improbability of nuclear criticality when disposing …improbability of nuclear criticality when...
TRANSCRIPT
Improbability of Nuclear Criticality When Disposing of TransuranicWaste at the Waste Isolation Pilot Plant
Rob P. RechardLawrence C. SanchezChristine T, StockmanHolly R. Trellue
Nuclear Waste Management Programs CenterSandia National LaboratoriesAlbuquerque, NM 87185-0776
Rob P. RechardMS-0776Sandia National LaboratoriesP.O. BOX 5800Albuquerque, NM 87185-0776505-844-1761 (V)505-284-3964 (F)mrecha@sandia.$?ovPages: 41 (30 tex~ 11 references)Figures: 21Tables: 7
1 June 1,2000
.,, .... . . .r,y--- ----- , --- -.- .... - —- . ..
DISCLAIMER
This report was prepared as an account of work sponsoredby an agency of the United States Government. Neitherthe United States Government nor any agency thereof, norany of their employees, make any warranty, express orimplied, or assumes any legal liability or responsibility forthe accuracy, completeness, or usefulness of anyinformation, apparatus, product, or process disclosed, orrepresents that its use would not infringe privately ownedrights. Reference herein to any specific commercialproduct, process, or service by trade name, trademark,manufacturer, or otherwise does not necessarily constituteor imply its endorsement, recommendation, or favoring bythe United States Government or any agency thereof. Theviews and opinions of authors expressed herein do notnecessarily state or reflect those of the United StatesGovernment or any agency thereof.
DISCLAIMER
Portions of this document may be illegiblein electronic image products. Images areproduced from the best available originaldocument.
,
ABSTRACT
Nuclear criticality was eliminated from performance assessment calculations for the Waste Isolation Pilot Plant
(WIPP), a repository for waste contaminated with transuranic (TRU) radioisotopes, located in southeastern New
Mexico, based onarguments presented inthis article. Afierdisposal and following aninadvefient human inwusion
into thereposito~ (anevent that must reconsidered because ofstie~ regulations), there isno credible mechanism
to counteract the natural. tendency of the material to disperse during transport. physical constr~nts on concen~ting
fissile material include low initial solid concentration of fissile material, small mass of fissile material transported
over 10,000 yr, and insufficient physical compaction; hydrologic constraints include the limited amount of brine
available to transport fissile material. Geochemical constraints on concentrating the fissile radioisotopes include
lack of sufficient adsorption and conditions conducive to precipitation. Hence, the probability of nuclear criticality
is low. Furthermore, before a criticality would have the potential to affect human health-assuming that a criticality
could occur—it would have to either (1) degrade the ability of the repository to contain radioactive waste or (2)
produce significantly more radioisotopes than originally present. Neither of these situations can occur at the WIPP;
thus, the consequences of a criticality are also low.
Keywords: radioactive waste disposal, criticality, risk assessment, performance assessment, Waste Isolation
Pilot Plant, 40 CFR 191
lxc 192000
,( I
L INTRODUCTION
To certify the compliance of the Waste Isolation Pilot Plant (WIPP), an operating facility owned by the U.S.
Department of Energy (DOE) for the geologic disposal of wastes containing transuranic (TRU) radioisotopes, the
U.S. Environmental Protection Agency (EPA) required estimates of the disposal system’s future behavior under
stylized conditions. The estimates were prepared by means of models that captured essential features, events, and
processes (FEPs) of the disposal system. As in any surface facility that handles fissile material, a potential event for
a deep geologic disposal system for radioactive waste is a sustained, nuclear chain reaction (i.e., a criticality). This
article describes the reasons why nuclear criticality was omitted from consideration as a modeled event w. The
work reported here supported the 1996 WJPP performance assessment (.2,a as part of the Compliance Certification
Application (CCA) @ submitted to EPA in October 1996 and approved in May 1998 @).
The likelihood of assembling a critical mass at a location in or near a repository depends upon the geologic
processes within the disposal system and the ability of these processes to concentrate fissile radioisotopes so that
they meet criteria necessary for criticality. Concern about criticality in TRU waste has never been great because of
the low initial concentration of fissile material; yet, the approach used can be applied to repositories with other types
of waste with higher initial concentrations of fissile material. To thk end, the article illustrates the use of
fundamental concepts of nucIear criticality and geologic aspects of concentrating minerals that support eliminating
the criticality events.
I.A. Location of WIPP
The WIPP is located in southeastern New Mexico, 42 km east of Carlsbad, New Mexico (Fig. 1). Although an
abandoned salt mine near Lyons, Kansas, was initially examined, the DOE, at the invitation of New Mexico civic
leaders, settled in 1974 on characterizing the 600-m-thick salt beds in the Delaware Basin for disposal of its
radioactive waste. Five years later, in December 1979, Congress authorized the DOE to build at the selected site @.
Construction of the repository was essentially complete in 1988. The WIPP received its first shipment of TRU
radioactive waste in March 1999 ~.
LB. Performance Assessment
The process of assessing whether a radioactive waste disposal system such as the WIPP meets a set of
performance criteria is a performance assessment (PA). The WIPP disposal system is discussed in Section IL The
following sections describe the performance criteria and analysis process.
LB. 1. Performance Criteria
The U.S. Congress, in the Nuclear Waste Policy Act (NWPA) of 1982 (!3, established the policy that the present
generation would bear the political and financial costs of developing a permanent disposal option for nuclear waste.
In 1985,1 in response to the NWPA, the EPA promulgated the standard, 40 CFR 191 (LQ),which had been under
study since 1977 @). In the WIPP Lund Withdrawal Act of 1992 (U), Congress directed that the EPA implement
this standard for the WIPP Project. In response, EPA promulgated 40 CFR 194 (.Q).
Neither 40 CFR 191 nor 40 CFR 194 provides specific guidance regarding the occurrence of criticality after
closure of the WIPP repository. Rather, risks associated with a critical condition are evaluated under the general
‘ The standard was remandedby the courts shortly thereafter but was repromulgated with only minor changes in 1993 Q).
3 June 1,2000
1 ,
provisions of the standard. The Individual Protection Requirement requires that the mean individual dose remain
below 15 mSv/yr. The Ground Water Protection Requirement requires that the activity from alpha-decaying
radioisotopes remain below 15 pCi/L The Containment Requirement requires an evaluation of probabilities of
c’umttlative release at the disposal system boundary (actually a complementary cumulative distribution function
[CCDFI). A distribution of release is required because uncertainty from various sources must be included in the
analysis,z where the uncertainty is usually expressed as a probability distribution (see, for example, Section IV.D on
volubility uncertainty).
/.9.2. FEP Selection and Screening
The process of performing a PA answers three basic questions (15, 16. 17): What unwanted features, events,
and processes3 (FEPs) or scenarios of these FEPs may occur? What is the probability of each FEP or scenario?
What are the consequences in terms of the performance criteria of each FEP or scenario? The first question is
answered through a scenario development process. The second and third questions are answered through a
modeling process. Hence, PA is intimately tied to the process of building scientifically plausible mathematical
models (18, 17. 14. 19). Several iterations of the PA process may be necessary to eliminate those FEPs or scenarios
with negligible influence and to improve the modeling of retained FEPs.
The formal selection and screening of FEPs for inclusion in modeling is an impontant step in PA and one of the
aspects that sets PA apart from typical scientific modeling or engineering analysis. The EPA set the guiding
philosophy for selection of FEPs, including the regulatory period, the type of natural and anthropogenic phenomena
to be considered, and criteria for omitting FEPs.
Regulatory Period. The EPA’s goal for the disposal standard was to provide protection for as long as the
wastes presented an unacceptable risk QQ, pp. 64-68). The EPA selected 10,000 yr as the period of regulation.4
Human Intrusion. The standard, 40 CFR 191, and the implementing regulation, 40 CFR 194, require the DOE
to demonstrate that the WIPP will comply with the performance criteria even after humans unknowingly intrude into
the repository with an exploratory drill hole using present day technology. The requirement by the EPA to address
human intrusion defeats, to a large degree, the advantage of the nearly ideal properties of salt in containing the
radioactive waste and focuses instead on other aspects of the WIPP disposal system, such as adsorption of
radioisotopes and the existence of nonpotable water in an aquifer above the salt beds. Hence, it is important to
emphasize that without inadvertent intrusion, the probability of criticality is practically zero; only with intrusion do
the arguments require more sophistication.
FEP Screening Criteria. In Appendix C of 40 CFR 191 G!) and in 40 CFR 194.32 (Q), the EPA allows
omission of categories of FEPs with probabilities of occurrence less than 10+ in 10,000 yr.s
2 Bemuse the performance criteria are probabitisric in the United Stat% a performance assessment (PA) and a probatitistic risk assessment(PRA) for a nuclear reactor we simikwin concept (Q ~3 br this context, a feature is an aspect or condition of the disposat system, an event is a short-term natural or anrhropogenic phenomenon, and a~ocess isalong-term natumtphenomenon (i.e., a phenomenon that occurs over a signific@ portion of the regulatory period).
The EPA stated tha~ because 10,000 Y is a short period of time geologically, changes in geological conditions were expected to be smatl @,p. 67). Time periods of 100and 1000 yr were rejected because few health effects were p~lcted for that rime span-many radioisotopes wouldstill be in transit toward the accessible environment and the probability of disruptive events would be low. Periods of time longer than 10,000 yrwere rejected because predictions of geological changes would not be rctiable.s For enghwred nuclear facilities, the Us. Nuclear Regulatory Co~.ssion (NRC) at onetimeattempted to define an incredible event as anevent with a frequencyof occurrence less than 104/yr. The EPA chose a more conservative frequency of 108/yr.
4 June 1,2000
LB.3. Human Intrusion Scenarios Evaluated in 1996 PA
Based on guidance in Appendix C of 40 CFR 191 and 40 CFR 194, human intrusion was considered by
assuming exploratory drilling for oil and gas deposits using present-day technology near the WIPP. In such a
situation, radioisotopes may be released in different ways and at different times. The short-term result of intrusion is
entrainment of wastes into the drilling fluid and their immediate release at the surface. At much later times, releases
through groundwater flow into the overlying Culebra may occur when brine flow and the volubility of the
radioisotopes are high. The later potential long-term releases were grouped into two categories based on whether a
pressurized brine reservoir in the Castile Formation (described in Section II) was intersected, and then further
grouped into two subcategories based on the time of intrusion. A third category was developed from the first two
categories. The results discussed in this article are primarily from scenario S5 (Table 1).6
In scenario S5, a human exploring for minerals drills through the repository. Later, the driller plugs the
borehole using concrete (present-day technology) and abandons it. Releases over the long term can occur from
selective degradation of the concrete and migration of radioisotopes to the brine aquifer in the Culebra (Fig. 2). As
suggested by regulatory guidance U.Q, ~, D, the concrete plug is conservatively assumed to degrade to the
properties of sand, without intrusion by salt or the precipitation of salts in the brine over the 10,000-yr regulatory
period. The assumption of no closure of the borehole (either through salt creep or precipitation of minerals from
solution) is very conservative over the 10,000-yr regulatory period; given the high likelihood of salt creep and salt
precipitation in a borehole, the assumption of no closure of the borehole for periods longer than 10,000 yr becomes
unrealistic.
I.C. Overall Approach
The performance criterion (i.e., CCDF) includes both a probability component and consequence component,
and the approach used herein is to use arguments of both these elements to demonstrate that criticality was
appropriately eliminated from the 1996 PA.
I.C. 1. Probability
Support for the low-probability argument (and the low-consequence argument discussed later) depends upon
constraints from two types of modeling. Nuclear criticality modeling is used to develop physical constraints on
fissile mass and geochemical constraints on fissile concentration. The nuclear criticality modeling is discussed
elsewhere &l, L m. The results, based on the WIPP disposal system, are summarized in Section III. The second
type of modeling, geophysical modeling is used to evaluate the feasibility of exceeding bounds on fissile
concentrations by examining whether physical, hydrologic, or geochemical constraints exist (Fig. 3). A description
of these geophysical processes is the primary focus of the remainder of this article. Specifically, the probability of
criticality is the conditional probability times the probabilities that physical, hydrologic, and geochemical constraints
are absent. The qualitative evaluation of these probabilities is used to organize the discussion in the article. That is,
P(C) = P{g] ● P(h) ● P(C I g n h n C) ● P{c}, where P{C} is the probability of a criticality even~ P{g) is the
probability of no geometric or other physical constraints, as discussed in Section IV; P{h} is the probability of no
6 tn addition to using scenarios, three replications of the entire calculation set, each with a different starting random number. were sun to evaluatepotential variation in results in the 1996 PA. Ilk article uses the first repticate, RI. For a complete description of the calculations, see ~ or z
5 June 1,2000
I ,
hydrologic constraints (also discussed in Section IV); P(C I g fl h n c) is the conditional probability of criticality
given no obvious physical, hydrologic, or geochemical constraints (Section V); and P{c} is the probability of no
geochemical constraints to criticality (Section Vi) (Fig. 3).
1.C.2. Consequences
In the final section, Section VII, two categories of consequences are evaluated through analogy with
anthro~ogexlic criticality accidentx (1) the potential increase in fission products and (2) damage to the disposal
system.
1.C.3 Use of General and Specific Arguments
When presenting arguments in Sections IV through VII, the arguments are differentiated by location, i.e., the
engineered barrier (repository) and the natural barrier (i.e., primarily the Culebra aquifer described in Section II),
because specific reasons for the absence of geophysical mechanisms ofien differ for each location. Furthermore, the
authors first discuss general constraints from geophysical modeling that argue against criticality at each location.
Although this approach does not always provide a definitive resultj it can be useful in conveying the arguments to a
wide audience. The authors then present results of the 1996 PA Q, 3 to confirm that the simulated conditions are
not conducive to criticality.
Il. BACKGROUND
11.A. Previous Studies of Criticality in Repositories
Prior to establishing deep geologic disposal of radioactive waste as United States policy, several studies were
conducted as to whether criticality was credible in a repository. Allen @ reported that the critical mass of
spherically shaped fissile material was surprisingly small; the masses of fissile material in the proposed containers of
defense high-level waste7 (HLW) either were only slightly smaller or, in a few cases, were of similar size to these
limits. Also in 1978, Winchester w hypothesized a mechanism to segregate 243Amfrom other actinides through
preferential adsorption; the 243Amcould then decay to ‘9Np and eventually to ‘9Pu. In 1979, Clayton u
suggested a positive feedback mechanism for a critical event in soil, based on calculations by Carter @j), who had
evaluated the potential for criticality as part of a safety analysis of radioactive waste disposal in trenches at the
Hanford reservation in Washington. To dismiss these hypotheses, Brookins @ reported on geochemical
constraints that prevented critical masses from accumulating in geologic repositories. Also in 1983, Stratton M
dismissed the allegations that a nuclear explosion had occurred at a waste disposal site in the Ural Mountains in
Russia..
When the option to place HLW and spent nuclear fuel (SNF) in the WJPP was blocked by Congress in 1979 @,
the concern for criticality diminished because of the low initial concentrations of fissile material in TRU waste.
Nonetheless, criticality has been listed as an event for consideration at the WIPP since 1979 in work supporting the
Environmental Impact Statement 031S)(29, 30,31, m. Based in part on studies of the WIPP, the International Atomic
Energy Agency (IAEA) has listed criticality for consideration since 1981 in its guidance to member countries that
desire to site repositories for radioactive waste @). The DOE conducted traditional nuclear criticality safety studies on
7Prior to Congressional authorization of the WIPP in December 1979 (Q), the DOE considend the site for disposal of HLW in addition to TRU
the storage and initial emplacement of the drums in the repository in the early 1980s (34, 35) and concluded that the
drum storage array in the WIPP would be subcritical. The same conclusion was independently ccmfirmed by the State
of New Mexico Environmental Evaluation Group (EEG), although the EEG analysis disagreed with some aspects of
the DOE analysis (36, 37). The EEG anaIysis also explorcd the possibility of dissolution, transpo~ and concentration
of the fissile material in a dolomite rock with 10% porosity. With assumed concentrations in the rock of 27.6 mM (6.6
kg/m3)’for %% and 4.7 mM (1.1 kg/m3) for ‘3U, a fissile mixture that was greater than 0.5 m tilck was supercriticals
~flz 1) @). However, the discussion noted that “the consequences to the accessible environment should not be very
high” @). The study also cited analogous aqueous criticality accidents reported by Stratton (30, 40,41,43 and the
natural reactors in the Oklo ore deposit in Gabon, Africa (43,42).
The preliminary screening of FEPs performed for the WIPP by Hunter @ and Galson and Swift @) retained
nuclear criticality for more thorough investigation. AIso, news articIes in 1995 @@ drew attention to speculative
scenarios proposed by Bowman and Venneri QZ) with regard to an atomic explosion occurring in the potential tuff
repository at Yucca Mountain, Nevada, speculation that probably received more attention than it deserved (48J.
Soon after, critics argued against the possibility of such a large energy release both qualitatively (49, 50) and
quantitatively @1. 52.53. 21). This article describes the difficulty of creating conditions conducive to any type of
criticality+xplosive or otherwise-with TRU waste.
ILB. Description of WIPP Disposal System
ILB. 1. Geologic Characteristics
The geology of southeastern New Mexico has been of great interest because of the potash and hydrocarbon
resources in the area and because of the academically interesting exposures of ancient reefs. Hence, extensive
literature exists on the general stratigraphy (see, for example, references in ~, Chapter 2; ~ 3). The lowest strata
mentioned here is the Bell Canyon Formation, which is composed of up to 330 m (1000 ft) of fine-grained sandstone
and siltstone with thin beds of carbonate and clay that contain hydrocarbon. The Bell Canyon Formation isolates the
overlying strata from deeper strata.
Castile Formation. Overlying the Bell Canyon Formation is the Castile Formation (hereafter shortened to the
Castile), which is found only on the interior of the Capitatt limestone reef (Fig. 1). The 500-m-tMck Castile consists
of three anhydrite (CaSOA) members and two interspersed halite (NaCl) members (Fig. 4). Within the kmd-
withdrawal boundary, a pressurized brineg reservoir has been intersected in the fractured Anhydrite III (uppermost)
layer of the Castile by one well (WIPP-12) and, outside the boundary, by 11 out of -100 nearby deep wells &l,
Figure 3-26) (28 out of 3406 wells in the New Mexico portion of the Delaware Basin E, Appendix DEL, Section
7.5]). As described earlier, the analysis of the WIPP assumes that a pressurized brine reservoir exists in the Castile
beneath a portion of the repository for the next 10,000 yr.
Salado Formation. The Salado Formation (hereafter shortened to the Salado), which overlays the Castile,
extends far beyond the Capitan reef. The 600-m-thick Salado hosts the repository about 655 m below the surface.
8 If n stray neutrons are introduced into a region of fissile material, the total number of neutrons that appear after r generations is n~, where k isthe multiplication h in rd systems). Three possible situations exist fork. k-d, subcritical; k=l, criticat; bl, supem-hicd.9 Herein brine refers to an wcous solution with total dissolved solids (TDS) greater than 30 kg/m3. For comparison, brackish water refers tosolutions with TDS between 3 rutd30 kglm3;fresh water refers to solutions with TDS less than 3 kg/m3.
7 June 1,2000
Near the repository, the Salado consists of nearly horizontal (cl 0 dip) halite and occasional interbeds of minerals
such as clay and anhydrites of the Late Permian Period (-255 million year old [255 Ma]) (Fig. 4).
Culebra Dolomite Member of Rustler Formation. The Rustler Formation overlies the Salado (Fig. 4). The
Culebra Dolomite Member of the Rustler Formation (hereafter shortened to the Culebra), consists of dolomite and
dolomitic limestone. Although radioisotope release via groundwater is unlikely overall Q, ~, the Culebra is the
most likely pathway relative to other pathways for lateral transport of radioisotopes because it is the most permeable
stratigraphic unit near the repository @lg. 2).
Within the WIPP Project, the Culebra has been divided into four units @). The uppermost unit, Culebra
Unit 1, averages 3.0 m in thickness. The small number of fractures tends to occur along bedding planes. The middle
Culebra Units, 2 and 3, are fairly similar except for the extent of fracturing. The fractures in both units typically
extend less than 5 cm and connect numerous vugs. Originally, the vugs were anhydrite pockets that hydrated to
gypsum during sedimentation; subsequent dissolution of gypsum left the vugs. The fractures in both units are either
open or gypsum-filled, with apertures up to 0.2 cm wide. Because of more extensive tlacturing, intact pieces of core
from Culebra Unit 3 are rare. Culebra Unit 2 is about 1.6 m thick Culebra Unit 3 is about 1.2 m thick. The
lowermost Culebra Unit 4 is typically 1.5 m thick near the repository. Small, open microvugs (<3 mm) are
common, but occasionally, large vugs and gypsum nodules (up to 8 cm in diameter) occur near the bottom of the
unit. Mercer and Orr @) reported that based on 1311tracer tests at H-3, all of the flow came tlom the lower portion
of the Culebra. Tracer tests in 1996 at H-19 confirmed that Culebra Unit 1 does not contribute to solute transport.
Therefore, the hydrologic effective thickness of Culebra is taken as -4 m (7 m minus the 3 m for Culebra Unit 1)
w.
1/.8,2. Transuranic Waste
Waste planned for disposal at the WIPP repository consists of a broad variety of materials, including inorganic
(e.g., iron and aluminum alloys, equipment, concrete, glass, firebrick, ceramics), organics (e.g., celhdosics, such as
paper, cardboard, laboratory tissues, wood, cloth, rubber, plastics), solidified materials (e.g., waste water treatment
sludge, cemented liquid waste, inorganic particles and soils), and solvents generated during the production of
nuclear weapons. Most of these wastes have been contaminated by alpha-emitting TRU radioisotopes. A small
portion of TRU are fragments of SNF that were examined in the laboratory. To be classified as TRU waste in the
United States, the average activity per mass in the container must be over 3.7 x 106 Bq/kg (100 nCi/g) from
radioisotopes with half-lives greater than 20 yr.
General Categories of TRU Waste. Two general types of TRU waste are (1) contact-handled transuranic (CH-
TRU) waste (-1.68 x 105m3 or -806,000 55-gal drums), which is TRU waste with an external dose rate of less than
5.6 x 10-7 Sv/s (200 mrernhr), and (2) remotely handled transuranic (RH-TRU) waste (7080 m3 or 4%), such as
fragments of SNF, which exceeds an external dose rate of 5.6 x 10‘7 SVIS but is less than 2.8 x 10-3 Svls
(1000 rernk). For both types, the final forms of the waste are grouped into 11 categories @lg. 5). Most of the
volume is in the Heterogeneous Waste catego~, the second largest volume category is Uncategorized Metal ~, ~,
Figure 8).
8 June 1,2000
,,-.,..,., ... ->-.:.-!.... - ,’. - . ;!-.,..:.V.*Z-Z-.,...--.-.>i“..,., T: .r -.. >:.X%7-.-Y!.!..,:...! ....-.,.~- .....Z.- - - --?-.TT..,.-
,
TRU F’issile Mass. The projected initial total mass of radioisotopes in the WIPP repository is -180 Mg. Of
“ ‘SU placed in the waste by Oak Ridge National Laboratories (Fig. 5) and elsewhere tothese, 80% (150 Mg) M
prevent criticality during storage. The masses of the two most important fissile materials, ‘5U and ‘9Pu, are 8 and
13 Mg, respectively (Table II). Note that the average enrichment of uranium at emplacement is 570. Over
10,000 yr, the enrichment increases slightly to 7% because of the decay of ‘9Pu. The initial average enrichment of
the combination of plutonium and uranium (in Pu fissile gram equivalents ~GEs]) at emplacement is 14%.
Containers. CH-TRU waste is shipped in standard 55-gal steel drums (0.89-m high, 0.60-m diameter) or steel
boxes (0.94 m high, 1.3 m wide, 1.8 m long). The drums are packaged in groups of seven for transportation and
disposal. The reference design for RH-TRU waste is a 6.35-mm-thick steel cylinder with an outside diameter of
0.65 m and length of 3 m. Neither type of container has any special feature that acts as an engineered barrier and so
the containers are considered part of the waste.
11.B.3. Design of the RepositoW
The WIPP repository is a 1.5 x 105-m2(38-acre) facility constructed in a single stratigraphic unit of the Salado.
The excavated disposal region is 4.35x 103m and is divided into eight panels plus two equivalent panels composed
of the central connecting drifts. All the openings are rectangular in cross section and 4 m high. The ‘7-pack” of
drums or individual boxes of waste are stacked three high and six wide across the rooms and connecting drifts
(Fig. 6). The design volume for the waste in the disposal region is 1.75 x 105 m3 or about 40% of the excavated
volume. A portion of the remaining volume will be backfilled with salt and magnesium oxide (MgO) such that the
average inhial porosity (including the porosity in the waste) is about 8590.
Ill. PHYSICAL CONSTRAINTS ON CRITICAL CONCENTRATION AND MASS
A critical condition depends not only on the quantity of fissile material but also on its concentration, shape, and
any other material mixed with or surrounding the fissiIe material. In addition, the temperature of the material has an
important influence. As demonstrated by Clayton m, setting limits for a heterogeneous mixture with specific
shapes and masses of tissile and other material is particularly complex. However, standard limits below which
criticality is impossible do exist for homogeneousmixtures of fissile material, and these limits can provide guidance
in more complex situations.
111.A.Experimental Critical Limits for Uranium- and Plutonium-Water Mixtures
In describing the possibilities of homogeneous mixtures of fissile material in the literature, the behavior of
fissile material immersed in water is most often presented as the fissile mass versus the fissile concentration @lg. 7).
For criticality to occur with uranium, the uranium mass and solid concentration must lie to the right and above the
‘9Pu/H@ mixture. For criticality to occur, the mass ofcurve. The same basic shape is observed for a 100 70wt
100 %wt ~9Pu must be greater than a 0.5 kg and the ‘9Pu solid concentration xmst be greater than 7 kg/m3 (Fig. 7b).
The minimum critical mass (or minimum critical volume) of ‘9Pu or ‘5U is occasionally used in arguments in
this reticle. However, the masses and volumes of fissile material collected in a certain region of geologic material
are dependent on time (e.g., flow rates) and therefore the regulatory period, unless a geometrkxd constraint on the
maximum mass or volume exists. The concentration limit, on the other hand, can often be more easily compared to
geologic processes such as dissolution, adsorption, and precipitation.
9 June 1,2000
r
111.B.Calculational Method and Data for Geologic Material
///.9. 1. Computational Tool
Although numerous criticality experiments have been performed in ideal material, criticality experiments with
common geologic material have not. Consequently, the work presented here is based on calculations with MCNPm
(Monte Carlo code for solving Neutron and Photon transport equations) (Los Alamos Version 4A, RS ICC version
CCC-200) m~” For ‘5U-HZ0 and ‘9Pu-H20 mixtures, the calculated criticality curves (using 105 samples to
evaluate ~ff) compare quite well with results of experiments compiled by Paxton et al. @) and Clayton and
Reardon ~ (Fig. 7).
///.9.2. Deposited Forms of Plutonium and Uranium
For the criticality calculations described herein, the deposited form of plutonium is resumed to be Pu02 (pg =
11580 kg/m3); the deposited form of uranium is assumed to be U02C03 (RuherfOrdine) (pg = 5724 Wm3), because
it is the thermodynamically stable form in a carbonate solution. Although the type of plutonium or uranium mineral
greatly influences volubility, as shown later, the mineral form has only a small influence on criticality limits.
///.9.3. Composition of Culebra Brine, Culebra Dolomite, and Ok/o Sandstone
In the criticality calculations, the elemental composition of the Culebra, Salado, and Castile brine used here is
that defined by Papenguth @) (Fig. 8). The intergranuhr Salado brine has substantially more magnesium and
potassium than the Castile brine. The element composition of Ctdebra dolomite used here is that defined by Siegel
et al. @ and Sewards et al. @) (Table HI).
Because it can serve as an analogue, the composition of the sandstone that hosts the Oklo uranium ore deposit in
Gabon, Africa, is also used herein (Section V). The deposit contains at least 16 natural reactors that operated
intermittently starting about 1.97 billion years ago (Ga) (QI, 1, 42, 65). The Oklo sandstone is high in silica
(84 %wt) (Fig. 9).
111.9.4. Porosity of Culebra Dolomite
Within the WIPP land-withdrawal boundary (Fig. 1), the porosity within the dolomite consists of intergranular
porosity, fractures, and vugs. Since 1991, the range for the intact matrix porosity in PA calculations has been set
between 0.10 and 0.25 @, Table 2.6-3). The median value assumed was 0.16 W, 66, %me 2.6-10
Prior to 1996, PA calculations assumed a range of 0.0001 to 0.01 @, p. 11-30) as the advective fracture
porosity, i.e., porosity through which fluid is conducted, based on an order-of-magnitude bracket around the tracer
test results at H-3 and H-11 (@. Results from the tracer tests at H-19, which estimated the advective porosity at
0.04 @), were used for the 1996 PA.
111.C.Calculated Criticality Limits for Geologic Media
In general, the addition of geologic media, such as Culebra dolomite, to the fissile/water binary system
substantially increases the mass and slightly decreases the solid concentration of Iissile material necessary to go
,10h MCNP the dist~ce between interactions, the fissions that occur, the loss by capt~re, or leakage am characteriti by Parameters such as,the reaction cress-section of the atoms of each materiat, the mean free path lengths between interactions, the distribution describing scattering,and the distribution of neutron energy. Although considered state-of-the-art because of its use of pointwise, continuous-energy cress-sections,MCNP cannot easily determine the point at which ~ is equrd to one. Rather, with MCNWs, the user must select parameters through trial-assd-ermr. Therefore, a prc-processor was developed for MCNP~ to rapidly generate input files and a post-pmccssor to estimate criticalconditions through interpolation of results. Sanchez et al. ~, Vol. 2, pp. 1-14) describe this analysis system in mote detail.
10 June 1,2000
critical. For a mixture of ‘9Pu02/Culebra dolomite (porosity [$]=16%) 100% brine saturated, the minimum
concentration limit for criticality is -3 kg/m3 (1250 ppm) @lg. 10a) and the minimum mass limit is 2.2 kg (based on
properties reported in Figs. 8 and 9 and Table III) Q, m. The concentration limit decreased about a factor of 2.3,
and the mass limit increased by a factor of 4.4 from those values shown for a Pu/HzO binary system (13g. 7). The
concentration limit for criticality increases to -10 kg/m3 when either Salado or Castile brine is the fluid, because of
the added chlorine and boron content @lg. 10b). The critical concentration for a ‘@u/halite/brine homogeneous
mixture is -53 kg/m3 and the minimum critical mass is 720 kg (Fig. 10a).
For uranium, as Rutherfordine (U02C03) (5% ‘5U), the limit concenmation is 9 kg/m3 and the minimum critical
mass is 50 kg Q., u @lg. 11a). Comparison of the criticality limit (left asymptote) for 3.68 %wt ‘5U and 93.2 %wt
‘5U shows that although there is a large change in required fissile mass to go critical (500 kg vs. 9 kg), there is only
a small change in required fissile concentration (9 kg/m3 versus 5 kg/m3)11@lg. 1lb). Furthermore, to go critical,
the concentration of uranium (1250 ppm) must reach levels that are considered to be economically mineable ore
bodies when located near the surface (-1000 ppm) and thus current knowledge about the formation of ore bodies is
discussed in Section V.
For a system composed of three components (e.g., uranium, water, Culebra dolomite), the critical curve of
fissile mass versus fissile concentration can have an asymptote at the right as well as the left of where the porosity of
the geologic material limits the possible mass of water and fissile material. This effect is seen for uranium in
dolomite with 16% porosity (Fig. 10a) and sandstone with 10’%porosity (Fig. 1lb).
111.D.Criticality Limits for Adsorption on Rust
Fksile material can adsorb onto corrosion products of the drums and metals in the waste within the repository,
most notably rust. In addition, adsorption on mst represents a reasonable maximum bound for adsorption on other
materials within the repository and elsewhere. Because goethite (et-FeOOH) represents the most likely form of rust
over the 10,000-yr regulatory period (as described in Section VI.A.3), the critical limits for homogeneous mixtures
of metallic uranium and plutonium with goethite were evaluated for hemispherical shapes12(l%g. 12). The porosity
was set at 20%. The critical concentration in pure water is about 45 kg/m3 and 20 kg/m3 for fissile ‘SU and ‘9Pu,
respectively.
111.E.Criticality Limits for Planar Shapes in Culebra
For a planar mixture of ‘9Pu02/Culebra dolomite (+ = 16%) and brine, the concentration limit for criticality is
-5 kg/m3 (Fig. 13), which is very similar to that for a spherical shape (i.e., 3 kg/m3, Fig. 10).
IV. GEOMETRIC AND HYDROLOGIC CONSTRAINTS ON CRITICALITY
IV.A. Compaction in Repository
The CH-TRU waste to be emplaced at the WIPP contains very low concentrations of fissile material (primarily
‘9Pu); thus, the possibility of criticality is extremely remote prior to closure. Just as important, the possibility of
criticality is also remote after closure because criticality requires that this emplaced fissile mass be substantially
11 me 3.68% Wt ‘U efichment is tie estimated enrichment when the OkIo naturaf reactors in sandstone we~ intefittentlY opemting. m
described further in Section V.n Sitim.nch (Uff,which is Vew si~lm to s~dstone m shown in Fig. 1lb, was used as the geologic m~lum so that tie ~Ysis could ~so be
used elsewhere.
11 June 1,2000
concentrated. Even when the possibility of salt creeping or precipitating into pores withh the waste is neglected, the
solid concentration is below the concentration limit required to go critical. To elaborate, the solid concentration
‘9Pu rust (goethite), and water will not gobelow which an infinite volume of a homogeneous mixture of pure ,
critical is 20 kg/m3 (Fig. 12). This limit is 25 times larger than the emplaced density of 0.79 kg/m3 if the
transportation limit of 0.2 kg of FGE ~9Pu per container is considered and assuming an initial porosity (~) of 0.848.
This density could be increased somewhat through compaction however, assuming compaction to an average
porosity of 0.08 (Fig. 14), without any salt creep into the waste layer, the fissile mass bulk density (pr) increases by
on]y a factor of 6 to 4.7 kglm3 (i.e., pipj = (1 – $r) / (1 – $j) = 6). Also, the combined mass of three containers is
0.6 kg of FGE ‘9Pu while the critical limit is 6 kg (Fig. 12).
IV.B. Depositional Space in Marker Beds, Disturbed Rock Zone, and Borehole
The average thickness of many of the anhydrite layers near the repository (e.g., anhydrite layers a and b) is
much less than 0.45 m, hence, criticality in these tlactured beds is impossible @lg. 13). (Although the 0.45 m
applies to dolomite [CaMg(CO&] rather than anhydrite [CaSOd], the influence on criticality would be similar
because of the similar influence of the elements on neutron behavior.)
The volume of disturbed halite is very large (e.g., the initial volume of the disturbed rock zone is likely larger
than the mined volume of the repository). However, the 700 kg of plutonium FGEs required in a homogeneous
mixture of salt (Fig. 10a) is 4 times greater than will be emplaced in any one room (-21 Mg FGE of ‘9Pu will be
emplaced in the repository [Table II], which has 118 equivalent rooms; thus -178 kg FGE of plutonium is emplaced
per room). Second, spherical diameters that are the same or greater than the original height of the waste (2.68 m)
(Fig. 6) are required for criticality in Salado salt (calculated from Fig. 10a). Consequently, if movement of fissile
material occurs from the repository into the disturbed halite, a mechanism must exist to concentrate the solution. In
general, the feasibility of concentrating the fissile material is not credible because no plausible geochemical
mechanism exists for selective movement of fissile material into only a small portion of the disturbed rock zone, as
discussed in Section VI.
Criticality within the borehole can be eliminated because insufficient space exists. The diameters used for
boreholes are 0.413 m (16-1/4 in.) or 0.444 m (17-1/2 in.) for oil and gas wells into the Culebra and 0.311 m (12-V4
in.) and 0.356 m (14 in.) into the salt (66, p. 4-45). Spherical diameters for fissile materialkdtlbrine mixtures are
much greater than these diameters (as calculated from Fig. 10a). Furthermore, if dolomite fell from above and filled
an empty borehoIe, all the original dimensions are less than the required spherical diameter or the 0.45 m required
thickness for criticality.
IV.C. Brine Flow through Repository
The rate of accumulation of fissile material in the fa fieId is dependent upon the brine flow (q) through the
repository (hydrologic constraint [see Fig. 3]) and total fissile concentration (Cd. Because of its higher permeability
relative to other strata in the area, practically all of the simulated dissolved radioisotopes migrate through the
Culebra in simulations, with only a maximum of -5.2 m3 of brine going above the Rustler Formation. This
maximum occurred for the S5 scenario on vector 50 @lg. 15), where vector 50 refers to the simulation with the 50th
set of parameters sampled through Latin Hypercube Sampling (LHS), a constrained Monte Carlo sampling method
~.
The maximum simulated brine flow in the Culebra was 3.6 x 104 m3/104yr @lg. 15). The source of the large
volume of brine in vector 23 is the marker beds, which, in turn, is the result of a high sampled value of Salado
permeability. Intrusions into the brine reservoir below the repository actually pressurize the repository to a very
small degree and so brine volumes are slightly reduced in scenarios S2, S3, and S4.
IV.D. Plutonium and Uranium Concentration Limits
IV.D. 1. Dissolved Concentration Limits
Brine Type Dependence. The range of Pu and U volubility (SD)is somewhat dependent upon the ionic strength
(i) (i.e., type of brine). Consequently in the 1996 PA, the volubility is estimated for both Salado (i = 6700 mM) and
Castile brines (i = 5300 rnM).
PI-ZDependence. The pH (i.e., activity of hydronium ion [H_’1)of the solution also influences the volubility of
Pu and U. For the 1996 PA, the DOE assumed, and the EPA subsequently required, enough MgO in the repository
such that any C02 generated by microbial degradation of organic material in the waste (e.g., paper) would form
MgC03, such that dissolved COZ(Hz.C03*)would not few, thus, the pH of the brine stays within the range of 9.4 to
9.9 in the repository. Narrowing the range of pH narrows the volubility range of various radioisotopes. This range of
concentrations is narrow in comparison to the range used in the 1991 WIPP PA, which incorporated volubility
values that were determined by an expert panel (66. 70). However, the range does center about the mean and
median values used in the 1991 WIPP PA and elsewhere @lg. 16).
Oxidation State Dependence. The actinides can exist in the III, IV, V, or VI oxidation states. Lower oxidation
states of actinides are generally less soluble than higher oxidation states. Although plutonium can be present as
either PU[l’,PUN, Puv, or Puvr, or often as a combination of several of these oxidation states, the WIPP Project
assumed plutonium would exist 50% of the time as a III valence state (Pu*l[)and 50% of the time as IV valence state
(Pu’v) ~.
Calculated Volubility. Volubility for an actinide oxidation state in solutions of high ionic strength, as occurs in
Castile and Salado brines, was calculated using a chemical equilibrium code (FMT, m, which accessed the WIPP
database of Pitzer interaction coefficients.13 Because experimental evidence suggests that actinide elements in the
same oxidation state exhibit the same chemical behavior (“oxidation state analogy”) ~, p. 536; m, it was possible
to use Pitzer coeftlcients experimentally evaluated for other actinides (e.g., Phzer coefllcients evaluated for ThW
were used to estimate the volubility of other IV actinides such as PUN). The calculated volubility in Salado and
Castile brines, respectively, was 5.5 x 10-5mM and 6.5 x 10-5 rnM for HI valence actinides; 4.4 x 10-3 mM and 6.0
131n~mctiw,~my intemctioncocfficien~~ quitesmallandcan be neglected however, if an iSnpOItiU3tinte~c~on is left ou~ the ~“lti%
stillity of a species and thereby the ovemfl element volubility will be in error. For the 1996 PA. many of the Pitzer coefficient were set to zerofor N vafence actinides in FMT. However, the catcrdated solubifity for PUNwas higher than the catculatcd volubility for U“’. Because of thksuspicious rcsulg the WIPP Project has continued to work on improving the estimation of Pitzer coefficients for FMT. One data point forestimating the interaction of Tfr(COJ54 with C~ has been obtained that results in the SOMSW of W actinid~ (e.g., PIS’”)being reduced by twoorders of magnitude (median solubllity of 2.0 x 10-5mM). However, the distribution of the plutonium used in the 1996 PA is used herein notonly because the original arguments were bud on thk vatue, but afso because a different interpretation may emerge from the continuingevacuationof plutonium volubility.
13 June 1,2000
x 10+ rrdvl for IV valence actinides; 2.3 x 10-3mM for V valence actinides in both brines; and 8.8 x 10-3 mM and
8.7 x 10-3mM for VI valence actinides Q, p. 4-39).
Uncertainly in Mubility. The uncertainty of volubility is potentially an important source of uncertainty in the
performance ‘measure (i.e., CCDF) Q, ~. Uncertainty in volubility is dependent on the uncertainty of the brine
chemistry, but this chemistry was not evaluated in the 1996 PA. Instead, a distribution for uncertainty was assigned
as follows: the one volubility value calculated by FMT was assumed equal to the median value (SD,SW). The
uncertainty about the median was then estimated using a sampled scale factor (fO)(i.e., SD= I@’ ● (SD5W), where
the piecewise linear distribution of ~Dwas established through analyst judgment. The minimum was 0.01 times
smaller than the median calculated value (i.e., fD = –2); the maximum, 25 times larger (i.e., fD = 1.4) (2, p. 5-15).
The same uncertainty distribution was used for both Pu and U but sampled separately, i.e., no correlation was
assumed in the uncertainty for Pu or U.
IV.D.2. Colloidal Concentrations
Four categories of colloids were considered: mineral, intrinsic, humic, and microbial ~. The mineral-type
colloids are mineral substrates that readily adsorb actinides; the intrinsic colloids are polymeric radioisotopes; the
microbial colloids are microbes that have adsorbed actinides; and the humic colloids are actinides cornplexed by
humic organics. The mineral colloid concentration (C.i.CJd)was fixed at 2.6 x 10-5rnM (X). Only Pu was assumed
to form intrinsic colloids at a fixed concentration (Ci.l) of 10+ mM (Q). The concentration of humic colloids was
assumed to be a fraction of the dissolved concentration (SD),where the fraction was a function of the ionic strength
(or brine type [Br]), oxidation state (Qx), and radioisotope (Rn) (i.e., c’hu~ic= ji~j,-JBr, Ox, Rn) ● SLJ. The humic
colloidal concentration was assumed to be bounded by 0.011 mM m. Experiments with WIPP-relevant microbes
(@ found that microbes do not actively move actinides through the cell membrane, but instead passively adsorb
acdnides extracellularly similar to mineral adsorption. Hence, the microbial colloidal concentration was also a
function of brine type, oxidation state, and the radioisotope.
IIZD.3. Total Concentration for Plutonium and Uranium
A total concentration of Pu and U was calculated as the sum of the dissolved species (SD) and the four
categories of colloids discussed above@, p. 4-39).
C’r= SD+ cminer~+Cinr-F chic -1-Cmf,yobe, (1)
2.6 X 10-5mM
lxlo+mhl
min {S’&~i~ 0.01 1333M)
min{S&&&, ().()68~} for PUU1and PUN
mh{~~m[~~obe,().Z1 ~} for UN and UV1
In Castile brine, the total concentration for PU1llranges from 10-5 to 2.4 x 10-2mM with a median of 1.69 x 10+
mM. In Salado brine, the total concentration for PUIUranges from 3 x 10-5 to 2.4 x 10-2 mM with a median of 7.32
x 10+ mM. In Castile brine, the total concentration for PU*Vranges from 10-5 to 3 x 10-3 mM with a median of 6.4
x 10-5MM. In Salacfo brine, the total concentration for PUNranges from 3 x 10+ to 4 x 10-1MM with a median of
1.57 X 10-2MM @lg. 16).
Information for UN and UVIis also tabulated in Table IV. The median Uw total concentration is 10-2 MM in
Castile brine. The maximum U“[ total concentration is 2.3 x 10-1MM (10+”6)in the Castile (Fig. 16).
IV.E. Constraints Posed by Volubility .
IV.E. 1. Volubility Concentration
A solution of pure ‘9Pu02 at a concentration of 3 kg/m3 (the level needed for criticality) corresponds to 12 MM,
a concentration 30 times greater than the maximum total concentration of 0.4 MM of PUN predicted in the previous
section (Fig. 16). Similarly, a solution of pure ‘5UOZ at a concentration of 10 kg/m3 corresponds to 37 MM.
Consequently, a solution of either dissolved plutonium (0.4 MM, maximum totrd concentration) or dissolved
uranium (0,23 MM, maximum total concentration) cannot go critical in the repository or elsewhere. Even maximum
solubilities of plutonium and uranium used in other studies are not critical @lg. 16).
The Individual Protection Requirement in 40 CFR 191 at the accessible environment (15 mSv/yr) is -680 pCtit
for a5U or -55 pCi/@for ‘9Pu. The Ground Water Requirement in 40 CFR 191 at the accessible environment is
even more stringent-15 pCi/! (see Section I.B.1). The equivalent radioisotope concentrations of 15 pCiA?for ‘5U
and ‘9Pu are 3 () x 10-5 MM and 1.0 x 10-9 MM, respectively. Hence the Individual Protection Requirement and
Ground Water Requirement for human health adequately protect against criticality near the boundary of the
accessible environment (even with iron-oxide adsorption, as described later).
lV.E.2. Maximum Uranium Release
Although releases above this limit do not necessarily imply criticality is possible (because the fissile material
must be concentrated), the volubility of the fissile material places a limit on the maximum radioisotope release over
10,000 yr. For uranium, the volubility required to release 50 kg of fissile uranium in 10,000 Y is 10+5 MM (i.e.,
50 kg c 5.1% ● [mole/O.235 kg]/[3.6 x 104m3 s m3/103 0]). This volubility is above the upper range on the uranium
‘1”2for UN and the upper range of 10concentration of 10 ‘“G for Uw anticipated at the WIPP (Table IV). Thus,
criticality with fissile uranium is precluded based on the anticipated maximum volubility and the crdcuIated
maximum fluid flow.
IV.E.3. Plutonium Reaching Marker Bed 139 and Culebra
For Marker Bed 139, the maximum flow is 2225 m3 flowing to the south and 1800 m3 flowing to the north in
vector 50, S5. The maximum flow of 2225 m3is not sufficient to move even 2.2 kg of ‘9Pu at the current maximum
volubility of 3 x 10-3MM [(3 x 10-3mole/m3) (2225 m3) (0.239 kg/mole= 0.16 kg c 2.2 kg)]-
In the 1996 PA, the fissile material reaching the Culebra, using the calculated flow rates (Fig. 15) and
concentrations of fissile material, is generally quite low. Only vectors 23 and 25 exceed the very conservative
2.5-kg mass limit for ‘9Pu using the maximum flow rate in S5 (Hg. 17); only vector 23 exceeds the 7-kg mass
limit.14
‘4 The possibly Iorgereduction in solubitity for PUNis sufficient to eliminate the possibility of enough plutonium reaching the Culebm in the first10,000yr to cause criticality.
,J
15 June 1,2000
V. GEOCHEMICAL CONSTRAINTS BASED ON ORE DEPOSITS
Once radioisotopes leave the repository, the general tendency is for them to disperse; only in specird
circumstances will radioisotopes concentrate. Because the required concentration of fissile mass to go critical is
similar to that found in an ore deposit (Section III), an examination of the genesis of uranium ore deposits provides
some insight into the probability of an ore-like concentration of fissile material forming at the WIPP.
V.A. Uranium Ore Deposits in United States
V.A. 1. General Conditions
Most uranium ore deposits worldwide are in sedimentary rock, with nearly one-third in sandstone. As reported
by Adler@, more than 95% of the uranium deposits in the United States that are potentially exploitable reserves
are in sandstone and found primarily in Wyoming and the Colorado Plateau. These sandstone deposits in the United
States represent -30% of the world uranium reserves @Q. The more concentrated portions of these ore deposits
vary between 1 and 5 kg/m3 (1000 to 2000 ppm). While a few uranium ore deposits formed simultaneously with the
sediments (placer or syngenetic ore deposits) (e.g., Ambrosia Lake near Grants, New Mexico) (Q, p. 116), the
uranium ore deposits in the United States were mostly deposited after the formation of the sandstone (i.e., epigenetic
ore deposits). The sandstone and other sedimentary rocks usually served as an aquifer that transported soluble U“]
species to a location, within or adjacent to the aquifer, that contained a chemical environment that caused deposition.
Occasionally, a completing agent such as vanadium is present in the sedimentary rock that produces an insoluble
uranium mineral, such as carnotite and tyuyamunite (82, 83). More often, deposition occurs when a soluble UV1
species (e.g., uranyl carbonate) is reduced to a UN species by organic material, pyrite, or HzS.
V,A.2. Uranium Deposits from Same Period Adjacent to WIPP
In Oklahoma, radioactive anomalies are present throughout several sedimentary sandstone formations deposited
during the late Permian Period. In the 1950s, about 13 tons of uranium ore was mined from the Rush Springs
Sandstone, which corresponds to Bell Canyon sandstone@ or older formations (.&). In addition, a Doxey shale,
which roughly corresponds in age to the Dewey Lake Red Beds Formation Q&l)has small lens of carnotite and
tyuyamunite uranium minerals deposited a distance of 46 m (150 ft) along a fracture in thin interbeds of sandstone
and sihstone of the shale.
V.A.3. Uranium Ore Deposit in Evapotites Adjacent to WIPP
As mentioned earlier, uranium ores in sedimentary rocks other than sandstone are uncommon (c5%) in the
United States. Conditions for the deposition of uranium in carbonate rocks, such as limestone or dolomite, are
especially unfavorable because uranyl carbonate complexes are highly soluble; hence, carbonate rocks typically
contain no more than 1 ppm uranium Q@. However, one small uranium ore deposit of the Jurassic Period occurs
near Grants, New Mexico, where the Entrada Sandstone aquifer is overlain by the Todilto Limestone formation
(similar to dolomite but lacking magnesium) that produced some oil. Highly ffactured areas of the limestone
formation allowed water carrying uranium to move into the limestone and deposit local accumulations of uranium
where HZS was present. Conversely, movement of some HzS frow. the limestone caused pockets of uranium
deposition in the sandstone (86. 79, 87).
V.B. Oklo Natural Reactors
The formation of the Oldo uranium ore deposit in Gabon, Africa, began when oxygenated water dissolved
uranium deposited in sandstone. This dissolved Uw was then reduced to UN in a saturated sandstone formation
through the presence of hydrocarbons in an overlying black shale formation. The Oldo ore, which contains
-17,300 Mg (17,300 metric tonnes as U203), consists mostly of a uniform low grade ore (0.2 to 1 %wt uranium);
however, a high grade ore (s20 %wt) was located at faults in the sandstone where greater amounts of hydrocarbons
were structurally trapped. The Oklo ore deposit is unique in that about 16 Ienticular regions have been identified
within the high grade that operated as natural reactors starting about 1.97 Ga, when natural uranium had a ‘5U
content of-3.68 %wt. The reactors operated intermittently for about 2 x 105 to 8 x 105yr (88, 89). The first six
natural reactors are typically 10 to 20 m in length and width, and less than 1 m thick (j&). Some of the natural
reactors are concentrated up to 55 Yowt,possibly because of added convective circulation of water heated by each
natural reactor, which supplied more uranium solution and dissolved silica in the sandstone to provide more
depositional space. The original porosity of the faulted zones is unknown but to provide adequate depositional
space and water moderation for criticality, the porosity must have been at least 10Yo(see Fig. 1lb) and was likely
between 20% and 30%.
V.C. Lack of Favorable Depositional Features in WIPP Disposal System
Several favorable features and conditions prompted the formation of the ore deposits in Gabon, Aiiic% and in
the United States. In general, the deposits were in sandstone that had (1) moderately high porosity for adequate
depositional space, (2) high permeability to allow sufficient water circulation, (3) oxygenated water to transport the
soluble UVIspecies, (4) an essentially infinite source of uranium in volcanic tuff or granitic source rocks, and (5)
either (a) reducing material to chemically reduce the uranium to insoluble UN species or (b) completing vanadium
to form uranyl vanadate minerals, such as carnotite or tyuyamunite.
These favorable features are absent in the WIPP disposal system. First, the naturally occurring brine cannot
oxidize the uranium or plutonium in order to mobilize the uranium and plutonium. Second, the pathways to the
accessible environment at the WIPP do not pass through permeable sandstone; the Bell Canyon sandstone lies
-590 m below the repository (Fig. 4) and the Dewey Lake Red Beds sandstone has a transmissivity too low to
measure and possibly is not saturated above the repository (X). Instead, any potential pathway to the accessible
environment must pass through evaporates such as carbonates where no evidence exists that either vanadium or
indigenous reductants/oxidants are present.
V.D. Natural Concentration Efficiency Based on Oklo
Besides demonstrating that the WJPP lacks favorable depositional features, Oklo can also be used to explore the
‘9Pu would be required to collect enough fissile material within 10,000 yr,implications of what volubility value for
with a probability greater than 10+ over that same 10,0OO-%period, to cause criticali~ Q, m. The answer is
conditional on the presence of proper geochemical conditions and proper physical conditions (i.e., P{ C Ip m h n c)
“P(h)) (see Fig. 3).
V.D. 1. Required Actinide Concentration Based on Maximum Flow
Of the 16 natural reactors, data are readily available in the literature for the first six (primarily “reactor 2“).
Based on these data, researchers have estimated that 6 Mg of ‘5U were consumed out of -800 Mg of high grade
uranium associated with the six natural reactors. Consumption of 6 Mg corresponds to -1028 fissions or -15,000
MW-yr (89. 91). Basically, six criticalities occurred per 800 Mg of U at an enrichment of -3’%, which is about the
average enrichment of U in the WIPP (Table II) (but four times less enriched than Pu). This rate, in essence, factors
in microscopic inefficiencies found in nature when macroscopic physical, hydrologic, and geochemical conditions
are generally favorable.
To be consistent with 40 CFR 191, the probability of criticality in the first 104yr must be evaluated, or as done
here, the plutonium volubility evaluated such that the probability of a criticality in the first 104yr is less than 10+.
The probability model is based on the failure-rate function defined by r(t) = -d hz[l-F(t)]/dt, where t is the time
elapsed since the disposal system was closed and F(r) denotes the cumulative distribution function for the first time,
T, when criticality occurs (i.e., F(t)= {Tslt}). The failure rate function can be integrated to give
Over a period of a year, F(t) must be less than 10-8(i.e., P{ C I p m h m c} ● P{h} < 10+). This fact, in turn,
implies r(t) s 108/yr. Yet the value of r(t) is bound by the product of the “microscopic efficiency” of natural
concentration processes as estimated from the Oklo natural reactors, (e = 6 events/800 Mg), unknown plutonium
concentration (here assumed to be an unknown volubility [SD]), the maximum cumulative flow rate over 104 yr
– 3.6x 104m3/104yr from Fig. 15), and the 104-yr regulatory time period (2) (i.e., r(t)sthrough the repository (q~m -
e~~u~. Solving for the plutonium volubility (Sf)) yields a value of 1.55 x 10-21mM (i.e., SD= [10-8 events/yr ●
800 Mg]/[6 events ● 3.6 x 104m3/104yr]).
V.D.2. Required Actinide Concentration Based on Range of Flows
The above arguments can be refined because the probability of the high flow rate (qd of 3-6 x 104m3/104yr is
no more than 0.01 based on the 100 samples (Fig. 15). That is, the cumulative distribution function (F in Eq. 2) is
actually conditional on the discharge (q) through the repository.
F(t Iq) = l-exp[-~r(t) t](3)
=1 -exp(-eS.qT),
but
F(t) = ~F(t Iq) p(q) dq
=~~-(l-exp(-eS.qT))p(q)@.
Ifp(q) is represented by its empirical cumulative distribution function, then
F(t) ~ X(1 -exp(-dDqiT))pi.
(4)
(5)
In the previous section, thk sum was bound by F’(r)= S 1 – exp (–eS~_71 Yet because the individual qi and
p, are available from the 1996 PA calculations @lg. 15), a better estimate is possible. Using the distribution in
‘“’ Thk maximum volubility is greater than the entire range ofFig. 15, the maximum volubility is -0.2 MM = 10 .
estimated uranium concentrations, and most of the range of plutonium concentrations.
V1. GEOCHEMICAL CONSTRAINTS ON CONCENTRATION MECHANISMS
As described in the previous section, the usual scenarios of ore formation are absent. In the following sections,
we speculate about less likely scenarios for deposition of plutonium and uranium although based on the same three
geochemical mechanisms and their probability P{C} (see Fig. 3): adsorption on mineral surfaces (e.g., ion exchange
or surface complexation), filtration of colloid material, and precipitation.
VI.A. Adsorption of Plutonium and Uranium
VLA. 1. Adsorption of F/ssile Material in Repository
Within the repository, conditions that might lead to potential adsorption15 of Pu or U on corrosion products or
dolomite in a localized area do not exist. Granted, the waste from each storage/generation site contains a wide
variety of potentially adsorptive material, such as soils, waste solidifies (e.g., bentonite), corroded metals, and
degraded glass; however, potentially adsorptive waste material such as IUS4clay, and solidifies’sis not concentrated
in a few drums (see Fig. 5). Rather, adsorptive materials will be fairly uniformly distributed throughout the
repository and preferential concentration of plutonium in one area is unlikely. Similarly, substrates and nutrients for
biofilms that might cause fissile concentration are also disseminated fairly uniformly throughout the repository.
Furthermore, separation of Pu to the exclusion of the large amounts of the other radioisotopes (e.g.,zsu) is not ‘
plausible even if localized concentration of adsorptive material did occur. If a facile technique for separating Pu
through adsorption were known, it would have been used in place of the expensive PUREX method for separating
fissile ‘9Pu from other radioisotopes. Likewise, a difficulty in proposing adsorptive backilll as an engineered
barrier in the WIPP repositow is that the adsorption on proposed backi311sis not specific and so large amounts of
other material that are in the waste also adsorb. This situation greatly decreases the likelihood of criticality because
many of these materials (e.g., nonfissile actinides, iron, nickel, and lead) readily absorb neutrons and halt a critical
excursion. Several quantitative arguments to support this general reasoned argument are provided below.
VI.A.2. Isotherm Adsorption Model
For modeling, adsorption in geologic media is most often expressed as the ratio of mobile material (here, the
mobile U and Pu in the brine) and material on the immobile rock (here, corrosion products in the repository or
dolomite in the Culebra). The distribution coefficient (KD) in the Freundlich isotherm model (e.g., n p. 353) is
defined as .dm = KDCS where x is mass of attached adsorbate on the solid, m is the mass of immobile solid, C, is the
equilibrium concentration of the adsorbate in the solution, and n is a nonlineru exponent. The exponent n is often set
to one in the equation, thereby not taking into account the potential to saturate the adsorption sites at high
concentrations of radioisotopes. Although assumed valid for evaluating adsorption of radioisotopes on dolomite at
is He~in, *@~n ~fem to the accumulation of materiat at the interface of ~orher materi~- Adsorption subsumes several differentmechanisms (e.g., surface complexation and ion exchange). Because geochernicat adsorption is not strictly a surface phenomenon somegcachemists prefer to use the more inclusive term “sorption.” However, thk article uses the mose tmditionrd terms adsorption and absorption.with the latter used when referring to neutron absorption (capture) by the nucleus.
19 June 1,2000
. .
low concentrations in transport calculations at the WIPP, setting n equal to one was not assumed valid for criticality
calculations where high concentrations were required to cause a criticality event.lc’17
VLA.3. Nonlinear Isotherm for Iron Oxides in Repository
The initial iron corrosion products of the steel withh the WIPP repository will likely be an amorphous
fernhydrite (am-F%Os ● HzO) that will progress, through dehydration, to more crystalline goetilte (a-FeOOH) and
then hematite (cx-FQOs) &, m. Payne et al. U Figure 3) experimentally evaluated the adsorption of Uw on
ferrihydrite (suspendkrl in a solution at 89 mg/1) over a range of pH and at three initial uranium solution
concentrations: 10-3 mM, 10-2 rnM, and 10-] mM. (The latter concentration is very close to the maximum uranium
solution concentration anticipated at the WIPP.) The extrapolated concentration in the fernhydrite colloidal
suspension is only 0.05 kg/m3 at the maximum UV1volubility of 10--6 rnM (i.e., 2.0 mole/kg ● M. “ 0.089 kg/m3)
(Fig. 18); this value is far below ~e 50 kg/m3 critical concentration. The ‘5U concentration on any one ferrihydrite
particle could be high-43 kg/m3 assuming 5.1% ~sU enrichment (i.e., 2.0 mole/kg ● lf~ ● p~[l - fl “ 5.1%) but still
not above the critical limit. Furthermore, transformation of ferrihydrite to the goethite might be fairly rapid (-10 yr)
CM)such that pure fernhydrite might not be present within the repository in large amounts (at least 20 kg) in a small
area. Finally, a critical concentration would require exclusive adsorption of uranium from the contaminated brine in
the repository-an unlikely occurrence.
Although isotherms of Pum adsorption on iron oxides at high solution concentrations of plutonium are not
available, an extrapolation based on two points at low concentrations @, Figure 1a) shows that a similar situation
occurs for plutonium whereby the estimated solid concentration on goethite cannot reach the critical limit, assuming
a ~9Pu FGE of 14% (i.e., 4 x 10-2mole/kg ● M,v● p~[1 – ~] ● 14’%0= 5 kg/m3) @lg. 18). Although this extrapolation
has large uncertainty, it is reassuring that the values for Puv adsorption on goethite lie near this same line @,
Figure 84) because Puv is thought to be reduced to Puw when adsorbed on goethite. While use of additional data
reported by Sanchez et al. m, Figures 1b and lc) for Puv results in a new line that predicts a solid concentration on
goethite of 26 kg/m3, it is only slightly above the 20 kg/m3 limit for rust surrounded by silica rich material (i.e., 2.3 x
10-1mole/kg” MW“ p~ [1 – @ ● 14%=26 kg/m3) (Fig. 18).
VI.A.4. Adsorption Sites Available on Dolomite
Criticality is unlikely for Pu adso~tion on goethite, as described in the previous section, and provides a
reasonable upper bound for adsorption on dolomite. However, the critical limit for dolomite is much less than for
rusu thus, additional arguments based directly on properties of dolomite are presented below.
The two cases in Fig. 17 showing a potential for criticality with plutonium in the Culebra can be eliminated for
three reasons based on geochemical and physical constraints of the system the density of adsorption sites on
dolomite would have to be very large, surface complexation at adsorption sites would not be limited to only ‘9Pu,
and the deposition of fissile radioisotopes and thus the geometry of the critical mass must be at the borehole (the
latter reason is discussed in Section VI.C).
16~ Rwhxd et ~. ~, ho~~ver, fie tin~ &.VdUSS were USd in combination with maximum soM@ v~ues m a sc~ning tool to ev~uate
whether a problem might exist.17Al~hou@ tie nonlinm Freund]ich model does not have an asymptote tike the Langsrmir model @.&p. 362), it does ~Iow for =tlY reduced
adsorption at higher concentrations for n c 1.
20 June 1,2000
Adsorption experiments for radioisotopes at high concermations are difficult for evaporates and have not been
done for dolomite. Hence, a bounding estimate of the required adsorption sites, and thus the maximum amount of
fissile material that can adsorb onto a small volume, was made for dolomite and compared to other highly adsorptive
natural materials.
The measured mass-specific surface area on carefidly crushed and lightly acid-washed samples of dolomite
from the WIPP Alr Intake Shaft (l%g.6), as evaluated by BET surface area analysis m, varies from 500 m3/kg for
indurated dolomite to 2600 m3/kg for silty dolomite @&);these values correspond to between 1 x 106and 6 x 106mz
of surface area per m3 of dolomite (assuming a porosity of 16% and dolomite grain density of 2850 kg/m3).
Dividing the adsorbed plutonium concentration of 3 kg/m3 (which corresponds to 7.6 x 1024atoms/m2 of dolomite)
by volume-specific surface area gives a site density of between 1 and 7 atoms/nm2. This range is similar to the
observed site density for many highly adsorptive natural minerals (about 2 sites/rsm2)18(Q).
Thus, obtaining a critical concentration requires the entire capacity of an adsorptive natural material (of which
dolomite in the Culebra is reasonably bound). Furthermore, the ‘9PuW would be competing with other plutonium
isotopes and other radioisotopes, e.g., ‘fh]v and UN, which would behave similarly to PUNfor the sorption sites, thus
increasing the required solid concentration limit (Fig. 10). More importantly, plutonium would be competing with
other metal cations (and even Ca and Mg, in solution). Hence, ‘9PuW would need to dislodge rdready adsorbed
material present on the dolomite.
The probability of criticality from adsorption of plutonium. becomes unlikely near the intrusion borehole,
because the critical concentration increases from 3 to-either -10 kg/m3 @lg. 10b) (-17 sites/nm2)*9as the result of
domination of the fluid composition by Castile brine or -19 kg/m3 (-32 sites/nm3) when dominated by Salado brine.
For uranium, the number of sorption sites is clearly too small. A concentration of 10 kglm3 of ‘5U corresponds
to 2.5 x 107 atoms/nm3, which in turn corresponds to 17 atoms/nm2. However, the uranium is only -5% enriched
(initially) and so the available adsorptive sites are much more likely to be filled with ~8U than ‘5U. To get 10
kg/m3of ‘5U at 5.l~o, enrichment would require 333 sites/nm2.
Consequently, adsorption, by itself, is an unlikely mechanism for concentrating massive amounts of fissile
material to obtain a critical solid concentration because of limitations in surface area in natural geologic systems. As
a precursor, adsorption, however, may potentially lead to the formation of ore deposits (e.g., 83:101, p. 244-249)
and is implicitly considered in the discussion of precipitation (Section VI.C).
VI.B. Concentration of Colloid Material in the WIPP Disposal System
The concern with colloids is (1) whether the concentration of colloids is critical or (2) whether there is a
mechanism by which actinide-bearing colloidal particles can be locally concentrated (e.g., filtration).
VI.B. 1. Limits on Dispersed Colloidal Concentration
Similar to adsorption, the uptake of fissile actinides by the various types of colloidal particles is governed by
equilibrium among dissolved actinides, solid actinide phases, and the colloidal particles. The link between colloidal
la Bmdy et ~. ~ ~po~ that if ~1 the Mg rissd& ions p~ent in a layer of dolomite were available for ion exchang% the m~mum densitywould be 30 sitednmz; however, not all Mg and Q ions are available and so the site density is probably much lower.‘9Synthetically prepared goethke (a-FeOH) has a site density of only -17 sites/nm2 (J,QQTable 3-l).
21 June 1,2000
actinide concentration and equilibrium thermodynamics produces a limit on the concentration of colloidal actinides
that cannot reasonably exceed by 30 times the dissolved concentration of fissile actinides and thereby go critical.
For example, the total uranium and plutonium concentration of the adsorption experiments discussed in Section
VI.C.2 (Fig. 19) includes the colloidal suspension and is not critical; the colloidal suspension of uranium on
ferrihydrite was only 0.05 kg/m3.
VLB.2. Fate of Colloids in Repository prior to Intrusion
The behavior of the colloidal particles themselves is best described as a steady-state process, in which new
particles are continuously generated and dispersed in the liquid phase but are gradually destabilized through a
variety of processes. Destabilization of mobile colloidal fissile actinides, followed primarily by gravitational
settling, could conceivably concentrate fissile material, but the mechanism is limited to local microenvironment
that are likely to be centimeter-sized or even smaller within the WIPP repository. After closure of the repository,
salt creep will crush the waste containers and compact waste material (Fig. 14). Colloidal particles could form
throughout the waste matrix however, the fluid columns in which agglomerated colloidal particles will settle are
likely to be short and poorly connected. Consequently, the local concentrations of colloidal agglomerates that
develop in the WIPP will remain uniformly distributed in the waste throughout the repository and will not
concentrate in a single location. By analogy, in a closed house, dust (i.e., aerosols) settles on the horizontal surfaces
in the house. The dust does not, however, concentrate at any single location.
VLB.3. Fate of Colloids in Repository after Intrusion
In a disturbed reposito~, flowing brine could result in filtration of any dispersed colloidal actinides, but
concentration of colloidal actinides will again be localized in small microenvironment. As described above, the
waste will contain many centimeter-sized microenvironment of small poorly connected pores. Flhration can take
place in interstices at the boundaries of marked changes in pore size, resulting in small concentrations of colloidal
actinides in those localized small areas.
Again, considering the dusty-house analogue, an air stream flowing through an open house is similar in concept
to brine flowing through the repository. In the dusty house, dust that has settled on horizontal surfaces is stirred up
and concentrated in local environments, such as comers or other areas where horizontal air velocity is small. In the
repository, local accumulations will develop in a similar fashion. However, the tight packing of the waste will
provide local filtration of colloidal particles. Hence, the fissile actinides associated with colloidal particles will
remain fairly uniformly distributed within the repository on a macroscopic scale. Furthermore, apart from these
qualitative arguments, the perfect filtration of iron oxides with adsorbed uranium cannot go critical and iron oxides
with adsorbed plutonium probably will not go critical, as previously discussed in Section VI.A.3.
VI,B.4. Colloidal Filtration in Culebra
Because the total mobile concentration of Puw is due partially to microbial and mineral colloids, filtering of
these colloids could lead to a possible means of Pu deposition in the Culebra. As argued above, it is probable that
colloids susceptible to filtering would be filtered during transit thro’tigh waste within the repository or in the
borehole backtill material; however, to be consistent with the assumptions for the 1996 PA, colloids were assumed
to have escaped the repository and borehole and then be filtered in the Culebra. The simulations in which large
.
releases of Pu are calculated, as occurred for Pun” in Salado brine, were used. In Salado brine, Pu associated with
mineral colloids and microbes accounts for only 7~o of the Pu total concentration (Table IV). Perfect filtering of 79Z0
of the maximum 110 kg released in simulation 23 (Fig. 17) results in 7.7 kg that must be filtered in less than 0.46 m3
from the borehole (e.g., sphere of radius 0.46 m) to obtain a minimum concentration of 19 kg/m3 for Salado brine
(Fig, 10). Although the 7.7 kg is similar to the 7-kg to 9-kg minimum mass calculated @lg. 10b), these masses
require a concentration of -30 kg/m3; furthermore, the 7.7-kg mass is smaller than the 13.4-kg minimum mass
suggested by the argument based on Oklo reactors (Section V.B). Therefore, colloidal filtering of Pu cannot cause
criticality by itself.
The total concentration of UN in Salado brine is partially due to filterable colloids (37%) (Table IV) but the
mass of UN brought to the Culebra is insufficient to cause criticality. While the mass of Uw brought to the Culebra
is higher than for UN, no portion of the volubility is due to colloids (Table IV).
VLC. Precipitation of Fissile Material
A necessary condition for precipitation of fissile material and other major components of the brine is a change
in the brine chemistry (i.e., the redox state and buffer capacity, pH, ionic strength, or solute concentrations) such that
the volubility of a component is reduced. As previously described, the deposition of U generally occurs when
soluble fissile material (e.g., Uw as a uranyl carbonate) is reduced by organic material, pyrite, or HzS. Because this
particular mechanism is improbable in the WIPP disposal system, as previously described in Section V, several
other speculative mechanisms were proposed in the following discussion but were also found to be improbable.
VLC. 1. Precipitation in Repository
The main determinant of oxidation states within the repository will be the rate of interaction with elemental
iron, Fe”, and the rate of microbial-facilitated reduction of oxidized species, such as nitrate and sulfate. The
resulting repository waters are expected to have relatively high concentrations of reducing species such as Fe+**and
S-ll. Hence, the most likely oxidation states are the insoluble PU”l and UN; however, there is some possibility of
more soluble PUIVand UV*. Because iron and organic waste are present throughout the waste, there is little chance
of great variation in the redox state of the brine from one place to another, and thus no driving force for dissolution
and precipitation to redistribute the fissile materials.
VI.C.2. Precipitation in Culebra
The calculated plumes of radioisotopes migrating through the Culebra over 10,000 yr do not show a tendency to
concentrate in a small area in the 1996 PA u a. Although the numerical grid size is not small enough to rule out
this possibility, the following arguments assume that, absent any strong indication otherwise, the most likeIy area to
precipitate the fissile material in sufficient concentration is at the borehole (similar to adsorption and colloidal
filtering). As a supplemental analysis to the 1996 PA, a simulation was run with a numerical mesh smaller than that
used in the 1996 PA analysis to evaluate the size of the zone in which dilution and potential precipitation might
occur because of changes in ionic strength, pH, or redox state.
Simulation of Miring Zone. The modeling system used has been explained elsewhere u 17. 19J. The
underlying mathematical models used have also been explained elsewhere (e.g., ~. Hence, only a cursory
explanation of the simulation is given, and the following discussion focuses on the results.
The supplementary simulation used information from S5, vector 23 (Fig. 15), and SECOFL2D (Version
3.OIZO) with a steady-state injection of 3.62 x 104m3 of Castile brine over 10,000 yr. Based on measured pumping
and slug tests, Beauheim and Holt w note that the Culebra behaves hydraulically as a single-porosity medium
where the transmissivity (T) is less than 2 x 10+ m2/s. Since the transmissivity values for WIPP-12, ERDA-9 and
H-3, located near the repository, are 1.1 x 10-7, 4.9 x 10-7, and 9.4 x 10-7 m2/s, respectively, the Culebra was
modeled as a single porosity medium (with a transmissivity of 4 x 10-7 m2/s) near the repository for evaluating the
fluid flow. For the transport calculations with SECOTP2D (Version 1.41), a small local grid, 26 m x 26 m with 961
cells (1 m or less on a side), was centered over a hypothetical borehole penetrating the center of the repository.
Similar to the original 1996 PA calculations, boundary conditions of hydraulic head (H) for the small local grid were
interpolated from the regional flow grid (Fig. 20).
Change in Ionic Strength. As a repository brine enters and mixes with Culebra brine, the ionic strength will
drop from between 6700 and 5300 mM to 800 mM. Volubility of most highly negatively charged species are
stabilized by high ionic strength so that a minertd whose volubility is controlled by highly charged species in
solution will have a higher volubility in high ionic strength brines. Hence, as the brine is diluted, the volubility could
decrease, Yet the concentration would also be reduced by the dilution, and so precipitation would occur only if the
change in volubility was greater than the change in concentration. The model of IV actinide volubility used in the
1996 PA showed the minus-six-charged pentacarbonate species dominating the volubility, and thus the volubility
could be dependent on the ionic strength of the brine. Similarly, the model of HI actinide volubility shows the plus-
2-charged hydroxy complex dominating the volubility and decreasing by an order of magnitude as the ionic strength
decreases from 6700 mM in Salado brine to 5300 rnM in Castile brine. Hence, if these models are accurate, they
predict some precipitation on dilution.
Consequently, a possible situation is the dilution of PuIVin a Salado-dominated brine to a Culebra-dominated
brine (Fig. 16). (Because insufficient PU1llreaches the Culebra in 10,000 yr, it is not considered here.) The
maximum calculated discharge of ‘9Pu1Vover 10,000 yr as calculated in the 1996 PA is considered, which is about
110 kg. To obtain a plutonium concentration of 9 kg/m3 or greater (see Fig. 10b), assuming a singIe porosity
medium, the volume of deposition of all the plutonium reaching the Culebra in 10,000 Y must be less than 12 m3 (a
sphere of about 1.4-m radius or a cylinder of l-m radius, assuming deposition throughout the 4-m thickness of the
Culebra dolomite).m A simple estimate of the volume of the dilution zone as the annulus between two concentric
cylinders @lg. 20b) results in a solid concentration 12 times lower than the 9 kg/m3 limit even for a total of 110 kg
brought to the surface over 10,000 yr.
Change in pH. The pH and COZ concentration in the repository is controlled by the addition of excess MgO.
Contaminated brines entering the Culebra will carry dissoIved Mg(OH)z, but little solid Mg(OH)z, so the pH could
drop from alkaline to near neutral, and the COZ concentration rise, as the Mg(OH)z was consumed by the Culebra
C02. The change in pH and COZ concentration could cause coprecipitation of plutonium or uranium with MgC03
within the Culebra pore spac~ however, changes in pH and availability of COZ would still be dependent on
m The maximum size of each cell was selected to avoid numerically diluting rhe fissile mass beyond thk concentration. If the mixing volume hadapproached Wk cell size, the anafysis would have beersmn with even smafler cells.
24 June 1,2000
sufficient mixing of the two brines which, as shown in Fig. 20b, does not occur in a small enough region to obtain a
critical concentration.
Change in Redox State. The Culebra brine has low concentrations of active reduction or oxidation species.
Consequently, the redox state (Eh)2*of the brine as it moves through the Culebra will change little, even up to 100-
fold dilution QQQ. Therefore, reduction of the UV1to UN will not occur (i.e., the normal situation in which a
uranium deposit is formed by means of reduction of soluble species does not apply). Likewise, the reduction of PUN
to Pull[would not occur.
Two hypothetical scenarios of precipitation from oxidation also seem unlikely (a) the oxidation of Pum to PUN
and (b) the oxidation of FeI*to Fell[ and coprecipitation with PUN. Although as Punl is removed from the Fe” in the
repository, it may slowly oxidize to PUN,in the 1996 PA, these conditions do not cause precipitation because PUNis
more soluble than PU1llin Salado brine.z Because FeI1lis less soluble than Fen, any oxidation occurring at the
borehole could cause precipitation of amorphous ferrihydrite (am-FezOs. HzO). Fissile material could be adsorbed
onto the ferrihydrite and coprecipitated. However, the Culebra brine or dolomite does not contain a strong oxidant
(or reluctant), as discussed earlier. The availability of sufficient oxygen tlom the surface through a de~aded plug
with the permeability of silty sand is unlikely.
Adsorption near Borehole. For the sake of the above precipitation arguments, adsorption in the supplemental
simulation was ignored, and deposition of fissile material was assumed to occur in the mixing zone between the
injected brine and Culebra brine because of the change in ionic strength, pH, or redox state. However, it is
instructive to briefly describe the model results if adsorption is considered. Based on batch adsorption experiments
~, the linear distribution coefficient (K~) on dolomite for Puw in the Salado and Castile brines in the 1996 PA was
assumed to be uniformly distributed between 0.7 and 10 m3/kg (104, 105). If one uses a linear Freundlich model for
dolomite, the linear KD must approach 0.7 m3/kg for the average fissile density to be greater than 9 kg/m3. The
average fissile density is less than 9 kg/m3 if one uses a nonlinear Freundlich adsorption model for goethite (as also
evaluated in Section VI.A.3) (Fig. 20c). Furthermore, the average ftssile density is even less assuming 14% FGE of
‘gpu and must approach 6.8 m2/kg for the linear Freundlich model.
In summary, some change in chemistry is likely as brine contaminated with fissile material mixes within the
Culebra; however, the supplemental simulation suggests these changes in chemistry will likely occur over a wide
area. Furthermore, the fact adsorption would occur in a different region than the precipitation @lg. 20b) and that the
amount of injected brine would actually vary over time as the borehole creeps shut or precipitates plug the porosity
would also tend to disperse the fissile material.
VII. CONSEQUENCES OF A CRITICALITY
For a criticality to have a significant effect on human health after repository closure, it would have to either (1)
produce more hazards than originally present (e.g., produce more fission products) or (2) degrade the ability of the
disposal system to contain radioisotopes (e.g., produce excessive heat or kinetic energy). These two categories of
consequences are examined below.
2’ The oxidizing or reducing potentiat (Eh) of the solution is measured in volts. Eh is formatly defined by the Nernst equation @ p. 405).n This precipitation mechanism is indeed possible with the more recent estimates of Puw solubitity because Pu’” is now more soluble. However,the sohdsilityof Pu’” is sufficiently low that a critical mass never reaches the Culebra in 10,000 yr.
25 June 1,2000
.
VILA Generation of Fission Products
VILA. 1. Burnup of Plutonium
The EPA performance measure (R) for radioisotope release is the sum of the quotients of the activity of
radioisotopes divided by the regulatory release Iimic
(6)
where fw is a waste unit factor equal to X W#106Ci; Wi is activity (Ci) for a-emitting TRU radioisotope i with half-
life (tln) 220 years; Li is the regulatory release limit (Ci) for radioisotope i specified in 40 CFR 191, Appendix A
~; nR is number of radioisotopes contributing to the releasq and Qi is cumulative release over 10,000 yr for
radioisotope i beyond a 5-km boundary.
Because R is a surrogate for health hazard, it can be used as a metric for the consequences of criticality wit.hh a
geologic repository as previously argued in Rechard et al. (@. For a waste consisting only of ‘%%, R is unity
when Qi =fJi. Assuming a unit value forfw Qi = Li = 100 Ci for ‘9Pu, which is equivalent to 1.6 kg. To identi~
the consequences of fissioning ‘9Pu, all 1.6 kg of ‘9Pu was converted to its fission radioisotope products (lQ@.
Then each fission product was decayed and its contribution to R calculated as a function of time. After 110 yr, the
‘9Pu. Thus, if a subsurface criticalitycontributions to R from the fission products are less than the initial 1.6 kg of
occurs with ~9Pu and the fission products are not substantially more mobile than ‘9Pu (which is valid at the WIPP),
then the hazard is less than before criticality. A similar benefit occurs when using Pu in nuclear reactors (or the
proposal to transmute Pu by particle accelerators), but the possibility of human exposure during operation or when
dismantling the facility is greater. .
VII.A.2. Fission Products Generated from Uranium Fission in the Repository
A reduction in potential health effects, as measured by the EPA units, from uranium fissioning does not occur as
for plutonium, but neither are the effects significantly detrimental. The placement of CH- and RH-TRU waste in the
WIPP has an initial (and maximum) heat power output of 136 kWx (Fig. 19), of which about 2.2% is fi-omdecay of
fission products that have contaminated the RH-TRU waste.24 This totrd 3 kW (2.2% of 136 kw) from fission
product decay is quite small; to provide perspective, it is equivalent to one-third container out of more than 7860
containers of commercial SNF (CSNF) planned for disposal at Yucca Mountain, Nevada assuming that each
container holds 21 pressurized water reactor (PWR) SNF awemblies (-8.2 MTHM/containeq -10 kW/container)
~, Table 3-3).
The 3 kW from fission products at the WIPP, assuming a burn-up of 40 GWd/MTHM, represen@ 2 x 102s
fissions (-102s”3). In comparison, the maximum number of fissions possible from the 8 Mg of ~5U (Table V) at the
WIPP, also assuming a 40 GWMMTHM bumup (consumption of about 4% of the fissile mass) with 2.8 x 10X
fissions/GWd, is 9 x 10~ (-1027) fissions. The comparison is more favorable when the entire heat power of 136 kW
u A maximum heat power of 197 kw from radioactive decay can be estimated assuming (a) the maximum 0.2 kg of FGEs of plutonium is placedin each 0.21-m3drum and 0.35 kg is placed in each 1.8-m3box, (b) the repository contains 3.7 x ld drums and 2.3 x 104boxes, and (c) ‘gpo
Emdum ‘at at2.4 ‘kg.For the 2.2% of the 97.8% of CH-TRU tiloisotopcs, 65% is from 238h enemt~ at the Savmfi River Plant (SRP) for power genemto~ in
space mrdelsewhere), 1870is from 9zw~, Syois fmm m, ~d 11~0 is from “’h (L m.
26 June 1,2000
is used. Again, assuming a current average of 40 GWd/MTHM, the equivalent 13.6 containers of CSNF represent
-lOU fissions-an order of magnitude less fissions than were initially placed at the WIPP.
VILB Potential Damage to Geologic Media
VILB. 1. Categories of Criticality
One important aspect of criticality is the degree of moderation with regard to the speed of neutrons and thereby
promotion of fission; moderation is most likely from water in a geologic setting. A second important aspect is the
rate that the fissile mass is assembled since it is related to the energy release. Potentially, the rate of assembly can
be slow, fast, or explosive, where slow refers to a process occurring over geologic time such as deposition from
precipitation, fast refers to events over seconds as might occur in an earthquake, and explosive refers to events over
microseconds. Using these two factors of criticality, six criticality categories are as follows (Q @: (1) high
moderation, slow assembly (e.g., Oklo thermal reactors w), (2) high moderation, fast assembly (e.g., aqueous
solution accidents), (3) high moderation, explosive assembly (e.g., Bowman-Venneri hypothesis ~), (4) low
moderation, slow assembly (e.g., fast breeder reactor), (5) low moderation, fast assembly (e.g., experiments/accident
with fissionable materials and analytically bounded with Bethe-Tait analysis m), and (6) low moderation,
explosive assembly (e.g., weapons).
Based on recent arguments ~, 50, 51, ~, 53, ~, @, the explosive assembly is highly improbable.
Furthermore, the average enrichment of fissile material is less than 14% so theoretical reasons also support the
improbability of nuclear explosion after assembly w p. 182). In the following discussion, the consequences of
fast assembly are used to bound potential kinetic energy releases of slow assembly (Category 1)~ (Section VII.B.3).
This assumption provides a conservative bound since fast rates of assembly are not likely at the WIPP because of the
diffuse distribution of fissile particles throughout TRU waste. Potential thermal energy releases from slow assembly
are directly estimated in Section VII.B.4.
VILB.2. Use of Criticality Accidents
In the discussion that follows, anthropogenic criticality accidents are used as an analogue for estimating energy
release from a criticality in or near a geologic repository. The validity of this approach was checked through two
types of calculations. First, a Bethe-Tait analysis of an unmoderated accident suggests a bound of 6 x 10]7fissions
or 18.7 MJ Q&, &j), which is approximately equal to the maximum number of fissions seen in accidents in
processing plants and laboratories throughout the world. More importantly, a detailed nuclear dynamics analysis of
a criticality in geologic media shows that the energy releases fkom nuclear criticality are minimized because the
reaction is quickly stopped by near instantaneous negative feedback u, Figures 4.3-1 and 4.3-2, 11.Q p. 10-41).
The feedback is from the heating of the fissile mass, which changes the resonance band (i.e., Doppler effects) of the
system; the negative feedback is not dependent upon the slower, thermally driven changes of the overall assembly,
such as evaporation of the water moderator. The maximum number of cumulative fissions (E) varied between 10IG
and 5.6 x 10*9during any one critical event and lasted between 3.6 x 103and 3.6 x 105s Q, p. 4-15) (Fig. 21). The
~ In a natumtsetting, a crdble method of slowly assembling the fissile materird without water or human intervention is unknown; thus theprobability of Category 4 is near zero; however, removing Category 4 fmm consideration is not necessary in the following discussion ofconsequences.
27 June 1,2000
maximum number of cumulative fissions could be expressed by the empirical equation: E = 2.73 x 1017&w3, where
s is &n/1 a+; &is initial reactivhfi 6 is reactivity (ratio of excess multiplication factor to multiplication factor) and
equal to (keM- l)/kcfl; m is fissile mass (kg); and @ is thermal feedback coefficient (K1) @, P- 4-4-s)-
VII.B.3. Disruption to Geologic Media
Disruption to Salado Salt. By analogy to criticality accidents, the potential damage to the salt from a fast
assembly of fissile mhterial is negligible. Specifically, the maximum energy release for accidents is -1020 fissions
(either moderated accidents from fissile solutions [Table VI] or unmoderated accidents with fissile metal [Table
VII]). This small amount of rapidly produced energy, released per event, would be unlikely to cause any significant
damage to the immediate rock situated 655 m (2100 ft) below the surface at the repository horizon. In rock of any
type, the potential void created would have a small radius. The radius of a potential spherical void (or camouflet)
formed beneath the surface (the void plus the debris) from an explosion is roughly r~ = 1.1053 w0x2 #35, where r~
is the radius (ft), d is depth below surface (ft), and w is weight of explosive (lbs) U. The explosive weight rather
than the energy from the explosive is used in the empirical expression because the exponent is so small (-1/4); the
type of rock is not used for rough estimates because it has a secondary effect. At mos~ potential voids (camouflets)
of <1.3 m radius would occur in the salt (42. 65), assuming the average accident energy release of 1018fissions (-4.5
kg TNT [10 lbs])~b
Disruption to Culebra Dolomite. If the average fissile content in the repository is less than 14%, then in order
for the percentage to approach the theoretical minimum of 35% required for an explosion u, p. 182), the fissile
content must more than double between the repository and its entry into the Culebra dolomite. Consequently, a
nuclear explosion is not feasible in the Culebra either. Furthermore, the potential damage to the dolomite from a
sudden assembly of fissile material is minor. For 4.5-kg TNT (1018fissions), the potential camouflet radius is only
slightly larger (1.4 m [4.8 ft]) at the shallower depth of 300 m than calculated in the Salado salt.
VILB.4. Heat Potentially Generated
Criticality Heat in Repository. The indefinite geometry of criticality in the salt after transport and precipitation
of plutonium (or uranium) requires making some arbitrary but reasonable assumptions. Here we assume a point
power source in a water-saturated geologic medium. For steady-state, the power is determined primarily by the
assumed operating temperature of the natural reactor and, secondarily, on the radial distance from the source to
where ambient temperatures are assumed. If a distance to the ambient temperature of 10 m is assumed (10 times the
typical thickness of the Ienticular natural reactors at Oklo) with a critical zone temperature of -700 K, then the
()steady-state power (Q.) is -40 kW (-1020 fissions/day) W, p. 248): Q. ~ 47cKAT ~ =40 kW, where r isr–a
radial distance to ambient temperature (10 m); a is diameter of critical sphere (0.75 m); AT is change in temperature
from surface of 0.75-m-diameter sphere to ambient temperature (700 -300 K); IXis thermal diffusivity = W&P; K is
u In 1961, the DOE detonated a 3.1-kiloton nuclear explosive 361 m below the surface in she Satado near Carlsbad, New Mexico w. Ahemispherical cavity of 22.6-m radius was created and increased the permeability in the satt about 100 m above the cavity. Such damage fromcriticality is not likely because of the nearly zero probability of a nuclear explosion being produced. The average fissile materiat enrichment ofWIPP TRU waste is only 14%, far below the theorcticat minimum of 35% enrichment required for an explosion w p. 182).
28 June 1,2000
thermal conductivity (5.4 W/m-K for salt at 300 K m and 4.3 W/m-K for dolomite/water at 300 K w, Table
5.2]); P is density; and CPis specific heat capacimnce.
The estimate of 40 kW for a 0.75-m-diameter sphere (compaction of room horn 84.8% porosity to 16% porosity
[Fig. 14]) or 23 kW/m2 is 2.5 orders of magnitude greater than the estimated power density of -0.1 kW/m2 at Oklo
~ and thus likely very conservative. This conservative estimate of heat power from criticality is only 7% of the
538 kW used for the EIS (31, 32) that resulted in an estimated 1.6°C risen in temperature at the center of the
repository u. Therefore, the heat generated during criticality is of no consequence.
Criticality Heat in Culebra. The maximum amount of plutonium carried to the Culebra in the 1996 PA is
110 kg (Fig. 17). The volume occupied by 110 kg of plutonium at a concentration of 3 kg/m3 is -37 m3. As noted
above, the amount of heat generated at Oklo has been estimated as 0.1 kW/m3 in the natural reactors w.
Consequently, the power produced.by a similar critical assembly in the Culebra would amount to only 3.6 kW.
Vlll. SUMMARY AND CONCLUSIONS
A critical event was omitted from consideration in the 1996 PA. The arguments were organized according to
the two aspects of risks-probability and consequences.
VIII.A. Probabilities of Criticality
Probability has been categorized by pertinent physical, hydrologic, and geochemical phenomena and with
regard to location within the disposal system. Prior to intrusion by exploratow drilling, a hydrologic constraint
exists in that the resaturation of the WIPP repository is not sufficient to move and assemble enough fissile material
for criticality. Furthermore, the fairly uniform distribution of material in the WIPP repository and the emplacement
of the fissile material far below the critical limit are two important physical constraints. The compaction of the
repository through salt creep is not sufficient to reach the minimum critical concentration. In addition, the uniform
distribution of material such as rust from corrosion of steel drums implies that preferential adsorption will not occur
at only one location in the repository and promote critical concentrations. Similarly, a fairly uniform material
distribution implies the brine chemistry is also fairly uniform such that changes in pH, ionic strength, or reduction-
oxidation potential do not vary and thus cannot promote preferential precipitation or colloidal filtering at only one or
a few locations.
After intrusion, two hydrologic constraints exist to prevent criticality in the Culebra. Firs4 the dispersion that
normally occurs while contaminants move through the Culebra aquifer suggests that the small amount of material
brought to the aquifer can reach a critical concentration only near the intrusion borehole. Second, the zone of
mixing fluid from the intrusion borehole and the Culebra is such that, at a minimum, 1000 kg must reach the
Culebra to provide conditions for criticality.
Several geochemical constraints exist to prevent criticality withh the disposal system. Firs4 the volubility of the
fissile material is insufficient to cause criticality. The low volubility of the fissile material also implies that vast
amounts of brine are required to transport a sufficient quantity of fissile material to cause criticality. In the
simulation for the 1996 PA of the WIPP, only one case out of 100 showed sut%cient material reaching the Culebra
27 me ~ 60C ~emwmmm fi~e fmm ~i~a~tive day is 16s thUII fhe 6°C temperature rise potentially caused by the exothemr.ic co~sion of
aluminum waste over 2 yr should sufficient brine be present in the repository(Q).
29 June 1,2000
to exceed the minimum fissile mass, and even then, the fissile material must be concentrated. Second, adsorption on
dolomite must substantially exceed the adsorptive capacity of synthetically prepared goethite in order to provide
sufficient adsorptive capacity for plutonium and similar actinides such as thorium and uranium. Because the fissile
material will be highly reduced within the repository, the traditional mechanism of forming uranium ore bodies
through precipitation by reduction of the more soluble high-oxidation-state actinides cannot occur. Furthermore,
neither a strong reductant nor oxidant in a localized area has been discovered withh the Culebrw thus, a sudden
change in the reduction-oxidation potential is unlikely. Therefore, any precipitation or coprecipitation would be
gradual and diffuse over an area too large to cause criticality.
VIII.B. Consequences of Criticality
Numerous procedures are put in place at the surface in a facility handling fissile radioisotopes to prevent
criticality and exposure of workers to radiation either directly during the events or later if decontamination of the
facility is necessary. However, prevention is less important after closure of a repository, assuming that criticality
could occur, because the criticality would have to either (1) generate significant amounts of fission products above
those already present or (2) degrade the ability of the disposal system to contain radioisotopes, before criticality
would have the potential to affect human health. Neither occurs. Criticality involving the most abundant fissile
material, ‘9Pu, actually decreases health hazards after 110 yr, based on the Containment Requirements criterion of
EPA’s standard, 40 CFR 191. The maximum number of fissions possible from uranium disposed of at the WIPP is
-1027 fissions This value is about an order of magnitude less than the equivalent amount of fission products
required to produce the heat power already present in the repository. Furthermore, the physical damage to the salt or
dolomite is negligible, Even if the average power released in a critical event were to be converted to an equivalent
explosive release, the damage to geologic media would not be significant. Finally, the amount of heat power
generated by decay of fission products from both plutonium and uranium fissions is also less (by an order of
magnitude) than the heat power initially present in the repository.
ACKNOWLEDGMENTS
This work was performed at Sandia National Laboratories, which is a muhiprogram laboratory operated by
Sandia Corporation, a Lockheed-Martin Company for the U.S. Department of Energy (DOE) under contract DE-
AC04-94AL8500. The authors wish to thank J.M. Chapman, S.K. Best, and D.J. Rivard of Tech Reps, Inc.,
Albuquerque, NM, for editorial, illustration, and word processing support, respectively. The authors also wish to
thank the reviewers at Sandia National Laboratories, Albuquerque, NM, M.S. Tiemey, M. Marten, P. Brady, and W.
Weart, for their helpful suggestions to the text. In addition, we wish to acknowledge several people who contributed
to various aspects of this article: R. Blaine, Ecodynamics Inc.; M. Wallace, The Plus Group; D.K. Rudeen, New
Mexico Engineering Research Institute (NMEFU);J.S. Rath, Sandia; and J.Il. Miller, Sandia.
REFERENCES
1. R.P. RECHARD, C.T. STOCKMAN, L.C. SANCHEZ, H.R. TRELLUE, J.S. RATH, and J. LISCUM-
POWELL, “RNT-l: Nuclear Criticality in Near Field and Far Field:’ FEP Screening Argument, Sandia
National Laboratories (1996).
2.
3.
4.
5.
6.
7.
8.
9.
10.
11.
12.
13.
14.
15.
16.
J.C. HELTON, J.E. BEAN, J.W. BERGLUND, F.J. DAVIS, K. ECONOMY, J.W. GARNER, J.D.
JOHNSON, R.J. MACKINNON, J. MILLER, D.G. OBRIEN, J.L. RAMSEY, J.D. SCHREIBER, A.
SHINTA, L.N. SMITH, D.M. STOELZEL, C. STOCKMAN, and P. VAUGHN. “Uncertainty and Sensitivity
Analysis Results Obtained in the 1996 Performance Assessment for the Waste Isolation Pilot Plant:’
SAND98-0365, Sandia National Laboratories (1998).
R.P. RECHARD, “Hktorical Background on Assessing the Performance of the Waste Isolation Pilot Plant:’
SAND98-2708, Sandia National Laboratories (1999).
“Compliance Certification Application for the Waste Isolation Pilot Plant:’ DOWCAO-1996-2184, United
States Department of Energy, Waste Isolation Pilot Plan4 Carlsbad Area Office, Vols. I-XXI (1996).
“Criteria for the Certification and Re-Certification of the Waste Isolation Pilot Plant’s Compliance With the
40 CFR Part 191 Disposal Regulations; Certification Decision: Final Rule: 40CFR194, Federal Register, 63,
27354 (1998).
“Department of Energy National Security and Military Applications of Nuclear Energy Authorization Act of
1980:’ Public Law 96-164,93 Stat. 1259.
R.P. RECHARD, “Milestones for Disposal of Radioactive Waste at the Waste Isolation Pilot Plant (WIPP) in
the United States,” SAND98-0072, Sandia National Laboratories (2000).
“Nuclear Waste Policy Act of 1982: Public Law 97-425,96 Stat. 2201; 42 U.S.C. 10101 et seq.
“Environmental Radiation Protection Standards for the Management and Disposal of Spent Nuclear Fuel,
High-Level and Transuranic Radioactive Wastes: Final Rule:’ 40CFR191, Federal Register, 58, 66398
(1993).
“Environmental Standards for the Management and Disposal of Spent Nuclear Fuel, High-Level and
Transuranic Radioactive Wastes: Final Rule;’40CFR191, Federal Register, 50,38066 (1985).
“waste Isolation Pilot Plant Land Withdrawal Act of 1992V Public Law 102-579, 106 Stat. 4777.
“criteria for the Certification and Re-Certification of the Waste Isolation Pilot Plant’s Compliance With the
40 CFR Part 191 Disposal Regulations: Final Rule;’40CFR194, Federal Register, 61,5224 (1996).
“Use of Probabilistic Risk Assessment Methods in Nuclear Regulatory Activities, Final Policy Statement:
Federal Register, 60,42622, U.S. Nuclear Regulatory Commission (1995).
R.P. RECHARD, “Historical Relationship Between Performance Assessment for Radioactive Waste Disposal
and Other Types of Risk Assessment:’ Risk Analysis, 19,763 (1999).
S. KAPLAN and B.J. GARRICK, “On the Quantitative Definition of Risk:’ Risk Analysis, 1,11 (1981).
J.C. HELTON, M.G. MARI171TA, and R.P. RECHARD, “Conceptual Structure of Performance Assessments
Conducted for the Waste Isolation Pilot Plant,” Proc. Symp. Scientific Basis for Nuclear Waste Management
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
27.
XVI, Materials Research Society, Boston, Massachusetts, November 30-December 4, 1992, C.G. Interrante
and R.T. Pabalan, Eds, Pittsburgh, Pennsylvania, 294, p. 885 (1993).
R.P. RECHARD, “An Introduction to the Mechanics of Performance Assessment Using Examples of
Calculations Done for the Waste Isolation Pilot Plant Between 1990 and 1992~ SAND93-1378, Sandia
National Laboratories (1995).
R.P. RECHARD, “CAMCON Computer System for Assessing Regulatory Compliance of the Waste
Isolation Pilot PlanL” Probabilistic Safety Assessment and Management, Proceedings of the International
Conference on Probabilistic Safety Assessment and Management (PSAM), Beverly Hills, California, February
4-7, 1991, G. Apostokikis, Ed., SAND90-2094C, New York, ~ Elsevier Science Publishers. 2, p. 899
(1991).
R.P. RECHARD, Ed., “User’s Reference Manual for CAMCON. Compliance Assessment Methodology
Controller, Version 3.0~’ SAND90-1983, Sandia National Laboratories (1992).
“Final Rule for High-Level and Transuranic Radioactive Wastesfl Background Information Document, EPA
520/1-85-023. Washington, DC: Environmental Protection Agency, Office of Radiation Programs (1985).
L.C. SANCHEZ, J.S. RATH, R. AGUILAR, H.R. TRELLUE, K. COCHRANE, L.L. TAYLOR, and J.R.
WILSON, “Nuclear Dynamics Consequence Analysis (NDCA) for the Disposal of Spent Nuclear Fuel in an
Underground Geologic Repository, Vols. l–3~’ SAND98-2208; INEEUEXT-098-00996; DOE/SNF/REP-
0033, Assistant Secretary for Environmental Management, U.S. Department of Energy (1998).
R.P. RECHARD, L.C. SANCHEZ, C.T. STOCKMAN, and H.R. TRJ3LLUE, “Consideration of Nuclear
Criticality When Disposing of Transuranic Waste at the Waste Isolation Pilot Plant: SAND99-2898, Sandia
National Laboratories (2000).
E.J. ALLEN, “Criticality Analysis of Aggregations of Actinides from Commercial Nuclear Waste in
Geological Storage: ORNL/TM-6458, Oak Ridge National Laboratory (1978).
J.W. WINCHESTER, “Long Term Geochemical Interactions of High Active Waste with Crystalline Rock
Repository Media:’ Scientific Basis for Nuclear Wrote Management, Proc. Symp. Science Underlying
Radioactive Waste Management, Materials Research Society Annual Meeting, Boston, Massachusetts,
November 28-December 1, 1978, G.J. McCarthy, Ed., New York, NY: Plenum Press. Vol. 1, p. 345 (1979).
E.D. CLAYTON, “Anomalies of Nuclear Criticality: PNL-SA-4868, Rev. 5; CONF791 103-114, Rev. 5;
Pacific Northwest Laboratory (1979).
R.D. CARTER, “Appendix N, Criticality Calculations:’ Nuclear Reactivity Evaluations of 216-L9 Enclosed
Trench, A.E. Smith, Ed., ARH-2915, Atlantic Richfield Hanford Company (1973).
D.G. BROOKINS, “Geochemical Constraints on Accumulation of Actinide Critical Masses from Stored
Nuclear Waste in Natural Rock Repositories; ONWI-17, Office of Nuclear Waste Isolation (ONWI), Battelle
Memorial Institute (1978).
.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
W.R. STRATTON, ‘The Myth of Nuclear Explosions at Waste Disposal Sites:’ LA-9360, Los Alamos
National Laboratory (1983).
F.W. BINGHAM and G.E. BARR, “Scenarios for Long-Term Release of Radionuclides from a Nuclear-
Waste ‘Repository in the Los Medaiios Region of New Mexico:’ SAND78-1730, Sandia National
Laboratories (1979).
F.W. BINGHAM and G.E. BARR, “Development of Scenarios for the Long-Term Release of Radionuclides
from the Proposed Waste Isolation Pilot Plant in Southeastern New Mexico,” Scienttj5c Basis for Nuclear
Waste Management, Proc. Symp. Science Underlying Radioactive Waste Management, Materials Research
Socie~ Annual Meeting, Boston, Massachusetts, November 27–December 30, 1979, C.J.M. Northrup, Jr.,
Ed., SAND79-0955C, New York, ~ Plenum Press. Vol. 2, p. 771 (1980).
“Draft Environmental Impact Statement, Waste Isolation Pilot Plant; DOEJEIS-0026-D, U.S. Department of
Energy, Vols. 1-2 (1979).
“Final Environmental Impact Statement, Waste Isolation Pilot Plant:’ DOE/EIS-0026, U.S. Department of
Energy, Vols. 1-2 (1980).
“safety Assessment for the Underground Disposal of Radioactive Wastes; Safety Series No. 56,
International Atomic Energy Agency (1981).
M.H. LIPNER, and J.M. IL4VETS, “Nuclear Criticality Safety Analyses for the Waste Isolation Pilot Plant
Project: WAES-TME-3025, Westinghouse Electric Corporation, Advanced Energy Systems Division
(1980),
W.A. BLYCKERT, and R.D. CARTER, “Criticality Parameters of 55-Gallon Waste Drum Arrays,” RHO-
SA-183; CONF-801 107-58, Rockwell International Corporation, Rockwell Hanford Operations (1980).
S.C. COHEN, “Supplement 2 Criticality Review for the Waste Isolation Pilot PkmL” Review of the WIPP
Draft Application to Show Compliance with EPA Transuranic Waste Disposal Standards, EEG-61;
DOWALJ58309-61, Environmental Evaluation Group S2-1 through S2-26 (1996).
S.C. COHEN, “Supplement 3: Reviews of WAESD-TR-83-O015 and TME-3025 Rev. l;’ Review of the
WIPP Draft Application to Show Compliance with EPA Transuranic Waste Disposal Standards, EEG-61;
DOE/AL/58309-61, Environmental Evaluation Group S3-1 through S3-3 (1996).
S.C. COHEN, “Supplement 4: Letter Dated January 18, 1984 to Environmental Evaluation Group on
Postulated WIPP Reconcentration Criticality: Review of the WIPP Draft Application to Show Compliance
with EPA Transuranic Waste Disposal Standards, EEG-61; DOWAIJ58309-61, Environmental Evaluation
Group S4-1 through S4-10 (1996).
W.R. STRA’ITON, “A Review of Criticality Accidents: LA-361 1, Los Ahunos Scientific Laboratory
(1967).
40.
41.
42.
43.
44.
45.
46.
47.
48.
49.
50.
51.
52.
53.
W.R. STRATTON and D.R. SMITH, “A Review of Criticality Accidents: DOWNCT-04, U.S. Department
of Energy (1989).
R.A. KNIEF, “Nuclear Criticality Safety: Theory and Practice:’ American Nuclear Society, 17-28 (1985).
R.P. RECHARD, M.S. TIERNEY, L.C. SANCHEZ, and M.A. MARTELL, “Consideration of Criticality
when Directly Disposing Highly Enriched Spent Nuclear Fuel in Unsaturated Tuft Bounding Estimates,”
SAND96-0866, Sandia National Laboratories (1996).
G.A. COWAN, “A Natural Fission Reactor: Scientific American. 235,36 (1976).
R.L. HUNTER, “Events and Processes for Constructing Scenarios for the Release of Transuranic Waste from
the Waste Isolation Pilot Plant, Southeastern New Mexico: SAND89-2546, Sandia National Laboratones
(1989).
D.A. GALSON and P.N. SWIFP, “Recent Progress in Scenario Development for the WIPP~’ High Level
Radioactive Waste Management Proceedings of the Sixth Annual International Conference, Las Vegas,
Nevada, April 30-May 5, 1995, SAND95-0117C, American Nuclear Society New York, N.Y. American
Society of Civil Engineers, p. 391 (1995).
G. TAUBES, “Blowup at Yucca Mountain:’ Science. 268,1836 (1995).
C.D. BOWMAN and F. VENNERI, “Underground Supercriticality from Plutonium and Other Fissile
Material;’ Science & Global SecuriV, 5,279 (1996).
C.G. WHIPPLE, “Can Nuclear Waste Be Stored Safely at Yucca Mountain?; Scientific American, 274, 72
(1996).
G.H. CANAVAN, S.A. COLGATE, O.P. JUDD, G.H. MCCALL, A.G. PETSCHEK, J.C. SOLEM, T.F.
STRATTON, W.R. STR4’ITON, and P.P. WHALEN, Comments on “Nuclear Excursions” and “Criticality
Issues: LA-UR-95-0851, Los Alamos National Laboratory (1995).
R.A. VAN KONYNENBURG, “Comments on the Draft Paper ‘Underground Supercnticrdity from
Plutonium and Other Fissile Material: written by C.D. Bowman and F. Venneri (LANL)fl Science and
Global Security, 5,303 (1996).
R. SANCHEZ, W. MYERS, D. HAYES, R. KIMPLAND, P. JAEGERS, P. PATERNOSTER, S. ROJAS, R.
ANDERSON, and W. STIL4TTON, “Criticality Characteristics of Mixtures of Plutonium, Silicon Dioxide,
Nevada Tuff, and Water; LA-UR-95-2130, Los Alamos National
Received from Authors, WPO 31283.) (1995).
R.H. KIMPLAND, “Dynamic Analysis of Nuclear Excursions in
Plutonium’ Science and Global Security, 5,323 (1996).Q
W.E. KASTENBERG, P.F. PETERSON, J. AHN, J. BURCH,
Laboratory (November 3,
Underground Repositories
1995 Draft
Containing
G. CASHER, P.L. CHAMBI&, E.
GREENSPAN, D.R. OLANDER, J.L. VUJIC, B. BESSINGER, N.G.W. COOJL F.M. DOYLE, and L.B.
34 June 1,2000
54.
55.
56.
57.
58.
59.
60.
61.
62.
63.
64.
HILBERT, Jr., “Considerations of Autocatalytic Criticality of Fissile Materials in Geologic Repositones~’
Nuclear Technology, 115,298 (1996).
A.R. LAPPIN, R.L. HUNTER, D.P. GARBER, and P.B. DAVIES, Eds., “Systems Analysis, Long-Term
Radionuclide Transport, and Dose Assessments, Waste Isolation Pilot Plant (WIPP), Southeastern New
Mexico; March 1989~’SAND89-0462, Sandia National Laboratories (1989).
R.M. HOLT, “Conceptual Model for Transport Processes in the Culebra Dolomite Member, Rustler
Formation; SAND97-0194, Sandia National Laboratories (1997).
J.W. MERCER, and B.R. ORR, “Interim Data Report on the Geohydrology of the Proposed Waste Isolation
Pilot Plant Site, Southeast New Mexicoj’ Water-Resources Investigations Report 79-98, U.S. Geological
Survey, 92 (1979).
L.C. MEIGS, R.L. BEAUHEIM, J.T. MCCORD, Y.W. TSANG, and R. HAGGERTY, “Design, Modelling,
and Current Interpretations of the H-19 and H-11 Tracer Tests at the WIPP Site; Field Tracer Experiments:
Role in the Prediction of Radionuclide Migration: Synthesis and Proceeding of an NEA/EC GEOT~P
Workshop, Cologne, Germany, August 28-30, 1996, SAND96-2796C, Nuclear Energy Agency, Organisation
for Economic Co-Operation and Development p. 157 (1997).
J.F. BRIESMEISTER, Ed., “MCNR A General Monte Carlo Code for Neutron and Photon Transport:’
Version 3a, LA-7396-M, Rev. 2 Manual, Los Alamos National Laboratory (1986).
H.C. PAXTON, J.T. THOMAS, D. CALLIHAN, and E.B. JOHNSON, “Critical Dimensions of Systems
Containing 235U, 239Pu, and 233U, TID-7028Y Los Alamos Scientific Laboratory (1964).
E.D. CLAYTON and W.A. REARDON, “Chapter 27: Nuclear Safety and Criticality of Plutonium~’
Plutonium Handbook: A Guide to the Technology, O.J. Wick, Ed., La Grange Park, Illinois, American
Nuclear Society, 1,875 (1980).
H.W. PAPENGUTH, “Attachment A Rationale for Definition of Parameter Values for Actinide Intrinsic
Colloids.” Sandia National Laboratories (In Parameter Record Package WPO 35852: Mobile-Colloidal-
Actinide Source Term. 2. Actinide Intrinsic Colloids.) (1996).
M.D. SIEGEL, S.J. LAMBERT, and K.L. ROBINSON, Eds., “Hydrogeochemical Studies of the Rustler
Formation and Related Rocks in the Waste Isolation Pilot Plant Area, Southeastern New Mexico:’ SAND88-
0196, Sandia National Laboratories (1991).
T. SEWARDS, A. BREARLEY, R. GLENN, I.D.R. MACKINNON, and M.D. SIEGEL, “Nature and
Genesis of Clay Minerals of the Rustler Formation in the Vicinity of the Waste Isolation Pilot Plant in
Southeastern New Mexicoy SAND90-2569, Sandia Nationrd Laboratories (1992).
F. GAUTHIER-LAFAYE and F. WEBER, “The Francevillian (Lower Proterozoic) Uranium Ore Deposits of
Gabom” Economic Geology and the Bulletin of the Society of Economic Geologists, 84,2267 (1989).
35 June 1,2000
...- . ..,. -— --- --- ---- ---- ,.-mr.~a.y-. .. ...~-- .,.- ._ , -. -.~... ;,-z,.,:>..
.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77.
R.P. RECHARD, M.S. TIERNEY, L.C. SANCHEZ, and M-A. MARTELL, “Bounding Estimates for Critical
Events when Directly Disposing Highly Enriched Spent Nuclear Fuel in Unsaturated Tuff~’ Risk Analysis, 17,
19 (1997).
“Preliminary Comparison with 40 CFR Part 191, Subpart B for the Waste Isolation Pilot PlanL December
1991. Volume 3fl Reference Data, R.P. Rechard, A.C. Peterson, J.D. Schreiber, H.J. Iuzzolino, M.S. Tiemey,
and J.S. Sandha, Eds., SAND91-0893/3, Sandia National Laboratories (1991).
R.P. RECHARD, W. BEYELER, R.D. MCCURLEY, D.K. RUDEEN, J.E. BEAN, and J.D. SCHREIBER,
“Parameter Sensitivity Studies of Selected Components of the Waste Isolation Pilot Plant Repository/Shaft
System:’ SAND89-2030, Sandia National Laboratories, 11-30(1990).
V.A. KELLEY and J.F. PICKENS, “Interpretation of the Convergent-Flow Tracer Tests Conducted in the
Culebra Dolomite at the H-3 and H-4 Hydropads at the Waste Isolation Pilot Plant (WIPP) Site:’ SAND86-
7161, Sandia National Laboratories (1986).
R.L. IMAN, W.J. CONOVER, “Small Sample Sensitivity Analysis Techniques for Computer Models, With
an Application to Risk Assessment” Communications in Statistics, A9, 1749 (1980).
K.M. TRAUTH, S.C. HORA, R.P. RECHARD, and D.R. ANDERSON, “The Use of Expert Judgment to
Quantify Uncertainty in Solubility and Sorption Parameters for Waste Isolation Pilot Plant Performance
Assessment” SAND92-0479, Sandia National Laboratories (1992).
R. WEINER, “Documentation Package for Oxidation State Distribution of Actinides in the Repository: SNL
Technical Memorandum, March 27, 1996, Sandia National Laboratories, (On file in SWCF as WPO 35194.)
(1996).
S.C. BABB and C.F. NOVAK, “User’s Manual for FMT, Version 2.0,” Sandia National Laboratories (On file
in Sandia WIPP Central Fdes [SWCFJ as WPO 28 119.) (1995).
D.L. LANGMUIR, “Aqueous Environmental Geochemistry;’ Upper Saddle River, NJ Prentice Hall (1997).
P.V. BRADY, H.W. PAPENGUTH, and J.W. KELLY, “Metal Sorption to Dolomite Surfaces;’ Applied
Geochemistry, 14,569 (1999).
H.W. PAPENGUTH and Y.K. BEHL, “Test Plan: For Evaluation of Colloid Facilitated Actinide Transport.”
TP 96-01, Sandia National Laboratories (1996).
H.W. PAPENGUTH and R. AGUILAR, “Attachment A. Rationale for Definition of Parameter Values for
Mineral Fragment Type Colloids.” Sandia National Laboratories (h Parameter Record Package 35850:
Mobile-Colloidal-Actinide Source Term Parameters. 1. Mineral Fragment Colloids.) (1996).
H.W. PAPENGUTH and R.C. MOORE, “Attachment A Rationale for Definition of Parameter Values for
Humic Substances: Sandia National Laboratories (In Parameter Record Package WPO 35855: Mobile-
Colloidal-Actinide Source Term. 3. Humic Substances.) (1996).
. .
78.
79.
80.
81.
82.
83.
84.
85.
86.
87.
88.
89.
H.W. PAPENGUTH, “Attachment A Rationale for Definition of Parameter Values for Microbes:’ Sandia
National Laboratories (In Parameter Record Package WPO 35856: Mobile-Colloidal-Actinide Source Term.
4. Microbes.) (1996).
H.H. ADLER, “Concepts of Uranium-Ore Formation in Reducing Environments in Sandstones and Other
Sediments: Proc. Symp. Formation of Uranium Ore Deposits, Athens, Greece, May 6-10, 1974,
STVPUEV374,International Atomic Energy Agency, p. 141 (1974).
J.W. BRINCK, “The Geochemical Distribution of Uranium as a Primary Criterion for the Formation of Ore
Deposits: Proc. Symp. Formation of Uranium Ore Deposits, Athens, Greece, May 6-10, 1974,
STUPUW374, International Atomic Energy Agency, p. 21 (1974).
A.E. SAUCIER, “Tertiary Oxidation in the Westwater Canyon Member of Morrison Formation: Geology
and Mineral Technology of the Grants Uranium Region 1979, Comp. C.A. Rautman, Memoir 38, New
Mexico Bureau of Mines&Mineral Resources, 116 (1980).
D.R. SHAWE “Significance of Roll Ore Bodies in Genesis of Uranium-Vanadium Deposits on the Colorado
Plateau;’ Proc. International Conference on the Peacefil Uses of Atomic Energy, Geneva, Switzerland,
August 8-20, 1955, Volume 6: Geology of Uranium and Thorium, New York, NY: United Nations, p. 335
(1956).
D.L. LANGMUIR, “Uranium Solution-Mineral Equilibria at Low Temperatures with Applications to
Sedimentary Ore Deposits: Geochimica et Cosmochimica Acts, 42,547 (1978).
K.S. JOHNSON, “Stratigraphy and Mineral Resources of Guadalupian and Ochoan Rocks in the Texas
Panhandle and Western Oklahoma: Geology and Mineral Deposits of Ochoan Rocks in Delaware Basin and
Adjacent Areas, Comp. G.S. Austin, Circular 159, New Mexico Bureau of Mines & Mineral Resources, 57,
(1978).
Z. AL-SHAIEB, R.W. OLMSTED, J.W. SHELTON, R.T. MAY, R.T. OWENS, and R.E. HANSON,
“Uranium Potential of Permian and Pennsylvanian Sandstones in Oklahoma” The American Association of
Petroleum Geologists Bulletin, 61,360 (1977).
K.G. BELL, “Uranium in Carbonate Rocks; Professional Paper 474-A, United States Government Printing
Office (1963).
R.Y. ANDERSON and D.W. KIRKLAND, “Ongen, Varves, and Cycles of Jurassic Todilto Formation, New
Mexico: Bulletin of the American Association of Petroleum Geologists, 44,37 (1960).
G.A. COWEN, E.A. BRYANT, and W.R. DANIELS, “Some United States Studies of the Oklo
Phenomenon; Proc. Symp. Le Phenomene d’Oklo/The Oklo Phenomenon, Libreville, Gabon, June 23-27,
1975, STI-PUB-405, International Atomic Energy Agency, p. 341 (1975).
F. GAUTHIER-LAFAYE, F. WEBER, and H. OHMOTO, “Natural Fission Reactors of Oklo:’ Economic
Geology and the Bulletin of the Socie~ of Economic Geologists, 84,2286 (1989).
90.
91.
92.
93.
94.
95.
96.
97.
98.
99.
100.
101.
J.W. MERCER, “Geohydrology of the Proposed Waste Isolation Pilot Plant Site, Los Medaiios Area,
Southeastern New Mexicofl Water-Resources Investigations Report 83-4016, Albuquerque, NM: U.S.
Geological Survey (1983).
G.A. COWEN and A.E. NORRIS, Eds., “Investigations of the Natural Fission Reactor Program October
1977-September 1978V LA-7536-PR, Los Alamos Scientific Laboratory (1978).
C-K.D. HSI and D. LANGMUIR, “Adsorption of Uranyl onto Ferric Oxyhydroxides: Application of the
Complexation Site-Binding Model:’ Geochimica et Cosmochbnica Acts, 49,1931 (1985).
R.W. SMITH, J.C. WALTON, and M. RAHMAN, “Effects of Recrystallization on Time Variant Sorption of
Radionuclides onto Steel Corrosion Products,” Proc. Symp. Scierztij7cBasis for Nuclear Waste Management
XVH, Materials Research $%ciety, Boston, Massachusetts, November 29–December 3, 1993, A. Barkatt and
R.A. Van Konynenburg, Eds., Pittsburgh, PA Materials Research Society, 333, p. 713 (1994).
T.E. PAYNE, G.R. LUMPKIN, and T.D. WAITE, “Chapter 2 UraniumVI Adsorption on Model Minerals:
Controlling Factors and Surface Complexadon Modeling,” Adsorption of Metals by Geomedia: Variables,
Mechanisms, and Model Applications. E.A. Jenne, Ed., San Diego, Califomi& Academic Press, 75 (1998).
A.L. SANCHEZ, J.W. MURRAY, and T.H. SIBLEY, “The Adsorption of Plutonium IV and V on GoethiteY
Geochimica et Cosmochimica Acts, 49,2297 (1985).
I.R.TRIAY, A. MEIJER, J.L. CONCA, K.S. KUNG, R.S. RUNDBERG, B.A. STRTETELMEIER, and C.D.
TAIT, “Summary and Synthesis Report on Radionuclide Retardation for the Yucca Mountain Site
Characterization Project” Yucca Mountain Site Characterization Program Milestone 3784M, LA-13262-MS,
Los Alamos National Laboratory (1997).
S. BRUNAUER, P.H. EMMETT, and E. TELLER, “Adsorption of Gases in Multimolecular Layers,” Journal
of the American Chemical Socie~, 60,309 (1938).
J.W. KELLY, ‘Surface Area Analysis Results: Conceptual Model for Transport Processes in the CWebra
Dolomite Member, Rustler Formation, R.M. Holt, SAND97-0194, Sandia National Laboratories, C-4 through
C-5 (1996).
J.A. DAVIS and D.B. KENT, “Surface Complexation Modeling in Aqueous Geochemistry:’ Mineral-Water
Inte@ace Geochemistry, M.F. Hochella, Jr. and A.F. White, Eds., Reviews in Mineralogy, 23, Mineralogical
Society of America, 177 (1990).
D.B. KENT, V.S. TRIPATHI, N.B. BALL, J.O. LECKI, and M.D. SIEGEL, “Surface-Complexation
Modeling of Radionuclide Adsorption in Subsurface Environments:’ SAND86-7175; NUREG/CR-4807,
Sandia National Laboratories (1988).
J-C. SAMAMA, “Ore Fields and Continental Weathering? New York, ~ Van Nostrand Reinhold
Company, 244 (1986).
102. J.C. HELTON, D.R. ANDERSON, B.L. BAKER, J.E. BEAN, J.W. BERGLUND, W. BEYELER, J.W.
GARNER, H.J. IUZZOLINO, M.G. MARIETTA, R.P. RECHARD, P.J. ROACHE, D.K. RUDEEN, J.D.
SCHREIBER, P.N. SWIFT, M.S. TIERNEY, and P. VAUGHN, “Effect of Alternative Conceptual Models in
a Preliminary Performance Assessment for the Waste Isolation Pilot PlanL” Nuclear Engineen.ng and Design,
154,251 (1995).
103. R.L. BEAUHElM, and R.M. HOLT, “Hydrogeology of the WIPP Site: Geological and Hydrological Studies
of Evaporates in the Northern Delaware Basin for the Waste Isolation Pilot Plant (WFP), New Mexico, Field
Trip #14, Guidebook, November 1-4, 1990, Leaders D. Powers, R. Holg R.L. Beauheim, and N. Rempe,
SAND90-2035J, Dallas Geological Society, 131 (1990).
104. L.H. BRUSH, “Ranges and Probability Distributions of Kds for Dissolved Pu, Am, U, Th, and Np in the
Culebra for the PA Calculations to Support the WIPP CCA~’ Laboratory Column Experiments for
Radionuclide Adsorption Studies of the Culebra Dolomite Member of the Rustler Formation, D.A. Lucero,
G.O. Brown, and C.E. Heath, SAND97-1763, Sandia National Laboratones, D-11 through D-108 (1998).
105. L.H. BRUSH, and L.J. STORZ, “Revised Ranges and Probability Distributions of Kds for Dissolved Pu, Am,
U, Th, and Np in the Culebra for the PA Calculations to Support the WIPP CCA: Memorandum to M.S.
Tierney, July 24,1996, Sandia National Laboratories (On file in SWCF as WPO 41561.) (1996)
106. F.W. WALKER, J.R. BARRINGTON, and F. FEINER, Nuclides and Zsotopes: Chart of the Nuclides, 14th
cd., San Jose, California, General Electric Company (1989).
107. R.P. RECHARD, Ed., “Update to Assessment of Direct Disposal in Unsaturated Tuff of Spent Nuclear Fuel
and High-Level Waste Owned by U.S. Department of Energy,”. SAND98-0795; DOE/SNF/REP-015;
INEEL/EXT-98-001 85, U.S. Department of Energy, Assistant Secretary for Environmental Management
under DOE Idaho Operations OffIce (1998).
108. H.A. BETHE, and J.H. TAIT, “An Estimate of the Order of Magnitude of the Explosion When the Core of a
Fast Reactor Collapses:’ U.S.-U.K. Reactor Hazard Meeting, 1956, RHM(56)/1 13, Harwell, Berks, England:
Atomic Energy Research Establishment p. 22 (1956).
109. R.L. MURRAY, Nuclear Reactor Physics, Englewood Cliffs, New Jersey: Prentice-Hall, Inc., 182 (1957).
110. R.P. RECHARD, Ed., “Performance Assessment of the Direct Disposal in Unsaturated Tuff of Spent Nuclear
Fuel and High-Level Waste Owned by U.S. Department of Energy:’ Volume 2 Methodology and Results,
SAND95-2563/2, Sandia National Laboratories (1995).
111. W.E. BAKER, J.J. KULESZ, P.S. WESTINE, P.A. COX, and J.S. WILBECK, “Manual for the Prediction of
Blast and Fragment Loadings on Stmctures: DOWI’IC-1 1268, San Antonio, TX Southwest Research Center
for Albuquerque Operations Office, U.S. Department of Energy, (1980).
112. D. RAWSON, C. BOARDMAN, and N. JAFFE-CHAZAN, “Project Gnome The Environment Created by a
Nuclear Explosion in Salt:’ PNE-107F, Lawrence Radiation Laboratory, University of California (1965).
113.
114.
115.
116.
117.
118.
119.
120.
121.
122.
123.
124.
H.S. CARSLAW, and J.C. JAEGER, Conduction of Heat in Solids, 2nd cd., Oxford: Clarendon Press; New
York, ~ Oxford University Press, 248 (1959).
J.N. SWEET, and J.E. MCCREIGHT, “Thermal Conductivity of RockSalt and Other Geologic Materials from
the Site of the Proposed Waste Isolation Pilot Plant:’ SAND79-1665, Sandia National Laboratories (1980).
F.F. SABINS, Jr., Remote Sensing: Principals and Interpretation, San Francisco, CA, W.H. Freeman and
Company (1978).
B.J. THORNE and D.K. RUDEEN, “Regional Effects of TRU Repository HeaL” SAND79-7161, Sandia
National Laboratories (1981).
H.W. PAPENGUTH and Y.K. BEHL, ‘“TestPlan Evaluation of Dissolved Actinide Retardation at the Waste
Isolation Pilot Plant:’ TP 96-02, Sandia National Laboratories (1996).
L.C. SANCHEZ, J. LISCUM-POWELL, J.S. RATH, and H.R. TRELLUE, “EPAUNI: Analysis Package for
Estimating Probability Distribution of EPA Unit Loading in the WIPP Repository for Performance
Assessment Calculations;’ Sandia National Laboratories (On file in SWCF as WPO#43898.) (1997).
‘Tlansuranic Waste Baseline Inventory Report (Revision 2):’ DOIYCAO-95-1 121, U.S. Department of
Energy, Carlsbad Area Office, (1995).
M.L. WILSON, J.H. GAUTHIER, R.W. BARNARD, G.E. BARR, H.A. DOCKERY, E. DUNN, R.R.
EATON, D.C. GUERIN, N. LU, M.J. MARTINEZ, R. NILSON, C.A. RAUTMAN, T.H. ROBEY, B. ROSS,
E.E. RYDER, A.R. SCHENKER, S.A. SHANNON, L.H. SKINNER, W.G. HALSEY, J.D. GANSEMER,
L.C. LEWIS, A.D. LAMONT, I.R. TRIAY, and A.S. MEIJER, “Total System Performance Assessment for
Yucca Mountain-SNL Second Iteration (TSPA-1993)7° SAND93-2675, Sandia National Laboratories,
Executive Summary, 1-2 (1994).
B.W. LESLIE, E.C. PEARCY, and J.C. PRIKRYL, “Oxidative Alteration of Uraninite at Noprd I Deposit,
Mexico: Possible Contaminant Transport and Source Term Constraints for the Proposed Repository at Yucca
Mountainfl Proc. Symp. Scientific Basis for Nuclear Waste Management XVI, Materials Research Socie@,
Boston, Massachusetts, November 30-December 4, 1992, C.G. Interrante and R.T. Pabalan, Eds., Pittsburgh,
PA Materials Research Society, 294, p. 505 (1993).
R.J. LEMIRE, and F. GARISTO, “The Volubility of U, Np, Pu, Th and Tc in a Geological Disposal Vault for
Used Nuclear Fuel:’ AECL-1OOO9, Pinaw% Manitoba Atomic Energy of Canada Limited, Whheshel]
Nuclear Research Establishment (1989).
J.J. CRAMER, and J.A.T. SMELLIE, Eds., “Final Report of the AECI..JSKB Cigar Lake Analog Study;’
AECL-10851, COG-93-147, SKB TR 94-04, Pinawa, Manitobx Atomic Energy of Canada Limited,r,
Whiteshell Nuclear Research Establishment (1994).
C.H. HO and N.H. MILLER, “Absorption of Uranyl Species from Bicarbonate Solution onto Hematite
.
125. R.C. WEAST, Ed., CRC Handbook of Chemistry and Physics, 55th cd., Cleveland, OH CRC Press, B-193 -
B-195 (1974).
126. L.H. BRUSH, “Test Plan for Laboratory and Modeling Studies of Repository and Radionuclide Chemistry for
the Waste Isolation Pilot Plan:’ SAND90-0266, Sandia National Laboratories (1990).
127. V.V. FROLOV, B.G. RYAZANOV, V.I. SVIRIDOV, and G.S. STARODUBTSEV, “A Review of Criticality
Accidents Which Occurred in the Russian Industry:’ ICNC ’95: The Fif/h International Conference on
Nuclear Criticali~ Safety, Albuquerque, New Mexico, September 17-21, 1995, University of New Mexico,
Department of Chemical and Nuclear Engineering, 1, p. 23 (1995).
128. H.C. PAXTON, “A History of Critical Experiments at Pajarito Site,” LA-9685-H, Los Alamos National
Laboratory (1983).
Fig. 1. Physical setting of the WIPP (after Ref. 17, Figure 2.1-1).
Fig. 2. Human intrusion scenario, S5, considered in 1996 WIPP PA (after Ref. 17, Figure 3.2+.
Fig. 3. The influence of criticality on the risk performance measure (here depicted as the expected value of theconsequence) depends on both the probability and consequence of the occurrence. In turn, evaluation of either theprobability or consequence depends on two types of simulation: nuclear criticality modeling and geophysicalmodeling.
Fig. 4. Stratigraphy above and below the WIPP repository (Ref. 3, Figure 3-3).
Fig. 5. Anticipated volumes of contact-handled transuranic (CH-TRU) waste per site (data from Ref. 119).
Fig. 6. The WIPP repository in bedded salt, showing transuranic waste disposal area with eight panels, 657 m belowsurface and 436 m below Culebra Dolomite Member of Rustler Formation (Ref. 17, Figure 2.0).
Fig. 7. Comparison of calculated and measured critical masses in a homogeneous, spherical shape as a function offissile solid concentration when mixed and reflected with water. Calculated results from MCNPm (after Ref. 21,Figures E.3-X1, E.3-X2, and E.3-X3): (a) measured data for 93.2 %wt ‘5U from Ref. 59, Figure 8, and (b) measureddata and analytically corrected for 100 %wt ‘9Pu from Ref. 60, Figure 27.7 and Ref. 59, F@re 27.
Fig. 8. Element concentrations in Culebra, Salado, and Castile brine (Ref. 117, Table 2).
Fig. 9. Percentages by weight of various element oxides that form after heating the Culebra dolomite and Oklosandstone (Ref. 63, Table IV-4, WIPP-12 at 836.2 ft (246.7 m) depth Ref. 89, Table 3, near natural reactors wheresilica has not been extensively dissolved).
~9Pu concentration when mixedFig. 10. Calculated critical masses of 100 %wt ‘9Pu and ‘9Pu02 or as a function ofwith halite, Culebra dolomite, and various brines (Ref. 22, Figure 16): (a) Culebra brine and (b) Salado and Castilebrines. (The material surrounding the spherical shape is the same as in the sphere but without fissile material.)
Fig. 11. Critical masses of ’51J as a function of ‘5U concentration when mixed with various geologic media (Ref.22, Figure 17): (a) Culebra dolomite and (b) Oklo sandstone and Yucca Mountain volcanic tuff.
Fig. 12. Calculated critical mass versus solid concentration of homogeneous solutions of fissile metallic ‘5U and‘9Pu in hemispherical shape of water and goethite at 20% porosity (Ref. 21, Figure E.3-26; Ref. 22, Figure 18).
Fig. 13. Critical thickness of plates of ‘9Pu as a function of ‘9Pu concentration in the Culebra (Ref. 22, Figure 19).Measured data from Ref. 59, Figure 30.
Fig. 14. Volume-average porosity in room (Ref. 66, Figure 5-20; Ref. 22, Figure 20)
Fig. 15. Cumulative brine flow up an intrusion borehole over 10,000 yr after an intrusion that occurs 1000 yr afterclosure for the S5 scenario (Ref. 22, Figure 14).
Fig. 16. Assumed total concentration of fissile material in Culebra for (a) plutonium and (b) uranium.
Fig. 17. Cumulative mass of ‘9Pu reaching the Culebra through intrusion borehole over 10,000 yr for S2, S3, S5,S6 for replicate 1 (Rl) (Ref. 22, Figure 22).
Fi .18. Variation of adsorbed radioisotopes on iron oxides as a function of solution concentration at neutral pH (a)$U in pure H20 on ferrihydrite (Ref. 94, Figure 3) and hematite (Ref. 95, Figure 1), (b) Puw in pure HzO (Ref. 95,
Figure 1) and Puv in J-13 well water from Yucca Mountain (Ref. 96, Figure 84) on goethite.
Fig. 19. Anticipated activity and thermal power of waste in the WIPP repository (after Ref. 118, as evaluated fromRef. 119). Activity of 2100 metric tonnes heavy metal (MTHM or Mg) DOE-owned spent nuclear fuel at Hanford
42 June 1,2000
and 8036 equivalent MTHM (as calculated by method in 40 CFR 191) high-level waste (DSNF/HLW) (primarilyfrom Savannah River Plant) shown for comparison (after Ref. 110, Figure 4-1 1).
Fig. 20. Calculated mixing of Castile brine with Culebra brine at intrusion borehole using discharge and flow fieldinformation horn S5, vector 23: (a) numerical mesh, (b) concentration contours, and (c) dishibution of mass andaverage density using various adsorption assumptions (Ref. 22, Figure 27).
Fig. 21. Typical result of nuclear excursion using point-reactor, kinetics model and range of total fissions and lengthof time (after Ref. 21, Figure 4.3-1 and 4.3-2).
.
.-laEl
32°00”New Mexim I I
Texas 104”00’ 103=4LY
o 10ml
+-l+o 5 10 15km ml-i3424a7a-3
Fig. 1.
44 June 1,2000
..
FutureExploratoryDrilling Rig
. ----------------- ,------------------,
L,
. >. . ..-. . . . .. . . .. , . . . .. . . . . . . ..?S.. .
% Boundayofkcessible ~a
Groundwater
\ Envimnmentfor : Releases
shan 40CFR 191 1,,
Lower IT /Sefd System
,
Anhydrite layers ‘a-end ‘b” ~ MB 138
g ~~~ / ,.F
0,
z(n
Y I ‘MB 139Potentially DisturbedRock Zone Around
1
,,
RepoQto~ ,,,
Not to Scaleg ,’ ,,’‘ /‘/’’’” ,
0’, Impermeable R& “ ●“ .“ ~“ ~“ Z“ .“ ~“ .“ ~’
~ ,’ ‘ ,,./ ‘. ~ Surrounding Brine ,~’ ,~’ ,~’ ,s’ ,1’ ,~’ ,~’ ,z’ ,~’ ,,’
~ ,,8’8 py~d ,.’,.’ ,e’ ,,.’,.’,,8’.,,.’,,$’ ,,0’ ,,~’ #,” ,,’’.,,” #,’’,,J’ ,,,. ,’ ,’,’,.’ . . . . . . . .
-TRI-6342-4B42-1
Fig. 2.
45 June 1,2000
.
Risk =
~1a) Specification of necessary‘1 physical conditions (5 111)
I
~i’!iszir’n1.L’IMinimum
mass ked
Ir
————-—- —
,0.$ I J,0.1 ,00 ,01 ,02 ,~3 ,(34 105
Fki!e Cmxntmtbn @#nPl
.~ b) Feasibilityof assernliy
I ● Physical constraints P{p} (5 IV)
,1 “ Hydrologic constraints P{h} ($ IV)
1 ● Microscopic constraints
~~ ● E!I%%! !%r%traintsP{c} (5 VI];. - absorption
‘,1 - colloidal filtering~l,az, ~ precipitation-— .~ -:-p-~.~+--~$~?l-. .., .-,;.-:.<,,.,:::, ~..,,-:.+. ‘?-’-.Ywfi.:.. .--:?~. .->...:.,:
x—.
-—
—
—
a) Number of fissions and energyrelease (~ Vll)
● analogy with accidents● dynamic nuclear simulation
Tim (s)
b) Specification of plausible initialconditions (~ Vll)c Slow assembly (with moderation]
- deposition● Fast assembly
(with and without moderation)- earthquake
Fig. 3.
-———1
NuclearModeling I
I
I
IIIIIII
——— — I.—— — 1Geophysical 1
Modeling I
IIIII
——— — -1TRI-6342-621O-2
‘
WIF
GatuikfSanta RosaFormations
Dewey LakeRadbeds
Formation
RustierFormation
SaladoFormation
CastilaFormation
Bell Can oniForma on
●Elevation
‘-12 (m)
ERyA-9 Cabin
=!===-=-.- ———...=
@
,...— —.— ...—— Murktons & Slltstone..— -——— .-. —,... — —... —— . . ...—. . .———-.,-—.
~. ~mef *9e@~&*/m&f
—
~nnfled
I Halite
T MB12=
. .
TMB 13~
5.5 m MB 139—4m
%
/Anhydfite Ill /
s&
Halite1
W......................................●Contact Elevations are
Referenced to Borehole ERDA-9
-1 p&.g
874.085S.7849.1823.4816.4
779.7
627.6
511.6
y#.:
178.1162.0TD
MeanSea Level
1100
1000
900
800
700
600
500
400
300
200
100
0
-1oo
-200
-300
TRI-6342.4843-4
Fig. 4.
47 June 1,2000
TRI-6342484G1
Fig. 5.
.
Salt HandIing ShaftSupport and Waste Handling Building
Exhaust Filter BuiIding
III
_ Experim~ntsl An% >
/ Stsndardwaste \ Waste’. . _- -. la
Fig. 6.
49
TRI-6346-4850.3
June 1,2000
.
105
104 ~
G? UOOweierreftector ~
11( J-1t , , ,,,,,1 , !,,,1, , t ,,,,,0 t , ,,,1,1 t ,,,,,1 , t ,,,
10”1 100 101 102 1,3 104 105
2*U Concentration (Q) (kg/m3)
1,5
102
.
‘Pu/H201 No Reflection (experimental)10.1
10-1 lfJo 101 102 103 104 1,5
2WPUConcentration (Cpu) (kg/m3)
(a)
Fig. 7.
50
TRI-6342-4033-5
(b)
June 1,2000
.
\ $elado I
-1 I
0.1 1 10 100 1000 10000
Concentration(mMormolelm3)
TRI-8342-4652-3
Fig. 8.
51 June 1,2000
SiO~
AIz03
K20
Na20
CaO
MgO
F~03
TiOz
Mn304
Loss on ignition
uo~
S03
0.001 0.01 0.1 1 10 iiiPercent Weight
TFU-6342-4879-I
Fig. 9.
June 1,2000
.
,: ... 3k#m~
....”. . . . . .
\
Maxhumicawbabn
%.. .. . .. .. . .. .. ~“,
Linlii(-1OOOWn) ~(~
pc6SwJw3h -
10-’Ici%pmay -
! ! t10-1 1(-JO 101 102 j 03 104 105
(a)
53
~~-y-Z&+: “:_
I
~ &’-y---,...Cukbid
COslaa t+’ine+=16%
—. —— ———CukixaMnO ;]., ” y~ --’w2!!’3— -$ “ ....... ......——-~ —-...........”
LPlm i. /....-.water .............
I 9 hpfd1 I II I I I . . ....1 d
10-1 100 lo~ 102 103 104 105
‘Pu Concentration (kg/m~TR1-6342-487&2
(b)
June 1,2000
.
104S&j&hci —I 1 I I
41———————
5%=u09Col /g?.;.cy‘ ‘“ E /Culebn brine
. . . . .I ~ +.10%
.? 101
i‘k
..... ................
.......... h4amum-- i
10-11 I I I I I WI10-1 100 101 102 103 104 105
Fig. 11.
‘U Concentration (kg/m3)
(a)
54
r 1 1 I I I --l
v’:yw~-
Uli,clmdt
*
/’.........-..................:.- .....
uc+@’3all&ooe =... ........$=10% ...........
‘U Concentration (kg/m3)
TRI-6342-4877-2
(b)
June 1,2000
105
1(34
100
E‘ “’’’’” ‘ “’’’’” ‘“’’’’”‘“’’’’”‘ “’’’’” “q~-------,.
!7‘. 93.2%=UI/ ‘. H20 (SW= 100%)/t’ ,
:Q
Goethite (.$.20%):,ti..r!. i
I or%:‘, H20 (SW= 100%]/ ,●.Tuflor Sandstone,’:
Pu/ H20/Goethiie 5
...{q:~yy).” :
\ tif--” :..: --- ....... .....
..... \.. ............ u~o
,o.ll~10s
Concentration (kg/m3)
-I-RI-3342-3228-O
55 June 1,2000
1.2 ,,,,, , I I I m
1.1 - Spheres of Pu$Q=lful:brdCu!ebra bdnee ~
g 1.0- PuO$l~~6b%dCulebra biine
\ ~ u:
J&&
~ 0.9-+.+%-..bPu0#f20
~ 0.8 - ~ .16%
g f307-Miiimum fhicknese for
1-1:
.. .“ :.. ..: :.:.:0.201 fracture pams”~
. . . . . ... ....p.:......: .
~ 0.6 -.. . . . .
%J:=Pu = 1:1/CulebrS/CulebrSbrine% 0.5- ~=1622~ 0.4-
g 0.3- \--
MinimumUkknees
g 0.2- ~.... for 0.04 fracture porosity
0.1- Puoz iliaO (expmfmenlel) -
I . . . . ...*... . .0.0 M.
,.O ,01 ,02 ,03 ,04 ,.5
239Pu Concentration (kg/m3)TRI-6342-S4795
Fig. 13.
60
r
t Initialporosity(S4.8%)
%!50~
I I I I
Scenm”oS5
~1
Iwo 100+30Time (years)
lWs3’12-4e21-2
Fig. 14.
57 June 1,2000
.
t I I 1 i I I i IScenafim RIS5
‘mar= >.--...-.””””””””””””””””””””””
‘aor%x=””:.----”---”---
I2$%:+20
537S01(macuan)
~
,
~- 10.2 I t I I I I t IItcc+332(Ilh)0123456 769 10X102
Time (yr)
TI%6342-546%2
58
.
) , ,,,hq # , BIllaj ‘1 ,8, w, v ‘ ‘ ‘“”l ‘ “’”Y ‘ “’’’”l T ‘ ‘ “’”t ‘ ‘ ““d ‘ “’’”l I ‘t 1 T ‘ “’””l 8“’’”l ‘ m
US WIPP Salt repository (TRU waste) Plutonium
1989 SEIS (Ref. 54, p. 4-27)10%
(5W.)
PU’JI I
1996 PA (includingcolloids)Pulll @~tile Bfinepullls~l~d~ BrinepulVCastileBfinepulVs~l~do Brfne
US Tuff repository (spent fuel)1993 PA (Ref. 120)
Oxidizing - fresh m
Canadian granite repository1994 Generfc EIS (Ref. 122)
I 1 ● 1
JJ.ulLJ, ~*~**J, ~~~~,~~, 18,4J,,1,,,.1, **,,JLIJJld t ,,*1J I Ultl*~*!IIIJ ~10-15 l(y.13 1011 10-s 10-7 10-5 10-3 10-1 101 103 105
(a) Plutonium total concentmtion (mM or moleslms)
I 8“’”q ‘ “’’”~ ‘ ‘ ““”1 ‘ “’”~ - ‘ “’’”q ‘ “’”q - ‘ “’””l ‘ “’’”~ ‘ ““u~ - ‘ “’”q ‘ “’””1 ‘ “’”MI ‘ “’’”q ‘ “’’”$ ‘ ‘ ““f ‘ “’”ul “-
Ill t**m41 ,WLl IIltlml IIULULI i II J I J .! , ,,WI I I.lu
us WIPP Salt repository (TRU waste) . Uranium
1989 SEIS (Ref. 54, p. 4-27)
199~~A and 1992 PA (Ref. 70)
Uvl
1996 PA (including colloids)IJV c~stile Brine ●
ulV Salado Brine
@l ca5tile Brine
@l Salado Brine
US Tuff repository (spent fuel)1993 PA (Ref. 120)
Oxidizing - fresh I I PxNepal I natural analogue (Measurad) (Ref. 121)
Oxidizing - fresh ●
Canadian granite repository1994 Generic EIS (Ref. 122)
Cigar lake natural analogue (Measured)(Raf. 123, Table 3.10)Reducing - fresh m I
10-15 10-13 10-11 10-9 10-7 10-5 10-3 10-1 101 103 105
(b) Uranium total concentration (mM or melee/m3)l-Rl-6342-4s51-3a
Fig. 16.
lo3p. Jill Jf. ,lJlll. rrll. .. l,. !.!,lrJ.]..li,l,,.lJ...~
[
0s2
g 102 ~“~”= ❑ s3b.!%
g ,.,0SS
“da 2s
/“
. . .. . . .. .. ... . . .. . . .. . . .. . . . . . . . . .. .. . .7k~Cm?
S..... ............................................
‘!
~w,tiyz;;.l ......
..... ...................... .... ................ .. ........................................................ ....G 100
~r=r&.
= 6’ 6
10-1 *+ +00
0.20
++
8
++ +
+
+ ++++
+
++
+++ +++ +Q+ ~+0 ‘+
++ ++
+
,0.7 ~
o 10 20 30 40 60 60 70 60 90 100Vector Number
TRI-6342-54E54
Fig. 17.
60 June 1,2000
.
Amount in solution per solution volume (Ce) (mM or mole/m3)
(a) Uranium
.-1“-12
lMo-11 lo-lolo-e 10410-7 lo~o-IJ J J I d
Amount in solution per solution volume (Ce) (mM or mole/m3)
(b) Plutonium
TRI-S2424Z?25-O
,Fig. 18.
61 June 1,2000
.
109
k /
8036 MTHMSRPHLW
.—-— -_,0’3 r 21CXJMTHM Honford ->
-~R:a:o~D:E\., ‘\\,07
‘. \. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...0..
i,06 CH-TRU-----
1\--
\7
,05RH-TRU “\.. ‘--—--
---- ---104
,03 ~
o Shlflqt) ,01 ,02. 103 104
lime (1)(y)
Fig. 19.
62
,rl~o 3inh-w) ,01 ,02 ,03 104
Time (t) (v)
TRI-6242-4880-O
June 1,2000
10
5
0
5
10
10 5 0 5 10
East - West Distance fmm Borehole (m)
(a) Computational mesh
1 ‘Transport Bounda~ Condltiol
F 10 Oirahlet= 42 kg.rm ,
d,,,,,,,7’
c; 10
1
Ma&r ci%slomDm=4.3 x lUIO m21s ,‘
io 5 0 5 10-I ...,.,.,,.. )j
East - West Distance from Borehole (m)
(b) Brine mixing zone
100
10?-
17TI,/f&”’awe100
=--- ....””> G~
.$;;22.2...<.-‘k”m’ ,,; 10 $y \ g-.0- /(.. 10m3/kg Average ~- ● m - K.= 6.8 m31kg-4A-. KD= 0.7 m3/kg
Density
-.+0-. KD= 0203, n= 0.707 (Goelhite noalinear)
0.1‘ I “ I0.10.1 1 10
Radius fmm Borehole Center (m)
(c) Average adsorption on dolomite
TRI-6342-6231-1
Fig. 20.
63 June 1,2000
.
105
104
~ 103
&3 10* 10” 100 10* 101 lox 10. d
E 102
10”110.2 10-1 100 101 102 103 104 105
1-Time (s)
Rangeoftimes q
Fig. 21.
TRI-6S42-6227-O
64 June 1,2000
TABLE I. Designation of Scenarios for 1996 PA
TABLE IL Anticipated Mass of Fissile Material in the WIPP Repository (as evaluated from Ref. 119)
TABLE III. Mineralogy of Salado Halite and Culebra Dolomite at the WIPP
TABLE IV. Example Calculation of the Median Total Concentration of Uranium and Plutonium in Salado and
Castile Brines
TABLE V. Summed EPA Units Versus Time (Ref. 65, Table II)
TABLE VI. Ten Largest Criticality Accidents in Processing Plants
TABLE VII. Six Largest Criticality Accidents Involving Moderated Metal and Oxide Systems (after Ref. 40)
TABLE IDesignation of Scenarios for 1996 PA
ID Short Description Description
S1 Undisturbed-or Base Case Predicted behavior of the WIPP disposal system when not disruptedby human intrusion or unlikely natural events
S2 El intrusion after 300 yr 300 yr after closure, inadvertent intrusion into the repository andpressurized brine reservoir in the CastilcZafter abandonment, sealsrapidly degrade
S3 El intrusion after 1000 yr 1000 yr after closure, inadvertent intrusion into the repository andpressurized brine reservoir in the Castile
S4 E2 intrusion after 300 yr 300 yr after closure, inadvertent intrusion into the repository
S5 E2 intrusion after 1000 yr 1000 yr after closure, inadvertent intrusion into the repository
S6 Eland E2 after 1000 yr Two inadvertent intrusions into the repository around 1000 yr afterclosure, one intrusion also intersects a brine reservoiq afterabandonment, seals degrade such that flow is forced through entirerepository
TABLE II
Anticipated Mass of Fissile Material
in the WIPP Repository
(as evaluated from Ref. 119)
Radioisotopea Mass (kg)
Uranium
232
“;
‘u22s
“:237u238u
Plutonium
‘bPu
28PU
‘9PU
‘%%
‘lPU
“12Pu
‘3PU
*PU
‘9Pu fissile gramequivalent (FGE)
Total (all radioisotopes)
1.21 x 10-3
2.01 x 102
8.13 x 101
8.07 X 103
6.64 X 106
7.36 X 10-7
1.49 x 105
1.96 X 10-s
1.53 x 102
1.28 X 104
9.44 x 102
2.38 X 101
3.07 x 102
1.23 X 10-18
8.46 X 10-5
2.11 x 104
1.81 X 105
a Bolding indicates tissi~ematerial.
.
TABLE IIIMineralogy of Salado Halite
and Culebra Dolomite at WIPP
Mineral Density %wt(kg/m’)
Culebra Dolomite’
Dolomite, CaMg (C03)2
Gypsum, CaSOq ● 2HZ0
Quartz, SiOz
Clay
CorrensiteCao.wNaOZ KO.GG(~3.13 F%.w
hfg1286) A1l.glSild.wOdO(OH)m
IlliteK ~ (S7Al OZO)(OH)g
Average density
Salado Halited
Halite, NaC1
Polyhalite, K2Mg Ca2 (S04)4 ●
2H20
Anhydrite, CaS04
Gypsum, CaS04 ● 2HZ0
Magnesite, Mg C03
Average density
2872
2320
2650
2950b’c
~c2850
2165
2775
2610
2320
~
2160
91.5
3
1.5
2.4
1.6
93.2
1.7
1.7
1.7
1.7
a Ref. 63, WIPP-l~ Ref. 62, Chapter 3.b Density of chlorite, which derives fmm corrcnsite.c Ref. 125, pp B-193-B-195.d Ref. 126, assumed absence of silica for PA modeling.
.
TABLE IVExample Calculation of Median Total Concentration of
Uranium and Plutonium in Salado and Castile Brines
Colloids Total FilterableDissolved
HumicConcen-
MicrobeColloid
(SD) Intrinsic Mineral Fractiontration Fraction(m Scale Factor Scale Factor c~ Colloid
f- f~ (% (m(CT)= (Microbe &Wvl’) Mineral)
Volubility in Salado Brine with MgO
Pu”[ 5.5 x 104 0.19
PUN 4.4 x 10-3 . 6.3(1996 PA)
PUN 2.0x 10-s 6.3(Current)b“N
4.4 x 10-3 6.3
1.0 x 104
1.0 x 104
2.6 X 10-5
2.6 X 10-5
2.6 X 10-5
2.6 X 10-5
2.6 X 10-5
2.6 X 10-5
2.6 X 10-5
2.6 X 10-s
2.6 X 10-5
2.6 X 10-5
8.5 X 104
3.4 x 10-2
0.35
0.87
0.23
0.40
0.3
0.3
1.0 x 10+0.3 1.8 X 104 0.88 0.18
0.0021
0.0021
0
0
3.2X 10-2
9.8X 10-3
0.86
0.11
1.1 x 10-3
4.5 x 10-3VIu 8.7 X 10-3 0.12
Volubility in C%stileBrinec with MgO
PI/” 6.5 X 10-s 1.37d
PUN 6.0 X 10+ 6.3(1996 PA)
PUN 4.0x 10-s 6.3(Current)bU!v
6.0 X 10+ 6.3
u VI8.8x 10-3 0.51
1.0x 10+
1.0x 10+
2.0 x 104
7.3x 10-5
0.3
0.3
0.68
0.92
0.23
0.38
0.3 1.OX 104 2.7X 10-5 0.99 0.95
7.0x 10-5
1.3 x 10-2
0.0Q21
0.0021
0
0
0.91
0.34
0.37
3.3x 10-3
n seeEm 1b Value; change substantially for all radioisotopes in +4 valence state but only plutonium is shown.c Volubility in the Culebra brine is assumed to be bound by the volubility in the Castile brine.d Sampled values ranged between 0.065 and 1.60; me&an used (Ref. 2, p. 5-14).
,
TA13LEVSummed EPA Units Versus Time’
(Ref. 65, Table IQ
Time Summed EPA Units of@r) Fission Productsb
o 1.0000
0.003 11.1790
10 8.8538
100 1.1047
110 0.8794
1000 0.0004
10000 0.0003
a Based on release limits (surrogate of healthhazard) in 40 CFR 191 for fission products ofone initial EPA unit of ‘9P0
b Release limit is 100 ClHalf-life is 7.594x 101’sMass is 1.60091kg
.
TABLE VITen Largest Criticality Accidents
in Processing Plants
Date PlantTotal
FEionsCause
6/16/5& Y-12, TN ~q ~ Iols ~U solution washed
into drum
1215/6tYMayak Enterprises, ~019Plutonium
the Urals concentration too high
in solution
l/25/61b ICPP, ID 6 x 1017 w soIution forcedinto cylinder by air
W’/62b Hanford Recuplex, 8.2 x 1017 Plutonium solution inWA sump sucked into tank
1/30/63’ Siberian Chemical 7.9 x 1017 HEU solution wasCombine divided and transferred
to diffenmt vessels
12/13/63’ Siberian Chemical 2.7 x 10’7 Vacuum VdVe to theCombme trap Wasshut Off
1216/65’ Mayak Enterprises, 7 x 1017 Uranium massthe Urals exceeded safety margin
12/10/68’ Mayak Enterprises, 5 x 10*7 P1udniumthe Urals concentration too high
10/17/78b lCPP, ID 3 x 10’* ‘% buildup due todiluted scrub solution
12/13/78a Siberian Chemicrd 3 x 10’S Plutonium mass inCombine containem too high
a Taken from Ref. 127.b Taken from Ref. 41 (Chapter 3, pp. 17-28), original reference,
Ref. 128. (Data values shown here agree with those in Ref. 40.)
,.< .
TABLE VIISix Largest Criticality Accidents Involving
Moderated Metal and Oxide Systems(after Ref. 40)
Date PlantTotal
F&osssCause
12/12/52 Chalk River 1.2x IOm Positive voidcoefficient
7/22/54Idaho National Lab 4.68 x 10’8 Planned transientextended
10/15/58 Vinca Yugoslavia 2.6x 10’6 Faulty powermonitoring
3/15/60 Saclay, France 3 x 1018 Removrdof absorberrob
1/03/61 Idaho National Jab 4.4 x 10’8 Removrdof control rod
11/05/62 Idaho Nationat Lab 1 x 10’s Planned transientexceeded