DRC-2009-002323 - 0901a0688011d267

[piC^J^CCPi-OD^P^aS Page lofl
Loren Morton - Groundwater Discharge Permit documents
From:
To:
Date:
Subject:
CC:
Attachments:
David Frydenlund <[e-mail]>
Loren Morton <[e-mail]>, Phillip Goble <[e-mail]>
6/5/09 4:23 PM
Groundwater Discharge Permit documents
Dane Finerfrock <[e-mail]>, Harold Roberts <HRoberts(gdenisonmines.com>, Ron
Hochstein <[e-mail]>, Steve Landau <[e-mail]>, David
Turk <[e-mail]>
Memo re GWCLs June 3 2009.pdf; Letter regarding Decon Pad 6.05.09.pdf; Letter regarding Seeps and
Springs Sampling 6.05.09.pdf; Memo re GWCLs June 3 2009.pdf; Letter regarding Decon Pad
6.05.09.pdf; Letter regarding Seeps and Springs Sampling 6.05.09.pdf
Dear Loren,
Attached are:
1. Memorandum from Dan Erskine regarding GWCLs.
2. Letter regarding the conditions to be satisfied before the Mill can use the new Decontamination Pad.
3. Letter requesting removal of the requirement to sample for SVOCs at the seeps and springs in the vicinity of the Mill.
Hard copies of the attached documents are being delivered to you by Federal Express, for receipt on Monday.
I will provide you with a mark-up of your draft revised GWDP and Statement of Basis, indicating our suggested wording changes
next week. I will be out of the office through Wednesday of next week, but will try to get the mark-up to you by Thursday or
earlier if I can.
Dave
David Frydenlund
Vice President. Regulatory Affairs and Counsel
t: [phone] | f: [phone]
1050 17th Street, Suite 950, Denver, CO 80265
DENISON MINES (USA) CORP
www.denisonmines.com
This e-mail is intended for exclusive use the person(s) mentioned as the recipient(s). This message and any attached files viiith it are confidential and may contain privileged or proprietary
information. If you are not the intended recipient(s) please delete this message and notify the sender. You may not use, distribute print or copy this message if you are not the intended
recipient(s).
file://C:\Documents and Settings\Lmorton\Local Settings\Temp\XPgrpwise\4A294693EQDOMAlNEQRAD... 6/9/2009
ffiffi
ffitH3ta
MEMO
Loren Morton (UDEQ DRC) and Phillip Goble (UDEQ DRC)
INTERA Incorporated
6000 Uptown Blvd, NE
Suite 100
Albuquerque, NM 87110
Telephone: [phone]
Fax [phone]
'ry
g safraton
fT'*"\/^"'- tD . -+A
I'*:ffiffi -g
it-,. -,!.*Tl"-..t 'f,
f,'rom: Daniel W. Erskine Ph.D. (INTERA) figgW"s
Date: 61512009
Re: Denison Mines (USA) Corp.
(GWCLS)
%gsi.
-- Determination of Ground Water Compliance Limits
Reference is made to the State of Utah Ground Water Discharge Permit No. UGW 370004 (the
"GWDP") for Denison Mines (USA) Corp.'s ("DUSA's") White Mesa Uranium Mill (the "Mill),
and to the following reports:
o Revised Background Groundwater Quality Report: Existing Wells for Denison
Mines (USA) Corp.'s White Mesa Mill Site, San Juan County, Utah, October
2007, prepared by INTERA, Inc. (.'INTERA") (the "Existing Well Background
Report");
c Revised Addendum - Evaluation of Available Pre-Operational and Regional
Background Data, Background Groundyyater Quality Report: Existing Wells for
Denison Mines (LISA) Corp.'s Wite Mesa Mill Site, San Juan County, Utah,
November 16, 2007, prepared by INTERA (the "Pre-Operational and Regional
Background Report"); and
o Revised Addendum - Background Groundtvater Quality Report: New Wells for
Denison Mines (USA) Corp.'s Wite Mesa Mill Site, San Juan County, Utah,
April30, 2}}8,prepared by INTERA (the "New Well Background Report").
The Existing Well Background Report, Pre-Operational and Regional Background Report and
New Well Background Report are referred to collectively as the "Backgtound Reports".
Reference is also made to the Summary of Work Completed, Data Results, Interpretations and
Recommendations for the July 2007 Sampling Event at the Denison Mines, USA, Wite Mesa
ffilniEraffiffi June 5. 2009
Uranium Mill Near Blanding, Utah,May 2008, prepared by T. Grant Hurst and D. Kip Solomon,
Department of Geology and Geophysics, University of Utah (the "University of Utah Study").
At a 5llll2009 meeting with Loren Morton and Phillip Goble of the Utah Department of
Environmental Quality ("UDEQ") Division of Radiation Control ("DRC"), DUSA committed to
provide DRC with a technical memo giving specific reasons to support setting Ground Water
Compliance Limits ("GWCLs") in the GWDP that exceed the State of Utah Ground Water
Quality Standards ("GWQSs") for constituents in monitor wells that were not a part of the
University of Utah Study. Additionally, DUSA committed to providing proposals for methods to
set GWCLs for constituents in monitor wells with naturally rising trends and in wells with data
sets of extremely low variance.
WELLS NOT INCLT]DED IN T]NIVERSIIY OF T]'TAH STT]DY
GWCLs that exceed the respective GWQSs have been proposed for the following constituents in
the following wells that were not included in the University of Utah Study:
Cadmium in VIW-12 and MW-28
Proposed GWCLs for cadmium of 7 1t"glL and 5.2 p,glL in MW-12 and MW-28, respectively,
exceed the GWQS of 5 pgll,. MW-12 and MW-28 were not included in the University of Utah
Study. However, the Background Reports and the following analysis demonstrate that cadmium
concentrations in these two wells are unlikely to have resulted from potential tailings seepage for
the followinq reasons:
o With respect to Cadmium in MW-12, there is no rising trend over time. The
relatively high GWCL is the result of two high sample results in the late 1980s;
however since that time sample results have been low or non-detect (see
Appendix D of the Existing Well Background Report). In Section 7.2-l of the
Existing Well Background Report we stated that:
ooThe proposed GWCLs for cadmium in MW-,l MW-2, MW-3, MW-5
and MW-12 exceed the GWQS for cadmium. In each case there is no
rising trend in cadmium, and in fact there are statistically significant
downward trends in cadmium in MW-l, MW-2, MW-3 and MW-5,
having Mann-Kendall Z values of -3.22, -3.9, -4.03 and -4.49,
respectively.
Page 2 of 20
ffimif;:taW June 5. 2009
From a review of the plots for cadmium in these wells (see Appendix D),
it is evident that, while there appears to have been some detections of
cadmium in samples taken in the late 1970s and early 1980s, cadmium has
been low or non-detect in these wells since that time. The one exception
appears to be in far downgradient MW-3, which has had detections at
approximately half the GWQS since then.
The early high detections of cadmium in these wells, which appear to have
systematically ended in the early 1980s for each of these wells suggests
that the high results could have been due to variations in sampling or
analytical techniques."
With respect to Cadmium in MW-28, the sample results are within the range
established for the site, and there is no systematic spatial relationship between
relatively high cadmium concentrations at the site and the Mill's tailings cells.
We concluded in Section 6.2.8 of the New Well Background Report that:
"There are relatively high concentrations of cadmium in new wells MW-
3A (2.23 pglL) and MW-28 (3.29 VelL). However, these concentrations
are well within the range established for the site. For example, far
downgradient well MW-3 and site well MW-5 have concentrations of 4.78
pglL arld3.24 pglL, respectively (see Figure 12). In no case does there
appear to be a systematic spatial relationship between cadmium
concentrations and the location of the tailings impoundments. We
therefore conclude that cadmium concentrations at the site are the result of
natural influences."
Cadmium is present in tailings solutions at a relatively low average concentration
of 3,400 p,glL, as stated in Table 5, Summary of Estimated and Measured IUC
Tailings Wastewater Quality, set out in the December 1,2004 Statement of Basis
for the GWDP (the "2004 Statement of Basis"). Cadmium is also typically
relatively insoluble in water except at low pH (Rai andZacharu,1984). For these
reasons, elevated cadmium concentrations from potential tailings seepage without
accompanying low pH, high chloride, and sulfafe that are lcnown to be present in
tailings solutions and are much better indicator constituents is unlikely. Neither
Page 3 of 20
iaffi
MiEITAffi June 5, 2009
MW-12 nor MW-28 has high chloride or sulfate or low pH (see Table 16 of the
Existing Well Report for MW-12 and Table 10 of the New Well Background
Report for MW-28), nor do they demonstrate rising trends in sulfate or chloride
(see Appendix D of the Existing Well Background Report for MW-12 and
Appendix B of the New Well Report for MW-28).
Cadmium concentrations in MW-12 and MW-28 are lower than concentrations in
MW-3A located far-downgradient of the Mill's tailings cells. MW-3A was
included in the University of Utah Study which concluded that groundwater from
that well has a different geochemical signature than tailings cell wastewater. In
addition, mean cadmium concentrations in samples from MW-28 and MW-12 are
3.3 and 1.4 pg/L, respectively. These values compare with mean concentrations
from nearby monitor wells MW-z (2.9 p.e/L) and MW-5 (3.2 1t'glL). MW-2 and
MW-5 were included in the University of Utah Study which concluded that
groundwater from those wells has a different geochemical signature than tailings
cell wastewater. The concentrations of cadmium in MW-12 and MW-28 are
therefore consistent with backeround levels.
As a result, we have concluded that cadmium concentrations in samples from MW-28 or MW-12
have not been impacted by Mill activities.
Iron in MW-32
The proposed GWCL of 14,060 pgil for iron in MW-32 exceeds the GWQS of 11,000 pgll-.
MW-32 was not included in the University of Utah Study. However, the Background Reports
and the following analysis demonstrate that the iron concentration in that well is unlikely to have
resulted from potential tailings seepage, for the following reasons:
o Iron concentrations in groundwater at the site are variable, including a rising trend
in iron in upgradient MW-1, which suggests that iron concentrations are not being
impacted by potential tailings seepage. In Section 11.11 of the Existing Well
Background Report we concluded that:
"There is a rising trend in iron concentrations in samples of groundwater
from MW-l and MW-5 and the proposed GWCL for iron in MW-32
exceeds the GWQS for iron. The highest observed value in upgradient
Page 4 oI 20
rlffilniEqaW June 5. 2009
monitoring well MW-l is 3,570 pg/L while, the highest observed
concentration of iron in MW-5 was 417 WglL; two orders of magnitude
lower than the GWQS. The variability of iron concentrations in samples
from all three wells suggests that colloidal iron may be influencing
concentrations. This may even occur as submicron particles entrained
during sampling disturbances in the well bore that are too small to be
filtered out. If so, concentrations of trace metals measured in samples
from these wells should be regarded with caution because trace metals are
known to adsorb on colloidal iron particles. The fact that iron
concentrations are flagged at upgadient locations as well as in monitoring
wells adjacent to the tailings impoundments suggest that they are not
related to seepage from tailings impoundments."
Iron is relatively immobile except at very low pH and would be unlikely to be an
indicator constituent. In Section 6.2.11 of the New Well Background Report we
concluded that:
o'Concentrations of iron at the site are variable and range from 8 StglL in
Well #37 to 7,942 p"glL in MW-32 (Figure 15). Iron concentrations in all
new wells are lower than the GWQS. Iron is relatively insoluble except at
very low pH, severely limiting the concentration of iron that can travel in
groundwater in the carbonate rich geologic environment beneath the
tailings impoundments. Therefore, it is unlikely that a potential tailings
cell leak would manifest itself in an increasing trend in iron in the absence
of increasing trends in chloride, sulfate, fluoride, and uranium (see the
discussion in Section 2.5.9)."
High iron concentrations in groundwater with neutral pH is rare, suggesting that
the samples may not have been filtered properly. In Section 2.5.9 of the New
Well Background Report we concluded that:
"There is a statistically significant rising trend in iron in MW-30;
however, the concentrations of iron in that well are very low with a mean
of 75.6 pg/L compared to a GWQS of 11,000 pgll. Samples of oxidized
water with pH values between 6.5 and 8.5 that contain iron in
Page 5 of 20
ffiMiETAW June 5, 2009
concentrations above a few micrograms per liter are rare, and higher
concentrations sometimes reported in such waters are generally
particulates small enough to pass through a 0.45 micron filter (Hem,
1992). This is a common sampling problem in wells that produce little
water or in wells with iron casings. While small variations in Eh and pH
can cause variations in iron concentration in groundwater on the order of
magnitude observed in MW-30 (Hem, 1992), iron is relatively insoluble
except at very low pH, severely limiting the concentration of iron that can
travel in groundwater in the carbonate rich geologic environment beneath
the tailings impoundments."
There is a significantly significant downward trend in iron in MW-32. See
Appendix D to the Existing Well Background Report.
There are no statistically significant upward trends in chloride, fluoride or
uranium or a significant downward trend in pH in MW-32 (see Table 16 and
Appendix D to the Existing Well Background Report). There is a statistically
significant upward trend in sulfate in MW-32, however, we noted in Section 11.2
of the Existing Well Background Report that:
"MW-32 has a high R2 value [for sulfate], due in large part to one data
point collected in September 2002,1ess than 2 months after the well was
developed. As a result, this low initial sample result may not be
representative. If this data point is removed the trend in MW-32 would no
longer be significant."
We also noted in Section ll.2 of the Existing Well Background Report that the
most significant rising trend in sulfate at the site is in upgradient well MW-18,
and concluded that
"[t]he fact that the most pronounced increasing trend in sulfate at the Mill
site is in upgradient well MW-18 is strong evidence that the other
increasing trends in sulfate at the Mill site are also due to natural causes."
Wells MW-5, MW-18 and MW-30 were included in the University of Utah Study,
which concluded that the groundwaters in those wells have different geochemical
Page 6 of 20
ffilnffiqaW June 5, 2009
signatures than tailings cell wastewater. This corroborates the conclusions above
that the rising trends in iron in MW-5 and MW-30 and the rising trend in sulfate
in MW-18 ate due to natural influences, and that the relatively high
concentrations of iron in MW-32 are also due to natural influences.
As a result, we have concluded that iron concentrations in samples from MW-32 have not been
impacted by Mill activities.
Manganese in MW-12, MW-L 7, IrvIW-24, NIW-25 lVlW-26, lVlW-28, and MW-32
Proposed GWCLs for manganese in MW-12, MW-17, MW-24, MW-25 MW-26, MW-28, and
MW-32 exceedthe GWQS of 800 p"glL,as follows:
None of these wells were included in the University of Utah Study. However, the Background
Reports and the following analysis demonstrate that manganese concentrations in these wells are
unlikely to result from tailings seepage for the following reasons:
o Manganese concentrations in MW-12 and MW-17 have been decreasing over
time. The proposed GWCLs exceeds the GWQS as a result of earlier higher
detections; current detections in each of these wells are relatively low for the site
(see Appendix D of the Existing Well Background Report);
o With respect to existing well MW-26, Section 11.6 of the Existing Well
Background Report provided the following analysis and conclusions:
"Although the proposed GWCL for manganese in MW-26 exceeds the
GWQS (see Table 16), there is not a significant upward rising trend
(Z:0.33). Lack of a rising trend and consistently elevated concentrations
of manganese indicate that the manganese in MW-26 is baseline for the
site."
Well MW-12 MW-17 MW-24 MW-25 MW-26 MW-28 MW-32
Proposed
GWCL
(pe/L)
2.088.80 9r5.4 7,507 1,806 1,610 r.837 5,594.9
Page 7 of 20
ffimffiqaW June 5. 2009
With respect to existing well MW-32, Section ll.6 of the Existing Well
Background Report provided the following analysis and conclusions:
"Manganese concentrations in groundwater samples from MW-32 exhibit
a significant upward trend (R2:0.61) from 3,660 ytglL in results from the
first sampling event in September of 2002 to 5,470 ltglL in results from
the most recent sampling event in June 2007. MW-32 is located a short
distance downgradient of identified elevated chloroform concentrations.
Manganese has been cited as an electron acceptor during the
biodegradation of chlorinated solvents (Wilson, 1996). Microorganisms
feed on the energy released from removal of electrons from the
chlorinated solvent and their transfer to electron acceptors. Such a transfer
would result in the reduction of relatively immobile oxidized manganese
(IV) to more mobile reduced manganese (II). Such a process may explain
the rising manganese trend in groundwater samples from MW-32.
A significant rising manganese trend also occurs in monitoring well MW-
l1 (R2:0.51); however, concentrations of manganese in samples from this
well are well below the GWQS and over an order of magnitude below
concentrations found in samples from other wells at the site (see Appendix
D). Further, there are no upward trends in chloride or fluoride in MW-l1.
For these reasons, we conclude that upward trends in manganese are not
caused by activities at the Mill."
With respect to new wells MW-24, MW-25 and MW-28, Manganese
concentrations in groundwater samples do not show any rising trends and are
consistent with natural variabilrty in background at the site. We made the
following conclusions in Section 2.5.1 of the New Well background Report:
"Table 3 indicates that the proposed GWCLs for manganese exceed the
GWQS in groundwater samples from the following monitoring wells:
MW-24; MW-25; MW-28; MW-29; and MW-3A. The proposed GWCLs
that exceed the GWQS are lower than the highest observed values in the
region in far downgradient well MW-22 (34,550 VglL) and in far
upgradient Well #38 (7,450 pglL). The proposed GWCLs for other new
Page 8 of 20
ffi
INIETAW June 5, 2009
wells (MW-30 and MW-31) are comparable to or lower than other wells at
the site. Additionally, the new well with the highest mean concentration
of manganese, MW-29 (5,028 Vg/L), was age dated by the University of
Utah using a tritium isotopic method and preliminary results indicate that
the water in MW-29 predates any milling activities at the site. There are
no statistically significant rising trends in Manganese in any of the new
wells. Manganese concentrations measured in new wells are therefore
consistent with background variability at the site.
Manganese concentrations that exceed the GWQS of 800 pglL ate
prevalent throughout the site and region (see the discussion in Section
6.2.5 below). Manganese concentrations in groundwater samples from the
new wells are consistent with natural variability in background."
Concentrations of manganese in the wells in question are within the background
ranges for the site and area, andthere is no systematic spatial relationship between
manganese concentrations and the tailings cells (see Figure 9 of the New Well
Background Report). We made the following additional observations in Section
6.2.5 of the New Well background Report:
'oThe highest observed average concentration of manganese in site
monitoring wells is now in new monitoring well MW-29 (5,028 pLgL)
followed by existing monitoring well MW-32 (4,922 VelL) (Figure 9).
However, concentrations in samples of groundwater from both wells are
within the regional background highs of 34,550 VglL in far downgradient
well MW-22 and 7,450 WglL in regional background Well #38.
Manganese values should be interpreted with caution because high values
often result from colloidal particles that are entrained in groundwater
samples during disturbances in well sediment caused by pumping. This
effect is particularly common in wells that do not yield much water such
as those on the west side of the tailings impoundments.
In no case does there appear to be a systematic spatial relationship
between manganese concentrations and the location of the tailings
impoundments."
Page 9 of 20
.rffiffiffiffim[ffiAW June 5, 2009
Average manganese concentrations in samples from MW-12, I|l[W-17, MW-24,
MW-25, MW-26, MW-28 and MW-32 are lower than concentrations in samples
from MW-29 (5,028 VglL) (see Figure 9 of the New Well Background Report for
a listing of the average concentrations of Manganese in the various wells). MW-
29 was apart of the University of Utah Study which concluded that groundwater
from the well has a different geochemical signature than tailings cells wastewater.
The concentrations of manganese in the wells in question are therefore consistent
with natural background at the site.
Elevated manganese concentrations from potential tailings seepage without
accompanying low pH, high chloride, and sulfate that are known to be present in
tailings solutions is unlikely. In Section 2.5.1 of the New Well Background
Report, we note that:
"The new wells with the highest concentrations of manganese, MW-29
and MW-24, with concentrations of 5,028 and 3,535 pgll-, respectively,
are not associated with high concentrations of chloride or uranium and are
associated with only moderate concentrations of fluoride and sulfate,
which are the best indicator parameters for potential tailings cell leakage.
Accordingly, wo do not consider the manganese at the site to have
originated from potential tailings cell leakage."
As a result, we have concluded that potential tailings seepage has not impacted manganese
concentrations in samples from these wells.
Uranium in MW-12 MW-23 and MW-26
Proposed GWCLs for uranium in MW-17, MW-23 and MW-26 of 46.66,32 aurrd 41.8 p'!L,
respectively, exceed the GWQS of 30 pgll. None of these wells were included in the University
of Utah Study. However, the Background Reports and following additional analysis demonstrate
that uranium concentrations in these wells are unlikely to have resulted from potential tailings
seepage for the following reasons:
. Rising trends in uranium in MW-17 and MW-26, which have contributed to the
proposed GWCLs for those wells exceeding the GWQS, as well as rising trends in
uranium in other site wells, are not related to Mill activities. With respect to
Page {0 of 20
ffiMiEIIAW June 5. 2009
MW-17 and MW-26, we make the following observations in Section 11.4 of the
Existing Well Background Report:
"There are increasing trends in uranium in MW-12, MW-14, MW-15 (Z
values of 5.08, 4.65 and 4.03 for MW-12, MW-14 and MW-15,
respectively), and in MW-17 and MW-26 (R'values of 0.11 and 0.38,
respectively). However, there are no increasing trends in chloride or
fluoride in any of those wells, which would be expected if these rising
trends in uranium were caused by seepage of tailings solutions (see
Section 9.0). In some circumstances, uranium may approach the mobility
of chloride in groundwater, but it cannot exceed it. It is, therefore, not
possible for uranium to be trending upwards in these wells without a
corresponding and linked increase in chloride. This fact, together with the
factthat uranium is increasing significantly in upgradient well MW-18 as
well as in far downgradient well MW-3 (which also are not associated
with increasing chloride), is conclusive proof that the uranium trends in
these wells are not being impacted by Mill activities. As discussed in
Section 7.4, since MW-26 is a chloroform pumping well, any trends in
MW-26 should be considered likely to be the result of such pumping and
should be discounted."
The elevated GWCL for uranium in MW-23 is within the range of background for
the site, and there is no increasing trend in uranium in MW-23. With respect to
MW-23, we make the following conclusions in Section 2.5.3 of the New Wells
Background Report:
"The calculated uranium GWCLs in monitor wells MW-23 (32 ltglL),
MW-24 (36 p,glL),MW-27 Qa p,glL), and MW-3A (35 pglL) are slightly
elevated above the GWQS of 30 pglL. However, while the mean uranium
concentration of MW-27 (31.40 VglL) is slightly elevated above the
GWQS, the mean concentrations in the other wells are all lower than the
GWQS (23.2 yt"glL,9.0 pglL, and 24.7 y'glL for MW-23, MW-24, and
MW-3A, respectively). These values are aLI within the range of regional
background values for uranium (for example, 48.5 pgll. and 42.8 ltglL in
upgradient Well #39 and MW-18, respectively and 41.7 ltglL and 31.4
Page 11 of20
ffimiErtaW June 5. 2009
pgll, for far downgradient l|r4W-22 and MW-3, respectively (see Figure 6).
None of these wells exhibits a statistically significant increasing trend in
uranium concentration over time. Furthermore, MW-18 and MW-3 are
among those wells that have University of Utah tritium isotopic age dates
indicating that the water in them predates any milling activities at the site.
For these reasons, the uranium concentrations in MW-23, MW-24, MW-
27, and MW-3A are not consistent with potential tailings seepage impacts
and are consistent with regional background values."
In Section 6.2.3 of the New Well Background Report, we noted that:
"Average uranium concentrations in MW-27, upgradient of the tailings
cells, are above the GWQS; however, the concentration of uranium in
samples of groundwater from all new monitoring wells falls within the
range of values from previously-existing site monitoring wells and
regional background values. The highest observed average concentrations
continue to be in samples of groundwater from MW-14 (59.8 FgiL) and
MW-15 $9.3 ltglL) followed closely by regional background Well #39
(48.5 pgll,), upgradient monitoring well MW-18 92.8 1tg/L), and far
downgradient monitoring well MW-22 (1.7 VglL) (Figure 6). Note that
samples of groundwater from MW-14 and MW-15 contain low
concentrations of chloride and fluoride and moderate concentrations of
sulfate, allowing the conclusion that uranium concentrations do not result
from potential tailings seepage impacts."
Average uranium concentrations in samples from MW-17,MW-23 and MW-26
are lower than average concentrations in samples from MW-14, MW-15 and
MW-18 (58.51, 43.42 and 42.8 pgll, respectively, averaged over 2006 and2007,
as shown on Figure 14 of the Existing Well Background Report). MW-14, MW-
15 and MW-18 were all part of the University of Utah Study which concluded
that groundwater from each of those wells has a different geochemical signature
than tailings cells wastewater, which provides further confirmation that uranium
concentrations in the ranges found at the site are natural background;
Page 12 of 20
ffimiffiaW June 5, 2009
The average value of uranium in tailings solutions is estimated in Table 5 of the
2004 Statement of Basis to be 93.6 mglL while the average value of chloride is
estimated in that Table to be two orders of magnitude higher at4,608.44 mglL. ln
general, uranium is retarded behind chloride during groundwater transport.
Therefore, uranium concentrations from potential tailings cell seepage impacting
groundwater would have to be accompanied by corresponding increases in
chloride concentrations. There are no increasing trends in chloride in MW-17,
MW-23 or MW-26 (see Appendix D of the Existing Well Background Report for
MW-17 and MW-26 and Table 10 of the New Well Background Report for MW-
23).
o Uranium concentrations are variable across the site but in no case does there
appear to be a systematic spatial relationship between uranium concentrations and
the location of the tailings impoundments. In Section 10.6 of the Existing Well
Background Report, we noted that:
o'It is noteworthy from Figure 14 that the highest concentrations of
uranium are distributed across the site, both upgradient and downgradient,
as well as close to the tailings cells themselves, which further supports our
conclusion that uranium concentrations at the site have not been impacted
by Mill activities."
As a result, we have concluded that uranium concentrations in samples from these wells have not
been impacted by Mill activities.
pH in Mw-28
The proposed GWCL for pH in MW-28 of 6.1-8.5 s.u. is outside of the GWQS range for pH of
6.5-8.5 s.u. MW-28 was not included in the University of Utah study. However, the
Background Reports and the following analysis demonstrate that pH in MW-28 is unlikely to
have resulted from potential tailings seepage for the following reasons:
o There is no systematic spatial relationship between low pH in monitoring wells at
the site and the Mill's tailings impoundments. In Section 6.2.10 of the New Well
Background Report we concluded that:
Page 13 of 20
ffitffi
INEE1AW June 5, 2009
"Average pH levels at the site range from a high of 8.9 in MW-20 to a low
of 6.7 in MW-28, with no particular spatial relationship that would suggest
potential tailings cell seepage (Figure 14). Decreasing trends in pH at the
site appears to be cyclical and typical of many wells at the site (see the
discussion in Section 2.5.6). For these reasons we conclude that pH levels
are the result of natural background influences."
The systematic decreasing trends in pH in a number of monitoring wells at the site
are not due to Mill influences. In Section 2.5.6 of the New Well Background
Report, we state that:
o'There are statistically significant decreasing trends in pH in MW-25,
MW-27, MW-28 and MW-3A with the most significant of these being in
MW-3A. However, in all cases, the pH levels typically fall within the
GWQS range of 6.5-8.5. It is extremely unlikely that low pH tailings
solutions could travel to the perched aquifer without being neutralized by
the calcareous soils underlying the cells. If that were possible, we would
expect to see rising trends in chloride, sulfate, fluoride, uranium, and other
metals that are mobile in low pH solutions. However, we do not see any
such trends.
It is also noteworthy that the most pronounced decreasing trend is in MW-
3,A' which is far downgradient of the Mill's tailings cells. It would be
extremely unlikely for low pH solutions originating in the tailings cell to
maintain their low pH characteristics over a distance of approximately
3,000 feet to MW-3A in a carbonate-rich geologic setting especially
without a much more dramatic decrease in pH being observed at any of
the monitoring wells on the downgradient edge of the tailings cells.
Furthermore, on a review of the pH time plots in all existing wells (see
Appendix D of the [Existing Well] Background Report), there appears to
be a general decreasing trend in pH in all wells. Figure 18 [of the New
Well Background Report] shows results of linear regression analyses for
all site monitoring wells over the same time period used for new wells.
Regression lines trend downward in all site monitoring wells and among
the existing wells the trends are statistically significant in MW-3, MW-12,
Page 14 of 20
ffimiffiaW June 5, 2009
MW-14 and MW-17. The fact that pH is trending downward in all site
monitoring wells indicates that statistically significant decreasing trends in
pH in MW-25, MW-27, MW-28 and MW-3A are not related to any
potential tailings seepage impacts. Instead there is a systematic process
occurring that affects the site as a whole. This process may be a natural
phenomenon related to regional changes or it could be some systematic
change in the way that samples are collected or analyzed. Since 2004
DUSA has been improving its sampling and analysis protocols at the
request of the UDEQ. The pH measurements recorded during this period
were all laboratory measurements. This period also coincides with the
observed decreases in pH suggesting a potential connection with some
laboratory process."
o Existing wells MW-3 and MW-l4 and New Wells MW-3A and MW-27 werc
included in the University of Utah Study, which concluded that the groundwater
in all of those wells have a different geochemical signature than tailings cells
wastewater. This is further evidence that the decreasing trends in pH at the site,
which have led to some relatively low pH readings at various wells including
MW-28, are not due to potential tailings cell impacts.
As a result, we have concluded that low pH in samples from MW-28 have not been influenced
by Mill activities.
PROPOSAL FORMETHOD TO SET GWCLS FORNATURALLYRISING TRENDS
The Background Reports have demonstrated that current concentrations of constituents in
monitor wells at the site represent natural background conditions. This conclusion has gained
additional support through the recent site specific University of Utah Study conducted by
researchers at the University of Utah that suggests that groundwater in site monitor wells
predates uranium milling at the site.
As stated in the Background Reports, for those constituents with a rising (decreasing pH) trend,
the Flow Sheet included as Figure 19 to the Existing Well Background Report (the "Flow
Sheet") indicates that a modified approach to determining the GWCLs should be considered in
order to recognize the fact that GWCLs set at absolute values are subject to being violated as a
result of such trends. solelv due to natural backsround causes.
Page 15 of 20
ffiWlnffiq&W June 5, 2009
Therefore, as contemplated by the Flow Sheet, DUSA proposes the following method for data
sets with naturally rising (decreasing pH) trends:
During each GWDp renewal review, each data set willbe evaluated for increasing
or decreasing trends;
Each statistically significant increasing trend (decreasing pH) will be evaluated to
determine if it is attributable to causes related to Mill operations. In performing
such an evaluation, consideration will be given to the behavior in the well of the
indicator constituents: chloride, sulfate, fluoride and uranium, among other things'
If there have been no statistically significant rising trends in any of the indicator
constituents, then that would be considered to be prima facie evidence that the
trend is due to natural influences. If one or more indicator constituents
demonstrates a significant upward trend, then a further analysis would be
performed to determine if the trend or trends are due to natural influences;
If the trend is determined to be unrelated to Mill operations, evidence for that
determination will be documented in a report attached to the renewal application;
o The report will include a graph of statisticatly significant trending data and an
extrapolation of that trend to the next renewal date; and
o The extrapolated value on the date of the next GWDP renewal will be set as the
GWCL for the period between the two renewals'
This method has the advantage of avoiding unnecessary and expected out-of-compliance
situations that might otherwise cause diversion of unnecessary time and resources by both
UDEQ and DUSA.
This method is protective of human health and the environment because it provides for increased
oversight and review of trending data sets. There is also very little chance that apossible tailings
cell leak would go undetected, given the large number of indicator constituents sampled at each
well. If there were to be a leak in the tailings cells between GWDP renewal periods, then the
first indication of the leak would be sharply increasing trends in the results for the indicator
constituents, such as chloride, sulfate, fluoride and for metals, uranium, regardless of the GWCL
that may have been set for any one constituent using the extrapolation method above' This will
Page 16 of 20
*xffiffilniEltaW June 5, 2009
new wells, it is also present for some constituents in the existing wells (for example, the mean
and standard deviation for chloride in MW-32 are 32.65 mglL and 1.37 mglL, respectively,
which results in a proposed GWCL of 35.39 me/L).
DUSA believes that the use of the Flow Sheet GWCLs in these circumstances could result in an
unwarranted out-of-compliance determination even if true concentrations do not change in the
well. The U.S. Geological Survey has stated that under optimum conditions, the measured
concentrations of major constituents may be within 2-10 percent of the true value and that
constituents present in concentrations greater than 100 mglL can generally be determined with an
accnracy of +/- 5 percent (Hem, 1992). They note that for constituents present in concentrations
of less than 1 mg/L, an accuracy of +/- 10 percent is considered good and as concentrations
approach the detection limit of the method used and in "a11 determinations of constituents that
are near or below the micro-gram-per-liter level, both accuracy and precision are even more
strongly affected by the experience and skill of the analyst."
Thus, if a sample from MW-30, for example, is assigned to a different analyst at the contract lab
for chloride analysis, his result could easily be outside of the 10 percent of the true value
returned under optimum conditions. Assuming that the current average of 125 mglL is the true
concentration in the sample, ffid adding the l0 percent variation possible under optimum
conditions, even an experienced analyst could potentially return a value of 137 mglL and that
result would exceed the GWCL by 9 mglL.
The USEPA recognizes this problem and, in guidelines for inorganic data review, sets limits on
the variability in duplicate analysis that is acceptable from their contract laboratories (USEPA,
2004). These limits are "A control limit of 20Yo for the Relative Percent Difference (RPD) shall
be used for original and duplicate sample values" > five times (5x) the Contract Required
Quantitation Limit (CRQL)." They go on to note that 'oThe above control limits are method
requirements for duplicate samples, regardless of the sample matrix type. However, it should be
noted that laboratory variability arising from the sub-sampling of non-homogeneous soil samples
is a common occuffence. Therefore, for technical review purposes only, Regional policy or
project Data Quality Objectives (DQOs) may allow the use of less restrictive criteria (e.g.,35o/o
RPD, 2xthe CRQL) to be assessed against duplicate soil samples."
USEPA also noted (EPA 1992)that:
Page 18 of 20
**wi'wM.lniffu
w June5,2OO9
'oUnfortunately, all of the available tests for Normality do at best a fair job
of rejecting non-Normal data when the sample size is small (say less than
20 to 30 observations). That is, the tests do not exhibit high degrees of
statistical power."
Based on the above observations, DUSA believes that calculation of GWCLs for data sets with
low variance and with fewer than 30 data points should be based on the highest of: (a) the Flow
Sheet approach; (b) the current fractional approach; and (c) the highest historical value, until 30
data points are available, at which time the GWCLs should be adjusted to reflect the Flow Sheet
approach. For purposes of this approach, we propose that a data set would be considered to have
a "low variance" if the Flow Sheet GWCL does not exceed 120% of the historic mean of the data
set.
This method is not perfect, because there still could be situations where the Flow Sheet GWCL is
the highest of these three indicia, yet the GWCL is very close the historic mean of the data set.
However, this method would be an improvement over the Flow Sheet method alone, and should
reduce the number of unwarranted out-of-compliance situations.
ANOTHER ISSUE
There is also a similar problem for some of the constituents where the GWCL has been set based
on the Flow Sheet, but the number of data points used to calculate the mean and standard
deviation has been small. In a number of these situations, the Flow Sheet GWCL would be set
based on an historic mean for the data set that is very low relative to the GWQS yet the standard
deviation is high relative to the mean. This has already resulted in a number of situations where
current sample results would exceed the GWCL, yet the sample result is very low relative to the
GWQS. The prime example of this is zinc in a number of wells. The proposed Flow Sheet
GWCL for zinc in MW-14 is 35.04 StglL relative to fractional GWCL of 2,500 pglL. The l't
quarter 2009 sample result for zinc in MW-14 is 51 pgil. Similar problems exist for zinc in
MW-28 and29. This is particularly troublesome, when the Flow Sheet approach would result at
the same time in a GWCL of 2,500 for zinc in other wells. By using the Flow Sheet approach for
constituents such as zinc, we are setting the GWCLs too tight in some wells relative to the
GWQS and relative to the GWCLs for the same constituent in other wells. In these situations,
we believe that the approach proposed above for low variance wells should be used until at least
30 data points are available.
Page 19 of 20
rr,Icffi ffilnffirya*.ffi June 5, 2009
REFE,RENCE
EPA (U.S. Environmental Protection Agency), 1992. Statistical analysis of ground-water
monitoring dataatRCRA facilities: Addendum to Interim final guidance, Office of Solid
Waste, Permits and State Programs Division, U.S. Environmental Protection Agency, 401
M Street, S.W. Washington, D.C.20460
Rai, D. and J. M. Zachara. 1984. Chemical affenuation rates, coefficients, and constants in
leachate migration. Volume I: a critical review. EPRI, EA-3356, Research Project 2188-
Figure 1. Sulfate Concentrations in Upgradient MonitorWell MW-{B
Oct-97 Sep-99 Aug-01 Jul-03 Jun-05
Tim e (Years)
May-07 Apr-09 Mar-11 Jan-13 Dc-14
^ 1500J
E5ovrdE
E tooo
Page 20 of 20