DRC-2025-001511

;:::::=-
--------ENERGYSOLUTIONs --------
May 7, 2025
Mr. Doug Hansen, Director
Div of Waste Management
and Radiation Control
MAY O 8 2025
Division of Waste Management and Radiation Control
195 North 1950 West
Salt Lake City, UT 84 I I 4-4880
CD-2025-097
Subject: Federal Cell Facility Application: Responses to Round 2 Request for Information (per DRC-
2024-005076)
Dear Mr. Hansen:
EnergySolutions hereby responds to the Utah Division of Waste Management and Radiation Control's March
28, 2024 (DRC-2024-005076) Requests for Information (RFI) on our Federal Cell Facility Application. A
response is provided for each request using the Director's assigned reference number.
Each RFI letter is addressed individually, with the text of the RFI quoted in bold italics followed by its
response. The following attachments are also provided:
1. Electronic input/output files in support of HYDRUS runs described below
2. Electronic input/output files in support of the Monticello model described below
The round 2 RFI responses have been addressed in four parts. The DU PA will be revised once, to v4.0, after
all responses are completed. This list of model changes will be updated in each part of the round 2 RFI
responses going forward. DU PA model changes are needed to address the following from earlier RFI
responses:
1. RFI O-29.a, uranium isotopic inventory update for GDP waste
2. RFI O-39.b, the deep time model
No additional DU PA changes are needed to address the responses contained herein.
Appendix 0: Federal Cell Facility Waste Characterization Plan
• 0-11.a: The RF/ response did not adequately demonstrate why thermally driven flows will not be realized
or important at the Federal Cell Facility, when they are considered a significant mechanism driving flow
in other cover systems in the region.
Please provide an analysis of the impact of thermal flow on the annual percolation rate, using methods
that have been validated based on experience in the region. This analysis should be used to draw a
quantitative and specific inference regarding the impact of thermal flow on the conclusion of the DU PA
v2.0 model.
An analysis of the impact of thermal flow on the annual percolation rate, following the methods and
assumptions outlined in Stantec (2024), is provided below. Stantec (2024) includes an evaluation of
299 South Main Street, Suite 1700 • Salt Lake City, Utah 84111
[phone] • Fax: (801 ) 880-2879 • www .energysolutions.com
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
thermally driven flow for the White Mesa facility in Blanding, UT. The methodology cited by Stantec
(2024) references work by Globus and Gee(] 995), which defines (non-diffusive) flow in porous
media as an additive function of chemical/matric potential and thermal potential through the
following expression:
qwt = -K(0) dµ/dx -Kt dT /dx (1)
where qwt is a time rate of flow, µ is chemical potential, T is temperature, K(0) is the isothermal
hydraulic conductivity function of water content (0), Kt is thermal water conductivity, and xis
distance. Globus and Gee (I 995) state that, for non-salinized soil,µ can be replaced with matric
potential. Therefore, the equation has two additive components, one for chemical (osmotic) or matric
(suction/tension) potential, and a second for thermal flow. While Stantec (2024) cites Globus and Gee
(1995), the equation that Stantec provides only includes the second half of equation (I) from Globus
and Gee, where it is defined that:
(2)
where qt is the thermal flux, Kt is the thermal water conductivity, Tis temperature, and xis distance.
While not explicitly stated, it is assumed the authors considered the matric potential portion of
equation ( 1) to be very small and thus could be ignored.
HYDRUS modeling simulations have shown that estimates of matric potential flux through the
evaporative cover system at Clive are indeed very low. Figure I shows four representative
simulations from DU PA v2.0 HYDRUS modeling (Neptune 2023), where cumulative fluxes range
between 2-4 cm after I 00,000 days (274 years). Adding the physics of thermal and vapor flow to the
HYDRUS modeling simulations suggested that flow induced from thermal gradients in the cover
system can impact net percolation. However, impacts to net flow through the cover were not
consistent for the simulations. In some model simulations (see 0005 and I 002 in Figure I),
cumulative flow decreased when thermal and vapor flow were included in the simulation, compared
to the base simulations that did not include thermal and vapor flow. In other words, adding thermally
generated flow resulted in a net upward additive contribution to flux in the system, canceling out
some of the downward flux predicted through matric forces, and leading to lower total percolation
through the cover system through time. In other model simulations (see 0551 and 0353 in Figure I),
the thermal component of flow contributed an additive downward component to overall flux through
the cover system, leading to higher total percolation through the system over time. In either case,
while cumulative flow through the cover system is noted to have been impacted from the
incorporation of thermally induced flow, the magnitude is not significant enough to impact the
general conclusions of the DU PA v2.0 Model.
Page 2 of 40
::::::=-
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
6..------------------------------------~
E 0
5
~ 3 J:J
111 > :iJ IQ
"S E ;;J u 2
1
.,. .....
--------_ ...
------------------
-1002 -base
1002 -Vapor+ Thermal
0551 -base
0551 -Vapor+ Thermal
0353 -base
0353 • Vapor+ Thermal
0005 -base
0005 • Vapor+ Thermal
.,, _________________ ; ......... ...... ----_ .. --
0-,...-------.---------,--------,-------...--------+
0 20000 40000 60000 80000
Time (days)
Figure 1. Cumulative fluxes at the bottom of the HYDROS ID profile for four simulations
from the DU PA v2.0, with temperature and vapor flow enabled.
100000
HYDRUS The fact that some simulations increase in cumulative flux as a result of adding thermal
and vapor flow to the model simulations, and others decrease in cumulative flux (Figure I), is a
consequence of the non-linear relationship of a layered heterogeneous cover system and
stochastic/variable inputs across model simulations. In a simple homogeneous system , when the near-
surface soil temperature is cooler than temperatures at depth, the thermal gradient, according to
equation (2), would be expected to generate upward fluxes. Conversely, when the soil temperature at
the near surface is warmer than at depth, the thermal gradient would generate downward fluxes.
However, in heterogeneous systems that employ evaporative storage areas and discrete contrasts in
materials to form capillary breaks, like the Federal Cell Facility (FCF) cover system, understanding
net flow through the overall system becomes much more complex due to the competing thermal and
matric potential flow contributions in equation (1 ). For example, simulations that include hi gher
storage parameters for the evaporative layer may retain more water during dry periods, allowing
greater availability of water for thermal gradients to move during these periods. Conversely, other
simulations may include lower storage parameters for the evaporative layer, and in these situations
evapotranspiration (ET) might be able to remove a greater portion of this stored water during dry
Page 3 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
periods, reducing the available water for thennal gradients to act upon and to move downward.
Across the hundreds of model simulations developed for the DU PA v2.0 model, HYDRUS
simultaneously solves for these types of non-linearities when predicting net percolation through the
cover system. The HYDRUS modeling results suggest that the net impact of thennal and vapor flow
on cover perfonnance is variable and does not clearly lead to increased, or decreased, net percolation
through the cover.
Stantec (2024) employs a much simpler method to estimate flow through the cover system, making
the assumption that the thennal flow tenn in equation (1) dominates, and can alone predict net
percolation through the cover system. No support is provided for this assumption, and the cited work
(Globus and Gee 1995) does not suggest that this is an appropriate assumption. Stantec (2024) also
does not consider other seasonal patterns that may give rise to the annual pattern in the percolation.
Seasonal changes in precipitation, evaporation, and transpiration that impact the matric potentials in
the cover (i.e., the first tenn in equation (I)) should equally be considered as possible detenninants of
the percolation. If the first tenn in equation (1) dominates, flow could occur contrary to the
temperature gradient. In Stantec (2024), the apparent correlation between the predicted percolation
using only the thennal gradient portion of equation (I), and the observed perfonnance of the cover
system, does not alone indicate a causal relationship.
Despite these concerns, the approach provided in Stantec (2024) is replicated here using data from the
test cell at the EnergySolutions Clive, UT site. From equation (2), the required parameters to estimate
thennally induced flow on a system are a thennal hydraulic conductivity value and a temperature
gradient. Figure 2 shows the recorded temperatures at White Mesa and the Clive test cell at various
depths. At both sites, during cooler months of the year, the temperature of soil at the near surface
(shown in green) is cooler than at greater depth (shown in orange and yellow). Conversely, during
warmer months of the year, the temperature of soil at the near surface is warmer than at depth. The
reversal in shallow and deep temperatures throughout the year results in the production of seasonal
negative (upward) and positive (downward) thennal gradients at the White Mesa and Clive sites.
Thermal gradients for the two sites are shown as green lines in each site's respective chart in Figure 3.
The thermal gradient at White Mesa is calculated using sensors at depths of 686 mm and 2743 mm,
and the thermal gradient at Clive uses depths of 421 mm and 2466 mm from the test cell, leading to a
similar magnitude gradient at both sites(+/-4°C/m).
From equation (2), thermal gradients are expected to have the ability to move water both upwards and
downwards in the cover system, depending on the season and the relationship between shallow and
deep soil temperatures. The computed thermal flux for the White Mesa site, however, clearly only
uses positive thennal gradients when estimating the cumulative flux. This is evident by the red line
becoming "flat", with zero change in the computed cumulative flux during periods when the thennal
gradient is negative. Using equation (2), and only using positive (downward) thermal fluxes, and
assuming the same Kt value of 2E-l 1 m2/s-°C, the estimated cumulative flux for the Clive test cell is
of similar magnitude (roughly 5mm) over similar time periods (approximately 7 years), when
compared to the White Mesa site (Figure 3).
Page 4 of 40
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
If the impacts on flow from both positive and negative thermal gradients are considered, the data
suggests that upward and downward fluxes largely cancel each other out, with upward fluxes being
slightly dominant over the multi-year period recorded for the test cell. Figure 4 provides the
computed cumulative flux when positive and negative gradients are considered, and also using a
range of K1 values from Globus and Gee ( 1995), who suggest the range of values for K1 is between
2E-l l and 8E-l l m2/s-°C. When only positive (downward) gradients are considered in equation (2),
the cumulative downward flux increases as a function of K1. That is, higher values of K1 lead to higher
cumulative fluxes through the system. However, when positive (downward) and negative (upward)
thermal gradients are considered together through time, this results in a cumulative net upward
thermal flux over time, meaning upward thermally induced fluxes are cumulatively larger compared
with the downward fluxes.
The net upward thermal gradients at the Clive site would facilitate moving water at depth in the cover
system to shallower depths, where evaporation and transpiration occur. Although ET and
transpiration demand is less in cooler seasons (when the upward thermal gradients are present in the
cover system), demand does not go to zero. While thermal gradients could potentially cause an
additive downward flux during portions of the year when thermal gradients are positive (downward),
the impacts of upward thermal gradients overall are expected to be larger over time.
Therefore, data from the test cel l suggests that thermally induced flow on the cover system at the
Clive site might be expected to enhance cover performance overall, leading to lower amounts of
percolation through the cover. However, the cover system is complex and heterogeneous with many
non-linear relationships with respect to climate, soil moisture, and material property uncertainty.
Page 5 of 40
(a)
(b)
40
35
(.) 30 0~
Q) ... 25 :::,
ii, ...
Q) 20 a. E ~ 15
·5
(/) 10
5
0
25
20
~ 15 .a ~ Q) a.
E ~ 10
·5 Vl
5
0
:::::=-
ENERGYSOLUTIONS
---457mm --6-----1524 mm --.1---2743 mm
-a-686mm -1829mm --Air On-Site
-+-914 mm --ir-2134mm
,.......'.;::::;:::;::::;:::;:::;::::;:::;::::;:::;::::;::;::::;::::;::::;:::;::::;:::;:::::;::;:::;::::;:::;::::;::;::::;::::;::::;:::;::::;:::;:::::;::;:::;::::;:::;::::;:::;::;:::;::::;::'......40
0 0 -30 ~
-40 "' ~
~ a,
r--. 00 (j) 0 ~ N C')
~ ~ ~ ~ ~ ~ -~ a; a; ~ a, a, en en en
• A (421 mm) B (1119 mm) • C (1868 mm) D (2466 mm)
00 "' 0 rl N rn .,. lfl u:,
0 0 rl rl rl rl rl rl rl
0 0 0 0 0 0 0 0 0 N N N N N N N N N --... --... --... --... --... --... --... --... --... N N N N rl rl rl rl rl --... --... --... --... ....._ ....._ ....._ ....._ rn "' "' "' "' "' "' "' "' ....._
00
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Figure 2. Recorded temperature at discrete depths along a vertical profile at (a) the White Mesa
site (reproduced from Stantec (2024)) and (b) the Clive site.
Page 6 of 40
(a)
~ E s
X :::, u:: ro E ,._
Q) .c I-,._
0
C 0 ·10
0 u ,._
Q) c..
Q) -~ ro :5 E :::,
~
ENERGYSOLUT/ONS
6.0 ........................... ..-.-..-.-..-.-................... ,....,....,....,....,........,,........,,........,,........,,........, .................................... _,..,
5.0
4.0
3.0
2.0
1.0
Computed
Thermal Flux
10
5
0
-5
Mr. Doug Hansen
CD-2025-097
May 7, 2025
-i :::r (D -,
3 !l1..
G) -,
Ol C. ar ::a. ......,,
()
3
u 0. 0 --""""'&.L ................. L....iu..,........L .......... _._._..i........ ........ ....i...1... .................. ...L...... ......... u..l...... -10
9/1/16 9/1/17 9/1/18 9/1/1 9 9/1/20 9/1/21 9/1/22 9/1/23
6
E .s 5
)( :I u::
t6 E 4
GI
f= ,._
0
5 3 i 15 u ,._
GI
Q.
GI 2 > ;::
C'CI :;
E :I u 1
10
8
6
4
2
-6
-8
0 .___....._ __ .,___ _ __._ __ ..,__ _ ___._ __ _.__ _ ___._ __ ...w -10
;i ,,
3
e!..
G) -, DI 5!, ,,
:I ...
9/2/08 9/2/09 9/2/10 9/2/11 9/1/12 9/1/13 9/1/14 9/1/158/31/16
(b)
Figure 3. Thermal gradient at (a) the White Mesa site (reproduced from Stantec (2024)) and (b) the Clive site is
shown as a green dashed line, and the computed thermal flux is shown as a red line. Measured percolation is
shown in blue for the White Mesa
Page 7 of 40
(a)
(b)
25
e
.§.. 20 )( :::, ;:;: .. e ..
f:. 15
5
C ~ ..
0 ~ 10 c.. .. > i :i E a 5
' ' ,,
t, i • ;
~
E ERGYSOLUTIONS
10
8
6
A I 2
0
-2
-4
-6
-8
0 -10
9/2/08 9/2/09 9/2/10 9/2/11 9/1/12 9/1/13 9/1/14 9/1/15 8/31/16
0
e -1
.§..
)( ~ -2 ..
E ~ -3
5
5 -4 1il 0 ~ ~ -5 .. > i
:i -6 E :::, u
-7
u r
I
10
8
6
4
2
0
-2
-4
-6
-8
~ .. 3 !!!..
Cl
iil c.. ;-a
n ..... 2-
.Q
2,
-8 .__ _ _._ __ '--_ _._ __ '----'----'----'----"--' -10
9/2/08 9/2/09 9/2/10 9/2/11 9/1/12 9/1/13 9/1/14 9/1/15 8/31/16
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Figure 4. Computed cumulative thermal fluxes for the Clive site, excluding negative gradients (a) and including
negative gradients (b). The red, green, and purple lines represent cumulative flu xes calculated with Kt values of
2E-ll, SE-11, and 8E-11, respectively. Higher values ofK, lead to higher estimates of cumulative fluxes when only
positive (downward) gradients are considered. When both positive and negative thermal gradients are considered
(b), the estimated cumulative flux through the cover system from thermal gradients overall is negative, meaning a
net upward flux in the system.
Page 8 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
■ 0-12.a: The initial RF/ requested appropriate documentation for the analysis conducted by Bingham
Environmental (1991) cited in the application. The documentation is necessary to justify the hydraulic
property inputs used in the unsaturated zone analysis. This justification should include an assessment of
potential borrow sources and their hydraulic properties to demonstrate the containment system can be
constructed as proposed in the application. Please provide detailed documentation of hydraulic properties
used in the DU PA v2.0 analysis.
EnergySolutions has elected to replace the Frost Protection Layer (FPL) specification with that of the
Type A Filter material within the Federal Cell Facility top slope cover. Type A Filter material is well
tested, well understood, and has been widely used in other embankment cover systems at the Clive
site. The use of Type A Filter for the FPL is expected to provide a stronger capillary barrier than the
FPL as originally specified based on its hydraulic properties and infiltration simulations described
below.
The Type A Filter is an engineered gravel material, with 90 percent of the particles being larger than
7.9 mm diameter. Sieve analysis data for Type A Filter material from past embankment cover
construction projects is shown in Figure 5; it is apparent that these samples have similar textures and
that this material is quite homogeneous.
100
-+-AF180605-1
-+-AF120618
--AF180405
--AF180420 80 -+-AF180524-01
---AF180604
---AF180605
---AF180710
;=: 60 '"' ----AF180716 .; ;: ----AF190423 >, .0
.; -+-AF190809
C: .;: --+---AF190530
i:: ., --.-AF190603 ~ 40 ., a.. -+-AF190604
-+-AF190606
-+-AF190607
-+-AF190610
20 -+-AF190611 3-1
-+-AF1906115-1
-+-AF1906134-1
-+-AF1906135-1
AF220210
0 -+-AF230728
1000 -+-AF240516 100 10
Sieve size (mm)
Figure 5. Sieve analysis data for Type A Filter samples.
Page 9 of 40
:::::=-
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Hydraulic properties for Type A Filter are reported in EnergySolutions (2020) for a single sample. In
order to incorporate uncertainty in hydraulic properties for the Type A Filter gravel material,
hydraulic properties for gravel reported in various literature sources were compiled. Weights were
assigned to the various materials, with the highest weight assigned for Type A Filter, and lower
weights assigned for the other gravel references. Type B Filter is another gravel material used at the
Clive site that is similar to Type A Filter and is weighted more heavily than other gravel sources, but
less than the weight for Type A Filter. Weights were also assigned to provide the best statistical fits to
the range of hydraulic properties. For example, lower weights were applied to parameters that were
inconsistent with Type A Filter parameters, thereby providing a range of values, but honoring the
single analysis for Type A Fi lter material reported in EnergySolutions (2020). The sources of gravel
hydraulic properties, the hydraulic properties, and weights assigned to each source are shown in Table
1.
Table 1. Sources of gravel hydraulic properties, the hydraulic properties, and weights assigned
to each source.
Gravel material Ksat !Res. WC Sat. WC alpha n Weight Weight
reference (cm/day) (11cm) (%)
Type A Filter,
Cover Test Cell, l.875E+06 0.03 0.37 0.3167 2.79 I 18.2%
EnergySolutions (2020)
Type B Filter,
Cover Test Cell, 9.072E+05 0.07 0.37 0.0569 10.42 0.75 13.6%
EnergySolutions (2020)
Benson et al. (2024),
Table 3, Bedding l.296E+04 0 0.25 0.1873 1.28 0.125 2.3%
(Naturalized)
Benson et al. (2024),
Table 3, Transition l.210E+04 0.05 0.25 0.0863 1.56 0.125 2.3%
(Naturalized)
Benson et al. (2024), 0.018 0.34 0.1069 1.8 0.5 9.1% Table 1, NW5-Gravel
Benson et al. (2024), 0.003 0.442 0.2059 4.4 0.5 9.1% Table I , Pea gravel
Mallants et al. (2003), 3.024E+04 0.005 0.42 4.93 2.19 0.5 9.1% Annex59, Table 59.1
Rockhold et al. (2015),
Table 4.8, Gravel l .728E+05 0.006 0.29 17.8 4.84 0.5 9.1%
drainage
Page 10 of 40
~
E ERGYSOLUTIONS
Gravel material Ksat IR.es. WC Sat. WC
reference (cm/day)
Tokunaga et al. (2002), 0.0176 0.4
Figure 5 (assumed)
Tokunaga et al. (2002), 0.0872 0.4
Figure 6 (assumed)
Tokunaga et al. (2002)
0.068 0.4
Figure 7 (assumed)
alpha n
(11cm)
2.2 2.13
0.7874 3.99
0.2545 4.35
Weight
0.5
0.5
0.5
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Weight
(%)
9.1%
9.1%
9.1%
Distribution development for hydraulic properties involved varying the weights for each source, and
testing distribution types (e.g., lognormal, gamma, etc.) until distributions were considered reasonable
and would honor the single property set reported for Type A Filter while providing a range of
properties for probabilistic modeling. To sample the data, 10,000 bootstrap replicates were pulled
from the data for each parameter using the weights in Table 1. Statistics for the distribution
development of hydraulic properties for the capillary barrier are shown in Table 2.
Table 2. Statistics for the distribution development of hydraulic properties for the capillary
barrier.
Parameter Distribution Mean Std Dev Median Min Max
Ksat Gamma l.121E+06 3.173E+05 l.091E+06 3.152E+05 2.585E+06 (cm/d)
van
Genuchten LogNormal 0.1975 0.0613 0.1885 0.0537 0.6041 alpha
(I /cm)
van
Genuchten Gamma 4.4453 0.9372 4.3748 1.9000 8.5925
n
Saturated
water Gamma 0.0361 0.0082 0.0355 0.0142 0.0726
content
Residual
water Gamma 0.3766 0.0103 0.3764 0.3390 0.4166
content
Page I I of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
I 0,000 replicates for each parameter were generated based on the statistical summary provided in
Table 2. Sorted replicates from lowest value to highest value were plotted for each parameter, along
with the actual values measured on gravel samples and reported in Table I. Plots showing the sorted
replicates along with the actual values are shown in Figure 6.
Random values of each of the developed parameters shown in Figure 6 were selected for each
HYDRUS simulation of the FCF cover design. I 00 sets of HYDRUS simulations were run at a time
due to the large computational burden of each simulation. The simulations were identical to those
reported in Neptune (2025), with the exception of replacing the hydraulic properties in Layer 3 with
the newly developed capillary barrier properties.
Due to the extremes in hydraulic properties, in particular between Layers 2 and 3 where saturated
hydraulic conductivity (Ksat) increases by five to six orders of magnitude going from the silty clay
layer to the gravel layer, the HYDRUS model experienced difficulties in reaching model
convergence. This extreme in hydraulic properties also results in a strong capillary barrier inhibiting
flow between Layers 2 and 3, and between Layers 3 and 4. In order to address the model convergence
challenges, many models were tested, with varying grid sizing, tolerances, maximum iterations,
timesteps, etc., none of which substantively improved model convergences. Therefore, all model
settings were left as originally set in the v2.0 model runs (Neptune 2025).
2300 simulations were run (23 sets of I 00 runs) in order to obtain I 00 converged model runs. Results
of the I 00 converged model runs are shown in Figure 7. The results are shown as sorted percolation,
from lowest to highest percolation. Sorted percolation from the DU PA v2.0 HYDRUS models are
also shown in Figure 7, where it is apparent that average percolation in the DU PA v2.0 model is
about three times higher than the results using a gravel distribution for the capillary barrier. Results
are also shown in which the single set of measured Type A Filter properties were used with I 00 sets
of varying properties for Layers 1, 2, 4, and 5 from the DU PA v2.0 model; 13 of 100 simulations
converged using the Type A Filter properties for Layer 3 for those I 00 simulations. These results
indicate that the use of the Type A Filter for the Layer 3 capillary barrier will generate less percolation
than that modeled in DU PA v2.0. The Type A Filter gravel will provide a strong capillary barrier in
the FCF cover design.
Page 12 of 40
l E•OG
~ 1
~
l E+OS
1 E•04
100
E 10
2
" .c g " C ! i? c3 C ~ 0.1
0.01
~
ENERGYSOLUTIONS
• •
•
• Type A F11l~r
• l')'pe 8 Filter
• Other Gr Meis •
-G1i1VCldtSlflt)tJt1on •
0 2000 4000 6000 8000 10000
Sorted Replicate
• Type A F1lte1
• TypeB F1lte1 ] • ♦ Othet Gr .1vels I G1ave~t10n • C • g
~ E • ~
! __.) C g
r i •
12
10
8
6
•
2
0
• Type A Filter
• Type B Filter
• Other Gravels
•
Mr. Doug Hansen
CD-2025-097
May 7, 2025
0 2000 4000 6000 8000 10000 0 2000 4000 6000 8000 10000
0,d$
e 0.4
'! ..,
E i 035
g
i 0.3
ll 5 ~ 0.25
0.2
0
Sot1Cd RepllC.Jte Sorted Replicate
0.1 • • TypeAf11ter • • • Type 8 F1lle1 • _ 0.08
"' ♦ Other Gravels E ~ -Gravel ct1stnbut1on E = 0.06 • i'; c 8
j 0.04
♦ " • lype AF11te, ;; :, • Type 8 F11te1 ,1 :l
• Other Gravels • " 0 02 •
-Gravel d1st11t>ut1on ' 0 •
2000 4000 6000 8000 10000 0 2000 4000 6000
Soned Replicate Soned Repl1ca1e
Figure 6. Sorted replicates of hydraulic property distributions, and gravel
measurements for the capillary barrier.
8000 10000
Page 13 of 40
> ..._
E
E
C 0 ·,;:;
<1) 0
:::'. Q)
0..
1
0.1
0.01
0.001
0
:::::=-
E ERGYSOLUTIONS
-DU PA ,2.0 {1000 "'"" lo GS model), '"" a 0.017 mm/y, I
~ Type A Filter (13/100 runs), avg = 0.004 mm/yr
--Gravel distribution (100/2300 runs), avg = 0.006 mm/yr
0 Percentiles (5, 25, 50, 75, 95)
100 200 300 400 500
Sorted replicate
600
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Gravel Distribution
Percentile mm/yr
Mean 0.0061
Minimum 0.0027
5th 0.0035
25th 0.0040
50th 0.0043
75th 0.0046
95th 0.0162
Maximum 0.0717
700 800 900 1000
Figure 7. Sorted percolation results for the DU PA v2.0 (blue), model testing using DU PA v2.0 and Type A Filter
properties for Layer 3 (green), and 100 results using random draws from a developed distribution of gravel
properties for Layer 3 (orange).
Water balance plots for a 5.5-year period (Day 443000 to 445000 of the 2000-year simulation) are
shown' in Figure 8. Plots for the 25th, 50th, 75th, and 100th (maximum) percentile percolation results
are presented. The 5th percentile percolation water balance plot looks similar to that of the 25th so it
is not included, and the 100th percentile plot is included instead of the 95th percentile percolation
plot. This period is a relatively wet period in the simulation, where percolation breakthrough occurred
for the highest percentile percolation simulations. Note the change in scale on the right-hand y-axis
for percolation for the I 00th percentile results.
Figure 9 shows cumulative percolation for the same 5.5-year period for the 5th, 25th, 50th, 75th, 95th,
and I 00th percentile percolation simulations. Breakthrough percolation events can be seen for the
95th and I 00th percentile percolation simulations only.
Page 14 of 40
~;-c E 2-
150
-P·('(.;1pllJIIOfl
-E·1:1pcvvt•on
~ ~ 120 ""'
-Tr~n"i~lfal1nn
Et :. e e,;. ~to ~ !! !? ~ ~ ~ .. ; t~ ~~ 0 0 ;z iE l:,;. ~~ ~ci ,,,,
0 C § •
u
if 2-H i! ,..;. E-e,;. 't i H ~i E , ~~ O 0 ~ z ½§
{j
ti a
-Ranor.
-S!oraft"
-SWE 90
-Perc0-1.111m
60
30
443000 443400
150
-P·~•pi\.ltlOo
-E..,opo,c;t"°"
120 -T1;,n-.pu1nion
90
60
30
Runoff
-s~orage
-SWE
-Pt!ICOl.'l!lon
0
443000 443400
~
ENERGY SOLUTIONS
75th Percentile Percolation
4"3eOO 444200 4M600
f1me{daysJ
100th Percentile P~colalion
443800 444200 444800
T1rnetdays)
0.15
0.12
£ E.
000 ~
l!
0.06
0.00
0.00
44,000
1000
800
600
• 00
2.00
0.00
4450CO
~ I u
£ E 1 ,;
0 ~ .t
~ i u
150
JI 2,:. ! ~ 120
Sl t~ ~ i 90 2 o !;t1 &• • ! ~i E C u ,
g~ e: C. ~ 2~ ~, . ,
ii a
60
30
0
-P·et.rp,uuon
-E.-opor.:t,on
-T,:m,-p11:1110n
-Aaioff
-S!orage
-SWE
-P!!-tCOIO!IOf'I
443000 443400
150
-P·eclp,1311on
Ir -E·Ji)POfDI""' I
2.§. -Tr!'lno;pmuu,.n n, 120 "" Ri.mof1 ~.
.§ -~
e'~ -S'.01agt>
ff -SWE
g 0
..,
~ o; -Pf"tCOMllon
i! ~ " ~,
E C ~ t 60
~~
21 ~! ii
30
0
0
443000 443400
2.Sth Percentile Percol.-tion
4'3a00 444200
50th Percentile Pcrcol..1fion
443800 444200
Tmlt(days}
Figure 8. Water balance plots for the 25th, 50th, 75th, and 100th percentile percolation simulations.
299 South Main Street, Suite 1700 • Salt Lake City, Utah 84111
[phone] • Fax: [phone] • wwwenergysolutions.com
...roe
Mr. Doug Hansen
CD-2025-060
March 27, 2025
0.15
0 12
£ s 000 g
" j
t
0.06 ~ ~ "
0.03
0,00
445000
0.15
0.12
e s
OO'l o j
i 0.06 § ~ u
0.00
0.00
444600 445000
-;:
"' 'O E E
C 0 ·~
0 u
OJ c..
1.0E-02
1.0E-03
1.0E-04
1.0E-05
443000
-5th%
-25th
-50th%
-75th%
-95tho/o
-100th%
~
ENERGYSOLUT/ONS
443500 444000
Time (days)
444500
Mr. Doug Hansen
CD-2025-060
March 27, 2025
445000
Figure 9. Cumulative percolation for the 5th, 25th, 50th, 75th, 95th, and 100th percentile percolation simulations.
Water content plots are shown in Figure 10 for the same percentile simulations that are shown in Figure 8.
Data are presented for water content between Day 44300 and 44500 of the simulations for HYDRUS
observation nodes located in the centers of Layers 1, 2, 3, and 5. Data are not presented for the node in Layer
4 because it is nearly identical to the data shown in Layer 5.
The use of the Type A Filter gravel for the capillary barrier in Layer 3 of the FCF cover design will provide a
strong capillary barrier that will inhibit infiltration and percolation to a greater extent than the FPL properties
used for Layer 3 in Neptune (2025). Accordingly, the hydraulic properties used in infiltration results
incorporated in DU PA v2.0 overestimate infiltration compared with those that would be expected for the
Type A Filter specification; and this aspect of the DU PA model does not need to be revised.
299 South Main Street, Suite 1700 • Salt Lake City, Utah 841 I I
(801 ) 649-2000 • Fax: (801 ) 880-2879 • www.energysolutions.com
0.45
0.40
035
c ~ O,JO ·;
" ~ ~ t " < " ~ § ~
·. C i g
~ t ~ ~
025
0 20
015
0.10
005
000
443000
0.45
0.A0
035
0.30
0.25
0 20
-UlSc.,,
-l.245C'TI
-L385c-n
-L.5152:.m
~
015 1:il lS<m .iJ !;
-L245cm 0.10
-LJBScrn
0.05 -L5152cm
0.00
443000
443400
443400
::::::=-
ENERGYSOLUTIONS
443800 4M200 444600 4"5300
T1m-e (days\
""3800 4M200 444600 4d5'.l00
Time ldaysl
0dS
0.40
035
030
0.25
0.dS
OAO
035
c i 03l
~ 0.25 e z ~ • ~ 2 ;;
0 :..'O
015
0.10
005
0.00
443000
-L115cm
-l245tm
-l.385cn•
-LS152cm
020 ._,J ~
0.15
0.10
0.05
000
443000
-u15cm
-L2.<1Scm
-L385c11'1
-l5152cm
443400
443400 443800 4M200
11m~{days}
·=• 4M200
Tltnf>(Oays.l
Figure 10. Water content plots for the 25th, 50th, 75th, and 100th percentile percolation simulations.
299 South Main Street, Suite 1700 • Salt Lake City, Utah 84 111
(80 1) 649-2000 • Fax [phone] • www.energysolutions.com
4418)0
444&)0
Mr. Doug Hansen
CD-2025-060
March 27, 2025
445300
4"5300
:::::=-
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
■ 0-13.a: The initial RF/ requested documentation regarding the assignment of unsaturated hydraulic
properties in the cover analyses, including how the assignment of hydraulic properties is consistent with
accepted industry standards. Additionally, the response uses data from the cover-test-cell but does not
acknowledge that the cover-test-cell was unvegetated. This unvegetated condition is unlikely to persist
over the intended service life of the cover for the Federal Cell Facility.
The response describes how some of the hydraulic properties used in the analysis are comparable to
those recommended in NUREG/CR 7288 on the topic, notably qs, Ks, a, and n and attempts to justify
using qr substantially greater than zero, as inherent in the Bingham data set, by stating the "effect of
using a zero value for 0, is to increase the storage capacity of the soil. In addition, a zero value of 0, will
also affect the water content at which percolation occurs." This does not adequately recognize that using
qr that is too large, compresses the effective saturation in the Mualem equation that defines the
unsaturated hydraulic conductivity, resulting in unsaturated hydraulic conductivities that are
unrealistically low and model predictions that underestimate percolation.
The pore interaction term remains unresolved. Please address this RF/ using the information provided
above.
A series of one-at-a-time (OAT) sensitivity analyses were run to address this RFI using the
information provided. Results of 100 converged model runs using a gravel distribution for Layer 3 are
shown in Figure 7. Also shown in Figure 7 are the percentile values of percolation for those 100
results. OAT simulations were conducted using the 5th, 25th, 50th, 75th, and 95th percentile
percolation simulations. For these five sets of percolation results, additional simulations were
conducted where residual water content (thetaR) was set to zero in Layers 1, 2, 4, and 5.
Additional simulations were conducted where the pore interaction term (PIT) was set to -1 and -2 in
Layers 1, 2, 4, and 5. Results of these OAT simulations are shown in Figure I I ,where it is apparent
that changing thetaR to zero has very little effect on percolation results. Changing PIT to -1 and -2
also has very little effect on percolation results. The only large change was that percolation decreased
using thetaR=0, and using PIT= -2 for the 95th percentile infiltration simulation.
299 South Main Street, Suite 1700 • Salt Lake City, Utah 84111
[phone] • Fax: [phone] • www.energysolutions.com
>-E
E
0.02
6 0.01 ·.;:;
~
0 ~
QJ 11.
0.00
~
ENERGYSOLUTIONS
■ Layer3 Gravel
■ Layer1245 thetaR=0
■ Layer1245 PIT=-1
Layer1245 PIT=-2
11 II
5th 25th
II
50th
Percentile of Drainage
75th
Mr. Doug Hansen
CD-2025-097
May 7, 2025
95th
Figure 11. Results of OAAT simulations with varying residual water content, and pore interaction term, in Layers
1, 2, 4, and 5.
• 0-14.a: The initial RF/ requested information regarding materials proposed for the frost protection layer,
which also serves to create a capillary break. This break is critical in controlling the percolation rate and
is a main driver for low percolation results in the DU PA v2.0.
The response indicated properties were estimated from publicly available databases, which are largely
populated with hydraulic properties for agricultural soils. Please provide relevant data for frost
protection materials intended for use at the Federal Cell Facility, and measurements of their hydraulic
properties including a detailed assessment of potential borrow sources and their properties, to
demonstrate the containment system can be constructed as proposed in the application.
EnergySolutions has elected to replace the Frost Protection Layer (FPL) specification with that of the
Type A Filter zone material within the Federal Cell Facility top slope cover. This material is well
tested, well understood, and widely used at the Clive site. The use of Type A Filter for the FPL is
expected to provide a stronger capillary barrier than the FPL as originally specified based on its
hydraulic properties.
To further evaluate reasonableness of the Clive FCF HYDRUS model results, a validation model has
been constructed for comparison with Monticello, Utah observational data.
Introduction
This analysis attempts to apply the setup and basic structure of the HYDRUS 1-D (HID) model for
the Clive, UT site to the test cover located in Monticello, UT. The objective is to validate the Clive
model by demonstrating the model's logic is capable of reasonably approximating observed data at
the Monticello site. Various sources of uncertainty include hydrologic parameters, initial moisture
conditions, plant community evolution, and soil pedogenesis. After installation, the cover system
would be evolving rapidly in terms of plant establishment, soil compaction, and moisture conditions.
Page 19 of 40
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Representing transient factors as steady-state parameters introduces uncertainties for the validation
effort, as the Clive model is structured to represent a long-term steady state of the cover system with
respect to these variables. Outputs from this effort should be considered only with respect to trends
and approximate values ( order of magnitude) when com pared to observations for the purposes of
validation. The primary output of concern is the cumulative percolation, as this is the output from the
Clive H 1 D that drives the fluxes to groundwater in the DU PA model.
Observation Data
Observation data for the Monticello test cover included the following:
• Daily soil temperature at six locations in the cover system.
• Daily soil volumetric water content at six locations in the cover system.
• Daily soil suction at six locations in the cover system.
• Daily meteorological data for the site, including precipitation, wind speed, wind direction,
minimum air temperature, average air temperature, maximum air temperature, minimum
relative humidity, average relative humidity, maximum relative humidity, and average solar
radiation.
• Water balance data, some of which are presumably calculated from the above data, including
cumulative precipitation, soil water storage in the cover system, cumulative percolation,
cumulative runoff, and cumulative evapotranspiration (ET).
Meteorological Data Inputs
Meteorological data inputs are of critical importance to the validation effort, because realistic outputs
like percolation cannot be expected unless the atmospheric forcings applied to the model accurately
represent the real meteorological history of the site.
The period ofrecord for the meteorological data provided from the site was 11/06/2000 to
10/11/2022. The period of record for the water balance data ( cumulative precipitation, percolation,
runoff, ET) and soil monitoring data (soil moisture, suction, and temperature) was 8/12/2000 to
9/30/2022. No explanation for the mismatch in these periods of record was provided, so it is not clear,
for example, how cumulative precipitation was computed for 8/12/2000 to 11/06/2000 given that
there are no raw precipitation data during this period in the site record.
No data quality information was provided, though the meteorological data spreadsheet does appear to
use a highlighting scheme to indicate data quality. Raw precipitation values highlighted in blue appear
to be verified in some fashion, as they compared well with nearby weather station data. Yellow
highlighting in the raw precipitation record appears to identify erroneous values that were not
included in the cumulative precipitation record. The raw precipitation record in 2017, for example,
contains implausible values including an entry of 405 mm on 6/12/2017. For context, the average
total annual precipitation for the NWS Monticello Station No. 2 from 2013 to 2024 was 404.6 mm .
Several other implausible values in excess of 50 mm were identified with yellow highlights and were
excluded from the cumulative precipitation record. Gi ven that some attempt at data validation appears
to have been applied to the "Cumulative On-Site Precip" data, this dataset was considered to be the
most reliable source for precipitation that was provided from the site record.
However, other abnormalities were still present in the cumulative precipitation record when
comparing selected periods to other local weather station data. For example, the cumulative
precipitation record contained numerous high precipitation days (computed for each day by
subtracting the cumulative precipitation value from the previous day) in which the nearby NWS
stations did not record any precipitation. One example is 2/22/2017, in which 37.4 mm of
precipitation was recorded in the site record, but no precipitation was recorded at the nearby NWS
Page 20 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Monticello Station No. 2 or the NWS Northdale, CO station 20 miles to the east. Another example is
provided in Figure 12, where the site daily precipitation for December 2018 is plotted alongside the
NWS Monticello station data and data acquired from Daymet. Daymet is a data interpolation tool
developed by Oak Ridge National Laboratory that incorporates nearby weather station data and
interpolates it on a 1-km grid. The data shown are for the grid tile containing the Monticello site. The
site data shows a very high precipitation day (about 35 mm or 1.4 inches of rain) while the other two
datasets show no precipitation. Other than this day, the datasets show only minor differences in the
magnitude and timing of precipitation events for the period depicted. Given the other extreme values
recorded in this same timeframe, confidence in the site recorded data is questionable during these
periods. It seems plausible that these values were not identified in the preceding data quality effort
described above in which yellow highlighted values were excluded from the cumulative precipitation
record.
8
E E -6 C: .~ "' 0 ~
&
QJ 4 -~ "' :i E ::, u
0
-Site Cumulative Percolation
Site Precipiation
Daymet Precipitation
NWS Monticello Precipitation
60
50
10
0
2018-12-01 2018-12-05 2018-12-09 2018-12-13 2018-12-17 2018-12-21 2018-12-25 2018-12-29
Figure 12. Comparison of the site precipitation record with NWS station and Daymet data for December 2018.
Another apparent issue with the site's cumulative precipitation record is the period preceding a significant
percolation event in the Spring of 2010. The site's cumulative precipitation record shows minimal
precipitation for the 7-month period leading up to April 2010 (1.3 mm total from 9/1/2009 to 4/1/2010).
Figure 13 shows the daily precipitation record from the site data compared with the NWS Monticello station
data and data from Daymet for this period. As can be seen in Figure 13, both the NWS station and Daymet
show significant precipitation events (totaling about 350 mm in the Daymet record) in this period leading up
to the percolation increase that occurred in April 2010, while the site cumulative precipitation indicates there
was virtually no precipitation in the entire 7-month period. As a result, simulating the percolation event in
April 20 IO would not be possible using the site meteorological record. This abnormality in the site
precipitation record also partially explains the difference in total precipitation that can be seen in the water
balance plots below (Figure 19 and Figure 20). It is possible that other events of this type exist in the site
record.
Page 21 of 40
8
E E -6 C 0 ·.:,
Ill 0 ~ &
~ 4
·~
:i E ::, u
2
0
2009-05
~
E ERGYSOLUTIONS
2009-07 2009-09 2009-11 2010-01 2010-03
Mr. Doug Hansen
CD-2025-097
May 7, 2025
-Site Cumulative Percolation
Site Precipiation
Daymet Precipitation
NWS Monticello Precipitation
60
50
10
0
2010-05 2010-07
Figure 13. Comparison of the site precipitation record with the nearby NWS Monticello station and Daymet
(2009-2010).
The data quality issues with the site precipitation record were revealed by nonsystematic spot checking and do
not constitute a thorough data review. However, it is clear from the 2009-20 IO example that the site's record
is not sufficiently reliable to be used for the validation effort. The dataset acquired from Daymet was used in
lieu of the site data because it showed good agreement with the local NWS Monticello station data while
having very few missi ng days compared to the NWS station data. While this was deemed the most expedient
and simple approach, it is recognized that differences in the Daymet data and the site record (where it is
reliable) could cause discrepancies in the model results compared to the observed outputs.
Meteorological data from Daymet were prepared as input to the HID model using the same method described
in the Unsaturated Zone Modeling for the Clive DU PA white paper (Neptune 2025). In short, the Hargreaves
equation was used to compute the potential evapotranspiration (PET), and the soil cover fraction was used to
define the split between potential evaporation and potential transpiration. Snow accumulation and snowmelt
were handled using the same method described in Neptune (2025).
Page 22 of 40
~
ENERGYSOLUTIONS
HlD Validation Model Setup and Parameterization
Boundary Conditions
Mr. Doug Hansen
CD-2025-097
May 7, 2025
The boundary conditions were assigned in an identical fashion to the Clive HID model. The top of the model
was set as an atmospheric boundary, which is informed by the meteorological record as described above. The
bottom boundary is set as a free drainage boundary condition. As described below, the base of the sand layer
was used as the bottom boundary of the model. Daily atmospheric boundary condition inputs were prepared
for model runs from 8/13/2000 through 9/29/2022.
Layering and Discretization
Figure 14 shows the cover profile for the ET cover at Monticello. For the HID model, the cover was divided
into five layers over a depth of 200 cm. The top four layers all use fine-grained soils as a principle
component, while the top layer (Soil/Gravel Admixture) and the third layer (Animal Intrusion Layer, also
referred to as the Bio-intrusion Layer) include the addition of coarse material. Waugh et al. (2008) stated that
the soil-gravel admixture is intended to limit wind and water erosion and enhance seedling emergence and
plant growth. Waugh et al. (2008) described theAnimal Intrusion Layer as a "layer of cobble-size rock above
the capillary barrier [that] is an added deterrent should deeper burrowers enter the area. Fine-textured soil fills
the interstices of this rock layer, preventing it from behaving like a second capillary barrier." Waugh et al.
(2008) also characterized the top four layers of the cover collectively as the "Water Storage Layer (Sponge)."
It was noted that there were some minor discrepancies between the depths to layer interfaces provided by
Benson et al. (2008) compared to those in Waugh et al. (2008); the values in Benson et al. (2008) were used
in the HID model. The bottommost layer of the model was the Sand Layer, as the geomembrane layer at its
base formed the drainage lysimeter for the test section (Benson et al. 2008).
Page 23 of 40
900mm
300mm
300mm
300mm
600mm
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Soil/Gravel Admixture
+--Water Storage and
Frost Protection
◄ Animal Intrusion Layer
(native pediment gravels)
◄ Geotextile Filter
11111 Sand Layer and Capillary Break
11111 1.5-mm Geomembrane
(high-density polyethylene)
11111 Clay Barrier
Figure 14. Monticello test cell ET cover profile, from Benson et al. (2008).
The layering scheme used in the H 1 D model is summarized in Table 3. The geotextile filter layer between the
fine soil layer and the sand layer was not represented in the model. The domain was discretized into 1001
nodes with node spacing varying from 0.036 cm at the top of the domain to 0.364 cm at the bottom. This is in
keeping with the Clive HID modeling approach where the top of the domain has increased node density in
order to resolve the high pressure and moisture gradients resulting from the atmospheric boundary condition.
The top slope of the cover was assumed to be 5% as stated in Benson et al. (2011 ).
Page 24 of 40
~
ENERGYSOLUTIONS
Table 3. Summary of HID layering.
Layer Layer description Layer Thickness (cm)
I Fine soil with gravel 20
2 Fine soil 90
3 Fine soil with cobbles/gravel 30
4 Fine soil 30
5 Sand 30
Hydraulic Properties
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Depth to bottom (cm)
20
110
140
170
200
Properties for the model were mainly taken from Benson et al. (2011 ), which included soil water
characteristic curve (SWCC) estimates for the fine soil layer on a variety of testing scales. For the top four
layers, identical SWCC parameters were used for the van Genuchten fitting parameters a and n, which were
taken from the largest-scale testing documented in Benson et al. (2011). This conceptualization conforms to
the description in Waugh et al. (2008) that the top four layers act like a single storage layer without any
capillary breaks between layers. The residual water content, 0,, was set to zero for all layers. The saturated
water content, 0,, for the unamended soil layers (layers 2 and 4) content was taken from the smaller scale
testing as it had the most samples and is less likely to depend on soil size. The value used, 0.41, was also
reported in early work (Benson et al. 2008).
For the amended layers (layers 1 and 3) the porosity was reduced to account for the amendments of gravel
(top layer) and gravel/cobbles (bottom layer). While the exact proportions of coarse material used for these
layers was not readily apparent in the references cited, it was assumed that the Animal Intrusion Layer (layer
3) had a much more significant portion of coarse material to accomplish its purpose. The saturated hydraulic
conductivity of this layer was also assumed to be lower by half than the unamended layers. A saturated
conductivity of 5 cm/day was applied for the top two layers as well as the fine soil layer under the Animal
Intrusion Layer (layer 4). This parameter is likely to be uncertain because of the soil pedogenesis processes
that were ongoing during the service period that is modeled, including the impacts of freeze/thaw cycles,
animal burrowing, vegetation establishment, etc. This value is meant to be a representative value that is also
in keeping with the Clive H 1 D model, which uses about 5.9 cm/day on average for the top storage layers in
the latest model runs to represent the long-term average condition.
For the sand layer, less information was available in a review of the provided references. The values used are
within typical ranges in literature for sands (see, e.g., Figure 11 of Benson et al. (2014)), and provide a
capillary break because of the relatively high values of a and n when compared to the fine soil layers. The
saturated hydraulic conductivity used is typical for a sand and substantially higher than the fine soil layers.
The pore interaction term (I) was set to 0.5 for all layers, as was done in the Clive HID model. Table 4
summarizes hydraulic properties used for all five layers.
Page 25 of 40
~
ENERGYSOLUTIONS
Table 4. Hydraulic properties for the HID validation model.
Layer er es a(l/cm) Ks n (cm/day
1 0.00 0.38 0.0114 1.24 5
2 0.00 0.41 0.0114 1.24 5
3 0.00 0.25 0.0114 1.24 2.5
4 0.00 0.41 0.0114 1.24 5
5 0.00 0.32 0.075 2.1 300
Initial Conditions
l
0.5
0.5
0.5
0.5
0.5
Mr. Doug Hansen
CD-2025-097
May 7, 2025
The initial conditions for the model were set at a suction of 30,000 cm in the top four layers, and
1,000 cm in the bottom layer. This produced water contents of around 10% in the fine soil layers,
about 6% in the Animal Intrusion Layer, and near zero in the sand layer. These water contents
broadly agree with the earliest data in the site record, though the recorded suction values at the site
differ significantly. An analysis of the site's recorded water contents vs. recorded suctions might yield
some insights about the materials' SWCCs that could help to understand this. There could also be
complications with how water content values should be interpreted for the layers that use coarse
admixtures. This relatively simple way to assign initial conditions was adopted to expediently
complete the validation exercise.
Vegetation-related Parameters
A soil cover fraction of 0.35 was used for the validation model, which attempts to represent a roughly
time-averaged value at the Monticello site. Vegetation cover rapidly evolved in the early part of the
service period, and the goal of 40% cover was achieved in roughly 2006 (Waugh et al. 2009) and was
exceeded thereafter. The lowered value accounts for the early years of low vegetation, though it is
recognized that there is significant uncertainty for this value, and that the Clive approach for steady-
state vegetation cover will be less applicable in early time for this validation model because of the
rapid evolution of the vegetation.
The root distribution in the model was based largely on Figure 21 of Benson et al. (2008), reproduced
as Figure 15 below. A step function was applied over three depth intervals to define the relative
amounts of transpiration assigned to each node:
• 0 cm to 5 cm: 0.45
• 5 cm to 65 cm: 0.10
• 65 cm to 100 cm: 0.025
Page 26 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
All other plant-related parameters ( e.g., root uptake as function of matric potential) were set to the
same values as in the Clive HID model.
Results
0.0
0.0
0.2
0.4 -E -.c a. '3 0.6
0.8
1.0
•o
•o
0 •
<a
0 •
0
Normalized Root Density
0.2 0.4 0.6 0.8 1.0
• 0
o West Area
• East Area
·1 . 2 '--'--l--'---1----1--1--+--1--'--'--.+---J'--'----l--l-...,__-I--I--I-~
Figure 15. Normalized root density (Benson et al. 2008).
Summary results for the validation run are provided in Figure 16, which shows the water content in
the five layers as well as modeled vs. measured percolations and soil water storage. The model
predicted about 28 mm of percolation in total, with the vast majority of it (-24 mm) occurring during
the first major breakthrough event in 2005. The measured percolation totaled 9.2 mm of percolation.
Thus, the validation model is overpredicting percolation by about a factor of three.
Page 27 of 40
0.40
0.35
;,: 0.30
~ t 0.25
Q.
C ~ 0.20 C e
~ 0.15
~ 0.10
0.05
0.00
:::::=-
E ERGYSOLUTIONS
-Layerl
2000 2004 2008 2012 2016 2020
-Modeled percolation (mm)
-Measured percolation {mm)
2000 2004 2008 2012 2016 2020
Date
Figure 16. Validation model results.
Mr. Doug Hansen
CD-2025-097
May 7, 2025
600~----------~
E §
500
~ 400
~ ~ .,
j 300
-~
-Modeled storage (mm)
-Measured/calculated storage (mm)
200
2000 2004 2008 2012 2016 2020
The pattern of percolation events in the model output is consistent with the measured data, occurring
in an episodic fashion following periods of high precipitation. Between these high percolation events,
the ET cover behaves as designed by storing and releasing moisture and not allowing moisture fronts
to reach past the capillary barrier. Increases in percolation correspond with spikes in the water content
of the sand layer (Layer 5).
The model predicts much higher total soil water storage than the measured data, but the trends and
oscillations follow the same pattern. This could possibly be due to the Hargreaves equation
underpredicting PET, or the rooting parameters unrealistically stifling transpiration. Calibrating the
coefficient in the Hargreaves equation for this location might yield a better fit to observations, and the
plant root uptake parameter for S-shaped rooting function, hso, could be adjusted upward to allow the
plant roots to operate more efficiently at high tensions, which might reduce water contents in a large
portion of the domain. However, neither of these adjustments were attempted, as the objective here
was to apply the Clive H 1 D approach "as-is" to the Monticello data. Uncertainty in the hydraulic
parameters can also impact the soil water storage. Lastly, the Monticello site's reported soil water
storage values are likely calculated based on discrete measurements of water content, and, therefore,
some interpolation/extrapolation was likely involved, which makes this calculated value somewhat
uncertain. However, no details of this calculation were provided.
The overprediction of the 2005 breakthrough is partially attributed to the fact that Daymet data
precipitation values were substantially higher than the site-recorded values for this period. Figure 17
shows the 30-day rolling summation of precipitation (i.e., each point represents the precipitation of
the prior 30-day period) for the site-recorded values alongside the Daymet and NWS station data.
Page 28 of 40
E E
8
-6 C: 0 ·;:;
"' 0 t:! &
~ 4
-~
:i E ::, u
~
E ERGYSOLUTIONS
0 +-----------'
2004-11 2004-12 2005-01 2005-02 2005-03
Mr. Doug Hansen
CD-2025-097
May 7, 2025
-Site Cumulative Percolation
Site Precipiation
Dayme.t Precipitation
NWS Monticello Precipitation
2005-04
200
E E
E ::,
150 ~
~ e
>-"' '?
100 g
c:·
0 ·;:;
l'l i5. ·v
50 f
0
2005-05
Figure 17. Precipitation leading up to and during the 2005 percolation event.
Another contributing factor to this overprediction in 2005 is that the model did not produce any
runoff, while about 28 mm of runoff was recorded at the site from December 2004 through February
2005. This event is associated with a large amount of snowmelt, and the model does not account for
dynamics like freezing and compaction of the snow layer that can produce runoff atop the snow/ice,
or the fact that frozen ground can have much lower hydraulic conductivity than when under temperate
conditions. The model results suggest that the saturated conductivity of the top layer is sufficient to
accept all of the snowmelt as infiltration, whereas the measured data indicates a significant amount of
runoff. While runoff is a minor component of the overall site water balance, it could have an outsized
impact during these events by reducing infiltration to the cover during a percolation breakthrough.
Coupled with the higher precipitation in the Daymet record, these factors produced a large
overprediction of percolation for the 2005 snowmelt event. This was the most significant percolation
breakthrough in both the model and the measured percolation.
Another event with a difference in percolation between the model and the measured data occurs in
Spring of 2019. Figure 18 shows the 30-day rolling summation of precipitation (i .e., each point
represents the precipitation of the prior 30-day period) for the site-recorded values alongside the
Daymet and NWS station data for this period. Here, the site-recorded precipitation leading up to the
event is higher than the Daymet precipitation, and the model consequently underpredicts percolation
for this event.
Page 29 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
-Site Cumulative Percolation
Site Precipiation
Dayrnet Precipitation
NWS Monticello Precipitation
8 ~------------7 200 !-----------------~-
E E -6 C 0 ·;:;
"' 0 ~ &
Cl/ 4 > ·;:; "' :i
E ::, u
2
0
2019-01 2019-02 2019-03 2019-04 2019-05 2019-06
Figure 18. Precipitation leading up to and during the 2019 percolation event.
e E
E ::,
150 ::;_
~
>-"' -0
100 ~
c·
0 ·;:;
i 'i:i ·u so !!!
0
2019-07
"-
Water balance plots for both the site-recorded data and the model output are provided in Figure 19
and Figure 20, and show similar behavior throughout the period ofrecord. In both, percolation and
runoff are minor components of the overall water balance, although, as noted above, the model did
not produce any runoff during the simulation. Precipitation is closely matched by evapotranspiration
in both the model and the observed data. Total evapotranspiration in the site record was 97% of
precipitation and 96% in the model output. Total precipitation was higher in the model inputs
(Daymet record) than in the site record. As noted above, some of this difference is attributable to gaps
in the site record like the 2009/20 IO event shown in Figure 13. However, some spatial variability in
precipitation patterns is expected as is shown in Figure I 7 and Figure I 8. In early time, the site record
suggests that cumulative ET exceeds the cumulative precipitation until the large
precipitation/percolation event in 2005, while precipitation always exceeds ET in the model.
Soil water storage oscillates over a range of about 30 cm in both water balances in response to
variations in precipitation and ET, although the absolute value of soil water storage was higher in the
model, as noted above. The water balance plot for the model also includes the snow water equivalent
(SWE) snow layer in the simulation and shows that percolation events often correspond to snowmelt
events. While no snowfall or snow depth data from the site was provided, comparison of the timing of
snowmelt events in the model water balance with the site water balance suggests these snowmelt
events often produce runoff.
Page 30 of 40
700
100
0
2000
Precipitation
Evapotranspiratio
2004
~
ENERGYSOLUTIONS
Monticello Site Recorded Water Balance
2008 2012
Time (days)
2016
Runoff
-Percolation
-Soil Water Storag~
2020
Figure 19. Water balance for the Monticello site based on site-recorded data.
800
0
2000
Precipitation
Evapotranspiratio
2004
Validation Model Water Balance
2008 2012
Time (days)
2016
Snow Layer
Runoff
P\ercolation
-Soil Water Storage
2020
Figure 20. Water balance for the HID validation model of the Monticello site.
Mr. Doug Hansen
CD-2025-097
May 7, 2025
so
40
10
0
60
so
10
0
Page 31 of 40
~
ENERGYSOLUTIONS
Conclusion
Mr. Doug Hansen
CD-2025-097
May 7, 2025
A validation model was created by applying the Clive HID modeling approach to the cover design
and historical meteorological forcings at the Monticello site in order to demonstrate that the Clive
approach can reasonably reproduce real-world observations. Parameters were estimated from the
available data and no calibration was attempted. Because the Clive HID model uses steady-state
conditions for hydraulic parameters and those related to vegetation, these were also held constant in
the validation model despite evidence that conditions were transient at the site. The model produced
reasonable output that followed the trend of the observed data and showed percolation breakthroughs
at the same events, though the model overpredicted percolation for the most impactful event. Water
balances for the site data and the model output compared well, though the model did not produce
runoff as described above. Despite a variety of sources of uncertainty associated with the inputs, the
cumulative percolation prediction is within a reasonable range (i.e., within an order of magnitude) to
demonstrate that the Clive HID modeling approach is valid.
• 0-17 .a: The RF/ response presented "a mechanistic explanation for the capillary barrier performance
based on the theory of unsaturated flow" and an illustration of different soil water characteristic curves
that were realizations from previous simulations, however, this does not adequately address the
underlying consideration of a sensitivity analysis discussing capillary break.
When conducting the sensitivity analysis, please consider systematically varying a and n for the frost
protection layer and rerunning the model in each case.
A series of one-at-a-time (OAT) sensitivity analyses were run to address this RFI using the
information provided. Results of I 00 converged model runs using a gravel distribution for Layer 3 are
shown in Figure 7. Also shown in Figure 7 are the percentile values of percolation for those I 00
results. OAT simulations were conducted using the 5th, 25th, 50th, 75th, and 95th percentile
percolation simulations. For these five sets of percolation results, additional simulations were
conducted where the 5th and 95th percentile values of van Genuchten alpha and n from the developed
gravel distribution were used. Results of these OAT simulations are shown in Figure 21 , where it is
apparent that changing alpha and n has very little effect on percolation results for the cases tested. The
only large change was that percolation increased using the 5th percentile of alpha for the 95th
percentile infiltration simulation. In this case, percolation increased by a factor of 2.6, to a value of
0.042 mm/yr. There were no results that converged using the 5th and 95th percentile values of alpha
and n for the 95th percentile percolation simulation.
Page 32 of 40
... >-
0.05
0.04
E o.o3
E
C .Q ...., ro 8 0.02 ai a..
0.01
0.00
~
ENERGYSOLUTIONS
■ Layer3 Gravel
■ L3 Gravel alpha 5%
■ L3 Gravel alpha 95%
L3 Gravel n 5%
■ L3 Gravel n 95%
I I I II
5th 25th
I II I
50th
Percentile of Drainage
II I
75th
Mr. Doug Hansen
CD-2025-097
May 7, 2025
95t h
Figure 21. Results of OAA T simulations with varying van Genuchten alpha and n in Layer 3.
• 0-18.a: The response to this RF/ does not adequately address the potential impacts of bioturbation and
biointrusion towards the earthen covers. Please consider impacts of burrowing mammals, birds and
insects simulated by varying Ks, including how mixing by bioturbation at the textural contrast forming
the capillary break affect the efficacy of the capillary break and percolation through the cover.
The initial response indicated the effects of bioturbation on hydraulic performance were not explicitly
modeled considering potential impact to the hydraulic properties of the Federal Cell Facility cover
system. However, a broad distribution for the saturated hydraulic conductivity (Ks) of the radon barriers
is used for unsaturated zone modeling.
Please justify the use of Ks in determining flow through the Federal Cell Facility cover system as it
pertains to biointrusion and bioturbation.
The potential impacts of bioturbation and biointrusion on earthen covers are di scussed below. The
discussion includes the following: 1) results from a review ofliterature and past field studies at the
Clive site that analyze the presence and impact of mammals, birds, and insects to identify the primary
drivers ofbioturbation and biointrusion at the site, and 2) an evaluation of the impact ofbioturbation
on the Ksat values used in modeling based on an approach that applies power-law averaging on data
obtained from the literature review. Overall, the field data suggest that bioturbation and biointrusion
are likel y to occur at the site, but the impact on the average Ksat va lue used for modeling the cover as
a whole is relati vely small and does not change the distribution of infiltration values used in the
GoldSim model.
Page 33 of 40
;::::::=-
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
A variety of mammals, birds, and insects are present at the Clive site. Associated diversity and
bioturbation data were collected during field studies performed at the site in 2010 and 2012 (SWCA
2011, 2012). Larger mammals observed at the site included coyotes, badgers, and jack-rabbits, but
observations of these large animal burrows from the SWCA field studies are limited. The density of
burrows related to these larger animals is small compared to the density of smaller animal burrows
and, notably, harvester ant mounds. Birds common to the region include owls, the greater sage-
grouse, the non-native chukar, and the Hungarian partridge (SWCA 2012). Most of these birds nest
near the ground surface, limiting their impact on the material properties of the cover, but owls may be
of higher concern due to their burrowing behavior. Field data on owls is limited to a single family that
was observed in the 2012 study. Most of the burrows observed and identified in the SWCA field
studies are from smaller mammals, including ground squirrels, mice, voles, and rats, specifically the
kangaroo rat, and it is likely that the majority of these small mammal burrows were created by deer
mice.
In general, bioturbation activity drops off significantly beyond 1 m of depth (Bjornstad and Teel
1993), and studies of larger mammals at Hanford show few observations of burrowing activity
beyond 25 cm depth (Cadwell et al. 1989). Among the smaller mammals, deer mice burrows are
usually less than 18 cm in depth (Weber and Hoekstra 2009). The depth of the capillary barrier is
about 0.61 m (42"), making it unlikely that smaller mammals will reach and disturb this layer. It is
also possible that the capillary barrier itself acts as a bioturbation barrier for these animals due to its
material properties, as test samples indicate about 90% of the material in this layer is gravel sized
(> 2 mm or 0.079"), which would require extra energy for animals to excavate compared to material
above it. The field data collected at the Clive site suggest that these burrowing animals will make less
of an impact to the capillary barrier than insects, specifically the harvester ant.
Harvester ants are the most common type of ant identified at the Clive site, and the density of their
mound habitats exceeds the density of any animal burrows observed at the site (SWCA 2012). Over
99% of ant specimens collected and identified in SWCA studies ( 1624 total) were identified as the
species Pogonomyrmex occidentalis. Ants in general increase permeability and decrease bulk density
as a result of building nests, but the degree to which this occurs is related to ant species (Cammeraat
and Risch 2008; Zhou et al. 2023). Mounds analyzed of the ant species Formica cinerea Emery
measured Ksat values as high as 44182 cm/day in nests where channels and voids occupy around 8%
of the mound, but ant nest structure is complex and can vary by species (Drager et al. 2016; Green et
al. 1999). Other controlled experiments on the direct impact of ant nest formation on hydraulic
conductivity measured a Ksat value of 308.94 cm/day in soil after nest development (Drager et al.
2016). Nests can have many openings at the surface that lead into sub-surface pathways. Individual
nests have been found with up to 60 entrances (Cammeraat and Risch 2008; Moreira et al. 2004).
Some species of ants exhibit behaviors at the surface that make it difficult to characterize infiltration
through mounds at the surface, like the species Lasius neoniger Emery which reacts to precipitation
and actively closes nest openings during precipitation events (Wang et al. 1996).
There is a high density of observed harvester ant mounds at the Clive site, and the nests of the
Pogonomyrmex occidentalis species can reach depths up to 2.7 m below the ground surface (Lavigne
1969). This makes them a primary concern in regard to bioturbation and biointrusion within the cover
at the Clive site. Data collected on this species of ant from literature and the SWCA field studies were
used to quantify and estimate a potential increase in saturated hydraulic conductivity (Ksat) derived
from the radon barrier layer at the site. The results show that, despite the relatively large density of
ant mounds observed, the impact on Ksat is minimal at the scale of the overall Federal Cell Facility.
The SWCA field studies in 20 IO and 20 I 2 estimate an average of 24 ant mounds per hectare ( I/ha),
an average individual ant mound area of 1,900 cm 2, and an average mound volume of 19,345 cm 3.
This density is consistent with estimates of this species made in a different location by Rogers et al.
( 1972) of 23 ± 3 per hectare. Mounds at the Clive site can also be observed and measured remotely
Page 34 of 40
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
from publicly available Google Earth imagery (Figure 22). Ant mound density near test sampling
locations of the capillary barrier material was estimated to be about 48 mounds/ha based on satellite
imagery from October of 2023. The estimates from field studies and this remote analysis provide an
overall range in the possible mound density at the site, as well as an estimate of the total area covered
by mounds. Assuming the average mound area of 1,900 cm2, a density of 24 mounds/ha corresponds
to a proportion of total area covered by the mounds (pa) of 4.56e-4, and a density of 48 mounds/ha
results in Pa= 9. l 2e-4.
Figure 22. Google Earth imagery (dated 10/23) with ant mounds marked with red dots at a location near Clive,
UT; red box is 47,156 m2 (4.7156 ha) in area (approx. 270 m length x 174 m width) and is estimated to contain
228 mounds.
It is assumed that the Ksat distribution used for unsaturated zone modeling is representative of areas
undisturbed by ant mounds (geometric mean of 3.37 cm/day), but that is probably not true in the areas
where mounds exist. These small areas will likely have a higher Ksat value due to the tunnel spaces,
pathways, and voids within the nest structure. A lower estimate of this Ksat value may be
308.94 cm/day, from the study by Drager et al. (2016), whereas an upper estimate of this Ksat can be
used from the study by Green et al. ( 1999) of 44182 cm/day. For reference, this lower bound is more
than an order of magnitude higher than the 99th percentile of the Ksat distribution used in modeling. A
power-law average between the mean value from the model distribution (kmean) and the extreme value
from this study (kmax) can be calculated to determine how these mound areas of high Ksat may impact
an overall Ksat value (kfina1). This traditional approach to upscaling is outlined in Zhang et al. (2021)
and is often applied to determine the overall permeability of a material with differing properties, such
as rock with fractures throughout it. The approach is sensitive to the exponent chosen (m) and is
related to the permeability structure of the material (Zhang et al. 2021 ). An exponent of m = 1 is
appropriate when flow is parallel to a layered structure, whereas m = -1 is representative when flow is
perpendicular to a layered structure. At m = 0 the equation produces the geometric mean. A
Page 35 of 40
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
reasonable value for the exponent in this application is likely near 1, representing flow that is mostly
parallel. That is, infiltration through non-impacted regions is parallel to infiltration through ant mound
impacted areas, so the contributions to the overall infiltration are additive. Because of this, a simple
area-weighted average of the Ksat values of impacted and non-impacted regions is adequate to
determine an overall average Ksat for the cover that can be used in 1 D calculations. Calculations
using m values of 0.9 and 1 using the upper and lower bound ofkmax derived from literature are
shown in Table 5 and Table 6 below.
The equation to calculate the power-law average (kfioa1) can be written as:
kfinal = [ (1 -Pa)(kmean)(m) + (pa)(kmax)<m)] 1/m
where (pa) represents a proportion of total area covered by the mounds, calculated using the data
provided above, kmean represents the geometric mean of the Ksat distribution in the model, and kmax
represents the maximum possible Ksat observed within the area of the ant mound itself.
Table 5. Calcul ated kfioal using m = 0.9.
Mound density (I I ha) max (cm/day) kfinal (cm/day)
24 308.94 3.4681
24 44182 12.7669
48 308.94 3.5664
48 44182 22.9982
Using a value form of 0.9 and a mound density of 24 mounds/ha, the calculated kfinal ranges from
3.468 I to I 2.7669 cm/day for the lower and upper bounds ofkmax derived from literature. Using a
value form of 0.9 and a mound density of 48 mounds/ha, the calculated kfinaJ ranges from 3.5664 to
22.9982 cm/day for the lower and upper bounds of kmax derived from literature.
Table 6. Calculated kfioal using m = 1.0.
kfinal
Mound density (1 / ha) max (cm/day) (cm/day)
24 308.94 3.5093
24 44182 23.5155
48 308.94 3.6487
48 44182 43 .6609
Using a value form of I .0 and a mound density of 24 mounds/ha, the calculated kfinal ranges from
3.5093 to 23.5155 cm/day for the lower and upper bounds ofkmax derived from literature. Using a
value form of 1.0 and a mound density of 48 mounds/ha, the calculated kfinal ranges from 3.6487 to
43.6609 cm/day for the lower and upper bounds ofkmax derived from literature.
Page 36 of 40
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
All kfinaI values calculated using a range ofkmax values from literature fall within the modeled
distribution of Ksat. In fact, even in the extreme case with m = 1, a maximum density of mounds, and
the upper bound ofkmax, the calculated ktinaJ value falls within the range of the modeled distribution of
Ksat where the 99th percentile of that distribution of Ksat is 52 cm/day.
Overall, using a power-law average approach to adjust the mean Ksat to account for ant nest tunnels
beneath mounds observed at the surface shows that the limited areas of ant mound impacted cover
expected for the Clive site is not sufficient to push the average Ksat of the cover beyond the
distribution limits used in the model. These estimates of adjusted Ksat values likely overstate the
impact of ant nests because they consider the area of the mound measured on the surface to be
representative of the cross-sectional area of the nest at depth, which is probably not the case
considering literature suggests a decrease in nest tunnel density over depth (Drager et al. 2016). In
fact, the nest below what is observed on the mound surface is likely smaller in cross-section.
Applying an upper kmax value chosen from literature to the entire area will then be an overestimate of
the impact of the nest on the overall Ksat value at the site.
In summary, mammals, birds, and insects present at the Clive site may impact the material properties
of the cover through bioturbation in the form of burrowing. Mammal and bird burrows are relatively
low in density and unlikely to affect the material properties of the capillary barrier at the scale of the
full cover system . Harvester ant nest density is high at the site in comparison , and there is evidence
that these nests can extend to depths beyond the capillary barrier. Data on the area and density of ant
nest mounds collected at the Clive site were used to calculate an adjusted Ksat value using a power-
law average, and this adjusted value is still well within the range of the Ksat di stribution used in
modeling.
Page 37 of 40
~
ENERGYSOLUTIONS
• REFERENCES
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Benson, C.H., et al., 2011. Engineered Covers for Waste Containment: Changes in Engineering Properties
and Implications for Long-Term Performance Assessment, NUREG/CR-7028, Volume 1, United States
Nuclear Regulatory Commission, Washington DC, December 2011
Benson, C.H., et al., 2024. Field Hydrology of Armored Earthen Final Covers with and without Vegetation,
Journal ofGeotechnical and Geoenvironmental Engineering 150 (I) 04023125 doi :
10.1061/JGGEFK.GTENG-l l l l0
Benson, C.H., et al., 2014. Estimating van Genuchten Parameters a and n for Clean Sands from Particle Size
Distribution Data. In From Soil Behavior Fundamentals to Innovations in Geotechnical Engineering, edited
by M. Iskander, et al., pp. 410--426, American Society of Civil Engineers (ASCE), Reston VA
Benson, C.H., et al., 2008. Hydraulic Properties and Geomorphology of the Earthen Component of the Final
Cover at the Monticello Uranium Mill Tailings Repository, Geo Engineering Report No. 08-04, University
of Wisconsin-Madison, Madison WI, April 2008
Bjornstad, B.N., and S.S. Teel, 1993 . Natural Analog Study of Engineered Protective Barriers at the Hanford
Site, PNL-8840, UC-510, prepared for United States Department of Energy, Pacific Northwest Laboratory,
Richland WA, September 1993
Cadwell, L.L., et al., 1989. Animal Intrusion Studies for Protective Barriers: Status Report for FY 1988, PNL-
6869, UC-11, Pacific Northwest Laboratory, Richland WA, May 1989
Cammeraat, E.L.H., and A.C. Risch, 2008. The Impact of Ants on Mineral Soil Properties and Processes at
Different Spatial Scales, Journal of Applied Entomology 132 (4) 285-294 doi : 10.l l l l/j.1439-
0418.2008.0l 28 l.x
Drager, K.I., et al., 2016. Effects of Ant (Formica subsericea) Nests on Physical and Hydrological Properties
of a Fine-Textured Soil, Soil Science Society of America Journal 80 (2) 364-375 doi :
10.2136/sssaj2015.08.0300
EnergySolutions, 2020. Radioactive Material Licenses UT 2300249 Cover Test Cell Deconstruction Study
Final Report, CD20-0123, EnergySolutions LLC, Salt Lake City UT, August 2020
Globus, AM., and G.W. Gee, 1995. Method to Estimate Water Diffusivity and Hydraulic Conductivity of
Moderately Dry Soil, Soil Sci. Soc. Am. J. 59 (1995) 684-689
Green, W.P., et al., 1999. Structure and Hydrology of Mounds of the Imported Fire Ants in the Southeastern
United States, Geoderma 93 (1999) 1-17
Lavigne, R.J ., 1969. Bionomics and Nest Structure of Pogonomyrmex occidentalis (Hymenoptera:
Formicidae), Annals of the Entomological Society of America 62 (5) I I 66-1175
Mallants, D., et al., 2003 . Parameter Values Used in the Performance Assessment of the Disposal of Low
Level Radioactive Waste at the Nuclear Zone Mol-Dessel, Volume 2: Annexes to the Data Collection
Forms for Engineered Barriers, Restricted Contract Report SCK•CEN-R-3521rev.l, SCK•CEN, Waste &
Disposal Department, Boeretang Belgium, December 2003
Page 38 of 40
::::::=-
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
Moreira, A.A., et al., 2004. Nest Architecture of Atta laevigata (F. Smith, 1858) (Hymenoptera: Formicidae),
Studies on Neotropical Fauna and Environment 39 (2) 109-116
Neptune, 2023. Clive DU PA Model-Response to DWMRC 1-19-23 Request for Information, NAC-
0183_R0, Neptune and Company Inc., Lakewood CO, March 2023
Neptune, 2025. Unsaturated Zone Modeling for the Clive DU PA, Clive DU PA Model v4.0, NAC-0015_R6,
prepared for EnergySolutions, Neptune and Company Inc., Los Alamos NM, April 2025
Rockhold, M.L., et al., 2015. Physical, Hydraulic, and Transport Properties of Sediments and Engineered
Materials Associated with Hanford Immobilized Low-Activity Waste, PNNL-23711, RPT-IGTP-004, Rev.
0, Pacific Northwest National Laboratory, Richland WA, February 2015
Rogers, L., et al., 1972. Bioenergetics of the Western Harvester Ant in the Shortgrass Plains Ecosystem,
Environmental Entomology 1 (6) 763-768
Stantec, 2024. White Mesa Uranium Mill Cell 2 Reclamation Cover 2023 Annual Performance Monitoring
Report, prepared for Energy Fuels Resources (USA) Inc., Stantec Consulting Services Inc., February 2024
SWCA, 2011. Field Sampling of Biotic Turbation of Soils at the Clive Site, Tooele County, Utah, prepared
for EnergySolutions, SWCA Environmental Consultants, Salt Lake City UT, January 2011
SWCA, 2012. Vegetated Cover System for the EnergySolutions Clive Site: Literature Review, Evaluation of
Existing Data, and Field Studies Summary Report, prepared for EnergySolutions, SWCA Environmental
Consultants, Salt Lake City UT, August 2012
Tokunaga, T.K., et al., 2002. Saturation-Matric Potential Relations in Gravel, Water Resources Research 38
(10) 1214 doi: 10.1029/2001WR001242
Wang, D., et al., 1996. Ant Burrow Effects on Water Flow and Soil Hydraulic Properties of Sparta Sand, Soil
& Tillage Research 37 (1996) 83-93
Waugh, W.J., et al., 2009. Sustainable Covers for Uranium Mill Tailings, USA: Alternative Design,
Performance, and Renovation, ICEM2009-16369, proceedings of the 12th International Conference on
Environmental Remediation and Radioactive Waste Management, October 11-15, Liverpool UK, 2009
Waugh, W.J., et al., 2008. Monitoring the Performance of an Alternative Landfill Cover at the Monticello,
Utah, Uranium Mill Tailings Disposal Site, Waste Management 2008 Conference, Phoenix AZ
Weber, J.N., and H.E. Hoekstra, 2009. The Evolution of Burrowing Behaviour in Deer Mice (genus
Peromyscus), Animal Behaviour 77 (2009) 603-609 doi : 10.1 0l 6/j.anbehav.2008 .10.031
Zhang, X., et al., 2021. Application of Upscaling Methods for Fluid Flow and Mass Transport in Multi-Scale
Heterogeneous Media: A Critical Review, Applied Energy 303 (2021) 1-21 doi:
10.1016/j .apenergy.2021 .117603
Zhou, W., et al., 2023. Effects of Ant Bioturbation and Foraging Activities on Soil Mechanical Properties and
Stability, Global Ecology and Conservation 46 (2023) e02575
Page 39 of 40
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-097
May 7, 2025
If you have further questions regarding these responses to the director's requests of DRC-2024-
005076, please contact me at [phone].
Sincerely,
Vern C.
Rogers
Vern C. Rogers
Digitally signed by Vern C. Rogers
DN: cn=Vern C. Rogers,
o=EnergySolutions, ou=Waste
Management Division,
email=vcrogers@energysolutions.c
om,c=US
Date: 2025.05.07 13:51 :32 -06'00'
Director, Regulatory Affairs
enclosure
I certify under penalty of law that this document and all attachments were prepared under my direction or supervision in accordance with a system
designed to assure that qualified personnel properly gather and evaluate the information submitted. Based on my inquiry of the person or persons who
manage the system, or those persons directly responsible for gathering the information, the information submitted is, to the best of my knowledge and
belief, true, accurate, and complete. I am aware that there are significant penalties for submitting false information, including the possibility of fine and
imprisonment for knowing violations.
Page 40 of 40