ML20206H700

From kanterella
Jump to navigation Jump to search
Speech Entitled Disturbed Zone & Groundwater Travel Time in NRC High Level Waste Rule (10CFR60), Presented on 860303-06 at Waste Mgt 1986 Conference in Tucson,Az
ML20206H700
Person / Time
Issue date: 03/03/1986
From: Codell R, Tanious M
NRC OFFICE OF NUCLEAR MATERIAL SAFETY & SAFEGUARDS (NMSS)
To:
Shared Package
ML20206H509 List:
References
NUDOCS 8606260269
Download: ML20206H700 (18)
v • d • e


Text

- ,

Conference: Waste Management '86 March 3-6,1986 Tucson, Arizona DISTURBED ZONE AND GROUND WATER TRAVEL TIME IN THE NPC HIGH LEVEL WASTE RULE (10 CFR 60)

Richard Codell and Natem Tanious Division of Waste Management Office of Nuclear Materials Safety and Safecuards U.S. Nuclear Regulatory Comission Washington D.C. 70555 ABSTRACT The hRC ground water travel time (GWTT) ob.iective of 1000 years is a quantitative but simple meas-ure of the hydrologic merit of the geolegic setting of sites for High Level Waste repositorie:. The disturbed zone is a surface surrounding the buried waste, which serves as the startino point for calcu-lation of GWTT. thereby avoiding complicated phenomena close to the waste. A distance of 50 meters from the waste is suggested as the minimum size of the disturbed zone by considering the effects of re-pository construction and heat generated by the decay of the waste. Alternative paths of potential ra-dionuclide transport must be explored for the site and the fastest one chosen. The GWTT along the fastest path will be a probability distributio't because of uncertainty, spatial variability and tempor-al variability. The criterion for satisfying the GWTT ob,iective is chnsen to represent travel times near the short end of the distribution. The bases for the choice of the disturbed zone definition and GWTT cbjective are discussed in detail in this paper.

INTRODUCTION Contr.18.d Ar Ground water is the most likely means by which Acc significant ouantitles of radionuclides could escape a omad sunece en ..mi.,

High Level Waste (HLW) repository. Release of radio- W hwwW "% we % m nuclices through ground water pathways is limited pri-maaily by th'ee " barriers": (1) the waste package and O"'*******

eNineered fa:111ty, (2) the rate of ground water flow ' G*ad**'*' F'** Pa'a l from the waste to the environment, and (3) geo- -

l chemical interaction of radionuclides with the rock g along the path of ground water movement. The present ,7 paper Leals only with the second barrier, y \ '*"*"*

The U.S. Nuclear Regulatory Commission (NRC) has Y b established performance objectives for the geologic setting of HLW repositories. One of these criteria, (Schematic - not to scale) commonly referred to as the " pre-waste-emplacement ground water travel time" or "GWTT" objective is stated as follows: rig,s . nign t ,,, w ,,,, ,,,,,,,,,,

"The geologic repository shall be located so that pre waste-emplacement ground water travel time along the fastest path of likely radionuclide travel from the disturbed zone to the accessible The ground water travel time performance objec-environment shall be at least 1000 years or such tive was established by NRC to serve as a quantita-other travel time as may be approved or specified tive, yet conceptually simple measure of the waste by the Commission"8 (10 CFR 60.113(a)(2)). isolation potential cf the geologic setting at candi-date repository sites. The travel time criterion was The " disturbed zone" cited in the above criterion

  • * ** #"""*
  • 9" **" '#

is the portion of the controIled area where changes '" ""* * "

caused by construction or the heat generated by radio- * " '" """*

active decay of the waste are great enough to have a significant effect on the performance of the reposito- The disturbed zone definition is intended to es-ry. The accessible environment" is defined as the tabli e the inner boundary from which GWTT is deter-atmosphere, land surface, surface water, oceans and mined. The evaluation of GVTT is simp 1tffed by the portion of the lithosphere that is outside of the avofding consideration of the effects of complicated controlled area. In this case, the deff ned con- and ill-defined processes related to the construction trolled area is consistent with the Final EPA high of the repository and the heat generated by the waste.

level waste rule' as extending no more than 5 kilome- This paper summarizes the NRC staff's guidance on the l ters from the orfginal emplacement of the waste in the GWTT objective and disturbed zone definition, which we anticipate will be contained in two draft Generic Tech-disposal system, with a maximum land surface area of nical Positions 8

  • The intent of the disturbed zone no more than 100 square kilometers. The disturbed is discussed first, followed by a description of the zone, accessible environment and path are depfcted in staff's attempt to place meaningful limits on its Fig. 1. size.

8606260269 860619 PDR MISC 8606260147 PDR

e INTENT OF DISTURBE0 ZONE that outside of this volume the pre waste-emplacement and post-waste-emplacement ground water travel times The volume of the rock which contributes to iso- would be identical. Unfortunately, it is likely that latice of HLW from the accessible environmeat clearly the disturbed zone defined in this way would be quite has its outer boundary at the accessible erstrormer.t. extensive and in some cases might extend beyond the For tho' 9rer boundary, the edge cf the ' underground boundary of the accessible environment. Effects such facility as shown in Fig. I would at first seem a sen- as buoyancy may have a significant impact on ground sible choice. However, the sta N 1as identified two water movement and radionuclide transport in terms of considerations in establishing tt e starting point for overall repository performance, and they must be taken GWTT calculations which preclude adoption of the edge into account in assessing total system performance.

of the underground facility as the origin. The staff considers that assessment of pre-waste-emplacement GWTT, representing the geologic component First, the travoi time criterion is intended to of the multiple barrier system, need only be based on ensure that the geologic setting provides significant pre-emplacement conditions, except where the intrinsic p~otection frt.N HLW releases to the accessible envi- properties of the rock which affect ground water flow ronmerc. Thus it constitutes a component of the are likely to be compromised as a result of HLW heat muittpie-barrier aspreach to HLW isolation. The staff generation or underground facility construction.

cMsiders that geologic barriers at a given site should not be permitted to depend exclusively or pre. For the purposes of evaluating the extent of the domfrantly on the favorable properties of the host disturbed zone at a given site, the staff will consid-rock directly adjacent te,the underground facility. er that a change in porosity by a factor of two con-Instead, the staff considers that an acce.9 table repos, stitutes a significant adverse e'fect on repository itory site would De one where the bulk of the sur, performance. Several theoretical relationships (e.g.,

rounding geologic settfrg contributes ts Jsolation of pa allel plate analogy) and laboratory studies demon-HLW.

strate that permeability can increase by an order of magnitute if porosity doubles. The increase in the

. 5dcond, the staf f considers that credit towards velocity of the grourd water is somewhat less, being the 1000 year pre-emplacement travel time should not partially offset by a lowering of the effective be taken within that portion of the current geologic porosity, setting (the "near-ffeld") that might be substantially Furthermore, the velocity in the affected region disturbed by construction of the facility or by the tight be controlled by external forces unaffected by thermal affects of emplacement of HLW (irrespective of the local changes in porosity. It is unlikely, however the possible offsetting benefits of engineered barri. that permeability or porosity could be characterized ers). Because of potential changes in the intrinsic within a factor of two at any of the proposed HLW rock properties, the geologic setting within the dis. Sites, because of the variability of the rock media as well as mezsurement inaccuracies.

turbed zone may not be well represented by pre. -

emplacement properties and corditions. Thus it my be difficult to predict the contributions of snis volume Based on the technical and policy considerations of rock to repository performance'. A pre-emplacement described above, the staff considers that the dis-analysis based on existing conditions within this lone turbed zone could be theoretically calculated through would not be an app spriate measure of the quality of evaluation of the spatial extent of changes to the in-the geologic settin; fcr the purpose of assessimt fu. trinsic hydraulic properties of the rock caused by ture performance. To avoid having to deal with .no i (1) stress redistfibution , (2) construction and uncertainties of characterizing the rock very close to excavation 4 (3) thermomechanical effects, and the emplaced waste, the disturbed zene was established (4) ther m hem cal effects.

as the inner boundary from which travel time calcula-tions are to be made, The discussion belcw is based on current concep-tual designs of HLW facilities, and considers only Shafts and surface boreholes are excluded from those rock types currently under consideration for the disturbed zone. The staff recognizes that the geologic repositories. It may not apply directly for rock immediately surrounding these cpanings will be greatly different designs, or sites with local geolog-disturbed to some extent, and may become a flow path ic anomalies. The approach demonstrates that the connecting the repository to the accessible environ. above phenomena can conservatively be assumed to be ment. Rathsr than assess the significance of these contained within a certain distance from the emplaced shaft for the GWTT determination, the performance of waste. The interactions of complex processes that .

shaft and borehole seals should be assessed as part of would be considered in identifying the spatial extent '

the overall performance assessment of the repository of intrinsic permeability and porosity changes have system. not yet been sufficiently studied to provide guidance in estimating their effect on the extent of the dis-Interpretation of the Disturbed Zone Definition turbed zone. ')

Having clarified the intent of the disturbed zone " '

concept, attempts to place meaningful limits on its Rock permeability may be significantly altered in size are now discussed. In the definition provided the region immediately surrounding repository openings earlier, the disturbed zone is described as the zone as a result of stress redistribution caused by the re-of physical or chemical property changes resulting moval of rock previously supporting a load. Tha imme-from underground facility construction or HLW heat gen- diste vicinity around the opening is most affected by eration that would significantly affect the perfor- the presence of the opening. The effect gradually de-mance of the repository. It includes the zone where creases with distance back to its pre-excavation coupled process interactions significantly perturb the condition. $1te-specific experimental results, pre-physical and chemical properties of the geologic dominately on small scale laboratory samples, relate setting.

permeability changes to stress changes. There are no The staff does not interpret the definition of universal relationships, however, which can be applied the disturbed zone to ful'y encompass temperature and * " " * " * ""I'* * " " " ' " "

1.e.,it may n t be possible to determine the stress buoyancy effects. It would be desirable of course to

  • ""I' # * #'"I' "'"** # * ""I' include these effects within the disturbed zone so l

, by a factor of two. Lacking a relationship betwIen openings and the position of the emplaced waste could stress change and permeability change, it is assumed change over time as a result of salt creep. The ex-that permeability will not change in the volume of tent of these changes should be considered in delin-rock beyond the surface of no stress change. There- eating the boundaries of the disturbed zone in salt fore, this boundary can be used to define the regio, media.

of no permeability change, i.e., the limit of the disturbed zone resulting from stress redistribution. A second important difference between salt and other rocks is the applicability of the concept of The distance to the contour of no stress change ground water flow. Traditional concepts of ground wa-from the edge of the opening wil' vary depending on ter flow through porous media and jointed rocks do not the size and shape of the opening and its orientation appear applicable to pure salt. However, changes in in the stress field. For a circular or semicircular permeability for the flow of brine or gas may occur as '

opening in an isotropic, homogeneous medium, a reason- a result of excavation, stress redistribution, and +

able estimata of this distance for the idealized case creep, t will be about three times the diameter of the opening'. For a noncircular opening, a reasonable es- For repositories constructed in salt, the staff timate is roughly five times the hetght. Other con- considers that the 50-meter envelope around repository siderations such as anisotropy of the rock, and the openings encompasses a sufficiently large zone to pre-degree of fracturing and subsequent yielding can po- vent the reliance on the portion of the host rock in-tentially increase this distance' '. mediately adjacent to the underground facility (i.e.,

within fifty meters of any opening) as the exclusive Taking the envelope of the above stress distribu- natural geologic barrier to radionuclide release. If 1 tions for a reasonable range of anticipated field con- geologic anomalies in the salt (e.g., gas pockets, l ditions, the no-stress-change contour could be brine cavities) are present, site specific analyses of conservatively estimated in many cases to be about 5 their potential impact on the size of the disturbed diameters for circular openings or 5 times the opening zone should be provided.  ;

j' height for noncircular openings. However, site- '

specific information must be utilfred in order to gain Construction and Excavation a realistic estimate of the extent of the disturbed ,

zone for a given site based on stress redistribution. The zone of mechanical damage caused by  !

construction-induced effects is usually smaller than '

Typical openings in the repository are in the that due to stress redistribution (with perhaps the range of about I to 10 meters in width, as illustrated exception of massive unjointed rocks). The extent of in the example presented in Fig. 2. Based on calcula- this zone depends on three factors: (1) the method of tions for homogeneous, isotropic, and linearly elastic excavation (blasting or boring); (2) the type of rock; media, and taking into consideration the effects of and (3) the extent of discontinuities in the rock, j host rock anisotropy, in-situ stress conditions, frac- Porosity and permeability changes caused by dewatering turing and yielding, it is estimated that the dis- of the facility should also be considered on a site-turbed zone caused by stress redistribution for the scocific basis. The extent of damage due to con- i simplified example described above may extend to a trolled blasting rarely extends more than 1.5 meters distance of 5 times the opening diameter, or 5 to 50 I or greater than half the opening diameter from the

, meters, from the edge of the opening depending on the edge of the opening'. Tunnel and shaft boring results

opening size. in little damage to the surrounding rock. Tnerefore, the approximately 50 meter distance from the edge of the opening, established as the minimum distance to t the boundary of the disturbed zone resulting from stress redistribution, should also conservatively en-i compass any effects of construction-induced changes in 3

- f f rock permeability.

, , Thermomechanical Effects

,,,,,,,,) 3 - - -

Heat generated by the radioactive decay of i

y ,,,,,,,,,,,,/ emplaced waste may thermally stress the host rock and

? 7 **"

surroeding e.edia. Thermal stresses can cause changes in permeability by creating new cracks and open or

, "b"*"******* close existing joints. In the far field, thermal

stresses may cause uplift and eventual subsidence ef-l a fects. The resulting permeability changes may be more i significant in salt than in hard rocks due to salt

( ) J creep and a relatively high coefficient of thermal expansion.

Experimental assessment of changes in rock perse-ability as a result of heat application has shown that Fle. 2 - Plecoment _Roosne ene Storese Heles _ '

permeability increases with temperature because of i

thermal cracking'. However, confining pressure can strongly reduce permeability increases caused by heat.

In a repository environment, permeability increases i

The mechanical properties of salt are signift- close to the repository wall may largely diminish as cantly different from other rocks and therefore a sep- confining pressure increases in the surrounding rocks, arate discussion is warranted. A unique feature of In jointed rock, some or all of the thermal expansion

salt is its ability to creep into and close excavated may be absorbed by the closure of microcracks and openings over periods of only a few years to hundreds joints resulting in overall reduction in the of years depending on the salt's mechanical proper- permeability.

ties, its depth, prevailing stresses, and other geo-logic phenomena and anos.altes. The shape of the i

h

. , - - - - - - , - - - - - ,- - ,., ,, , . , ,-, -, ,,- _--n i. , in -, w a n..m.-. -nn --. ,, .,-..n

-. .- - . __. - ... -. . . - - , - . -. - - - . _-- - ~

i i

  • a
  • j =* ,

j Considering the above discussion, the disturbed INTERPRETATION OF WTT OBJECTIVE

zone which covers the effect of stress redistribution may generally be expected to include the zone of sig- Assessment of the GWTT objective requires (1) nificant thermomechanical effects on permeability. proper characterization of the pre waste-emplacement l environment and its potential spatial and short-term .

Thermochemical Effects temporal variabilities, (2) identification of the l paths of likely radionuclide travel, (3) calculation lI Thermochemical reactions include alteration of of the appropriate travel time along the likely paths, pre-emplacement minerals, dehydration, and precipita- and (4) picking the fastest travel time. Compliance ]

l tion of secondary minerals. A change in porosity and with the objective requires a demonstration that the permeability of the host rock may result from altera- fastest GWTT exceeds 1000 years. The steps necessary j tions of pre emplacement minerals caused by increased to demonstrate compliance are discussed below.

I temperature and/or altered ground water composition.

These reactions can result in either increases or de- Identiffeation of Pre-Waste-Emplacement Environment i

creases in solid volume.

Pre waste-emplacement pertains to conditions Oehydration reactions may result in a not in- which exist prior to significant disturbance of the l l crease in porosity of the host rect, especially for hydrological setting by construction activities and i

, hydrous minerals, and in the region surrounding the major testing activities capable of serinusly disturb-1 HLW canisters where temperatures are highest. This ing the geologic setting. Restriction of the GWTT re-

' will be of particular importance to a repository lo- quirement to pre-disturbance conditions is consistent cated in the unsaturated zone. Dehydration of zoo- with the original intent of 10 CFR 60 to establish a lites can occur at temperatures as low as 85'C. straightforward criterion which is easily defined and determined. The staff recognizes the importance of Redistribution of solid phases (e.g., dissolution post weste-emplacement effects. Evaluation of ground ' I close to the waste and precipitation at a distance) water and radionuclide eavement under post weste- .

results in porosity increases in one portion of the emplacema t conditions will be required as part of the '

repository and decreases la another. The not effect demonstration of overall compliance of the repository on the flow may be favorable or unfavorable. Generic with the EPA s.tandards as implemented by NRC.

calculations for the redistribution of silica * ", a common mineral found at at least two of the HLW sites The determination of GWTT will be for present day 1_

currently under consideration, indicate that dissolu- environmental conditions only. Short-term changes to tion is not likely to be significant beyond the previ- the environment, (e.g., tens of years) which can be ously discussed distance of 50 m for mechanical reasonably inferred from records in the vicinity of disturbance * ". It is apparent, however, that the the site and any other factors that may explain tran-distance to the edge of the thermochemically disturbed sient responses in the hydrologic system should be 4

zone is strongly dependent on the thermal loading of considered when developing the conceptual model(s) for the repository and the ground water flux in the host determining GWTT whenever practicable. For example, rock. The staff recommends that analyses of thermo- decreases in ground water levels caused by pumping for chemical effects be performed with site-specific val- 1rrigation may emplain transient responses in nearby

, ues of the hydrological, mineralogical and chemical aquifers. Other esemples include cycles of wet and parameters for the repository system. These calcula- dry years, local flooding and changes in land use.

tions will roughly indicate the suitability of the 5

suggested minimum distance based on stuss If present-day conditions have changed markedly redistribution in encompassing the zone of thermo- uver the period of record, the investigator must ques-chemical chariges at the site, tion whether inappropriate credit is being,taken for increased ground water travel times caused by these j Sumary for Disturbed Zene changes. For example, if a cone of depressicn has i formed as a result of large ground water withdrawals,

! The NRC staff considers that establishment of ge- it could reduce or even reverse the direction of an neric and easily evaluable guidarice on the disturbed unfavorable hydraulic gradient. This could lead to

! zone is desirable in order to simplify the demonstra- the possibility that GWTT calculated on the basis of tion of compliance with the ground water travel time present-day gradtents are overly optimistic and not criterion. Based on the information provided above, it representative of the geologic setting. In this case, appears that a distance of five opening diameters from it would be prudent to consider the effects of an any underground opening, escluding surface shafts and otherwise-likely hydraulic gradient absent the effect surface boreholes, would be a reasonably conservative of the cone of depression. The rational _e behind this distance for the extent of the mechanically-disturbed philosophy is to avoid the appearance of taking credit zone in some cases. Given current conceptual designs for processes for which there could be no assurances for underground MLW facilities, this would imply a of long-term reliability, e.g., continued ground water distance of roughly fifty meters from the underground withdrawals maintaining the favorable gradient.

openings. The limit of one process (silica dissolu-tion) contributing to the thermochemically disturbed Identification of Paths of Likely Radionuclide Travel zone, based on a simplified evaluation, appears to be less than the mechanically-disturbed distance from the The paths from the disturbed zone to the accessi-I- underground facility. However, the thermochemically ble environment are to be described in a macroscopic disturbed zone at a site should be calculated on a sense. Paths must be potentially capable of carrying site- and design-specific basis, taking into account a significant quantity of ground water. In the sense the hydrochemical, geochemical, hydrologic and thermal employed here, "significant" is relative to the abso-conditions for each site. The impact of stress redis- lute quantity of ground water moving by the site. In t

tribution, construction and thermomechanical effects crystalline or fractured rocks, paths may consist of should also be considered on a site- and design- fractured, weathered, brecciated or porous zenes. In specific basis where practicable, porous media, paths will generally consist of layers of porous, permeable sediments. Paths may also con-sist of fractured zones in consolidated porous media.

In a uniform medium, a distinct conduit for radio-9 1

-_ y __.,_,_ , _ - , , , . _ . _ _ - ,,m.,..,., ,,.p,y , y ,yy -m,,,y .r., , y

s nuclide transport may not be evident, in which case contrasts in hydraulic conductivity leading to perched the path will be defined by the considering the hy- water tables. The possibility of perched water under draulic gradient and hydraulic conductivity. reasonably conceivable conditions (e.g., a series of wet years that could occur under present climatic con-Cata from the site could support several alterna- ditions) should be considered, even if such conditions tive conceptual models for the repository setting, currently do not exist at the site. possible connec-each of which might determine a different path for tions of perched water to fractures or other structur-radionuclide transport. For example, limited borehole al features of the site, which would provide taformation sight indicate the presence of permeable preferential pathways for radionuclide migration zones, but the investigator may be unable to determine through the unsaturated material in which the repos1-whether or not these zones were connected in such a tory would be placed, should also be considered.

may that they represent a path.

The analysis of GWTT, therefore, should consider all paths for radionuclide transport defined by alter-native conceptual models, unless they can clearly be C'" ass a .

@ U O '* DMO f V^M-  %

demonstrated to be unlikely or insignificant. Collec- -

tion of data at the site must be directed to identify- ""m uan A ing these paths, establishing the validity of the p,..,w,,.,,.,,,

vant.e, P in conceptual models for interpreting and sleulating the __  :

hydrogeology, and making a reasoned determination that o e r.a. ,

potentially faster paths have not been overlooked. , , , , , , , r% - F<=i-.

Underground facilities located in saturated media will usually be emplaced in a rock unit of low p w. p wa.,

  1. ---- ( um , p. n permeability. More permeable units may underlie and over11e the repository horizon, however, as shown in Fig. 3. Some of these horizons may intersect the dis- namuanc turbed zone. While there may be little movement of ground water in the host rock under natural condi- uneveen tions, factors which could cause the movement of _P.".! 't.P.*. i!28.*

radionuclides from the disturbed Zone to these more -

permeable horizons should be taken into account.

Paths of transport between horizons could occur via fracture or porous flow under natural hydraulle nam uan o gradlents.

D  ?

.~. _

n.e ..., v u. . ...n l ---4.-**'.**.'.***----------- _

, , , , , , , Determination of Travel Time Ground water travel time must be deterinined for e,ui ,. each of the likely paths determined in the previous step. The GWTT.is a distributed variable rather than bh.1, P.th a fixed Quantity. Several reasons exist for the dis-sa. e e _

tributed nature of travel time. These are mechanical dispersien, molecular diffusion, data uncertainty, the distributed nature of the source and accessible envi-ronment, and the transient nature of ground water sa**+8 c f

/

,N 3

\bkdv'"a flow. These phenomena are discussed below. It is helpful in the following discussion to introduce the concept that GWTT may be determined 3 simulating the

/ g  % release and migration of inert, infinitesimal tracer

""*""""*'_ particles which behave identically to the molecules of

{ (D*'**+8 2*a* water:

j

'"**** * \ / o Dispersion - Ground water travels only in the

  • % / '"*'P"" open spaces (pores, fractures) in the rock.

There would be numerous possible particle trajec-Fle. 3 - Paths in Saturated Medium tories within each path, each with a different travel time. In any real situation, natural spa-J tial variability in the properties of the medium Identification of paths for repository sites in such as porosity and hydraulic conductivity, unsaturated media, as illustrated in Fig. 4, will will cause the tracer particles to disperse, differ from those in saturated media. The flow is Tracer particles moving in the ground water will likely to be predominantly in the downward direction follow trajectories governed by the hydraulic until the water table it, reached. In some cases, the prcperties of the medium at their location. The path may be defined in terms of the direction of the less uniform the medium, the less parallel will gradient, unless there are barriers to flow such as De the trajectories and the greater the range'of

, e

.. . , )

l i

the arrival times of the tracer particles at the variabilities of the data are usually taken into ac-boundary of the accessible environment. count by obtaining many solutions, each one based on a different statistical realization of the parameter o Molecular diffusion - Random movement on a very set. Such techniques are generally known as " Monte small scale allows water and solute molecules to Carlo" simulations, diffuse into openings in the rock where there may be little or no not flow of the water. Molecular Each Monte Carlo run requires the solution of the diffusion may be important in cases where there hydraulic head and velocity field for the values of is appreciable matriz porosity and generally slow parameters and boundary conditions chosen randomly, movement of ground water". but following certain statistical rules based on the data. The solution is generally accomplished by solv-o Uncertainty - Measurement error and limits of da- ing the partial differential equations (PDE's) that ta necessary to characterize the site add uncer- describe ground water flow using techniques such as tainty to the hydrologic parameters used to finite differences or finite elements. Once the ve-calculate GWTT, which is translated into uncer- locity field is known, travel time distributions can tainties in the GWTT estimates. This uncertainty be calculated by simulating the release of tracer par-can be combined into the probability distribution ticles from single or multiple locations along the of arrival times for tracer particles, disturbed zone boundary and counting their arrival times as they reach the accessible environment" ".

o 01stributed source - The disturbed zone and ac-cessible environment are defined as surfaces Stochastic models deal with the variability and

, rather than points. Tracer particles released at uncertainty of the data in a more direct way". The different points along the disturbed Zone surface coefficients and variables of the equations are treat-will reach the boundary of the accessible envi- ed as random fields rather than deterministic quanti-ronment at different times, ties. The PDE's are solved indirectly in terms of the maents of the dependent variables (e.g., mean and o Temporal variability - Hydrologic and hydro- variance). This technique has the advantage of re-geologic data within recorded history of the site quiring only one solution rather than the numerous might indicate that the ground water velocities Monte-Carlo solutions for the deterministic approach.

are fluctuating. This fact afght be particularly Direct stochastic approaches to modeling are much less important at sites built in unsaturated media. developed than Monte Carlo techniques, although it is For example, it is conceivable that a period of an area cf rapid development.

unusually heavy precipitation for several years (unreleated to a global climatic change) could With either type of model, the computations must increase hydraulic heads, decreasing travel times be performed with parameters inferred from the avail-along a normally slow pathway. A transient GWTT able data in order to generate the GWTT distribution.

should be weighted according to its frequency and The types and quality of data available will determine duration. In addition, a path which changes di- how the computations are to be performed. For exam-rection or length over time as a result of vari- ple, if only a few data points are available for a able fluxes of ground water should be considered particular parameter, a conservative estimate of that to be a single path for the purposes of GWTT cal- parameter may have to be made and carried through the culations. This allows the low probability, fast calculations. With more data, a mean and variance of GWTT's to be fairly weighed with the high proba- theparameterscanbecalculatedandusedwithasim-bility, slow GWTT's. pie sampling approach. If the site is well-characterized, spatially varying properties of the Analysis of the GWTT for a repository must depend parameters can be employed".

on methods of indirect inference fror observations of hydrogeologic data at the site. Artificial tracers Ground water travel time can be represented as a are useful in some cases, but the time (thousands of distribution for each of the paths in terms of a Cumu-years) and distance (kilometers) scales are too great lative Distribution Function (C0F), shown in Fig. 5.

for direct characterization of GWTT at HLW repository This CDF combines all spatial variability, temporal by such methods. Naturally occurring isotopes and variability and quantifiable uncertainty of the GWTT those produced from atmospheric weapons testing and into a single curve for each of the likely paths. The nuclear reactors can be used for ground water age dat- CDF itself, however, is assumed to contain na uncer-ing to support estimates of travel time distributions tainty. It is important to note that the CDF does not for real sites. Such techniques should be used when- exclude the existence of uncertainty, but that all ever practicable, although investigations must usually Quantifiable uncertainty is incorporated into the CDF.

resort to mathematical models of the repository for Compliance with the 1000 year GWTT objective would be predictions of performance. demonstrated if it could be shown that aiong any of the paths, a tracer particle leaving the disturbed Values of travel time from the disturbed zone to zone has a (100-X)5 or greater probability of arriving the accessible environment are usually obtained from at the accessible environment in a time greater than mathematical models consisting of the equations gov- 1000 years, where X is a small number. The choice of erning the hydraulic potential, flow of ground water, X is discussed in the next section. i and transport of a tracer. There are many models for l ground water flow in various media which are based on The GWTT computed using this general guidance j the equations at steady state or transient conditions will be sensitive to the degree of character 12ation of in one, two or three dimensions". Currently, the the site. That is, investigators of poorly-most prevalent method of determining the GWTT distri- characterized sites will be forced to use conservative' bution is by the use of deterministic models with or at least overly wide distributions to represent the inputs generated randomly" ". large uncertainties in the input parameters. Sites that have been characterized on the basis of defensi-Deterministic models consist of equations whose ble conceptual models will facilitate the development solution is based on the assumption of known parame- of a more tefensible GWTT distribution function by re-ters, e.g., hydrogeologic properties, initial condi- ducing the variance of parameter distributions.

tions and system geometries. Uncertainties and

J e

to 3 i i , , o the same time, t' or (2) the particles arrive gradual-ly after t'. Case 2 would oaviously be more favorable in terms of repository performance, but this fact n - -

ei could not be determined if the "first particle" or ze-ro percentile criterion for the GWTT distribution had been chosen. The choice of a higher percentile would u - -

u give proper credit to Case 2.

A finite percentile also avoids some of the dif-p _ _

n ficulties caused by model inaccu.~acies and inadequa-cies. For example, calculations for solute

, , concentrations are frequently made using the 5u 3

n53 convective-diffusion equation. This equation can pre-dict small but non-zero concentrations of solute ap-8 pearing instantaneously at any point in the '

k" y

~ ~

"5 l computational field, in violation of the laws of phys-tes. Similar inaccuracies can be caused by approxi-5 i mations made in the numerical solution of the 5 9 equations. These extremely small concentrations are "E

  • recognized as erroneous and are not usually cause for concern, but the "first particle" criterion would present difficulties which could be easily avoided by 83 - -

88 the choice of a larger percentile.

A choice for the percentile which is too high, sa - -

os say the median, would be undesirable because it may be

'" ~~ ~ ~- ~ ~ ~ - - ~ ~ ~ - ~ ~ ~ ~ ~ ,,, insensitive to the variance of the GWTT distribution.

This can be demonstrated for the hypothetical example ci - I -

es depicted in Fig. 7. The two CDF curves of GWTT in I this figure have the same median of 1000 years, but I different variances. Under the median GWTT criterion, e ' ' ' ' '8

  1. # '# # # # # # sites which exhibit a wide variance of the travel time distribution for reasons such as great spatial vart-onous,owaren raavst rwa n.w a'u111ty, dispersion, matrix diffusion, or uncertainty, would be treated as equals. This leaves open the pos-Fig. 5 - GWTT Oletribution sibility that in the case of the curve with the wider variance, a substantial fraction of the tracer parti-cles could arrive at the accessible environment in Percentile (X) Criterion for GWTT Distributior: less than 1000 years. A smaller percentile justifiably favors the site which has the smaller variance in the At the upper and lower limits of the GWTT distri- GWTT distribution. If the wider variance is due to bution for each of the paths, there will be ground wa- quantifiable uncertainty i (e.g., lack of data), the ter travel times which, although possible, are so smaller percentile would serve as an incentive to fur-unlikely that they are inappropriate measures of GWTT. ther characterize the site. A smaller percentile As an example, consider the two hypothetical cases favors the site which has the smaller variance in the shown in Fig. 6 for which (1) all particles arrive at GWTT distribution.

i------

Narrow D6stribution

2 i I

= Wide Olstribution o . .

E  ; n, ig o s ------------ ----  ?-----

! E i O 5 i 3 i i o

u I

c is --- -----

~~~

0 OM - - - - - - - - ---

1o0o m rius l

I Fig. 7 - Cumulative oletributione of GWTT with

, i 0 ,

rwa same Median but Different verlence Fle. e

  • Cua ulative Oletitutione of GWTT with Same Minimum but Olfferent Verlence

s The percentile for the C0F as the criterton for Pre waste-emplaenent pertains to conditions at GWTT is presently unspecified, but the rationale from the site prior to any significant disturbance of the the above two examples suggests that a value greater hydrological or geological setting such as construc-than a few percent and considerably less than 50% tion activities or the effects of radioactive waste, would be destreable. The determination of the percen- and whose spatial and temporal variability can be rea-tile for the GWTT criterion should also be based on sonably inferred from historical records at or near considerations of " reasonable assurance". Licensing the site. Testing activities capable of altering the considerations to be made in connection with GWTT in- pre-waste-emplacement environment should be taken into volve substantial uncertainties, many of which are consideration.

unquantifiable (e.g., those pertaining to the correct-ness of the models used to evaluate GWTT). Such un- CONCLUSIONS certainties can be accommodated within the licensing process only if a qualitative test such as reasonable The ground water travel time objective was estab-assurance is applied for the level of confidence that 11shed as a simple but meaningful measure of the merit the numerical performance objective is expected to of the geologic setting of a high level waste repos1-achieve. Both the quantifiable uncertainties incorpo- tory. The disturbed Zone was defined to avoid the no-rated in the GWTT distributton and the unquantifiable cessity of quantifying complicated processes which are uncertainties which are not included must be consid- likely to occur close to the emplaced waste, and to ered together in reaching a finding of reasonable as- prevent reliance on a relatively small thickness of surance. It might, for example, be proper to select a rock to satisfy the GWTT objective.

different percentile criterion for a relatively well-understood, easily modeled site (where unquantifiable The staff has endeavored to present a workable uncertainties are small) than would be appropriate for definition of the pre waste emplacement ground water a site with larger unquantifiable uncertainties. travel time objective to be used for HLW repository Stated another way, selection of the percentile crite- licensing. The definition will assist the staff in rien is a qualitative judgement, and is part of a evaluating compliance of a specific site with the per-larger set of judgements necessary to reach a finding formance objectives of 10 CFR 60. This paper is how-of reasonable assurance that the performance objec- ever intended to be guidance only. It reflects the tives will be a:hteved. staff's current interpretation of the disturbed zone definition and GWTT objective. It does not prevent Note that the user is not required to generate a others from advancing alternative interpretations.

detailed CDF of the GVTT distribution. A simplified approach would be acceptable, provided that achieve- The draft Generic Technical Positions on which ment of the 1000 year GdTT objective could be demon- the present paper is based'

  • have not yet been made strated with reasonable assurance. Such a simplified available for public comment. This paper may not approach could for example, define a conservatively therefore reflect the final NRC positions on the dis-short path along which tne travel time of a single turbed zone and ground water travel time. The authors particle could be calculated using Darcy's law with encourage comments and suggestions on the topics cov-conservatively chosen coefficients of hydraulic con- ered in this paper.

ductivity, gradient and effective porosity.

ACKNOWLEDGEMENTS Summary for GWTT Large portions of this paper were drawn from Ground water travel time is a measure of the mer- Refs. 3 and 4 The authors therefore acknowledge Mat-it of the geologic setting of a high level waste re- thew Gorden, John Bradbury and Linda Kovach for their pository. The staff recognizes that there may be written contributions. The text of this paper was ca-alternative conceptual models of the site because of pably reviewed by Malcolm Knapp, Michael Bell, Myron the inability to completely characterize it with the Fliegel, John Bradbury, Michael Weber, Seth Coplan, available data. This inability may lead to a multi- Daniel Fehringer and James Wolf. Special thanks go to plicity of paths for likely radionuclide travel. The Michael Weber for his many useful suggestions.

ground water travel time along the paths will be a distributed quantity because of spatial variability, REFERENCES temporal variability, the distributed nature of the disturbed zone and accessible environment, and model 1. U.S. Nuclear Regulatory Commission, Rules and or data uncertainty. Ground water travel time should Regulations, Title 10, Part 60, " Disposal of therefom be represented as a et.mulative distribution, High-level Radioactive Waste in Geologic Reposito-although a single valued GWTT would be acceptable if ries", Code of Feceral Requations (1986).

it were derived from appropriately conservative models and coefficients. The " pre waste emplacement ground 2. U.S. Environmental Protection Agency, Rules and water travel time along the fastest path of likely Regualtions, Title 40, Part 191, " Environmental Stan-radionuclide travel" should be represented as a per- dards for the Management and Disposal of Spent Nuclear centile of all travel times' contained in the Cumula- Fuel, High-Level Waste and Transuranic Radioactive tive Distribution Function (CDF) for the fastest of Wastes; Final Rule", Code of Federal Regulations the identified paths. Compliance with the 1000 year (1986).

GWTT objective would be demonstrated if it could be shown that along any of the paths, a tracer particle 3. R.B. Codell, " Draft Generic Technical Position on leaving the disturbed zone has a (100-X)% or greater Ground Water Travel Time", Division of Waste Manage-probability of arriving at the accessible environment ment, Office of Nuclear Materials Safety and Safe-in a time greater than 1000 years, where X is a small guards, U.S. Nuclear Regulatory Commission (1986).

number. A numerical value of X has not been specified by the staff at this time.

)

. . , ,, l 4

4. M. Gordon, N. Tantous. J. Bradbury. L. Kovach and 11. C.R. Faust, and J W. Mercer, " Ground Water Model-R. Codell, "Oraf t Generic Technical position: Inter- ing: Numerical Models", Groundwater Vol.18, p. 395 pretation and Identification of the Extent of the Dis- (1980) turbed Zene in the High Level Waste Rule (10 CFR 60),"

Division of Waste Management, Office of Nuclear Mate- 12. J.W. Mercer, and C. Faust, " Ground Water Model-rials Safety and Safeguards, U.S. Nuclear Regulatory ing: An Overview", Grosndwater, Vol.18, p.108 (1980)

Commission (1986).

13. L.F. Smith, and F. Schwartz, " Mass Transport:
5. E. Hoek, and E.T. Brown. " Underground Excavation 1. A Stochastic Analysis of Macroscopic Dispersion",

in Rock", The Institution of Mining and Metallurgy, Water Resources Research, Vol.16, no. 2, pp. 303-313 London, England (1980). (1980)

6. R.E. Goodman, Introduction to Rock Mechantes, 14. P.M. Clifton, " Ground Water Travel Time Uncer-John Wiley & Sons, New Yort (1980). tainty Analysis - Sensitivity of Results to Model Go-ometry, and Correlations and Cross Correlations among
7. 0.F. Coates, " Rock Mechanics Principles", Mines Input Parameters," Report no BWI-TI-256, Rockwell Branch Monograph 874, Department of Energy, Canada Hanford Operations. Hanford WA (1984)

(1970).

15. R.W. Nelson, " Evaluating the Environmental Conse-
8. P.C. Kelsall, J.B. Case, and C.R. Chabaness "A quences of Ground Water Contamination Parts I-IV, "

Preliminary Evaluation of the Rock Mass Disturbance Water Resources Research, Vol. 19, no.3 pp. 409-450 Resulting from Shaft Tunnel, or Borehole Excavation." (1978)

Technical Report OWI-411 Office of Nuclear Waste Isolation, Battelle Memorial Institute, Columbus, Chio 16. L. Gelhar, and C.L. Axness, "Three-Ofmensional (1982). Analysis of Macredispersion in Aquifers", Water Resources Research, Vol.18, no. 1, pp. 161 W 1983)

9. J.J.K. Daemen, " Rock Mass Sealing - Experimental Assessment of Borehole Plug Performance", 17. ONWI, "A Proposed Approach to Uncertainty Analy-NUREG/CR-3473 U.S. Nuclear Regulatory Commission sis," ONWI-488., Office of Nuclear Waste Isolation, (1983). Battelle Memorial Institute Columbus, OH (1983)
10. J.W. Braithwaite, and F.B. Nimick, "Effect of Host-Rock Olssolution and Precipitation on Permeabili-ty in a Nuclear Waste Repository in Tuff,"

$AN084-0192, Sandia National Laboratories, Albuquerque NM (1984).

4

Conference: Waste Management '86 March 3-6,1986 Tucson, Arizona DISTURBED ZONE AND GROUND WATER TpAVEL TIME IN THE NRC HIGH LEVEL WASTE RULE (10 CFR 60)

Richard Codell and Nefem Tanious Division of Waste Management i Office of Nuclear Materials Safetv and Safeguards U.S. Nuclear Regulatory Commission Washington D.C. 70555 ABSTRACT The NRC ground water travel time (GWTT) ob.iective of 1000_vears is a quantitative but simple meas-ure of the hydrologic merit of the geologic setting of sites for High Level Waste repositories. The disturbed zone is a surface surrounding the buried waste, which serves as the startino point for calcu-lation of GWTT. thereby avoiding compitcated phenomena close to the waste. A distance of 50 meters from the waste is suggested as the minimum size of the disturbed zone by considering the effects nf re-pository construction and heat generated by the decay of the waste. Alternative paths of potential ra-dionuclide transport must be explored for the site and the fastest one chosen. The GWTT along the fastest path will be a probability distribution because o' uncertainty, spatial variability and tempor-al variability. The criterion for satisfying the GWTT objective is chosen to represent travel times near the short end of the distribution. The bases for the choice of the disturbed zone definition and GWTT objective are discussed in detail in this paper.

INTRODUCTION c ,.n., 4, Ground water is the most likely means by which Acc m.

significant quantities of radionuclides could escape a High Level Waste (HLW) repository. Release of radio-

    • ""#*** N $'"'"*"'

nuclides through ground water pathways is limited pri-

"Nwu # ^%i m e%

marily by three " barriers": (1) the waste pactage and **'""*"* -~

engineered facility, (2) the rate of ground water flow ' G *""** * F* Ph l from the waste to the environment, and (3) geo- -

l chemical interaction of radionuclides with the rock

  • j along the path of ground water movement. The present -

paper deals only with the second barrier, g . 7, The U.S. Nuclear Regulatory h,mmission (NRC) has

  • N **** **

established performance objectives for the geologic setting of HLW repositories. One of these criteria, (senemonic - no . ,ces.)

commonly referred to as the " pre waste-emplacement ground water travel time" or "GWTT" objective is rio,i . wien t.,,, ,,,,,n stated as follows: ,,,,,,,,

"The geologic repository shall be located so that pre waste-emplacement ground water travel time along the fastest path of likely radionuclide travel from the disturbed zone to the accessible The ground water travel time performance objec-environment shall be ac least 1000 years or such tive was established by NRC to serve as a quantita-tive, yet conceptually simple measure of the waste other travel time _as may be approved or specified isolation potential of the geologic setting at candi-by the Commission 8 (10 CFR 60.113(a)(2)).

date repository sites. The travel time criterion was The " disturbed zone" cited in the above criterion * * ** #"" " E **'U" **"* '#

is the portion of the contro.11ed area where changes "*"" ' """ " "I ""

caused by construction or the heat generated by radio- * ** **'"* "" "*"

  • active decay of the waste are great enough to have a significant effect on the performance of the reposito- The disturbed zone definition is intended to es-ry. The " accessible environment" is defiried as the tablish the inner boundary from which GWTT is deter-atmosphere, land surface, surface water, oceans and mined. The evaluation of GWTT is simplified by the portion of the lithosphere that is outside of the avoiding consideration of the effects of complicated controlled area. In this case, the defined con- and ill-defined processes related to the construction trolled area" is consistent with the Final EPA high of the repository and the heat generated by the waste.

level waste rule

  • as extending no more than 5 k11eme- This paper summarizes the NRC staff's guidance on the ters from the original emplacement of the waste in the GWTT objective and disturbed zone definition, which we anticipate will be contained in two draft Generic Tech-disposal system, with a maximum land surface area of nical Positions' *. The intent of the disturbed zone no more than 100 square kilometers. The disturbed is discussed first, followed by a description of the zone, accessible environment and path are depicted in staff's attempt to place meaningful limits on its Fig. 1. size.

~ -

INTENT OF O!STURBED ZONE that outside of this volume the pre waste-emplacement and post waste-emplacement ground water travel times The volume of the rock which contributes to iso- would be identical. Unfortunately, it is likely that lation of HLW from the accessible environment clearly the disturbed zone defined in this way would be quite has its outer boundary at the accessible environment, extensive and in some cases might extend beyond the For the inner boundary, the ecge of the underground boundary of the accessible environment. Effects such facility as shown in Fig. I would at first seem a sen- as buoyancy may have a significant impact on ground sible choice. However, the staff has identified two water movement and radionuclide transport in terms of considerations in establishing the starting point for overall repository performance, and they must be taken GWTT calculations which preclude adoption of the edge into account in assessing total system performance, of the underground facility as the origin. The staff considers that assessment of pre-waste-emplacement GWTT, representing the geologic component First, the travel time criterion is intended to of the multiple barrier system, need only be based on ensure that the geologic setting provides significant pre-emplacement conditions, except where the intrinsic protection from HLW releases to the accessible envi- properties of the rock which affect ground water flow ronment. Thus it constitutes a component of the are likely to be compromised as a result of HLW heat multiple-barrier approach to HLW isolation. The staff generation or underground facility construction.

considers that geologic barriers at a given site should not be permitted to depend exclusively or pre. For the purposes of evaluating the extent of the dominantly on the favorable properties of the host disturbed zone at a given s.te, the staff will consid-rock directly adjacent to the underground facility. er that a change in porosity by a factor of two con-Instead, the staff considers that an acceptable repos- stitutes a significant adverse effect on repository itcry site would be one where the bulk of the sur. performance. Several theoretical relationships (e.g.,

rounding geclogic setting contributes to isolation of parallel plate analogy) and laboratory studies demon-HLW.

strate that permeability can increase by an order of magnitute if porosity doubles. The increase in the Second, the staff considers that credit towards velocity of the ground water is somewhat less, being the 1000 year pre-emplacement travel time should not partially offset by a lowering of the effective be taken within that portion of the current geologic porosity, setting (the "near-field") that might be substantially Furthermore, the velocity in the affected region disturbed by construction of the facility or by the might be controlled by external forces unaffected by thermal effects of emplacement of HLW (irrespective of the local changes in porosity. It is unlikely, however the possible offsetting benefits of engineered barri- that permeability or porosity could be characterized ers). Because of potential changes in the intrinsic within a factor of two at any of the proposed HLW rock properties, the geologic setting within the dis- sites, because of the variability of the rock media as turbed zone may not be well represented by pre- well as measurement inaccuracies.

emplacement properties and conditions. Thus it may be difficult to predict the contributions of this volume Based on the technical and policy considerations of rock to repository performance. A pre-emplacement described above, the staff considers that the dis-analysis based on existing conditions within this zone turbed zone could be theoretically calculated through would not be an appropriate measure of the quality of evaluation of the spatial extent of changes to the in-the geologic setting for the purpose of assessing fu- trinsic hydraulic properties of the rock caused by ture performance. To avoid having to deal with the (1) stress redistfibution , (2) construction and uncertainties of characterizing the rock very close to excavation , (3) thermomechanical effects, and the emplaced waste, the disturbed zone was established (4) thermochemical effects, as the inner boundary from which travel time calcula-tions are to be made. The discussion below is based on current concep-tual designs of HLW facilities, and considers only Shafts and surface boreholes are excluded from those rock types currently under consideration for the disturbed zone. The staff recognizes that the geologic reoositories. It may not apply directly for rock immediately surrounding these openings will be greatly.different designs, or sites with local geolog-disturbed to some extent, and may become a flow path ic anomalies. The approach demonstrates that the connecting the repository to the accessible environ- above phenomena can conservatively be assumed to be ment. Rather than assess the significance of these contained within a certain distance from the emplaced shaft for the GWTT determination, the performance of waste. The interactions of complex processes that shaft and borehole seals should be assessed as part of would be considered in identifying the spatial extent the overall performance assessment of the repository of intrinsic permeability and porosity changes have system. not yet been sufficiently studied to provide guidance in estimating their effect on the extent of the dis-Interpretation of the 01sturbed Zone Definition t"'D'd 28"'-

Having clarified the intent of the disturbed Zone " "

concept, attempts to place meaningful limits on its Rock permeability may be significantly altered in size are now discussed. In the definition provided the region immediately surrounding repository openings earlier, the disturbed zone is described as the zone as a result of stress redistribution caused by the re-of physical or chemical property changes resulting moval of rock predously supporting a load. The imme-from underground facility construction or HLW heat gen- diate vicinity around the opening is most affected by eration that would significantly affect the perfor- the presence of the opening. The effect gradually de-mance of the repository. It includes the zone where creases with distance back to its pre-excavation coupled process interactions significantly perturb the condition. Site-specific experimental results, pre-

, physical and chemical properties of the geologic dominately on small scale laboratory samples, relate setting, permeability changes to stress changes. There are no The staf f does not interpret the definition of universal relationships, however, which can be applied the disturbed zone to ful'y encompass temperature and to all rocks under wide ranges of stress. changes; buoyancy effects. It would be desirable of course to 1.e. it may not be possible to determine the stress include these effects within the disturbed zone so change which would cause the permeability to change

- _ _ . _ _ . - - ~. - -. , -

1 l

l l .

i I by a factor of two. Lacking a relationship between openings and the position of the emplaced waste could stress change and permeability change, it is assumed change over time as a result of salt creep. The ex-that permeability will not change in the volume of tent cf these changes should be considered in delin-rock beyond the surface of no stress change. There- eating the boundaries of the disturbed zone in salt fore, this bodndary can be used to define the region media.

of no permeability change, i.e., the limit of the disturbed zone resulting from stress redistribution. A second important difference between salt and other rocks is the applicability of the concept of The distance to the contour of no stress change ground water flow. Traditional concepts of ground wa-from the edge of the opening will vary depending on ter flow through porous media and jointed rocks do not the size and shape of the opening and its orientation appear applicable to pure salt. However, changes in in the stress field. For a circular or semicircular permeability for the flow of brine or gas may occur as opening in an isotrcpic, homogeneous medium, a reason- a result of excavation, stress redistribution, a>d able estimate of this distance for the idealized case creep, will be about three times the diameter of the opening'. For a noncircular opening, a reasonable es- For repositories constructed in salt, the staff timate is roughly five times the height. Other con- considers that the 50-meter envelope around repository siderations such as anisotropy of the rock, and the openings enccmpasses a sufficiently large zone to pre-degree of fracturing and subsequent yielding can po- vent the reliance on the portion of the host rock im-tentially increase this distance' '. mediately adjacent to the unde ground facility (i.e.,

within fifty meters of any opening) as the exclusive Taking the envelope of the above stress distribu- natural geologic barrier to radionuclide release. If tions for a reasonable range of anticipated field con- geologic anomalies in the salt (e.g., gas pockets, ditions, the no stress-change contour could be brine cavities) are present, site specific analyses of conservatively estimated in many cases to be about 5 their potential impact on the size of the disturbed diameters for circular openings or 5 times the opening zone should be provided.

heig5t for noncircular openings. However, site-specific information must be utilized in order to gain Construction and Excavation a realistic estimate of the extent of the disturbed zone for a given site based on stress redistribution. The zone of mechanical damage caused by con _struction-induced effects is usually smaller than Typical openings in the repository are in the that due to stress redistribution (with perhaps the J range of about I to 10 meters in width, as illustrated exceptionofmassiveunjointedrocks). The extent of in the example presented in Fig. 2. Based on calcula- this zone depends on three factors: (1) the method of tions for homogeneous, isotropic, and linearly elastic excavation (blssting or boring); (2) the type of rock; media, and taking into consideration the effects of and (3) tha extent of discontinuities in the rock.

host rock anisotropy, in-situ stress conditions, frac- porosity and permeability changes caused by dewatering turing and yielding, it is estimated that the dis- of the facility should also be considered on a site-turbed zone caused by stress redistribution for the specific basis. The extent of damage due to con-simplified example described above may extend to a trolled blasting rarely extends more than 1.5 meters distance of 5 times the opening diameter, or 5 to 50 or greater than half the opening diameter from the meters, from the edge of the opening depending on the edge of the opening'. Tunnel and shaft boring results opening size. in little damage to the surrounding rock. Therefore, the approximately 50 meter distance from the edge of the opening, established as the minimum distance to the boundary of the disturbed zone resulting from stress redistribution, should also conservatively en-

-i ' compass any effects of construction-induced changes in

-'f rock permeability.

  • --a== Thermomechanical Effects 4

3 Heat generated by the radioactive decay of

  • X - ~~g emplaced waste may thermally stress the host rock and 7

surrounding media. Thermal stresses can cause changes in permeability by creating new cracks and open or

//**-* close existing joints. In the far field, thermal stresses may cause uplift and eventual subsidence ef-

[ t f I fects. The resulting permeability changes may be more Y significant in salt than in hard rocks due to salt

(

s

) J creep and a relatively high coefficient of thermal expansion.

e-Experimental assessment of changes in rock perme-ability as a result of heat application has shown that Fig. 2 - Placement Rooms and 8torage Holes . permeability increases with temperature because of thermal cracking'. However, confining pressure can strongly reduce permeability increases caused by beat.

In a repository environment, permeability increases The mechanical properties of salt are signift- close to the repository wall may largely diminish as cantly different from other rocks and therefore a sep- confining pressure increases in the surrounding rocks.

arate discussion is warranted. A unique feature of In jointed rock, some or all of the thermal expansion salt is its ability to creep into and close excavated may be absorbed by the closure of microcracks and 1 openings over periods of only a few years to hundreds joints resulting in overall reduction in the of years depending on the salt's mechanical proper- permesbility. j ties, its depth, prevailing stresses, and other geo- '

logic phenomena and anon.alies. The shape of the 1

l t

l

i Considering the above discussion, the disturbed INTERPRETATION OF GWTT OBJECTIVE zone which covers the effect of stress redistribution may generally be espected to trclude the zone of sig- Assessment of the GWTT objective requires (1) nificant thermomechanical effects on permeability. proper characterization of the pre waste-emplacement environment and its potential spatial and short-term Thermochemical Effects temporal variabilities (2) identification of the paths of likely radionuclide travel, (3) calculation Thermochemical reactions include alteration of of the appropriate travel time along the Itkely paths, pre-emplacement minerals, dehydratton, and precipita- and (4) picking the fastest travel time. Compliance tion of secondary minerals. A change in porosity and with the objective requires a demonstration that the permeability of the host rock may result from altera- fastest GWTT exceeds 1000 years. The steps necessary tions of pre-emplacement minerals caused by increased to demonstrate compliance are discussed below.

temperatu*e and/or altered ground water composition.

These reactions can result in either increases er de- Identification of Pre-Waste-Emplacement Environment creases in solid volume.

Pre waste-emplacement pertains to conditions Dehydration reactions may result in a net in- which exist prior to significant disturbance of the crease in porosity of the host rock, especially for hydrological setting by construction activities and hydrous minerals, and in the region surrounding the major testing activities capable of seriously disturb-HLW canisters where temperatures are hfghest. This ing the geologic setting. Restriction of the GWTT re-will be of particular importance to a repository lo- quirement to pre-disturbance conditions is consistent cated in the unsaturated zone. Dehydration of zeo- with the original intent of 10 CFR 60 to establish a lites can occur at temperatures as low as 85'C. straightforward criterion which is easily defined and determined. The staff recognizes the importance of Redistribution of solid phases (e.g., dissolution post waste-emplacement effects. Evaluation of ground close to the waste and precipitation at a distance) water and radionuclide movement under post waste-results in porosity increases in one portion of the emplacement conditions will be required as part of the repository and decreases in another. The net effect demonstration of overall compliance of the repository on the flow may be favorable or unfavorable. Generic with the EPA s.tandards as implemented by NRC.

calculations for the redistribution of silica * ", a common mineral found at at least two of the HLW sites The determination of GWTT will be for present day currently under consideration, indicate that dissolu- environmental conditions only. Snort-term changes to tion is not likely to be significent beyond the previ- the environment (e.g., tens of years) which can be ously discussed distance of 50 m for mechanical reasonably inferred from records in the vicinity of disturbance * ". It is apparent, however, that the the site and any other factors that may explain tran-distance to the edge of the thermochemically disturbed sient responses in the hydrologic system should be Zone is strongly dependent on the thermal loading of considered when developing the conceptual model(s) for l the repository and the ground water flux in the host determining GWTT whenever practicable. For example, rock. The staff recommends that analyses of thermo- decreases in grcund water levels caused by pumping for chemical effects be performed with site-specific val- 1rrigation may explain transient responses in nearby ues of the hydrological, mineralogical and chemical aquifers. Other examples include cycles of wet and parameters for the repository system. These calcula- dry years, local flooding and changes in land use.

tions will roughly indicate the suitability of the suggested minimum distance based on stress If present-day conditions have changed markedly redistribution in encompassing the zone of thermo- over the period of record, the investigator must oues-chemical changes at the site. tion whether inappropriate credit is being.taken for increased ground water travel times caused by these Summary for utsturbed Zone changes. For example, if a cone of depression nas formed as a result of large ground water withdrawals.

The NRC staff considers that establishment of ge- it could reduce or even reverse the direction of an neric and easily evaluable guidance on the disturbed unfavorable hydraulic gradtent. This could lead to i

zone is desirable in order to simplify the demonstra- the possibility that GWTT calculated on the basis of tion of compliance with the ground water travel time present-day gradients are overly optimistic and not criterion. Based on the information provided above, it representative of the geologic setting. In this case, appears that a distance of five opening diameters from it would be prudent to cont? der the effects of an I

any underground opening, excluding surface shafts and otherwise-likely hydraulic gradient absent the effect surface boreholes, would be a reasonably conservative of the cone of depression. The rational _e behind this distance for the extent of the mechanically-disturbed philosophy is to avoid the appearance of taking credit zone in some cases. Given current conceptual designs for processes for which there could be no assurances for underground HLW facilities, this would imply a of long-term reliability, e.g., continued ground water distance of roughly fifty meters from the underground withdrawals maintaining the favorable gradient.

t openings. The limit of one process (silica dissolu-

[ tion) contributing to the thermochemically disturned Identification of Paths of likely Radlonuclide Travel zone, based on a simpliffed evaluation, appears to be less than the mechanically-disturbed distance from the The paths from the disturbed zone to the accessi-underground factitty. However, the thermochemically ble environment are to be described in a' macroscopic disturbed zone at a site should be calculated on a sense. Paths must be potentially capable of carrying site- and design-specific basis, taking into account a significant quantity of ground water. In the sense the hydrochemical, geochemical, hydrologic and thermal employed here, "significant" is relative to the abso-conditions for each site. The impact of stress redis- lute quantity of ground water moving by the site. In I tribution, construction and thermomechanical effects crystalline or fractured rocks, paths may consist of should also be considered on a site- and design- fractured, weathered, brecciated or porous zones. In speciff e basis where practicable. porous media, paths will generally consist of layers of porous, permeable sediments. Paths may also con-sist of fractured zones in consolidated porous media.

In a uniform medium, a distinct conduit for radio-

,,--,,w-- ,-ee - - -,m.

- - , - -_.--,,v-,-e,,- - - ,m-w- _w, ,,, , , , - - - - - - - ,

\

l l

  • 1 nuclide transport may not be evident, in which case contrists in hydraulic conductivity leading to perched the path will be defined by the considering the hy- water tables. The possibility of perched water under '

draulic gradient and hydraulic conductivity. reasonably conceivable conditions (e.g., a series of )

wet years that could occur under present climatic con- 1 Data from the site could support several alterna- ditions) should be considered, even if such conditions '

tive conceptual models for the repository setting, currently do not exist at the site. Possible connec-each of which might determine a different path for tions of perched water to fractures or other structur-radionuclide transport. For example, limited borehole al features of the site, which would provide information might indicate the presence of permeable preferential pathways for radionuclide migration zones, but the investigator may be unable to determine through the unsaturated material in which the reposi-whether or not these zones were connected in such a tory would be placed, should also be considered.

way that they represent a path.

The analysis of GVTT, therefore, should consider all paths for radionuclide transport defteed by alter- C**d 5d"'

native conceptual models, unless they can clearly be M M W ' W W " M._lWO ' WMA' h demonstrated to be unlikely or insignificant. Collec- ""*""""

tion of data at the site must be directed to identify-ing these paths, establishing the validity of the p.. ,,,en.4 w..., u.i.m , e.,n conceptual models for interpreting and simulating the ----------i - - -

hydrogeology, and making a reasoned determination that ow az potentially faster paths have not been overlooked. " butwo

,,,,y,,,

Underground facilities located in saturated media

["tWNU" g i will usually be emplaced in a rock unit of low

g u,,,,,,,,

permeability. More permeable units may underlie and overlie the repository horizon, however, as shown in Fig. 3. Some of these horizons may intersect the dis- m ctumic turbed zone. While there may be little movement of ground water in the host rock under natural condt- umw e.m tions, factors which could cause the movement of 1 " d *1 "**** *L _"-. ------- -_-

radionuclides from the disturbed zone to these more permeable horizons should be taken into account.

Paths of. transport between horizons could occur via fracture or porous flow under natural hydraulic n cn una o gradlents.

O  ? . _

ss.e. ama n w .v ume e.m

_ _ _ m . m ,._ _ _ _ _ - _ _

,,,,,,,, Determination of Travel Time Ground water travel time must be determined for

,.e,,. each of the likely paths determined in the previous step. The GWTT.is a distributed variable rather than un.ev e.m a fixed quantity. Several reasons exist for the dis-we u _ tributed nature of travel time. These are mechanical dispersion, molecular diffusion, data uncertainty, the distributed nature of the source and accessible envi-ronment, and the transient nature of ground water flow. These phenomena are discussed below. It is

/N une e.* helpful in the following discussion to introduce the

    • '**.c / i concept that GWTT may be determined _by simulating the

/, j  % release and migration of inert, infinitesimal tracer

{

"" i'* T O ( C*****

2*a* particles which behave identically to the molecules of water:

g j

    • '**'\ / o Dispersion - Ground water travels only in the
  • %/ "'**** open spaces (pores, fractures) in the rock.

There would be numerous possible particle trajec-Fle. 3 - Pathe in Saturated M*Wum tories within each path, each with a different travel time. In any real situation, natural spa-tial variability in the properties of the medium Identification of paths for repository sites in such as porosity and hydraulic conductivity, unsaturated media, as illustrated in Fig. 4, will will cause the tracer particles to disperse.

differ from those in saturated media. The flow is Tracer particles moving in the ground water will likely to be predominantly in the downward direction follow trajectories governed by the hydraulic until the water table is. reached. In some cases, the properties of the medium at their location. The path may be de'ined in terms of the direction of the less uniform the medium, the less parallel wi11 gradient, unless there are barriers to flow such as be the trajectories and the greater the range of

. i I

the arrival times of the tracer particles at the variabilities of the data are usselly taken into ac-boundary of the accessible environment, count by obtaining many solutions, each one based on a different statistical realization of the parameter o Molecular diffusion - Random movement on a very set. Such techniques are generally known as " Monte small scale allows water and solute molecules to Carlo" simulations, diffuse into openings in the rock where there may be little or no net flow of the water. Molecular Each Monte Carlo run requires the solution of the diffusion may be important in cases where there hydraulic head and velocity field for the values of is appreciable matrix porosity and generally slow parameters and boundary conditions chosen randomly, movement of ground water', but following certain statistical rules based on the data. The solution is generally accomplished by solv-o Uncertainty - Measurement error and limits of da- ing the partial differential equations (PDE's) that ta necessary to charactertre the site add uncer- describe ground water flow using techniques such as tainty to the hydrologic parameters used to finite differences or finite elements. Once the ve-calculate GWTT, which is translated into uncer- locity field is known, travel time distributions can tainties in the GWTT estimates. This uncertainty be calculated by simulating the release of tracer par-can be combined into the probability distribution ticles from single or multiple locations along the of arrival times for tracer particles. disturbed zone boundary and counting their arrival times as they reach the accessible environment " ".

o Distributed source - The disturbed zone and ac-cessible environment are defined as surfaces Stochastic models deal with the variability and rather than points. Tracer particles released at uncertainty of the data in a more direct way". The different points along the disturbed zone surface coefficients and variables of the equations are treat-will reach the boundary of the accessible envi- ed as random fields rather than deterministic quanti-ronment at different times, ties. The POE's are solved indirectly in terms of the moments of the dependent variables (e.g., mean and o Temporal variability - Hydrologic and hydro- variance). This technique has the advantage of re-geologic data within recorded history of the site quiring only crie solution rather than the numerous might indicate that the ground water velocities Monte-Carlo solutions for the deterministic approach.

are fluctuating. This fa;t might be particularly 01 rect stochastic approaches to modeltag are much less important at sites built in unsaturated media. developed than Monte Carlo techniques, although it is For example, it is conceivable that a period of an area of rapid development.

unusually heavy precipitation for several years .

(unreleated to a global climatic change) could With either type of model, the computations must increase hydraulic heads, decreasing travel times be performed with parameters inferred from the avail-along a normally slow pathway. A transient GWTT able data in order to generate the GWTT distribution.

should be weighted according to its frequency and The types and quality of data avajlable will determine curation. In addition, a path which changes di- how the computations are to be performed. For exam-rection or length over time as a result of vart- ple, if only a few data points are available for a able fluxes of ground water should be considered particular parameter, a conservative estimate of that to be a single path for the purposes of GWTT cal- parameter may have to be made and cerried through the culations. This allows the low probability, fast calculations. With more data, a mean and variance of GWTT's to be fairly weighed with the high proba- the parameters can be calculated and used with a sia-bility, slow GWTT's. ple sampling approach. If the site is well-characterized, spatially varying properties of the Analysis of the GWTT for a repository must depend parameters can be employed".

on methods of indirect inference from observations of hydrogeologic data at the site. Artificial tracers Ground water travel time can be represented as a are useful in some cases, but the time (thousands of distribution for each of the paths in terms of a Cumu-years) and distance (kilometers) scales are too great lative Distribution Function (CDF), shown in Fig. 5.

for direct characterization of GWTT at HLW repository This CDF combines all spatial variability, temporal by such methods. Naturally occurring isotopes and variability and quantifiable uncertainty of the GWTT those produced from atmospheric weapons testing and into a single curve for each of the likely paths. The nuclear reactors can be used for ground water age dat- CDF itself, however, is assumed to contain no uncer-ing to support estimates of travel time distributions tainty. It is important to note that the CDF does not for real sites. Such techniques should be used when- exclude the existence of-uncertainty, but that all ever practicable, although investigations must usually quantifiable uncertainty is incorporated into the C0F.

resort to mathematical models of the repository for Compliance with the 1000 year GWTT objective would be predictions of performance. demonstrated if it could be shown that along any of the paths, a tracer particle leaving the disturbed Values of travel time from the disturbed zone to zone has a (100-X)% or greater probability of arriving the accessible environment are usually obtained from at the accessible environment in a time greater than mathematical models consisting of the equations gov- 1000 years, where X is a small number. The choice of erning the hydraulic potential, flow of ground water, X is discussed in the next section.

and transport of a tracer. There are many models for ground water flow in various media which are based on The GWTT computed using this general guidance the equations at steady state or transient conditions will be sensitive to the degree of characterization of in one, two or three dimensions". Currently, the the site. That is, investigators of poorly-most prevalent method of determining the GWTT distri- characterized sites will be forced to use conservative bution is by the use of deterministic models with or et least overly wide distributions to represent the in,)uts generated randomly' ". large uncertainties in the input parameters. Sites that have been characterized on the basis of defensi-Deterministic models consist of equations whose ble conceptual models will facilitate the development solution is based on the assumption of known parame- of a more defensible GWTT distribution function by re-ters, e.g., hydrogeologic properties, initial condt- ducing the variance of parameter distributions.

tions and system geometries. Uncertainties and

.9 .

18 i , , , , i 8 the same time, t' or (2) the particles arrive gradual-ly after t'. Case 2 would obviously be more fr.vorable in terms of repository performance, but this fact as - - ei could not be determined if the "first particle" or Ze- 1 ro percentile criterlon for the GWTT distribution had been chosen. The choice of a higher percentile would

,, _ ,, give proper credit to Case 2.

A finite percentile also avoids some of the dif-

.3 ficulties caused by model inaccuracies and inadequa-cies. For example, calculations for solute

, , concentrations are frequently made using the 5 ,, ,,$ convective-diffusion equation. This equation can pre-j _ .

dict small but non-zero concentrations of solute ap-8 8 pearing instantaneously at any point in the E computational field, in violation of the laws of phys-y

~ ~

"g e ics. $1m11ar inaccuracies can be caused by approxi-motions made in the numerical solution of the i

5 2 equations. These extremely small concentrations are

~ ~

"E recognized as erroneous and are not usually cause for concern, but the "first particle" criterion would present difficulties which could be easily avoided by 83 ~ ~

" the choice cf a larger percentile.

A choice for the percentile which is too high, en - - e' say the median, would be undesirable because it may be e ,. ___ _ _ _ _ _ _ _ _ _ _ e os insensitive to the variance of the GWTT distribution.

This can be demonstrated for the hypothetical example si . 1 - es depicted in Fig. 7. The two C0F curves of GWTT in I this figure have the same median of 1000 years, but I different variances. Under the median GWTT criterion,

' ' ' ' ' ' 18 e

  1. # # # # # # # sites which exhibit a wide variance of the travel time distribution for reasons such as great spatial vart-omou=owaran tnavn tmas iv.** ability, dispersion, matrix diffusion, or uncertainty, would be treated as equals. This leaves open the pos-'

Fig. 5 - GWTT Oletribution sibility that in the case of the curve with the wider variance, a substantial fraction of the tracer parti-cles could arrive at the accessible environment in Percentile (X) Criterion for GWTT Distribution less than 1000 years. A smaller percentile justifiably favors the site which has the smaller variance in the At the upper and lower limits of the GWTT distri- GWTT distribution. If the wider variance is due to bution for each of the paths, there will be ground wa- quantifiable uncert,ainty (e.g., lack of data), the ter travel times which, although possible, are so smaller percentile would serve as an incentive to fur-unlikely that they are inappropriate measures of GWTT. ther characterize the site. A smaller percentile, As an example, consider the two hypothetical cases favors the site which has the smaller variance in the shown in Fig. 6 for which (1) all particles arrive at GWTT distribution, t__----

Narrow 04stributlen t

E I

. Wide oistr6tuteen

, ;es ____-

. .i . -

2 E $ s a U j i 3 a l o

I e is --- -_---

i l

8 en - - _ _ _ _ - - , _ _

TIME g

I Fle. 7 - Cumulative Oletributtone of GWTT with

, t e e reus Same Medien but Olfferent Vertence Fle e - Cumulative Distitutione of OWTT with Game Minimum but Olfferent Verlence n.- -

The percentile for the CDF as the criterton for Pre waste-emplacement pertains to conditions at GWTT is presently unspecified, but the rationale from the site prior to any significant disturbance of the the above two examples suggests that a value greater hydrological or geological setting such as construc-than a few percent and considerably less than 50% tion activities or the effects of radioactive waste, would be desireable. The determination of the percen- and whose spatial and temporal variability can be ree-tile for the GWTT criterton should also be based on sonably inferred from historical records at or near censiderations of " reasonable assurance". Licensing tne site. Testing activities capable of altering the considerations to be made in connection with GWTT in- pre waste-emplacement environment should be taken into volve substantial uncertainties, many of which are consideration.

unquantifiable (e.g., those pertaining to the correct-l ness of the models used to evaluate GWTT). Such un- CONCLUSIONS '

certainties can be accommodated within the Itcensing process only if a qualitative test such as reasonable The ground water travel time objective was estab-assurance is applied for the level of confidence that 11shed as a simple but meaningful measure of the merit the numerical performance objective is expected to of the geologic setting of a high level waste repost-achieve. Both the quantifiable uncertainties incorpo- tory. The disturbed zone was defined to avoid the no-rated in the GWTT distributton and the unquantifiable cessity of quantifying complicated processes which are uncertainties which are not included must be consid- likely to occur close to the emplaced waste, and to ered together in reaching a finding of reasonable as- prevent relf ance on a relatively small thickness of i surance. It might, for example, be proper to select a rock to satisfy the GWTT objective.

different percentile criterion for a relatively well-i understood, easily modeled site (where unquantifiable The staff has endeavored to present a workable uncertainties are small) than would be appropriate for definition of the pre weste-emplacement ground water a site with larger unquantifiable uncertainties. travel time objective to be used for HLW repository Stated another way, selection of the percentile crite- licensing. The definition will assist the staff in rion is a qualitative judgement, and is part of a evaluating compliance of a specific site with the per-larger set of judgements necessary to reach a finding formance objectives of 10 CFR 60. This paper is how-of reasonable assurance that the performance objec- ever intended to be guidance only. It reflects the tives will be achieved. staff's current interpretation of the disturbed zone definition and GWTT objective. It does not prevent Note that the user is not required to generate a others from advancing alternative interpretations.

detailed CDF of the GWTT distribution. A simplified approach would be acceptable, provided that achieve- .

The draft Generic Technical Positions on which ment of the 1000 year GWTT objective could be demon- the present paper is based'

  • have not yet been made strated with reasonable assurance. Such a simplified available for pubite comment. This paper may not

. approach could for example, define a conservatively therefore reflect the final NRC positions on the dis-short path along which the travel time of a single turbed zone and ground water travel time. The authors particle could be calculated using Darcy's law with encourage comments and suggestions on the topics cov-conservatively chosen coefficients of hydraulic con- ered in this paper, ductivity, gradient and effective porosity.

ACKNOWLEDGEMENTS Summary for GWTT Large portions of this paper were drawn from Ground water travel time is a measure of the mer- Refs. 3 and 4. The authors therefore acknowledge Mat-it of the geologic setting of a high level waste re- thew Gorden, John Bradbury and Linda Kovach for their pository. The staff recognizes that there may be written contributions. The text of this paper was ca-alternative conceptual models of the site because of pably reviewed by MaleMa Knapp, Michael Bell, Myron the inability to completely characterize it with the F11egel, John Bradbury, Michael Weber, Seth Coplan, available data. This inability may lead to a multi- Daniel Fehringer and James Wolf. Special thanks go to p11 city of paths for likely radionuclide travel. The Michael Weber for his many useful suggestions, ground water travel time along the paths will be a distributed quantity because of spatial variability, REFERENCES temporal variability, the distributed nature of the i

disturbed zone and accessible environment, and model 1. U.S. Nuclear Regulatory Commission, Rules and or data uncertainty. Ground water travel time should Regulations, Title 10, Part 60, " Disposal of therefore be represented as a cumulative distribution, High-level Radioactive Waste in Geologic Reposito-although a single valued GWTT would be acceptable if ries". Code of Federal Recuations (1986).

I it were derived from appropriately conservative models l and coefficients. The " pre waste-emplacement ground 2. U.S. Environmental Protection Agency, Rules and water travel time along the fastest path of likely Regualtions, Title 40, Part 191 " Environmental Stan-l radionuclide travel" should be represented as a per- dards for the Management and Disposal of Spent Nuclear centile of all travel times' contained in the Cumula- Fuel, High-Level Waste and Transuranic Radioactive tive 01stributton Function (COF) for the fastest of Wastes; Final Rule", Code of Federal Regulations i the identified paths. Compliance with the 1000 year (1986).

GWTT objective would be demonstrated if it could be shown that along any of the paths, a tracer particle 3. R.B. Codell, " Draft Generic Technical. Position on leaving the disturbed zone has a (100-X)% or greater Ground Water Travel Time", Division of Weste Manage-probability of arriving at the accessible environment ment Office of Nuclear Materials Safety and Safe-in a time greater than 1000 years, where X is a small guards, U.S. Nuclear Regulatory Commission (1986).

number. A numerical value of X has not been specified by the staff at this time.

I __i.

._ _ m _ _ . _ _ ._ . _ _ . . _ _ _ _ . _ . __

i, The percentile for the C0F as the criterion for Pre weste-emplacement pertains to conditions at

. GWTT is presently unspecified, but the rationale from the site prior to any significant disturbance of the the above two examples suggests that a value greater hydrological or geological setting such as construc-than a few percent and considerably less than 50% tion activities or the effects of radioactive waste, would be desireable. The determination of the percen- and whose spatial and temporal variability can be rea-tile for the GWTT criterton should also be based on sonably inferred from historical records at or near considerations of " reasonable assurance". Licensing the site. Testing activities capable of altering the considerations to be made in connection with GWTT in- pre-waste-emplacement environment should be taken into volve substantial uncertainties, many of which are consideration.

unquantiflable (e.g., those pertaining to the correct-ness of the models used to evaluate GWTT). Such un- CONCLUSIONS certainties can be accommodated within the licensing process only if a qualttative test such as reasonable The ground water travel time objective was estab-assurance is applied for the level of confidence that 11shed as a simple but meaningful measure of the merit the numerical performance objective is expected to of the geologic setting of a high level waste reposi-achieve. Both the quantifiable uncertainties incorpo- tory. The disturbed zone was defined to avoid the no-

, rated in the GWTT distribution and the unquantifiable cessity of quantifying compitcated processes which are uncertainties which are not included must be consid- likely to occur close to the emplaced waste, and to ered together in reaching a finding of reasonable as- prevent reliance on a relatively small thickness of surance. It might, for example, be proper to select a rock to satisfy the GWTT objective, different percentile criterion for a relatively well-understood, easily modeled site (where unquantifiable The staff has endeavored to present a workable uncertatnties are small) than would be appropriate for definition of the pre-waste emplacement ground water a site with larger unquantifiable uncertainties. travel time objective to be used for HLW repository Stated another way, selection of the percentile crite- Itceesing. The definition will assist the staff in rion is a qualitative judgement, and is part of a jvaluating compliance of a specific site with the per-larger set of judgements necessary to reach a finding formance objectives of 10 CFR 60. This paper is how-of reasonable assurance that the performance objec- ever intended to be guidance only. It reflects the tives will be achieved. staff's current interpretation of the disturbed zone definition and GWTT objective. It does not prevent Note that the user is not required to generate a others from advancing alternative interpretations.

detailed CDF of the GWTT distributton. A simplified approach would be acceptable, provided that achieve-1 The draft Generic Technical Positions on which ment of the 1000 year GWTT objective could be demon- the present paper is based'

  • have not yet been made strated with reasonable assurance. Such a simplified available for public comment. This paper may not

, approach could for example, define a conservatively therefore reflect the final NRC positions on the dis-short path along which the travel time of a single turbed zone and ground water travel time. The authors

, particle coald be calculated using Darcy's law with encourage comments and suggestions on the topics cov-conservatively chosen coefficients of hydraulic con- ered in this paper.

ductivf ty, gradient and effective porosity.

ACKNOWLEDGEMENT 5

_Semary for GWTT Large portions of thf s paper were drawn from Ground water travel time is a measure of the mer- Refs. 3 and 4. The authors therefore acknowledge Mat-it of the geologic setting of a high level waste re- thew Garden, John Bradbury and Linda Kovech for their pository. The staff recognizes that there may be written contributtons. The text of this paper was ca-alternative conceptual models of the site because of pably reviewed by Malcolm Knapp, Michael Bell, Myron the inability to completely characterize it with the F11egel, John Bradbury, Michael Weber, Seth Coplan,

, available data. This inability may lead to a multi- Daniel Fehringer and James Wolf. Special thanks go to I

plicity of paths for like'. radionuclide travel. The Michael Weber for his many useful suggestions, ground water travel time along the paths will be a

, distributed quantity because of spatial variability, REFERENCES

temporal variability, the distributed nature of the disturbed zone and accessible environment, and model 1. U.S. Nuclear Regulatory Commission Rules and or data uncertainty. Ground sa:er travel time should Regulations Title 10. Part 60, "Of sposal of therefore be represented as a cumulative distribution, High-level Radioactive Weste in Geologic Reposito-although a single valued GWTT would be acceptable tf ries", Code of Federal Recuations (1986).

It wers derived from appropriately conservative models and cM fficients. The " pre waste-emplacement ground 2. U.S. Environmental Protection ' Agency, Rules and

water travel time along the fastest path of likely Regualtions. Title 40 Part 191, " Environmental Stan-radionuclide travel" should be represented as a per- dards for the Management and Otsposal of Spent Nuclear centile of all travel times' contained in the Cumula- Fuel, High-Level Waste and Transurante Radioactive

' tive Distribution Function (CDF) for the fastest of Wastes; Final Rule" Code of Federal Regulations the identified paths. Compliance with the 1000 year (1986).

GWTT objective would be demonstrated if it could be shown that along any of the paths, a tracer particle 3. R.B. Codell, "Oraft Generic Technical. Position on I i leaving the disturbed zone has a (100-X)% or greater Ground Water Travel Time", Otvision of Weste Manage- j i

probability of arriving at the accessible environment ment, Office of Nuclear Materials Safety and Safe- )

in a time greater than 1000 years, where X is a small guards, U.S. Nuclear Regulatory Commission (1986).

l number. A numerical value of X has not been specified by the staff at this time.

P 4 M. Gordon, N. Tantous. J. Bradbury, L. Kovach and 11. C.R. Faust, and J.W. Mercer, " Ground Water Model-R. Codell, " Draft Generic Technical Position: Inter- ing: Nunierical Models", Groundwater , Vol.18, p. 395 pretation and Identification of the Extent of the Dis- (1980) turbed Zone in the High level Waste Rule (10 CFR 60),"

01 vision of Waste Management. Office of Nuclear Mate- 12. J.W. Mercer, and C.R. f aust, " Ground Water Model-rials Safety and Safeguards, U.S. Nuclear Regulatory ing: An Overview", Groundwater Vol.18, p.108 (1980)

Commission (1986).

13. L.F. Smith, and F. Schwartz, " Mass Transport:
5. E. Hoek, and E.T. Brown, " Underground Excavation 1. A Stochastic Analysis of Macroscopic Otspersion",

in Rock", The Institution of Mining and Metallurgy. Water Resources Research, Vol. 16, no. 2, pp. 303-313 London, England (1980). (1980)

6. R.E. Goodman, Introduction to Rock Mechanics, 14 P.M. Clifton, " Ground Water Travel Time Uncer-John Wiley & Sons, hew York (1980). tainty Analysis - Sensitivity of Results to Model Ge-ometry, and Correlations and Cross Correlations among
7. 0.F. Coates, " Rock Mechanics Principles", Mines Input Parameters," Report no. 8WI-TI-256, Rockwell Branch Monograph 874, Department of Energy, Canada Hanford Operations, Hanford WA (1984)

(1970).

15. R.W. Nelson, " Evaluating the Environmental Conse-
8. P.C. Kelsall, J.B. Case, and C.R. Chabaness, "A quences of Ground Water Contamination Parts I-IV, "

Preliminary Evaluation of the Rock Mass Disturbance Water Resources Research, Vol.19, no.3 pp. 409-450 Resulting from Shaft, Tunnel, or Borehole Excavation." (1978)

Technical Report ONWI-411. Office of Nuclear Waste Isolation, Battelle Memortal Institute, Columbus, Ohio 16. L. Gelhar, and C.L. Amness, "Three-01mensional (1982). Analysis of Macrodispersion in Aquifers", Water ResourcesResearch,Vol.18,no.1,pp.161!!W~(1983)

9. J.J.K. Daemen, " Rock Mass Sealing - Experimental Assessment of Borehole Plug Performance", 17. ONWI, "A Proposed Approach to Uncertainty Analy-NUREG/CR-3473, U.S. Nuclear Regulatory Commission sis," ONWI-488,, Office of Nuclear Weste Isolation, (1983). Battelle Memorial Institute, Columbus, OH (1983)
10. J.W. Braithwaite, and F.B. Nimick, "Effect of Host-Rock Dissolution and Precipitation on Permeabili-ty in a Nuclear Waste Repository in Tuff "

SAND 84-0192, Sandia National Laboratories, Albuquerque NM (1984).

1 1

1 1 _

J

Retrieved from "https://kanterella.com/w/index.php?title=ML20206H700&oldid=3058033"