ML19332E862

Speech Entitled, Estimating Radionuclide Releases from Low Level Radwates, Presented at 871123-25 Near Field Assessment of Repositories for Low & Intermediate Level Wastes in Baden,Switzerland

ML19332E862
Person / Time
Issue date:	11/23/1987
From:	Weber M NRC
To:
References
NUDOCS 8912130080
Download: ML19332E862 (15)

Text

y N

Conference

Title:

_Near-Field Assessment of Repositories for' Low'and g:+ 4 Y Intermediate Level Wastes f

4 f Date & Place:

' November 23-25, 1987. Baden, Switzerland I

ESTIMATING RADIONUCLIDE RELEASES FROM LOW-LEVEL RADI0 ACTIVE WASTES Michael F. Weber i

U.S. Nuclear Regulatory Commission Washington, D.C., U.S.A.

ABSTRACT

)

Performance' assessments of low-level radioactive waste (LLW) disposal facilities are particularly sensitive to assumed radionuclide source terms and-release models.

The U.S. Nuclear Regulatory Commission (NRC) is currently evaluating' alternative approaches for estimating radionuclide releases from i

LLW. As part of a' more comprehensive performance assessment methodology, the NRC is - developing source. term models to estimate radionuclide releases for

. safety assessments of LLW disposal facilities.

These models range from simple, empirical approaches to more complex models that attempt' to simulate the processes of. leaching, dissolution, sorption, corrosion diffusion, and advection.

Preliminary results indicate that relatively simple, deterministic i

approaches are preferable provided that they can be demonstrated to be reasonably. conservative.

The NRC staff intends to select an approach for estimating radionuclide releases that provides demonstrably conservative estimates and sufficient flexibility to consider releases on a

i

facility-specific basis.

RESUME Les estimations de la ' performance des installations destin$es au

'traitment ' des r$sidus faiblement radioactifs (abbrdviation US:

LLW, pcur

" Low-Level Waste") sont-particul1&rement sensibles aux modEles de simulat'on du :digagement: et-aux termes prssum6s de la. source des radionuclides.

Actuellement, la Commission R6g,ulatrice Nucl6 aire. Am6ricaine (U.S. Nuclear j

Regulatory Commission:

NRC) etudie d'au".res msthodes pour, estimer les j

_dsgagements de radionuclides 6manant des LLW. C'est dans l' ensemble d'un plus grand ' programme d' etude de la m6thodologie de s tinite i

estimer cette performance que le personnel du NRC s' applique l' d6velopper des - modeles concernant les ' termes de la source afin d'estimer les d6gagements de radionuclides en vue de d6finir la s6curitt des usines de traitement de LLW.

Ces modiles partent des m6thodes simples et empirique jusqu'a des modEles plus complexes - qui tentent de simuler les processus de filtration, dissolution,

@y adsorption, corrosion, diffusion et advection.

Les resultats pr$11minaires indiquent que des methods simples et d6terminstiques sont pr6ferables - tant 6 s qu'il est possible de d6montrer qu'elles sont raisonnablement conservatrices.

le personnel de la NRC a l' intention de s61ectionner une m6thode pour estimer i

les dsgegements des radionuclides qui puiss fournir des evaluations conservatrices et une suopless suffisante pour pouvoir etre modifise pour chaque installation.

8912130080 871123 PDR MISC 89.12130080 PDC

/

y

1

.,z Y

~

e M

INTRODUCTION-

+

The U.S.

Nuclear Regulatory Commission (NRC) is responsible for

- regulating the disposal of commercial low-level radioactive waste (LLW) % the United States

.NRC's regulations in 10 CFR Part 61 provide a comprehensive set of licensing requirements for land disposal of LLW.

These regulations provide several performance objectives, including one that requires license applicants to demonstrate that concentrations of radioective materials that may be released to the general environment will not exceed appropriate dose limits [1]..

To. demonstrate compliance with this performance objective, license applicants.need to assess the performance of proposed LLW disposal facilities.

Such assessments should consider any significant pathway of radionuclide release, transport, and exposure to. humans [2].

The NRC staff considers the groundwater transport pathway as one of the most significant pathways re:ulting in potential human exposure to radionuclides released from LLW disposal facilities [3).

In addition to evaluating the environmental transport of radionuclides af ter they reach the groundwater, a realistic performance assessment of a LLW disposal facility should include an evaluation of the release of radionuclides from the waste form as the source term for the transport evaluation.

.The processes-and conditions that control radionuclide releases from LLW are among the poorest understood components of performance assessments of LLW disposal facilities. Although a substantial amount of source term modeling and experimental studies have been performed, these studies have not significantly reduced uncertainties associated with the estimation of radienuclide releases from LLW [4).

Nevertheless, such estin ates are essertial to performance assessments of LLW disposal facilities.

Previous performance assessments have either assumed radionuclide-releases a priori or utilized extremely simple-models.to estimate radionuclide releases.

These-

~

aporoaches,

however, do not provide demonstrably conservative or best estimates of radionuclide releases.

In addition, such approaches do not provide' sufficient flexibility to account for enhanced radionuclide containment by engineered barriers and stab,ilized waste forms.

Therefore, additional effort is warranted to improve current performance assessment

- capabilities to estimate radionuclide releases from LLW disposal facilities.

After a brief review of previous efforts to estimate radionuclide releases from LLW, this paper d9 scribes several approaches being considered by NRC and its contractors for estimating radionuclide releases from the waste form into the disposal unit.

The paper briefly reviews the diverse characteristics of LLW streams in the United States and suggests that this diversity currently precludes complicated approaches for estimating radionuclide releases.

Based on this information, the paper presents an approach and computer code entitled RELEASE that the NRC staf f is currently evaluating.

The paper concludes by describing planned future work to test and refine the capabilities of RELEASE and alternative approaches.

PREVIOUS WORK The relative importance of radionuclide release estimates in performance assessment for LLW disposal facilities has been recognized for many years [5].

Without releases of radionuclides into the aqueous phase, groundwater and 4

n

, =

g i

'Isurface' water pathways would not be significant in considering the performance N

of LLW disposal facilities. In recognition of the importance of radionuclide 1'

releases from LLW numerous attempts have been made to minimi:e or eliminate thereleasesfromcertaintypesofLLW[6]. However, these attempts have only been partially. successful because they have not been able to qucntify the small, but finite release rates of the radionuclides.over the long periods of' In addition, not all types of LLW can be stabit':ed and treated to concern.

minimize the. release of radionuclides.

Considerable ef fort has also been invested in development of standard laboratory leaching tests to screen waste form characteristics and demonstrate their suitability for disposal [7-10].

The validity of extrapolating the results from these tests to estimate radionuclide releases in actual LLW disposal facilities has not been demonstrated.

In _ addition, the technical community has not been able to select a standard leaching test to represent the full range of environmental and waste conditions anticipated at LLW disposal facilities.

Thus, the principal utility of the laboratory testing has been to screen waste form characteristics and expand. the data base about radionuclide releases rather than to estimate long-term radionuclide. release rates. 'This data base has supported the development and validation of empirical and phenomenological models for estimating radionuclide releases.

Previous investinators have also pursued development of computer models to estimate long-term releases of radionuclides from: LLW.

For example, NRC developed an empirically-based generic approach in the Impacts Methodology used to support the development of 10 CFR Part 61. The source term algorithm of_ the comprehensive Impacts Methodology was developed based on the best information about radionuclide releases available when NRC promulgated these regulations in 1982.

The algorithm employs an integer-based approach to calculate a waste form and package barrier factor, which is a multiplier in the overall equation used to estimate radionuclide concentrations in the L

environment.

The source term algorithm of the impacts Methodology accounts for leachability (i.e., waste form), chelating agents, waste segregation, and the accessibility of radionuclides embedded in activated metals [11].

This

approach, however, does not ettempt to estimate radionuclide releases by

~

explicitly simulating any of the processes. or phenomena that may control radionuclide releases (e.g.,

advection, diffusion, solubility, sorption).

Instead, the approach provides release estimates based on simple empirical relationships and radionuclide concentrations observed in leachate at two commercial LLW disposal facilities: West Valley, New York, and Maxey Flats, Kentucky.

1 More recent approaches have attempted to simulate the processes and phenomena that control radionuclide releases from LLW [12-14].

For example, Reference 14 develops an ovaluation methodology that explicitly simulates moisture migration and groundwater flow in and near the LLW, as well as transfer of the radionuclides into the aqueous phase, degradation of waste packages, and transport of radionuclides after they are released.

Under L

contract to the NRC, this effort should continue for the next couple of years i

L and develop an approach to simulate the processes that significantly af fect release and near-field transport of radionuclides at LLW disposal facilities.

Such complicated analytical approaches, however, may not be compatible with the amounts and types of information that are available for LLW disposal

p<q o, -

'1 facilities. Therefore, implementation of relatively complicated approaches for source estimation may be precluded by data limitations.

DIVERSITY OF LLW.

i The release of radionuclides from' LLW is controlled at least in part by the characteristics of. the - waste.

Thus, it is important to assess the relative amounts and characteristics of LLW expected to be. received at LLW disposal facilities. Figure 1 compares typical radionuclide concentrations in untreated, high volume waste streams in the United States.

This figure illustrates the diversity of waste characteristics in terms of radionuclide i

types and concentrations.

It is imoortant to recognize that the figure-provides 'nformation on only a few radionuclides, whereas LLW may contain many other radionuclides.

It is-also imoortant to recognize that radionuclide 4

concentrations are often lower in the large volume LLW streams thar, in smaller

{

- volume waste streams.

For example, typical radionuclide concentrations in

{

decontamination resins may be orders of magnitude greater than concentrations in the large volume waste streams depicted in Figure 1.

Further, these concentrations reflect average concentrations in " typical" waste streams.

Actual radionuclide concentrations may vary considerably as a function of many

. parameters, including generation process, operational cycle, age of facility, fuel burn-up, housekeeping practices, waste treatment, and age of waste.

The diversity of radionuclide concentrations is only one factor that complicates development and use.of approaches for estimating radionuclide releases from LLW.

Although ' the diverse characteristics of LLW complicates analysis of radionuclide releases, careful record-keeping of the characteristics of the l

waste in disposal units is essential to the development of defensible sot.rce i

term models.

In addition to radionuclide concentrations, radionuclide releases from LLW - may depend upon'~many other variables, including composition and characteristics of the waste form, leachant residence

times, and the physico-chemical characteristics of the leachant.

For example, if the percolation rate of water through the waste is slow, release of radionuclides from _ the waste may be dominated by diffusion. However, if the flow rates are S

very slow, radionuclide concentrations in the leachate may be constrained by solubility _ limits. Alternatively, if - percolation rates are relatively rapid, then release of the radionuclides may be controlled by the dissolution rate of the waste form [15]. Table I summarizes the factors that are known or suspected to affect the release of radionuclides from LLW [4].

Given the

{

current limitations of analytical approaches for estimating radionuclide i

releases from LLW, these complications currently preclude explicit simulation of all. of the processes.or phenomena that may be important in quantifying the releases.

Therefore, approaches to estimate the releases should rely on relatively simple and conservative analyses that attempt to approximate the results of release processes rather than on complicated algorithms derived from fundamental equations (e.g., complicated multi-dimensional flow modeling

)

coupled with geochemical equilibria and speciation modeling).

I t

I

~,

.lm

,3 4,"N

-a

~

=..

10 --

~i!

LEGEND

~

~ '

~~~~~\\_

P-CONCLiO.

/

E i

./

/

8-CONCLIO

- s i y - _.

/

...... B-FSLU DGE i

/

[f i '.

f'

./

c g5 lll

\\\\ ','.

l

-l

_ ___ ___ P - C OTHAS H :

'3E

[

S p=

.oi

,. f I

,e

\\i *.

.8..

P-f JCTRASH

g

/

T

.o_

.f wo

- i.f -

,/

s-coTRASH ox ji -

/

z=

.ooi

=

V.

o.j=

,\\

~ /.ff

g \\1 /,/'

1 --

NN

,//

\\. \\j j

//

xN oooi s

N/ /

N. /

x N

/,

.//

N

'N.\\ //

'V'

.ooooo. =

\\./

.O000o0 H-3 C-14 Fe-SS Co-60 Sr-90 Cs-137 RADIONUCLIDE' 1

Figure 1-Typical radionuclide concentrations _ in 'high-volume low-Ieve1 radioactive waste streams. ' P= Pressurized Water Reactor, l

B= Boiling Water Reactori CCNCLIQ: Concentrated I.iquid Waste,'

~ FSLUDGE=FiIter SIudge, 07 MASH =Ccmpressible Trash, NCTRASH=Non-campressible Trash.(Based on Ref. 18).

I y

s.-

.f.

.2

.u..

m u

.,_,m_.-

m

.s

=

m

[hi 4

H" 7 pg N.

Js Table I.

-Factors that may_ influence 'the release of radionuclides from LLW-(Modified'from. Table 1.1, pg. 2, of Ref. 4).

System Factors.-

_Radionuclide Factors E

~

Temperature Radionuclide y

Pressure Solubility J

Radiation-field-Solid spet.iation and k

Leachant resident time form Leachant contact

~ Atomic-radius Ratio of waste form surface Physical state area to'leachant volume p

Waste Form Factors.

Leachant Factors i

Composition teachant composition e

Surface characteristics Complexing agents

-Effective diffusivity pH and Eh Corrosion rate Percolation rate

_ Porosity Biologic activity Structural integrity Curing time Mixing. ratio (waste / matrix and matrix / water)

Biodegradation RELEASE This paper focuses on a source term release approach and computer code -

entitled RELEASE. RELEASE is being developed and assessed by NRC staff along -

with 'other -alternative approaches, including site-specific application of ' he t

source term algorithm of the Impacts Methodology [11] and a diffusion-limited

' approach developed by investigators at Brookhaven National Laboratory _for

estimating releases from LLW [12]. The objective.of the NRC staff evaluation is to - select-a preferred approach -or set of approaches for estimating radionuclide releases from LLW.

The RELEASE approach has been incorporated into a computer program of the i

same name, which was written by the author in advanced. BASIC for execution on a personal computer.. RELEASE interactively quertes the user for input information and calculates incremental and cumulative fractional releases of individual radionuclides -and annual releases in curies.

The author is currently revising RELEASE to enhance its ' computational efficiency and user

_ options.

This section' briefly reviews RELEASE's analytical approach, input L

. requirements, _ and output characteristics, and compares th9 performance of RELEASE with other source term estimation approaches for a sample problem.

Analytical Approach RELEASE is intended for site-specific application using the L

characteristics of the waste, disposal facility, and disposal environment.

RELEASE provides limited flexibility to consider engineered barriers and other types _ of disposal facility enhancements intended to improve LLW isolation.

l I

p i

The analytical approact used in RELEASE assumes tMt radionuclice releases from LLW are dominated by rinsing of surface contamiration from LLW trash and diffusion of radionuclides out Of solidified waste forms.

In addition, i

RELEASE approximates radionuclide releases from activated metals.

RELEASE combines these processes to estimate total reieases of a given radionuclide from individual LLW disposal units. The equations used in RELEASE to describe these prucesses were developed based on laboratory experiments and field i

observations at several LLW disposal facilities in the United States (4).

The analytical approach used in RELEASE is based un several fundamental assumptions.

RELEASE assumes that O) advective release of radionuclides from LLW trash can be represented by a rinsing model, (2) radionuclide releases 1

from solidified wastes (e.g., cementitious or bituminous waste forms) are centrolled by diffusion through the solid matrix, and (3) releases of radionuclides from activated metals can be approximated as a constant fraction j

of the diffusional relaases.

RELEASE also assumes that no liquid or gaseous wastes are included in the disposal units and that wastes do not degrade form ~ liquid or gaseous wastes uther than leachate within the disposal unit.

to In addition, RELEASE does not explicitly simulate corrosion of waste packages or activated metal components.

Further, RELEASE does not consider 3

biodegradation of' the LLW-or secondary release mechansims, such as gaseous tritium transport with subsequent i

deposition in the unsaturated zone and transport into the saturated zone.

The equations used in RELEASE were developed in References 4, 12, ard 14.

The rinse model was developed in Reference 14 by Brookhaven National Laboratory to simulate releases of radionuclides from trash in carbon steel drums at the Sheffield LLW disposal facility. surface contaminate is intended to simulate the mixing of leachant The model that occurs within a waste package after the package has been initially breached by corrosion or mechanical failure, but while the waste package still limits the exposed surface area of the waste.

dependant upon the sum of the volumes of the individual waste each infiltration event For during the year, RELEASE assumes that the wetting front propagates through the waste and displaces a finite volume of leacnate from the waste packages.

radionuclide releases occur during the lifetime of the waste pa the waste packagas fail completely, Once advection of radionuclides through the bulk waste n'ay dominate over the simple mixing process within individual w packages.

Therefore, the rinse model in RELEASE is best suited for estimatin releases from unstabilized, low activity trash containing relatively short-lived radionuclides.

Based on the equations derived in References 4 and 14 for the rinse model, radionuclide releases from LLW trash may be approximated as follows:

An = X2(1-T)(

)

, where An = activity release per cycle,

= number of rinse cycle, n

Z = ratio between the rinse volume (L) and standing bath volume [L), and X

= source activity corrected for C

radioactive decay = X(exp(-Lt)),

where X = initial activity, f

L=decayconstant[1/T),and t = age of waste [T).

L s

" RELEASE assumes that each infiltration event results in a constant displacement of mixed leachate from all waste packages.

In addition, the model also 61Wmes that the radionuclides are contained in the trash as surface contamination.

Thus, each infiltration event releases an ever decreasing incremental release of the radionuclide; RELEASE sums these incre ental releases to approximate the annual activity release r, ate.

RELEASE does not incorporate the " tilt" factor discussed in the original rinsemodel[14]. The tilt factor was originally used to decrease the annual releases proportional to the cosine of the angle of the long axis of the waste packages with respect to the vertical.

Because of uncertainty about waste package configuration in disposal units, the reduction of releases by the tilt factor does not appear to provide conservative estimates of radionuclide releases and is, therefore, not included in RELEASE.

RELEASE conservatively assumes that all waste packages are vertically oriented, which maximizes the rinse volume per infiltration cycle.

In addition to estimating radionuclide releases from unstabilized LLW trash, RELEASE can be used to estimate releases from stabilized LLW that has been solidified. As previously diset.ssed, RELEASE assumes that releases from solidified wastes are dominated by diffusion of the radionuclides through the waste matrix.

Once they reach the outer surface of the waste form, RELEASE assumes that they are instantaneously relessed into the leachant.

RELEASE partially accounts for containment by waste packages by assuming that radionuclide releases are proportional to the exposed surface area of the waste, which increases as a function of time.

Unlike the rinse model, however, the dif fusion model in RELEASE does not account for alternating wetting-drying cycles.

Such cycling has been shown to increase releases of certain radionuclides under specific conditions [12].

Future testing u the RELEASE approach will include evaluation of the need to consider wet-dry cycling in estimating radionuclide releases.

The equation used in RELEASE to estimate radionuclide releases via diffusion is developed in Reference 14 and may be summarized as:

CFR = (A/S)(C X + C X), where CFR = cumulative fractional release, 3

2 A

= $ + d t', where A = surf ace area function, S= initially exposed surface area [L),

K = container area constant, and t=leachingtime[T],

S

= surface,g ea of waste form [L],

X

= (S/V)(J0t), where V = waste form volume [L],

D = ef fective dif fusivity

[L/T),and C

= 1.3441 and C2 = -0.4416.

3 This equation is based on a quadratic power series fit for the solution to the infinite plane sheet diffusion equation as used in ANS Standard 16.1 [8, 14].

r The incremential fractional release rates may be calculated as IFR = (CFR(t) - CFR(t-1))(exp(-Lt')), where L = radioactive decay constant (1/T),and t'= tirte since emplacement cf waste.

The decay corrected CFR's are then calculated by summing the IFR's; incremental releases are calculated by multiplyug the IFR's by the initial activity of the radionuclide in the waste. Intheoriginalreference[14],t' was assumed to be significantly greater than t because leaching only occurs during a small fraction of the total time (t') since emplacement of the waste, However, RELEASE invokes the conservative assumption that leaching time ecuals time since emplacement.

Although it may be possible to estimate t on a site-specific basis using complicated moisture migration 'nodeling within the

waste, the uncertainties associated with predicting moisture migration in a highly heterogenous disposal unit may preclude such estimates.

Therefore, RELEASE assumes that radionuclide diffusion through the waste matrix begins upon emplacement of the LLW in the disposal unit.

RELEASE also estimates releases from activated metals in the LLW Ly multiplying the IFR's for solidified waste by the initial activity contained in the activated metals and by a constant ratio of 0.01.

Although the rates of radionuclide release from activated metals are not well understood, most approaches assume that these rates are controlled by the corrosion rate of the activated metal components.

These rates are generally low because the components are constructed out of corrosion resistant metals.

Many of the activated metal co'mponents are composed of high-alloy materials (materials with a high non-ferrous metallic component).

The radionuclides contained within the activated metal components are gradually released as the metal matrix slowly corrodes.

Corrosion of these materials occurs at low rates such as 7.3E-6 g/cm/yr for high allow stainless steels [11, 16, and 18).

Although release rates based on metal corrosion are a function of geometry of the metallic component, these rates are generally lower than those estimated for solidified wastes using RELEASE.

Thus, RELEASE estimates radionuclide releases from activated metals by multiplying the diffusion-dominated release rates by a constant

{

factor of 0.01 to account for the slower rate of radionuclide release from activiated metals.

The author has incorporated several features in RELEASE to consider the increased isolation of LLW provided by engineered barriers.

For example, covers that minimize infiltration into the waste in conventional trenches can be considered by reducing rinse volume.

Special waste packages intended to contain the waste and minimize its contact with leachant can be considered by delaying initial release time and reducing or eliminating the initial surface RELEASE can also consider structural components such as concrete vaults area.

and monoliths by delaying the initial release time and reducing rinse volumes.

Further, these barriers may provide additional protection by containing leachate after radionuclides are released from the LLW and before the leachate is released to the general environment.

Selection of these input parameters should be based on accelerated leach and durability testing in the laboratory; field observations i

of actual waste form, waste package, and disposal unit 1

t 0

~*

performance; and information on the reliability of engineered components such as waste packages.

Input Requirements and Output Characteristics bers enter a variety of input data into RELEASE in response to a series of interactive queries. The input requirements are listed in Table II..The input requirements have been minimized so that RELEASE can be ef fectively implemented in assessing potential LLW disposal sites given the types and amounts of information avc11able for the sites.

Based on input supplied by the user, RELEASE estimates radionuclide releases from the waste form into the disposal unit in terms of incremental and cumulative fractional releases and annual activity releases.

RELEASE provides separate output for releases from LLW trash, solified waste, and activated metals.

The releases are then summed to estimate the annual activity releases into individual disposal units.

These estimated releases may then be used as the radionuclide source term for evaluations of transport out of the disposal units and into the surrounding environment.

Table II.

Input Requirements for RELEASE Input Type Units Radionuclide name Decay constant 1/yr Site name/ unit number Initial activity Ci Fraction of activity in trash Fraction of activity in activated metals Time of first pit / crack yr Total volume of waste packages m

Rinse volume per cycle m

Initial surface area m

Total surface area of solidified waste m

Effective diffusivity em/s Comparison of Results Figure 2 compares releases of cesium-137 from a hypothetical LLW disposal unit estimated using RELEASE and - four other computer codes.

Two of the codes, IMPACTS and DIFFUSE, are based on alternative approaches [11 and 12] for estimating source term releases and were written by the author as computer programs in BASIC to support the comparison of the approaches.

The " IMPACTS" model is basically the submodel used in NRC's generic rational for 10 CFR Part 61 and is based on the algorithm used by the Impacts Analysis Methodology to calculate the waste form and pa;:kage barrier factor [11].

The " DIFFUSE" model solves the infinite sheet diffusion-limited release model (12].

The

!)

3 i

1 fn.

[.

3:

i eE b]

i h

o]

RINSE l

[

b" RELEAst

~1

)

i t

p.

eg

• • 1 u

1 x

9-o PACK

.$ "l.

.T o-1 A.'

OlFFUsE t

• ..,'O *.. g f i

g

,, o %.. 0"o..

l

-r -

o-1

, W - - -,- e- -6 4 4 4.,6,,, - 60"...

Efo" IMPACTS A ',1

-1 8

e t

o]

~

- t.

.r, o]

s T

~

i o\\

L o~1

\\\\

. i, s.

l.

9 -

'. 's t

j-o

.4

-1

\\ ',

u'a 3

O

'o _

l.

i 6

e i.

. i ii,si i'

10 10 10'

,,iii' 10 1

Time (yr)

Figure 2.

Estimated releases of Cesium-137 frcrn a hypothetical LIM disposal unit. See text for discussion of models RELEASE, RINSE, PACK, DIFFUSE, and IMPACTS.

e

'I

last two models listed in limited) submodels that have been incorporated in RELEASE. Figure 2 are th Their results have been included in the figure to assess the relative release contribution from LLW trash and solidified waste for the sample problem.

The input information i

for RINSE and PACK has been modified from that for the sample problem to represent all waste as LLW trash for the RINSE model and as solidified LLW for the PACK model.

These modifications account for the large differences in the estimated releases between RINSE, PACK, and RELEASE.

i Figure 2

indicates that 14 LEASE provides relatively conservative

)

estimates of radionuclido releases compand with the other models. The RINSE model provided the most conservative estimates, as expected, because it i

assumes that the entire radioraclide inventory is contained in the LLW as surface contamination.

This approach is clearly conservative if much of the cesium-137 inventory is contained in solidified waste forms and is not

)

readily accessible.

Release estimates from RINSE decrease below the other estimates after 100 years because most cf the waste activity has already been released.

The IMPACTS model provides the lowest estimated source term releases for the first 100 years.

It is important to recognize that the radionuclide releases estimated by IMPACTS for the sample problem are not

]

necessarily representative of the results of generic applications of the Impacts Analysis Methodology. Although the source term releases estimated by

)

IMPACTS are lower than the estimates from the other models, the relative conservatism of the Impacts Analysis Methodology is influenced by conservatisms introduced by such components as disposal unit design and performance, and environmental transport.

The comparison of releases in Figure 2 does not account for these components because the models described only estimate releases.from the waste form into the disposal unit.

EUTURE WORK The RELEASE model described in this paper is being developed and tested-by the NRC staff to provide necessary capabilities for estimating or bounding radionuclide releases from LLW. RELEASE represents one of several alternative approaches that are currently being evaluated by the NRC staff.

Although preliminary comparisons of the performance of RELEASE with other approaches indicates that RELEASF may provide relatively conservative estimates, i

additional work is necessary to determine whether RELEASE provides radionuclide release estimates that are adequate and defensible in reviewing license applications for LLW disposal.

Ongoing NRC studies of commercial LLW disposal facilities may provide information about actual releases of t

radionuclides that may be compared with releases estimated using RELEASE, IMPACTS, and other models.

For example, NRC staff and contractors are i

!~

developing transport models for the

inactive, commercial LLW disposal facility at Sheffield, Illinois, which may be useful in estimating actual releases by inverse modeling.

These estimates could then be used in calibrating and val uating RELEASE and other approaches for estimating radionuclide releases from LLW.

Further evaluation of the performance of RELEASE and other approaches is warranted to ensure that the approaches provide reasonably conservative or best estimates of radionuclide releases L

that are consistent with observations of radionuclide releases from LLW.

i L

t

i n

The purpose of the staff's evaluation is to select a preferred approach that can be integrated into a comprehensive performance assessment methodology for LLW disposal facilities.

In the future, the NRC staff may propcse such an approach in regulatory guidance or as part of standard review procedures for LLW disposal license applications.

Prior to selection or rejection of RELEASE as an estimation approach, the NRC staff will further assess the performance of RELEASE to evaluate its capacity to provide reasonably conservative or bust citimates of radionuclide releases from LLW based on reasonable input informa tion.

In particular, the

' staff intends to compare the estimates from RELEASE with laboratory and field measurements of releases from LLW.

The staff also intends to evaluate alternative and more rigorous approaches to assess the release of radionuclides from activated metal components, which comprise most of the projected activity in LLW in the United States. Further, the staff intends to assess the importance of processes that RELEASE does not currentiy consider such as biodegradation and wet / dry cycling.

The goal of these assessments is to enhance the capabilities of RELEASE or alternative approaches to provide reasonably conservative or best estimates of radionuclide releases from LLW.

SUMMARY

Estimation of the source term for radionuclide releases to the environment is required to assess long-term performance of disposal facilities i

for LLW. The processes and phenomena that centrol radionuclide releases from LLW are relatively poorly understood.

The diversity of waste, facility, and environmental characteristics that influence radionuclide releases currently preclude use of complicated approaches to estimate releases.-

Simple approaches that can be damonstrated to provide reasonably conservative estimates are, therefore, preferable to more complicated approaches. Based on laboratory leach tests and actual performance of LLW disposal units, the NRC staff and its contractors have developed an approach for for estimating radionuclide releases. This approach has been developed into a computer code entitled RELEASE.

The NRC staff is currently assessing the capability of RELEASE to provide reasonably conservative or best estimates of radionuclide releases from LLW.

Preliminary comparison of results from RELEASE with two other approaches indicates that RELEASE can provide relatively conservative estimates of radionuclide releases.

Prior to selection of a preferred approach for estimating radionuclide releases, the NRC staf f intend to further evaluate the performance and capabilities of RELEASE and other approaches.

i The author wishes to express his appreciaticn for NRC staff and contractor support of this effort.

Because of the preliminary nature of the work described in this paper, the conclusions provided above are those of the author and do not necessarily represent those of the U.S. Nuclear Regulatory Commission or its staff.

I REFERENCES 1.

U.

S. 14uclea r Regulatory Commission: " Licensing Requirements for Land

-Disposal of Radioactive Waste", 10 CFR Part 61, 1982.

2.

U. S. Nuclear Regulatory Commission: " Standard Review Plan for the Review of a License Application for a Low-Level Radioactive Waste Disposal Facility",

NUREG-1200, 1987.

g e.

e e

' 3.

Siefkin, D.,

Pangburn, G.,

Pennif111, R.,

and R.

J. Starmer: " Branch Technical Position on Site Suitability, Selection, and Characterization", U.S.

J Nuclear Regulatory Commission, NUREG-0902, 1932.

4.

Sullivan, T., and C. R. Kempf: " Low-Level Waste Source Term Evaluation:

i l

Review of Published Modeling and Experimental Work, and Presentation of Low-LeWi Waste Source Term Modeling Framework and Prelimina ry Model Development", U.S. Nuclear Regulatory Commission, NUREG/CR-4897, 1987.

5.

V.

S.

Nuclear Regula tory Commission:

" Draft Environmental Impact Statement on 10 CFR Part 61 ' Licensing Requirements for Land Disposal of Radioactive Waste'", NUREG-0782, 1981.

6.

U. S. Nuclear Regulatory Commission: " Final Environmental Impact Statement on 10 CFR Part 61 ' Licensing Requirements for Land Disposal of Radioactive Waste'", NUREG-0945, 1982.

7.

International Atomic Energy Agency:

" Conditioning of Low-and Intermediate-Level Radioactive Wastes",

Technical Reports Series No. 222, 1983.

B.

American Nuclear Society: " Measurement of Leachab111ty of Solidified Low-Level Radioactive Wastes", American Nuclear Society Standards Committee, Final Draft of Standards ANS-16.1, 1984 1

9.

Dougherty, D. R., Fuhrman, M., and P. Columbo: " Accelerated Leach Test (s)

Program Annual Report", Brookhaven National Labora tory, BNL-51955, 1985.

10. Hespe, E. 0.: " Leach Testing of Immobilized Radioactive Waste Solids: A Proposal for a Standard Method", Atomic Energy Review, v.

9, pp 195-207, 1971.

11.

Oztunali, O. I., Re, G. C, Moskowitz, P. M., Picazo, E. D., and C. J.

Pitt: " Data Base for Radioactive Waste Management:

Impacts Methodology Report", U.S. Nuclear Regulatory Commission, NUREG/CR-1759, v.3,1981.

12.

Arora, H.,

and R.

Dayal: " Leaching Studies of Cement-Based Low-Level Radioactive Waste Forms", U.S. Nuclear Regulatory Commission, NUREG/CR-4756,

1986,

13. Essington, E. H., Fuentes, H. R., Polzer, W. L., Lopez, E. A., and E. A.

Stallings: " Leaching of Solutes from lon-Exchange Resins Buried in Bandelier Tuff",

U.S. Nuclear Regulatory Commission, NUREG/CR-4592, 1986.

14.

MacKenzie, D.

R.,

Smalley, J.

F., Kempf, C. R., and R. E. Barletta:

" Evaluation of the Radioactive Invetory in, and Estimation of Isotopic Releases f rom, the Waste in Eight Trenches at the Sheffield Low-Level Waste Burial Site", U.S. Nuclear Regulatory Commission, NUREG/CR-3865, 1985.

15.

International Atomic Energy Agency:

" Deep Underground Disposal of Radioactive Wastes:

Near-Field Effects", Technical Report Series No. 251, 1985.

l l'!

p[, w eV, ' '

M+,

f.

s.

..U.

S.. Nuclear Regulatory Commission:."Oraft Environmental' Impact J

U

'16.

Statement on 10 CFR - Part 61 ' Licensing Requirements for Land 01sposal: of Radioactive Waste': Appendices G-Q", NUREG-0782, v. 3 1931.

.17.

Oztunali O.

I.,

and G. W. Roles: " Update of Part 61 Impacts Analysis Methodology",,U.S.f Nuclear Regulatory Commission, NUREG/CR-4370, v.

1, 1986.

18.

Wild,- R. E., 0ztuna11, O. I.

Clancy, J. J., Pitt O. J.,. and i E.

D.

-Picazo: - "Data-: Base for Radioactive Waste Management:

Waste - Source Options

, Report",o u;S. Nuclear Regulatory Commission, NUREG/CR-1759. v. 2,1981.

o f

fY l u~

U.

p

[,:

.i j.

i 1

e-6

,d,..

f}':

1 e

1 a

5 j

i.,,

y Q~

[...

q-