DRC-2025-001078

~
Div of VVa.;t'" Management
and Rad1at1on Control
MAR 2 7 2025
--------ENERGYSOLUTIONs --------
March 27, 2025
Mr. Doug Hansen, Director
Division of Waste Management and Radiation Control
195 North 1950 West
Salt Lake City, UT 84114-4880
CD-2025-060
Subject: Federal Cell Facility Application: Responses to Round 2 Requests for Information (per DRC-
2024-005227, DRC-2024-005438, DRC-2024-005545, and DRC-2024-005704)
Dear Mr. Hansen:
Energy Solutions hereby responds to the Utah Division of Waste Management and Radiation Control's April
17, 2024 (DRC-2024-005227), April 25, 2024 (DRC-2024-005438), May 14, 2024 (DRC-2024-005545), and
May 31, 2024 (DRC-2024-005704) Requests for Information (RFI) on our Federal Cell Facility Application.
A response is provided for each request using the Director's assigned reference number. Note that Director-
assigned RFI numbers O-40.a and O-40.b are repeated in letters dated April I 7, 2024 and April 25, 2024, with
different topics.
Each RFI letter is addressed individually, with the text of the RFI quoted in bold italics followed by its
response. Multiple follow-up RFI letters are addressed in this response, with each letter preceded by a solid
line.
The following attachments are electronically provided on the attached compact disc:
1. REPORT: Geotechnical Modeling for the Clive DU PA -Clive DU PA Model v3.0 (NAC-0025_R6)
(Neptune 2024)
2. REPORT: Deep Time Assessment for the Clive DU PA -Deep Time Assessment for the Clive DU
PA Model v4.0 (NAC-0032_R7)(January 9, 2025)
3. FOLDER: VMINTEQfiles
4. FOLDER: references for RF/ responses
• distributable references ((the materials may be freely duplicated or transmitted)
• non-distributable references (the copyrights of these materials prevent secondary duplication or
transmission without specific permission of the author/publisher)
Appendix 0: Federal Cell Facility Waste Characterization Plan
• 0-39.a: (from DRC-2024-005227): The Division has reviewed the literature and reports provided by
EnergySolutions/Neptune regarding data scaling, and while the importance is understood, documented
justification for the approach has not been located. The Division is not convinced data scaling is being
conservatively applied in NAC-0032_R6 of the DU PA (Deep Time Assessment). The response to the
request for justification was "This is a technically correct choice." (NAC-0147_R0, Section 2.3.1).
To avoid underestimating Deep Time impacts, it is important that the amount of aeolian deposition
estimates be conservate and defensible. The response needs to be supported by providing the
mathematical derivation as to why it is the technically correct choice. Please provide a White Paper or a
299 South Main Street, Suite 1700 • Salt Lake City, Utah 84111
[phone] • Fax: [phone] • www.energysolutions.com
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
peer reviewed reference that includes the technical basis taken in NAC-0032_R6for aeolian deposition
data scaling as the correct approach.
In general, deep time sedimentation impacts are modeled in the DU PA as a combination of deep lake
deposition, intermediate lake deposition, and aeolian deposition. Relevant data can be extracted from
the physical record of deep cores in the Bonneville Basin. This historical record suggests total
deposition per I 00 ky (Milankovitch) cycle to be around 15 m. The deposition rates in the model are
effectively calibrated to that value, which represents the historical record, or the best science and data
available. For example, in Neptune NAC-0032_R6 :
Brimhall and Merritt (1981) reviewed previous studies that analyzed sediment cores of Utah
Lake, a freshwater remnant of Lake Bonneville that formed at approximately 10 ka. They
suggest that up to 8.5 m of sediment has accumulated since the genesis of Utah Lake,
implying an average sedimentation rate of 0.85 mm/y or 850 mm/ky. Within the Bonneville
basin as a whole the major lake cycles resulted in substantial accumulations of sediment
based upon th e depth of the cores analyzed (e.g., a 110 m core that corresponds to the past
780 ky, or four deep lake cycles (Oviatt et al. 1999)). This accumulation averages about
140 mm/ky. Einsele and Hinderer (1997) indicate that sediment accumulation in the
Bonneville basin occurred at a rate of 120 mm/ky during the past 800 ky. The Knolls Core
suggests that there has been 16.8 m of sediment formed in the last glacial cycle, or nearly
170 mm/ky.
The Burmester core indicates a rate of 14 m per I 00 ky. The work of Einsele and Hinderer (Einsele
and Hinderer 1997) on the Knolls core suggests a rate of at least 17 m for the last I 00 ky cycle.
Consequently, the overall sedimentation rate in the model represents these data. Note that the
Burmester "data," for example, are averaged over 780 ky. Averaging and scaling are the same in this
context. The model is calibrated against an average total deposition rate, rather than what might have
occurred every year, or every 1 ky, etc. In this instance the data are already essentially scaled to the
scale of the model. This is not always the case with input data, and sometimes statistical instead of
physical approaches to scaling (averaging) are needed. Note that the unit of interest for deep time
sedimentation after the current I 00 ky cycle ends is m/100 ky. This unit is used because it seems
sufficient to model the distant future and matches the natural Milankovitch cycles. Consequently, the
future cycles model total sedimentation, and do not model the three separate components. The current
cycle is modeled differently; it separates the three components because of the need to model the
impact of the first returning intermediate lake. Given the total data, the system would not be modeled
correctly if any of the three components of sedimentation were under-estimated. To match the total
sedimentation rate in the historical record, this would require adjusting by over-estimating another
one of the components. This risks creating inconsistencies in the way the system is modeled. Where
aeolian deposition is concerned, the best available data have been used, and a distribution has been
developed that statistically or mathematically averages data in a similar way to how the Burmester
core physically averages data. The scale of the data and the scale of the model need to match,
otherwise the model given the data will not make sense.
Regarding the aeolian deposition rate distribution, in response to O-39.b a mistake is acknowledged.
Perhaps this mistake addresses the overall concern about aeolian deposition. The wrong distribution
has been used in the model, and the correct distribution, although available in NAC-0032_R6, is not
included in the upfront table.
Page 2 of28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Nevertheless, based on field study measurements by Oviatt and Crowe, as well as those described by
Oviatt (Appendix A ofNAC-0032), a constant sedimentation rate of 5.3 x I 0-5 m/yr was applied. For
the same area, Brimhall and Merritt (198 I) reported deposition rates of 8.5 x 10-4 m/yr. In Oviatt
(2024), the author synthesized several field studies in the area and reported deposition rates ranging
from 9 x 10-5 to 2.6 x 10-4 m/yr. The modeled aeolian deposition rate, 5.3 x 10-5 m/yr, represents a
low deposition rate relative to those reported in the literature. However, the distribution used is based
on site-specific data, and hence is used in preference to other data sources.
Regarding possible references for scaling sedimentation rates, there is not a lot ofrelevant literature,
but there is some. A paper by Wikle and Berliner (2005) describes the reasons why scaling is
necessary when developing complex, mostly physical-based models. Not many studies have been
performed for sedimentation rates in deserts or even other environments, and essentially no such
sedimentation rate studies considered such long-term consequences as the DU PA. Note that the way
long-term PAs are built is to use existing knowledge and project it into the future. The existing
knowledge is based on past and present data -there are no future data. However, in the
sedimentation literature there are some references that point to data that have been averaged, and
Schumer and Jerolmack (2009) provide a spatio-temporal scaling methodology consistent with the
representation of sedimentation rates in the Deep Time Model. Shum er and Jerolmack demonstrate
that average accumulation rates approach a constant value at very long time scales, which applies to
the Deep Time Model simulation period of 2+ million years, and also applies to the approximately
60 ky until the first intermediate lake might arrive in the current model. In addition to those
references already cited, average sedimentation rates are also discussed in Neff et al. (2008), Rea and
Leinen ( I 988), Inoue and Naruse ( 1987), and Reheis et al. ( 1995).
• 0-39.b: (from DRC-2024-005227): The approach taken for Intermediate Lake sedimentation in NAC-
0032 _ R6 was based upon the interpretation of lake cycles and sediment thickness from a Clive pit wall
description by C. G. Oviatt, PhD, Professor of Geology, Kansas State University, as documented in NAC-
0032 _ R6, Table 3 (Lake cycles and sediment thickness from Clive pit wall interpretations (C. G. Oviatt,
personal communication)). The Division has reviewed the data documented in NAC-0032_R6 and
compared it to other data, such as the lake sedimentation data provided in NAC-0105_R0. Based on
review of the data provided by Oviatt (Oviatt, et al 1999, Oviatt 2019), Jewell (Jewell 2014), and Neptune
(Neptune 2018, Neptune 2021), it is concluded that the Clive DU PA Intermediate Lake sedimentation
model does not represent existing lake sedimentation data and is non-conservative.
The Division finds that in the use of an intermediate lake sedimentation rate based on the total lake
cycles, Neptune's document NAC-0032_R6 speculates that there would be multiple short-term
transgressions and regressions in lake elevations at Clive and constructed a heuristic model to evaluate
short-term variations. To avoid underestimating the Deep Time impacts, it is important that the estimates
of the amount of aeolian deposition and lake sedimentation be conservate and defensible. Please provide
documentation that the intermediate lake sedimentation model is conservative as presented or revise the
intermediate lake sedimentation model to be conservative with existing lake sedimentation data.
A mistake has been identified in the development of intermediate lake sedimentation rates. The Deep
Time Assessment technical report (NAC-0031 _ R6) indicates in Table I that the "sediment thickness
for each intermediate lake event" is modeled using a lognormal (LN) distribution with a geometric
mean (GM) of 2.82 m and a geometric standard deviation (GSD) of 1.7 I m. However, at the bottom
of page 40 of the same document, sediment thickness for individual lake events is modeled as
Page 3 of28
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
LN(GM=0.75 m, GSD=l.4 m). Subsequently, on page 38 it is indicated that the distribution for all
intermediate lake events (in a I 00 ky cycle) is modeled as LN(2.82 m, 1. 71 m). The two distributions
should have been used for different purposes in the deep time model. Note that the LN(2.82, I. 71)
distribution has units of meters per 100 ky, whereas the LN(0.75, 1.4) distribution has units of meters
per lake event. The LN(2.82, 1.71) distribution should have been applied only to total sedimentation
in a 100 ky cycle (and then added to the aeolian and deep lake sedimentation components). The
LN(0.75, 1.4) distribution should have been applied to the first lake event. However, the LN(2.82,
1.71) distribution was incorrectly applied to both cases.
As a consequence, the DU PA model in GoldSim will be updated so that the first lake event is subject
to the LN(0.75, I .4) distribution, representing sedimentation per lake event, and the LN(2.82, 1. 71)
distribution will only be applied to sedimentation per 100 ky cycle. The technical report has been
updated and is included with this response. Because the DU PA model will need to be updated for
other round 2 RFI responses, the revised model is not included with this response. Once all RFI
responses have been determined, those which require changes to the DU PA will be entered into the
model. The model will then be re-run, and updated results will be presented in the final report.
The LN(2.82, 1.71) distribution arises from a calibration of the intermediate lake sedimentation rate
to the overall 100 ky sedimentation rate. It roughly corresponds to four intermediate lake events per
100 ky cycle. The mean ofLN(0.75 m, 1.4 m) is about 0.8 m, and the mean ofLN(2.82 m, 1.71 m) is
about 3.2 m (geometric means are always less than means for lognormal distributions). In effect, DU
PA v3.0 has roughly four times as much sedimentation for the first lake event as was intended. This
will be corrected in DU PA v4.0.
• 0-40.a: (from DRC-2024-005227): NAC-0025_R4 (Geochemical Modeling) states that literature studies
were selected which best represent the site conditions. While this is an understandable approach, little
evidence is provided (Section 4) to support the statement. In addition, because the sorption coefficients
are empirical, it is not generally recommended to use a Kd value determination from one site to another
site. The use of the same empirical coefficients at different sites introduces a large amount of uncertainty
into the model due to the variability in the coefficients and geochemical conditions.
Please provide supplementary evidence that the selected range of partitioning coefficients from the
literature were collected under geochemical conditions and with comparable mineralogy to the
conditions at the Clive site. Additionally, please provide information of the overall technical approach of
selecting partitioning coefficients from the literature as opposed to simulating them using site specific
conditions.
Neptune agrees that incorporating research data developed from sites unrelated to the Clive site
introduces uncertainty in the inputs. This uncertainty is incorporated into the Kd values as input
distributions that are wider than the literature data minimum and maximum, which were assigned to
be the 5th and 95 th percentiles, respectively, as described in Section 3.0 ofNAC-0025_R4. Sensitivity
analysis helps identify which inputs affect the results most significantly and can lead to decisions to
refine input distributions if those inputs affect results to an unacceptably large degree.
Page 4 of28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
In Section 3.0 ofNAC-0025_R4 1, Neptune defines the process to select literature data for Kd input
distributions which best fit the conditions found at the Clive site as follows:
" ... general literature were screened to retain studies relevant to the Clive site, any value
within the range of those studies was deemed to be "equally viable," given th e uncertainty
associated with the various soil and water characteristics for th e site."
Note that the literature reviewed includes Kd data collected for the site, such as Bingham
Environmental ( 1995) and Adrian Brown Consultants (I 997). Because pH and mineral texture are
generally important factors that affect adsorption, these parameters were considered carefully when
they were provided in the literature studies. Some literature studies, like Sheppard and Thibault
(I 990) and, to some extent, the EPA series (EPA 1999a, 1999b, 2004 ), did not include pH values for
all of their Kd data but these studies were included in the input distributions. The use of the range of
those Kds potentially makes the Kd input distributions wider than they would otherwise be if pH
information were stated and data were culled based on pH. Most references clarified mineralogy or
mineral classes, such as sand, silt, and clay, as presented in Sheppard and Thibault (1990), for
example. Other references, such as Seme (2007), were considered that have pH values and organic
carbon contents similar to the Clive site, as noted in Section 4.0.
In Section 4.0 ofNAC-0025 _R6 (Neptune 2024), the last paragraph has been edited to provide clarity
in the statements comparing conditions in literature data with the Clive site conditions (quoted here in
redline/strikeout to make the revisions readily apparent):
"Of note is that the Hanford soils are slightly acidic (pH 6.2 to 7.8), with organic content of
0.5 to 1.5% organic carbon, which is slightly higher than that of the Clive location, which
has organic carbon contents of approximately 0.3% to 1%. Serne (2007) also reviews a
number of studies that are equally viable to th e Clive facility and the range of K d values
provided are useful as a first comparison."
Table 1 summarizes characteristics of the experimental samples and site conditions in the various
references for Kds, along with the characteristics of the Clive site. Where it was simple to extract pH
values and other information from references that contain great amounts of information, those values
are given in the table. Where pH ranges were large, those are noted in the table. Blank cells indicate
that insufficient information was present to provide data in the table cells.
1 NAC-0025_R6 (Neptune 2024) has been prepared in response to the follow-up RFls and is included with this
response. The quoted text remains unchanged in thi s revision. NAC-0025 _ R5 was prepared in December 2023 to
update uranium solubility for reproducibility of geochemical modeling. The questions raised in these round 2 RFis,
based on version NAC-0025_R4, remain applicable.
Page 5 of28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Table 1. Geochemical conditions for Kd references as compared to the Clive Site.
Comparison of References for Kds to Clive Site Conditions
Reference pH Soil matrix Eh Bicarbonate
Unit 2 (clay with
silty sand layers)
6.62 to Unit 3 (silty sand -43 .70 to Clive Site 7.69 with layers of clay 500 102 to 350
and silt)
Unit 4 (silt and
clay)
Whetstone Associates (2009) Soil density
1.566 g/cm3
Bingham Environmental (1995) 7.52 Silty sand (Unit 3) 160 243
Bingham Environmental (1996) 7.52 Silty Sand (Unit 3) 103 192
Scism (2006) 8.2 to 8.7 Alluvium (particle 189 to 3.0E-03 to
size 75-2000 um) 212 4.7E-03
Sheppard and Thibault (1990) Sand, silt, clay
Triay et al. (1997), Fortymile 6.9 Tuff units 340 143 Wash
Triay et al. (1997), Yucca Mt 6.5 to 7.5 Paintbrush Tuff -143 to "difficult to
600 quantify"
Triay et al. (1997), Midway Paleocarbonates between Yucca Mt and 6.7 under Tuff 360 698
Fortymile Wash
Decker and Papelis (2003) Variable Various
6.37 to Silty clay (Unit 2) 220 to Envirocare (2000) 7.87 and silty sand (Unit 400 3)
EPA (1999a) Wide Various Range
EPA (1999b) Wide Various Range
EPA (2004) Wide Various Range
Organic C 0.5
Serne (2007) 6.2 to 7.8 Various to 1.5% (Clive
0.3to1%)
Last et al. (2004) 7.66 to Various 8.17
Krupka et al. (2004) Used data Various, especially
~8 sand and gravel
Page 6 of28
~
ENERGY SOLUTIONS
Comparison of References for Kds to Clive Site Conditions
Reference pH Soil matrix
Adrian Brown Consultants
(1997) 7.5 to 7.6 Silty sand (Unit 3)
Bingham Environmental (1991) 6.3 to 8.5 Silty clay/sand
(Units 2-4)
Glover et al. (1976) 7.7 to 8.2 Silty sand to sand
7.29 to 70-97% sand -Um et al. (2009) 7.92 3-30% silt
Blank -Information not readily available in the literature.
Eh
160 (test
sample
based
upon
average
of 4 well
samples)
45-430
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Bicarbonate
197 to 322.5
4 to 11
(groundwater)
1.3 to 3% by
weight as
CaCO3
Regarding using literature to inform Kd input distributions rather than simulation of adsorption
coefficients using site-specific conditions, simulation of adsorption using geochemical computer
codes, such as Phreeqc (Parkhurst and Appelo 2013), requires input values that are not easily obtained
for a site. Adsorbent properties of site minerals, such as number and type of adsorption sites and
surface area of minerals, are not known data for various materials at the Clive site and this could
create great uncertainty in adsorption modeling results -greater uncertainty than when using
literature references. Assumptions can be made in order to run adsorption models, and these types of
models can be helpful for understanding changes in water chemistry for a fixed material adsorption
capacity, for example. The use of literature data to inform Kd distributions allows for the acquisition
of data and distributions that span the likely Kd for the site. Data that are from experiments in
conditions similar to the site were reviewed and minimum and maximum values for each element in
that collection of relevant sites were expanded slightly in the development of the log-uniform input
distributions. This approach also includes uncertainty and variability for each element and has a basis
in experimental results rather than pure modeling results.
■ 0-40.b: (from DRC-2024-005227): In NAC-0025_R4 (Geochemical Modeling), many of the partition
coefficients and the solubility of constituents described in Sections 4 and 5 are based on literature
reviews. The Division has been unable to verify their applicability to the Clive site and recommend
simulating the exact speciation under site-specific conditions using the geochemical parameters provided
in Tables 7 and 8 of NA C-0025 _ R4 to provide better site applicability. This effort would be useful to
ensure that the partitioning coefficients and solubility value ranges selected in this report match the
geochemical conditions under which the literature values were collected. (Note: Visual MINTEQ was
discussed for modeling some of the uranium speciation in this report.)
Using the values in Tables 7 and 8, please provide site specific speciation simulations for each of the
constituents considered in the report.
Page 7 of 28
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
In the initial version ofNAC-0025 _R4 (Geochemical Modeling), the partition coefficients (Kd) and
solubility estimates of constituents discussed in Section 4 and Section 5 were based on a
comprehensive literature review, with the exception of modeling UO3 and U3Os for uranium
solubility in the waste. Recognizing the Division's concerns regarding the direct applicability of
literature-based parameter estimates to the specific geochemical conditions at the Clive site,
additional detailed modeling was conducted to strengthen the basis of these estimates and to support
site-specific reliability.
The performed modeling employed the geochemical modeling software Visual MINTEQ (VM),
chosen in part due to its streamlined application. VM enables simulations under controlled,
representative geochemical conditions, which can be reflective of site-specific groundwater
chemistry, redox conditions, and pH levels, all of which are identified from Clive site groundwater
monitoring data. While the modeling is limited by the extent of available thermodynamics databases
within VM and assumptions of mineral forms ofradionuclides, it enhances the confidence in
solubility and adsorption parameter distributions for some of the key elements. This supplemental
modeling specifically addressed constituents for which sufficient thermodynamic and mineralogical
data were available. Elements lacking sufficient database coverage within Visual MINTEQ only show
results for literature-derived values due to current constraints in modeling capabilities.
Visual MINTEQ was used to simulate solubility coefficients and to provide input parameters needed
to model adsorption coefficients. Visual MINTEQ was used over other popular geochemical
modeling software such as PHREEQC or The Geochemist Workbench due to its simplicity in
speciation calculations compared to the more complex setup needed in other modeling software
(Khalidy and Santos 2021). Visual MINTEQ was also used for prior uranium solubility modeling
(NAC-0025_R6 Sections 5.1.14.1-5.1.14.3) and is thus consistent with the previous analysis.
However, Visual MINTEQ has limited chemical elements and solid phases, and thus is not able to
perform a complete analysis of all chemical elements listed in Table 2 ofNAC-0025_R4. The VM
database of components includes Am, Cs, I, Pb, Np, Pu, Sr, Th, and U, and these elements were
considered for both solubility and adsorption coefficient modeling.
Solubility Modeling
To model solubility, the solid phase present, i.e., the mineral that is dissolving or precipitating, is
needed as a model input. For the uranium solubility analysis previously completed, U3Os and UO3
minerals were specifically modeled to get those solubility distributions, knowing those minerals are
present as the waste form of uranium. For other elements, the exact solid phase or mineral that will
control solubility at the Clive facility is not known . Choices made for the solid mineral present can
result in orders of magnitude difference for solubility. To provide a range of modeled solubility,
several solid phases were selected for each element, based on the geochemical conditions and the
solubility discussion in NAC-0025_R6 Section 5.1. The default Visual MINTEQ database for
thermodynamics, the NIST Critical Stability Constants database, was used . This database has a
limited number of solid forms, and does not include any mineral species for Ac or Cs. The solubility .
analysis was therefore limited to Am , I, Pb, Np, Pu , Sr, and Th. A list of the solid phases selected for
modeling is provided in Table 2 below.
Page 8 of28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Table 2. Mineral Phases Selected for Solubility Modeling of Various Elements Using Visual
MINTEQ.
Element Mineral
Am Am(OH)3 (am)
Am(OH)3 (c)
Am2(C03)3
AmF3
I Pb'2
Pb Hydrocerrusite
Chloropyromorphite ( c)
Chloropyromorphite
(soil)
Cerrusite
Anglesite
PbCl2
Pb(OH)2
Pb(OH)CI
PbCl(OH)
Pb2(C03)Cl2
Pb2( OH )3CI( s)
PbBrF(s)
PbFCI
Np Np02C03
K3Np02(C03)2
KNp02C03
NaNp02C03:3.5H20
Pu Pu(OH)4 (am)
Pu02(0H)2(s)
Pu02C03(s)
Sr Strontianite (SrC03)
Celestite (SrS04)
SrF2
Th Th02(aged)
Th02 (c)
Th02 (fresh)
NAC-0025 _R4 Tables 7 and 8 were used to define geochemical conditions in simulating equilibrated
mass distributions of a given solid dissolved in solution. For each mineral species, Visual MINTEQ
was used to simulate solubility coefficients at three pH levels (6.6, 7.15, and 7.7) and two bicarbonate
concentrations ( 100 mg/Land 350 mg/L), representing the range of measured site values. The redox
potential was fixed at 200 m V, representing the mean site value. Mean values of anions and cations
were used from Table 8 ofNAC-0025_R4 for groundwater concentrations.
Page 9 of28
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Results of the solubility modeling are presented in Table 3 and Figure 1 below, along with the
distributions from the comprehensive literature review. The modeled range of americium solubility
across Am-containing minerals was higher than the literature estimate. The literature solubilities for
Am were based on Triay et al. (1997), which references Nitsche et al. (1993) laboratory experiments.
These data are for AmOHCO3 mineral phase, which has a solubility of about 1 E-9 mol/L at 25C. This
mineral phase is not in the MINTEQ database. The modeled range of solubility coefficients for
neptunium, lead, plutonium, strontium, and thorium were all within the ranges of the literature
estimates. The modeled range of solubility for strontium was smaller (6.1 0E-04 to 6.64E-03 mol/L)
than the range estimated based on literature (6.81E-7 to l.47E-3 mol/L). The modeled range of
solubility for thorium (3.42E-14 to l.68E-07 mol/L) was larger than the range estimated based on
literature (7.74E-9 to l.29E-6 mol/L). Iodine solubility was modeled but resulted in an unrealistically
low solubility value due to the only available solid phase for iodine in the Visual MINTEQ database
being Pbl, which would not precipitate without Pb in solution. With Pb provided as cerrusite as an
infinite solid, iodine solubility ranged from 9.65E-03 to 1.46E-2 mol/L.
Several constituents were not modeled due to limitations within Visual MINTEQ's components
database (e.g., Ac, Pa, Ra, Rn, and Tc) and thermodynamics database (e.g., Ac, Cs). Literature-
derived solubility values remain the recommended basis for these elements without an extensive
modeling review oflikely solid minerals at the Clive site and exploring other solubility databases or
modeling tools.
Table 3. Comparison of Literature-Derived and Modeled Solubility Coefficients
for Key Constituents at the Clive Site.
Solubilities Literature Modeled
Chemical Min (mol/L) Max (mol/L) Min (mol/L) Max (mol/L) Element
Ac 6.81 E-09 1.47E-05 --
Am 6.81E-10 1.47E-06 2.46E-06 9.98E-02
Cs 6.81 E-03 1.47E+01 --
I 5.99E-05 1.67E+00 9.65E-03 1.46E-02
Np 6.81 E-06 1.47E-02 1.04E-06 2.15E-02
Pa 6.81 E-09 1.47E-05 --
Pb 6.81 E-09 1.47E-05 9.23E-08 8.41 E-03
Pu 5.27E-11 1.90E-05 1.46E-1 0 1.69E-04
Ra 5.99E-10 1.67E-05 --
Rn 7.74E-04 1.29E-01 --
Sr 6.81 E-07 1.47E-03 6.10E-04 6.64E-03
Tc 7.74E-05 1.29E-02 --
Th 7.74E-09 1.29E-06 3.42 E-14 1.68E-07
U* 3.58E-06 2.79E-03 --
U as LJ3Qa** 4.81E-17 2.40E-10 --
U as UO3** 1.74E-04 5.45E-03 --
Page 10 of28
T
I 0
I
~
ENERGYSOLUTIONS
T T
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Il ~ 10·3 I I r :::: 0 .s -C
(V ·c::;
!E (V
0 (.)
>, ....,
.0 :::,
0 Cf)
I I T I I
Il
I
I I ..L l I I h I 10•7 I
I
I I
I I
.L I
I
10·11 I
Literature I
I Model I
.L
Am Np Pb Pu Sr Th
Chemical Element
Figure 1. Comparison of Modeled Solubility Distributions with Literature-Based Estimates for
Key Elements at the Clive Site.
Partition Coefficients
The following elements have site-specific data included in the modeled distributions.
• cesmm
• iodine
• neptunium
• technetium
• uramum
For this response, partition coefficients were estimated using a linear adsorption isotherm, with input
information provided by Visual MINTEQ via surface complexation models (Ayawei et al. 2017).
Similar to solubility coefficients, we are limited in use of available elements within the Visual
MINTEQ databases (e.g., Am, Cs, I, Pb, Np, Pu, Sr, Th, and U). For a given element, Visual
MINTEQ was used to simul ate adsorption coefficients at three pH levels (6.6, 7.15, and 7.7), two
bicarbonate concentrations (100 mg/L and 350 mg/L), and with a fixed redox potential of 200 m V.
Mean groundwater concentrations for anions and cations were provided by Table 8 of the
Geochemistry white paper (NAD-0025_R4). The mean solubility reported in Table 2 (NAC-
0025_R4) was used as the concentration for each element of interest, with the exception of uranium.
A lower uranium concentration of l .00E-5 mol/L was required in order to achieve model
Page 11 of 28
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
convergence. Under each of those combinations, a surface complexation model was run to provide an
estimate of the total dissolved and total sorbed concentration of each element. Those outputs of the
surface complexation models were input into the linear adsorption isotherm equation to provide a
modeled estimate of Kd (mL/g). In the linear adsorption isotherm model, Cs is the concentration of a
given element in the solid phase, and c is the concentration of a given element in the fluid phase:
Two different surface complexation models were used to provide estimates of Cs and c to the linear
adsorption isotherm model: 1) the Gibbsite-DLM (diffuse layer model) (Karamalidis and Dzombak
2011 ); and 2) the fixed-charge site model. Both were run using a single complexation surface. The
Gibbsite-DLM complexation model was used for Pb, Th, and U, and the fixed-charge site
complexation model was used for Am, Cs, and Sr. The use of the Gibbsite-DLM versus fixed-charge
site model was driven by the availability of data for elements of interest for the Clive DU PA Model.
For the Gibbsite-DLM runs, to highlight the upper and lower bounds of uncertainty in Kd
calculations, each element was run using both 1 g/L and 10 g/L of gibbsite as an input concentration
(Irunde et al. 2019). This concentration of adsorption sites is consistent with gibbsite concentrations
found in the literature for adsorption modeling in Visual MINTEQ. It might be low in terms of solid
to liquid ratios expected in soils, making it likely conservative. For the fixed-charge site models,
upper and lower uncertainty bounds were determined by running the model with input fixed-charge
site concentrations of 100 mmol/L, 200 mmol/L, and 500 mmol/L. The fixed-charge site
concentrations were selected based on the range of cation exchange capacity values reported at the
site of 10 to 20 meq/100 g (NRCS 2009), which approximately corresponds to 100 to 200 mmol/L.
An upper bound of 500 mmol/L was also included to represent possible sorption values for soils with
a high clay fraction or other general negative adsorption sites.
The modeled range of adsorption coefficients for lead and thorium were all within the ranges of the
literature estimates (Figure 2; Table 4). The modeled range of adsorption coefficients for uranium was
lower (5.78E-04 to 2.49E-01 mL/g) than the range estimated based on literature estimates (3.44E-01
to 6.63E+0l mL/g) across all soil fractions (sand, clay, silt fractions). VM had challenges with
convergence of results using concentrations of uranium higher than lE-5 M, indicating the lower
confidence in the range ofresults for uranium. The modeled ranges of adsorption coefficient values
were lower relative to the literature estimates for the elements modeled with the fixed charge model:
americium (Literature: 4.32E+0 1 to l. l 4E+03 mL/g; Modeled: 6.80E-03 to 4.30E-02 mL/g), cesium
(Literature: 2.70E+00 to 2.39E+02 mL/g, Modeled: l.25E-02 to 2.54E-02 mL/g), and strontium
(Literature: 2.70E+00 to 2.39E+02 mL/g, Modeled: 4.37E-02 to 1.22E-0I mL/g). The applicability of
the fixed-charge model to the Clive site is not clear and was used because it produced non-zero
adsorption results, rather than because there is clear applicability to the site.
Summary and Discussion
The results demonstrate both the feasibility and value of the modeling efforts undertaken, while
affirming that the literature-sourced estimates of solubility and adsorption provide an optimal basis
for robust and defensible geochemical parameter inputs for the Clive DU PA Model. The surface
complexation models rely on adsorption databases for those specific models, such as the diffuse-layer
model and the fixed-charge model. These databases for these surface complexation models and the
Page 12 of28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
choices made for surface complexation model are not specific to the Clive site .. The close range of
results of the Gibbsite-DLM and the literature sources give some confidence in the ranges of Kds.
The large ranges reported in modeled results illustrate the sensitivity of the model to soil type and
types of adsorption sites, which currently lack detailed site-specific inform ation. As such, the best
way to parameterize these values in the Clive DU PA Model is to use the literature-deri ved ranges
which also give a wide range of results and have a basis in laboratory experiments.
Table 4. Literature estimates of the minimum and maximum Kd coefficients, across all sand,
clay, and silt fractions, for each element. Modeled estimates of minimum and maximum Kd
coefficients for select elements, and the complexation model used to estimate that value, are
shown.
Literature Modeled Chemical Complexation
Element Min (ml/g) Max (ml/g) Min (ml/g) Max (ml/g) Model
Ac 1.57E+01 2.99E+03 ---
Am 4.32E+01 1.14E+03 6.80E-03 4.30E-02 Fixed-charge site
Cs 2.70E+00 2.39E+02 1.25E-02 2.54E-02 Fixed-charge site
I N(4.28e-1 , 6.05e-1) --
Np 3.92E-01 8.11 E+01 ---
Pa 8.32E+00 1.56E+03 ---
Pb 2.70E+00 2.39E+02 7.38E+00 4.39E+03 Gibbsite-DLM
Pu 6.69E+01 6.21 E+03 ---
Ra 3.87E-01 1.41 E+03 ---
Rn 0.00E+00 0.00E+00 ---
Sr 2.70E+00 2.39E+02 4.37E-02 1.22E-01 Fixed-charge site
Tc N(1.02e-1 , 1.45e-1) --
Th 1.92E+01 2.36E+03 9.06E+01 4.49E+03 Gibbsite-DLM
u 3.44E-01 6.63E+01 5.78E-04 2.49E-01 Gibbsite-DLM
Page 13 of28
102
~
Cl :::i E ~ 100
"Cl ::.:::
10·2
~
ENERGYSOLUTIONS
T
I
I I
I
I
I
I
I
.1
T
I
I
I T
I I
I ...I..
Am Cs Pb Sr
T
I
I
I
I
..J..
T
I
I
I
..1.
Th
Literature
Model
Mr. Doug Hansen
CD-2025-060
March 27, 2025
T
I
I
I
I
I
I
I
..J..
u
Figure 2. Comparison of Literature-Derived and Modeled Adsorption (Partitioning) Coefficients (Kd) for
Key Constituents at the Clive Site.
■ 0-40.a: (from DRC-2024-005438): NAC-0025 R4 indicates the chemical conditions (high ionic strength)
in the saturated zone at Clive are not considered favorable for colloid transport. The Division agrees with
the assessment that discussions and assertations regarding colloid facilitated radionuclide migration are
complex. However, the potential influence of colloids and the ionic strength of Clive groundwater needs
to be documented. While higher ionic strengths can facilitate aggregation and reduce colloid transport,
there are conditions where colloids can move depending on the colloid composition, pH, and presence of
counter ions which can reduce aggregations. A summary of previous colloid transport work is included,
but the studies are not explicitly compared to the conditions at the Clive site. Thus, more detail is
necessary to support the assumptions presented.
Please submit an evaluation of the role of colloid transport in terms related to the specific conditions at
the Clive site.
Additional information on two of the references cited in Section 2.0 ofNAC-0025 _ R42 (Neptune
2024) is included below. Evidence in these references and their experimental results strongly indicate
that the high ionic strength at the Clive site will not support colloid transport ofradionuclides. These
2 NAC-0025 _R6 (Neptune 2024) has been prepared in response to the follow-up RFls and is included with this response.
The quoted text remains unchanged in this revision. NAC-0025 _ RS was prepared in December 2023 to update uranium
solubility for reproducibility of geochemical modeling. The questions raised in these round 2 RFis, based on version
NAC-0025_R4, remain applicable.
Page 14 of 28
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
references and the comparison to conditions at the Clive site support the assumptions made in NAC-
0025 _ R4 related to the evaluation of colloidal transport.
From CRWMS (2000), cited in Section 2.0 ofNAC-0025_R4:
"A high fraction of colloids is present in low ionic strength solutions, and the concentration
of radionuclide-bearing colloids decreases significantly as the ionic strength is increased
(absence of colloids noted for ionic strengths above ~0.05 mol/kg)."
They note that high ionic strength solutions destabilize and promote aggregation and flocculation of
the colloids, decreasing their mobility.
The Clive site has average groundwater ionic strength of about 0.8 moVL, which is much greater than
the 0.05 mol/kg threshold, even given salt water density is slightly greater than 1 kg/L (noting
saturated salt solution has a density of 1.2 kg/L). Figure 3 relates plutonium colloids with relative
ionic strength, visually displaying this effect.
8E-8
• ♦202A_2K
■202A_20K
• 131A_2K
6E-8 e 131A_20K
..... ::e ....,, • 1 4E-8 •
• ~
2E-8 •• •
• OE +O +---...,.__--.----->-----,11------.--tr---.--..... ---,
0.000 0.020 0.040 0.060 0.080 0.100
Relative Ionic Strength (molft<g)
NOTE: OTN: LL000122051021 .116; DTN: LL000123351021 .117
Figure 3. Plutonium Concentration from Plutonium-Bearing Colloids as a Function of Ionic
Strength for Corrosion Tests on SRL 202A and SRL 131A at 2,000 and 20,000/m (at 90° C)
[Figure 8 of CRWMS (2000)].
From Degueldre et al. (2000), also cited in Section 2.0 ofNAC-0025_R4:
"This test clearly demonstrates that the salinity and hardness of the water play an important
complementary role on colloid stability. The attachment factor of these clay colloids
approaches unity for total salt concentrations of about 1 0e-2 Mand 1 0e-4 Min alkaline
earth elements."
Page 15 of 28
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
The attachment factor is "the ratio of colloids which attach to those which collide." See Figure 4,
which reproduces Figure 4 from Degueldre et al. (2000). These salt concentrations are much smaller
than the ionic strength at the Clive site of 0.8 M. For this aspect of their study, the pH of solution was
about 8. Their pH is slightly higher than that of the Clive site groundwater (pH 6.6-7.7) and similar
enough to that of the Clive site to apply their same conclusions. The conclusions of this paper also
support unfavorable conditions for colloid transport at Clive.
-----. _I
' . _,,. ... -~,. -~ ~... ---.. -... -...
' ' ' ' ' --r .. , ~ .. .. ---... --: ...
0 I
' I
. --.. ___ _ 3.5
-3
2.5
0 Ca and Na e Mg and K
!-'lg_ 4. i he clfcct o 'a, Mg and l\.a. K al pH 8 on the ·olloid atla hmcnt factor. Condttmns: consid=d colloids: mon1monlloni1c
s11.r > 100 nm_ the arrows denote th.c dcc:rc,ase of stab1hly. "ith. pOC < 4.3 and pH .0. prc-c1smn 10-20''•·
Figure 4. Effects of salinity (Na+K) and hardness (Ca+Mg) on colloid attachment, indicating
high concentrations result in an attachment factor of 1 (log(0)=l) [Figure 4 ofDegueldre et al.
(2000)).
• 0-40.b: (from DRC-2024-005438): The following statement is found on Page 9 of NAC-0025 R4: "Using
the data from the average of all wells shown in Table 7, the stoichiometric ionic strength is calculated at
0. 73 M (mol/L). "It appears there are some errors in calculating the average values. The ionic strength
of 0. 73 M matches that for GW-19A with the highest Na+ and Cl-concentrations but is not the
stoichiometric ionic strength calculated for all wells shown in Table 7.
Please verify the stoichiometric ionic strength calculated for all wells shown in Table 7.
The stoichiometric ionic strength average was recalculated. The average is corrected as 0.82 mol/L in
Section 2.2 ofNAC-0025_R6 (Neptune 2024). Additionally, the text ofNAC-0025_R4 references
Table 7; however, the correct table is Table 8, which shows the average mol/L for each ion. The
average mol/L is determined by the average of all ions. The last sentence of the first paragraph on
page 9 will now read as follows, with corrections shown in redline/strikeout:
Page 16 of 28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Using the data from the average of all wells shown in Table 8, the stoichiometric ionic
strength is calculated at 0.82 M (mol/L) using the equation:
where
I= ~x L Ciz 2
i
C is the average mol/L of anion or cation,
z is the charge of the ion, and
i is the number of ions included in the total ionic strength.
■ 0-40.c: (from DRC-2024-005438): In section 5.1.9 of NAC-0025 R4, the solubility of radium is reported
to be controlled by RaSO4(s) using surrogate values for barite (BaSO4(s)) and as a co-precipitate
(Ra,Ba(SO4)(s)). The assumption that radium can form a solubility limiting phase may not be justified or
necessary if the concentrations of counter ions in the phase are insufficiently high to cause saturation.
Please provide relevant barium and sulfate concentrations at the site to indicate which phases can be
formed.
The sentence represented a hypothetical situation that may not reflect actual conditions at the Clive
site. Therefore, the following sentence has been removed from Section 5.1.9 ofNAC-0025_R6
(Neptune 2024).
■ 0-40.d: (from DRC-2024-005545): Please provide an evaluation of the minerology of hydrostratigraphic
units 1 through 3. Please use the same methodology used for unit 4 in Table 6. Additionally, please
clarify that the bedrock was not included in the simulations and the rationale for excluding it.
The information listed in Table 6 ofNAC-0025_R4 is taken from the Revised Hydrogeologic Report
for the Envirocare Waste Disposal Facility, Clive, UT, August 2004 (Envirocare 2004), Sections 5.2.1
and 6.0, except for the addition of the mineralogy. In Table 6 ofNAC-0025_R6 (Neptune 2024), Unit
4 references to "quartz, feldspars, clay minerals (kaolinite, smectite, and illite/mica) trace gypsum"
have been removed as these minerals are not relevant to the derivation of solubility and sorption
parameters. The mineralogy is not relevant to the model and so is removed.
In NAC-0025 _ R6, Table 6 has been revised to only include the information as presented in
Envirocare (2004), Section 5.2.1. Redline/strikeout of the revisions is provided below.
Page 17 of28
~
ENERGYSOLUTIONS
Table 6. Soil Type within the Four Hydrostratigraphic Units.
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Unit Number Soil Type Unit Description
4
3
2
Silt and Clay
Silty sand, interbedded with silty and
clay layers.
Clay with occasional silty sand
interbeds.
Silty sand interbedded with clay and
silt layers.
From 6 to16.5 ft thick with an average thickness
of 10 ft. Unsaturated.
7 to 25 ft thick with an average thickness of 15 ft.
Largely unsaturated, with lower portion saturated
in western part of site. The unconfined water-
bearing zone in Unit 3 and the upper part of Unit
2 has been designated as the shallow aquifer.
2.5 to 25 ft thick with an average thickness of 15
ft. Unit 2 is saturated below the Clive facility.
Begins at a depth of approximately 45 ft bgs.
The thickness of Unit 1 is unknown. Locally
confined aquifer, designated as the deep aquifer.
In Section 2.0, paragraph 6, the migration of radionuclides is not expected to reach the deep aquifer
due to a natural upward gradient at the facility. For this reason, Unit 1 is not included in the PA
model. Unit 2, Unit 3, and Unit 4 do not encounter bedrock. The following sentence has been added
to paragraph 6 ofNAC-0025_R6 (Neptune 2024).
"Since migration is not expected to reach the deep aquifer, Unit 1 will not be included in the
PA model and therefore, bedrock will also not be included. "
■ 0-40.e: (from DRC-2024-005545): Please provide a comparison of the partitioning coefficients from the
references used in NAC-0025 R4 to the site-specific conditions found in hydrostratigraphic units 1
through 4.
Neptune agrees that incorporating research data developed from sites unrelated to the Clive site
introduces uncertainty in the inputs. This uncertainty is incorporated into the Kd values as input
distributions that are wider than the literature data minimum and maximum, which were assigned to
be the 5th and 95th percentiles, respectively, as described in Section 3.0 ofNAC-0025 _R4. Sensitivity
analysis helps identify which inputs affect the results most significantly and can lead to decisions to
refine input distributions if those inputs affect results to an unacceptably large degree.
In Section 3.0 ofNAC-0025_R43, Neptune defines the process to select literature data for Kd input
distributions which best fit the conditions found at the Clive site as follows :
" ... general literature were screened to retain studies relevant to the Clive site, any value
within the range of those studies was de emed to be "equally viable," given the uncertainty
associated with the various soil and water characteristics for the site."
Note that the literature reviewed includes Kd data collected for the site, such as Bingham
Environmental (1995) and Adrian Brown Consultants (I 997). Because pH and mineral texture are
3 NAC-0025_R6 (Neptune 2024) has been prepared in response to the follow-up RFis and is included with this response.
The quoted text remains unchanged in this revision. NAC-0025_R5 was prepared in December 2023 to update uranium
solubility for reproducibility of geochemical modeling. The questions raised in these round 2 RFis, based on version
NAC-0025_R4, remain applicable.
Page 18 of 28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
generally important factors that affect adsorption, these parameters were considered carefully when
they were provided in the literature studies. Some literature studies, like Sheppard and Thibault
(I 990) and, to some extent, the EPA series (EPA 1999a, 1999b, 2004 ), did not include pH values for
all of their Kd data but these studies were included in the input distributions. The use of the range of
those Kds potentially makes the Kd input distributions wider than they would otherwise be if pH
information were stated and data were culled based on pH. Most references clarified mineralogy or
mineral classes, such as sand, silt, and clay, as presented in Sheppard and Thibault (1990), for
example. Other references, such as Seme (2007), were considered that have pH values and organic
carbon contents similar to the Clive site, as noted in Section 4.0.
In Section 4.0 of NAC-0025 _ R6 (Neptune 2024), the last paragraph has been edited to provide clarity
in the statements comparing conditions in literature data with the Clive site conditions (quoted here in
redline/strikeout to make the revisions readily apparent):
"Of note is that the Hanford soils are slightly acidic (pH 6.2 to 7.8), with organic content of
0.5 to 1.5% organic carbon, which is slightly higher than that of the Clive location, which
has organic carbon contents of approximately 0.3% to 1%. Serne (2007) also reviews a
number of studies that are equally viable to the Clive facility and the range of K d values
provided are useful as a first comparison."
Table 5 summarizes characteristics of the experimental samples and site conditions in the various
references for Kds, along with the characteristics of the Clive site. Where it was simple to extract pH
values and other information from references that contain great amounts of information, those values
are given in the table. Where pH ranges were large, those are noted in the table. Blank cells indicate
that insufficient information was present to provide data in the table cells.
Table 5. Geochemical conditions for Kd references as compared to the Clive Site.
Comparison of References for Kds to Clive Site Conditions
Reference pH Soil matrix Eh Bicarbonate
Unit 2 (clay with
silty sand layers)
6.62 to Unit 3 (silty sand -43.70 to Clive Site 7.69 with layers of clay 500 102 to 350
and silt)
Unit 4 (silt and
clay)
Whetstone Associates (2009) Soil density
1.566 g/cm3
Bingham Environmental (1995) 7.52 Silty sand (Unit 3) 160 243
Bingham Environmental (1996) 7.52 Silty Sand (Unit 3) 103 192
Scism (2006) 8.2 to 8.7 Alluvium (particle 189 to 3.0E-03 to
size 75-2000 um) 212 4.7E-03
Sheppard and Thibault (1990) Sand , silt, clay
Triay et al. (1997), Fortymile 6.9 Tuff units 340 143 Wash
Page 19 of 28
~
E ERGYSOLUTIONS
Comparison of References for Kds to Clive Site Conditions
Reference pH Soil matrix
Triay et al. (1997), Yucca Mt 6.5 to 7.5 Paintbrush Tuff
Triay et al. (1997), Midway Paleocarbonates between Yucca Mt and 6.7 under Tuff Fortymile Wash
Decker and Papelis (2003) Variable Various
6.37 to Silty clay (Unit 2)
Envirocare (2000) 7.87 and silty sand (Unit
3)
EPA (1999a) Wide Various Range
EPA (1999b) Wide Various Range
EPA (2004) Wide Various Range
Serne (2007) 6.2 to 7.8 Various
Last et al. (2004) 7.66 to Various 8.17
Krupka et al. (2004) Used data Various, especially
~8 sand and gravel
Adrian Brown Consultants
(1997) 7.5 to 7.6 Silty sand (Unit 3)
Bingham Environmental (1991) 6.3 to 8.5 Silty clay/sand
(Units 2-4)
Glover et al. (1976) 7.7 to 8.2 Silty sand to sand
7.29 to 70-97% sand -Um et al. (2009) 7.92 3-30% silt
Blank -Information not readily available in the literature.
Eh
-143 to
600
360
220 to
400
160 (test
sample
based
upon
average
of 4 well
samples)
45-430
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Bicarbonate
"difficult to
quantify"
698
Organic C 0.5
to 1.5% (Clive
0.3to 1%)
197 to 322.5
4 to 11
(groundwater)
1.3 to 3% by
weight as
CaCO3
Page 20 of 28
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Regarding using literature to inform Kd input distributions rather than simulation of adsorption
coefficients using site-specific conditions, simulation of adsorption using geochemical computer
codes, such as Phreeqc (Parkhurst and Appelo 2013), requires input values that are not easily obtained
for a site. Adsorbent properties of site minerals, such as number and type of adsorption sites and
surface area of minerals, are not known data for various materials at the Clive site and this could
create great uncertainty in adsorption modeling results -greater uncertainty than when using
literature references. Assumptions can be made in order to run adsorption models, and these types of
models can be helpful for understanding changes in water chemistry for a fixed material adsorption
capacity, for example. The use of literature data to inform Kd distributions allows for the acquisition
of data and distributions that span the likely Kd for the site. Data that are from experiments in
conditions similar to the site were reviewed and minimum and maximum values for each element in
that collection of relevant sites were expanded slightly in the development of the log-uniform input
distributions. This approach also includes uncertainty and variability for each element and has a basis
in experimental results rather than pure modeling results.
Note to reviewers: RFls relating to Appendix AB will be addressed under separate cover together
with responses addressing DU inventory, so that updates to the inventory are carried through to the
RESRAD-OFFSITE model. RFls relating to NAC-0025_R4 are addressed below.
■ 0-40.f: (from DRC-2024-005704): NAC-0025 R4, Page 16, indicates [little] information was identified
for the determination of sorption values for Protactinium. The Division concurs with sorption
va lues/data for protactinium being understandably sparse but would like EnergySolutions to review
reference "Sakamoto, Y., et al., Sorption characteristics of Actinium and Protactinium onto soils;
Journal of Nuclear Science and Technology, 2002, 39(sup 3): P. 481-484." and provide additional
discussion on this matter.
Although protactinium and actinium are considered on the inventory species list and have Kd values
developed in NAC-0025 _R44, their inventory is zero based on analytical data (Neptune (2023), Table
2).
Sakamoto et al. (2002) was reviewed to better understand their characterization of sorption of
protactinium in five soil types (loam, yellowish soil, sand-A, sand-B, and tuff) in Japanese soil. This
study was designed to describe sorption behavior under influences of anion ligands, such as
carbonates, in groundwater. Protactinium adsorption was measured using a batch method and then the
sorbed forms were studied using sequential extraction technique of the soils. The Kd values were
measured as a function of carbonate concentration to study the influence of carbonate on sorption
behavior. Actinium was also studied. Protactinium Kd values "varied widely among different kinds of
soils, because the sorption of the 233Pa was mainly based on irreversible sorption on amorphous Fe
and Mn oxides surfaces." The range ofKd values for 233 Pa was found to be from 0.7 m3/kg (sand-B)
to 52 m3/kg (loam) with a pH range of 6.2 to 8.9. The higher Kd values correspond to higher
concentrations of carbonates. These Kd values are much lower than the distribution of Pa adsorption
values presented in NAC-0025 _ R4 . The difference between the Sakamoto et al. (2002) results and Kd
4 NAC-0025_R6 (Neptune 2024) has been prepared in response to the follow-up RFis and is included with this
response. The quoted text remains unchanged in this revision. NAC-0025_R5 was prepared in December 2023 to
update uranium solubility for reproducibility of geochemical modeling. The questions raised in these round 2 RFis,
based on version NAC-0025_R4, remain applicable.
Page 21 of 28
~
E ERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
values cited in many references in NAC-0025 _R4 make the Sakamoto et al. (2002) results
questionable for their applicability to Clive. Sakamoto et al. (2002) used very small amounts of soil
(0.8 g) and 5.6E-14 mol/dm 3 of Pa, which could potentially affect results that are near zero
adsorption. As well, data on amorphous Fe and Mn oxides surfaces at the Clive site is unknown.
Therefore, the data from Sakamoto, et al., will not be included in the modeling.
• 0-40.g: (from DRC-2024-005704): The following statement is found in section 4.1.2 (Americium), on
page 13 of NAC-0025 R4: "This rare earth element will have a large sorption coefficient." Americium is
an actinide element, not a rare earth element. Please refer to americium as an actinide element in section
4.1.2 and along with any other placement of the reference.
The sentence has been corrected as follows (in redline/strikeout):
"This actinide element will have a large sorption coefficient."
A word search ofNAC-0025_R6 (Neptune 2024) did not locate any other mention of americium as a
rare earth element.
• 0-40.h: (from DRC-2024-005704): In section 5.1.9 of NAC-0025 R4, the solubility of radium is reported
to be controlled by RaSO4(s) using surrogate values for barite (BaSO4(s)) and as a co-precipitate (Ra,
Ba(SO4)(s) ). The assumption that radium can form a solubility limiting phase may not be justified or
necessary if the concentrations of counter ions in phase are insufficiently high to cause saturation.
Please provide relevant barium and sulfate concentrations at the site to indicate which phase can be
formed.
The sentence represented a hypothetical situation that may not reflect actual conditions at the Clive
site. Therefore, the following sentence was removed from Section 5.1.9.
• 0-40.i: (from DRC-2024-005704): In section 5.1.11 of NAC-0025 R4, the solubility of strontium is
reported to be controlled by comparable phases to calcium minerals including celesite (Sr, SO4(s)) and
strontianite (SrCO3(s) ). Please indicate whether sulfate and carbonate concentrations are sufficiently
high to approach saturation of these minerals.
Applying average values of geochemical parameters from Table 8 ofNAC-0025, with an average pH
of 7.2 and middle value of bicarbonate (226 mg/L) from Table 7 ofNAC-0025, analysis using Visual
Minteq indicates that there is sufficient carbonate in the system for precipitation of strontianite (see
Table 6 below). In equilibrium with strontianite, only a small percentage of the carbonate, 0.3%, is
complexed with strontium. That amount of carbonate converted to a concentration is smaller than I%
of total bicarbonate shown in Table 7 ofNAC-0025. Bicarbonate is 76% of total carbonate and other
solution species make up the rest of total carbonate.
Page 22 of 28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Table 6. Visual Minteq Species Distribution Output for Carbonates and Strontium in Equilibrium with
Strontianite.
Species
Com onent % of total concentration name
CO3-2 0.233 CO3-2
NaHCO3
10.328 (aq)
0.015 Mg2CO3+2
76.074 HCO3-
H2CO3*
6.468 (aq)
MgCO3
0.127 (aq)
3.277 MgHCO3+
2.542 CaHCO3+
CaCO3
0.158 (aq)
0.327 SrHCO3+
0.444 NaCO3-
Sr+2 79.064 Sr+2
14.259 SrCI+
5.731 SrSO4 (aq)
0.018 SrCO3 (aq)
0.922 SrHCO3+
Applying average values of geochemical parameters from Table 8 ofNAC-0025, with an average pH
of7.2 from Table 7 ofNAC-0025, analysis using Visual Minteq shows that there is sufficient sulfate
in the system for precipitation of celestite (see Table 7 below). In equilibrium with celestite, only a
small percentage of the sulfur in solution, 0.2%, is complexed with strontium. Sulfate (SO4-2) is 58%
of total solution sulfur and other solution species make up the rest of total sulfate.
Page 23 of28
~
E ERGYSOLUTIONS
Table 7. Visual Minteq Species Distribution Output for Sulfates and
Strontium in Equilibrium with Celestite
Species
Com onent % of total concentration name
SO4-2 57.964 SO4-2
MgSO4
6.32 (aq)
CaSO4
4.975 (aq)
0.179 SrSO4 (aq)
29.816 NaSO4-
0.746 KSO4-
Sr+2 79.158 Sr+2
14.289 SrCI+
5.937 SrSO4 (aq)
SrCO3
0.011 (aq)
0.598 SrHCO3+
Mr. Doug Hansen
CD-2025-060
March 27, 2025
• 0-40.j: (from DRC-2024-005704): In section 5.1. 7 of NAC-0025 R4, the reported dominant aqueous
Plutonium species is tetravalent hydroxycarbonate complex Pu(OH)2(CO3)2-3, but section 4.1. 7 of NAC-
0025 R4 indicates that Pu(V) and Pu(VI) are the most likely states. Assumed aqueous species for sorption
partitioning and solubility control should be consistent. Please clarify the discrepancy between the
Plutonium species reported in section 5.1. 7 and section 4.1. 7 of NAC-0025 R4.
Section 5.1 .7 refers to the solubility of plutonium with a reported dominant species of
Pu(OH)2(CO3)22-in groundwater, for which Pu is in the +4 oxidation state. Section 4.1.7 refers to
the sorption of plutonium with reported dominant redox states of Pu(V) and Pu(VI). Section 4.1 .7 was
corrected to include consideration of Pu(IV) as a dominant redox state of Pu. Pu(IV) had already been
considered for adsorption distribution development, so no revisions to these distributions are needed.
For example, note that the Glover et al. (1976) reference cited in the white paper studied Pu(IV)
sorption. That section now reads (redline/strikeout shown to clarify the edits made):
References
The most likely states are as Pu(IV), Pu(V), and Pu(VI), both as cations and complexed with
hydroxide and carbonate. Pu(IV) may be present in the slightly reducing conditions of the
saturated zone and localized areas of the unsaturated zone due to surface-mediated reduction
of Pu(V).
Adrian Brown Consultants, 1997. Response to UDEQ Kd Interrogatories, Report 310 IB.970422,
prepared for Envirocare of Utah, Adrian Brown Consultants Inc., Denver CO, April 1997
Ayawei, N., et al., 2017. Modelling and Interpretation of Adsorption Isotherms, Journal of Chemistry
(Article ID 3039817) 1-11 doi: 10.l 155/2017/3039817
Bingham Environmental, 1991. Hydrogeologic Report, Envirocare Waste Disposal Facility, South
Clive, Utah, prepared for Envirocare of Utah, Bingham Environmental Inc., Salt Lake City UT,
October 1991
Page 24 of28
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Bingham Environmental, I 995. Project Memorandum. Summary of Results, Radionuclide Kd Tests,
Envirocare Disposal Landfills, Clive, Utah, Bingham Environmental Inc., Salt Lake City UT, August
3, 1995
Bingham Environmental, 1996. Project Memorandum. Summary of Results, Radionuclide Kd Tests,
Envirocare Disposal Landfills, Clive, Utah, Bingham Environmental Inc., Salt Lake City, UT,
January 25, 1996
Brimhall, W.H., and L.B. Merritt, 1981. Geology of Utah Lake-Implications for Resource
Management, Great Basin Naturalist Memoirs 5 (Article 3) 24--42
CRWMS, 2000. Colloid-Associated Radionuclide Concentration Limits, ANL-EBS-MD-000020
REV 00 ICN 01, Civilian Radioactive Waste Management System, 2000
Decker, D., and C. Papelis, 2003. Evaluation of Surface Complexation Models for Radionuclide
Transport at the Nevada Test Site: Data Availability and Parameter Evaluation, DOE/NV/13609-14,
Publication No. 45185, United States Department of Energy, National Nuclear Security
Administration, Nevada Site Office, Las Vegas NV, May 2003
Degueldre, C., et al., 2000. Groundwater Colloid Properties: A Global Approach, Applied
Geochemistry 15 (2000) 1043-1051
Einsele, G., and M. Hinderer, 1997. Terrestrial Sediment Yield and the Lifetimes of Reservoirs,
Lakes, and Larger Basins, Geo) Rundsch 86 (1997) 288-310
Envirocare, 2000. Submittal of Report titled "Metals Distribution Coefficient Values (Kd) Relevant to
the Envirocare Site." Memorandum and Report to the U.S. Nuclear Regulatory Commission,
Envirocare of Utah Inc., Salt Lake City UT, September 2000
Envirocare, 2004. Revised Hydrogeologic Report for the Envirocare Waste Disposal Facility, Clive,
Utah, Version 2.0, Envirocare of Utah Inc., August 2004
EPA, I 999a. Understanding Variation in Partition Coefficient, Kd, Values, Volume II: Review of
Geochemistry and Available Kd Values for Cadmium, Cesium, Chromium, Lead, Plutonium , Radon,
Strontium, Thorium, Tritium (3H), and Uranium, EPA 402-R-99-004B, United States Environmental
Protection Agency, Washington DC, August I 999
EPA, 1999b. Understanding Variation in Partition Coefficient, Kd, Values, Volume I: The Kd Model,
Methods of Measurement, and Application of Chemical Reaction Codes, EPA 402-R-99-004A,
United States Environmental Protection Agency, Washington DC, August 1999
EPA, 2004. Understanding Variation in Partition Coefficient, Kd, Values, Volume III: Review of
Geochemistry and Available Kd Values for Americium, Arsenic, Curium, Iodine, Neptunium,
Radium, and Technetium, EPA 402-R-04-002C, United States Environmental Protection Agency,
Washington DC, July 2004
Glover, P.A., et al., 1976. Plutonium and Americium Behavior in the Soil/Water Environment: I.
Sorption of Plutonium and Americium by Soils. In Proceedings of Actinide-Sediment Reactions
Page 25 of 28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Working Meeting at Seattle, Washington on February 10-11 , 1976, BNWL-2117, edited by L.L.
Ames, pp. 225-254, Battelle Pacific Northwest Laboratories, Richland WA
Inoue, K., and T. Naruse, 1987. Physical, Chemical, and Mineralogical Characteristics of Modern
Eolian Dust in Japan and Rate of Dust Deposition, Soil Science and Plant Nutrition 33 (3) 327-345
doi: 10.1080/00380768.1987 .10557579
lrunde, R., et al., 2019. Visual MlNTEQ Simulation for Prediction of the Adsorption of Arsenic on
Ferrihydrite. In Environmental Arsenic in a Changing World, edited by Y. Zhu, et al., pp. 435--436,
CRC Press, London
Karamalidis, A.K., and D.A. Dzombak, 2011. Surface Complexation Modeling: Gibbsite, John Wiley
& Sons, New York NY
Khalidy, R., and R.M. Santos, 2021. Assessment of Geochemical Modeling Applications and
Research Hot Spots-A Year in Review, Environ Geochem Health 43 (2021) 3351-3374 doi:
I 0.1007 /s I 0653-021-00862-w
Krupka, K.M., et al., 2004. Geochemical Data Package for the 2005 Hanford Integrated Disposal
Facility Performance Assessment, PNNL-13037 Rev. 2, prepared for United States Department of
Energy, Pacific Northwest National Laboratory, Richland WA, September 2004
Last, G.V., et al., 2004. Vadose Zone Hydrogeology Data Package for the 2004 Composite Analysis,
PNNL-14702, Rev. 0, Pacific Northwest National Laboratory, Richland WA, July 2004
Neff, J.C., et al., 2008. Increasing Eolian Dust Deposition in the Western United States Linked to
Human Activity, Nature Geoscience I (2008) 189-195 doi: I0.1038/ngeol33
Neptune, 2023. Radioactive Waste Inventory for the Clive DU PA, Clive DU PA Model v3.0, NAC-
0023 _R6, Neptune and Company Inc., Los Alamos NM, December 2023
Neptune, 2024. Geochemical Modeling for the Clive DU PA, Clive DU PA Model v3.0, NAC-
0025_R6, Neptune and Company Inc., Los Alamos NM, November 2024
Nitsche, H., et al., 1993. Measured Solubilities and Speciations of Neptunium, Plutonium, and
Americium in a Typical Groundwater (J-13) from the Yucca Mountain Region Milestone Report
3010-WBS 1.2.3.4.1.3.1, LA-12562-MS, UC-802, Los Alamos National Laboratory, Los Alamos
NM, July 1993
NRCS, 2009. Web Soil Survey (WSS), Tooele Area Soil Survey Utah, Version 5, Natural Resources
Conservation Service, United States Department of Agriculture (USDA), accessed September 2,
2009, available from https://websoilsurvey.nrcs.usda.gov/app/HomePage.htm
Oviatt, C.G., 2024. Late Neogene and Quaternary Lacustrine History of the Great Salt Lake-
Bonneville Basin, Geosites 51 (2024) 1-16 doi: 10.3 l 711 /ugap.v5 I i.133
Oviatt, C.G., et al., 1999. Reinterpretation of the Burmester Core, Bonneville Basin, Utah, Quaternary
Research 52 (2) 180-184
Page 26 of28
~
ENERGYSOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
Parkhurst, D.L., and C.A.J. Appelo, 2013. Description oflnput and Examples for PHREEQC Version
3-A Computer Program for Speciation, Batch-Reaction, One-Dimensional Transport, and Inverse
Geochemical Calculations, Techniques and Methods 6-A43, U.S. Geological Survey (USGS), Denver
co
Rea, D.K., and M. Leinen, 1988. Asian Aridity and the Zonal Westerlies: Late Pleistocene and
Holocene Record of Eolian Deposition in the Northwest Pacific Ocean, Palaeogeography,
Palaeoclimatology, Palaeoecology 66 (1988) 1-8
Reheis, M.C., et al., 1995. Quaternary Soils and Dust Deposition in Southern Nevada and California,
GSA Bulletin 107 (9) 1003-1022
Sakamoto, Y., et al., 2002. Sorption Characteristics of Actinium and Protactinium onto Soils, Journal
ofNuclear Science and Technology 39 (Supplement 3) 481 -484
Schumer, R., and D.J. Jerolmack, 2009. Real and Apparent Changes in Sediment Deposition Rates
through Time, Journal of Geophysical Research 114 (F00A06) 1-12 doi: 10.1029/2009JF00l266
Scism, C.D., 2006. The Sorption/Desorption Behavior of Uranium in Transport Studies Using Yucca
Mountain Alluvium, LA-14271-T, Thesis, Los Alamos National Laboratory, Los Alamos NM,
February 2006
Serne, R.J., 2007. Kd Values for Agricultural and Surface Soils for Use in Hanford Site Farm,
Residential, and River Shoreline Scenarios, Technical Report for Groundwater Protection Project-
Characterization of Systems Task, PNNL-16531, prepared for Fluor Hanford Inc. and U.S.
Department of Energy, Pacific Northwest National Laboratory, Richland WA, August 2007
Sheppard, M.I., and D.H. Thibault, 1990. Default Soil Solid/Liquid Partition Coefficients, KdS, for
Four Major Soil Types: A Compendium, Health Phys 59 (4) 471-482
Triay, LR., et al., 1997. Summary and Synthesis Report on Radionuclide Retardation for the Yucca
Mountain Site Characterization Project, Yucca Mountain Site Characterization Program Milestone
3784M, LA-13262-MS, UC-903 and UC-940, Los Alamos National Laboratory, Los Alamos NM,
October 1997
Um, W., et al., 2009. The Effect of Gravel Size Fraction on the Distribution Coefficients of Selected
Radionuclides, Journal of Contaminant Hydrology 107 (1 -2) 82-90
Whetstone Associates, 2009. Technical Memorandum to EnergySolutions from Whetstone
Associates, Subject: Uranium Fate & Transport Modeling, 10,000 years, for EnergySolutions Class A
Cells, Whetstone Associates Inc., Gunnison CO, October 30, 2009
Wikle, C.K., and L.M. Berliner, 2005. Combining Information Across Spatial Scales, Technometrics
47 (1) 80-91
Page 27 of 28
~
ENERGY SOLUTIONS
Mr. Doug Hansen
CD-2025-060
March 27, 2025
If you have further questions regarding these responses to the director's requests of DRC-2024-
005227, DRC-2024-005438, DRC-2024-005545, and DRC-2024-005704, please contact me at (801)
649-2000.
Sincerely,
Vern C.
Roger s
Vern C. Rogers
Digitally signed by Vern C. Rogers
ON: cn=Vern C. Rogers,
o=EnergySolutions, ou=Waste
Management Division,
email:::[e-mail],
c=US
Date: 2025.03.27 12:20:34-06'00'
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 28 of 28