DEQ-CFW_00083210

UMTED", ST.A'TES ENV�R�'�N'��l'�.-'.'��IN-�'A�.,
August 31, 2017
SUBJECT: Laboratory PFAS Results for NC DEQ Cape dear Watershed. Sanipling:
Preliminary Non. -Targeted Analysi,
a'y
FROMa Timothy J. BuckleyD).ir-, -.or
u Expos're Methods and 116-)surernents Djvisiori
THRU: Jennifer Orme-Zavaleta, Director
National ExposureResearch Labor
)r
TO: Linda Culpepper, Deputy Director
Division of Water .Resources
North Carolina Del wtuent of Environmental Quality
Enclosed please find our fourth report of'PFAS concentrations in Cape -Fear River water
samples collected tinder the direction of NC DEQ. 'I'his report includes preliminary rindings
from oux non -targeted analyses. These results were presented and discussed during your visit
August 28, 2017 to our Laboratory in. Rcsea�rch Triangle Park-, N.C.
Thank you for inviting us to be a part of this effort that addresses a very important public
health concern in North Carolina. These results represent the effort of many within our lab, but I
would especially like to acknowledge Drs. Mark Strynar,.Aiidy Lindstrom, James McCord, and
Seth Newton in conducting the laboratory analyses. Dr.My'riam Medina -Vera who provided
invaluable support and coordination, and Ms. Sania Tong Argao who supported and oversaw
quality assurauce.
If you have any questions or concerns., do not hesitate to contact rn. c at. (919 541-2454 or
--v, , I look for,�Nand to OUr continued work together. email b� �,ckley
Enclosure
CC: Beck y.B. Allenbach, USEPA. Region 4
Jeff Mo-iYis, USEPA OPPT
Bets. Behl, USEPA. OW
Pete; Gr cvatt., USEPA, OW
DEQ-CFW-0008321 0

Summary of Results
Our preliminary non -targeted results are limited to samples from the Chernours outfiall
and finished water from the Sweeney Water Treatment Plant for weeks 1 6, We chose these
sites because we believe the concentrations observed bound this portion of die watershed.
Furthennore, we did not want to dekty our relx)rtflig due to the additional time required to
assemble and interpret results from. the other locations. We are, continui.ng to ivoek on a
comprehensive report that will include targeted and non -targeted analysis results at all locations
over the seven weeks, of sampling,
We include five anal�tes in this initial non -targeted analysis report (Table 1). An
important limitation to ournon-targeted analysis resitfis is that these result-, are, considered scini-
quantitative, We cannot know the exact concentration because no a�ithentic '.;Iandards arc
available for these chemicals. However, we are very confident of the chemical identitv based on.
Table 1. Analytes Measured Non -Targeted LC/TOFMS Analysis
Monoisotopic
Short Name Chemical Naine Formula CAS no
Mass (Da)
PFESA
tetrat
Byproduct 2
(trifluoroniethoxy)acefic acid)
acid
311,9680
acid
We provide -senil-quantitative "concentrations" in. two fornis (Fable 2). The first is the
peak area that is associated with the monoisotopiemass.for each compound. The peak area is
generally proportional to theanalyte concentration and it is use.ful in interpreting changes in
concentration over time and bctwcen locations -for a given analyte, For example, 1.1"or PFMOA'A_
measured in S�veeney Finished water, we see the peak -area change frorn -4.5 million to 3.000
counts fTom week I to 6This can be interpreted as roughly a I.,500-fold decrease in
concentration. vvithout knowing the exact concentration.The second wa-y,-,'�,e providea selni-
quantitative estimate of concentration is toscale the non-targetedamdyte based on the measured
concentration of GenX.
[N7A]= *NTApj
----
2
OEQ-CFVV_00083211

August 31,:2 17
[GenX] is the concentration of Germs. (rrg./t,)
I Tr PA is the integrated peak area for the non -targeted arr lyf°te
Ge11XPA is the integrated peals area for GenX
In essence, we are assuming that the mass spectrometer responds to the non -targeted
arralyte as if it were GenX. The actual instrurr ent response may he weaker or stronger resulting
in an reader- or over -estimation of the note -targeted cortcenb ation. Our experience with this; cuss
f analytes suggests that estimates of this fashion are, accurate to within —I -fold of the estimated.
value.
I`iie non -targeted arralw=te estimated concentrations ns are particularly uncertain at the
Chemours outfail during weeis 1-3. Concentrations were so high that even alter samples were
,diluted 20X, we exceeded our calibration curve for GenX and were also likely saturating the:
mass spectrometer for bath Germs.: and non -targeted an lytes, The semi -quantitative estimate for
the non —targeted analytes are particularly uncerta n are likely underestimated. These results are
shovm inTable 2 and. have been flagged accordingly.
NXrhether considering peak area or estimated concentration, the non -targeted results show
two very different time profiles. For three of the analytes, concentrations at the orrtfail and
Sweeney finished water showw, a precipitous drop very similar to what was observed for Gen
(Figures l n- ). These results suggest that whatever mitigation n strategy used to reduce uce GenX -was
also effect w w aor these three chemicals. The second time proffile is for two periluoroethersulfunic
acid (l FEStk) byproducts. We believe these:, chemicals are a byproduct of l at`on production. In
contrast to the 6ren -related c:hen'.iicals, l'cr. these two chemicals, we do Trot observe a clear
decreasing tread in concentration (.Figures 5 & 6). These results suggest the discharge of these:
chemicals wa�" iinalfectcd by whatever strategies were used to mitigate GetLX discharge.
Concentrations of the PFESAs range from 2,900 to 73, 00 'rg/1.. at the Chernours outfall and 5
to 7;860 z g/L in Sweeney -finished drinking water. Note that these concentrations are in the same
range as GenX originally noted in Bran et al., 201:0
In FigUre 7, the Blots show the two different types of time profiles for the six analytes.
Each analyte is graphed as a relatiN e percentage of its maximum intensity over the sampling
period. For the Pf ESAs byproducts, this maximum period occurred in the twiddle of sampling,
while 'for t:he other analytes, the maximuni was during the first week.
As with GenX. our QA/QC results for the non -targeted results are within expected
tolerances. We did not detect,any of the analytes in field blanks, indicating that no field or lab
contamination tools place. Because there are no standards for these anal), tes, we haw.e no
assessment of accuracy, but duplicate analyses were within 20 percent. The laboratory rnethods.
for the results reported here are described in Sun et aL, 2016' andtrynar° of aL, 201 52
..
! Swi w1, Arc: alo I : Str1+nsti M LindstFo?Fl A Richardson M, Kearns B, t' .:kett A SrFF,tiF.C; tisFappc frrrt,l. LLf;n aErd Fina.rgin-,
r efflw o lk,A Sut)SWICCS .wrio rr ponant Drinkir3I, wsaler t omar,inarits in flic Cape Fear River Wate r: i d of North Ca:tifina €'m,,itonsa ental
�ieEr:: '.'e :ttra;tlds L :t it . 2016
Strl'n.ar IMDar3'F.F,tE'S Nickl¢3t: m Ri i Yadn;?::S, r.indstrt. m A, And pawn ',,.NiMrtIfflan L' M, Fk�'rrer t, B4 t ldwovttf a3t on of 'ado-vol
Pc:rfluomajlky€ E'it1CF Carryoxyfiw Adds, Er=I r:c:' w, f and Sldrfcmiw A,-,idlf fPFFSN,,) in N<F11 ral W,xtcrs U,3i E3g Accar itL E,�C EFEF -L?i-f r3Ll7i �1 i:F :C
sp7odt TiFA3 dy ('M MS)l'YiL'tF6:n SO eCr3H o 2011.
DEQ-CFW 00083212

Attgus1 ;1, 2017
Table 2. Semi -Quantitative Estimates of GenX andon-T r t mal te Concentrations
Measured at Chemo rs Outfall and Sweeney Finished Drinking Water During Sampling
Weeks I -- 6.
PF�wT€OAA
amours Outfall 002
63,712,278
10,363,496
21, 60
134r000
1
PFC2H.xA
0hemours Outfall 002
1
1.82,599,647
1.0363.;496
21,760
383,000
1.
PF030A
01€emours Outfall 002
1
51,940,.394
10,363,496
21,760
109,000
1
GemiX
Chemours Outfall 002
1
10,363,496�
10,363,4 6
__..21 760
_ 21,800
1
I PFESA Byproduct 1
6l>emours Outfall 002
1
1,380,791 1
1:0,3631,496
21,760
2,900
1
PFFSA Byproduct 2
0hemours-Cutfall 002
10,3 33,496
21,760
29,500
1
R"I"'w1 Yi
C1.tall 002.
2lerr�rs
37,33,851.
8,34,860
15,250
68,300.
PF02HxA
0hermours Outfall 002
2
71,331,553 8,345,860 15,250 130,000 1
PF030
Chemours Outfall 002
2
19,111,355
8,345,860
15,250
34,900 1
....... ............
GenX
Chemo rs uffall 002
2
«....�.. «..�.�...�...�
8,345,860
..............................
8,34�s,860
15,250 '
____
15,300 1
........
PFFSA Byproduct.1
..-..._..............................
........ .......................
0h ou, s Outrall 002
2
1,995,442
8,345,860
w ._W ..._.
15,250
�..0 u.uu..u.u.............. «..........
3,460 1
.........,.�..,,.. _.M...... ... ..._.....
PFFS Bypr€ duct 2
......... ......................
6hemours Outfall 002
2
13,230,172
......,,_ _...........
8,345,860
_ ...........................
........ .........._.............
... 1S,2.50
24,200 1
DEQ-CFW 00083213

For eek#3, there was insufficient sample available fora Sweeny finished water analysis..
For creek#4, there was ins€<fficient sample available for a Chernours outfall 002 water anal
PFMcl k
PF02HxA
Chemours Outfall 002
�_....�
Chemours Outfall002
�.
5
m.. 558,33x
3,66,856
287,302
287,302
713
..m.....
t 713
� 1,390
............................................
910
3 .......
3
PF030A
Chemours Outfall 002
5
175,874
287,302
713'
43
3
CenX
Chernours Outfall 002
5
287,302
287,302
t 713
713
3
PFESA Byproduct 1
Chemours Ou:tfall 002
9
1,797,348
281,302
713
4,460
3
PFESA Byproduct'2
Chemours
«««««
15,762,943
u 287,3132.u«««.k«<uuu
713..
39,100 u«< u
3 «<
Chemours Cutfall 002
6
1:1.3,443
1.6,637
102
696
mm .___PFMOA _______________
PF32lxlerrs
,,,,,,,,, ,,,u rs _ I1' 0-0 —2
-O�tfall 002
.._..-,..-.._.
6
70,333
16,637
— _..._.._ .... .....
102
......... ........ .. .........
431
Pt 03CA
Chernours SOutta€l 002
6
1.4,038
16,637
102
86
enX
Chemours OutfaH 002
6
16,637 «<
««««<16,637 «<
102.
1.02
PFESA Byproduct 1Chemours
Outfall 002 �
6
2,569,948
1.6,637 �
10
15,800
PFESA Byproduct 2
Chemours Outfall 002
6
12,036,574
16,637
�.....
102 1
73,900
Flag
1 = Sample
2 = Sample
Sample
diluted 20X and diluted sample exceeded the calibration curare for Gen
diluted 5X
dilute: 20
5
DEQ-CFW 00083214

Augu,,st3 \ 2017
..Figure / Gen/ Concentration (n g/l Profile
.......... .R
20OU
» GmA «ems uK
. .
§
DEQ CFW-0 0083 ]5

Figure 3. PF02HxA Concentration. L Profile
5E t ?
7000
a
iW
200000
is
xxxxxx
x�Er
20001
i7E 000
XXXI
50000
a
2 3 5
Sample Week
Figure k PF030A Concentration(_ -) Profile
120,000
3000
xxxxxxx
Sweeney Chem
Ut ?000
5 C,
2000
aa
i3}
xxxxxxx
3
2 4 56
Wo.mpk%e
DEQ-CFW 00083216

4i
21
I
August 3 1, 2017
Figure 5.
PROS A Byproduct 1 Concentration (nW.I,) Profile
.8 C, 0 0
1.60,00
14,MO
1M,00
—60,
........ ........
0
M
xx�
X
1000
X,
I.
2 3 4 5 6
Sarnple week
Figure, 6.
PFESA Byproduct 2 Concentration (ngfL) Profile
8DOOO
MWX)
Cii
40000
0
..... ........
300
X
70000
........ ....
10000
3 4 5 6
Sample wepk
8
DEQ-CFW-00083217

August 31, 2017
Figure 7. Relative change (compared to highest measured value) in PFAS concentration
over weeks I — 6 for GenX and. NTAs at the Chemors outfall and Sweeney- Finished
Drinking Water. GenX and NTAs in Panels A,D,E, & F show a consistent decreasing
profile. The PFESA Byproduct concentrations are variable and do not show a clear trend.
1100
75
50
25
D
0
100
Z
50
25
0
2 4 6
Sampling Week
9
Chemours OutfaH 002
SweOney
DEQ-CFW-00083218