Saturated Zone Colloid-Facilitated Transport REV 00 ICN 01

ANL-NBS-HS-000031

November 2001


1. PURPOSE 
The purpose of the Saturated Zone Colloid-Facilitated Transport Analysis and Modeling Report 
(AMR), as outlined in its Work Direction and Planning Document (CRWMS M&O 1999a), is to 
provide retardation factors for colloids with irreversibly-attached radionuclides, such as 
plutonium, in the saturated zone (SZ) between their point of entrance from the unsaturated zone 
(UZ) and downgradient compliance points. Although it is not exclusive to any particular 
radionuclide release scenario, this AMR especially addresses those scenarios pertaining to 
evidence from waste degradation experiments, which indicate that plutonium and perhaps other 
radionuclides may be irreversibly attached to colloids. 
This report establishes the requirements and elements of the design of a methodology for 
calculating colloid transport in the saturated zone at Yucca Mountain. In previous Total Systems 
Performance Assessment (TSPA) analyses, radionuclide-bearing colloids were assumed to be 
unretarded in their migration. Field experiments in fractured tuff at Yucca Mountain and in 
porous media at other sites indicate that colloids may, in fact, experience retardation relative to 
the mean pore-water velocity, suggesting that contaminants associated with colloids should also 
experience some retardation. Therefore, this analysis incorporates field data where available and 
a theoretical framework when site-specific data are not available for estimating plausible ranges 
of retardation factors in both saturated fractured tuff and saturated alluvium. The distribution of 
retardation factors for tuff and alluvium are developed in a form consistent with the Performance 
Assessment (PA) analysis framework for simulating radionuclide transport in the saturated zone. 
To improve on the work performed so far for the saturated-zone flow and transport modeling, 
concerted effort has been made in quantifying colloid retardation factors in both fractured tuff 
and alluvium. The fractured tuff analysis used recent data and interpretation from the C-wells 
reactive tracer testing complex in the saturated zone of Yucca Mountain. As no data regarding 
colloid transport have been developed by the Yucca Mountain Site Characterization Project 
(YMP) for the alluvial system, a theoretical analysis based on studies performed in other alluvial 
systems is developed. The parameters derived in this AMR are developed in a manner consistent 
with the PA methodology and can be readily integrated into that analysis. 
The work activities documented in REV 00 of this AMR are governed by the Work Direction 
and Planning Document (CRWMS M&O 1999a) for abstraction of colloid facilitated plutonium 
transport. The purpose and scope of the activity is to abstract colloid-facilitated transport 
parameters for use in TSPA analyses. While the general scope of this activity has remained the 
same, specific tasks have been modified to better address the evolving needs of TSPA. The 
codes RTA V1.1 (Software Tracking Number (STN): 10032-1.1-00) and GoldSim V6.03 (STN: 
10296-6.03-00) were used in this AMR rather than FEHM, which was originally planned 
(CRWMS M&O 1999a). RTA was developed with a specific goal of interpreting colloid 
transport parameters in fractured tuff at Yucca Mountain. GoldSim, which is a standard code 
used in the Yucca Mountain PA analyses, was used for analysis of retardation factors in alluvium 
because large ranges in uncertainty had to be addressed due to the lack of site-specific data for 
this process. This approach replaced the 1-D simulations that were originally planned (CRWMS 
M&O 1999a). 
November 2001 7 ANL-NBS-HS-000031 REV 00 ICN 01
This ICN of the AMR was prepared under Technical Work Plan for Saturated Zone Flow and 
Transport Modeling and Testing (BSC 2001a) to reflect updates in the qualification status of 
input data, and to address TBVs associated with REV 00 of the AMR. The scope of the TBV 
resolution actions which necessitated ICN 01 of this AMR is described in Technical Work Plan 
for: Integrated Management of Technical Product Input Department (BSC 2001b, Addendum B, 
Section 4.1). 
November 2001 8 ANL-NBS-HS-000031 REV 00 ICN 01
2. QUALITY ASSURANCE 
The activities documented in REV 00, ICN 00, of this Analysis/Model Report (AMR) were 
evaluated in accordance with QAP-2-0, Conduct of Activities, and were determined to be subject 
to the requirements of the U.S. DOE Office of Civilian Radioactive Waste Management 
(OCRWM) Quality Assurance Requirements and Description (QARD) (DOE 2000). This 
evaluation is entitled Conduct of Performance Assessment (CRWMS M&O 1999b). 
The Activity Evaluation for ICN 01 is contained in the technical work plan (BSC 2001a). This 
AMR has been prepared in accordance with procedures listed in the technical work plan. 
November 2001 9 ANL-NBS-HS-000031 REV 00 ICN 01
3. COMPUTER SOFTWARE AND MODEL USAGE 
The computer software codes used in this AMR are as follows. 
1. Software: Reactive Transport Application (RTA), V1.1, (STN: 10032-1.1-00) 
Used for: Analysis of colloid retardation in the C-wells tracer test 
The software was obtained from Configuration Management, is appropriate for the 
application, and was used only within the range of validation in accordance with APSI.
1Q. 
2. Software: GoldSim, V6.03 (STN: 10344-6.03-00) 
Used For: Theoretical model of colloid retardation factors in alluvial material 
The software was obtained from Configuration Management, is appropriate for the 
application and was used only within a range for which it has been validated. 
3. Software: Microsoft Excel, 97 SR-1 
Used for: Plotting graphs 
Only built-in standard functions in this commercial software were used. No software 
routines or macros were used to prepare this report. No numerical models were used in 
report preparation. 
ANL-NBS-HS-000031 REV 00 ICN 01 November 2001 10
4.1 DATA AND PARAMETERS 
Data Description 
Bullfrog Tuff tracer test data 
Prow Pass reactive tracer test 
field data 
Interpretations of Bullfrog Tuff 
tracer test data 
Bullfrog test production 
rate/recirculation rate 
Prow Pass test production 
rate/recirculation rate 
Bullfrog solute mean resisdence 
time 
Prow Pass solute mean 
residence time 
Table 1. Input Data Sources 
ANL-NBS-HS-000031 REV 00 ICN 01 
4. INPUTS 
Data Sources 
Fractured Tuff Analysis 
Table 2, Assumption 21 
DTN: 
LAPR831231AQ99.001 
DTN: 
LA9909PR831231.003 
DTN: 
MO0111BFROGPRO.000 
MO0111BFROGPRO.001 
MO0111BFROGPRO.002 
MO0111BFROGPRO.003 
MO0110BFROGREC.001 
DTN: 
GS010799992315.001 
MO0111PROWPASS.000 
DTN: 
LA9909PR831231.003 
Table 2, Assumption 20 
11 
Location in this Document 
Fig. 1 
Fig. 2 
Fig. 1, Table 4, Table 6, Fig. 
3, Fig. 4, Fig. 5, Table 2 
(Assumption 21) 
Table 3 
Table 3 
Table 3 
Table 3, Fig. 3, Fig. 4, Fig. 5 
November 2001
Data Description 
Specific discharge distribution 
and porosity distribution in tuff 
Table 1. Input Data Sources (Continued) 
Data Sources Location in this Document 
Fig. 6 DTN: 
SN0004T0571599.004 
Location in this Document Data Description 
Porosity distribution in alluvium DTN: 
Specific discharge (Flux) (q) 
(m/yr) in alluvium 
Alluvial bulk density is 1.27 
g/cm3 
Colloid size distribution 
4.2 CRITERIA 
ANL-NBS-HS-000031 REV 00 ICN 01 
Although DTN: LL991109751021.094 has been superceded by DTN: LL000905312241.018, 
the superceded DTN was retained in this ICN. The superceding DTN contains revisions to 
figures and tables to reflect new Pu values. These changes did not impact the colloid size 
determination used in this AMR, although the plot used in the original DTN has been modified 
in the superceding DTN. The difference is minor and insignificant to this AMR, given the 
uncertainties in the measurements. 
No system description documentation (SDD) criteria are available at this time. This AMR 
complies with the DOE interim guidance (Dyer 1999). Subparts of the interim guidance that 
apply to this analysis or modeling activity are those pertaining to the characterization of the 
Yucca Mountain site (Subpart B, Section 15), the compilation of information regarding 
hydrogeology and geochemistry of the site in support of the License Application (Subpart B, 
Section 21(c)(1)(ii)), and the definition of hydrogeologic parameters and conceptual models used 
in performance assessment (Subpart E, Section 114(a)). 
Table 7, Figs. 7–10, Section 
6.1.5 
November 2001 
Alluvial Material Analysis 
Data Sources 
SN0004T0571599.004 
DTN: 
SN0004T0501600.004 
Section 6.1.4, Table 7, Figs. 
6–10 
Table 7, Figs. 7–10 
Section 6.2.2 DTN: 
LA0002JC831341.001 
DTN: 
LL991109751021.094 
12
The underlying assumptions of the radionuclide transport model are outlined in this section as 
the first step toward developing the conceptual, mathematical, and computational models needed 
Category 
Table 2. Assumptions 
Assumption 
Flow only in fractures. Colloids have 
diffusion coefficients at least three orders 
of magnitude smaller than solutes, and 
because of this and their large size, they 
will be not be able to negotiate significant 
matrix tortuosity. 
Constant aperture and flow rate 
throughout fracture model domain. 
Colloids experience the same mean 
residence time and hydrodynamic 
dispersivity in fractures as solutes. 
Colloid attachment and detachment 
described by first-order rate expressions 
(fracture only). 
Carboxylate-modified-latex microspheres 
are suitable analogs for natural colloids. 
Filtration and detachment rates of colloids 
to/from immobile surfaces is fast enough, 
relative to the porewater velocity, that the 
process can be modeled with an 
equilibrium formulation. 
Results from Prow Pass Tuff and Bullfrog 
Tuff tracer tests are equally weighted. 
The two pathways identified in the 
Bullfrog Tuff analysis are equally 
weighted. 
Basis 
Colloid-size range is larger than mean pore 
size of fractured volcanic tuffs. 
Philosophy was to use models that were 
only as sophisticated as knowledge of the 
system could support. Model complexity 
was introduced incrementally and only as 
necessary to match the field data. An 
example of additional complexity was the 
introduction of multiple flow pathways to 
match the Bullfrog test field data. 
Standard assumption in colloid transport 
modeling (e.g., Harvey and Garabedian, 
1991). 
RTA model simulates attachment and 
detachment with first-order kinetics and 
matches field observations. Data do not 
warrant more complex forms of the 
equation. 
Engineered microspheres were selected to 
optimize transport. Therefore, using them 
as an analog for natural colloids should be 
conservative. 
Damkohler number analyses conducted in 
Section 6.1 and 6.2 of this AMR, which 
show that the rates are sufficiently fast 
relative to water velocities that there is no 
difference between kinetic and equilibrium 
formulations, or that, at worst, the kinetic 
formulation will be conservative. 
Radionuclides will encounter the saturated 
zone in volcanic units beneath the 
repository. Both the Prow Pass and 
Bullfrog Tuffs are present at the water table 
beneath the repository. 
Differences in observed mass distribution 
between the two flow paths were more 
likely due to density effects in the borehole 
rather than natural phenomena. 
in PA calculations. 
No. 
1
2
3
4
5
6
7
8 
Location 
in this 
AMR 
Colloid location Sections 
6.1.1 and 
6.1.2. 
Flow domain Sections 
6.1.1 and 
6.1.2. 
Colloid transport Sections 
6.1.1 
Filtration 
parameters 
Sections 
6.1.1, 
6.1.6, and 
Fig. 2. 
Section 6. Applicability of 
microsphere 
studies 
Local equilibrium 
assumption 
Sections 
6.1.4 and 
6.2.4. 
Section 
6.1.3. 
Development of 
probability 
density function 
for colloid 
retardation factor 
Section 
6.1.3. 
Development of 
probability 
density function 
for colloid 
retardation factor 
ANL-NBS-HS-000031 REV 00 ICN 01 
5. ASSUMPTIONS 
13 November 2001
No. 
9 
10 
11 
12 
13 
ANL-NBS-HS-000031 REV 00 ICN 01 
Location 
in this 
Category AMR 
Section 
6.1.3. 
Development of 
probability 
density function 
for colloid 
retardation factor 
Section 
6.1.3. 
Development of 
probability 
density function 
for colloid 
retardation factor 
Pu-colloid form Section 6. 
Colloid density Sections 
6.2.2 and 
6.2.3. 
Alluvial grain 
size distribution 
Table 7, 
Fig. 7, 
Fig. 8, 
Fig. 9, 
Fig. 10. 
Additional 
ranges 
used in 
uncertaint 
y analysis 
in Section 
6.2.5. 
Table 2. Assumptions (Continued) 
Assumption 
Within a given pathway in the Bullfrog 
Tuff, the probabilities of colloids having 
transport parameters associated with 
each subpathway is assumed to be 
proportional to the mass fractions of 
colloids associated with those 
subpathways. 
The two microsphere responses in the 
Prow Pass test are equally weighted. 
For Pu-bearing colloids reaching the SZ, 
the Pu is irreversibly attached to the 
colloids. Although the retardation factors 
computed in this AMR are applicable to 
ALL colloids. The assumption only 
relates to the reason that this AMR does 
not address Pu-colloid reactions. 
Assumed to be equivalent to 
montmorillonite. 
Ranges from 0.02 to 0.11 cm (fine, 
medium, and course grain sands). 
[Size correlation to sand type based on 
the United States Department of 
Agriculture size fractions (Marshall et al. 
1996; Fig. 1.2, p. 4)] 
14 
Basis 
The splitting of colloid mass into 
subpathways was necessary to capture the 
complexity of the colloid response within 
each pathway (Reimus et al. 1999) 
Only one pathway is evaluated. The two 
different microsphere types are equally 
likely analogs for natural colloid transport. 
Indications from the Colloid-Associated 
Radionuclide Concentration Limits AMR 
(CRWMS M&O 2000a, Sec. 6.2) are that 
the plutonium is embedded within the 
released colloids as well as reversibly 
sorbed onto them. 
Density of plutonium-bearing colloids from 
waste form has not been measured, but 
has been identified as clay. 
We are not aware of any completed 
analyses of grain size distributions in the 
Yucca Mountain valley fill alluvium. 
Therefore, we use analyses from Yucca 
Flat as a surrogate (knowing that YMP will 
soon have material analyses that can be 
used to supercede these). Bechtel (1998) 
reported on the weight fraction of sand, silt, 
and clay in two alluvial boreholes in Yucca 
Flat. In the samples collected, sand was 
the dominant material with a weight fraction 
greater than 80%. Thus, we conservatively 
assume that the YM valley-fill alluvium 
system is sand. 
Marshall et al. (1996, Fig. 1.2, p. 4) show 
that the grain sizes represented by fine, 
medium, and coarse sands range between 
.02 and .11 cm. 
Further review of a Bechtel data report 
(Fulkerson 2001) indicates that their grain 
size distribution for sand ranges from 0.005 
to 0.2 cm. Therefore, sensitivity analysis in 
this AMR compares this slightly larger 
range to the range used for the PA 
calculations. 
It should be noted that horizons with grains 
much larger than a coarse sand are likely to 
have their interstitial spaces filled by finer 
material, and horizons with grain sizes 
much smaller (e.g., clays) are unlikely to 
transmit much water. 
November 2001
14 
15 
16 
17 
18 
ANL-NBS-HS-000031 REV 00 ICN 01 
Colloid filtration 
in alluvium 
Sections 
6.2.1, 
6.2.2, 
6.2.3, and 
6.2.8. 
Colloid filtration 
in alluvium 
Sections 
6.2.2 and 
6.2.3, 
Table 7, 
Fig. 7, 
Fig. 8, 
Fig. 9, 
Fig. 10 
Colloid filtration 
in alluvium 
Sections 
6.2.2 and 
6.2.3. 
Colloid filtration 
in alluvium 
Sections 
6.2.2 and 
6.2.3, 
Table 7. 
Colloid filtration 
in alluvium 
Section 
6.2.2 and 
6.2.3 
Table 2. Assumptions (Continued) 
Attachment rate can be predicted with 
kinetic filtration theory. 
Distribution range for Alpha parameter 
(collision efficiency factor) in kinetic 
filtration theory, is between 0.005 and 
0.025. 
Detachment rate range should cover the 
range of detachment rates of colloids and 
microspheres from fracture surfaces. 
Detachment rate range should also cover 
the range measured for microspheres and 
bacteriophages in the Borden Sand 
aquifer. This extends the lower range of 
detachment rates that were estimated for 
for fracture surfaces down from 1E-4 to 
1E-5. 
Water temperature is 25° Celsius
15 
In the absence of YMP data, the Cape Cod 
site is used as an analog (Harvey and 
Garabedian 1991, Table I, p. 179). Kinetic 
filtration theory was used to predict colloid 
attachment to immobile grains in that study 
(Harvey and Garabedian 1991, Eqs. 1, 2, 6, 
and p. 184) 
This is the range of collision efficiency 
factors found by Harvey and Garabedian 
(1991, p. 181, Table II) for bacterial and 
microsphere colloids in the Cape Cod sand 
aquifer. This is probably a conservative 
range since collision efficiency factors may 
be higher in Yucca Mountain alluvium if the 
material is less well-sorted than the uniform 
sand at Cape Cod. 
Although detachment has not been 
measured for Yucca Mountain alluvium, the 
process of detachment from alluvial grains 
may be similar to the process of 
detachment from fracture surfaces. 
Detachment rates of microspheres and 
bacteriophages were measured in a limited 
experiment at the Borden Sand aquifer 
(Bales et al. 1997, Table 3, parameter k2) 
In the absence of Yucca Mountain specific 
data, the range of detachment rates of 
colloids from porous media surfaces should 
include the values from the Borden aquifer. 
Sass et al. (1988) [DTN: 
GS950408318523.001] measured surface 
water temperatures to be about 20 degrees 
Celsius and a geothermal gradient of about 
25 degrees per 1000 meters. Alluvial flow 
and transport are considered to occur only 
in the shallow part of the aquifer, so a 
temperature of 25 degrees represents the 
temperature at about 200 meters depth. 
November 2001
19 
20 
21 
ANL-NBS-HS-000031 REV 00 ICN 01 
Assumption 19 
not used 
Table 3, 
Fig. 3-5 
Prow Pass 
solute mean 
residence time 
Fig. 1 Bullfrog Tuff 
tracer test data 
Table 2. Assumptions (Continued) 
The calculated hydrologic properties, rock 
properties and filtration characteristics 
contained in DTN LA9909PR831231.005 
are suitable for use in describing 
generalized tracer and colloid transport 
behavior. 
The colloid-sized-microsphere tracer test 
data contained in DTN 
LA0007PR831231.001 are suitable for 
use in describing generalized tracer and 
colloid transport behavior. 
16 
The calculated hydrologic properties are 
based on measurements of samples taken 
from the stratigraphic interval for which 
generalized tracer and colloid transport 
behavior are determined in the AMR. 
Although the exact depths from which the 
samples were taken were not adequately 
documented, the samples are considered to 
be representative of the stratigraphic 
intervals which they were assigned (see 
DTN:MO0011DQRCWELC.001), and 
further precise depth determinations are not 
needed for calculating hydrologic 
properties, rock properties, and filtration 
characteristics required for the description 
of generalized tracer and colloid transport 
behavior. 
DTN LA0007PR831231.001 includes data 
acquired over a period of nine months 
during microsphere tracer tests at the CWells 
complex. These data are qualified 
but not yet verified. The data were 
acquired during a series of tracer tests at 
the C-Wells complex which were conducted 
for the Yucca Mountain Project and 
performed under approved test procedures 
and appropriate QA requirements. DTN: 
LA0007PR831231.001 supercedes DTN: 
LA0002PR831231.001, which is qualified 
and verified and contains tracer test data 
other than the microsphere tracer test data. 
These data are unchanged in the 
superceding DTN. The microsphere tracer 
test data were collected continuously over 
the nine-month period in order to obtain 
breakthrough times, i.e., the times after 
injection at which the microspheres were 
detected in the sampling well, and the timedependent 
concentrations of the 
microspheres in the samples. The results 
were compared with other tracer results 
from the same suite of tests (i.e., 
nonsorbing solute tracer data, which are 
qualified and verified; DTN: 
LA9909PR831231.003) and used to derive 
colloid filtration rate and detachment rate 
constants. The microsphere and 
nonsorbing solute tracer responses were 
sufficiently similar and consistent that most 
of the transport factors (e.g., flow velocity 
and dispersion coefficient) for the 
microspheres were obtained from the 
interpretation of the nonsorbing solute 
tracer responses. Therefore, the data in 
DTN: LA0007PR831231.001 are of 
sufficient quality and are appropriate for use 
in the development of retardation factors for 
the SZ colloid model. No further 
confirmation is needed. 
November 2001
6. ANALYSIS/MODEL 
The primary cause of colloid retardation is attachment and detachment from immobile surfaces. 
This analysis demonstrates the development of parameters necessary to estimate attachment and 
detachment of colloids and, hence, retardation in both fractured tuff and porous alluvium. Field 
data are used for the analysis of colloid retardation in fractured tuff. Due to the lack of any field 
data for colloid transport in alluvial material, a theoretical analysis developed at another site is 
implemented in this analysis. 
6.1 COLLOID TRANSPORT IN FRACTURED TUFF 
This report considers the migration of plutonium-colloids where, based on CRWMS M&O 
2000a, Sec. 6.2) the plutonium is assumed to be irreversibly attached to colloids (Table 2). 
6.1.1 Background 
The colloid transport parameters developed for this AMR were derived from field tracer tests 
conducted in fractured volcanic tuffs at the C-wells, UE25-C#1, #2, and #3 (Reimus et al. 1999). 
The colloid tracers used at the C-wells were fluorescent carboxylate-modified latex (CML) 
polystyrene microspheres (Interfacial Dynamics Corp., Portland, Oregon). Table 3 summarizes 
the colloid tracer tests and microspheres used at the C-wells. Carboxylate-modified latex (CML) 
microspheres were chosen as field tracers because of the following: (1) they can be tagged with 
different fluorescent dyes that allow them to be quantified even at concentrations several orders 
of magnitude below background colloid concentrations; (2) they have a negative surface charge 
similar to rock surfaces and presumably natural colloids; and (3) they are more hydrophilic than 
other microspheres, which makes them more resistant to attachment and flocculation, even at 
higher ionic strengths (hence, making them good “conservative” colloid tracers). The pKa of 
carboxylic acid is around 5, and the C-wells groundwater pH was 7 to 8; therefore, the 
microspheres should have been negatively charged in the field tracer tests. Also, because of their 
hydrophilic surfaces, the CML microspheres are relatively stable even at ionic strengths 
approaching 1 molar (in monovalent cations), which is advantageous when injecting them with 
ionic solute tracers. CML microspheres behaved more conservatively (less attachment) than 
silica microspheres and non-modified carboxylated latex microspheres (hydrophobic surfaces) in 
previous laboratory experiments (Reimus 1995). The use of conservative colloid tracers was 
expected to result in conservative filtration parameter estimates, which in turn should result in 
conservative predictions of large-scale colloid transport. 
In the C-wells field tracer tests, the microsphere responses were interpreted by comparing them 
to the responses of nonsorbing solute tracers that were injected either simultaneously with or just 
after the microspheres. The responses of the nonsorbing solutes were fitted using a semianalytical, 
dual-porosity mass-transport computer model, Reactive Transport Application (RTA 
V1.1), to obtain estimates of the fractional mass participation, tracer mean-residence times, 
dispersivities, and matrix-diffusion mass-transfer coefficients in the flow system(s). The 
fractional mass participation is simply the fraction of the injected tracer mass accounting for the 
observed response or for a portion of the observed response (e.g., one of two peaks). Fractional 
mass participation could be physically caused by (1) a finite volumetric flow rate into the matrix 
November 2001 17 ANL-NBS-HS-000031 REV 00 ICN 01
that acts as an sink for tracers because its velocity is much too low for tracers to appear at the 
production well; (2) stagnation point(s) resulting from the weak recirculation in the tests, or 
Table 3. Summary of C-Wells Tracer Tests Involving Microspheres 
Prow Pass Tuff Bullfrog Tuff Test Parameter 
UE25-c#2 / c#3 
800-900 
568 
0.033 
37, 995 
Br-, PFBAInjection/
Production Wells 
Interval, m below surface(1) 
Production Flow Rate, L/min 
Fraction Recirculated(2) 
Mean Solute Residence Time(3), hrs 
Nonsorbing Solutes(4) 
360 / yellow Microsphere Diameter, nm / Dye 
UE25-c#3 / c#2 
630-705 
19.5 
0.3 
1230 
Br-, Cl-, PFBA- 
280 / orange(5) 
280 / yellow(5) 
640 / blue 
DTNs: Bullfrog test flow rate: MO0111BFROGPRO.000, MO0111BFROGPRO.001, MO0111BFROGPRO.002, 
NOTES: 
b
1 ¶S 
¶t 
- k filt C + k resS = 0 
MO0110BFROGREC.001, MO0111BFROGPRO.003; Prow Pass test flow rate: GS010799992315.001, 
MO0111PROWPASS.000; Bullfrog test mean residence time: LA9909PR831231.003; Prow Pass test mean 
residence time: Assumption 20 
(1) Intervals are approximate because packer locations were different in each well. Water table is 
~400 m below surface. 
(2) Fraction of produced water recirculated into the injection well. 
(3) Two mean residence times are listed for the Bullfrog Tuff, one for each of the two tracer peaks. 
(4) PFBA- is pentafluorobenzoate. 
(5) The 280-nm yellow and orange microspheres were the same spheres, but they were tagged with 
(Eq. 1) 
(Eq. 2) 
November 2001 
different fluorescent dyes because they were injected at different times. 
(3) loss of tracer due to a portion of the relatively dense injection solution “sinking” out of the 
zone of influence of pumping. The matrix-diffusion mass transfer coefficient is a lumped 
parameter (porosity times square root of matrix diffusion coefficient divided by average fracture 
half-aperture, time-1/2) that describes the diffusive mass transfer rate of solutes between fractures 
and matrix. 
It was assumed that the fractional mass participation, the mean-residence times, and 
dispersivities obtained for the solutes also applied to the microspheres (e.g., Harvey and 
Garabedian, 1991). However, because of their large size and small diffusivities compared to the 
solutes (approximately 3 orders-of-magnitude smaller), it was assumed that there was no matrix 
diffusion of microspheres (matrix-diffusion coefficient was set equal to zero). The advectiondispersion 
equation with appropriate terms for a single reversible first-order reaction to account 
for mass transfer between mobile water and immobile surfaces (filtration and detachment) was 
used to model microsphere transport: 
¶C ¶C ¶ 2C 
¶t 
+ V ¶x 
- D ¶x 2 + k filt C - k resS = 0 
18 ANL-NBS-HS-000031 REV 00 ICN 01
where 
C = colloid concentration in solution, no./L 
S = colloid concentration on fracture surfaces, no./cm2 
V = flow velocity in fractures, cm/sec 
D = dispersion coefficient, cm2/sec 
kfilt = filtration rate constant (1/sec) = lV, where l = filtration coefficient (1/cm) 
kres = detachment rate constant, 1/cm-sec 
x, t = independent variables for distance and time, respectively. 
b = fracture half aperture (cm). 
The values for V and D in Equations 1 and 2 were obtained from the interpretation of the 
nonsorbing solute tracer responses; therefore, the filtration and detachment rate constants were 
the only parameters adjusted to match the microsphere responses. Details of the interpretation 
procedure for both the Bullfrog Tuff and Prow Pass Tuff tracer tests are provided in Reimus et 
al. (1999). 
6.1.2 Interpretations of Microsphere Responses in Tracer Tests 
The interpretation of the Bullfrog Tuff tracer test was complicated by the fact that the 
breakthrough curves of all tracers were bimodal. This problem was addressed by treating each 
“peak” as being the result of a separate set of flow pathways, each with its own set of transport 
parameters, which, when added together, resulted in the observed composite response. The term 
“pathway” is henceforth used in this AMR to refer to such sets of flow pathways (loosely defined 
as pathways that, when lumped together, result in a tracer response that can be described using a 
single set of parameters in the advection-dispersion equation). Using the relatively simple model 
represented by Equations 1 and 2, it was not possible to match the 360-nm microsphere response 
with a single set of filtration parameters for either peak. Instead, the microsphere mass in each 
pathway was divided into subsets that were assumed to experience different detachment rates 
due to both physical and chemical heterogeneities in the system. These subsets of mass are 
referred to as “subpathways” because they represent a fraction of the total transport within a 
given “pathway.” All of the microsphere mass in each of the two primary pathways was 
assumed to experience the same forward filtration rate (defined as the “forward” component of 
reversible filtration). 
Figure 1 shows the resulting composite fit to the microsphere response in the Bullfrog Tuff test 
along with the predicted contributing responses of each “subpathway.” The filtration and 
detachment rate constants associated with the subpathways are listed in Table 4. Note that the 
detachment rate constants for subpathways 1A and 2A are the maximum values that can be used 
without degrading the overall fit to the response at late times. That is, if larger detachment rate 
constants were used, the tail of the predicted response would be raised higher than the data. 
However, the overall fit to the data would be just as good if kres in subpathways 1A and 2A were 
set equal to zero (i.e., irreversible filtration); hence, this interpretive procedure can only establish 
maximum detachment rate constants. These maximum values should yield the most 
conservative (fastest) colloid transport predictions. 
November 2001 19 ANL-NBS-HS-000031 REV 00 ICN 01
Sources: Assumption 21 for data, DTN: LA9909PR831231.003 for model input data. 
NOTE: Normalized concentration is 109 times particles/L divided by the total number of particles injected. 
Figure 1. Composite Fit to the 360-nm Diameter Microsphere Response in the Bullfrog Tuff Tracer Test 
November 2001 20 ANL-NBS-HS-000031 REV 00 ICN 01
Path 2C 
0.1 0.423 0.115 f, mass fraction 
0.04 0.04 0.2 kfilt, 1/hr 
0.0133 0.0133 0.00247 l(1), 1/cm 
Table 4. Filtration and Detachment Rate Constants for the Microspheres in each 
Subpathway of the Bullfrog Tuff Tracer Test 
Path 2A Path 1A Parameter Path 2B 
0.067 
0.04 
0.0133 
0.4 
Path 1B 
0.005 
0.2 
0.00247 
3.33 0.008 0.000404(3) 0.00025(3) bkres
 (2), 1/hr 
DTNs: LA9909PR831231.003 (for mass fractions and detachment rate constants for paths 1B, 2B, and 2C). 
(1) NOTES: l = filtration coefficient, calculated as kfilt/V, where V = average linear velocity determined from mean 
fluid residence time. 
(2) b = fracture half aperture, cm. The fitted detachment rate constant is this lumped parameter. 
November 2001 
(3) Maximum detachment rate constant; cannot distinguish between this value and zero. 
The interpretation of the Prow Pass Tuff tracer test was much more straightforward than that for 
the Bullfrog Tuff test. A single set of filtration parameters could adequately explain the 
responses of both microspheres. Figure 2 shows the resulting fits to the microsphere responses, 
and Table 5 lists the filtration parameters corresponding to these fits. Note that the 280-nm 
yellow microspheres never appeared at the production well. These spheres were injected in the 
same solution as the solute tracers, which had an ionic-strength solution of ~0.4 M due to the 
high solute masses/concentrations necessary to ensure quantifiable responses at the production 
well. It is suspected that the spheres quickly attached to fracture surfaces or to each other 
because of the destabilizing effect of the high ionic strength. The filtration and detachment rate 
constants reported for these microspheres are the minimum and maximum values, respectively, 
that result in no predicted response of these colloids. That is, any smaller value of the filtration 
rate constant or larger value of the detachment rate constant would result in a small predicted 
response that was not observed. Because the yellow microspheres were injected in such a high 
ionic-strength solution, they were not considered in the development of cumulative probability 
density functions for colloid transport parameters (see below). Their inclusion in this analysis 
would raise the probabilities of high-filtration-rate constants and low-detachment-rate constants, 
which would raise the probability of high retardation factors. Thus, their omission from the 
following analysis is considered conservative. The yellow spheres were injected with the solute 
tracers primarily because this was the only practical way of investigating the effect of solution 
ionic strength on colloid transport in the field tests. 
21 ANL-NBS-HS-000031 REV 00 ICN 01
100.0 
640-nm 
280-nm 
640-nm Fit 
280-nm Fit 
1.0 
0.1 
900 1200 1500 1800 2100 2400 2700 3000 300 
particles injected. 
280-nm Yellow 
0.2(1) kfilt, 1/hr 
0.08 l, 1/cm 
10.0 
bkres
(2), 1/hr 0.001 
600 
640-nm Blue 
0.043 
0.017 
0.000154 
Output data - DTN: LA9912PR831231.006 
NOTES: 
0 
Parameters 
Table 3. 
Time, hrs 
DTN: LAPR831231AQ99.001 for data. Output data - DTN: LA9912PR831231.006 for model fits. 
NOTES: The zig-zag appearance of the fits at late times is the result of minor instabilities in the inversion algorithm 
in the RTA model. Normalized concentration is 109 times particles/L divided by the total number of 
Figure 2. Fits to the Microsphere Responses in the Prow Pass Tuff Tracer Test 
Table 5. Filtration and Detachment Rate Constants for the Microspheres in the 
Prow Pass Tuff Tracer Test 
Microspheres 
280-nm Orange 
0.07 
0.028 
0.000251 
(1) Minimum value that is consistent with the lack of appearance of these spheres at the production well. 
The actual filtration rate constant could be much higher. 
(2) Maximum values; cannot distinguish between these values and zero. See also footnote (2) of 
November 2001 22 ANL-NBS-HS-000031 REV 00 ICN 01 
Normalized Concentration, 10^-9/L .
6.1.3 Development of Cumulative Probability Density Functions 
The assumptions associated with the development of cumulative probability density functions for 
colloid-filtration-rate constants, detachment-rate constants, and effective retardation factors, 
based on Reimus et al. (1999), are stated in Table 2. 
The assumptions effectively divided the two tracer tests into four equally-weighted “trials.” 
1. The first peak of 360-nm-diameter microspheres in the Bullfrog Tuff test. 
2. The second peak of 360-nm-diameter microspheres in the Bullfrog Tuff test. 
3. The response of 280-nm-diameter orange microspheres in the Prow Pass test. 
4. The response of 640-nm-diameter blue microspheres in the Prow Pass test. 
No weights were assigned to the different sizes of microspheres. Thus, the 360-nm spheres used 
in the Bullfrog Tuff test implicitly were given a higher weighting than the other spheres because 
they were used in two of the four “trials.” 
Figures 3 and 4 show the resulting cumulative probability density functions for the microspherefiltration 
and detachment-rate constants, respectively. Figure 5 shows the cumulative probability 
density function for the effective microsphere retardation factor. The points plotted in Figures 3 
through 5 are listed in Table 6. The retardation factors, R, in Figure 5 and Table 6 were obtained 
by applying the following equation in each pathway of each test: 
kfilt (Eq. 3) R = 1 + 
bkres 
Thus, the R values were based on actual observations in each pathway; they were not developed 
by independently randomly sampling the two probability density functions in Figures 3 and 4. 
The former approach effectively captures any correlations that may exist between the filtration 
and detachment rate constants (forward and reverse reactions, respectively), whereas in the latter 
approach, any correlations will be lost. No analyses were attempted to establish formally such 
correlations between forward and reverse rates. 
November 2001 23 ANL-NBS-HS-000031 REV 00 ICN 01
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1
0 
Cumulative Probability . 
0 
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1 
SOURCES: DTN: LA9909PR831231.003 and Table 2, Assumption 20 for model input data. 
Figure 3. Discrete Cumulative Probability Density Function for Microsphere Filtration 
Rate Constants in the C-Wells Tracer Tests 
Kfilt, 1/hr 
0.1 
0
0.0001 
SOURCES: DTN: LA9909PR831231.003 and Table 2, Assumption 20 for model input data. 
Figure 4. Discrete Cumulative Probability Density Function for Microsphere 
Detachment Rate Constants in the C-Wells Tracer Tests 
b*Kres, 1/hr 
ANL-NBS-HS-000031 REV 00 ICN 01 
Cumulative Probability . 
0.2 0.15 0.1 0.05 
10 1 0.01 0.001 
November 2001 24
Cumulative Probability . 
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1
0 
1 
Group 1 
Filtration Only 
Probability kfilt, 1/hr 
0.25 0.04 
0.5 0.043 
0.75 0.07 
1 0.2 
SOURCES: DTN: LA9909PR831231.003 and Table 2, Assumption 20 for model input data. 
Figure 5. Discrete Cumulative Probability Density Function for 
Microsphere Retardation Factors in the C-Wells Tracer Tests 
Table 6. Values Used for Cumulative Probability Density Functions Shown in Figures 3 through 5 
Retardation Factor 
Group 2 
Detachment Only Retardation Factors with Coupled Rates 
Distribution 
Probability R 
0.25 1.06 
1.1 0.7395 
6 0.91875 
100 0.961 
280 0.9895 
280 1 
800 
ANL-NBS-HS-000031 REV 00 ICN 01 
1000 100 10 
Group 3 
Probability bkres, 1/hr kfilt bkres 
0.000154 0.0105 3.33 0.2 
0.039 0.4 0.04 0.00025 
0.08125 0.008 0.04 0.000404 
0.2605 0.0004 0.04 0.008 
0.5102 0.000251 0.07 0.4 
0.7605 0.000154 0.043 3.33 
1.0 0.00025 0.2 
November 2001 25
6.1.4 Validity of the Local Equilibrium Assumption in Estimating Retardation for 
Saturated Fractured Tuff 
The length and time scales associated with colloid transport in saturated, fractured tuff will both 
be significantly larger than those sampled during the forced gradient tests from which these 
parameters are derived. Therefore, the time scales associated with colloid filtration and 
detachment in the saturated tuff have been assumed to be small relative to fluid residence times 
(an assumption necessary for the local equilibrium assumption to be valid). To evaluate the 
validity of this assumption, a simple analysis can be performed with non-dimensional Damkohler 
numbers. The Damkohler number is simply the rate constant, k (1/hr), multiplied by a 
representative residence time, T (hr), Da=kT. Bahr and Rubin (1987, p. 440, Equation 12) 
demonstrate that the mass balance equation describing solute transport can be differentiated into 
an equilibrium and kinetic component. The smaller the kinetic component, the more accurate are 
the retardation factors based on the local equilibrium assumption. 
For evaluation of colloid behavior, Damkohler numbers, Daatt and Dadet, can be computed for 
attachment and detachment of colloids, respectively, using kfilt and kres. The magnitude of the 
kinetic component is inversely proportional to Daatt+Dadet. Thus, the larger the sum of the two 
Damkohler numbers, the more appropriate the assumption of equilibrium. Bahr and Rubin 
(1987, p. 450) found that equilibrium was well approximated when the sum of the two 
Damkohler numbers is greater than 100 and reasonably well estimated when the sum is greater 
than 10. Valocchi (1985, p. 813, Figure 2) had a similar result, although he only used the reverse 
rate to compute a Damkohler number similar to Dadet in this analysis. Bahr and Rubin (1987) 
point out that the kinetic term can only be completely separated when the sum of the two 
Damkohler numbers is used. 
To compute the residence time, T, the length (L), specific discharge (q), and porosity (q) are 
needed (T=L/qq), where porosity is the fracture volume fraction of the domain. A representative 
length scale, L, for the Yucca Mountain saturated volcanic tuffs is 16,000 meters, but specific 
discharge and porosity are variable parameters drawn from distributions. The attachment and 
detachment rates are also drawn from a distribution (Table 6, group 3). Therefore, a GoldSim 
V6.03 (STN: 10032-1.1-00) model is used to compute cumulative distributions for Daatt, Dadet, 
and the sum of the two for this evaluation of the validity of the local equilibrium assumption in 
estimating retardation factors. The porosity ranges between 0.00001 and 0.1 with a log-uniform 
distribution, and the specific discharge is characterized with a discrete distribution spanning low, 
medium, and high rates (0.06 m/yr 24% of the time, 0.6 m/yr 52%, and 6 m/yr 24%) (CRWMS 
M&O 2000b, DTN SN0004T0571599.004). The low end of the porosity range is well below the 
porosity estimates from the C-wells, but it is used in this exercise to allow a finite probability of 
very rapid flow and transport pathways through the fractured tuff. The attachment and 
detachment rate distributions (Table 6, Group 3) are correlated with the retardation factor 
probabilities in Group 3 of Table 6. 
Figure 6 shows the cumulative distributions of Daatt, Dadet, and Daatt +Dadet resulting from a 5000 
realization GoldSim calculation. The sum of the two Damkohler numbers drops below 100 in 
less than 5 percent of the realizations and below 10 in less than 1 percent of the realizations. In 
these few cases, the assumption of local equilibrium may not be valid. However, for those cases, 
November 2001 26 ANL-NBS-HS-000031 REV 00 ICN 01
the residence times are very small due to very low porosity and high specific discharge. 
Specifically, for the simulations in which Daatt +Dadet <100, travel times through the tuff aquifer 
estimated from these parameters are less than 20 years, which is much less than the presumed 
regulatory time frame of 10,000 years. Therefore, a slight overestimate of travel time as a result 
of assuming local equilibrium for these few cases should not have an impact on PA results. 
Also, it should be noted that the largest filtration rate constant observed in the C-wells field tests 
(0.2 hr-1) was associated with the shortest travel time in any of the tests. Although it is not 
possible to establish a correlation between filtration rates and travel times from this single 
observation, if an inverse correlation exists, then the above analysis would be conservative. 
In about 30 percent of the realizations Dadet is less than 10, indicating that its contribution may 
not be modeled well with the equilibrium approximation. However, in such cases, using an 
equilibrium approximation leads to errors on the side of conservatism with this term because 
retention of colloids on the fracture surfaces is underestimated. The attachment terms, as shown 
in Figure 6, are always greater than 10. If there is any error in the attachment term (for those less 
than 100), the equilibrium approximation is not conservative, but as stated above, the errors can 
only lead to overestimating retardation by about 20 years, which is an error that will have no 
consequence in the PA calculations (Recall also that Bahr and Rubin (1987) caution that both 
Damkohler numbers, Daatt+Dadet, should be used to assess the kinetic term). 
Finally, a particularly conservative aspect of this analysis is that there is no probability of a zero 
detachment rate constant. In reality, it is very likely that some colloids will irreversibly attach. 
Although only a small fraction of the injected microspheres were recovered in the C-wells tracer 
tests (Figures 1 and 2), the retardation factors derived from the kinetic parameters that were used 
to fit the data will lead to predictions of 100% recovery. 
November 2001 27 ANL-NBS-HS-000031 REV 00 ICN 01
DTN: SN0004T0571599.004 (for flow rates). Output Data – DTN LA0004AW12213S.001. 
Damkohler Number (Da) 
Figure 6. Damkohler Number Distributions for Attachment and Detachment Kinetic Rates (Table 6) for 
Fractured Tuff 
6.1.5 Uncertainty in Assumptions 
Appropriate caution should be exercised when using the probability density functions of Figures 
3 through 5 to represent the behavior of radionuclide-bearing colloids in fractured volcanic tuffs. 
One reason for this caution is that it is not known how well the polystyrene microspheres 
represent such colloids. Although they were chosen for their similarity in size, surface charge, 
and hydrophilicity to colloids that may facilitate radionuclide transport, the microspheres 
certainly differ from such colloids in their surface chemistry, shape, and density. The density of 
the microspheres is 1.055 g/cm3 (almost neutrally buoyant), as compared to natural or wasteform 
colloids, which should have a density of 2.0-2.5 g/cm3. It can be shown that the latex 
spheres should settle (by gravity) at about the same rate as a 2.5 g/cm3 colloid that is ~5.2 times 
smaller in diameter (see last term of Equation 6 in Section 6.2.2). Waste-form colloids are about 
120-160 nm in diameter (DTN: LL991109751021.094), so the microspheres used in the C-wells 
tests (ranging from 280- to 640-nm diameter) should have actually settled more slowly than 
waste-form colloids, which is conservative. The larger microspheres would also have diffused 
more slowly than waste-form colloids (diffusivity is inversely proportional to diameter), which is 
conservative. However, the microspheres would have been more prone to inertial collisions with 
fracture surfaces than waste-form colloids (interception is proportional to diameter squared), 
which is nonconservative. 
Another reason for exercising caution in using the probability density functions of Figures 3 
through 5 is that all C-wells observations were made over only a ~30-meter travel distance and 
for less than one year. Extrapolation to longer distances and time scales is an exercise in 
ANL-NBS-HS-000031 REV 00 ICN 01 November 2001 28 
Cumulative Probability 
1 
0.9 
0.8 Dadet 
0.7 
0.6 Daatt 
0.5 
Daatt + Dadet 
0.4 
0.3 
0.2 
0.1
0
0.01 1000000 10000000 100000 10000 1000 100 10 1 0.1
uncertainty. The flow conditions in the field tests were, by necessity, perturbed from natural 
flow conditions with significant forced gradients being imposed to induce tracer movement. 
Also, there were three significant flow interruptions during the Prow Pass Tuff test, of which at 
least the first two appeared to result in brief, large jumps in the colloid concentrations upon 
resumption of flow at approximately 1000 and 1250 hrs (see Figure 2). These spikes in colloid 
concentration were ignored when estimating colloid transport parameters. The retardation factor 
distribution generated in this analysis is based on the assumption that the Bullfrog Tuff and the 
Prow Pass Tuff are equally probable formation types through which colloids would travel in the 
saturated zone (Table 2). If the Prow Pass Tuff parameters were not considered due to the higher 
permeability in the Bullfrog Tuff (based on an argument suggesting that pathways through the 
saturated tuff are more likely in higher permeability units), then the resulting retardation factor 
distribution would show higher probabilities for lower values. The net result of dropping the 
Prow Pass data would be to double the probabilities of the smaller retardation factors and the 
larger detachment rate constants. 
6.1.6 Model Summary: Fractured Tuff Colloid Retardation Factors 
At the field scale, colloid retardation factors in fractured tuff can be estimated as attachment rates 
of colloids to fracture walls divided by detachment rates of colloids from fracture walls. The 
transport model, RTA V1.1, is parameterized such that it is used to estimate these attachment and 
detachment rate parameters by matching model predictions to the microsphere transport data 
from the Bullfrog and Prow Pass tuff tracer tests at the C-wells complex. Since multiple tests 
were conducted, each providing different parameters, a distribution of retardation factors is 
developed to represent all of the different testing conditions and flow intervals tested. This 
distribution of retardation factors is used in predictive models that may be used in PA 
calculations. 
6.1.7 Model Validation Summary: Fractured Tuff Colloid Transport 
The kinetic model of colloid attachment and detachment in fractured tuffs (Equations 1 and 2) is 
validated by matching the field observations of microsphere transport at the C-wells (Figures 1 
and 2). However, the use of the derived probability distribution of retardation factors in fieldscale 
predictions requires that the local equilibrium approximation be valid. Therefore, an 
additional analysis (Section 6.1.4) evaluates the validity of the equilibrium approximation at the 
field scale and demonstrates that it is either accurate or conservative. 
November 2001 29 ANL-NBS-HS-000031 REV 00 ICN 01
6.2 COLLOID TRANSPORT IN ALLUVIAL MATERIAL 
6.2.1 Background 
Retardation factors for colloids, natural or synthetic, have not been measured in the alluvial 
aquifer downgradient of Yucca Mountain. There is a good possibility that such parameters will 
be measured within the next few years if the currently planned alluvial tracer testing is 
implemented. However, TSPA requires estimates of retardation factors for colloids in alluvium 
now. Therefore, this portion of the AMR provides a theoretical approach based, in part, on field 
tests at other alluvial aquifers. Those alluvial aquifers considered are a sand aquifer at the 
Canadian Forces Base, Borden, Ontario, studied by Bales et al. (1997) and a sand aquifer on 
Cape Cod, Massachusetts, studied by Harvey and Garabedian (1991). Direct comparison 
between these aquifers and the valley fill alluvium in the Yucca Mountain flow system is 
impossible because the Yucca Mountain alluvium has not to been characterized to date. 
However, alluvial material in Yucca Flat, which is less than 50 miles from Yucca Mountain, is 
characterized mostly by sand with finer material, including clay and silt, accounting for less than 
25 percent of the weight fraction in the samples analyzed by Fulkerson (2001). Compared with 
the clean, well-sorted sands of the other study sites, it is likely that the Yucca Mountain 
alluvium, like the Yucca Flat alluvium, is less well-sorted and that the range of grain sizes is 
larger. Further, the interstitial pore spaces between larger grains, rocks, and cobbles are likely to 
be filled with smaller grain sands and silts. Still, the theoretical approach taken here accounts for 
variable grain sizes. Therefore, because the Yucca Mountain alluvium, like the other two sites 
where colloid transport was studied, is characterized by flow-through porous media (as opposed 
to fractured tuff), the approach taken by Harvey and Garabedian (1991) is the best approach, 
considering the absence of Yucca Mountain specific data. The approach allows specifically for 
the grain size distribution to be specified. Thus, the method tested with site colloid transport data 
at Cape Cod is extended, incorporating estimates of the grain sizes specific to the Yucca 
Mountain system. 
6.2.2 Parameters for Calculation of Retardation Factors 
The retardation factor, R, by definition implies equilibrium conditions. Thus, the rate of 
accumulation of colloids in the attached phase is zero and R, for colloid transport in alluvial 
material, is approximated, similarly as in Equation 3, as follows: 
where
a 
q k 
R = 1 + 
ra k f
r 
r denotes the density of the alluvial material 
q is the porosity 
k f is the rate of colloid attachment onto the immobile material 
k r is the detachment rate. 
30 ANL-NBS-HS-000031 REV 00 ICN 01 
(Eq. 4) 
November 2001
Equation 4 is directly derived from Harvey and Garabedian (1991, Equation 1 and Equation 5) or 
Bales et al. (1997, Equation 3). Colloid attachment, k f , is estimated with colloid filtration 
theory, which considers the diameter of the porous media grains, the size of the colloids, the 
single-collector efficiency (the rate at which colloids strike a porous media grain divided by the 
rate at which colloids move toward the grain), and the collision efficiency factor (the efficiency 
with which collisions between colloids and immobile grains results in immobilization of the 
colloid). The standard colloid filtration theory equation for the rate of attachment is given by 
Harvey and Garabedian (1991, Equation 2) as: 
k 
d 2 
where 
h can be estimated by h = hD + hI + hG = : 
2 / 3 
0.9 + 1.5( d 
18 mv 
where
D
I 
é  kT ù  
ë  ê mdpdv û  ú  
h is the colloid collector collision caused by Brownian motion 
h is the colloid collector collision caused by interception 
G is the colloid collector collision caused by settling 
k is the Boltzmann constant 
T is the solute temperature 
m is the fluid viscosity 
dp is the colloid diameter 
d is the diameter of the porous media grains 
r is the fluid density 
h
p is the colloid density 
g is gravity. 
f = v 
3 (1 - q) ah, 
v is the fluid velocity 
d is the diameter of the porous media grains 
a is the collision efficiency factor 
h is the single-collector efficiency. 
p / d )2 + 
(rp - r)gd p 
2 
. 
31 
r 
The collision efficiency factor, a, is estimated from field experiments by Harvey and Garabedian 
(1991, Table II) and ranges between 0.005 (bacterial colloids) and 0.025 (microsphere colloids). 
The collision efficiency factor represents the number of colloid-collector collisions that occur 
before attachment (e.g., a value of 0.005 indicates that 500 collisions occur before attachment). 
Whereas, very small values represent unfavorable conditions for filtration, larger values indicate 
ANL-NBS-HS-000031 REV 00 ICN 01 
(Eq. 5) 
(Eq. 6) 
November 2001
favorable conditions (when a=1, filtration occurs with every collision, and colloids are least 
mobile). 
If all of the parameters in Equations 5 and 6 were well known and non-varying, then computing 
kf would be trivial. However, there is uncertainty in the grain size, the colloid size, the alluvium 
porosity, the specific discharge (flux), and the collision efficiency factor. Therefore, those terms 
are considered with distributions spanning their range of uncertainty and are reported in Table 7 
(sources reported in Section 4). The terms in Equations 5 and 6 that are held constant are k, T, m, 
r, rp, and g. T, as reported in Section 4, is held constant at 25 degrees Celsius (DTN: 
GS950408318523.001), thus removing and temperature dependent variability in other 
parameters in this analysis. 
In addition to kf, the calculation of a retardation factor in Equation 4 requires kr. However, 
filtration theory does not provide a method for estimating detachment of colloids from grains 
with such measurable parameters as particle size, grain size, or fluid velocity. Therefore, the 
detachment rate is also treated as an uncertain parameter. The range for the distribution of kr is 
estimated from reported field experiments of colloid transport in the Borden aquifer (Bales et al. 
1997, Table 3, p. 645) and from the detachment rates of colloids from fractured tuff at Yucca 
Mountain (Section 6.1 of this report). The range of detachment rates used in this analysis spans 
nearly six orders of magnitude. This range represents our uncertainty due to the lack of data for 
the system under consideration. It captures both ends of the spectrum for what are considered to 
be plausible or possible values for the saturated alluvium downgradient of Yucca Mountain. 
November 2001 32 ANL-NBS-HS-000031 REV 00 ICN 01
Max Min 
Grain size (d) (cm) (1) 
Colloid Size (dp) (cm) 
Table 7. Parameters for Retardation Factor Calculations and Uncertainty Range Distributions 
Parameter 
Uniform 0.11 0.02 
Uniform 4.5E-05 
Uniform .025 
6.0E-07 
.005 Alpha (a) 
Porosity (q) 
Flux (q) (m/yr)(2) 0.2 
0 (min limit) 1 (max limit) Truncated Normal (mean = 0.18, s=0.055) 
20
3.33 
Discrete (0.2,24%; 2.0,52%; 20.0,24%) 
Log-Uniform 1.0E-05 Detachment Rate (kr) (1/hr) 
DTNs: SN0004T0571599.004 (for porosity), SN0004T0501600.004 (for flux), LL991109751021.094 (for colloid size). 
Other sources: Marshall et al. (1996), Fulkerson (2001), and Bechtel (1998) (for grain size), Harvey and 
Garabedian (1991) (for alpha), and Bales et al. (1997) and Table 6 of this report (for detachment rate). 
Section 5. 
NOTES: (1) Additional ranges of grain sizes are considered in Section 6.2.5. Basis for this assumption discussed in 
November 2001 
(2) The velocity (v) in Equations 5 and 6 is computed from the flux and porosity; v=q/q 
6.2.3 Calculation of Retardation Factor Distribution 
Using the parameter distributions in Table 7, a distribution of retardation factors for colloids in 
saturated alluvial material (Equations 4 through 6) is computed with GoldSim V6.03 (STN: 
10296-6.03-00). Also, the groundwater temperature is set at 25 degrees Celsius in these 
simulations as described in Table 2 (DTN: GS950408318523.001). The model draws from the 
prescribed parameter distributions to compute a new retardation factor for each realization. 
Figure 7 shows the GoldSim schematic for computing retardation factors using Equations 4 
through 6. Figure 8 shows the results from five different 10,000 realization simulations. 
Although each simulation shown in Figure 8 results from sampling independently the 
distributions for each uncertain parameter listed in Table 7, all five simulations are virtually 
identical as a result of 10,000 realizations in each simulation. The curve in Figure 8 provides a 
basis from which PA can create a discrete probability distribution of retardation factors for their 
simulations. 
33 ANL-NBS-HS-000031 REV 00 ICN 01 
Distribution
Figure 7. Schematic of GoldSim Model of Equations 4-6 for Alluvial Colloid Retardation Factors 
November 2001 34 ANL-NBS-HS-000031 REV 00 ICN 01
DTNs: SN0004T0571599.004 (for porosity), SN0004T0501600.004 (for flux), LL991109751021.094 (for colloid size). 
Other sources: Marshall et al. (1996), Fulkerson (2001), and Bechtel (1998) (for grain size), Harvey and 
Retardation Factor 
Cumulative Distribution 
NOTES: 
Garabedian (1991) (for alpha), and Bales et al. (1997) and Table 6 of this report (for detachment rate). 
Output Data – DTN LA0004AW12213S.001. 
November 2001 
6.2.4 Validity of the Local Equilibrium Assumption in Estimating Retardation for 
Whereas high porewater velocities in the fractured tuff brings the assumption of local 
equilibrium for colloid retardation into question, there appears to be no such issue in the alluvial 
aquifer. The Damkohler numbers for the alluvial system, as shown in Figure 7, are computed 
using a length scale of 5000 meters. Due to the high porosity, Daatt+Dadet is always greater than 
1000 for the distributions of specific discharge (flux) and porosity in the alluvium, discussed 
earlier in this section. Figure 9 shows the Damkohler number distributions (computed with 
GoldSim V6.03 (STN: 10296-6.03-00) as shown in Figure 7) associated with the estimation of 
retardation factors computed for Figure 8. Although Daatt never drops below 100, Dadet drops 
below 100 in about 10 percent of the realizations. For these cases, estimating retardation factors 
with the local equilibrium assumption leads only to consevative errors as retention on the 
immobile surfaces would be larger if modeled with a kinetic formulation. 
ANL-NBS-HS-000031 REV 00 ICN 01 
Figure 8. Retardation Factor Distribution For Five Simulations 
Saturated Alluvium 
35 
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
Simulation 1 
Simulation 2 
Simulation 3 
Simulation 4 
Simulation 5 
0.1
0 
1000000 100000 10000 1000 100 10 1
Results of all five simulations are essentially superimposed.
As mentioned in Section 6.1.4, the use of retardation factors is conservative because there is no 
probability of a zero detachment constant. In reality, it is very likely that some colloids will 
irreversibly attach to immobile surfaces. 
DTNs: SN0004T0571599.004 (for porosity), SN0004T0501600.004 (for flux), LL991109751021.094 (for colloid size). Other sources: 
Marshall et al. (1996), Fulkerson (2001), and Bechtel (1998) (for grain size), Harvey and Garabedian (1991) (for alpha), and 
Bales et al. (1997) and Table 6 of this report (for detachment rate). Output Data – DTN LA0004AW12213S.001. 
November 2001 
6.2.5 Uncertainty in Assumptions 
Although the GoldSim V6.03 (STN: 10296-6.03-00) simulation shown in Figure 8 captures 
ranges of uncertainty in the parameter distributions leading to the computation of alluvial 
retardation factors, there is still uncertainty in the ranges and shapes of those distributions due to 
the paucity of data for the alluvial system. One of the most important parameter distributions 
affecting the shape of the simulated retardation factor distribution is the grain size distribution. 
For the purposes of this AMR, a mixture of fine, medium, and course sand has been assumed to 
be representative of the alluvial material (Table 2). However, if the valley fill deposits are 
actually coarser or finer, then the retardation factor distribution will be shifted. Although the 
valley fill deposits of the system under consideration are only now being characterized and have 
not been documented in a referenceable format for this AMR, there is the possibility that the 
material may be poorly sorted and that the interstices between pebbles and cobbles will be filled 
with fine-grained material. Under such conditions, the fine-grained material surface area may 
increase the collector efficiency. Similarly, without more detailed understanding of the system 
under consideration, there is some possibility that the valley fill material may actually be 
ANL-NBS-HS-000031 REV 00 ICN 01 
Figure 9. Damkohler Number Distributions for Attachment and Detachment 
Kinetic Rates (Table 6) for Alluvium 
36 
Damkohler Number (Da) 
Cummulative Probability 
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 Dadet 
0.3 Daatt+Dadet 
0.2 Daatt 
0.1
0
0.01 1000000 10000000 100000 10000 1000 100 10 0.1 1
characterized by only coarser material, particularly in prominent flow pathways and subsurface 
channels. Therefore, Figure 10 compares the simulated retardation factor distribution for four 
different grain size distributions: (a) a range representing fine grain material, between 0.002 and 
0.02 cm ; (b) the range considered in Figure 8, between 0.02 cm and 0.11 cm; (c) the range for 
sand reported by Fulkerson (2001), which is between 0.005 and 0.2 cm; and (d) a range 
representing very coarse grains ranging only from 0.2 cm and 2.0 cm. Clearly, the distribution of 
retardation factors is sensitive to the grain-size distribution. Yet, the range for all cases is still 
large, spanning between 5 and 7 orders of magnitude, thus indicating significant uncertainty in 
the process due to the lack of site-specific data. The range used in Figure 8 and the range from 
Fulkerson (2001) lead to very similar distributions, with the Bechtel range providing slightly 
smaller retardation factors due to the coarser grains. However, a detailed distribution accounting 
for exact percentages of each subrange in the grain-size distribution, including the silts and clays, 
might actually lead to increases in R. Such an analysis is only warranted once the Yucca 
Mountain alluvium grain-size distributions have been documented. 
One must remember that this is strictly a theoretical assessment of the colloid filtration process. 
Reduction in uncertainty will require detailed measurements of actual colloid transport process in 
the alluvial material of interest with field and laboratory experiments coupled with additional 
sensitivity analyses. Although case (b) described above may be plausible, it is conservative in 
the absence of site-specific data to assume (Table 2) the grain-size distribution of case (a) 
because it leads to lower retardation factors than case (b). Another parameter that needs to be 
better assessed for this system is the collosion efficienty factor, alpha. Although it has much less 
impact on the resulting retardation factor distribution, its range is acquired strictly from other 
systems in the current analysis. 
November 2001 37 ANL-NBS-HS-000031 REV 00 ICN 01
1 
0.9 
0.8 
0.7 
0.6 
0.5 
0.4 
0.3 
0.2 
0.1
0 
10 1 
DTNs: SN0004T0571599.004 (for porosity), SN0004T0501600.004 (for flux), LL991109751021.094 (for colloid size). 
Other sources: Marshall et al. (1996), Fulkerson (2001), and Bechtel (1998) (for grain size), Harvey and 
Garabedian (1991) (for alpha), and Bales et al. (1997) and Table 6 of this report (for detachment rate). 
Output Data – DTN LA0004AW12213S.001. 
NOTES: PA range is the original distribution of retardation factors developed in this study and used by Performance 
Assessment. Bechtel range is the grain-size range for sand reported by Fulkerson (2001). For the extremely 
coarse grains, D, Equations 5-6 may not be appropriate and, hence, Equation 4 may not properly represent 
colloid retardation due to attachment for this extreme size of grains. However, there is no evidence to 
suggest that such coarse grains are representative of the Yucca Mountain alluvium. 
Figure 10. Comparing GoldSim Simulations of Retardation Factors for Four Different Grain-Size 
Diameter Distributions 
A 
A)Silt, very fine, and fine sand [0.002-0.02 cm] 
B) PA range [0.02 - 0.11 cm] 
C) Bechtel range [0.005 - 0.2 cm] 
D) Very coarse sand and gravel [0.2 - 2cm] 
6.2.6 Conceptual Model Summary: Alluvial Colloid Retardation Factors 
Conceptually, this model links attachment rates and detachment rates of colloids together to yield 
a colloid retardation factor in alluvium. For large enough time and space scales, the ratio of 
attachment rates to detachment rates added to one yields an equilibrium retardation factor that 
approximates the processes that inhibit migration of colloids in alluvial material. This model 
bases the attachment rates on classic colloid filtration theory, which predicts the attachment rate 
ANL-NBS-HS-000031 REV 00 ICN 01 
Cumulative Probability 
D 
B 
C 
100 1000000 100000 10000 1000 
R 
November 2001 38
using such measurable quantities as colloid size, grain size, water flux, colloid density, etc. 
(Equation 6). Because all of these quantities are variable, the distribution of attachment rates is 
created by drawing on distributions of all the quantities. Similarly, the detachment rate is 
represented with a distribution rather than a fixed value. Finally, the retardation factor 
distribution is computed by drawing from the distribution of attachment and detachment rates. 
6.2.7 Model Summary: Alluvial Colloid Retardation Factors 
The model in this portion of the AMR represents the processes associated with retardation 
factors for alluvial colloid transport to be used in other analyses such as Yucca Mountain PA. 
The model represents the retardation factor as a function of attachment rate of colloids to and 
detachment rate of colloids from immobile grains. The attachment rates are estimated with 
classic colloid filtration theory. Such theory does not provide for evaluation of detachment rates. 
Therefore, detachment rates are approximated based on detachment rates from fracture surfaces 
(this AMR) and from field observations in a sand aquifer at another location (Bales et al. 1997). 
Attachment rates are uncertain due to the range of uncertainty in the input parameters used in the 
model. Therefore, a distribution of attachment rates is developed via Monte Carlo simulation, 
which samples the ranges of uncertainty in all parameters. Similarly, observed detachment rates 
are uncertain. The model linking attachment rates and detachment rates to retardation factors 
draws from the distributions of attachment and detachment rates and produces a distribution of 
retardation factors. This distribution, then, represents the range in uncertainty of alluvial colloid 
retardation factors. It is in a form such that PA calculations can draw from it to capture the 
uncertainty in alluvial colloid retardation factors. 
6.2.8 Model Validation Summary: Alluvial Colloid Transport 
The development of retardation factors for alluvial colloid transport in this AMR is classified as 
a model because a set of equations are used to represent the physical and chemical phenomena 
associated with such retardation. It is not a predictive process model in that it is not used in this 
AMR to predict travel times and concentrations of colloids or radionuclides. Rather, it is a 
model specifically for estimating a transport parameter based on system properties. This model 
cannot currently be validated against Yucca Mountain laboratory or field observations because 
those data have not yet been collected and analyzed. However, the equations used and, hence, 
the model by definition have been validated through technical review and publication in the open 
literature. Specifically, the equations used to estimate colloid filtration (Equations 5 and 6) are 
well established in the literature. Harvey and Garabedian (1991) used the theory in a field 
demonstration of colloid transport in a sand aquifer, demonstrating the appropriateness of the 
equations. Those equations are adopted in this analysis and utilized. The primary differences 
between the analysis of Harvey and Garabedian (1991) and this AMR are: (1) The input 
parameters are site specific (Equation 6); that is, the groundwater flux, grain sizes, colloid sizes, 
etc., are unique to the system under consideration. (2) We assume equilibrium conditions, which 
is a necessary condition for Yucca Mountain PA models. Thus, there is no accumulation of 
attached colloids due to irreversible (or very slow) kinetics. This assumption is tested and 
evaluated in Section 6.2.4 of this AMR and demonstrated to be either accurate or conservative. 
Until the appropriate data are collected and analyzed for the Yucca Mountain system, this model 
represents the best use of established, published theory for estimating retardation factors for 
colloids in alluvium. 
November 2001 39 ANL-NBS-HS-000031 REV 00 ICN 01
7. CONCLUSIONS 
This AMR addresses the retardation mechanisms associated with the transport of colloids in the 
fractured tuff and alluvial systems of the Yucca Mountain saturated zone. In the fractured tuff, 
tracer experiments under forced gradient conditions have utilized synthetic microspheres as 
analogs of contaminant-bearing colloids. Analysis of the breakthrough curves of these 
microspheres provides estimates of the ranges of retardation factors that may be associated with 
colloid migration over much larger distances and time scales than were sampled during the tracer 
tests. In this analysis, the kinetic attachment and detachment rates estimated for the microsphere 
tracer tests are abstracted into a single retardation factor. The validity of this abstraction is 
demonstrated by showing that (1) colloid transport predictions using the retardation factor and 
the full kinetic attachment and detachment rate expressions do not differ over almost the entire 
range of travel times expected for the fractured tuffs, and (2) the retardation factor provides 
conservative predictions for very short travel times when the kinetic and equilibrium predictions 
differ. 
Whereas field and laboratory studies of colloid and/or microsphere transport have been 
conducted for saturated fractured tuff, there are currently no similar studies that have been 
performed for the saturated alluvial system. Therefore, in this AMR, a theoretical model is 
employed to estimate the attachment of colloids onto immobile alluvial material. The model 
uses a well-established methodology that has been applied to colloid transport in other saturated 
porous media systems. The theory, however, does not supply a method for estimating the 
detachment of colloids from immobile porous media. Therefore, rates estimated for detachment 
from fractures as well as detachment rates estimated in another instrumented aquifer are used. 
The resulting estimated retardation factor distribution spans over six orders of magnitude, thus 
capturing both ends of the spectrum of possible retardation factors for the currently untested 
system. 
The PA method for computing saturated-zone colloid transport requires an equilibrium-based 
retardation factor. For both the fractured tuff and alluvial aquifer, the retardation factors have 
been estimated from kinetic rates of attachment and detachment of colloids onto immobile 
surfaces. However, this local equilibrium assumption may not be accurate when residence times 
are small relative to the rates of attachment and detachment of colloids from immobile surfaces. 
A simple Damkohler analysis shows that assuming local equilibrium is generally suitable for 
both the fractured tuff and the alluvial aquifer. There are a few cases in the fractured tuff when 
assuming local equilibrium leads to overestimates of the attachment rate of colloids to fracture 
surfaces, which, in turn, leads to nonconservative overestimation of retardation. However, for 
those cases, the retardation factors and the residence times are so small that the overestimation 
results, at most, in tens of years additional residence of colloids in the fracture tuff, which is an 
insignificant time period relative to the time scales impacting PA. 
Although this analysis represents a reasonable approach for estimating field-scale retardation 
factors for colloid transport, there are inherent uncertainties that should be considered when 
evaluating the confidence in results obtained with these parameters. For fractured tuff, this 
November 2001 40 ANL-NBS-HS-000031 REV 00 ICN 01
analysis used extensive field data. However, the data were collected under stressed conditions 
that differ from those associated with the ambient system. Namely, transport of colloids was 
measured under conditions involving forced groundwater gradients induced by well pumping. 
Further, the spatial scale of the field-scale experiment is only about 100 meters, and the 
measurements were conducted over a time period less than 1 year. The differences between the 
test scales and those considered for the entire system should always be considered when 
evaluating model confidence. Whereas an extensive set of field experiments were conducted for 
colloid transport in fractured tuff, such experiments have not yet been conducted for alluvial 
material. Therefore, determination of colloid retardation factors for alluvial material was 
dependent on an established theoretical model but estimated site-specific properties. The 
uncertainty in property parameters was accounted for with distributions which were then 
sampled in a Monte Carlo fashion for the calculation of a retardation factor distribution. The 
range associated with the resulting retardation factor distribution spans six orders of magnitude, 
indicating the uncertainty associated with site-specific retardation factors for Yucca Mountain 
alluvium. Although the shape of the distribution curve changes, depending on assumptions 
about input parameter ranges, the large range in the resulting distribution clearly identifies the 
uncertainty. However, the PA use of this distribution is appropriate. By sampling the entire 
distribution, PA calculations will yield a large range of results, thereby indicating the uncertainty 
with the alluvial retardation parameter. 
The data and model developed by this analysis are included in DTNs: LA0004AW12213S.001, 
LA0002PR831231.003, and LA9912PR831231.006. 
November 2001 41 ANL-NBS-HS-000031 REV 00 ICN 01
8. INPUTS AND REFERENCES 
8.1 DOCUMENTS CITED 
Bahr, J.M. and Rubin, J. 1987. "Direct Comparison of Kinetic and Local Equilibrium 
Formulations for Solute Transport Affected by Surface Reactions." Water Resources Research, 
23, (3), 438-452. Washington, D.C.: American Geophysical Union. TIC: 246894. 
Bales, R.C.; Li, S.; Yeh, T.-C.J.; Lenczewski, M.E.; and Gerba, C.P. 1997. "Bacteriophage and 
Microsphere Transport in Saturated Porous Media: Forced-Gradient Experiment at Bordon, 
Ontario." Water Resources Research, 33, (4), 639-648. Washington, D.C.: American 
Geophysical Union. TIC: 246892. 
Bechtel Nevada 1998. Data Interpretations. Volume 1 of Hydrogeologic Characterization of the 
Unsaturated Zone at the Area 3 Radioactive Waste Management Site. DOE/NV/11718-210. Las 
Vegas, Nevada: Bechtel Nevada. TIC: 247707. 
BSC (Bechtel SAIC Company) 2001a. Technical Work Plan for Saturated Zone Flow and 
Transport Modeling and Testing. TWP-NBS-MD-000001 REV 02. Las Vegas, Nevada: Bechtel 
SAIC Company. ACC: MOL.20010924.0269. 
BSC 2001b. Technical Work Plan for: Integrated Management of Technical Product Input 
Department. TWP-MGR-MD-000004 REV 02. Las Vegas, Nevada: Bechtel SAIC Company. 
ACC: MOL.20010814.0324. 
CRWMS M&O 1999a. Abstraction of Colloid-Facilitated Pu Transport Modeling, Rev. 01. ID: 
B2005; Activity: SPP1075. Work Direction and Planning Document. Las Vegas, Nevada: 
CRWMS M&O. ACC: MOL.19990707.0110. 
CRWMS M&O 1999b. Conduct of Performance Assessment. Activity Evaluation, September 
30, 1999. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.19991028.0092. 
CRWMS M&O 2000a. Colloid-Associated Radionuclide Concentration Limits: ANL. ANLEBS-
MD-000020 REV 00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20000329.1187. 
CRWMS M&O 2000b. Uncertainty Distribution for Stochastic Parameters. ANL-NBS-MD- 
000011 REV 00. Las Vegas, Nevada: CRWMS M&O. ACC: MOL.20000526.0328. 
DOE (U.S. Department of Energy) 2000. Quality Assurance Requirements and Description. 
DOE/RW-0333P, Rev. 10. Washington, D.C.: U.S. Department of Energy, Office of Civilian 
Radioactive Waste Management. ACC: MOL.20000427.0422. 
Dyer, J.R. 1999. "Revised Interim Guidance Pending Issuance of New U.S. Nuclear Regulatory 
Commission (NRC) Regulations (Revision 01, July 22, 1999), for Yucca Mountain, Nevada." 
Letter from Dr. J.R. Dyer (DOE/YMSCO) to Dr. D.R. Wilkins (CRWMS M&O), September 3, 
November 2001 42 ANL-NBS-HS-000031 REV 00 ICN 01
1999, OL&RC:SB-1714, with enclosure, "Interim Guidance Pending Issuance of New NRC 
Regulations for Yucca Mountain (Revision 01)." ACC: MOL.19990910.0079. 
Fulkerson, R. 2001. “Submittal of ‘Hydrogeologic Characterization of U-3at Collapse Zone: 
Data Report”. Email from R. Fulkerson to R. Wemheuer, November 28, 2001, with attachment. 
ACC: MOL.20011128.0105; MOL.20011128.0107. 
Harvey, R.W. and Garabedian, S.P. 1991. "Use of Colloid Filtration Theory in Modeling 
Movement of Bacteria Through a Contaminated Sandy Aquifer ." Environmental Science and 
Technology, 25, (1), 178-185. Washington, D.C.: American Chemical Society. TIC: 245733. 
Marshall, T.J.; Holmes, J.W.; and Rose, C.W. 1996. Soil Physics. 3rd Edition. Pages 207-212. 
New York, New York: Cambridge University Press. TIC: 246638. 
Reimus, P.W. 1995. The Use of Synthetic Colloids in Tracer Transport Experiments in 
Saturated Rock Fractures. Ph.D. dissertation. LA-13004-T. Los Alamos, New Mexico: Los 
Alamos National Laboratory. TIC: 240694. 
Reimus, P.W.; Adams, A.; Haga, M.J.; Humphrey, A.; Callahan, T.; Anghel, I.; and Counce, D. 
1999. Results and Interpretation of Hydraulic and Tracer Testing in the Prow Pass Tuff at the 
C-Holes. Milestone SP32E7M4. Los Alamos, New Mexico: Los Alamos National Laboratory. 
TIC: 246377. 
Valocchi, A.J. 1985. "Validity of the Local Equilibrium Assumption for Modeling Sorbing 
Solute Transport through Homogeneous Soils." Water Resources Research, 21, (6), 808-820. 
Washington , D.C.: American Geophysical Union. TIC: 223203. 
8.2 CODES, STANDARDS, REGULATIONS, AND PROCEDURES 
AP-2.14Q, REV 2, ICN 0. Review of Technical Products and Data. [Washington, D.C.]: U.S. 
Department of Energy, Office of Civilian Radioactive Waste Management. ACC: 
MOL.20010801.0316. 
AP-SI.1Q, Rev. 2, ICN 4. Software Management. OCRWM Procedure. Washington, D.C.: 
U.S. Department of Energy, Office of Civilian Radioactive Waste Management. ACC: 
MOL.200000223.0508. 
QAP-2-0, Rev. 5, ICN 1. Conduct of Activities. Las Vegas, Nevada: CRWMS M&O. ACC: 
MOL.19991109.0221. 
8.3 SOFTWARE 
Los Alamos National Laboratory. 1999. Software Code: RTA V1.1. V1.1. SUN. 10032-1.1- 
00. 
Sandia National Laboratory. 2001. Software Code: GoldSim. V6.03. 10344-6.03-00. 
November 2001 43 ANL-NBS-HS-000031 REV 00 ICN 01
8.4 SOURCE DATA, LISTED BY DATA TRACKING NUMBER 
GS010799992315.001. Injection and Production Flow Rates for the LANL ProwPass Test, 
11/30/98 Through 1/27/99. Submittal date: 07/25/2001. 
GS950408318523.001. Temperature, Thermal Conductivity, and Heat Flow Near Yucca 
Mountain, Nevada. Submittal date: 04/21/1995. 
LA0002JC831341.001. Depth Intervals and Bulk Densities of Alluviums. Submittal date: 
03/08/2000. 
LA0002PR831231.001. Bullfrog Reactive Tracer Test Data. Submittal date: 02/04/2000. 
LA0007PR831231.001. Bullfrog Reactive Tracer Test Data. Submittal date: 07/21/2000. 
LA9909PR831231.003. Interpretations of Bullfrog Reactive Tracer Test Data - Modeling Data. 
Submittal date: 09/02/1999. 
LA9909PR831231.005. Interpretations of Tracer Data - Modeling Data. Submittal date: 
09/02/1999. 
LAPR831231AQ99.001. Prow Pass Reactive Tracer Test Field Data. Submittal date: 
02/10/1999. 
LL000905312241.018. Data Associated with the Detection and Measurement of Colloids 
Recorded in Scientific Notebook 1644. Submittal date: 09/29/2000. 
LL991109751021.094. Data Associated with the Detection and Measurement of Colloids in 
Scientific Notebook SN 1644. Submittal date: 01/10/2000. 
MO0011DQRCWELC.001. Data Qualification Report for C-Well Core Used in Sorption, 
Mineral and Hydraulic Studies For Use On The Yucca Mountain Project. Submittal date: 
11/29/2000. 
MO0110BFROGREC.001. BullFrog Test Recirculation Flow Rate Data. Submittal date: 
10/17/2001. 
MO0111BFROGPRO.000 Bullfrog Test Production Rate Data in Tabular Format in SEP Table 
S01097_001. Submittal date: 11/08/2001. 
MO0111BFROGPRO.001 Bullfrog Test Production Rate Data in Tabular Format in SEP Table 
S97545_001. Submittal date: 11/08/2001. 
MO0111BFROGPRO.002. Bullfrog Test Production Rate Data in Tabular Format in SEP Table 
S97239_001. Submittal date: 11/08/2001. 
November 2001 44 ANL-NBS-HS-000031 REV 00 ICN 01
MO0111BFROGPRO.003. Bullfrog Test Production Rate Data in Graphical Format in Figure 
22 of MOL.19980122.0412. Submittal date: 11/08/2001. 
MO0111PROWPASS.000. Prow Pass Test Production and Recirculation Rate Data in Tabular 
Format. Submittal date: 11/08/2001. 
SN0004T0501600.004. Updated Results of the Base Case Saturated Zone (SZ) Flow and 
Transport Model. Submittal date: 04/10/2000. 
SN0004T0571599.004. Uncertainty Distributions for Stochastic Parameters Revision to Include 
New U Sorption Coefficients in the Alluvium and Supporting Electronic Files. Submittal date: 
04/10/2000. 
8.5 OUTPUT DATA, LISTED BY DATA TRACKING NUMBER 
LA0002PR831231.003. Probabilities from C-Wells Microsphere Data. Submittal date: 
02/17/2000. 
LA0004AW12213S.001. Input and Output Modeling Data for Goldsim Calculations Associated 
with AMR: ANL-NBS-HS-000031. Submittal date: 04/11/2000. 
LA9912PR831231.006. Simulations of Microsphere Tailing in Bullfrog and Prow Pass Tests. 
Submittal date: 12/14/1999. 
November 2001 45 ANL-NBS-HS-000031 REV 00 ICN 01
ANL-NBS-HS-000031 REV 00 ICN 01 46 November 2001