DRC-2001-001124 - 0901a06880adecdb

Srmr* c$urmh cALITY
6)ilov z 8 2001
Michael O. l*avitt
Governor
Dianne R. Nielson, Ph.D.
Executive Dircctor
William J. Sinclair
Director
DEPARTMENT OF ENVIRONMENTAL QU
DTVISION OF RADIATION CONTROL
168 North 1950 West
P.O. Box lrl4850
Salt Lake City, Utah 84114-4850
[phone]
[phone] Fax
(801) 536-.1414 T.D.D.
www.deq.state.ut.us Web
Re:
November 28,2OOl
Mr. Harold Roberts
Vice President, Corporate Development
International Uranium (USA) Corporation
Independen ce Plaza, Suite 950
1050 17th Street
Denver, CO 80265
December 31, 1998 Knight Piesold Report on Seepage Flux from Tailings Cetl3 Liner,white Mesa Uranium Mill: Request for Additionar Information.
Dear Mr. Roberts:
Pursuant to your request during our November 14,2001 telephone call, I am providing this
written request to relay several questions that arose during my review of the Decembei 31, 1998Knight-Piesold (KP) report entitled "Methodology for Calculation of Flux Through the Cell 3
Liner, White Mesa Mill". As we discussed previously, review of the Kp report was undertaken,
because predictions from it were used as model inputs in the September ZS,ZOO| Hydro Geo
Chem (HGC) Report on monitoring well effectiveness. Please provide the additional
information requested below.
December 31. 1998 KP Report
1. Composite Liner - a claim is made in the KP report that the liner geometry under Cell 3constitutes a "composite liner", as defined by Giroud and Bonap arte (1213-1198 Kp Report, p.1). Review of the technical literature shows that a composite liner is defined as a Fleiible -
Membrane Liner (FML) that is immediately underlain by a clay with a permeability of less
than 1.0E-4 cm./sec, but usually in the range of 1.0E-6 to l.0E-8 cm/seilBonaparte, et. al.,
1989, p. 18). Review of the March, 1983 Energy Fuels Nuclear (EFN) Cell 3 As-Built
Report shows that the FML bedding layer was constructed of material with the consistency of
"coarse sand" (ibid., p. 3-4). In some cases, the EFN construction used "washed concrete
sand to fill voids crated during rock removal operations" (ibid., p. 3-5). The permeability of
these liner bedding sands would likely fall into a range that is higher than l.0E-4 cm./sec.
Consequently, DRC staff see no support for the KP claim that a iomposite liner exists underCell3.
Mr. Harold Roberts
November 28, 2001 Page2
2. Liner Bedding Permeability: North. East.West Sideslopes and Cell Floor - the KP Report
assumed that the liner bedding material under the North, East, and West sideslopes, and the
Cell 3 floor had a permeability of 1.0E-6 cm/sec. Concern about this low permeability
assumption was raised previously (ll2ll99 DRC letter,p.2). In response to this DRC
concern, IUC responded that there was no documentation available to justify the 1.0E-6
cm/sec liner bedding assumption and that this value was based solely on "engineering
judgment" (2112199 KP Response, p. 2). However, after review of the March, 1983 EFN
Cell 3 As-Built Report, described above, it is very unlikely that the permeability of the liner
bedding material is this low. Available technical literature suggests the permeability of
"coarse" sand should be greater than 1.0E-4 cm./sec, as follows:
A. Clean Coarse Sand - on the order of 1.0E-1 to 1.0E+0 cm/sec (Freeze & Cherry, p.29);
B. Coarse Sand Filter - about 3.58-2 cm/sec (100 fUday, Moulton, p.5Z),
C. Well Graded (SW) Sand - between 4.98-4 to 4.88-2 cm/sec (1.4 to 137 ftlday, Moulton,
p.48).
D. Coarse Sand (repacked) - average of 5.19E-2 cm./sec (1,100 gpdlf(,as determined from
158 samples, Morris and Johnson, p. D20).
Please revise the FML bedding permeability assumption to include a value greater than 1.0E-
4 cm/sec, or provide additional justification for 1.08-6 cm/sec value used.
3. Liner Design Case and Equations - previously the DRC asked for additional justification of
the liner design case used by KP in selection of the equations that govern seepage flux thru
defects in FMLs (ll2ll99 DRC letter , p. 2). In response to this request, ff eipiained that
(2112199 KP letter, pp. 2 and 3):
A. The spreadsheet model used in the KP report was based on Geomembrane Liner Design
Case 3a, as found in Schroeder et. al.
B. The KP model ignored the low permeability tailings above the FML,
C. The appropriate design case is determined by the "controlling" soil layer in the profile,
D. The equations KP used apply equally well regardless of whether the controlling soil layer(1.0E-6 cm/sec) is a tailings layer immediately above the FML, or a bedding layer
immediately below the FML.
Regarding this IUC response, DRC staff have made the following findings:
E. The referenced Geomembrane Liner Design Case 3a is defined as follows, descending
order (Schroeder, et. al., pp.79 and 95):
A high permeability soil layer (K >= 1.OE-l cm/sec),
The FML, and
1)
2)
Mr. Harold Roberts
November 28,2001
F.
Page 3
3) A low permeability soil layer (K < 1.0E-4 cm/sec).
We agree that the Giroud and Bonaparte equations still apply when the controlling soil
layer is immediately above the FML. However, the tailings layer referred to is NOT in
direct contact with the FML.
Other HELP model Geomembrane Liner Design Cases appear to better represent the field
conditions under Cell 3. The February 12, 1999 KP response did not account for the
presence of the slimes drain layer and associated piping network constructed immediately
above the Cell 3 FML to de-water the overlying tailings. As a result, the tailings cannot
be used as the "controlling" soil layer for purposes of assigning a liner design case, or
determining governing equations to predict FML leakage. Depending on the field
permeability of the slimes drain layer, a hydraulic discontinuity, or head break could exist
in the profile below the tailings; especially if this layer is pumped to remove tailings
leachate from the system. Based on DRC review of the Schroeder et. al. document, it
appears that Geomembrane Liner Design Cases 2a,2b, or 2c would be more applicable to
Cell 3, as summarized below (with layers described in descending order):
Design Case 2a: (1) a medium permeability soil layer (1.0E-4 to 1.0E-l cm/sec),
(2) the FML, and
(3) a high permeability soil layer (>= 1.0E-1 cmlsec)
Design Case 2b: (1) a high permeability soil layer (>= 1.0E-1 cm/sec)
(2) rhe FML, and
(3) a medium permeability soil layer (1.0E-4 to 1.08-1 cm/sec)
Design Case 2c: (1) a medium permeability soil layer (1.08-4 to 1.0E-1 cm/sec)
(2) the FML, and
(3) a medium permeability soil layer (1.0E-4 ro 1.0E-1 cm/sec)
Please revise the equations used in the December 31, 1998 KP report to incorporate
equations from Liner Design Case 2, as defined by Schroeder, et. al.
A relatively high permeability is suggested for the slimes drain layer by the May, 1981
DAppolonia Consulting Engineers (DCE) Cell 3 design report, which shows this layer
was to be constructed of "coarse" tailings, 1.5 feet thick over the sideslope areas, and 1.0
foot thick over the Cell 3 floor area (ibid., Sheet 4 of 5). Later, the March, 1983 EFN
Cell 3 As-Built Report explained that (ibid., pp.3-i and 8):
1) At the time of construction the "coarse" tailings available were only enough to
cover about 30Vo of the cell floor.
2) EFN covered the remainingTj%o of the Cell 3 liner area with excavated soil from
stockpiles located East and West of Cell 3, and
G.
,Mr. Harold Roberts
November 28,2001
4.
Page 4
3) These cover materials were placed on the FML using front end loaders and769
CAT haul trucks after construction of a haul ramp in the Southwest corner of the
cell.
No information was provided in the March, 1983 EFN Cell 3 As-Built Report to
document the gradation or permeability of the cover materials used from the nearby soil
stockpiles. However, because the layer was designed to de-water the overlying tailings, it
is plausible that the permeability of this material is rather high, perhaps greater than 1.0E-
3 cm/sec. Please increase and justify the permeability assigned to the FML bedding layer
and recalculate the seepage flux from the cell 3 disposal facility.
Recommendation for Use of EPA HELP Model - because the tailings layer in the Cell 3
profile will continue to limit the amount of leachate flux made available to the slimes drain
layer, which in turn accumulates on the FML, the spreadsheet equations in the Decemb er 31,
1998 KP Report need to be modified to: 1) add predictions of seepage flux from the tailings
layer, and 2) provide predictions of resulting head on the FML to be used in calculation of
FML leakage rates. To simplify this effort, DRC staff recommend IUC consider use of the
EPA HELP model for this purpose. We also recommend that a meeting be held to discuss
the construction of a conceptual model and other input values for this simulation.
Vapor Diffusion: Equivalent PVC FML Permeability - the December 31, 1998 KP Report
lists an equivalent permeability for the PVC membrane, 4.42E-8 inch/day or l.3E-12 cm/sec(ibid., Table I and Appendix B). Unfortunately, this equivalent permeability value is
unjustified. Information provided by Giroud and Bonaparte (1985) and Schroeder, et. al.
(Table 8,p.77) demonstrates that the equivalent permeability of PVC is z.OB-ll cm/sec,
which is about 15 times more perneable than the-December 3l, 1998 KP Report value.
Please correct this permeability value and revise the seepage flux calculationi for Cell 3.
Assumed Flaw Rate: Pinholes and Installation Defects - the December 12,1998 Kp Report
cited research by Giroud and Bonaparte (1989) and assumed the following FML flaw arias
and rates of occurrence (ibid., pp.2,4, andTable 1):
5.
6.
Flaw Tvoe
Pinholes
Installation defects
Radius Circular Area Defect Rate
l fladacre
2 flaws/acre
0.02 in (0.05 cm) 0.0013 jn2 (0.00g cm2)
O.22in(0.55 cm) 0.15 in2 (l cm2)
However, careful review of the Giroud and Bonaparte reference shows that installation
defects of l0 flawslacre or more are warranted when FML construction quality assurance islimited to an engineer spot-checking the work of a geomembrane installei (ibid., pp. 6a-65).
Schroeder and others (1994b) also reinforce this recommendation and add that the 1
flaw/acre rate is only applicable with "intensive quality controVquality assurance monitoring"(ibid., p. 78). After review of the May, 1981 DCE Cell3 Design andihe March, 1983 EFNCell 3 As-Built Reports, DRC staff have concluded that a installation defect flaw rate of l0
flaws/acre or more is appropriate, based on the following findings:
Mr. Harold Roberts
November 28,2001 Page 5
A. Limited FML COA: Destructive Testing - FML construction quality assurance was
limited to destructive testing of the FML membrane on infrequent intervals. A suite of
peel, elongation, ten-sile strength, tear, and other destructive tests were performed once
for every 250,000 ft2 of factory fabricated FML liner (5/81 DCE Report, Appendix B,
Table 1 and p. 3-7). In addition, single tear strength tests were conducted on field seam
samples on a basis of 1 for every 100,000 ft2. Based on an approximate 3.5 million
square feet of FML surface under Cell 3, this would suggest that 14 suites of tests would
have been conducted on the factory fabricated liner material and 35 tear strength tests on
the field seams constructed. For a disposal cell of such large size, it appears that low
number of destructive tests qualifies as spot-checking.
B. Limitations of Air-Lancing - non-destructive testing of field-constructed seams was
limited to air lancing (3/83 EFN Report, p. 3-6). Unfortunately, this technique can only
find a seam defect if it is exposed at the front edge of a seam and is described as "strictly
a contractor/installer's tool to be used in a construction quality control (CQC) manner." It
is not recorlmended for construction quality assurance purpor"s (Koerner, pp. a99-500).
In addition, air lancing has the potential to provide a false negative response, where a
pocket or channel-shaped defect in the seam adhesive could easily occur behind the front
edge of the seam. Such a defect could run along the seam for a considerable distance and
never be detected by the air lance. In turn, if the upper surface of the seam above this
pocket or channel were to be punctured or encounter another defect, an avenue would be
created for leakage to pass thru the FML. Such areas of incomplete or poor seam
adhesion pose points of weakness where defects could form later, particularly after the
FML is loaded. On this basis, DRC staff have concluded that air lancing does NOT
qualify as "intensive quality assurance monitoring", and therefore the KP assumed defect
rate of 2 installation flaws/acre is currently unsupported.
C. Puncture Potential During Installation - a significant potential exists for FML puncture
during installation. The original engineering design called for the slimes drain layer to be
made from coarse sand-sized tailings discharge from the mill and segregated by a cyclone
separator (5/81 DCE Report, Appendix B, p. 3-7). However, the March, 1983 EFN As-
built Report stated that there was only enough coarse tailings available at the time of
construction to cover about 30Vo of the Cell 3 floor area. Instead, EFN constructed the
remaining TOVo of the slimes drain layer with soils derived from the Cell 3 excavation
(ibid., pp.3-7 and 8). Unfortunately, no information is provided in the March, 1983 EFN
As-built Report regarding several critical FML construction issues, including:
1) Soils excavated from the foundation of Cell 3 could easily contain angular rock
fragments that could puncture the FML during placement. Although the May,
1981 DCE design report stipulated that the slimes drain layer not contain any
sharp, angular pieces (ibid., Appendix B, p. 3-7), no description was included in
the March, 1983 EFN As-Built Report to document how the excavation soil
stockpiles were screened or otherwise treated to remove or eliminate angular rock
fragments. These types of defects, caused by FML cover soil placement, cannot
Mr. Harold Roberts
November 28,2OOI
D.
Page 6
be observed by construction quality assurance personnel (Giroud & Bonaparte,
1989, p. 64).
2) No effort was made in either the May, 1981 DCE design report or the March,
1983 EFN as-built report to determine the maximum pressure or load that could
be applied to the FML without damage or puncture. Consequently, the thickness
of the slimes drain layer needed to protect the FML from static and dynamic loads
from haul trucks, front-end loaders, or bull dozers appears to have never been
quantified. Determination of the thickness of this "protective cushion" is essential
to avoiding punctures during construction, and is especially important for PVC
membranes that are much more prone to point source puncture than other FML
materials (EPA, p. 31). Again, FML damage caused by such equipment traffic
cannot be observed by construction quality assurance personnel (Giroud &
Bonaparte, 1989, p. 64).
3) Potential for impact damage from apparent dumping of slimes drain cover soils.
Little description was provided in the March, 1983 EFN as-built report to explain
how slimes drain soil was supplied to the low ground pressure bulldozer used to
spread a progressive pad of soil. Apparently, front-end loaders and haul trucks
were employed to bring the excavated soil or coarse tailings to where they were
needed by accessing the Southwest comer of Cell 3. Apparently, no liner was
built in this area at the time of the haulage to avoid FML damage by repetitive
truck traffic (ibid., p. 3-8). However, from the photographs provided in the As-
Built Report (Appendix E), it is apparent in at least I photo that windrows of
slimes drain cover soils have been end dumped on the FML, either by truck or
front-end loader. Such dumping has the potential to create large dynamic stress
and punctures thru the PVC liner material; especially if angular rock fragments
are found in the excavated soils for cover material. As before, FML damage
caused by dropping loads of cover soil cannot be observed by construction quality
assurance personnel (Giroud & Bonaparte, 1989, p.64).
Apparent Lack of COA/OC Controls for FML Wrinkles at Seams - no construction
specifications were provided for FML Wrinkles. Review of both the May, 1981 DCE
design report and the March, 1983 EFN as-built report show no mention made of
preventing wrinkles in the FML during construction. This is especially important near
field seams, where if a FML wrinkle were to impinge on a seam at angle ard become
incorporated into the seam, a bypass conduit could be created that would allow tailings
leachate to be discharged.
Effects of FML Aging - no consideration was given to the effects of FML aging on liner
defect rate. The December 31, 1998 KP Report does not include any discussion of the
effects of FML aging. Plasticizer compounds used in the manufacture of PVC liners are
prone to leaching (Koerner, p. 510). The loss of the plasticizer in turn makes the FML
more brittle and susceptible to damage. Under this scenario stress cracks can develop in
E.
Mr. Harold Roberts
November 28,2001
7.
8.
PageT
a FML. Add to this the increased loads on the liner as tailings are continuously disposed
into Cell 3, and it is possible that additional FML flaws could develop.
F. Poor Chemical Resistance Effects - no consideration is given in the December 31, 1998
KP Report to chemical resistance of PVC in the presence of the tailings contaminants. In
general, PVC liners exhibit poor resistance to petroleum hydrocarbons and chlorinated
solvents (Koerner, p. 389). Historically, significant amounts of kerosene/diesel fuel and
small quantities of chlorinated solvent have been discharged to the IUC tailings cells. In
addition, no information was provided in either the May, 1981 DCE design report or the
March, 1983 EFN As-Built Report on the chemical resistance of the PVC adhesive to
these same chemicals. Adverse reactions of these organic compounds with the PVC liner
material or seam adhesive could easily cause the formation of additional liner defects.
After consideration of the above factors, it appears that the FML installation defect rate of 2
flaws/acre is grossly under-estimated, and should be increased to at least 10 flaws/acre, if not
more. Please revise the December 31, 1998 KP Report accordingly and re-submit.
FML Bedding Thickness: North. East. and West Sideslopes - the December 31, 1998 KP
Report cites a 6-inch thickness for the bedding layer under the FML for the North, East, and
West sideslope areas of Cell 3. [n contrast, the May, 1981 DCE design report called for a 1-
foot thick bedding layer for these sideslopes (ibid., Sheet 4 of 5). Please revise your model to
include the correct thickness.
Justification of Extrapolation to Cells I and 2 - the spreadsheet model presented in the
December 31, 1998 KP Report focused specifically on the physical characteristics of Cell3.
Previously IUC has made claims that the Cell 3 seepage predictions are applicable to Tailings
Cells 1 and2 (Ill23l98 KP letter, pp. 10-11). However, after consideration of the myriad of
independent design and construction details, it appears that the extrapolation of the Cell 3
analysis to these other 2 disposal cells is unwarranted. Please provide a justification for why
any Cell 3 analysis is applicable to the other 2 cells, after careful consideration of several key
issues, including, but not limited to:
A. Gradation and permeability of component layers, including but not limited to the FML
bedding layer, slimes drain layer, etc.
B. Applicable geomembrane liner design case,
C. Cell geometry, including total depth, internal slopes, layer thickness, grade and shape of
cell floor, etc.
D. Effects of differing cell geometry on average head on FML, different load on FML and
resulting soilJiner contact, etc.
Mr. Harold Roberts
November 28,2001 Page 8
E. Construction techniques used to excavate and prepare final grades, prepare FML bedding
layer, emplace FML cover layers or other overlying material or equipment without
damage to underlying geomembranes, etc.
F. Techniques to measure and monitor construction progress and compliance with
engineering specifications (e.g. gradation tests, soil permeability tests, etc.).
G. FML construction techniques, including but not limited to methods, equipment, and
training for: FML transport, placement, wrinkle control, seam construction, and FML
destructive and non-destructive quality assurance/quality control.
H. Effects of FML aging to leaching of plasticizers, or chemical interaction of the FML or
seam adhesives with tailings leachate contaminants.
I. Pumping rates from the slimes drain layer (Cell 2), or leak detection layer from either
Cells I or 2.
9. Need to Submit Sensitivity Testing Results - the December 31, 1998 KP Report described
sensitivity testing conducted on the spreadsheet model and summarized the results thereof.
Unfortunately, the report failed to include the results of this sensitivity testing. For future
simulations, please provide the input values and output results for all sensitivity test work
conducted.
Please resolve the above information request in order to allow completion of our review of the
September 25,2001 HGC Report. If you have any questions or comments, please call me at
[phone]. I appreciate your assistance in this matter.
Respectfully,
A il""k-
bn B. Morton
LBM:lm
Attachments (1)
cc: Stewart Smith, HGC
Bill von Till, NRc-Washington, D.C.
F:\...\Cell3Flux.doc
File: IUC Infiltration Modeling Reports
Mr. Harold Roberts
November 28,2001 Page 9
References
Bonaparte, R., J.P. Giroud, and B.A. Gross, 1989, "Rates of Leakage Through Landfill Liners",
from Geosynthetics Conference Proceedings, Vol. 1, February 2l-23,1989, San Diego,
CA, pp. 18 - 29.
DAppolonia Consulting Engineers, Inc., May, 1981, "Engineer's Report Second Phase Design -
Cell 3 Tailings Management System", unpublished consultants report, approximately 20
pp., 1 figure, 3 appendices.
Energy Fuels Nuclear, March, 1983, "Construction Report Second Phase Tailings Management
System", unpublished company report, l8 pp., 3 tables,4 figures, 5 appendices.
Freeze, R.A., and J.A. Cherry,lg'19, Groundwater, Prentice Hall, Englewood Cliffs, NJ, 604 pp.
Giroud, J.P., and R. Bonaparte, 1985, "Waterproofing and Drainage: Geomembranes and
Synthetic Drainage Layers" from Geotextiles and Geomembranes - - Definitions,
Properties, and Design - Selected Papers, Revisions, and Comments, 2nd Ed., Industrial
Fabrics Association lnternational, St. Paul, MN.
Giroud, J.P. and R. Bonaparte, 1989, "Leakage thru Liners Constructed with Geomembranes - -
Part I. Geomembrane Liners", Geotextiles and Geomembranes, Elsevier Science
Publishers, Vol. 8, No. 1, pp.27-67.
Knight Piesold LLC, November 23, 1998, "Evaluation of Potential for Tailings Cell Discharge -
White Mesa Mill", unpublished consultants letter report from Messrs. Samuel Billin and
Roman Popielak to Anthony Thompson, l5pp., 6 figures.
Ifuight Piesold LLC, December 31, 1998, "Methodology for Calculation of Flux Through the
Cell 3 Liner, White Mesa Mill", unpublished consultants report, 5 pp., I figure,I table, 3
appendices.
Knight Piesold LLC, February 12,1999, "Response to UDEQ Comments on Methodology
Assumptions", unpublished consultants letter report ,4 pp.
Koerner, R.M., 1990, Designing with Geosynthetics, 2nd Ed., Prentice Hall, Englewood Cliffs,
NJ, 652 pp.
Morris, D.A. and A.I. Johnson,1967, "Summary or Hydrologic and Physical properties of Rock
and Soil Materials, as Analyzed by the Hydrologic Laboratory of the U.S. Geologic
Survey 1948-60", usGS Geological Survey water-Supply paper 1g39-D, 42 pp.
Moulton, L.K., August, 1980, "Highway Subdrainage Design", West Virginia University for the
Federal Highway Adminisrration, FHWA-TS-90-224, 162 pp.
Mr. Harold Roberts
November 28,2001 page l0
Schroeder, P.R., T.S. Dozier,P.A.zappi. B.M. McEnroe, J.w. Sjostrom, and R.L. payton,
September,1994b, "The Hydrologic Evaluation of Landfill Performance (HELP) Model,
Engineering Documentation for Version 3", IJ.S. Army Corps of Engineers for U.S.
Environmental Protection Agency, EPA/600/R -941 168b, I I 6 pp.
U.S. Environmental Protection Agency, August, 1989, "Requirements for Hazardous Waste
Landfill Design, Construction, and Closure", Technology Transfer Seminar Publication,
EP N 625 I 4-89 1022, 127 pp.
Utah Division of Radiation Control, January 2I, !999, "Methodology Assumptions Used for
Calculation of Flux Through the Cell 3 Liner, White Mesa Uranium Mill", agency
request for additional information, 3 pp.
FEBrf1ge
Michael O. Leavitt
Govcmor
Dianne R. Nielson, Ph.D.
Exeutivc Director
William J. Sinclair
Diretor
DEPARTMENT OF ENVIRONMENTAL QUALITY
DIVISION OF RADIATION CONTROL
168 Norrh 1950 West
P.O. Box 144850
Salt Lake City, Utah 841144850
(801) 5364250 Voice
(801) 5334097 Fax
(801) 5364414 T.D.D.
e/lO/
February ll,1999
David C. Frydenlund
Vice President and General Counsel
lnternational Uranium (USA) Corporation
Independence Plaza, Suite 950
I 050 Seventeenth Street
Denver, CO 80265
SUBJECT:February 4,1999 Letter to Mr. Don Ostler
Director - Division of Water Quality
Utah Department of Environmental Quality
Dear Mr. Frydenlund:
The Utah Department of Environmental Quality, Division of Radiation Control (DRC) has
received the subject letter via facsimile on February 4,1999. As we indicated in our meeting
with you on December I l, 1998, the DRC has many concerns related to groundwater protection
from potential seepage from the tailings impoundments at the White Mesa Mill. These concerns
were further clarified in subsequent letters to International Uranium Corporation (lUC) on
January 8, 1999 and January 21,1999. As requested by you in the subject letter, the DRC's
concerns are stated again below.
Tailings Impoundment Liner Systems
The DRC is not convinced that the bottom liner systems for tailings impoundment cells l, 2, and
3 at White Mesa are adequate for minimizing discharge of tailings leachate to groundwater.
DRC staffreviews of the November 23,1998 and December 31, 1998 Knight Pi6sold modeling
reports indicated that a number of assumptions were made in the modeling effort without
appropriate supporting documentation. As stated in the January 21, 1999letter to IUC, these
assumptions have critical implications associated with the analytical model inputs and
corresponding output liner leakagapredictions. Without the supporting documentation, these
assumptions and the corresponding model predictions cannot be confirmed. As we indicated in
the January 21,1999letter, the DRC cannot veriff the predictions rendered by the modeling
effort without the requested information. In addition, the DRC does not believe that a best-case
scenario for liner leakage is valid as assumed in the Knight Pi6sold modeling effort. A more
realistic approach should be employed which considers sensitivity analyses of key model input
parameters to provide a range of possible predictions instead of a single best-case scenario.
David C. Frydenlund
February 11,1999
Page2
Leak Detection Systems
Similarly, the DRC does not have confidence in the efficiency of the leak detection systems for
tailings impoundment cells 1,2, and 3 at the White Mesa Mill. The leak detection systems have
a high potential for undetected leakage for two primary reasons. First of all, an efficient leak
detection system must have a secondary low-permeability barrier below the primary low-
permeability liner to accumulate and divert leakage to the leak collection pipe. However, the
leak detection systems for these cells consists of a primary 3O-mil PVC geomembrane on top of a
6-inch thick layer of reworked sandstone bedrock which is supposed to function as a secondary
low-permeability barrier. In the December 31, 1998 Knight Pidsold modeling report, the
reworked sandstone bedrock material is assigned a saturated hydraulic conductivity of lxl0-6
centimeters per second. Because the reworked bedrock layer beneath the PVC geomembrane is
the controlling soil layer, there needs to be some quantitative justification for using this value.
Secondly, should a leak occur that is large enough to pool and accumulate on top of the
reworked sandstone bedrock material, it would have to travel over a long horizontal distance to
reach the collection pipe and be detected at the downslope end of the cell. During this horizontal
travel path across the impoundment, vertical seepage losses through the reworked sandstone
material will further reduce the effectiveness of the detection system to report small leaks.
Consequently, only the largest catastrophic leaks will be detected by the current leak detection
systems for these cells. Non-catastrophic seepage from these disposal cells will travel vertically
through the vadose zone with the potential for reaching the water table aquifer. Once reaching
the water table, Ieachate contamination will not be detected until reaching the groundwater
monitoring wells which could take many years to occur.
Fracture Flow Potential
Accelerated travel times of tailings fluid leakage via secondary permeability from joints and
fractures was not addressed in either the November 23, 1998 or the December 3 l, 1998 Knight
Pi6sold reports. As reported in the February 1993 UMETCO Groundwater Study of the White
Mesa Facility (Peel Environmental Services, 1993) fluid travel times to the perched aquifer from
pond liner leakage were estimated based on site-specific boring and well test data. These data
indicate that it is likely that seepage under positive pressrue could be in direct contact with
vertical joints at the base of the ponds. In this case, seepage would occur as localized saturated
flow through joints within the Dakota Sandstone into the Burro Canyon perched aquifer.
Consequently, travel times for tailings pond leakage to the perched aquifer could be as short as a
few weeks through joints directly in contact with tailings solutions to approximately 60 years for
partially saturated flow conditions (Peel Environmental Services, 1993). This is in sharp contrast
to the 1,300 year travel time estimated in the November 23,1998 Knight Pi6sold report.
David C. Frydenlund
February 1 l, 1999
Page 3
Deficient Groundwater Monitoring Program
Another concern the DRC has is the groundwater monitoring program which we find to be
inadequate. Presently, the groundwater detection monitoring program employed at the mill
analyzes only for the inorganic constituents of chloride, potassium, nickel, and uranium. Based
on the constituents that are typically present in I le.(2) byproduct material from acid leach
processing of natural uranium ores, other conservative more mobile "smoking gun" leakage
parameters such as antmonia, nitrate, nitrite, molybdenum and sulfate should be included. In
addition to inorganics associated with acid leach processing of natural uranium ores, IUC has
introduced a number of additional organic constituents from alternate feed materials such as the
Ashland 2 FUSRAP material which are not common constituents of 11e.(2) byproduct material
from natural ores. The current groundwater detection monitoring program at the mill does not
include any organic compounds and is therefore inadequate for detecting releases of these
compounds to the perched aquifer. As indicated by analytical results of soil samples in the
Remedial Investigation Report, pre-excavation sampling activities, and receipt sampling
activities at the mill, there are a wide range of volatile and semi-volatile organic compounds
mixed with the Ashland 2 material including chlorinated solvents. Chlorinated solvents have
much different chemical characteristics than petroleum hydrocarbons which make them a serious
threat to groundwater systems. In particular, the high density and low viscosity of chlorinated
solvents enables them to migrate downward through vertical fractures in bedrock systems such as
the one beneath the White Mesa tailings impoundment.
I hope this letter has clarified our concerns to IUC regarding Utah DEQ's request for a groundwater
discharge permit. The State will notify you prior to taking any formal enforcement action against
IUC. If you have any questions about this letter, please call me or Rob Herbert at [phone].
Sincerely,@wWilliam J. Sincla{r, pirector
Division of Radiahdn Control
cc:Fred Nelson, Utah Attomey Generals Office
Don Ostler, P.E., Director,OEQ-DWQ
Dianne Nielson, Ph.D., Executive Director, UDEQ
F: RHERBERT\wP\W}fl TE MEsA\FRYDENLUND.LTR
J
qrh
t'JAh{ 2 i lBgg
Michael O. Leavitt
Govemor
Dianne R. Nielson, Ph.D.
Executivc Dircctor
William J. SinclairDiretor
DEPARTMENT OF ENVIRONMENTAL QUALITY
DIVISION OF RADIATION CONTROL
168 North 1950 West
P.O. Box I44850
Salt Lake ciry, utah 841144850
(801) 5364250 Voice
(801) 5334097 Fax
(801) s3644r4 T.D.D.
January 21,1999
Michelle R. Rehmann
Environmental Manager
International Uranium (USA) Corporation
Independen ce Plaza, Suite 950
I 050 Seventeenth Street
Denver, CO 80265
SUBJECT:Methodology Assumptions used for Calculation of Flux Through The Cell 3 Liner
White Mesa Uranium Mill
Dear Ms. Rehmann:
The Utah Department of Environmental Quality, Division of Radiation Control (DRC) has
received the subject report prepared by Knight Pi6sold LLC and dated December 3 l, 1998. A
review of this report by DRC staff indicates that a number of assumptions were made without
appropriate supporting documentation. These assumptions have critical implications associated
with the analytical model inputs and corresponding output liner leakage predictions. Without the
supporting documentation, these assumptions and the model predictions cannot be confirmed.
To enable the DRC to proceed with a review of the modeling effort and verifu the predictions
rendered, please provide the following information .
. The geomembrane defect frequencies and sizes used in the modeling effort assumed
intensive quality assurance/quality control (QA/QC) monitoring during liner construction.
To validate this assumption, extensive documentation of construction QA/QC is needed.
Please provide the DRC with the construction QA/QC documentation to ensure the
following:
- Quality control was provided by the geomembrane installer following a rigorous
construction quality control manual;
- Quality assurance was provided continuously by an third party independent firm;
All geomembrane panel seams were tested after installation to find and repair all
seam defects;
Description and documentation of steps were taken in preparation of the soil
subgrade below the 30-mil synthetic PVC liner. In particular, please provide:
Michelle R. Rehmann
January 21,1999
Page 2
l) Maximum and average particle size allowed on the soil subgrade prior to
installation of the 3O-mil synthetic liner. Please provide gradation testing
results to support said claims.
2) Description of equipment and methods used to remove over-sized
materials (e.g. rock clasts, soil clods) from the soil subgrade prior to
placement of the 30-mil synthetic liner.
- monitoring of moisture, ambient temperature, seaming temperature, seam
contamination by dust or dirt, and remedial activities were conducted and
documented; and
- all connections between geomembranes and appurtenances were tested to find and
repair defective connections.
' As stated in the Summary of Model Assumptions on page I of the subject report,"The
soil layer underlying the geomembrane has a saturated hydraulic conductivity ranging
from lxl0't (fo, sand) to lxl0'6 cm/s (or reworked bedrock materials)." Because the soil
layer beneath the geomembrane is the controlling soil layer, there needs to be some
quantitative justification for using these values, particularly for the reworked bedrock
materials of the Dakota Sandstone. Please provide the DRC with documentation for
quantitative results of permeability and compaction tests to justiff the hydraulic
conductivity values used in the analytical modeling effort.
' As indicated above, the DRC questions the validity of the hydraulic conductivity used for
the soil layers underlying the geomembrane. Consequently, the DRC questions whether
the appropriate Geomembrane liner Design Case and corresponding equations of
Schroeder and others (1994) was applied in the modeling effort. Please justify the Design
Case that was used in the leakage analytical modeling effort.
. Accelerated travel times of tailings pond leakage via secondary permeability from joints
and fractures was not addressed in either the November 23, 1998 or the December 31,
1998 Knight Pi6sold reports. However, site-specific well test data from a previous
groundwater study of the White Mesa mill indicated the presence ofjoints and fractures .
Please justify why the potential effects ofjoint and fracture flow were not incorporated in
the seepage analytical modeling effort.
Michelle R. Rehmann
January 21,1999
Page 3
We appreciate the opportunity to review the Knight Pi6sold report and look forward to working with
you in the future. If you have any questions about this letter, please call me or Rob Herbert at (801)
536-4250.
k^ffi;F
Division of Radiation Control
WJS:RFH:rh
cc: Don Ostler, P.E., Director, DEQ-DWQ
F:RHERBERT\WP\WHTE MESA\PTESoLD.LTR
ffi
UDEO RFD
Michrcl O. l.cavitt
0ovcmgt
Diannc R. Niclton, Ph.D.fxcutivr Dirrcror
Wlllianr J, Sinclcir
Diroctor
OL
it:t
ii
ii
CONTR ID:801-533-4097 FEE(o
DEPARTMENT OI: ENVIIIONMENTAI.. QUALITY
DIVISION OF RADIAI'ION CON'I'ROL
168 Nonh 1950 Wcrt
P.O. 8ox 1d4850
Salt Ldic City, Utuh 8c I 14.4850
(801) 515-{250 Voicc
(E0l) 5314097 Fu
(80r) 53644r4 T.D.D,
03 '99
o
8 :06 No .001
oltll 2 t Ugt
P .02
';/
,*4
January 21, 1999
Michelte R. Rchmann
Environmcntal Manager
International Uranium (USA) Corporation
Indepcndcnce PIaza, Suite 950
I 050 Seventeenth Strect
Denver, CO 80265
SUBJECT; Methodology Assumptions used for Calculation of Flux Through The Cell 3 Liner
White Mesa Uranium Mill
Dear Ms. Rehmann:
The Urah Department of Environnrental Quality, Division of Radiation Control (DRC) has
received the subject report preparod by Knieht Pidsold LLC and dated December 3l, 1998. A
review of this report by DRC staff indicates that a nunrber of assumptions were made without
appropriate supporting documentation. These assumptiorts have critical implications associated
rvith thc analytical model inputs and corresponding output lincr lcakagc predictions. Without the
supporting documentation, these assumptions and the modelpredictions cannot be confirmed.
To cnable the DRC to proceed with a rcvicw of thc modcling effort and verify the predictions
rendcred, please provide the following information .
. The geomembrane defect frequencies and sizes used in the modeling effort assumed' intensive quality assurance/quality control (QA/QC) monitoring during liner cortstruction.
To validatc this assumption, extensive documentation of construction QA/QC is necdcd.
Please provide the DRC with thc consuuction QA/QC documentation to ensure the
following:
Quality control was provided by the geomembrane installer following a rigorous
construction quality'control manual I
Quality assurance was providcd continuously by an third party independent finn;
All geomembrane panel seams wqre tested after installation to find and repair all
seam defects;
Description and documentation of stcps were takcn in preparation of the soil
subgrade below the 30-nril synthctic PVC lincr. In particular, plcase providc:
UDTO RRD CONTROL
'a
80 1 -533- 4097
,
ID:FEB 03 '99
o
8 :07 No .001 P .03
Michcllc R. Rehnrann
Jnrruary 21, 1999
Page 2
t) Maximum and avcragc particlc sizc allowcd on the soil subgradc prior to
irrstallation o[ the 30-mil synthetic liner. Please provide gradation testing
results to support said claims.
2) Description of equipment and nrethods used to rcrnove over-sized
materials (e.g. rock clasts, soil clods) fronr the soil subgrade prior to
placement of the 3O-mil synthetic liner.
- monitoring of moisture, ambient temperature, seanrirrg ternperature, seam
contaurination by dust or dirt, and rcmcdial activities were conducted and
documented; and
- all connections between geomcmbranes and appurtenBnces were tested to firrd and
repair defective connections.
As statcd in thc Summary of Model Assumptions on page I of the subject report, "Ihe
soil layer underlying the goomembrane has a saluraled hydratiic conductivity ranging
from lxttt (for sand) to lxlO6 cmls (or reworked bedrock malerials)." Bccause the soil
Iayer beneath the geomcmbrane is the controlling soil layer, there needs to be some
quBntitative justification for using thcse values, particularly for the reworked bedrock
materials of the Dakota Sandstone. Plcasc provide the DRC with docunrentation for
quantitative results of pernteability and compaction tasts to justify thc hydraulic
conducrivity values used irr the analytical modeling effort.
As indicated above, the DIIC questions thc validity of the hydraulic conductivity used for
the soil layers underlying thc gcomernbranc. Conscqucntly, the DRC questions whether
the appropriate Geomembrane liner Design Case and corrcsponding equations of
Schrocdcr and othcrs (1994) was applied in the rnodeling effort. Pleasc justify the Design
Case that was used in the leakage analytical nrodeling cftbrt.
Accelerated travel times of tailings pond leakage via secondary pcrmeability from joints
and fractures was not addresscd in eithcr the November 23, 1998 or the Dccember 3 l,
1998 Knight Pidsold reports. I'lowever, sitc-spccific well test data fronr a previous
groundwater study of the White Mesn nri[[ indicated thc presence ofjoints and fractures .
Please justify wlry the potcntial effbcts ofjoint and fracturc flow wcrc not irrcorporated in
tlre seepage analytical nrodeling effort.
UDE.O RRD CONTRT]L 80 1 -533- 4097oID:
(
FEE 03'99o
8 :07 No .001 P .04
Michcllc R, Rehnrann
January 21,1999
Page 3
We appreciate the opportunity to review the Knight Pidsold report and look forwarcl to working with
you in the future. If you have any questions about this letter, please call me or Rob Herbert at (E0l)
535-4250.
Sinccrcly,
WJS:MH:rh
ccr Don Ostler, P.8., Direclor, DEQ-DWQ
t:kllBt0EtflwAwruf0 MEtA\nEsoLD.LYt
FAX
DATE: February 3, 1999 TIME: 8:02am PAGES: 5 + COVEr
TO: Michelle Rehmann FROM: Rob Herbert
IUSA
FAX: [phone]
Utah Division of Radiation
Control
168 North 1950 West
Salt Lake City, Utah 84116
[phone] VOX
SUBJ: Requested associated with White Mesa Mill
Hi Michelle. Enclosed are the following letters that Bill asked me to fax you:
(1) January 2l,1999letter to Michelle Rehmann from Bill Sinclair regarding Methodology
Assumptions used for calculation of flux through the cell 3 liner at the Whid Mesa uraniummill.
(2) January 29, lggg letter from Dianne Nielson to Utah State Senate President Lane Beattie
concerning the White Mesa uranium mill.
I will send file copies of the originals to you via mail.
INrnnNRrro*o" I
UneNruu (use)
ConponATroN
lndependence Plaza, Suite 950 . 1050 Seventeenth Street r Denver, CO 80265 .
February 3,
303 628 7798 (main) . 303 389 al25 (fax)
1999
VIA FEDERAL EXPRESS
William J. Sinclair, Director
Division of Radiation Control
State of Utatr Department of Environmental Quality
168 North 1950 West
P.O. Box 144850
Salt Lake city, UT 84114-4850
,'{' fiB r';,,,
'!^"ti'nuu
l,u;orii ),'uonrroj.,
'<il - ;-'-'/'
Re: Your letter of January 21, 1999 Regarding Methodology Assumptions for Calculation of
Flux Through the Cell 3 Liner, White Mesa Uranium Mill
Dear Mr. Sinclair:
As discussed in our telephone call on Tuesday, February 2, I999,IUC had not received the referenced
letter. We are now in receipt of a faxed mpy ofyour letter, which we received the morning of Wednesday,
February 3, 1999.
We will reply to 0re questions and request for additional information in the letter by Friday, February 12,
1999. I can be reached at (303) 389-413 L
MRR/dm
Attachment
cc: Earl E. Hoellen
David C. Frydenlund
Harold R. Roberts
WilliamN. Deal
Ron E. Berg
Sam Billen (Ituight Pi6sold)
Tony Thompson
Dave Bird
CF: LELEAMhite Mesa MilUGroundwater Discharge Permit
CF: LELE/Knight Pi6sold Report
L Sincerely,
,fL_z-lr*@
Michelle R. Rehmann,
Environmental Manager
IurnnNauoNAL I
UneNruu
ConponATroN
Independence Plaza, Suite 950 . 1050 Seventeenth Street . Denver, CO 80265 . 303 628 7798 (main) . 303 389 a125 (f.r.x)
February 12,1999
VIA FEDERAL EXPRESS
William J. Sinclair, Director
Division of Radiation Control
State of Utah Department of Environmental Quality
168 North 1950 West
P.O. Box 144850
Salt Lake City, UT 84114-4850
Re: Response to your letter of January 21,1999 Regarding Methodology Assumptions for
Calculation of Flux Through the Cell 3 Liner, White Mesa Uranium Mill
Dear Mr. Sinclair:
This letter responds to your letter of Janutry 21,1999 regarding methodology assumptions for
calculation of flux through the Cell 3 liner, White Mesa Uranium Mill, which we received by
fax on February 3,1999. We appreciated your sending us a faxed copy of the letter.
As we indicated we would do during our telephone call of February 2, 1999, we hereby
transmit responses to the questions in your letter of January 21, 1999. We hope that this
provides your engineering staffwith all necessary information to complete their review of the
tailings cell performance review and modeling performed by Knight Pi6sold, LLC, as reported
on November 23, 1998. I can be reached at [phone].
Sincerely,
:\,. .''.\w' :\.-i'( ,,'l
. ,,1';,1'
&,,@
ul
;.O !;r. .a
!.@-:
?'l*'
tr,l-
./a:. l,\J
al. .
%
',-!'-. , a':r
;- I/,v/v'/o)
'vlse-- it{I\fl.
-Y4,,-LIAM
Michelle R. Rehmann,
Environmental Manager
MRR/dm
Attachment
Mr. William J. Sinclair
February 12,1999
Page2 of 2
cc: Ron E. Berg
Sam Billen (Knight Piesold)
Dave Bird
William N. Deal
David C. Frydenlund
Earl E. Hoellen
Harold R. Roberts
Tony Thompson
CF: LELEAilhite Mesa MilUGroundwater Discharge Permit
CF: LELElKnight Piesold Report
FIUSERS\STAFNMRR\LETTER99\SINCL2 I I.DOC
Xnight Pidsold LLC
CONSULTING ENGINEERS ANO ENVIRONMENTAL SCIENTISTS
Fehruan ll. 1999
1050 Sev'enteenth Street, Suite 500
Denv e r, C olo rado 8026 5 -0 5 00
Telephone ( 303 ) 629-8788
Telefar [phone]
YouF REFEFENCE t l0lC
ouR FEFERENCE UDEQ3.wp.l
lvlichelle Rehmann
International Uranium (USA) Corporation
1050 Seventeenth Street, Suite 950
Denver. CO 80265
Re: Response to UDEQ Comments on lvlethodolo_ey Assumptions
Dear lvlichelle:
At your request, rve have revierved the letter tiom the Utah Department of Environmental Quality
(UDEQ) dated January ?1. 1999. This letter contarned tbur comments regardin_e the UDEQ's revierv
of modeling r,ve recentlv completed tbr the White iVlesa Uranium ivlill. The purpose ol our modeling
et'tort lvils to estimate the rvater tlu.t that could reasonably be expected to pass through Cell 3, a
PVC-[ined impoundment at your tacility. Previous cell modelin_e by others urilized hyporhetical
cases involving unrealistic assumptions of massive liner tailure. Eighteen years of operation have
indicated that these hypothetical assumptions are unrvarranted. Our objective has been to revierv
avaiiable data and approximate actual site conditions. We have used en_sineerin,e judgement to
quantity the h.v-draulic conductivity ol the soils benearh the PVC liner. We int-er rhat UDEQ
generally agrees rvith the modeling but is questionin-s specilic input values used in rhe model.
Additionall.v-. UDEQ seems to purport that unsaturated tlorv in the underlying Dakota Sandsrone is
tiacture controlled. We have summarized the UDEQ comments and our responses as tbllorvs:
Comment l: UDEQ questions the conclusion that the liner was installed under intensive quaiity
assurance/quality control (QA/QC) and, therefore, our assumptions regarding liner
defect tiequencies are invalid.
Response l: Our review and analysis of cell construction activities as reported in our letter to
Anthony Thompson, dated November 23, 1998 concluded that rhe liner was, in fact,
installed in accordance with intensive QA/QC procedures. This report cites numerous
specifications, construction reports, Nuclear Re_eulatory Commission (NRC)
inspections, and third party reviews used to arrive at this conclusion. Should UDEQ
question our engineering review of the QA/QC documentation, these documents are
part of the public record and can be reviewed by UDEQ as required. These reports
contain the factory seam tests, quality control tests, field seam tests, bedding gradation
tests, and liner repair reporrs requested by UDEQ.
MEMBER OF
AMERICAN CONSULTING
ENGINEERS COUNCIL
xnlsfuPidsoA
GROUP
Kniebt Pidsold
-v-
ivlicheUc R,-.hnrrnn
lnternational Uraniunr t US,\r Corptrr;ltion
Februarv l 999
A.s tt,c srated m our letter report titled ,\[ethotlolog)" Jbr Colculution of Flu.rThrotryh
the CelL .J Liner. dated December i l. t998:
"Sensitivity analyses rvere conducted to determine the ettect of dei'ect
assumptions. Increasing rhe tiequency ol pinholes and installation
def'ects by an order of magnirude (i.e., t 0 times) resulted in only a 30%
increase in the estimates tbr averase tlu.r through the liner. These
analy'ses indicate that pinhole and det'ect tlux tiequencies are a minor
tactor in the estimation of rotal volumetric tlux throush the Liner."
Based on our revierv olconstruction documentation, rvejud,ee it improbable that there
could be 10 rimes the insrallation det'ects we assumed. Thus, although UDEe
questions the QA/QC assumptions. these parameters do not signihcantly change our
conclusions.
Comment 2: UDEQ questions the assumed hydraulic conductivity of rhe re.sraded materials beneath
the liner.
Response 2: No documentation is available tbr the sarurated hydraulic conductivities of dike or
bottom materials underlying the geomembrane. In our et'torts ro approximate actual
seepace r"'e used engineering .iudgement to estimate the hydraulic properties of the
hner bedding material. We assumed that the saturated hydraulic conductivity of the
l2-in sand layer behind the liner on the south dike of Cell3 rvas l.xl0'3 cm/s because
this is a typical value for the clean sand that was used tbr the underdrain marerial. The
value of lxl0'6 cm/s was used tbr the compacted soils behind the other three sides
(dikes) of Cell3. This same value also was used for the compacred, reworked Dakota
formation beneath the bottom of Cell 3. However, as shown by our response to
Comment 3. these assumptions are not critical to the estimated flu.x values calculated.
Comment 3: UDEQ comments that a change in assumed hydraulic conductivity would require
modeling rhe sysrem under a ditferent Design Case.
Response 3: The model we applied provides for six Design Cases as det-rned by Schroeder and
others (1994). These Design Cases vary depending on the arrangement of thej
composite liner and the hydraulic conductivity of its constituents. Our model
conservatively ignored the low conductivity tailings overlyin_e the geomembrane. The
appropriate Design Case for this arrangement is Design Case 3a. This case is formed
Gtt 60051 I 526c\wp\Uo€@.wpa
Knielzt Piisold
-u-
Comment.l.
Response 4:
\lichelle Rchmann
lnternational Uranium (USA) Corporation
February' ll. 1999
by a high conducrivitv material (pure rvarer) overlying the geomembrane w.ith a lorv
conductivity lay'er (reworked Dakota bedrock) underlying the geomembrane. In rhis
case. the liner beddin_s material acts as the controlling sorJ.
The UDEQ is correct that changing the assumed hydraulic conductivity tbr the liner
beddrn-s material rvould change the applicable Design Ciuie. However, as the UDEe
points out. the appropriate design case is determined by rhe conrrolling soil. If the
UDEQ t'eels that the hydraulic conductivity of the highly compacted liner beddin_e is
greater than 10'6 cm/s, the lorv conductivity tailing overlying the liner would become
the controlling soil.
Our engineering experience and the observed pertormance of the existin_e tailin_e
underdrain indicate that this tailin-s is trnely ground with resulring hydraulic
conductivities most likely ',vell below 10'6 cm/s. This case is most appropriarelv
modeled by Design Case -la. Design Case -la is a mirror image of our modeled case
rv'ith a lorv conductivity lay'er (tailing) ol'erlying the _ueomembrane ancl a hi_eh
conductivity la,v-er (rervorked Dakota bedrock) underlying the _eeomembrane. The tlu.x
equations tbr both Design Cases 3a and -la are identical, as are the heads on the
-seomembrane used in the tlu.x model. Theretbre, rhe Design Case used tbr the llux
model is correct no matter which assumptions are used tbr the saturated hvdraulic
co nduc tivity o f the tailin-e/be drocU,ee o membrane laye rs.
Derivation of the tlux model requires rhar one of the soils (i.e.. upper or lorver) be the
controliing soil. [n this case. the tlux is controlled by eirher the tailing above or rhe
bedding material beneath. Regudless of the assumption. the model indicates the same
llu.x rate and travel time tbr both Design Cases. As such, protracted discussions with
respect to proper hydraulic conductivity estimate do not change the conclusions of our
study.
UDEQ asks tbr justilication as to why fracture tlorv rvas not incorporated into the
travel time modelin-s.
Fracture flow was not incorporated into the tlow modeling because our review of
boring logs, pumping tests, and previous hydrogeologic reports gave no indication that
any significant fractures exist. We are awiue that questions regarding bedrock
tiactures have been raised in the past. Our review of available data concurs with the
conclusion reached in Titan Environmental's 1994 report titled Hydrogeologic
Evaluation of Wite Mesa Uranium Mill:
G:\t 5005\l 62&\wD\UOEO2.wpd
Xnj1lrt Pid s old
Nlichellc Rehm.rnn
lnternational Uraniunr (USA) Corptlration
February I 999
"lt could be postulated rhar a hyporhetical tiacrure beneath the wet
tailings cell rvould reduce the time of intiltration through the vadose
zone. Horvever, no si-eniJrcant tiacture/joints have been documented in
the subsurtace in the approximately 45 wells and borings at the site. In
addition, Disposal Cell No. 2 has been in operation tbr over t4 years
with no evidence of constituents migratin,e tlrou,eh the vadose zone."
(Titan, Pa_ee -10)
Our intent has been to model actual conditions and not elevate the hypotherical ro
reality. Fracture t'low was not considered in our model because we tbund no basis ro
believe that it exists. The UDEQ comment ret'ers to "site-speciJic well rest data". If
UDEQ is arvare of well testin-s that indicates tiacture tlow. it would be benetlcial tbr
them to cite their ret'erence.
It is important to lsalizs that minor adjustments to model assumptions do not signilicantly chanee
the estimated I .300 years required betbre any tiu.r throu-eh the liner could reach the perched warer
zone. Changing model results by even a few hundred years does not ne_sate the conclusion that Cell
3 overlies several layers of extremely low conductivity bedrock thar severely limit the potentiai tbr
tailings solution to reach the perched water zone or impact the deep re-eional aquifer.
We are pieased to assist you in respondin-s to UDEQ questions re_earding our morlelin_e et'forrs. As
ahva.v-s. t'eel tiee to call if you should need turrher assistance.
Sincerely.
eK/,,JL,L
ames R. Kunkel,P.E., Ph.D.
Senior Engineer
illin, P.E.
Glt 6005\t 526c\wp\UOEQ2.wpd
INrBnxeuor.ref
UneNluna (usl)
ConponATIoN
lndependencePlaza,Suite950r1050SeventeenthStreetoDenver,CO80265.3036287798(rnain) '303389a125(fa.r)
December 7,1998
William J. Sinclair, Director
Division of Radiation Control
State ofUtah Department of Environmental Quality
168 North 1950 West
P.O. Box 144850
Salt Lake City, UT 84114-4850
Dear Mr. Sinclair:
Enclosed is a copy of a report prepared by Knight Pi6sold entitled "Evaluation of Potential for
Tailings Cell Discharge - White Mesa lvfiU", dated November 23, 1998. We are enclosing this report
for your review prior to our meeting this Friday, December 11, 1998.
In additiorq we expect to fa,x you a draft of the listed hazardous waste review protocol by the
afternoon ofDecember 8, 1998.
. Sincerely yours,
I
\
Yl"-er^
Michelle R. Rehmann
Environmental Manager
MRR/dm
F"'ou"t\/;: h n -.\tE$rq" /*!r"n *y
{,ruight Pidsold LLC
CONSULTING ENGINEERS AND ENVIRONMENTAL SCIENTISTS
November 23,1998
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
2300 N Street, N.W.
Washingtort D.C. 20037 -128
1050 Seventeenth Street, Suite 500
D enver, C oLo rado 8026 5 -0 5 00
Telephone ( 303 ) 629 -8788
Telefax [phone]
YOUR REFERENC' rcz6c
OUR REFEHENCE EVALUAT3
Re: Evaluation of Potential for Tailings Cell Discharge - White Mesa Mill
Dear Mr. Thompson:
In response to your request, we have conducted an evaluation of tailings cell performance at the
White Mesa Nfill of your client, International Uranium (USA) Corporation (ruC). This independent
review was commissioned to analyze the potential for discharge of tailings water from this facility.
Our evaluation has included the following:
1.
2.
J.
4.
Review of tailings cell construction,
Review of liner leakage monitoring,
Modeling of hypothetical discharge of tailings water from Cell 3, and
Extrapolation of Cell 3 modeling to Cells 1 and 2.
This evaluation indicates that no discharge of tailings water to the underlying perched water zone in
the Burro Canyon Sandstone is likely to occur during the operational life of the cell. Reclamation
of tailings would eliminate the potential for future discharge. Should the cells be reclaimed with
retained water, our modeling indicates that discharge to the perched water zone is not possible for
approximately 1,300 years after closure. Even theq discharge of chemical constituents is not likely
due to microfiltration by the low permeability liner and attenuation in the vadose zone.
We hope that this review proves beneficial in evaluating your client's standing with regard to the
potential for discharge of tailings water. Please call if we can be of further assistance.
Sincerely,
PIESOLD LLC
Roman S.
AssociateMEMBER OF€$KLisht Pidsold.
GROUP
KNIGHT
<vl,'Jztr-.-<- / /f-
Popielak,P {
AMERICAN CONSULTING
ENGINEERS COUNCIL
.l
{glkU t'i4soA
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Tailings Cell Construction
Facility Summary
The White Mesa Mill has constructed four below-grade tailing disposal cells. These cells are
summarized by the following:
l. Cell 1 is constructed with a 30-mil PVC liner covered with earthen material. This cell was
completed in 1981 and is used for the evaporation and storage of process solution.
2. Cell2 is constructed with a 30-mil PVC liner covered with earthen material. This cell was
completed in 1980 and is used for the storage of barren tailing sands. This cell has received
an interim cover and presently receives no liquid effluent from the mill.
3. Cell 3 is constructed with a 30-mil PVC liner covered with earthen material. This cell was
completed in 1982 and is used for the storage of barren tailing sands and associated solution.
4. Cell 4 is constructed with a 40-mil HDPE liner. This cell was constructed in 1990 and
presently receives no tailings from the mill. Tailing solution was initially stored in this cell
but was later removed. A detailed analysis of liner performance will be conducted prior to
any process use of this cell.
Foundation Conditions and Excavation
The cells have similar foundation conditions, namely, variable thickness of cohesive clay (ML to CL)
overlying sandstone and claystone bedrock. Cells were excavated into the bedrock, but cell dikes
incorporated in-situ soils unless they were found to be calcareous. Some calcareous soils in the
vicinity of Cells I andZwere excavated for this reason and replaced with non-calcareous soil. The
soil excavated to form the cell bases was generally used in dike construction.
In general, bedrock was excavated by ripping and dozing to design grade, although some hard zones
were encountered in all cells. The rock was excavated to a final surface that slopes toward the
midpoint of the downslope (south) dike in each cell. After the last bedrock lift was excavated,large
rock fragments and claystone were removed from the underlying surface; other fragments down to
coarse sand were left in place for construction of the liner bedding layer.
Dike Construction
Dikes were constructed of cohesive (ML, CL) soils. D'Appolonia (1982a) reports that the soil was
placed and compacted in lifts to at least gOVo Modified Proctor dry density, or at least 115 pcf.
Cohesive soils compacted to this dry density have substantial strength, low permeability, and
essentially no liquefaction or settlement potential. Test results in D'Appolonia ( 1982a) show l -97o
maximum volume change in consolidation tests with acid pore Iiquid, demonstrating that these soils
are not susceptible to weakening and collapse in the event of liner leakage. Harrison and Abt ( 1980)
state that QC field density testing was performed frequently, averaging once per 1,000 cubic yards
(cy). This frequencyexceeds NRC requirements as stated in theconstruction specifications. Fillthat
failed testing was reworked and retested until it passed. This observation is confirmed by
C:\ I 6O0s\ I 626c\wp\EVALUAT3. u'pd
l
frlilrt Pu'ra
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
D'Appolonia (1982a, Appendix B) for dikes for cells 1, 2, and 3. All inspections by Abt
( 198O)reported no deficiencies in construction or QC practices or results. The records provide high
confidence that these dikes were well constructed and should remain intact under any failure or
leakage scenario likety to be encountered at the White Mesa Mill site.
Base Preparation and Bedding
The cell bottoms were prepared for liner installation by crushing, then compacting, the last lift of
ripped rock, less claystone and large sandstone blocks. The final excavated rock surface, on both
the cell bottom and the slopes excavated in rock, was picked free of loose +6-inch rock fragments
so that no rock protruded more than four inches above the general level of excavation. The small
broken rock was ripped, then crushed to a consistency of sand using compactors. This material was
placed on top of the remaining rock and rolled by a smooth-drum compactor until the surface was
free of fragments protruding above the rolled surface, as documented by visual inspections by all
parties (D'Appolonia, Energy Fuels Nuclear, and Goodrich orWatersaver) and photographs recorded
in D'Appolonia ( 1982a) and Energy Fuels Nuclear ( 1983). The finished bedding, which covers rock
surfaces on both the cell bottoms and side slopes, has a maximum size of 1.0 inch, less than 20 Vo
clay, and gradations (D'Appolonia,1982a; Fig. l3) consistent with a well graded medium to coarse
sand. The bedding material conforms to the specifications for this material in the design
(D'Appolonia, 1981). Cell4,constructedof HPDEin l990,wasfurnishedwitha 1-footlayerofclay
underlying the HDPE.
Underdrain System
The underdrain system consists of a l2-inch sand drain on the inslope of the south dike of each cell,
with a 3-inch diameter slotted PVC pipe buried in the downslope end of the sand drain connected
to a Driscopipe riser that connects to the top of the inslope. During construction some modifications
were made in the pipe connections to facilitate construction. The underdrain system was designed
to intercept and bleed off any moisture that might penetrate the liner on the downstream (south) dike
of each cell.
Although this system was originally intended to ensure that the dikes would not become saturated
with acidic solution that would compromise their structural integrity, the underdrain is also
hydraulicatly connected to the liner bedding, which is in direct contact with (directly underlies) the
12-inch thick sand drain of the underdrain system along the south inslope of each cell. Therefore,
there is also direct hydrautic connection between the liner bedding layer and the 3-inch PVC pipe
in the underdrain system, making the underdrain system also a leak detection system for the entire
liner. A more extensive underdrain was incorporated into Cell4 construction. However, Cell4 is
not in use and will not be modeled in this review.
Liner
The liner for cells t,2 and 3 is 30 mil PVC supplied and installed by B.F. Goodrich for cells I and
2 and Watersaver Company for cell 3. D'Appotonia ( 1982a) and Energy Fuels Nuclear ( 1983) have
C:\ I (rOOs\ l6 2(rc\rvpUSV ALU A1'.l.rvpd
T
xrlilrt Piat,n
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
documented that liner materials supplied by these companies met or exceeded specifications. These
reports also contain descriptions of ground preparation and daily inspections of the bedding surfaces
prior to installation, pointing out that the liner installation contractor had to be satisfied with the
surface before liner was installed.
D'Appolonia(lgS2a,Appendices C and D) and Energy Fuels Nuclear (1983, Appendices B and C)
document the seaming procedures used to join liner panels as well as the results of field and
laboratory tests performed on the liner seams. Additional documentation on liner installation and
test results is contained in Harrison and Abt ( 1980), Goodrich General Products Division ( 1980), and
D'Appolonia, (1980). The records contained in these documents demonstrate that QC protocols for
assurance of liner material quality and installation were followed rigorously. This record establishes
the basis for high confidence that the liner was installed correctly and would, therefore, be expected
to function as designed.
Liner Cover and Slimes Pool Drain System
Thelinerwascoveredwith 12to lSinchesofqualifying(non-calcareous)soilinwhichaslimespool
drain system was installed in cells 2 and3. The original design called for the Iiner cover to consist
of coarse tailings; however, insufficient volume of coarse tailing was available early enough to
construct the liner cover entirely of this material, so other qualifying soil was used in liner cover
locations where no slimes pool drain pipes were installed. A graded sand was used to fill over and
around the slimes pool drain pipes in Cell2. Coarse tailings were used as pipe bedding material for
all other cells. This drain system, intended to facilitate dewatering of fine tailings (slimes), consisted
of a rectangular grid of slotted PVC pipe wrapped in Mirafi 140 filter cloth and connected to a
Driscopipe riser ar the middle of the south dike of each tailing cell, the low point in the cell bottom.
The design is documented in D'Appolonia ( 1981). The actual grid pattern of pipes installed in Cell
3 (Energy Fuels Nuclear, 1983, Figure 4) differed from the design (D'Appolonia, 198 l, Sheet 3) to
better ensure gravity flow to the riser location.
Monitoring Plan
The monitoring plan (D'Appolonia, 1982b) covers inspection of operations, training of personnel,
supervision, lines of communication and responsibility, and documentation relating to design,
construction and operations of the tailings cells. [t was prepared in recognition of the fact that
diligence should not end at the end of construction but continue during operations. It calls for
inspections to be performed at regular intervals, ranging from daily to yearly.
Daily Inspections a-re to be made of each active tailing disposal area, the slurry pipeline (including
slurry flow and line pressure) and slurry discharge location, the evaporation pond (Cell 1), and the
sump and drain systems. Three levels of response are defined, classified according to the urgency
of the required response.
Weekly inspections include pond surface elevations, flow in sump and drain lines, and liquid levels
in underdrairt risers.
G:\ I 6O0s\ I (r26c\wp\EV At-U AT3 wPd
{:gl:nt PUso.L!
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Monthly inspections are conducted of the surface water diversion and retention structures, and the
pipeline is surveyed for wear (erosion of wall thickness) using ultrasonic methods.
Quarterly inspections are made of emergency spillways and post-construction changes outside the
disposal area. A review of operating and maintenance procedures is also conducted.
Yearly inspections include surveys of the dike crests and slopes, technical evaluation of inspection
reports, and a summary of inspection observations-
This monitoring plan provides regular, timely examination of the key indicators of cell and liner
function assuring that leaks substantial enough to saturate the bedding layer will be detected under
this program during the daily or weekly inspections.
C :\ I 60Os\ t (r26c\rvp\EVALU AT3.rvPd
Xuj,!lrt I'iis,,A
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Leak Detection System Monitoring
History
The uranium tailings cells (numbers 1,2 and 3) were built in the early 1980's. Since the inception
of their operation, there has been no indication that cells were or are discharging tailings liquid to
the leak detection system or underlying aquifer. Also site records indicate that operators of the
White Mesa Mill have followed inspection protocols requiring inspections of all tailings cell leak
detection risers. Data reviewed by Knight Pi6sold indicate that there has been no detection of water
in either of the LDS sumps from Cells 1 or 3. However, water was encountered during the
construction of the Cell} LDS sump. Additional water was later detected in a previously dry Well
J-2,inJune of 1980. This well was located between the Cell2 Dike and the Cell 3 Safety Dike in
an area which would later become the floor of Cell 3-
An October 1980 monthly report indicated that the water quality of Well 7-2 was similar to that of
the Fly Ash Pond. In December 1981 water was detected in the Cell2 LDS sump. After laboratory
analysis this water was once again determined to be unrelated to tailings liquids. Subsequent
analyses throughout the 1980s continued to corroborate that the LDS for Cell 2 was intercepting
ponded waters in the FIy Ash Pond. Therefore, although some waters are being collected by the Cell
2 LDS, several years of analyses and evaluations support the conclusion that no tailings cell leakage
has been detected in any of the LDS sumps for Cells I,2, or 3.
In August 1989, Umetco proposed a Detection Monitoring Program which was incorporated into the
pre-1997 license conditions. These conditions originated from the desire to detect any statistically
significant trends which would indicate that tailings liquids are present in the Cell 2 LDS sump.
Atthough this procedure is applicable to all cells, at no time has water been detected in the LDS
sump of either Cell 1 or 3. This program is summarized as the following:
l. lrak Detection Systems are to be checked weekly for presence of liquids. Any liquids found
are to be removed.
2. Determination of "significant leakage" will trigger increased sampling frequency from
selected compliance rnonitoring wells. Significant leakage was defined as flow greater than
one gallon per minute. Should flows exceed one gallon per minute, an automatic pumping
system would be installed.
3. Lrakage would be analyzed and evaluated for statistically significant trends. Should this
evaluation indicate that water removed from the LDS was originating from the lined facility,
Umetco would characterize the extent and degree of contamination and report to the NRC.
However, should water removed from the LDS originate flrom other sources (i.e., Mill Area
Sedimentation Pond) no additional work would be required.
C:\t 6OOs\l 626c\w'p\E,VALUAT-1.rv1xl
{*!g!rt t'i4soA
6
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Present Condition
The 1997 license renewal modified the detection monitoring program to evaluate the chemical
characteristics of any water found in the LDS, thus focusing the program on chemical analytes.
lnternational Uranium Corporation requested (January 9 and February 26, 1998) that the program
be restored to the pre-l991permit conditions which included an evaluation of flow rate as well as
analytes. Such a program is more likely to detect leakage through a damaged liner than consideration
of chemical analysis alone. The NRC concurred (1998) and issued an amended materials license
in 1998 restoring the intent of the pre-1997 conditions. Although the amended materials license
varies slightly from conditions proposed by Umetco in 1989, these variations are minor and do not
change the overall monitoring program as outlined in the previous section.
C :\ I 6O0s\ I 626c\wp\EVALU AT3.rvpd
{glklrt t'i|soA o
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Modeling of Potential Volumetric Flux
Purpose
This section assesses the hypothetical volumetric flux from tailings Cell 3 at White Mesa Mill.
Modeling of Cell 3 was determined to be the most conservative case to model as its saturated depth
and area are much greater than those of Cells I or 2, hence, the liner in the Cell 3 is under greater
stress than in the remaining cells. This modeling assesses the quantity of potential volumetric flux
through the liner of Cell 3 as well as the effect such flux might cause upon underlying strata. Results
of these analyses for Cell 3 were extrapolated to Cells 1, and2.
Background and Scope
Tailings Cell 3 was constructed in 1982 in accordance with the standards and requirements of the
U.S. Nuclear Regulatory Commission (NRC), which approved both the design and construction. As-
built records of the Cell 3 facility indicate that it is lined with 30-mil-thick polyvinyl chloride (PVC)
plastic underlain by a 6-inch (in) compacted soil layer, except along the south embankment where
the PVC liner is underlain by a l}-rn layer of sand drain material containing a 3-in diameter
perforated plastic pipe. A generalized schematic of Cell 3 is shown on Figure l. Monitoring of the
drain material since construction of Cell 3 has indicated no detectable water in this embankment
underdrain system. Cell 3 will be used for an additional two to three years and then will be
reclaimed.
In the absence of any evidence of leakage occurring from Cell 3, hypothetical modeling to evaluate
potential environmental effects if the leakage were to occur frqm this cell was performed. This
section presents results of the following:
l. modeling of volumetric flux through the PVC liner of Cell 3 based on historical measured
water levels in the cell provided by IUC;
2. modeling of water retention within the unsaturated zone between Cell 3 and a perched water
zone approximately 110 feet beneath the cell; and
3. modeling of the rate of water movement in the unsaturated zone beneath Cell 3 and the time
it would take for water to reach the perched water table under assumed future operating and
closure conditions in Cell 3.
Volumetric Flux through Cell 3 PVC Liner Under Historical Operation
Unlike water flow through a porous soil, water transmission through a PVC liner can only occur
because of vapor diffusion and density discontinuities (EPA 1988). The discontinuities may be
present as pinholes and installation defects. Vapor diffusion involves the transmission of water
vapor through the liner on a molecular scale and is controlled by the permeability of the liner, its
thickness, and the pressure head on the fluid. Pinholes and installation defects could serve as
C:\ I 60Os\ I 626c\wp\EVALUAT3. w'pd
xrjil,t t,tOt,n
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23, 1998
passageways for liquids. The combined flow through the discontinuities and vapor diffusion is
henceforth termed volumetric flux.
The passage of water through a liner also is dependent upon the thickness and hydraulic conductivity
of the materials immediately above and below the liner. Giroud and Bonaparte (1989) provide
procedures for calculating flux rates through liners, taking into account the characteristics of the
materials above and below the liner, potential installation defects, as well as the available hydraulic
head on the liner. The total volumetric flux across the liner calculated in this review includes
potential flux from vapor diffusion across the intact liner, flux through pinholes, and flux through
the installation defects. Giroud and Bonaparte ( 1989) indicate that typical geomembrane liners have
about 0.5 to 1.0 pinholes per acre from manufacturing defects. Additionally, good to excellent liner
installation results in less than I defect per acre. To be conservative, the flux rate analyses in this
review assume I pinhole and 2 defects per acre. Review of model results indicates that pinhole and
defect flux rates are a minor factor in the calculation of total volumetric flux through the liner.
Volumetric flux through the Cell 3 liner was calculated in three parts due to the geometry and the
underlying compacted soil/drainage layers; I ) flux through the south dike liner, 2) flux through the
three remaining dike liners, and 3) flux through the cell bottom liner. The flux rates for these three
were multiplied by their respective liner areas to give a total volumetric flux througli the Cell 3 liner.
The Cell 3 PVC liner equivalent hydraulic conductivity was taken from liner data published by the
U.S. Environmental Protection Agency (EPA, 1988)-
The total flux rate and the associated volumetric flux, based upon effective liner areas, were used
to calculate the volume of water which would enter and be retained by the underlying unsaturated
zone, as well as the time for the unsaturated zone to reach a water content which would begin to
initiate unsaturated flow downward toward the perched water table. Figure 2 presents a time series
of the historic Cell 3 water-surface elevations and also shows the calculated volumetric flux rates
through the liner for those water-surface elevations.
Average Cell 3 water-surface elevation during the 190-month periods of record was 5,595.57 ft
above mean sea level (famsl), with a minimum and maximum elevation, respectively of 5,580.23
and 5,605.41 famsl. These water surface elevations were used to calculated the hydraulic heads
acting on the liner.
Based upon the calculated flux rates shown on Figure 2 and the liner areas over which they apply,
the total volumetric flux across the Cell 3 liner is estimated to have averaged 50 fd/d over the 190-
months. Figure 3 shows the cumulative volumetric flux of water that could have passed across the
Cell 3 liner since January 1983. This volume is approximately 290,000 ft3. Of this volume,
approximately 79 percent is from vapor diffusion across intact liner surfaces, less than I percent is
from hypothetical pinholes, and approximately 20 percent is from potential installation defects.
Clearly, a majority of the seepage across the Cell 3 liner is from vapor diffusion across the intact
liner surfaces.
G:\ I 60Os\ I 626c\rvp\EVALUAT.l.rvPd
X"!g!rt t'i|toA
9
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
This hypothetical flux is very low and equates to less than 5 gallons per acre of liner per day (gpad).
Acknowledging that vapor diffusion through PVC liners occurs, the Environmental Protection
Agency (EPA) recommends that liner seepage be limited to de minimis quantities. This term refers
to the insignificant quantity of water vapor that may permeate a PVC liner. Although this rate is
calculated for site-specific conditions, EPA has proposed that 5 to 20 gpad is representative of a liner
installed with a high level of quality assurance (EPA, 1988). Our estimated flux rate, including
potential installation defects, is less than 5 gpad and is indicative of a well-constructed, functional
PVC liner.
Water Retention in the Unsaturated. Zone Beneath Cell 3 Under Historical
Operation
In unsaturated materials the pores are only partially filled with water, with the remaining pore space
usually occupied by air. Additionally, unsaturated flow can occur only if enough of the pore volume
has water in excess of moisture retained in storage by the forces of attraction. This threshold
volumetric water content is called "specific retention" and is the water content at which essentially
no water moves downward under gravity flow.
In the Dakota Sandstone and Burro Canyon formations underlying Cell 3, the rocks are unsaturated
for a depth of I l0 ft, until a perched water-bearing zone in the Burro Canyon Formation is
encountered. Data published by Titan Environmental (1994) indicate that within this 110-ft
unsaturated zone the average water content of the rocks is less than the moisture retention. This
means that some volume of water can be stored in the unsaturated zone before initiation of
unsaturated flow by gravity. This ability to permanently store additional water and the configuration
of the strata underlying Cell 3 are shown on Figure 4.
The documented volumetric water content of the I 1O-ft unsaturated zone in the Dakota and Burro
Canyon formations is 3.4 percent. Because the specific retention for this same thickness is 5.5
percent, 2.1 percent by volume is available for water storage prior to downward unsaturated water
movement (Titan Environmental, 1994). Applying this potential storage volume to the footprint of
Cell 3 (an area of 3,375,913 ft2', approximatelyTT.5 acres) results in aresidual storage volume of
about 7.8 million cubic feet for the 11O-ft thick unsaturated zone.
Assuming that2.1 percent residual water storage volume was available in January 1983, and the
seepage from Cell 3 between January 1983 and present was approximately 290,000 ft3, indicates that
approximately 4 percent of the residual pore volume in the unsaturated zone could have been filled
since Cell 3 began operation in January 1983. This means that an additional 7.5 million cubic feet
of water would have to discharge from Cell 3 just to bring the average water content of the
underlying Dakota and Burro Canyon formations to moisture levels adequate to initiate unsaturated
downward flow.
Cell 3 will be used for an additional2to 3 years, at which time it willbe capped and reclaimed. We
have estimated the volumetric flux for the remaining years oIoperation by conservatively assuming
C:\l 600s\l (r26c\u,p\EVALUAT3.wpd
{*!g!rt t'i4soW
10
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
that the water-surface elevation in Cell 3 would be at a constant maximum level of 5,603 famsl. At
this elevation the volumetric flux rate from the entire Cell 3 would be approximately 80 ft3/d or 0.4
gallons per minute (gpm). After that time, Cell 3 would be capped and reclaimed. Hence, volumetric
flux of tailings water from Cell 3 would have resulted in no discharge of tailings solution to the
underlying perched water zone during its operation life.
The time to bring the Dakota and Burro Canyon sandstones to a volumetric moisture content of 5.5
percent would occur far into the future after Cell 3 closure and reclamation if drainable liquids
remained in the cetl. Model results indicate that an additional 7.5 million ft3 of residual storage
would still be available to store future fluxes across the Cell 3 liner after closure and reclamation.
Conservatively assuming that water remains in the cell and an effective cell cap eliminates the addition
of water to the cell, the amount of liquids available for seepage would be limited to that which was
in the Cell 3 tailings at the time of closure. We estimate that the tailings within the cell have a specific
retention of 75Yo (Hoffrnan and Cellan, 1998 and Vick, 1990). Using this relationship we have
modeled a decreasing saturated level within the tails after capping. Projections of future water
surface levels and liner flux rates are shown on Figures 5 and 6. These data indicate that the residual
storage in the underlying Dakota and Burro Canyon formation would be filled to a volumetric
moisture content of 5.5 percent in approximately 400 years after Cetl 3 closure and reclamation.
After that time, additional volumetric flux from Cell 3 could begin to move downward toward the
perched water table at a very slow rate determined by the unsaturated hydraulic conductivity of the
underlying formation. At the inception of unsaturated flow, volumetric flux from the cell would be
34 trld (Frgu.e 6), and would require approximately 900 additional years to reach the perched water
table 110 ft beneath Cell 3. In summary, a total of i,300 years would be needed for volumetric flux
from Cell 3 to reach the perched water table after closure of the cell.
Water-Quality Implications of Liner Seepage
A majority of the potential flux from the cell would result from vapor diffi.rsion through the intact
liner. PVC liners do not appear to be permeable by ions with the possible exception of hydrogen
(EPA 1988). Because of this, a majority of the seepage would have a water chemistry much lower
in dissolved solids (virtually absent) than the water seeping through the liner via pinholes and
installation defects.
Transmission of water through soil or rock does not necessarily include the transmission of potential
pollutants contained within the fluid. Several physical and chemical processes result in the attenuation
of many chemical constituents. These processes include mechanical dispersioq adsorption to soil
particles, cation exchange, and oxidation-reduction reactions. As a result of these processes, not only
would it take approximately 1,300 years for volumetric flux to potentially reach the perched water
zone, but such volumetric flux could be expected to be relatively free of most contaminants.
Extrapolation of Cell 3 Modeling to Cells 1 and 2
Modeling of Cell 3 was determined to be the most conservative case to model as its saturated depth
and area are much greater that those of Cells I or 2. All three cells were lined with the same
G11600s\1626c\wp\DVALUAT3.wd
a
fulgU t,idsoA
u
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
materials in the same fashion. Our review of construction reports indicates that all cells were
constructed to the same general level of quality control - excellent. As flux through the liner is
directly proportional to the head above the liner, estimated flux rates from Cells 1 and 2 will be
consistently lower than for Cell 3. Therefore, modeling of Cell 3 results in the most conservative
estimates of potential impacts to the perched water zone.
G:\ I 600s\1626c\wp\EVAI-UAT3.wpd
Knfuht I'iisolde(r-
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
Summary of Conclusions
From the above review of cell construction and analyses of Cell 3 liner seepage during and after
operation, we offer the following conclusions:
l.Since the cells were constructed in the early 1980's there have been no indications that tailing
cells were or are discharging tailings liquid to either the leak detection systems or the
underlying formation;
Water observed in the Cell2 LDS sump has been thoroughly analyzed and determined not
to be a component of the tailings water,
Recent modifications to the operating permit are based on sound engineering principles and
are more likely to detect leakage through a damaged liner than consideration of chemical
analysis alone;
Modeling of potentially occurring volumetric flux through the Cell 3 PVC liner during the
period between January 1983 and October 1998 may have reached an average rate of 50 ftr/d
(0.25 gpm). This rate is considered "de minimi.r" and inherent for PVC liners by the EPA.
Based on our modeling, the total volumetric flux since beginning of cell use would represent
only 4 percent of the specific retention (i.e., permanent pore storage) in the underlying
sandstone. Hence, 96 percent of the permanent pore storage would be available for future
moisture if any were to migrate below the cell's liner;
Cessation of the discharge of any liquids upon termination of cell operating life and
reclamation of tailings will result in the gradually diminishing rate of volumetric flux during
the post-operation period;
If the status quo were to continue, the volumetric flux through the Cell 3 liner, based on our
modeling would require at least 400 years after closure to fill remaining sandstone pores such
that unsaturated flow downward toward the perched water zone could commence;
Unsaturated flow, if it were to exist, based on our modeling, would require an additional 900
years to travel the 110 vertical feet to the perched water-bearing zone after sandstone
moisture is raised to a degree facilitating downward movement of moisture. In other words,
a total of 1,300 years would be required before any potential volumetric flux from a
reclaimed cell could reach the perched water zone below the site;
Dissolved metals in tailings water are unlikely to be transported through the I lO-ft vadose
zone due to significant attenuation from a number of potential processes documented to exist
when moisture moves at a very slow rate through a very low permeability rnedia. These
processes include a combination of microfiltration through the PVC liner, adsorption to soil
t2
2.
3.
4.
5.
6.
7.
G:\ I 6(X)s\ I 626c\wp\EV ALU AT3.rvpd
Krtis'ht I'idsold
-(r-l3
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
particles, cation exchange, horizontal and vertical dispersion due to heterogeneities of rock,
and oxidation-reduction processes.
9. Since Cell I and2 are smaller and the hydraulic heads of liquids present in those cells are
also lower, estimated flux rates from Cells I and2 will be correspondingly lower than those
which may occur for Cell 3.
C:\ I 6O0s\ I 626c\wp\EVALUAT3.wpd
{:tignt ri4soW
t4
Anthony J. Thompson
Shaw, Pittman, Potts, & Trowbridge
November 23,1998
References
We have reviewed and/or cited the following documents in preparation of this review:
Abt, S.R., 1980, Trip Report of Inspection of Embankment #3 on 42,4/80.
Brooks, R.H. and A.T. Corey,lg64,Hydraulic Properties of Porous Media, Hydrology Papers No.3,
Fort Collins, Colorado State University, March,27 p.
D'Appolonia, 1980, Lrtter Report on PVC Liner and Underdrain Installation dated 8/8/80.
D'Appolonia, 198l, Engineer's Report, Second Phase Design - Cell3 Tailing Management System,
Wite Mesa Uranium Project.
D'Appolonia, 198 l, [rtter Report, Leak Detection System Evaluation, White Mesa Uranium Project.
D'Appolon ra,1982a, Construction Report, Initial Plmse - Tailing Management System, White Mesa
Uraniunt Project.
D'Appolonia, 1982b, Monitoring Plan, Initial Phase - Tailing Management System, Wite Mesa
Uranium Project.
Energy Fuels Nuclear, 1983, Construction report, Second Phase - Tailing Managentent System,
White Mesa Uranium Project.
Environmental Protection Agency, 1988, Lining of Waste Containment and other Impoundment
Facilities.
Giroud, J.P. and R. Bonaparte, 1989, Itakage through Liners Constructed with Geomembrane
Liners Parts I and II and Technical Note, Geotextiles and Geomembranes, vol. 8, no. l, pp 21-
6'l,vol.8, no.2,pp71-11[, and vol.8, no.4, pp.337-340-
Goodrich General Products Division, 1980, Inboratory Test Report of 3/6/80 on PVC Liner
Material.
Harrison, H.C., and Abt, S. R., 1980, IE Inspection Report No. 40/8681/80-01 of 6/6/80 on 4/15/80
Inspection of Dike #4 and Cell #2 Liner Placement.
Hoffman, G.L. and R.R. Cellan, 1998, Slime Dewatering at the Homestake Grants Proiect,In:
Tailines and Mine Waste '98, Proceedings of the Fifth InternationalConference on Tailings and
Mine Waste '98, Fort Collins, Colorado, USA, 26-29 January, 1998, Brookfield, VT, USA: A.A.
Balkema Publishers, 965 p.
G:\ l 600s\ l 626c\rvp\EVALU AT-1.u'pd
{*!t:!rt t'i4soA
l5
Anthony J. Thompson November 23,1998
Shaw, Pittman, Potts, & Trowbridge
McWhorter, D.B. and J.D. Nelson, 1980, Seepage in the Partially Saturatedhne BeneathTailings
I mp o undnrcn /s, Mini ng Engineering, April, pp. 432-439 -
McWhorter, D.B and J.D. Nelson, 1979, Unsaturated Flow BeneathTailings Impoundments, Journal
of the Geotechnical Division, ASCE, vol. 105, no. GTl1, pp. 1311-1334.
Tindall, J.A. and J.R. Kunkel, 1999, Unsaturated Zone Hydrology for Scientists and Engineers,
Upper Saddle River, NJ: Prentice Hall,Inc., 602 p-
Titan Environmental Corporation,1994, Hydrogeologic Evaluation of White Mesa Uranium Mill,
Report Prepared for Energy Fuels Nuclear, [nc., July,5l p.,5 tables, 19 figures, references,
Appendices A through G.
Umetco Minerals Corporation, 1989, Ittterfront D.K. Sparling and l.S. Hamrick regarding the
proposed Detection Monitoring Program for Cell 3.
U.S. Environmental Protection Agency, (EPA), 1988, Lining of Waste Containment and Otlrcr
Impoundment Facilities,EPN600/2-88/O52,Risk Reduction Engineering [.aboratory, Cincinnati,
oH.
U.S. Nuclear Regulatory Commission (USNRC), 1980, Letterfront G.D. Brown to Energy Fuels
Nuclear regarding inspections by Abt and Harrisort.
U.S. NuclearRegulatoryCommission (USNRC), 1998, Materials License, SUA'1358, Amendment
No.8.
Vick, S.G., 1990, Planning, Design, and Analysis of Tailings Datns, Vancouver, B.C., Canada:
Bitech Publishers, Ltd.,342 p.
G :\ I (rfi )s\ I 626c\u,p\EVALU AT3.wpd
6kG)(tEE-ET'-odfiE.5bBgB
d$o.x
Eg
8.A
o
9E'a.- =t6!!E9'H2E
H€
tr(\toL!)EItrD
E,
GIa
€trIan
(J
A
EI
f.)
6.9oEa, c,Pb
FE
o:)
E
.=
r!
l
(E
Lo
(u
Eo=L
rieooYrtTlI] El fr]oooc.i.jo
\t' (r| (oootttrq sl rqqqo.Ho\&
Qlntpi alqu rnlf, PetBInTIBJ
v.r r rr) rn rnooooolrlllrurqHHrqooooot'- \O ,.i + c.;
O\nOrrlO\n-o\o\oor-\orrnr\nr\nrnrrnri
(ry) uqlulelu etB3uns reluil\ IBc.rrolslH
g6-rrBf
L6'uet
96-qed
s6-qa{
?6-qe{
t6-qed
z6-qe{,
I6-qe{ E6e
06-qe{
68-qed
88-qed
z8-qed
98-qed
98-rq/{
,8-qe{
t8-trBf
OrnF- \Orrnir
orr)HO\o \olr} ltl
a
G
a
{
at,tlg-\rL.
\
/bEiB'tB9c-, -95 [rloo)().B
ti
a
$
ry
,c
?
)d
bo 6q
(
,ryffi
E
)
r
d
J
\
L
I
h
l'tr
ffi
tr?
,q
#
l
ruo
x)oqIi'
Deao(
I
tr
I
l
ft
db@
ffin
I
Y
9roc
F F.lFcaX-E8
)
D
\h
d
r
trtl*
{
)
rooQoa
oo6tsoa=J
$epaEx.:l-=LE:oo!!EEra..,E
-e6Eo
af
E
.=c6Ll6
o
(E
E0)c
oooorr')
ooodocn
ooooraca
(rU) rnu rlrleurnloA e^puJnurnc
ooooooooooooddddrnoroC{c{-H
oooooooooOO\€F-\Orn$ciN
(XBp/eiJ) rnlf ,lrlotunJo1 pelBlnrpJ
86-u?f
L6:rlgL
96-ge{
s6-qe{
,6-qef,
t6-ged
z6-wt
(l)I6-qed En
06-qed
68-qe{
88-qed
t8-qed
98-qe{
98-rel l
?8-qed
t8-rr?I
\
Tx
Ir
(.)
Lo
o
\
{,
{
"4r4
\
'r3s't^:ct$
II\
ooco
)o
o
$#
)ro
loooo
l\')to, ol
&80too
\a
\
Qoo,
{
oa,
Po
PoO
fir
\
roo oa
(
\
b
)cit
4
\
Dio
oI
,
000000
x
t\o
(D
E
o
c)
cl
(J
\
7
\
oo
LL6= Is-E
= CDFo;.rt>E3
1F E6x>of
't
'u6o2
d
I8
EC)od.)
EIEoo>'
8E
EoZ'5;EEH,(A-Or'H9ra'€ (-)
,Hga=8t
-rqEE9ct'5b 6&s
GiincrlN
(v)
(1)O
Bo
3{io
cll
(.)troN
o
GI
BrloEo
o)A
s\or-
>'
qo
aP.E4i6 c)-^H A=' e xtl rnEl O\3.2tr9gE
ado
oh9.Hli aA50
sE.
_vs-doAF.(HOoor (E
EleEO-t.a.1,
-v6rrE!
80
€$€.!le
= Qo\! E-:
IE"EE
g,
&ooc'€*EId6
A
!saE s$.a'E t,i
EE3
$EB
'ds
PN=s
tra
d
Eohtro6
o
-o.EaEOtso'Eof;Ei5E}E8pk(!,do
=F
=
=(Uoo
=o
8E'6'
(L
eo
C)
a:)
E
.=E6Lf
roIrqo
tt-
(r)otHo
\c;
{Sty$ e}su xntf petBIncIBJ n Xt",oooolltt14 14 r4 Eloooo,r; $ cd e.i
rnot
14q
oo+E]oo
9n9Z_f,e1^t
rcNZ-ENY
9ErZ-^oN
zStz-qet
LZEZ-f,EIN
q)
ZLZZ-E$Y E
l,IZZ-^oN
t9tz-UpL
80tz-rdv
ESozlnt
o€\osNOO\O\O\O\\o\\nrrr\nrn\n\n(g) uopu,reJfl erByns relBil\ ernlnt
o00O\ oOirr)\n rn
xro
E'5(D F{f;coE=
€.3?6E63FTI{ F
866t-}cO
oC'E
"86
=6A=,oo(!EO3p
E(!
=c=olL6:Eo ltJo
o
=9,o
o
(Utx
=lL
al
E
J.E
El
6co
.U
Eos
(-ry) xng tlrlerunlo1 e^llBJnrlrnJ\f,
ooooooooooooooooooooodo"o"o^o^o^o"o^o^o^oooooooooooooooooooodo"o"o"o"o"o"o^o"o"rrO\OOf-\Oltf,CiC{HO
9V9Z-fiey\r
rcvz-ar(Y
9t?z-^oN
zStz-qet
LZTZ-f,EIN
c)
zLzZ-EnY A
LIZZ-^ON
E9rz-ueI
Sotz-rdv
€soz-Ft
ooooooo@F-\Olr|<tcaN
(Xsplrg) xnld rrrlerunlo1 pefulntpJ ernlnd
x
tr,{o
Io)
o
866I-lco
o
LEbFE2JyOeE>Etrpde
EENi! Ic'':
=lLoo