et al., 2006). A later study, fol-lowing intervention by water providers using activated carbon,found a maximum concentration of PFOA of 68 ng L�1, and a med-ian < 10 ng L�1 (Wilhelm et al., 2010). In Wahington, West Virginia,USA, a fluoropolymer manufacturing facility impacted the watersupply of several surrounding regions with PFOA. This area hasbeen studied extensively as part of the C8 project (C8 SciencePanel, 2011). Prior to intervention PFOA was found in the LittleHocking water supply area at concentrations of several 1000 ng L�1

(Dupont, 2006; Emmett et al., 2006). Again water concentrationswere reduced following the introduction of granulated activatedcarbon filtration (Bartell et al., 2010).

Being relatively new compounds under scrutiny from waterproviders, very few guidelines are available for PFAAs in drinkingwater. The guidelines which are available are for PFOS and PFOAspecifically, due to the high frequency of their detection, and thehigher number of studies on their toxicological properties. TheUS EPA sets provisional health based guidelines of 200 ng L�1

for PFOS and 400 ng L�1 for PFOA (US EPA, 2009b). The GermanDrinking Water Commission sets a similar health based guidelinevalue of 300 ng L�1 for combined PFOS and PFOA concentrations(Trinkwasserkommission, 2007). Provisional health based valueshave also recently been proposed for shorter chain PFCs (C4–C7), at values of 300–7000 ng L�1 (Wilhelm et al., 2010),although in some instances these are based on standardized pre-cautionary approaches, rather than compound specific data. Theconcentrations presented here are several orders of magnitudebelow all of these values, suggesting any risks posed by theselevels according to the best of accepted current knowledge arevery low to negligible and considered acceptable.

3.4. Exposure to the Australian population via drinking water

Recently attempts were made using a simple pharmacokineticmodel to estimate intakes based on pooled serum samples col-lected from the general population of south east Queensland, Aus-tralia (Thompson et al., 2010). While there are many assumptionsunderlying the model, the intakes provided are plausible whenviewed alongside similar estimates (Fromme et al., 2009). Takingthe average of these values across the various age groups, the esti-mated daily intakes were 0.8 ng kg�1 b.w d�1 for PFOA and1.4 ng kg�1 b.w d�1 PFOS (Thompson et al., 2010). For an adult withan average body weight of 70 kg, this equates to an estimated totalintake of 56 ng d�1 for PFOA and 96 ng d�1 PFOS.

Assuming a daily water consumption rate of 1.4 L d�1 (US EPA,1997), an average body weight of 70 kg and the concentrations ofPFOS and PFOA found in this study Table 1 we can obtain an esti-mated daily intake from drinking water. The calculated daily in-takes from water ranged from 0 to 11 ng PFOS/day and 0 to13 ng PFOA/day, with a majority of locations providing estimatedintakes between 1 and 3 ng PFOS/day and <1 ng PFOA/day. As apercentage of estimated total daily intake, the daily water intakesare generally quite small, with an average for both PFOS and PFOAof 2.2% of estimated daily intake. In the locations with higher con-centrations, these values ranged up to 22% for PFOS and 24% forPFOA.

The limitations of this analysis is that, for one, we are usingmodeled estimates of total intake, themselves based on a numberof assumptions and potentially containing some error. Secondlythis study has already suggested some temporal variability in PFAAconcentrations in drinking water, introducing another source ofuncertainty. We are also assuming similar total intakes acrossthe entire country, despite the serum data informing the modelused in Thompson et al., 2010 being from a more limited geograph-ical area (predominantly QLD and NSW). The variation observed inthe water concentrations in this study may also be reflected invariations in other sources across the country, such as food, andin this case total estimates would also vary. Despite these uncer-tainties this data gives us a first estimate at the significance ofdrinking water to human exposure to PFAAs in Australia.

Other studies have attempted to estimate total exposure bycombining measured concentrations in various media, with as-sumed patterns of ingestion or exposure. Fromme et al., 2009 usedthis approach to calculate a mean adult daily intake for PFOS andPFOA of 1559.8 pg kg�1 b.w and 2857 pg kg�1 b.w, of which 1.5%and 0.7% were estimated as arising from drinking water. This esti-mate is similar to the 2% averages calculated above. Trudel et al.,2008 did not separate food and water in their estimates, and calcu-lated percentage intakes from both as 30–80% for PFOS, and 10–20% for PFOA. The concentrations in drinking water used in theirestimates were typically an order of magnitude lower than theconcentrations in food, when present.

4. Conclusions

The data presented here provides the first published picture ofPFAA concentrations in Australian drinking water. The concentra-tions were well below the currently available provisionary guide-lines suggested by the US EPA, as well as those set by theGerman Drinking Water Commission and other internationalauthorities. In terms of other studies, the concentrations were onpar with those measured in other countries such as China, USAand Brazil, particularly in comparison with the lower ranges ofconcentrations measured in these international studies. The con-centrations observed in this study suggest drinking water is onlya minor contributor to the daily intake of these chemicals in the

J. Thompson et al. / Chemosphere 83 (2011) 1320–1325 1325

Australian population, although it may be more significant in somelocations.

Acknowledgements

The authors wish to thank Shalona Anuj and Steve Carter ofQHFSS for providing access to the HPLC/MS/MS. We acknowledgefinancial support from an ARC Linkage Grant (LP 0774925). Entoxis jointly funded by The University of Queensland and QueenslandHealth Forensic and Scientific Services. Jack Thompson receives anAPA PhD scholarship. The authors would also like to gratefullyacknowledge all volunteers who provided us with samples.

Appendix A. Supplementary material

Supplementary data associated with this article can be found, inthe online version, at doi:10.1016/j.chemosphere.2011.04.017.

References

Bartell, S.M., Calafat, A.M., Lyu, C., Kato, K., Ryan, P.B., Steenland, K., 2010. Rate ofdecline in serum PFOA concentrations after granulated activated carbonfiltration at two public water systems in Ohio and West Virginia.Environmental Health Perspectives 118, 222–228.

C8 Science Panel, 2011. C8 Study Results – Status Reports. http://www.c8sciencepanel.org/contact.html (retrieved 16.03.11).

Dupont, 2006. 1Q06 Public Water Supply Results, West Virginia. DuPontWashington Works, Washington, WV. Letter from Andrew Hartten, Dupont, toChad Board, West Virginia Department of Environmental Protection, WestVirginia, EPA Docket. U. EPA, EPA (dated 04.05.06).

Emmett, E.A., Shofer, F.S., Zhang, H., Freeman, D., Desai, C., Shaw, L.M., (, 2006.Community exposure to perfluorooctanoate: relationships between serumconcentrations and exposure sources. Journal of Occupational andEnvironmental Medicine 48 (8), 759–770.

Fromme, H., Tittlemier, S.A., Volkel, W., Wilhelm, M., Twardella, D., 2009.Perfluorinated compounds – exposure assessment for the general populationin western countries. International Journal of Hygiene and EnvironmentalHealth 212 (3), 239–270.

Giesy, J.P., Kannan, K., 2002. Perfluorochemical surfactants in the environment.Environmental Science and Technology 36 (7), 146A–152A.

Heidler, J., Halden, R.U., 2008. Meta-analysis of mass balances examining chemicalfate during wastewater treatment. Environmental Science and Technology 42(17), 6324–6332.

Hundley, S.G., Sarrif, A.M., Kennedy Jr., G.L., 2006. Absorption, distribution, andexcretion of ammonium perfluorooctanoate (APFO) after oral administration tovarious species. Drug and Chemical Toxicology 29, 137–145.

Jin, Y.H., Liu, W., Sato, I., Nakayama, S.F., Sasaki, K., Saito, N., Tsuda, S., 2009. PFOSand PFOA in environmental and tap water in China. Chemosphere 77, 605–611.

Lau, C., Anitole, K., Hodes, C., Lai, D., Pfahles-Hutchens, A., Seed, J., 2007.Perfluoroalkyl acids: a review of monitoring and toxicological findings.Toxicological Sciences 99, 366–394.

Mak, Y.L., Taniyasu, S., Yeung, L.W.Y., Lu, G., Jin, L., Yang, Y., Lam, P.K.S., Kannan, K.,Yamashita, N., 2009. Perfluorinated compounds in tap water from China andseveral other countries. Environmental Science & Technology 43 (13), 4824–4829.

Naille, J.E., Khim, J.S., Wang, T., Chen, C., Luo, W., Kwon, B.-O., Park, J., Koh, C.-H.,Jones, P.D., Lu, Y., Giesy, J.P., 2010. Perfluorinated compounds in water,sediment, soil and biota from estuarine and coastal areas of Korea.Environmental Pollution 158, 1237–1244.

NICNAS, 2007. NICNAS Alert No. 6 Perfluorooctanoic Acid (PFOA) &Perfluorocarboxylic Acid (PFCA). Department of Health and Ageing, Sydney,Australian Government, p. 2.

NICNAS, 2008. NICNAS Alert No. 8 Perfluorooctane Sulfonate (PFOS) &Perfluoroalkyl Sulfonate (PFAS). Department of Health and Ageing, Sydney,NICNAS, pp. 1–5.

OECD, 2002. Co-operation on Existing Chemicals: Hazard Assessment ofPerfluorooctane Sulfonate (PFOS) and Its Salts. Co-operation on Existing Titles.OECD, OECD, pp. 1–362.

Quinete, N., Wu, Q., Zhang, T., Yun, H.S., Moreira, I., Kannan, K., 2009. Specificprofiles of perfluorinated compounds in surface and drinking waters andaccumulation in mussels, fish, and dolphins from southeastern Brazil.Chemosphere 77, 863–869.

Quinones, O., Snyder, S.A., 2009. Occurrence of perfluoroalkyl carboxylates andsulfonates in drinking water utilities and related waters from the United States.Environmental Science & Technology 43 (24), 9089–9095.

Seacat, A.M., Thomford, P.J., Hansen, K.J., Olsen, G.W., 2002. Subchronic toxicitystudies on perfluorooctanesulfonate potassium salt in Cynomolgus monkeys.Toxicological Sciences 68, 249–264.

Sinclair, E., Kannan, K., 2006. Mass loading and fate of perfluoroalkyl surfactants inwastewater treatment plants. Environmental Science and Technology 40 (5),1408–1414.

Sinclair, E., Mayack, D.T., Roblee, K., Yamashita, N., Kannan, K., 2006. Occurrence ofperfluoroalkyl surfactants in water, fish, and birds from New York State.Archives of Environmental Contamination and Toxicology 50, 398–410.

Skutlarek, D., Exner, M., Farber, H., 2006. Perfluorinated surfactants in surface anddrinking waters. Environmental Science and Pollution Research 13 (5), 299–307.

Takagi, S., Adachi, F., Miyano, K., Koizumi, Y., Tanaka, H., Mimura, M., Watanabe, I.,Tanabe, S., Kannan, K., 2008. Perfluorosulfonate and perfluorooctanoate in rawand treated tap water from Osaka, Japan. Chemosphere 72, 1409–1412.

Taniyasu, S., Kannan, K., So, M.K., Gulkowska, A., Sinclair, E., Okazawa, T., Yamashita,N., 2005. Analysis of fluorotelomer alcohols, fluorotelomer acids, and short- andlong-chain perfluorinated acids in water and biota. Journal of ChromatographyA 1093, 89–97.

Thompson, J., Lorber, M., Toms, L.M., Kato, K., Calafat, A.M., Mueller, J.F., 2010. Use ofsimple pharmacokinetic modeling to characterize exposure of Australians toperfluorooctanoic acid and perfluorooctane sulfonic acid. EnvironmentInternational 36 (4), 390–397.

Toms, L.-M.L., Calafat, A., Kato, K., Thompson, J., Harden, F., Hobson, P., Sjodin, A.,Mueller, J.F., 2009. Polyfluoroalkyl chemicals (PFCs) in human blood serum fromchildren and adults in Australia. Environmental Science and Technology 43 (11),4194–4199.

Trinkwasserkommission, 2007. Aktuelle gesundheitliche und gewässerhygienischeBewertung perfluorierter Verbindungen (PFC). Bekanntmachung desUmweltbundesamtes. <http://www.umweltdaten.de/wasser/themen/trinkwasserkommission/fazit-hbm-studie-pft.pdf> (retrieved 15.03.11).

Trudel, D., Horowitz, L., Wormuth, M., Scheringer, M., Cousins, I.T., Hungerbuhler, K.,2008. Estimating consumer exposure to PFOS and PFOA. Risk Analysis 28 (2),251–269.

US EPA, 1997. Exposure Factors Handbook, U.S. Environmental Protection Agency,National Center for Environmental Assessment, Washington, D.C.

US EPA, 2009a. 2010/2015 PFOA Stewardship Program. Pollution Prevention & Toxics.<http://www.epa.gov/oppt/pfoa/pubs/stewardship/index.html> (retrieved17.11.10).

US EPA, 2009b. Provisional Health Advisories for Perfluorooctanoic Acid (PFOA) andPerfluorooctane Sulfonate (PFOS). US EPA Water. <http://water.epa.gov/action/advisories/drinking/upload/2009_01_15_criteria_drinking_pha-PFOA_PFOS.pdf> (retrieved 02.02.10).

Wilhelm, M., Bergmann, S., Dieter, H.H., 2010. Occurrence of perfluorinatedcompounds (PFCs) in drinking water of North Rhine–Westphalia, Germanyand new approach to assess drinking water contamination by shorter-chainedC4–C7 PFCs. International Journal of Hygiene and Environmental Health 213,224–232.