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