IAEA-TECDOC-319

EFFECTS OF HEAT FROM HIGH-LEVEL WASTE ON PERFORMANC DEEF EO P GEOLOGICAL REPOSITORY COMPONENTS

REPORT OF A TECHNICAL COMMITTEE MEETING ON EFFECTS OF HEAT FROM RADIOACTIVE WASTE IN DEEP GEOLOGICAL REPOSITORIES ORGANIZEE TH Y DB INTERNATIONAL ATOMIC ENERGY AGENCY AND HELD IN STOCKHOLM, 28 AUGUST - 2 SEPTEMBER 1983

A TECHNICAL DOCUMENT ISSUEE TH Y DB INTERNATIONAL ATOMIC ENERGY AGENCY. VIENNA, 1984 EFFECTS OF HEAT FROM HIGH-LEVEL WASTE ON PERFORMANCE OF DEEP GEOLOGICAL REPOSITORY COMPONENTS IAEA, VIENNA, 1984 IAEA-TECDOC-319

Printed by the IAEA in Austria November 1984 PLEASE BE AWARE THAT ALL OF THE MISSING PAGES IN THIS DOCUMENT WERE ORIGINALLY BLANK The IAEA doe t maintaisno n stock f reportso thin si s series. However, microfiche copies of these reports can be obtained from

IN IS Clearinghouse International Atomic Energy Agency Wagramerstrasse5 P.O.Bo0 x10 A-1400 Vienna, Austria

Orders shoul accompaniee db prepaymeny db f Austriao t n Schillings 80.00 in the form of a cheque or in the form of IAEA microfiche service coupons orderee whicb y hdma separatel ClearinghouseS I y N froI e mth . FOREWORD

This report discusse effecte th sdeee th f heapso n geologicato l repository systems and its different components. It reviews the experimental dat d theoreticaaan l effectmodele th f f heaso o s t botn o h the behaviou engineeref o r d naturadan l barriers. This document aimt a s supporting another IAEA publicatio relatea n i n d subject, viz. "Deep Underground Disposal of Radioactive Waste - Near Field Effects", Technical Reports Serie n preparation)s(i lattee Th . r deals witl hal those effects occurring in the near field which affect the mobilization of radionuclide d theisan r subsequent r fieldmigratiofa e Th e .th o t n document deals mainly with high-level wast d spenean t fuel considert bu , s other types of waste as well, whenever appropriate. The present report, however, is focussed specifically on effects due to thermal energy release solely from high-level waste or spent fuel.

This outcome repor IAEn th a s Af i te o Technica l Committee Meetinn go "Effects of Heat from Radioactive Waste in Deep Geological Repositories", held in Stockholm from 28 August to 2 September 1983. The initial draft repore o th fmads Mssrst wa y Bourk eb . J . .P e Werm. L (UK d e)an (Sweden). After the Technical Committee Meeting, Mr. Werme revised the draft of the report.

e responsiblTh e office e IAEr thith Afo s t ra repor s twa Tsyplenko. S . V . Mr v froWaste th m e Management SectioDivisioe th f no f no Nuclear Fuel Heinone. CycleU e wore . completes th J Th . kwa f . no Mr y b d same Division. CONTENTS

1. INTRODUCTION ...... 2. DECAY HEAT PRODUCTION ....»••*•«•»*«««•»••••*••»••••••.»••••••••••. • 2.1. Heat release characteristics of waste form ...... 11 2.2. Mean heaf so t transfer ...... •...... *...... 4 1 . 2.3. Thermophysical properties ...... »...... «.*...... 16 2.4. Repository design variables ...... J 1 . 2.5. Representative temperature field ...... 8 1 . 3. THERMAL EFFECTS ON THE ENGINEERED BARRIER COMPONENTS 3.1. Effect wastn so e form prio dissolutioo t r n ...... 9 1 . 3.1.1. Borosilicate glass ...... 9 1 . 3.1.1.1. Liquid-liquid phase separation ...... 19 3.1.1.2. Crystallization ...... 20 3.1.1.3. Crackin glasf go s ...... 0 2 . 3.1.2. Spent fuel ...... 1 2 . . 3.2. Dissolutio waste th ef no for m ...... 2 2 . 3.3. Corrosion of containers ...... 23 3.3.1. Consideration choicr containea sfo f eo r ...... 4 2 . 3.3.2. Thermodynamically stable containers ...... 25 3.3.3. Kinetically stable container ...... 27 3.3.4. Sacrificial containers ...... *...... 9 2 . 3.4. Backfill, buffer materials ••••••••*••••••••••••••••••••• 30 3.4.1. Swelling clays •••••••••••••••••••••••••••••••••• 30 3.4.2. Zeolites ...... 31 3.4.3. Crushed salt ...... 32 3.4.4. Granular material...... 2 3 . 3.4.5. Cementatious materials...... 32 4. THERMAL EFFECTS ON THE SURROUNDING ROCK 4.1. Thermomechanical effects ...... 33 4.2. Change waten si r flow pattern ...... 7 3 . 4.3. Thermally induced geochemical effects ...... 40 5. NUMERICAL EVALUATION OF THE EFFECTS OF HEAT 5.1. Introduction ...... 45 5.2. Numerical model describo t s processee th e s ...... 7 .4 5.2.1. Temperature distribution ..,..,..,.,...... ,7 .4 5.2.2. Thermal stress distribution ...... 48 5.2.3. Thermally induced groundwater convection •••••••5 5 • 5.2.4. Thermally accelerated chemical reactions •••••••6 5 • 5.2.4.1. Waste form dissolution .*...... 6 .5 5.2.4.2. Container corrosion ••••••••••••••••••• 57 5.2.4.3. Alterations in backfill ...•*...•...... 57 5.2.4.4. Water-rock system •••••••••»•••*••••*•8 5 • 5.3. Experiments to provide data for the models«•••..».....»•»•• 58 5.3.1. Parameters to be determined ••••••••••••••••••••• 53 5.3.2. Laboratory tests »...••»»«,»,»»•»,,»,,,•.....,.,9 5 » 5.3.3. Field tests ...... 9 .5 5.4. Verificatio validatiod nan n ...... 0 .6 5.4.1. Temperature distribution ••••••••••«....»...... • 61 5.4.2. Thermal stress distribution ...... 4 .6 5.4.3. Thermally induced groundwater convection ...... 6 .6 5.4.4. Thermally accelerated chemical reactions ...... 6 .6 . 6 DISCUSSIO RECOMMENDATIOD NAN N 6.1« Introduction «..««..«.....«..«...«.«.«»....««..«...... 67 6.2. Thermal limitations ...... 67 6.3. Mitigation strategies ...... 68 6.4. Recommendations for further research ...... 68 REFERENCES •••••••*«••••«••«*«••«••««•«•»•••«•••••»••••«••••••••• Y1 LIST OF PARTICIPANTS ...... 79 1. INTRODUCTION

High-level radioactive wastes produce nucleae th y b dr power industrie sy whic muswa disposee a htb n assurei f o d s that radionuclides will not reach the human environment in concentrations and quantities that could resuln unacceptabla n i t e detrimen publio t c healtt a h present or in the future.

Disposa high-levef lo l waste conditionen i s d carefulla for n i m y designed underground repository constructed a selectedee n i p d geologic environmen present a s mose i t th tt widely accepted approach.

e safTh e disposa radioactive th f lo e waste depende th n o s performance of the overall waste disposal system. This system may include both engineered and natural barriers. The isolation effectivenes f eaco s h componen disposae th f o t l system dependa n o s number of natural and emplacement-induced effects.

Near—field effects associated wite emplacementh h f highlo t y radioactive waste deea n psi geological repositor e discussear ye th n i d IAEA-TECDOC on "Underground Disposal in Deep Rock Formations, Near-Field Effects". A major consideration in both near and far fields is the heat generated by the waste due primarily to the presence of large amounts of concentrated beta and gamma emitting fission products. Many of the technical uncertainties in the design of a high-level waste disposal system are related to this thermal energy release, the resulting temperature field, and the impact this temperature field has on the waste disposal system. The heat generated by the nuclear wastes will eventually flow from the waste form, through any container and buffer roune surroundin waste th th do t e g rock d the ,an e transferre b n d through this rock and will be ultimately dissipated to the atmosphere.

e rat d duratioTh ean heae th t f releasno n affececa t some physical and chemical propertie barriere th a repositorf o sn i s y system. Transport by groundwater is by far the most likely mechanism by which any eventual release of nuclides to the environment may occur.

Thermally induced changes in the waste form, containers or buffer materials could affect their integrit r effectivenesso y r exampleFo . , waste glass may undergo devitrification which can eventually increase dissolution containee ratesth s A . r surface temperature increaseo s doesusceptibilits it s corrosiono t y . earlien Thi a lean o sca t d r contact with groundwater, if present, than if the container were not generating heat.

e stabilitTh openingf o yheatee th n dsi formatio importann a s ni t consideration. If the structure of the rock is affected, it will change groundwatea r movement transpore patterth d nan t pathwar yfo radionuclide environmente th o st .

systematiA c assessmen thermae th f to l loadinrepositore th f go s i y require assuro t d e thaenerge th t y generate emplacee th y b d d wastes will hav econtrollabla e effect. This includes studieeffecte th f sheaf o s o t on the structural stability of the host rock and radionuclide release and migration from the conditioned waste forms to the environment. That should lea establishino t d g maximum acceptable thermal load maximud san m temperature for underground repository components that will not violate safety, engineering, and operational requirements. Well defined thermal limits of the various disposal system components do not yet generally necessarilt no y existma d basee yb ,an d exclusivel absolutn yo e physical criteria attempn A . derivint a t g such criteri limitr ao howeves swa e rth subjec studa f n "Admissiblto yo e Thermal Loadinge "th carrien i t dou framewor Radioactive th f ko e Waste Management Programm Commissioe th f eo n of the European Communities [1].

This document deals wit effecte hth entire th heaf so n eto repository system reviewt I . experimentae sth l dat d theoreticaaan l effectmodele th f heaf so so behavioute th bot n o h engineeref ro f o d an d the natural barriers.

materiale Th s considere engineeres a d d barriers are:

borosilicate glass and spent fuel for the waste form; - steel, copper, stainless steel, high nickel alloys, titaniud man titanium alloy containersr sfo ; - clays, cement, zeolites crushed ,an d rock backfillr sfo buffersd san . widA e variet rocf o y k types available throughou worle n th t ca d constitute effective natural barriers to radionuclide migration. Granite, basalt, tuff, sald argillaceouan t s rock e consideresar thin i d s document.

The report also contributes to identifying some subjects for experimental work and for theoretical modelling and suggests some thermal design guidance for repositories. 2. DECAY HEAT PRODUCTION

e effecte temperaturth th l f so Al erepositore fielth n o d y system components can be attributed to the heat generation by high-level radioactive waste. The rate of heat release by the waste will be initially hig d wilhan l decrease decae th wit th f o ho yt time edu radionuclides. Hence, the increase in rock temperature will rise to early maxima and thereafter fall towards the local ambient value.

The variation of the transient temperature field will depend on: - heat release from the waste - the means of heat transfer - the relevant thermophysical properties of the waste form, its packaging buffee ,th r material surroundin waste th ge eth ford man rock - the size and shape of the waste package, quantity of waste per package, pitch, and layout of the array, and the total heat load.

2.1. Heat release characteristics of waste form

The rates at which heat is released from the waste form will in general depend on: type reactof th o e whicn - i rnucleae th h r fue s emplacedlwa ; - compositio solidifiee th f no d waste form; time th e betwee - n fuel discharge froreactoe th m d emplacemenan r n i t repositorya .

Heat generation results radionuclidee frodecae th th mf o y s incorporated in the waste form being solidified, high-level reprocessing waste or encapsulated spent fuel. A typical composition for the vitrified waste from light water reactors is given in Table I. Corresponding figures for spent fuel are given in Table II. The major heae th part f releaso t e from both waste forms results from beta/gamma emitters. The heat generation from those beta/gamma emitters, mainly frodaughtee th m r product decaf so f Sr-9o y d Cs-130an 7 (having approximately 30 years half-life), will decrease rapidly in the firs hundrew fe t d years after discharge froreactore th m . After this period mainly alpha emitter e responsiblsar hear fo et generation.

11 TABL . I E COMPOSITIO F SOLIDIFIEO N D HIGH-LEVEL REPROCESSING WAST WEIGHN I E CENR ] PE T[2 T

Glass Ceramic Glass SON GlasN sSO Glass Ceramic Glass Glass Glass Glass 78/7 Bl/3 2 U . 20 58. .30 64. 19. 20. F3 C31.3EC 209 189 VG 98/3 Pamela Fission Product oxides 14.55 22.69 13.85 15.20 9.75 9.58 15.54 28.10

Waste oxides a/ b/ Ç/ a/ d/ d/ Fe203 1.51 0.60 5.90 1~47 2.73 2.68 0.70 0.5 Cr2°3 0.43 0.20 0.50 0.42 0.56 0.55 0.24 0.20 NiO 0.24 0.10 0.20 0.23 0.36 0.36 0.21 0.20 MgO 6.34 6.23 0.40 A1203 0.10 5.11 5.03 2.40 U308 0.47 3.60 0.90 0.46 0.06 0.058 1.21

ALSO: a/ 2.2% Na20, c/ 5.9% Gd2O3, d/ 0.1% S04, 0.4% ZnO and 0.23% P205 , b/0.60% P205

Glass formers a/ £/ 50.88 41.51 41.84 Si02 28.00 43.60 • 44.20 34.75 P 0 48.50 B203 6.40 19.00 17.30 4.14 11.12 21.87 10.48 2 5 22.25 Fe 0 15.2 Na20 1.60 9.40 11.50 1.12 8.30 7.68 2 3 Li20 2.40 0.98 3.99 3.69 9 . A14 203 1.20 A1203 12.80 10.28 CaO 4.00 3.84 2.32 BaO 14.80 14.48 3.52 Ti02 4.56 2.80

ALSO: e/ Zr02 0.8%, ZnO 3.60%, As2O3 0.40%, MgO 1.2%, f/ Zr02 0.8%, ZnO 4.88%, As203 0.48%, MgO 1.44% TABLE II. EXAMPLES OF COMPOSITION OF LIGHT WATER REACTOR FUEL

(PERCENTAG WEIGHTY B E DISCHARGN O ) E FROE REACTOMTH j [3 R

BWR, 28,000 PWR, 38,000 MWd/tU MWd/tU

Fission products 2.9 3.9

Uraniu f whicm(o h

0.6%-0.8% fissionable) 96.2 95.1

Plutonium (approx% 75 . fissionable) 0.8 0.9

Other heavy nuclides 0.06 0.09 Due to the higher actinide content of spent fuel, the decrease of heat generation is slower than that of reprocessed high-level waste . (Fig1) .

Spent fuel from heavy water reactors Canadiae , th suc s ha n CANDU reactor, hav heaea t generation simila ligha o t rt wate R reactoPW r p ru to about 200 years, but a major difference is noticeable after 1000 years heae Th .t generation ratefunctioa s sa timf botr no fo eh spent fuels normalized to unity at 10 years after reactor discharge are shown togethe Fign i r . 2 .

The specific heat generation in a repository is dependent on the time between discharge of the fuel elements from the reactor and emplacement of waste in the repository. Ten years after discharge from the reactor, the heat generation has decreased by a factor of 10. Additiona year0 5 l s intermediate storage will lowe heae th rt output with a facto abouf o r (se3 t e Fig. 1) .

2.2. Means of heat transfer

Heat transfer through the waste form, its container and surrounding materials and the rock will be mainly by conduction with a theoretical possible contribution from convectio roce watey th kb n ni r movement either naturally occurring or thermally induced. Heat transfer by radiation and convection is negligible after backfilling. During the operational period ventilatio galleriee th f no y removsma ea portio f no the heat emitted by the waste.

Various assessments have been made of the convective contribution to heat transfe evaporitesn i r , clay d fracturesan d crystalline rocksr Fo . evaporites and clays, convective currents are negligible. Convection currents are possible in fractured hard rocks, if groundwater is present. However, this phenomenon will not greatly affect the temperature field.

Therefor temperature th e e fiel r thesfo d e hoste b rock n sca predicted without major error assuming the transfer to be by conduction» For long-term predictions, the heat transfer at the surface must also be taken into account.

14 Decay heat, W/ton uranium K>* PWR Burnujs i p 33.000 MWd(t)/tU

' \ "\ Reprocessing afte year$\• 2 r X " TV B<"; S j \ ( 10 - ! \ N Kl ! \ i

1 , ! | 1 1 1 1C» 10OO Time after discharge from reactor, Yuan Decay heat, W/ton uranium

High-level waste

ReprocewJng afte yearr2 s ^ years

j____I 1 10 1OO 1OOO Time after discharge from reactor, years

FIG. 1. Decay heat in spent fuel and high-level waste ] [4 R BW d an froR PW m

15 r»-133OflNG DW L 1 1.0

•CANDU

PWR

O.I c u> z -.CANDU u o

—

Ui PWR-

10 10* \03 TIME OUT OF THE REACTOR (YEARS)

FIG. 2. Relative heat generation races of PWR and CANDU spent fuels as a function of cime [5].

Fo a repositorr y locateunsaturatee th n i d d zone abov watee th e r table, vaporization may have a significant effect on heat transfer. For this case boiling is controlled by atmospheric conditions. Vaporization/ condensation behaviour of the water could account for a redistribution of the latent heat resultin a significan n i g t increas effective th n ei e thermal conductivity. Such behaviour has been identified for tuff.

2.3. Thermophysical properties

e temperatureTh e neath r n i fiels d strongly depend upoe th n properties of the material used in this near field: waste form, container and backfill materia d hoslan t rock. Thervarieta s i e materialn i y s used in the near field of the numerous repository designs proposed, thus resulting in a wide range of values for the physical properties.

16 e temperaturTh e distributio particulae th n i n r host rock depends only on the physical properties of the rock itself. Numerous studies of the thermophysical properties of various host rocks have been made and representative values of the physical properties are given in Table III.

TABLE III. TYPICAL INTERVALS OF THERMOPHYSICAL PROPERTIES OF ROCKS AT ROOM TEMPERATURE

Thermal Specific heat Host rock conductivity Density capacity (W/m.K) (ton/m3) (J/kg.K)

Clay 1.5 - 1.7 1.9 2.09- 4 91 1700- 0 Granite 2.0 - 2.5 2.6 2.95- 0 0 88 72 0- Salt 5. 5.20- 9 2.16 - 2.20 0 92 - 0 85 Tuff 1.4 - 1.8 2.20 910

2.4. Repository design variables

The near-field temperature distribution is primarily determined by the amount of heat produced in each package and the size and shape of the package. However when arrays of large numbers of packages are buried at e plac on temperature eth e fields produce e individuath y db l packages will coalesce after some characteristic time, thus producing a higher temperature at every position than would have occurred otherwise. If the distance between the packages is large enough, then the near-field temperature will have reache maximus it d m before this characteristic time. Hence the interaction of the individual thermal sources will not influenc maximue th e m near-field temperature r repositoriefo s s witt a h least this minimum distance between the packages. In practice this distance is dependent on the pre-cooling time of the waste, the fission product content, the size of the waste package and the rock type.

For repositories where the spacing between the individual packages s smallei r than this distance analysie maximu,e th th f so m local temperature has to take into account the interaction of the temperature fields produced by the individual packages. This means that for a repository wit t leasha t this minimum spacin a fixe d dan g desige th f no waste packages and backfill, the maximum near-field temperature can be controlle amounde th onl heaf y to b y t produce eacn di h package.

17 The temperatures in the repository between the individual packages and in the host rock in the emplacement area are determined by the total amount of heat produced in the repository.

Give amounn na wastf o t e ther e threear e way reduco st e eth temperature field.

- Reducing the heat production by postponing the disposal by means of a long intermediate storage. - Dispersing the waste more widely in a larger number of smaller size packages in each repository, thus reducing the near-field temperatures. - Distributing the waste in smaller quantities between a larger number of repositories, thus reducine temperaturesth l al g .

2.5. Representative temperature field

Heat emission frowaste th m e forms dependcharacteristice th n so s of the waste, and decreases with time as the fission products and other heat generating waste products decay. During the period of significant heat generation ,variabla e temperature distribution, bot n timi hd ean space, develop waste th en i s for surroundine th m n itseli d an fg materials. The actual heat distribution is very complex and depends on thermal properties of the waste form and surrounding materials and the desigrepositorye th f no .

In the general situation a rapid temperature rise in the disposal zone occurs, resulting in a maximum temperature in the repository region after 50-100 years. This heat pulse slowl ytemperature decayth d san e approaches that of undisturbed rock after 300 - 10,000 years. At the surface of the waste form, the heat pulse is always more pronounced; the maximum temperature is reached after some decades (10 - 50 years) and durin firse centuriew th g0 years fe t30 decae - )th relativels 0 i ( ys y rapid. After that period, the temperature of the waste form approaches that of the surrounding media.

18 3. THERMAL EFFECTS ON THE ENGINEERED BARRIER COMPONENTS

During the heat pulse, the elevated temperature will affect the engineered barriers, including the waste form. Some of the effects on the waste form may take place prior to loss of containment, and may have altered the properties of some of the materials in the waste isolation time systeth e t nuclida m e release occurs. Paralle r followino o t l g these phenomena, corrosio metallie th f no c barriers will proceed an d eventually dissolution* of the waste form will start. Other heat effects (reversible or irreversible) may influence container corrosion, the rate of waste form dissolution, and radionuclide transport.

3.1. Effects on waste form prior to dissolution

3.1.1. Borosilicate glass

Three heat effect y degradsma performance th e borosilicatf o e e glass used for immobilization of high-level wastes. Two of these effects involve phase transformations and can be classified as:

pha_se_ _id _ £epa£at_ioinj e growtTh _ f non-crystallinho e phases differin composition i g n frooriginae th m l glass.

^: The growth of crystalline phases in the glass. These phase havy same sma th ee compositio originae th s na l glass.

An overview of these phenomena can be found in textbooks of glass science or in review articles, e.g., [6] and [7].

The third heat induced effect involves £.rac:]cing_pf^ the_ e}-ass_ matrix.

3.1.1.1. Liquid-liquid phase separation

Liquid-liquid phase separation is a relatively common phenomeno glassesn ni borosilicate Th . e system exhibits phase separation. The constraints imposed by the vitrification technology and

* Throughout this document, the word 'dissolution' has been adopted to describ reactione th e s between waste ford groundwateran m , resultinn i g releas radionuclidesf eo .

19 waste compositio a finay lea o nma t d l glass tha n somi t e temperature interva susceptibls i l o phast e e separation, whic y changhma e th e physical and chemical properties of a glass.

Borosilicate glasses have relatively large area f immiscibilitso y for temperatures below about 750 C. However, at low temperatures, the increased viscosity of the glass acts as a kinetic barrier against phase separation. Prolonged storage at elevated temperatures will promote phase separation in the nuclear waste glass, should it be sensitive to such alterations. The maximum permissible temperature depends on the glass composition. Temperatures below 400-50 0shoulC sufficiene b d t to prevent phase separation.

3.1.1.2. Crystallization

Crystallizatio glasa f o ns e precedesysteb y a phas mma y b de separation. The crystallization of a glass is controlled mainly by the same factors that control phase separation. Consequently, the kinetics for crystallizatio extremels i n moderatr o yw lo slo et wa temperatures. Normally, temperature consideree ar s C belo 0 w40 d safe.

3.1.1.3. Crackin glasf o g s

Emplacemen f borosilicato t e glass-filled containera n i s repository will produc significana e t thermal gradient betwee glase th n s centerline and its periphery due to the cooler surrounding rock. The resulting thermal stresses may be of sufficient magnitude to cause the glass to crack, and substantially increase the surface area. The increased area would be expected to increase the dissolution rate of the matrix, oncmetallie th e c containe s beerha n breeched, assuming ? sufficient water is available to allow the dissolution to proceed.

A "threshold" maximum temperatur crackinr glase fo e th f so n i g 400-500 C has been identified [1J. Appropriate treatment during fabrication, suc s annealingha , would reduc ever o e n eliminate th e cracking phenomenon.

20 3.1.2. Spent fuel

Some temperature-related effect sconsideree b nee o e t d th n o d direct disposa unreprocessef lo d spent fuel e sucOn .h effecs i t potentially importan e casth e n i that t credenc presence s givei th e o t n e of intact Zircaloy fuel cladding as a barrier to radionuclides release from the irradiated fuel matrix. Most modern fuel assemblies used in light water reactors employ a pressurized inert gas - typically helium - insid e sealeth e d claddin o improvt g e heat transfer characteristicd an s retard fuel pellet oxidation. This gas is present at pressures up to approximately 5 MPa at ambient temperatures. It is therefore necessary to limit the allowable temperature of the fuel rods to prevent over pressurization and consequent stress rupture of the cladding. Theoretical and experimental studies indicate that maximum allowable temperatures for limited time periods (decades) in the range of 350-400 C should be observed to avoid this failure mechanism [8].

A second adverse effec potentiae n spen o tth „ fue U0 s ti l r fo l oxidation to U 0 with an attendent significant volumetric increase J0 o0 and tendency to powder when heated in the presence of an oxidizing atmosphere. The increased surface area available for dissolution effects will increase the potential release rates. Thus it is important to assure that low temperatures - less than about 200 C [9] should be maintaine regioe th f spenn no i d t fuel components afte metallie th r c containers have been breeched and the fuel pellets are accessible to wate n f oxidatiothera i r s i e n source, suc s oxygeha n evolvey b d radiolysi waterf so , availabl fuele th .o t e

3.2. Dissolution of the waste form

After failure through corrosion of the container, dissolution of the waste form will occur if moving groundwater contacts the waste. This dissolution may, depending on the point of time for container failure, take place or start during the heat pulse period.

Throughou pase th tt fifteen years, several type f high-leveo s l waste glasses have been developed e presenth t A t. stag borosilicate th e e glasses are technically most advanced. However, the interaction of these glasses wit aqueoun ha s fluid phas extremels i e y exaccomplee th td xan

21 nature and mechanism of these chemical reactions depend to a large degree on the manner in which glass dissolution experiments are carried out. For example so-callen ,i d 'dynamic' dissolution experiments (e.g. Soxhlet or flow-through systems) waste glasses have dissolution rates at room _C _O _O _T e d th [2] d m c ,an g 0 1 temperatur- 0 range 1 th f eo n i e temperature dependenc n leaceo h rates would appea folloo t r Arrheniun wa s type relationship wit apparenn ha t activation energ arounf o y d 15 kcal mol whic, e h would imply dissolutio abouC n0 ratet10 t sa 30-50 times higher than at room temperature.

In the context of modelling waste form dissolution in a repository environment, such dynamic test e inappropriatsar e becaus f limiteeo d groundwater flow rate. Solubility of waste species will, under repository conditions, be rate limiting and Arrhenius extrapolations of low temperature behaviour will be invalid because of competing dissolution and precipitation reactions. The solubilities of waste species wil affectee lb e compositio th y b de groundwater th f no e ,th presence of complexing anions, etc. For example, in clay environments, dissolutio 5 time1 o st greatep u n e b rate ry sthama n distilleni d water [10], whereas salt brines appea reduco t r e dissolution rates sa compare distilleo t d d water higt A .h temperatures, borosilicate glasses break down rapidly but surface layer alterations and crystalline precipitates [11] limit waste solubilities [12].

At present mant , no thery e fielar e d experiment glasn so s dissolution reported, but the general trend for experiments related to nuclear waste managemen thas i t t dissolution rate considerable sar y lower tha laboratorn i n y experiments [13].

Thus, the rate of radionuclide release will be a function of solubility and groundwater flow rate at any given temperature, and possibly als oa functio reactioe th f no nglase rateth sf o swit h groundwater.

For these reasons, it seems reasonable to limit the maximum temperature to the range below 150 C or perhaps, conservatively, to below 100 C at the time when water is expected to reach the waste form. These temperature achievee b n sca d eithe isolatiny b r waste th g e from water durin e thermath g y limitinb lr o pulse hea t th gi r part o ef o t generation in the waste by using a sufficiently low disposal density.

22 For other candidate waste forms, such as multiphase ceramics (e.g.

Synroc spenr )o t fuel (mainly single phase, microcrystalline U0?)e ,th situation is in many respects similar to that for high level waste glass. Although debated whethe leace th r h rate r Synrofo s t a c temperaturesame th e onle n ar s founrangi y i C abouo t s 0 dea p 10 tsu for the best borosilicate glasses [14], Synroc seems to be more resistant to hydrothermal alterations [15]. However, present thera s i e t relatively little information availabl dissolutioe th n eo ceramicf no n i s groundwater and salt environments; particularly on waste loaded ceramic materials.

Spent fuel may be regarded as a ceramic waste form. It is a single phas O eU ceramic s beeha nt altere,bu d considerably e whilth n i e reactor presentt A . r lesfa s , y b datther e a ear availabl n speno e t fuel tha vitreoun no s waste form d datsan a from different laboratoriee sar difficul compareo t . However e leacth , h rate e similaar s thoso t r e foun other fo d r waste form d temperatursan e sensitiv n similai e r ways [16].

3.3. Corrosion of containers

e purposeTh f containerso provido t e sar packaga e e aroune th d nuclear waste in order to facilitate handling and at least more important ensuro ,t e zero releas f radioactiviteo y during some required period. The primary cause for container failure during normal operational conditions is corrosion. The types of corrosion to be considered are: general corrosion, crevice corrosion, pitting corrosion, stress corrosion cracking wels i lt I know. n that elevated temperatures generally cause accelerated corrosion alss i ot I .possibl e that typef o s corrosion that are not anticipated at lower temperatures must be considered if the temperature is increased.

In discussing the corrosion of containers, the materials most frequently considered, can be subgrouped as:

. 1 Thermodynamically stable containers 2. Kinetically stable containers. . 3 Sacrificial containers.

23 A completely thermodynamlcally stable metallic container is unrealistic, since it would call for the use of gold or platinum. However, partial stabilit obtainee b n specifia ca yn i d c environment, e.g., copper in reducing (oxygen-free) groundwater. Ceramic containers have been proposed, but to date these concepts are less advanced than the concepts using metallic containers, and will not be discussed.

The kinetically stable containers refer to overpacks that, although not thermodynamically stable, are protected by their own corrosion products, which for a mstabl esurface layeth mord n o rean r les eo s inhibit further corrosion by, for instance, acting as a diffusion barrier for oxidants. Examples of this type of container are: stainless steel, titaniualloyss it d ,an m nickel base alloys etc.

e sacrificiaTh l container e overpacksar s thae neithear t r thermodynamically nor kinetically stable, but corrode at a known rate and by their thicknesses ensur n absolutea e isolatio wasta e r th fo ef no desired period of time. Such materials are mild steels, cast iron etc.

3.3.1. Considerations for choice of a container

In choosing container material, the desired service life of the container shal consideree b l d together witcharacteristice th h e th f o s nea rwaste fielth ef o dpackage e mosTh .t important parametee th s i r water chemistry e compositioTh . groundwatee th f e disposao n th t a r l area, its pH and E, , are of fundamental importance, as are the processe d mechanisman s s controlling them A temperatur. e gradient caused e thermabth y waste l th loa n ei dwil l ten o shif t de existin th t g equilibria such that temperature induced flows and possible evaporation effect y creatma s e environments aroun waste th d e package varying with time after disposal n thaI .t respect e containeth , e y b hav rma o t e designed to withstand corrosion in a range of temperature and groundwater conditions in a hard rock or clay repository. For salt repositories, where n ordinara flo n i w y t expectesensno s i e d under normal circumstances, these environmental variation e importantb y sma .

addition I o thermat n l effectsurroundine th n o s g chemical system, radiation effects will also have to be considered. Gamma-radiolysis will produce H , H CL and 0 . Since H„ is expected to escape from

24 the near field remainine ,th g effec radiolysif o t s n increaswila e lb n i e the oxidizing capacit groundwatere th f o y e radiolysiTh . t no s si temperature independent. Both the production of radiolysis products and subsequent recombination reactions are temperature dependent. For pure water radiolysie ,th less si s sever elevatet ea d temperatures [17]. However, for more complex groundwater systems, very little is known about the effects of temperatures on radiolysis. The stability of the radiolysis products must also be taken into account; at 300 C HO rapidly decomposes. Increased temperatures also strongly affece th t diffusion constants for most radiolysis products, which can then more easily escape from the waste package near field. However, if radiolysis s considerei problea e b o mt d this could r instance,fo overcome ,b y b e choosin sufficientlga y thick container.

3.3.2. Thermodynamically stable containers

The only container material that may be considered as thermodynamically stable as a metal in oxygen—free water is copper. It has mainly been considered for hard rock, where in particularly granitic rocks, the redox potential (E,) is typically 0.25 V and the pH is in the range of 8-10. The reactions controlling these properties are of cours reactione th e s between groundwate surroundine th d ran g rock. However thin ,i s contex s importani t i t noto t e thaconcepe th ta f o t 'system which 'E h assumes tharedol al t x reaction aqueoue th n si s phase are in equilibrium is a gross over simplification [18], since the oxidation stat f groundwateo e s perturbei r y biologicab d l activitd yan the kinetically slow oxidatio d reductionan lighf no t elements sucs ha C,H,0,N,S A simplifie. wate n diagraH i p C - r copper 5 , 2 fo mE d t ra is shown in Fig. 3. With increasing temperature, the CuS and Cu„S phases become less stable (CuS disappears at 150 C), while the window of stable Cu-metal is unchanged in size, although somewhat displaced within the pH-Eh diagram [19].

During the operational, or post sealing period, the corrosion requires transpor oxidantsf to , suc s oxygeha r sulphideno coppee th o rt s containers. Thucorrosioe sth n rat mors i e e controlle accese th y f sb o d corrodants and by thermodynamics, than by temperature dependent corrosion rate mechanismsr so , provided thawaste th t e packag stores i ea t a d

25 temperature where the container materials retain their partial thermodynamic stability [20].

addition I theso t n e effects sulphate-containinn ,i g waterse ,th kinetic sulphate-sulphide barrieth r fo r e reactio overcomee b n ca n . This reaction, whic root ha m temperature temperaturer o s s below 100°C only can be catalyzed by bacteria, is known to proceed at temperatures above 200 C. A consequence of this is a new supply of oxidants. Depending on the sulphate content of the groundwater, this source of oxidants may well be rate determining for the copper corrosion and finally set the service life for the container at temperatures above 100 C [20].

20. . i.o

, 0.5

. 0.0

Cu(0) metallic copper

us r-» f 2 — S

FIG. 3. Copper phases at 25 C in 10~ M bicarbonate-, 10 M sulphate- and 10 ' M silicate solutions [19].

26 3.3.3. Kinetically stable containers

The kinetically stable materials owe their corrosion resistance to forming a thin impermeable oxide layer on the surface (or more generally a fil corrosiof o m n products). This protective film will then prevenr o t considerably slow down further corrosion attack.

The major candidate kineticallr fo s y stable container materials are:

stainless steels, titanium, titanium alloys, and high nickel alloys.

These alloys retai greana t dea f theilo r resistanc generao t e l corrosion and have been used widely as industrial materials. Especially nickel and titanium base alloys (Hastelloy, Inconel, Ti-cod etc., e12 ) show good corrosion resistanc elevatet ea d temperatures.

Common type f stainlesso s steel e reportesar suffeo t d r from crevice corrosio d pittinnan temperaturet ga s eve n. HoweveC belo 0 w w10 lo r carbon stainless steels such as SUS 304 L, SUS 304 EL are expected to show good corrosion resistance.

The nickel base alloys are less corrosion resistant than Ti below 100 C and are also reported to be susceptible to stress corrosion cracking and crevice corrosion above 100°C in groundwater [21].

At present, the most promising materais seem to be titanium and titanium base alloys particula n ,allo- i Mo , y Ni calle ra d Ti-cod E diagra - 2 [22] H e1 titaniur p fo me .Th shows i m Fign i n. .4 As can be seen the metal is passive over the entire range where ordinary groundwaters are found. Corrosion rates for Ti in different types of water have been measured over extended periods of time (several years) both under oxic and anoxic conditions. At low Cl contents (4000 ppm) the corrosio 0.0- 1 rang e 0. ,um/ th f n eo n temperaturet a i ratea e sar s in the range of 100-130°C [24]. Even in brines (15-20% Cl~) the C [25Jmicronw 0 fe rang e yea25 r a . th spe t f ra e n o i rate e sar Against general corrosion a titanium container, 6 mm thick should last 3 5 10 - 10 years, provided the passive film does not break down.

27 pH ——

H diagrap _ r titaniu fo £ m FIG. 4 .m '.23•

However, titaniu d titaniuman m alloy e sensitivsar localizeo t e d corrosion, crevice corrosion seeming to be most severe. The most important agent r producinfo s g crevice corrosio titaniun ni l C e mar moderatt A . an F de temperaturel C moderatd w san lo o t e concentrations, titanium is normally considered as immune to crevice corrosion t cleapresentt A no . s ri whict ,i h temperatured san ~ concentrationCl e permissiblesar thert ,e reportbu e ar eth n i s literature that crevice t corrosioa C 0 y occu10 nma t ra Cl~ concentrations as low as 1% [26]; the sensitivity to crevice corrosion in F containing solutions are generally considered as far higher e immunit Th .i fro T mf o y crevic e corrosio illustrates ni n i d Fig. 5. This refers to commercially pure titanium; the Ti-code 12 alloy i smorr reportefa ee b resistan o t d crevico L t e corrosion n salI .t

28 formations, commercially pure titanium can therefore be considered for applications where the surface temperatures are less or about 100 C, while the Ti-code 12 alloy has a potential for applications at higher temperatures.

3.3.4. Sacrificial containers

Mild steels and cast iron can have relatively low corrosion rates in anoxic groundwaters. It is conceivable that as sacrificial materials, 10-2 thicm c 0 k would provid controllea e d barrier against radioactive release for some thousand years in hard rock at medium to low temperatures t thermodynamicall no s iro A .s ni y stabl purn i e e watere ,th direct reaction between iro waterd nan , resultin hydrogen gi n production, considerede b o t s ha . From general experience, this reactio knows i n o nt be kinetically inhibited at lower temperatures and the reaction rate approaches zero. At 100 C this reaction is still considered as slow, but the reaction rate increases strongly with temperature, particularly presence th n i chloridesf eo .

When using iro milr no d steel sacrificias sa l material oxygen i s n containing waters, the strong temperature dependence of the reaction rates must be considered. Assuming an Arrhenius type temperature dependenc expectee th d ean d activatio nreactione energth r fo y n ,a increas reaction ei e ordee nordeon th magnitudf rato f ro rf o e e would nounlikelye tb t 1 , when increasin temperature th gC 0 12 o t e C fro 5 2 m [27]

v —— : Presenk wx t t- v (PH7) ^ \ —— 'Feige c V c (immunity | Öl ICrevice corrosion! o c o (J

— -i 1 ^^~—- ——————— 1 ——————— ! _ oi 50 100 150 200 250 TemoeraturetX; )

FIG. 5. Immunity of cicanium from crevice corrosion [26].

29 3.4. Backfill, buffer materials

e buffeTh r materia uses i l filo space t d th l e betwee waste nth e hose packagth t d rocean k afte waste th s rbee eha n emplacede Th . backfill materia uses li refilo t d excavatee th l d portione th f o s repository (tunnel d shafts)san .

The main functions of these materials are: - to restrict the access of water to the waste packages. to locally modify the chemistry of the groundwater in order to reduce the potential for corrosion of the container and for the transfe radionuclidef o r s throug buffere th h . cushioo t waste - nth e packag d limi package strese an th th t n so e casn i rocf eo k movemen around an emplacemene n th di t t hole. retaro t movemene - th d f radionuclideo t s released frowaste th m e package. - to transfer heat from the packages into the host rock.

A number of backfill buffer materials are currently discussed. For hard rocks, swelling clays or clay minerals seem at present to be most favoured but other materials such as cementatious materials, zeolites etc. are also considered, as well as mixtures of these materials with sand and/or crushed rock. Whehose th nt roc clas ki r sal buffee o y th t r material usually produced during the construction of the repository is the excavated rock.

Becausproximite th waste f th eo ef o ypackages , buffer materiale sar the most influence y thermab d l effects. Some temperature limits muse b t observe orden i d conservo t r e desirablth e e propertie f theso s e materials, particularl w permeabilitylo e th y .

3.4.1. Swelling clays

Swelling clays, mainly bentonite, have been proposed for backfilling waste repositories. These materials consist of sheets of alumino-silicates. By binding water onto the sheets as ordered molecules, the spacing can be varied, thus giving the mineral a swelling capacity compactea f I . d bentonit access eha confineds i wateo st d ran , it excerts a considerable pressure on the walls of the confinement as it

30 saturates with water. The advantage of this is that a high degree of uniform density is preserved as well as a self-sealing capacity, should the backfill suffer local mechanical damage. This also guarantees a good contact betwee backfile th n d hoslan tunifor a rocd w permeabilitkan lo m y in the backfill. The structure of the mineral also gives it an ion-exchang adsorption io n wels a ea s la n ability.

The desired physical and chemical properties of bentonite or other swelling clays can only be retained if the structure of the material is maintained. At elevated temperatures, bentonite loses pore water at 100-150°C. If during the early stages of the heat pulse, water has not saturate bentonitee th d , temperature range th 100-30f o en i sn ca 0C caus losea swellinn i s g capacity. Shorter exposures (several hourst )a temperatures up to 600 C have indicated that the structural changes of e smectiteth e reversiblb y sma o somt e e extente responsth t o ,bu t e long-term heat treatment of dry bentonites is largely unknown [31]. In the presenc waterf o e , hydrothermal reaction y changsma structure th e f o e smectites. A leaching of silicates and other components like alkali metals (potassium) from the minerals are to be expected. If potassium is present, and the concentrations of Ca, Mg and Na are low, a change from smectite o illitet s y occursma . This change result lossen i s n i s swelling capacity and also to reduction in volume of the clay, leading to increased permeabilit backfile th f o y l material. Provide conditione th d s for these alterations are favourable concerning supply of potassium, temperatures as low as 100 C may be sufficient for the reaction to proceed at reasonable rates. In the absence of potassium, temperatures causn ca eC abov cementatio0 e10 10e 0th Cf no ca n cause cementation of the smectites due to the replacement of Si by Al in the structure,

3.4.2. Zeolites

Zeolites have been proposed either as a backfill material or as n additiva otheo t e r backfill materials mainly becaus f theio e r excellent ion-exchange capacity. However, zeolitt;s are known to lose water at temperatures exceeding 300 C and losing their ion-exchange capacity f28J. Furthermore, in dry the jcate, zeolites have low thermal. conductivities. Concerning long-term stability, geological evidence indicates that zeolites are unstable at higher temperatures. At increasing pressure they are easily converted to feldspars. However, cementing propertie f zeoliteso e providesar Referencn i d e f30J. 3.4.3. Crushed salt f crusheo e us d e salTh t wil restrictee lb backfilline th o t d f o g repositories in salt formations. In this case, the main function of this material ensur o wilt e lb e proper heat dissipation frowasteo e t th m d ,an provide mechanical support. Crushed salt will not be unfavourably affecte y temperatureb d .

3.4.4. Granular material

Granular materia uses li modifo t d y groundwater chemistro t y saturate it, for instance for glass components the Si and Fe containing material can be used.

3.4.5. Cementatious materials

A variety of these materials have been considered to be used in different ways in constructing a repository system. One of the specific aspect consideree b o t s usinn i d g these material heae th t s releasi s e frohydratioe th m f concreteno majoe Th . r effec elevatef o t d temperatures (above 100 C) on hardened concrete would be attempts of free wate evaporateo t r risA . vapoun ei r pressure could conceivably disrupt the concrete and increase its permeability. Concerning the hydrostatic pressure in a repository, this seems very unlikely to occur at temperatures moderately exceedin . C Difference 0 10 gthermae th n i sl expansion coefficient wastr sfo e container, concret hard ean d rocs ka wel differences la n thermai s l gradients durin thermae th g l pulsy ema also cause cracks to open in the concrete with locally increased permeabilit consequencea s a y avoio T . d dehydratio d thermanan l stresses, whic y lea microcrackino hma t d g followe reducey b d d structural properties, temperatures below 150 C are normally recommended [28]. However, these problems may be solved by using selected ballast materials and special cements. Considerin long-tere th g m thermodynamic stability of cements, a compilation of available thermodynamical data can be found in [29].

32 4. THERMAL EFFECTS ON THE SURROUNDING ROCK

Due to the very interactive nature of the phenomena involved with heat propagation in rock masses, their only meaningful description would ba e'system-approach ' one. This implies consideratioe th f no interactions of all components. However, and only for the sake of simplification, a sequential presentation will be performed below; the effects will be somewhat arbitrarily classified under three main headings;

thermomechanical effects changes in water flow patterns thermally induced geochemical effects

4.1. Thermomechanical effects

When thermae heateth y b dwastle th flu ef xo container s most rock types tend to expand. The state of confinement of these rocks at depth prevents their free expansio d generatenan s thermal stressese ,th development of which can be studied by principles of rock mechanics. It cleas i r thathermae th t l stresseresponse roce th theso th d t k f san eo e thermal stresses will depen presence dth upo) absencr (a neo f eo discontinuities stiffnese respective th th rocke ) ) th (b ,(c ,f so e properties of solid rock and interstitial water, and (d) disturbances buildine cause repositorye th th y b df o g additionn I . effecte ,th s will be different accordin distance th o t g e froheae th mt sources d wil,an f lo cours time-dependente eb .

These thermal stresses should be considered together with the 'primary1 state of stress in the rock mass, due to the weight of the overburde possiblo t d nan e tectonic e localoadinth ld gan relie f o f stress conditions due to the mine excavations. The resulting stresses must be kept at admissible levels so as not to impair confinement properties of the host rock at the time of the 'heat pulse', otherwise, any damage might prove to be irreversible when the rock cools down again.

Salt

Although some heterogeneous layers can be found in salt deposits (e.g. anhydrite, carnallite) salt can be considered as a continuous medium. Salt expands under the influence of a temperature rise, but the

33 build-up of stresses of confined salt is somewhat relieved by its ability to deform plastically, even under constant load: this phenomeno knows i n n as 'creep'. Tests show that this ability to 'flow increases with 1 increasing temperature.

vicinite heae th th n t I f o ysources creee , th f salo p t will result in rapid closure of the inlined disposal holes and thus a restoration of the confinement pressure on the waste packages. Therefore, the disposal holes should be cased if the waste containers are to be retrievable. During the period before backfilling of galleries takes place, the convergenc acceleratee b walle th y rat hease ma f o eth t y db propagatin g from the waste, without this resulting in failure at the walls.

After closure of the repository, the salt formation and overlying strata will temporarily show some uplift followe somy b d e later subsidence.

Excessive differential displacement more th e n compaci s t 'cap-rock1 (e.g. anhydrite y resul)ma somn i t e fracturin thif o g s cover layer. Continuous uplift due to diapirism is a reason for assuming the cap-rock to be fractured already. In the bulk of the salt itself, any generation w fracturesone f re-openinr ,o existinf o g g consideree onesb n ,ca s a d unlikely creepo t e ,;du moreover, eve f fractureni s generated e werb o t e , geological evidence indicates that they would 'self-heal o longen d ran ' behave as discontinuities.

Most of these effects can be predicted by numerical simulations with a reasonable degree of confidence. Large-scale, in-situ tests are carried out in several countries; for instance, the extensive research programmes in USA [32] and in the Federal Republic of Germany [33], which exampler ,fo , includes several heating experiments associated with measurements of borehole and gallery closure rate, changes in rock

salt pressure etc. These tests allow validatio f predictivno e computer models (see Section 5.3 below).

Granite

vicinite Heatinwaste th roce th n th i k ef f o yo g cause expano t t i sd and push out the cooler surrounding rock. Due to the rock stiffness,

34 this will induce tension coolee th n rsi rocadditionad kan l compressive stresse hottee th n rsi rock.

depthe Ath t s selecte r repositoriefo d graniten i s , there shoult no d be any serious risk of rock failure at the wall of open galleries or disposal holes. After closurrepositorye th f eo magnitude ,th e th f eo stresses developed will practicen ,i , depend upo extene nth whico t e th h rock behave eithes sa continuura m with homogeneous mechanical properties fracturea r ,o d lattemediume th n rI .case , expansioe th f no rock near the heat-sources is taken up by closing of fractures locally; coolee ith n r rock d particularl,an y between ground e surfacth d ean repository area, fractures will ten opeo t d n under tension.

These effects could therefore be beneficial in that compression will tend to close cracks in the hotter region, thereby reducing the flow of repositore d frowaste wateth an th mn o ei t r y area. They may, however, also be harmful in that the tensions will open the existing fractures between ground surface and the repository area, thus increasing the permeability in the far field of the rock cover. If these tensile stresses are too high and not counteracted enough by the existing state of stress, they could even induce creatio w fracturene f no s which might remai s permanenna t defects afte heae th rt puls overs i e .

Althoug difficuls i t hi predico t t these effects globally, some general principle identifiede b n sca appreciablf I . e expansios ni confined to only the volumes of rock containing the waste, i.e. if their dimensions are small compared with the depth, the effects can be considered likely to be small. In any case, the effects can be controlled by keeping the temperature rise at moderate levels, i.e. by suitable waste characteristics and/or repository lay-out.

Most of these effects are amenable to computer modelling. In-situ heating tests at the Stripa mine [34] and in the Climax granite [35] allowed the quantification of these near-field effects. Damages to the rock close to the heat sources were observed only at temperatures much higher than those currently envisaged for disposal in granite formations.

35 widA e rang f thermomechanicaeo l propertie coveree sar d undee th r generic term 'clay'. They can range from plastic materials with high water content to more rigid rocks with appreciable strength and lower water content. The mechanical response of clay layers subjected to a temperature rise under repository condition clearlt no s i sy knowns ,a mos presenf o t t in-situ heating test e beinsar g performed clos grouno et d surface [36].

In the immediate vicinity of the heat sources, drying out of the clay could occur if there were some way for water to migrate and/or to vaporize; this drying-out would induce shrinkage of the clay, therefore resulting in some degree of fracturing. It is, however, highly unlikely that dryin woult ou g d occudistanca t a r e froheae th mt sources once eth gallerie e backfilledsar potentiae ,th r 'large-scalelfo ' fracturinf go deep clays under heating seeming to be very low in the repository area.

The behaviour of heated clay will depend upon the respective proportion porf o s e wate d soliran d matrix plastir Fo . c clays with high water-content, heating will induce expansio pore th e f waterno e th s ;a solid matrix will also expand, but to a lower extent, there will be an increase in pore pressure, inducing some water flow away from the heated zone. The magnitude of this excess pore pressure would of course depend upo e permeabilitth n clae th y f o layery t presen A .s difficul i t i t o t assess the consequences of this phenomenon. For more competent clays however, the heating could produce effects similar to those described above for granite, and below for tuff and basalt.

In-situ testing is planned, e.g. in the underground laboratory at plastie th Mon i lc boom clawild an yl allow quantificatioe th f no behaviour patterns presented here.

Tuff

Available information mainly concern Yucce th s a Mountain site, USA [37]. Investigations show that this tuff (consolidated, welded volcanic ash) may be considered as a porous, hard rock, where some natural fracturing is present; it must be pointed out that other sites may reveal different properties. For the Yucca Mountain site,

36 calculations of thermally induced stresses were performed using data from laboratory testing. They show thaadditionae th t l stresses wile lb absorbe limitey b d d movements alon existine th g g fractures; neithee th r disposal holes or galleries, nor the overlying strata above the repository would be affected by new fracturing.

An 'in-situ water migratio heaten- r test performes 'wa thin o d s site, which concluded thaactuae th t l effect heatinf so g were even less pronounced than those predicte computey b d r models [38].

Basalt

Fro poine s thermomechanica th mvieit f o t f wo l properties, basalt can be considered as a hard, moderately jointed rock; therefore, most of the conclusions presented for granite also apply for basalt.

Specific investigations have been performed on the basalt flow of the Columbia Plateau in the Hanford Reservation [39] including a near-surface heating test. In this specific case a very pronounced horizontal stress fiel presens depte i d th ht ta envisage e th r fo d repository. As a consequence a 'flat' elliptic shape was given to the galleries and the disposal holes are horizontal, so that the total (thermal + excavation induced) stresses do not endanger the rock in the repository area. As regards the larger-scale effect, some stress-relief should take place along the existing joints which give the basalt its typical 'columnar' structure.

4.2. Change waten si r flow pattern

Heating of the host rock may cause changes in the water flow pattern changiny sb propertiee th g solie th d liqui dan f so d phase [40,41]

For the liquid phase:

- changes in physical properties depending on temperature (density, viscosity) - change physican i s l properties dependin waten o g r compositions (density, viscosity changes caused by dissolution reactions see Section 3.2).

37 solie th dr phaseFo :

- changes in permeability and porosity of rock due to thermomechanical effects (see section 3.2) - changes in permeability and porosity of rock due to dissolution/precipitation reactions in the groundwater bedrock system (see Section 3.2)

If the repository is saturated, changes in density of water modify or induce groundwater flow. Consequence large b y es ma convective cells depending on temperature and average permeability and porosity of the rock (i.e. the existence of interconnected fractures).

The water travel time to the surface will be modified and this effect persists unti thermae l th mos f o tl waste effecth s eha f to vanished.

The thermomechanical response of the overburden rock could result in the openin f existingo g fractures neasurface th r mainle edu o t y nonlinear uplift e resultanTh . t increas nean ei r surface fracture permeability may also affect the waterflow patterns.

Vaporizatio n importana e b y nma t mean f removinso g heawated an t r from the rock surrounding the underground openings during the operational period or for repositories located above the water table.

Salt

Salt generally contains very little water, less than 0.3 volume % in diapiric sald slightltan y highe bedden i r d salt. Inclusion brinf so n ei rock salt ten migrato t d e toward heae th st sourcee thermath f i sl gradient is sufficiently high. However, the inflow of brine will be very limited (in the order of a few litres per canister) and will cease after a few decades [42].

inclusione Ith f s also contain gas, tha liquid-gas i t s inclusions, the directio migratiof no s inverteinclusione ni th d an d s move away from the heat source, again as long as the thermal gradient is sufficiently high.

38 Granite

Granite harde sar , massive rocks with closed joint d fracturesan t sa depths matrie Th . x permeabilit granitef o y usualls si y very lowd ,an relative to fracture permeabilities, can be taken as zero. If the fractures in the rock surrounding the repository contain water, heating may induce flow due to convection for the reasons discussed above. Such convection cells will modify the travel times to the surface. However, this phenomeno onln nca y happen when fracture e interconnectedsar . Because the convective flow is very small, it has virtually no influence on the temperature field or on mechanical effects.

Clay

In clay, permeabilities are so low that convection effects will not majoe bth e r consideration.

If temperatures in the host rock are limited to 100 C, the only effects expected are related to the thermal expansion of interstitial fluids and solid particles forming the matrix of the rock. This is expected to result in a pressure gradient and some flow of pore water away from the heat sources.

temperature Ith f e were allowe exceeo significantlyC t d 0 10 d , then vapour might be formed and pressures might reach much higher values n thiI .s caspossibilite th e crackinf magnitude o y th d gan f eo the phenomenon would e investigatedb nee o t d addition I . t na temperatures in excess of 100 C the possibililty of hydrothermal reactions with associated mineralogical and other physico-chemical changes would need to be investigated.

Tuff

e concep placo On t s repositorea i t unsaturatee th tufn i yn i f d zone (above the water table). In this case boiling behaviour is controlled by atmospheric conditions. Heating will cause substantial increases in water vapour pressure wit potentiae hth dryinr lfo largt gou e portionf so the rock mass surrounding the waste. Vapour movement also may account fo largra e transfe removao latene t heaf th e ro f tdu tlo heaf to

39 vaporizatio subsequene th d nan t redepositio condensatioy nb a t na different location.

The direction of water movement in tuff is determined by the boundary conditions. If the waste emplacement boreholes are 'open1 the heated water vapour moves into the cavities where it is carried out of the syste ventilatioy mb condenset i systee th r no f md o flow san t sou alon fracturee th g s [38].

If the waste emplacement boreholes are closed (for example, with a swelling bentonite backfill) pressure ,th e gradient force watee sth r vapour radially outward from the heat source where it condenses in the rock at temperatures below 100 C. This results in increasing the water content of the rock (in some cases approaching saturation) and increasing the effective thermal conductivity.

Of equal importance is the rate at which the dry regions of rock re-acquire water as the thermal pulse decays away. It is only then that corrosion of the waste container and leaching of the waste form can begin. The driving forces for re-wetting the rock are primarily the infiltration rate and the matrix capillary forces.

Basalt

hydraulie Th c conductivit basaln i y t depends largeldegree th n eyo of fracturing present changee Th . s cause heatiny db muce gar h similao rt those already described for granite. Following emplacement of high-level waste, the temperature of the rock will rise and the water contained near the repository will be driven away. After certain time resaturation will take place, thereafter the waterflow depends on the density and orientatio fracturesf no , includin originae th g l fracture wels sa s la those induce y thermo-mechanicab d l respons rocke th .f o e Flow direction will depend on the local hydraulic gradient [43].

4.3. Thermally induced geochemical effects

The heating of a repository host rock from the ambient geothermal gradient to temperatures of 100 C or more through the emplacement of high-level radioactive waste may locally have profound effects upon the

40 chemical compositio groundwatee th mineralogicae f o nth d an r l constitutio hose th t f rocko n , whic turn i h n will affect backfill buffer stability, container corrosion, the nature of waste form dissolution and radionuclide ratardation in the near field. The nature of these chemical changes will depend upon the host rock, its mineralogy, the chemical compositio e groundwateth f no r fluio r de temperature th phas d ean t ,bu adequatele b y ma y simulate y closeb d d system hydrothermal laboratory experiments [11], or by geochemical mass transfer computer programs [44J.

The thermal pulse will also perturb any microbial populations contained within the original undisturbed system (rock groundwater) and those introduced during construction of the repository. Biological activity within the repository environment may have considerable effects upon groundwater chemistry, but their interactions with physical and chemical system complee sar x [45]. Site specific dat microbian o a l population e necessarsar evaluato t y e effectth e f higo s h temperatures upon their behaviour.

No treatment of the temperature dependence of sorption processes in the rock will be given here since radionuclide retardation by these processes is seen as being predominantly relevant to the far-field (i.e. thermally unperturbed) portion of the repository system [45]. However, where thermally induced change mineraf so l assemblage affecy sma t radionuclide retardation in the near field, then these will be mentioned below.

Salt

Salt deposits predominantly consist of halite (NaCl) with smaller, additional proportion mineralf so s suc anhydrits ha e (CaSO,),

polyhalit MgCK ( e a (SO,),.2 , carnallitH0) e (MgKCl3.6H20), kieserite (MgSO,.H„0), and various carbonates, clay and bitumen. The important geochemical effects resulting from thermal load revolve around the thermal stabilities of these constituent minerals and also the tendency of fluid inclusions contained within the deposits to migrate towards a heat source. The minerals above which are particularly sensitiv thermao t e l change e polyhalitar s d kieseritean e which liberates water at very high temperatures (greater than 300 C) and rarnallile which begin dehydrato t s about wheC ea 0 n8 t exposeatmosphère th o t d e

41 or at about 170 C for the in-situ confined conditions [47]. Obviously, maximue ith f m temperatur repositora n i e n suci y h rock types exceed these values muse ,on t take into accoun effece th t dehydratiof to n proportion sucf so h minerals contained within them.

Granite

Granite consist quartzf so , feldspar micd san a with small proportion accessorf so y minerals suc apatits ha d iron-titaniuean m oxideswhicf o e stabll har , al e phases unde y heatino rdr t p u g temperatures of at least 300 C. However, at repository depths an aqueous fluid phase (groundwater) will be present 'in fractures and intergranular pores chemicae Th . l compositio groundwatee th f no r will change when going from ambient underground temperatures up to 100 C or more through thermally driven, irreversible mineral-fluid chemical reactions. The exact nature of these chemical reactions will depend upon the initial groundwater composition (e.g. some granitic groundwaters may be unusually saline, such as those investigated from the Canadian Shield general n i [48 t ]bu silice ,th a contengroundwatee th f to r will increase (quartz solubility is strongly temperature dependent [49]) and the pH and ionic content of the groundwater will be altered through dissolution of the original silicate minerals and precipitation of secondary silicates. Because of the pronounced thermal gradients away from waste containers during the first 100 years after disposal, silica may be transported away from the immediate container zone by convective groundwater movement to be precipitated in cooler zones. Experimental investigatio thesf no e processe reveales sha decreasa d permeabilitn ei y by 10-100 fol thosn i d e zones wher precipitatioe th e s takenha n place [50]. Increased alkali (potassium) content in the groundwater through dissolutio potassiuf no m silicates (e.g. feldspars, micay )ma also accelerate the conversion of smectite in the backfill to illite, depending on the groundwater flow direction. The alteration of the granitic mineralogy along paths of water movement may ameliorate the sorption characteristics [51].

Clay

Clays (and argillaceous rocks in general) show great variability in chemical and mineralogical composition, and water content. Most clays

42 contain significant amount non-claf so y minerals suc quartzs ha , carbonates, sulphides and organic matter. The most important aspects of the heatin clayf o g s concer thermae nth l stabilit f theio y r constituent mineral potentiae th d san l dewaterin formatioe th f o g n and/oe rth generatio corrosiva f no e fluid phase l claAl .y minerals lose water from their structure upon heating, which may be complete at a temperature of 400°C. Heating will losa lea o plasticitf t so d potentiae th d an yf lo transformation of smectite to illite or chlorite. The pH of the fluid phase may decrease dramatically through mineral breakdown reactions and the increased fugacities of C0„ and S» through breakdown of carbonates, sulphide d organisan c matter e sorptivTh . e capacite th f o y rock will decreas f temperatureei hige sar h enough such that mineral transformations of the type clay —»mica take place. In general, temperatures shoulC belo 0 w15 d 15ensur0 Ce that detrimental geochemical changes will be minimal.

Tuff

Tufporoua s i f s volcanic rock consistin glassf o g y shardd san pumice. Many tuffs are devitrified such that the mineralogy consists of quartz, feldspar, zeolites and minor amounts of smectite clay. If the tuff is water-saturated, increased temperatures may lead to dissolution of primary minerals and a concomitant change in groundwater composition. e silicTh a contengroundwatee th f to r will increas valueo et s appropriate to quartz or amorphous silica saturation at the temperature attained and pH will be in the region of 8-9 [52]. Zeolites break down at temperature excesn i s 200°f so C [53], which would mean poorer sorptive capacit highet a y r temperatures e sealinTh . porf d go fractur an e e voidag silicy eb a precipitatio coolen i n r region y significantlsma y reduce permeabilit near-fiele th n i y d zone.

Basalt

Basalt is a volcanic rock resulting from the cooling of flows of molten materia d consistlan pyroxenef so , olivine, plagioclase, glass, accessory iron-titanium oxide secondard san y smectite clay. These primary phases woul stable b d temperatureo t e s wel excesn li f thosso e considered for waste disposal, so that the most important geochemical effects cause heatiny b d e thosgar e resulting froe interactioth m e th f no

43 mineral phase glasd san s with groundwater. Again exace ,th t naturf eo these reactions will depend upon the chemical and mineralogical compositio e basaltth f e compositiono ,th groundwatee th f no d an r temperature e interactioTh . f basalno d groundwatetan s beerha n studied experimentall temperatureo n [54]t yi resulte C .Th 0 30 thif so f so s study reveale formatioa d f secondarno y zeolite d increasesan e th n i s silic d potassiuaan me contenth groundwatere th f o f H o tp e Th . groundwater showed a slight decrease, from ambient to pH = 6-7. Comments made about permeability decreases through silica precipitation for granit d tufean f will also appl basalto t y e formatioTh . f zeoliteno s through reaction of groundwater with the basalt would increase the sorptive capacit neae th r f fieldo y .

44 . 5 NUMERICAL EVALUATIO EFFECTE TH F NO HEAF SO T

5.1. Introduction

e processeTh s whic releasa lean o hca t d radionuclidef eo s froa m repositor influence th d an yf temperatur eo n theso e e processee sar describe e thermaTh . Chaptern l4 i d effectd componente an th s3 n so f so the system are summarized in Table IV.

TABL . PROPERTIEIV E PROCESSED SAN S AFFECTE TEMPERATURY DB E

______Component______Thermal Effect______

waste form glas- s matrix crystallization phase separation cracking dissolution rate

- spent fuel cladding integrity oxidation

container stresses corrosion rate

backfill/buffer saturation with water swelling pressure permeability nuclide retention

host rock geochemistry stresses fracture opening fracture distributions water transport uplif f surfaco t e

For most components models are available to describe their behaviour. Mos f theso t e model n incorporatsca effece th e f heat o te collectio Th . n of relevant generaf n theso i dat s e bottleneci e aus th modell e th r s fo k in performance analyses.

45 However, accurate models (conceptual and mathematical) for describing the interactions of the near field components (waste form, containers, backfill and host rock) as well as the effects of temperature on these interactions are still being developed. Another field for further improvement e stud th f influenc o ys i s f roceo k stressen so fracture d therebpermeabilite san th n o y water fo y r [55]. Natural stresses stressee ,th s inducebuildine repositore th th y b df o g wels a y l as the stresses caused by heat release from the waste have to be considered.

As noted e introductionabovth n i e heae ,th t release froe th m high-level radioactiv factore eth wastf o s onls e i whic yon h might influence the performance of a waste repository. Different disposal concepts will rely to a different degree on the individual components of the multibarrier system.

During recent years the following effects of the released heat have been subjecte boto t d h measurement d theoreticasan l modelling:

Temperature distributions - Thermally induced stresses - Thermally induced groundwater convection - Thermally accelerated chemical reactions.

The thermal effects are important in several aspects. The thermally induced stresses, dependin whethen o g r compressio r tensionno causey ,ma , in the first case, borehole and tunnel decrepitation, when combined with the excavation-induced stresses. Induced tensile stresse n initiatsca e new fractures and reduce the normal stress component across fissures in the rock mass, thus causing increased permeability.

The heat released froradioactive th m e waste will cause somn ,i e cases, significant changes in the flow regime in the near field. The decreast no y travee ma th echanger o l y tim sma f radionuclide o e s froe th m repositor e biosphereth o t y .

s importanIi t develoo t p long-term predictive model wastr sfo e form leachin d corrosiongan whicn ,i h temperature dependence b n yca influencf wilincludedt o i e lb r fa eo ,s see limitee nth d time periof o d the heat pulse relative to possible isolation periods.

46 An extensive review of computer programs to be used for the analysis processee oth f s listed abov gives i e n [56]i n n thiI .s reviee th w status of the programs with respect to the important aspect of verification and validation is also given.

An important aspect is the compatibility of the data with the models whicr datfo e e neededth haar meane Th .datf so a acquisition shoule b d considered, as well as the treatment of the data, and eventually, the algorithm used to convert the primary data into data for the computational models.

5.2. Numerical model describo t s processee th e s

5.2.1. Temperature distribution

The calculation of the temperature distribution in the repository and the host rock can be performed accurately if:

e thermophysicaTh - l propertie wele sar l known, includine th g temperature dependency.

- The boundary conditions are well defined, in particular for long-term prediction.

The finite difference as well as the finite element method are used to solve the heat transfer equation for complex geometries with constant or temperature dependent properties.

For constant material propertie d idealizesan d geometries ha e sus been made of analytical solutions. In [57] such an approach is used for a detailed 3-D temperature calculation of a multi-layer high-level waste repositor sala n ti y dome.

Comparisons between calculations of rock temperature considering the temperature dependent properties and calculations based on constant properties have been carried out [58, 59J. Deviations in calculated temperatures remain withi w percenmaximue fe th na f o tm value, provided the temperature range is less than 200°C and the constant properties are chosen at ,an average temperature.

47 In réf. [60] a finite difference code has been used to calculate the temperature distribution as a function of depth at various times in a granitic repository assuming constant thermal propertie rocke th .f so e resultcalculatione Th th f so illustratee sar Fign i d. .Usinb a g similar code, the effects of the repository design have been investigated [61] and they are illustrated in Fig. 7 for a one-level repository and in same th e r amounfo Fig 8 wastf .o t e equally distribute two-levea n i d l repository with 100 m between the levels. The figures also illustrate the differences betwee modelso ntw e simplifie:on d mode periodr lfo s over seee b n nn i years0 D mode6 ca 3- o r times t A yeara . fo l0 6 p sd u san two-levee th Fig, 8 . l repository design will maxime o resulth tw n n ai i t temperature-time dependency. The relative height of these maxima were dependene founb e repositoro t dth n o t y design parameters.

Temperature distributions have been calculated in the near field of horizontally emplaced canisters surrounded by bentonite. Temperature dependent properties have bee bentonite nwaste th useth r ed fo d ean package [62]. Similar calculations have been performed for other rocks as well.

5.2.2. Thermal stress distribution

The rock respons thermao t e l loadin s beegha n investigaten i d several experiments [63-66]. The theoretical modelling has been either couple theso t d e experiment beer so n performed independentlye Th . approaches have been analytical solutions to an idealized repository [67], or the applications of finite element methods for numerical solutions [63, 68].

analyticae Ith n l solution n [67hease i ]th t producing waste (10,000 kW initial heat production and half-life = 30 years) is spread uniformly throughou repositorvolume e th th t f eo hara n di y rock, which is taken to be spherical. Heat is given out uniformly throughout the spherical region and a linear elastic theory is used to predict the stresses induceheate e rocth Th s treate.y ki b d continuoua s a d s medium, i.e. fissure d faultsan disregarde e sar d whic simplifyina s hi g assumption e stresTh . s distributio calculates ni domaia n i d n with cylindrical symmetry calculatioe resulA .th f o t shows ni Fign ni . .9

48 Temperature C * 0 6 0 4 20

100 years 1960 years 9960 vears

Repository

Geothermal gradient

2000 Depth below surface

FIG. 6. Temperature distributions in the granitic repository e wholth n ei formatio d an t differena n t times after disposal of the waste [60] •

49 5 10 20 W 100 200 «M 1000 10 COO ICO 000 TIME AFTER DEPOSITIO flNAiN I . STORACc YEARS TEMPERATURE ON CANISTER SURFACE ————— —— MAXIMUM TEMPERATURE IN ROCK •—-—•— — MEAN TEMPERATUR ROCN I E K THICK LINES 30-MODEL

FIG. 7. Variation of température at centrum [61].

c '.00

I nui 7i'C 30 , ^ •• — "' » — S. i fl ! l U t X Uu ^\^~~ £•--^ S J3 i« J'CAI5I-t*s^*'T^ -t' 20 •""' *""

————— -^ 3 1 5 10 3 U 100 200 UM 1000 10000 100000 TIME AFTER CEPCSITIO FINAN I N L STORAGE YEARS —•—— MAXIMUM TEMPERATURE IN socx _._.—.— MEAN TEMPERATUR ROON I E C 2 PUNES OIST. 100« THICK LINES 3Q-MOOEI

FIG. 8 Variatio. f temperaturo n t centrua e m [61].

The general conclusion from the calculations is that after about 0 years10 , nett tensile stresse roce occun th k sca n overlyini r a g repository. At the same time, compressive stresses occur at the centre repositorye oth f .

At this moment programs are under development in which a proper modelling of the fissures and faults is taken into account.

50 FIG. 9. Time dependence of principal hoop stresses âc various depch (z) above ehe repository (radius = 250 m, depth of the repository = 1000 ra, thermal conductivity 2.51 W ra-1 K-l) [67].

The model n [64si ] discus resulte th stwo-dimensionaa f so l finite-element model applied to a Canadian immobilized waste vault. The vault geometry is shown in Fig. 10. The power per waste package is assumed to be 269 W. The horizontal stress field after 30 years is shown in Fig .whil1 1 e Fig .show2 1 horizontae sth l stres depts sv h after seee b nn 0 years botca excavatio20 e s hth A . d thermanan l loading tend to reinforc e stateth horizonta f eo l compressio roce th k n masni s above the repository. A conclusion drawn in [68] is that the analyses have not detected evidence of thermal fracturing. Thus, it must be concluded that the possible development of nett tensile stresses is dependent on the repository design.

51 7.5 m.

E, —— o ~~~~N ••»•••• ^

£ 0

v-4< ••» E X, J M Y. !WI X! ^- E x: jRl X!

W ROO l PLA MF 0 N

FIG. 10. Geometry of immobilized waste room and container [68],

For salt repositories the stresses, mechanical as well as thermal, are difficult to model. This is due to the fact that rock salt cannot be treated as a linear elastic material. At ambient and at elevated temperatur deformatioe th e n behaviou strongls i r y time dependent. This behaviour can be described with creep equations as is done in [66]. The time and temperature dependent deformation problem can be solved using the finite element metho dons a dn [66] ei . finitA e element programme uses analyso wa t d behavioue th e f locallo r y heated salt e sals Th .twa modelle elastic-plastis da c material with secondary creep comparisoA . n with experimental results [62] showed thamethode th t s giv gooea d prediction of the thermomechanical effects in rock salt, at least with respect to near field and relatively short time spans.

52 FIG. 11. Horizontal stress (MPa) distribution at 30 years [68],

53 HORIZONTAL STRESSES (MPa) 0 8 0 6 0 4 20 100 120

\

0.2

u h. «0.4 OMBINED STRESS (THERMAL STRESS + IN-SITU AND Q EXCAVATION STRESSES) O . 6 s o tJ ClJ 03 £0.8 THERMAL u \ a STRESS

1.0 REPOSITORY HORIZON

1.2

. FIGHorizonta12 . l stres n immobilizea s dept v sr fo h d waste repositor 0 year20 t sa y after waste emplacemen a 2-dimensiona n o t l finite element analysis [68] •

o avoiT d high computation costs special attention mus paie tb o t d the coupling of the transient temperature distribution and the time dependent deformation analysis.

n [69I ] speciaa l purpose prograanalysie e th s useth r mwa f fo dso stresses in a salt dome where a repository with a total initial heat generatio f 27.7s locatedno i s concludei W 5M t I . d thae totath t l stresses in the salt dome remain compressive.

With respect to the modelling of the thermal stresses in clay, only preliminary work has been done and programs are still under development [70].

54 5.2.3. Thermally induced groundwater convection

Two problems are related to the modelling of the thermally induced groundwater flow.

- The validity of Darcy's law for very low permeable media such as clay. - The characterisation of the permeability and porosity of hard rocks like granite havin matriga xpermeabilitw witlo ha d yan fractures with a high permeability.

Models for the thermally induced water flows in a hard rock repository have been develope severay b d l institutions; examplee sar given in [40,41,70,71,72,73]. As for the thermomechanical model, both analytical solutions [41] and numerical computations [40,71] are used.

The analytical solutions in [41] treat an idealized repository [67]n i analysie e .Th on similae th so rt assume s that fractured rocn kca be treate equivalenn a s a d t porous medium permeabilite Th . roce th k f o y is assumed to be low; i.e. the dominant heat transfer mechanism will be heat conduction throug roc e convectiod th hkan neglectede b n nca n I . this model, the temperature field is deduced from the heat conduction equatio watee th rd flonan s thewi n determined usin temperaturee th g s sa driving forces e flo assumeTh s . wi describee b o t d Darcy'y b d w sla throug hsaturatea d homogeneous isotropic porous medium resule Th .f to such computations are of course dependent on permeabilities, regional flows etc., but a general conclusion is that the thermal convection of water near a repository could be as important as regional flows for several thousand years.

In [41 ]mathematicaa l mode describeds li , which consist couplef so d non-linear partial differential equation hea r groundwated sfo tan r flow. The system of equations is solved numerically using the finite element threr o o methoetw dimensionsn i d e fractureTh . d roc treates ki : das overlappino Tw . 1 g continua representine on ; networe th g fracturef ko s and one representing the solid blocks of rock.

55 2. A single equivalent medium.

Anisotrop permeabilite th n i y alss yha o been taken into accounn i t the model. The model is illustrated for cases where the initial pressure and geothermal gradients are parallel and for cases where the repository is situated unde crese hila th rundef r lo o t hillslopera examplel Al . s show thaheae th tt released frorepositore th m y havyma esignificana t floe impacth w n regimo t e aroun . Eveit d ver t na y moderate increasen si temperature (10 C), there is a significant effect on the flow patterns.

5.2.4. Thermally accelerated chemical reactions

Realistic modellin temperature th f go e dependenc glasf eo s leaching and container corrosion requires a qualitative understanding of the physical and chemical interactions in the repository near field, includin backfille th g . Eve detailef ni d modellin s probabli g y impossible, simplification y reducs ma tase manageablo th et k e proportions. However, although much dats beeaha n accumulate o fars d , the interpretations of the data are in doubt. In the following a very brief overview wil modelline e statu givee lth th b e f f th o so n f o g dissolutio containee th d nan r corrosion.

5-2.4.1. Waste form dissolution

The status of modelling of the interactions between simple glasse aqueoud san s foune reviea b medi n n i d aca w pape glasn ro s science (see e.g. réf. [74]). However nuclear ,fo r waste glasses, whic vere har y complex in composition, there may not be one model, but in the extreme a mode eacr lfo h nuclide, dependin e interactioth n gnuclideo e th f no s with the aqueous medium [75]. At present, the temperature dependence of leachin assumes i g folloo t d Arrheniun wa s type relation (see e.g. Fig. 13) [10, 76]. It may be concluded that at present, the major efforts in modelling concern the mechanisms themselves, and their temperature dependence are of secondary importance.

At present, only attempts at modelling the interactions between spent fuel and water have been made [77]. It is important that the temperature dependency of spent fuel dissolution is incorporated into such models.

56 200 no «0 90 n 40

/fft z 0 l 3 c 1 fM L 's r 9 u * i & . g A. 10 ' • à A.

10

0 3 8 2 1 06 2 24 2 4 ^3 2

10/T

. 13 Arrheniu . FIG se specifiplotth f o s c weight lossef o s 6 different glasses after 7 days in distilled water [10.

5.2.4.2. Container corrosion

The same statement Section i s sa n 5.2.4.1 generalle .ar y valid alsr containeofo r corrosion, althoug corrosioe hth f containerno s i s perhaps less complex than waste form dissolution Arrheniue Th . s type relatiokinetice reactione th th r f nfo s o e normall sar y assumee b o t d valid.

5.2.4.3. Alteration backfiln si l

nowo Ut p , changebackfile th n si l cause y increaseb d d temperature e describesar d qualitativel d sometimean y s quantitativeln i y conceptual models [78] computeo .N r models havbeet eye n developed. However, as mentioned before there is a need for modelling of the integrated near field. Important temperature dependent processee th n i s backfill are transport of water through backfill, diffusion, convection d sorptioan f radionuclideno backfille th n i s .

These effectcoursf o e esar relevan clay-typr fo t e materialsr fo ; other backfills, such as crushed salt in salt repositories, other effects coul considerede db .

57 5.2.4.4. Water-rock system

Increases in temperature due to the emplacement of radioactive waste will potentially affect both the chemistry of naturally occurring compounds and of compounds originating from the waste. It is quite probable that small quantities of radioactive material will leave the waste package during the first 300 years after closure when temperatures higheste th e ar . Nevertheless chemicae th , l behaviou radionuclidef o r s release modellee b n ca d d wit e sam chemistre hth e th e theors th i f s o ya y naturally occurring compounds, although this modelling requires additional data.

The chemical modellin higf o g h temperature natural aqueous systems is based on equilibrium speciation and reaction pathway tracking models used in groundwater chemical evolution studies. In these models the mass of each elemen distributes i t e systeth n i md amon aqueoue th g d solisan d phases available and potentially available. Kinetics have been treated to some degree by using an extension of reaction coefficients and an attemp s beeha t n mad incorporato t e e rate constants t temperatureA . s Van', d abouC betwee an 0 t5 C t Hoff5 2 n s rulbees eha n used successfully to account for temperature effects above the standard state. Above this temperature, heat capacity function e oftesar n necessar modeo t y l speciation d particularly,an modeo ,t l precipitation and dissolution. Model f thesso e processe highet sa r temperaturo t p (u e several hundred degrees centigrade) have been develope describo t d e th e geochemical evolution of hydrothermal systems [79]. The chemical processes occurring within the heated zone can be modelled as if the system were some hundred meters deeper alon geothermae th g l gradient.

5.3. Experiments to provide data for the models

5.3.1. Parameter determinee b o t s d Most of the local effects of heating can be simulated in laboratory tests e maiTh .n parameter determinee b e o t sth e ar d sensitivit o temperaturt y : eof

- dissolution of the waste form - degradation of containers mechanica- l properties (e.g. strength, expansivity, creepf )o

58 the rock - thermochemical properties of the chemical components

extensive th Som f eo e effect f thermaso l loadinamenable b y gma o et field investigations. These probably include the ability of pre-existing fractures in the rock, to relieve stress and interrupt the propagation of deformation in fractured rock and the thermoplastic and creep properties of some rocks. It will inevitably be impractical to simulate experimentally more far-ranging effects, suc extensivels ha y induced thermal flow.

5.3.2. Laboratory tests

The local effects listed above can be investigated in laboratory tests. Som f theseo y requirema e bote temperaturth h d statean f o e stress to be comparable with real values, in which case high-pressure vessel techniques will have to be used. In some cases it will additionall necessare b y applo t y y mechanical stressed san vapour-pressure conditions representativ in-sitf eo u values during experiments.

Inclusio watef no r wit chemistrha y identica in-site th o lt u chemistry, intexperimene oth t would increas applicabilite eth e th f o y results for realistic modelling. Also the backfill to be used in the repository should be considered. Such integrated experiments have been develope d wil an dusee l b severan i d l laboratorie derivo st e realistic parameter valuedevelopmene modelth r sfo r welfo s sa s morf la o t e realistic models [79,80,81]. Care shoul takee b d n thalaboratore th t y conditions will be representative to measure the effects under in-situ conditions.

5.3.3. Field tests

Major underground field programmes are in progress or planning: for hard rock in the USA [32], Sweden [34], Switzerland [82], Canada [83], India [84] and Japan [85] with minor supporting programmes in France [86] and the U.K. [87]. for clay in Belgium [88] and Italy [89]. for salt in Germany, Fed. Rep., [33] and USA [90].

59 In-situ measurements in these programmes are being made for studies on heat transfeeffecte th d temperaturf so ran n corrosioneo , creep, deformations, and other mechanical properties of the rock. For example, Germae th a par f no t programm Asse th e n sali e t mine includee th ) (a d development of a "Standard Probe", to measure the convergence of a heated simulated disposal hole, and (b) the measurement of salt pressure on a steel casing inserted in the hole. Good agreements were found between predictio actuad nan l field measurements [49] . summarA curren e th f yo t statu researcf o s d repositorhan y development including underground test facilities in OECD/NEA countries is presented in Table V [42]

5.4. Verificatio validatiod nan n

Verificatio d validationan e definenar d accordin e IAEth A o t g Radioactive Waste Management Glossary [91 followss ]a :

"A computer code is 'verified' when it is confirmed that the conceptual e reamodeth l f lo syste adequatels i m y representee th y b d mathematical solution. Verificatio n thu carriee nca b s d r outfo , example by intercomparison of codes and by comparison of numerical codes with analytical solutions.

"A conceptua lcomputee modeth d lan r code derivede ar fro t i m "validated 1 confirmes whei t ni d thaconceptuae th t e l th mode d lan derived computer code provid gooa e d representatio actuae th f lno processes occurring in the real system. Validation is thus carried oucomparisoy b t f calculationno s with field observationd san experimental measurements."

modele th r s Fo described abov e statuth e f theio s r verificatiod nan validatio discusseds i n .

Analyses using models being verifie validated an d d only produce correct resultboundare th f i s y condition e probleth e definen so ar m d correctly e temperatureth r Fo . , deformatio d floan n w problemse ,th initial conditions and the conditions at the physical boundaries must be handled with care, as they can lead to completely wrong results.

60 TABL . V E CURRENT STATU REPOSITORF O S Y DEVELOPMEN OECN I T D COUNTRIES [42]

TEST FACILITIES AT DEPTH

•5 ^

""" « LOCATION n —i u. COUNIR V î u ê j j 3 tjj TION S SURVE Y O F SITE S SURFAC E INVESTIGA - TES T BOREHOLE S GENERI C STUDIE S Crystalline Rocks USA X X Switzerland XXX Sweden X X X X

Granite Finland XXX' Canada X X X X X IVhiteshell FrancX e X X X Japan X X Spain X X SwedeX n X X X X Stripa SwitzerlanX X d X X X X Grimsel Pass . UK X X X X USA X X X X X X Climax

Other

Gabbro Sweden X X Canada X XX

Diabase Japan X X

Evapontes

Salt Diapirs Denmark X X X X F.R. X German X y X X X Gorleben (2) F.R. Germany X X X X X Asse II (2,3,4) Netherlands XXX USA X X X X X Avery Island

Bedded Salt Spain X X USA X X X X X Lyons USA X X WIPP (2)

Anhydrite Switzerland X X Felsenau USA X X

Other Sedimentary Rocks

Clay BelgiuX m X X X X X Hoi (2) Italy XXX X Switzerland X UK X X X X

Shale USA X X Japan X X Spain X XX

Other

Mixed Marine Sed. F.R. Germany X X Konrad (2,3) Sequence

Basalt USA X XX X NSTF Hanfor) (2 d

Tuff USA X X X X NTS (2) Japan X X

. 1 Tesb Facilitie deptt a s h withou ta shaf t being sunk implie e facilitth s s i y n iexistina n g mine. 2. This facility or another facility at this site has accepted or is planned to accept waste. 3. Low and intermediate waste only; no high level waste. Asse Th e min. 4 situatea i e sala n ti d anticline.

5.4.1. Temperature distribution

In [58] a verification is reported of the temperature codes by comparin resulte differente th gth f so codes with each othe d witan rn ha

61 analytical solution leading to the conclusion that the codes do a good job. Validatio meany b n f fielso d test clan reportese i yar n [89i dd ]an [92].

In-situ heating experiments were and are being performed in the Asse salt mine as part of the CEC R&D Programme on Radioactive Waste Management [42] so-callee .Th d 'Temperatur performes e wa Tes' a IV tt a d verticaa dept ; m abouf o h0 l75 t borehole deepm diameterm c 5 ,1 5 ,3 s ,wa drilled in the floor of a gallery; in this hole, heating sources with controllable power output were inserted in the bottom 6 m. Temperature of the heater was increased up to 270 C during 3000 hours, then kept at a constant value (80 C) during almost one and a half years. The temperatures aroun e heateth d r were recorde y thermocoupleb d s insertet a d various distances froheae th mt source. Good agreemen s obtainetwa d between the measurements and predictions using computer codes. Another experiment, called 'Temperature Test V' is also being performed in the same mine, not far from the Test IV, but in a zone where salt includes some heterogeneous layer polyhalitef so heatere Th . e insertesar a n i d horizontal borehol wale gallerya th f l o n heatine i e Th . g tes stils i t l in progress, but it seems clear that, once again, the computer codes used allowe gooa d d predictio actuae th f lno test data presently recorden o d the site [93].

A heating experiment was carried out at approximately 20 m below the mid level of a quarry, situated about 50 km south west of Mol, in an area where the Boom clay formation outcrops [92]. The experiment ran for one year at 1000 W output and for another 6 months at 1500 W. A thermal conduction code, taking also into account the water movements into the boreholes s use ,interpreo wa t d experimentae th t l resultst fi e Th . between numerical and experimental data was quite good, see Fig. 14.

An in-situ experimen s alstwa o carrieshallot a t ou dw deptn a n hi open clay quarry in the area of Monterotondo, near Rome [89]. The , W 0 50 d experimenan 0 powestageo e 25 tw th f s carrieo rt n sa i twa t ou d respectively. A thermal conduction code, elaborated by Belgian staff, was used for comparisons between measured and predicted temperature increases. The clay was considered as a continuous, homogeneous and isotropic medium, where heat dissipate conductiony sb . Clay thermal properties were assume s temperatura d e independent e temperaturTh . e

62 50 • —— measured values

= 1.15 m ——— calculated values

5 20 = 2.65

10 - 4.15 m

FIG. 14.

Temperature diagram (horizonta d planemi l t severaa ) l radial distance) (r s from the heater [92]•

increases measured in clay fit quite well the theoretical values obtained by the numerical code.

Durin perioe th g d 1980-1983 ,larga e scale in-situ tess fieldewa t d in quartz-monazite (granite) rock at the Climax stock in Nevada, USA [94]. This test employed both electrical heaters and actual spent fuel assemblie s thermasa l sources therma7 totaA .3 f lo l sources were emplaced in an array of three parallel drifts which was designed to simulate a portion of a repository. The test depth was 420 m below the abovm watee 0 surfacth e15 rd tableean . Extensive thermal calculations were made [95] using both finite differenc d finitean e element codes sa well as superposition solutions for analytic formulations. These calculations also took accounl heaal tf o ttransfe r mechanismd san allowe r heafo dt removae testh t y lb ventilatio n system.

A tota abouf lo 0 precisio 60 t n temperature recordings were made 3 throughout the test area, which included about 10,000 m of rock where measurable temperature effects were observed. Peak rock temperatures up

63 to 200°C were observed. Very good agreement was found between the measured and calculated temperature fields. The maximum variation was less tha oveC three 5 nth r e year duratio teste th .f no

The effects of a heater placed in a partially saturated, fractured, porous welded tuff are described in [38]. The location of the experiment was about 425 m below the surface in G-tunnel on the Nevada Test Site. The heater was operated at powers of l kW and 750 W. The rock surface was heated to approximately 240 C. A finite element thermal conduction code, which also accounte latene th r t fo dhea vaporizinf o t g wated ran radiative heat transport from the heater to the rock was used to compare wit experimentae hth l results generaln I . agreemene ,th t betweee nth calculate measured an d d value goos swa d [96].

Full-scale heater tests have been performed in the Pomona basalt at the Near-Surface Test Facility on the Hanford Reservation, Washington in the USA. Based on 270 days of heating, during which heater power levels were raised in steps to 3.0 kW, it has been concluded that the thermal propertie basale th e isotropi f tar o s horizontaa n i c l plane, with good agreement between predicted and measured temperatures. Although more difficult to evaluate, thermal response in the vertical direction is not appreciably different than in the horizontal direction. A thermal conductivit 9 watt 1. metrr f degrer spe o y pe e e centrigrade seem o best s t response accounth r fo te observed [43].

5.4.2. Thermal stress distribution

e computeTh r code analyso t s thermalle eth y induced stressee sar either general purpose commercial finite element program speciar so l purpose programs e verificatioTh . generae th f no l program beins i s g carried out. The verification of the special purpose programs is not always performed, and it seems that some work has to be done here by means of benchmark analysis. The validation of the programs has been performed by means of in-situ heater experiments.

modele Th n [63i s ] have been develope guido t ddesige th e d nan interpretatio n in-sita f no u heater test experimen e Stripth n ai t mine. Two experiments are modelled; one full-scale test comparable in size and powehigh-levea o t r l waste containe e time-scaleon d ran d test. Linear

64 y—Temperature dependent properties 0.5

iCn. G 5-1.0

to Öl

H. 5 100 0 30 20O 400 5OO Time (days)

FIG. 15. Plot showing measured rock displacements. Also included are displacements predicted using constan s wela ts temperatura l e dependent propertie e rocth k f o [71]s .

thermoelasticit d temperaturyan e independent therma mechanicad lan l rock properties are assumed. For the full-scale experiments, radial symmetry is assumed, while for the time-scaled experiments three-dimensional models have been used. An evaluation of the models is given in [71]. Comparisons between measured and predicted displacements in the rock have shown tharoce th tk movement r lesfa s e thasar n predicted from linear thermoelasticity for intact rock. Upon including temperature dependency for the material properties, better agreement could be obtained (see Fig. 15) .speciaA l purpose structural analysis prograundew no rs i m development [72].

The experiment describe n [66 i dalsn ]regardeca e ob a s a d validatio modele th f sno useharr fo dd rock.

No verificatio d validationan modele th dons ni n seo describine th g thermal stresses in clay.

65 5.4.3. Thermally induced groundwater convection

The verification and validation of the models used to describe the thermally induced groundwater convection has been initiated. Models effece oth f f stresseo t f fractureso e permeabilitth d san f roco ye ar k being developed. Verificatio d validationan n will need much attention and benchmarks shoul definee b d d [42.97].

5.4.4. Thermally accelerated chemical reactions

The model geochemicaf o s l processes describe Section i d n 5.2.4. have been validated with some succes n hydrothermasi l systemst ,bu validatio t beeno n s performenha combiner fo d d precipitation/dissolution, corrosion and alteration of waste package components.

Geochemical groundwater evolution models have been successfully used to describ chemistre th e wida f eo y variet naturaf o y l aquifer systems. The models have been verified and validated in a number of geothermal systems in the United States, Canada, and Europe [42,97].

66 . 6 DISCUSSIO RECOMMENDATIOD NAN N

6.1. Introduction

Thermal effect manf o onle ye sar factoron y s which influence th e performanc repositorya f o e performancr Fo . e analyses, integrated systems modellin prerequisitea s i g . Different repository concepts rely tdifferena o t individuae degreth n eo l systems components. Therefore, temperatures whic havy hma e detrimental effect singln so e componentsy ,ma not or only slightly affect the performance of a repository in some concepts. Validation of the modelling is needed to increase confidence e performancth n i repositorya f eo . Therefore site-specific data havo t e be collected for use in and for validation of the models. In addition, integrated experiment needede sar .

These integrated experiments should combin differene th e t components orepositora f evaluato t y interactionse th e , e.g. waste form, container corrosios oit r n products, backfill d sit,an e specific host roc wels ka l as site-specific water.

6.2. Thermal limitations

It is possible to establish two types of thermal limits or criteria on the various repository components. These are technical limits based absolute onth e behaviou n individuaa f ro l componen d practicatan r (o l engineering) limits based on the possibility to predict the behaviour of the component in a specific environment. A technical thermal limitation ensures that the desired properties of the component remain unchanged for the duration of the time for which it is expected to function, and it may or may not depend on the site-specific environment. Technical limitations have been summarized in [28,97 and 98]. Examples of such limitations, among many, are maximum temperatures for glass waste forms to prevent devitrification, maximum temperature r spensfo t fueo t l minimiz emaximud oxidatioan , mO U temperaturf no bentonitr efo e backfill to prevent irreversible loss of swelling capability due to collapse of the crystal lattice. The difficulty with such limitations is tha t necessaril no the o d y y addres kinetice e long-terth sth r sfo m alterations associated witthermae th h l effect nuclean si r waste repositories, they do not address the special environments associated

67 with specific repositories and they do not account for possible interactions between various repository components.

developmene Th practicaf to l thermal limit componentr sfo s wile lb required for engineering design of a site-specific repository. The practical limits should be based on a total system analysis which includes all aspects of component interaction and repository performance. The practical thermal limits for components will also include adequate factor conservatisf so alloo t muncertaintier wfo n si the properties of the materials as well as the variability in properties of the components, especially the naturally occurring elements such as host rocgroundwated kan r chemistry.

6.3. Mitigation strategies

Ther mainle ear y three way influenco st heae eth t release froma given accumulatio wastef no :

- Postponing buria lsame whicincreasins th ea s hi g interim storage. This will decrease the temperature maxima in the waste in its direct environmen wels ta s temperaturla fieldr fa e e maxim.th n i a

- Dispersing the waste over a larger area of over a larger volume of host rock. This will mainly reduce temperature maxima in the waste environment, and to a lesser extent the temperatures in the far field.

- Distributing of the waste over a larger number of repositories. This wil changt l no temperaturee eth neae th r n fieldsi ; howevet ri might reduce the far-field temperatures in some cases.

Beside the strategies mentioned before, technical concepts for the near field will influenc larga o et e exten near-fiele th t d temperatures, e.g. dimensions of waste form and container as well as properties and dimensions of the backfill.

6.4. Recommendations for further research

) (a Although generic researc calculationd han carriee b y sma d oute ,th part played by temperature in the overall performance of disposal

68 systems mus assessee b t site-specifin o d d concept—specifican c bases r instanceFo . extene ,th f fracturino t vary gma y froe mon graniti otherce th sit o ;t e and, ever similafo n r wasted san repository lay-outs, the thermally induced effects on stresses and groundwater flow may be quantitatively and qualitatively different, witpotentiae th h l impact waten so r geochemistry d late,an n ro waste form degradation processes. Therefore, it is recommended thay assessmenan t f temperaturo t e effects uses high quality, specific data and that the conclusions drawn from studies on a given site be not transferred to other sites without careful examination. significanA ) (b t improvemen theoreticaf o t l capabilities shoule b d made, both at the level (1) understanding and modelling of the behaviour of individual repository components (i.e. container, backfilling material), and (2) coupling and linking of individual models into integrated, comprehensive models, where temperature influenc adequatels ei y taken into account givo T .examplen ea : s seeItwa n that heatinn influenca e flos th gha w n eo patter n i n fractured and porous rocks, and that groundwater geochemistry will largely affect the degradation rate of the waste forms. Coupling of groundwater flod geochemicawan l model aren a as si requiring much additional work.

This leads to a second example: For fractured rock, most of the flow calculation e performesar d without taking into accoune th t changes in fracture permeability due to thermal stresses. This is howeve ra typica l exampl couplef eo d processes wels i lt i r ,fo known that even a minute change in fracture aperature can induce drastic changeroce th k n permeabilitysi . Future assessmenf to heat effects on groundwater flows should therefore be performed by coupled heat-stress-permeability hydraulic, mass transport calculations, for which adequate concepts and coupled computer codes should be important.

In addition, attention should be given to effects accompanying the cooling-down of the whole disposal system. Unfortunately the time-span necessar coolinr fo y g dow orders i n magnitudf so e longer than that for heating-up, so the rate at which these effects would occu extremels i r y slow.

69 (c) Lastly, the need for 'systems-scale1 testing (i.e. interactions in multi-component set-up whicn si h temperature play rolea s ) muse b t stressed, both for the production of reliable input data and for the validatio couplee th f no d theoretical models. Although these kind experimentf so e amenablsar laboratoro t e ye testingth o t e ,du better contro f teslo t condition n achieveca e son t turn ,i t sou thamose th tt representative conditions coul foune b d actualn i d , in-situ experiments, for instance in underground laboratories.

70 REFERENCES

] [1 COMMISSIO EUROPEAF NO N COMMUNITIES, Admissible Thermal Loadinn i g Geological Formations - Consequences on Waste Management Methods, Rep 8179R .EU , CEC, Luxembourg (1983).

] MARPLES[2 , J.A.C., LUTZE SOMBRET, ,W. , 'Th,C. e leachinf o g solidified high-level waste under various conditions', Radioactive Waste Managemen d Disposatan l (ProcFirse th f t.o European Community Conference, Luxembourg 20-23y ,Ma , 1980) (SIMON, ,R. ORLOWSK Eds, IS. ) Harwood Academic Publishers (1980).

[3] KBS, Handling and Final Storage of Unprocessed Spent Nuclear Fuel, Vol. I General, Stockholm (1978).

] [4 KBS, Handlin f Spengo t Nuclear Fue d Finalan l Storag Vitrifief eo d High Level Reprocessing Waste, Vol ,SafetV I y Analysys, Stockholm (1978. )24

[5] BEYERLEIN, S.A., CLAIBORNE, H.C., The possibility of multiple temperature maxim geologin i a c repositorie r spensfo t fuel from nuclear reactors, ORNL/TM-7024 Ridgk ,Oa e (1980).

[6] VOGEL, W., Glaschemie, VEB Deutscher Verlag für Grundstoffindustrie, Leipzig (1979).

] TOMOZAWA[7 , 'Phas,M. e separatio glass'n ni , Treatis Materialn eo s Scienc Technologyd ean , Vol , (TOMOZAWA.17 DOREMUS, ,M. , R.H., Eds) Academic Press, New York, (1979).

[8] LÖNNERBERG, B., LARKER, H., AGESKOG, L., Encapsulation and Handlin Spenf o g t Nuclear FueFinar lfo l Disposal, SKBF/KBS, Stockholm, TR 83-20 (1983).

] KUZMICHEVA[9 , E.U., KOVBA, L.M., IPPOLITOVA, E.A., Oxidatiof no uranium dioxide at temperatures below 270°C, Radiochimia, (1971, 6 . )No 852-857, Vol13 . .

ISEGHEMN [10VA ]TIMMERMANS, ,P. BATISTE D , , 'Chemica,W. ,R. l stabilit f simulateo y W formdHL contac n i s t with clay media , 1 Scientific Basis for Nuclear Waste Management V, Berlin 1982, (LUTZE , Ed.),W. , North-Holland Yorkw ,Ne , (1982).

[11] SAVAGE CHAPMAN, ,D. , N.A., 'Hydrothermal behaviou simulatef o r d waste glas waste-rocd san k interactions under repository conditions', Geochemistry of Radioactive Waste Disposal (BIRD, G.W., FYFE, W.S., Eds) Chem. Geol. (1982).

[12] GRAMBOW , 'Th,B. e rol metaf eo solubilitn io l leachinn i y f o g nuclear waste glasses', Scientific Basis for Nuclear Waste Management V, Berlin 1982 (LUTZE, W., Ed.) North-Holland, New York, (1982).

[13] BENCH, L.L., LODÛING, A., WERKE, L.O., 'Analysis of one year in situ buria nucleaf o l r waste glasse Stripan i s , Proc. 1 85th Annual MeetinAmericae th f go n Ceramic Society, April 1983 (TO BE PUBLISHED).

71 [14] MACEDO, P.B., BARKATT SIMMONS, ,À. , J.H. floA ' , we modeth r lfo kinetics of dissolution of nuclear waste forms; A comparison of borosilicate glass, Synro high-silicd can a glass', Scientific Basis for Nuclear Waste Management V, Berlin 1982 (LUTZE, W., Ed.) North-Holland, New York, (1982).

[15] OVERSBY, V.M., RINGWOOD, A.E., Leach testin SYNROf o g glasd Can s d 200°Csamplean 5 ,8 t Nuclsa . Chem. Waste Manag. 2_ (1981) 201.

[16] JOHNSON, L.H., BURNS, K.I., JOLING MOORE, ,H. , C.J.e ,Th Dissolutio f Irradiationo n IK>2 Fuel under Hydrothermal Oxidizing Conditions, AECL Whiteshell Nuclear Research Establishment, TR-128 (1981).

[17] JENKS, G.H., Effect Reactof so r Operatio HFIR-coolantn no , ORNL-3848 Ridgk ,Oa e Natl. Lab. (1961).

[18] STRUM MORGAN, ,W. , J.J., Aquatic Chemistry, (2nd edn), John Wiley and Sons, New York (1981).

[19] SKYTTE-JENSEN, B., Stability Diagrams for Copper at Different Temperatures and Exposed to Varying Groundwater Conditions, SKBF/KBS, Stockholm AR.82-3S ,KB 9 (1982).

Corrosioe [20Th ] n Resistanc Coppea f eo r Caniste Spenr rfo t Nuclear Fue Follol- SKBF/KBS, wup , Stockholm 83-2R T S 4, KB (1983).

[21] Disposa Radioactivf lo e Waste into Geological Formations: Studies on Crystalline Rock, European Appl. Res. Dept., Nucl. Sei. Technol. _4_ (1983) 1015.

[22] BRAITHWAITE, J.W., MOLECKE, M.A., Nuclear Waste Canister Corrosion Studies Pertinen Geologio t c Isolation, Sandia Natl. Lab. SAND 79-1935 (1979).

[23] DUDSON, P.J., CHAPMAN, N.A., Geochemical Factors Controlling the Nuclide Release Source Ter Graniten mi : Engineered Barrierr sfo High-Level Waste, Inst Geologicaf .o l Sciences, Harwell, ENPU 81-7 (1981).

[24] HENRIKSON, S., DE POURBAIX, M., Korrosionsprovning av olegerad tita simuleradi n e deponeringsmiljoe upparbetar fo r t kärnbränsleavfall. SKBF/KBS, Stockholm, Slutrapport 79-1R T S 4,KB (1979).

[25] RUPPEN J.A., GLASS, R.S., MOLECKE, M.A., Titanium Utilization in Long-Term Nuclear Waste Storage, Sandia National Lab. SAND 81-2466J (1981).

[26] SHIMOGORI SATO, ,K. TOMARI, ,H. Crevic, ,H. e Corrosio Titaniuf no m in NaCl Solution Temperature th n si e250°Co t Rang 0 . ,J e10 Japan. Inst. Met. 4_2_ (1978) 567.

[27] MEGEDER HEITZZU Korrosionsverhalte, , ,D. E. unlegierten nvo m Stahl, Stahlguss and Gusseisen als Endlagerbehälterwerkstoff in wasserführendem Granitgestein, NAGRA Technischer Bericht 82-08 (1982).

72 [28] SCOTT, J.A. , Limits on the Thermal Energy Release from Radioactive Wastes in a Mined Geologic Repository, ON1-4, Office of NWTS Integration, Battelle Memorial Institute, Columbus (1983)H ,O .

[29] SASKAR, A.K., BARNES, M.W., ROY, D.M., Longevity of Borehole and Shaft Sealing Materials: Thermodynamic Propertie Cementf o s d san Related Phases Applied to Repository Sealing, ONWI-201, prepared e PennbTh y . State Univ Officr .fo Nucleaf eo r Waste Isolation, Battelle Memorial Institute, Columbus (1982)H ,O .

[30] JACOBSSON ShorA , t,A. Formatione Revieth f wo , Stabilityd ,an Cementing Propertie f Naturaso l Zeolites, SKBF/KBS, Stockholm, KBS TR-27 (1977).

[31] JACOBSSON , PUSCH,A. Propertie, ,R. Bentonite-Basef so d Buffer Materials, SKBF/KBS, Stockholm, KBS TR-32.

[32] STEIN, R., COLYER, P.L., 'Pilot Research Projects for Underground Disposa Radioactivf lo e WasteUnitee th n dsi State Americaf so 1, Radioactive Waste Management (Proc. Int. Conf. Seattle 1983y ,Ma ) Vol IAEA, 3 . , Vienna (1984) 311.

[33] KÜHN, K., VERKERK, B., 'Disposal in salt formations', Radioactive Waste Managemen d Disposatan l Firs(Proce th tf . o European Community Conference, Luxembourg, May 20-23, 1980) (SIMON, R., ORLOWSK Eds, IS. ) Harwood Academic Publishers, (1980).

[34] OECD/NEA, Geological Disposal of Radioactive Waste: In Situ Experiments in Granite (Proc. Workshop, Stockholm, 1982) OECD, Paris (1983).

[35] PATRICK, W.C., RAMSPOTT, L.D., BALLOU, L.B., 'Experimental and calculational results from the spent fuel test-climax1, Geological Disposal of Radioactive Waste: In Situ Experiments in Granite, Proceeding Workshoa f so p hel Stockholn i d 1982n i m , OECD, Paris (1983).

[36] HEREMANS , "Possibilité,R. s d'évacuatio déchets nde s radioactifs dans une formation profonde d'argile plastique , Radioactive Waste 1 Management (Proc. Int. Conf. Seattle, May 1983) Vol. 3, IAEA, Vienna, (1984) 243.

[37] JOHNSTONE, J.K. al.t e ,, Unit Evaluatio Yucct na a Mountain, N.T.S.: Summary Report and Recommendations, Report SAND 83-0372, Sandia Natl. Lab., Alberquerque Mexicw ,Ne o (1983).

[38] JOHNSTONE, J.K., HADLEY, G.R., WAYMIRE, D.R., In Situ Tuff Water Migration/Heater Experiment: Final Report, SAND 81-1918, Sandia Natl. Lab., Alberquerque, New Mexico (IN PRESS).

[39] Site Characterization Basale Reporth r t tfo Isolatio n Program, Report DOE-RL 82-3, November (1982).

[40] THUNVIK, R., BRAESTER, C., Hydrothermal Conditions around a Radioactive Waste Repository SKBF/KBS, II & ,I , Stockholm, KBS TR 80-19 (1980); III, SKBF/KBS, Stockholm, KBS TR 82-01.

73 [41] HODGKINSON, O.P. Mathematica,A l Mode Hydrothermar lfo l Convection around a Radioactive Waste Repository in Hard Rock, UK Atomic Energy Authority, AERE-R 9149 (1979).

[42] OECD/NEA, CEC, Geological Disposal of Radioactive Waste. An overview of the current status of understading and development, OECD, Paris (1984).

[43] ANDERSON, W.J., "Conceptual Design Requirements for Spent Fuel, High-Level Waste, and Transuranic Waste Packages", Rockwell Handford Operations RHO-BW-ST-25P.

[44] DEUTSCH, W.J. Geochemical Modellin Nuclear-Waste th f go e Repository System. A Status Report. Pacific Northwest Laboratory Report PNL-3518 (1980).

[45] WEST, J.M., McKINLEY, I.G., CHAPMAN, N.A., Microbes in deep geological system d theisan r possible influenc radioactivn eo e waste disposal, Radioact. Waste Manage. Nucl. Fuel Cycle, _3_ (1982) 1-15. ~

[46] OECD/NEA, Sorption Modelling for Nuclear Waste Disposal Studies, OECD, Paris (1983).

[47J JOCKWER, N., "Laboratory Investigations on Waste Content Within Rock Salt and its Behaviour in a Temperature Field of Disposed High-Level Waste", Scientific Basi Radioactivr fo s e Waste Management, Boston 1980.

[48] FRITZ FRAPE, ,P. , S.K., 'Saline groundwater Canadiae th n si n Shiel firsA - d t overview1, Geochemistr Radioactivf o y e Waste Disposal (BIRD, G.W., FYFE, W.S., Eds) Chem. Geol., J36_ (1982) 179-190.

[49] RIMSTIDT, J.D., BARNES, H.L., The kinetics of silica-water reactions, Geochim. Cosmochim. Acta (1980_ ,44 ) 1683-1699.

[50] MOORE, D.E., MORROW BYERLEE, ,C. , J.D., Chemical reactions accompanying fluid flow through granite held in a temperature gradient, Geochim. Cosmochim. Acta, 47_ (1983) 445-454.

[51] KAMINENI, D.C., DUGAL, J.J.B., 'A study of rock alteration in the Eye-Dashwa Lakes Pluton, Atikokan, Northwestern Ontario, Canada', Geochemistry of Radioactive Waste Disposal (BIRD, G.W., FYFE, W.S., Eds) Chem. Geol (1982_ .36 ) 35-57.

[52] OVERSBY, V., IAEA Technical Committee on Studies Relating to Near-field Effect Deen i s p Geological Disposal Systems, Otaniemi, Finland, 22-26 August 1983.

[53] BIRD, D.,K., HELGESON, H.C., Chemical interaction of aqueous solutions with epidote-feldspar mineral assemblages in geologic systems. II. Equilibrium constraints in metamorphic/geothermal processes, Amer Sei.. .J 1 (1981,28 ) 576-614.

[54] WOOD, M.I., ADEN, G.D., LANE, D.L., Evaluation of Sodium Bentonite and Crushed Basalt as Waste Package Backfill Materials. Rockwell Hanford Operations Report RHO-BW-ST-121P (1982).

74 [55] TSANG, C.F., NOORISHAD, J., WANG, J.S.Y., 'A study of coupled thermomechanical, thermohydrological and hydromechanical processes associated wit nucleaha r waste repositor fracturea n i y d rock medium1, Scientific Basis for Nuclear Waste Management - VI, (BROOKINS, D.G., Ed.) North-Holland Yorkw ,Ne , (1983).

DirectorA [56] Computef o y r Program Assessmenr sfo Radioactivf to e Waste Disposa Geologican li l Formations. Study performey b d ATKINS R. and Epsom D., UK. CEC Report EUR 8669, Luxembourg (1983).

[57] HAMSTRA KEVENAAR, ,J. , J.W.A.M. ,procedurA ' PRIJ , ,J. r fo e detaile D analyse3- d s applie temperaturn o d e rise multi-layen i s r W repositorieHL sala n ti s dome1, Underground Disposaf lo Radioactive Waste (Proc. Symp. Otaniemi, 1979) Vol. 2, IAEA, Vienna, (1980) 189.

[58] KEVENAAR, J.W.A.M., PLOUMEN, P., JANSSEN, L.G.J., WINSKE, P., Compariso f temperaturno e calculation arbitrarn a r sfo W HL y disposal configuratio saln ni t formation. ECN-79-63 (1979).

[59] PLOUMEN , STRICKMAN,P. Berechnun, ,G. zeitlicher gde d nun räumlichen Temperaturverteilun säkularer de i gbe n Lagerung hochradioaktiver Abfälle in Salzstöcken. Aachen, Rheinisch-Westfälische Technische Hochschule (1977).

[60] TARANDI, T., Temperaturberäkningar för slutförvar för använt bränsle, SKBF/KBS, Stockholm, TR-46 (1978).

[61] TARANDI, T., Calculated temperature field in and around a repositor spenr fo y t nuclear fuel, SKBF/KBS, Stockholm, KB83-2R ST 2 (1983).

[62] HOPKIRK, R.J., GILBY, D.J., SCHWANNEK, J., Preliminary calculations of the temperature distribution around a type C (highly active) nuclear waste repository. Nagra, Baden, Switzerland, NTB 83-20 (In Press).

[63] CHAN, T., COOK, N.G.W., Calculated Thermally Induced Displacements and Stresse Heater sfo r Experiment Stripat sa , Sweden, Swedish-American Cooperative Progra Radioactivn mo e Waste Storage in Mined Caverns in Crystalline Rock, SKBF Sweden, Lawrence Berkeley Lab., LBL-7061, SAC-22 (1979).

[64] BUTKOVICH, T.R., As-Built Mechanical and Thermo-mechanical Calculation Spenf so t Fuel Tes Climan i t x Stock Granite, Lawrence Livermore National Laboratory, Univ Calif.f .o , UCRL-53198 (1981).

[65] LEHNHOFF, T.F., THIRUMALAI, K., KRUG, A.D., 'Modeling approach to determine short- and long-term thermal and thermomechanical effects of waste emplaced in a repository in basalt', Scientific Basi Nuclear sfo r Waste Managemen Berli, tV n 1982 (LUTZE Ed., ,W. ) North-Holland, New York, (1982).

[67] PUDEWILLS, A., MULLER, R., KORTHAUS, E., KÖSTER, R., 'Thermomechanical in-situ experiments and finite element computations', Scientific Basis for Nuclear Waste Management V, Berlin 1982 (LUTZE Ed., ,W. ) North-Holland (1982).

75 [67] HODGKINSON, O.P., Deep Rock Disposal of High Level Radioactive Waste: Initial Assessment of the Thermal Stress Field, U.K. Atomic Energy Authority, Harwell, AERE-R 8999 (1978).

[68] LEE, C.F., SIMMONS, G.R., 'Thermal effects on rock and fluid flow1, Environmental and Safety Assessment-Proc. of the Ninth Nuclear Fuel Waste Management Information Meeting, Jan. 20-21, 1981, compile Hilto. J y b dn Wright, Atomic Energ Canadf o y a Ltd, WNRE TR79 (1981).

[69] KLOK, J., PRIJ, J., Calculations on the closing of boreholes for nuclear waste disposa saln i l t domes, Contract Report WAS-226-81-7N, Sub-Contract 7-61(3), Commissio Europeae th f no n Communities, (1982).

[7Ü] LADOUCEUR MILLAR, ,B. , CalculeDA. s thermomécanique soln se s argileux, Rapport CEA.ÜEMT.SMTS.83/009.

[71] WITHERSPOON, P.A., COOK, N.G.W., GALE, J.E., Progress with Field Investigation Stripat sa , Swedish-American Cooperative Progran mo Radioactive Waste Storage in Mined Caverns in Crystalline Rock, SKBF Sweden, Lawrence Berkeley Lab. LBL-10559, SAC-27 (1980).

[72] COOK, N.G.W., WITHERSPOON, P.A., WILSON, E.L., MYER, L.R., 'Progress with thermomechanical investigations of the Stripa Site', Geological Disposal of Radioactive Waste: In Situ Experiments in Granite, (Proc. Workshop, Stockholm, 1982) OECD/NEA, Paris (1983).

[73] RAE, J., ROBINSON, P.C., NAMMU: Finite Element Program for Coupled HeaGroundwated an t r Flow Problems AtomiK ,U c Energy Authority, Harwell, AERE-R 9610 (1979).

[74] DOREMUS, R.H., 'Chemical durabilit glass'f o y , Treatisn eo Materials Science and Technology, Vol. 17 (TOMOZAWA, M., DOREMUS R.H., Eds) Academic Press, ...... (1979).

[75] CLAASSEN, H.C., Conceptual models governing leaching behaviours and their long-term predictive capacity, Nucl. Chem. Waste Manag. _2 (1981) 307.

[76] ALTENHEIN, F.K., LUTÜE, W., 'Long-term radioactivity release from high-level waste - Part II, Parametric study of waste form properties, temperature and time1, Scientific Basis for Nuclear Waste Management-VI, Boston 1982 (BROOKINS, D.G., Ed.) North-Holland Yorkw ,Ne , (1983).

[77] GOODWIN, B.W., LEMIRE, R.J., JOHNSON, L.H., 'A stochastic model for dissolution of irradiated U02 fuel1, Experimental Methodologies in Radionuclide Sorption Studies, (Proc. Workshop, Paris, 1982) OECÜ/NEA, Paris (1983).

[78] PUSCH, R., StabLlity of Deep-sited Smectite Minerals in Crystalline Roc Chemica- k l Aspects, SKBF/KBS, Stockholm, KB83-1R ST 6 (1983).

76 [79]' MULLER, A.B., LANGMUIR, D., DUDA, L.E., The formulation of an integrated physicochemical - hydrologie model for predicting waste nuclide retardation in geologic media, Mat. Res. Soc. Symp. 15 (1983) 547-564.

[80] HELGESON, H.C., KIRKHAM, D.H., FLOWERS, G.G., Theoretical prediction of the thermodynamic behaviour of aqueous electrolites at high pressures and temperatures - Parts I through IV, published originall foun i y r separate part 1974n i s , 197 d 1986an n 1i American Journal of Science, reprint.

[81] NAKAMURA TASHIRO, H. , , (Eds.,S. ) Progress Repor Safetn o t y Research of High-Level Waste Management for the Period April 1982 to March 1983, JAERI-M 83-076 (1983).

[82] PFISTER, E., NOLD, A.L., 'The Grimsel rock laboratory for in-situ experiments in crystalline rock', Radioactive Waste Management (Proc. Int. Conf. Seattle, May 1983) Vol. 3, IAEA, Vienna, (1984) 329.

[83] SIMMONS, G.R., et al. "The Canadian underground research laboratory", Radioactive Waste Management (Proc. Int. Conf. Seattle 1983y ,Ma ) Vol IAEA, .3 , Vienna, (1984) 291.

[84] GOSSE, V.B.. "Revieal t geologicaf ,e o w l criteri assessmend aan t techniques user sitfo de investigation disposar fo s f o l radioactive wastes in India", Seminar on the Site Investigation Techniques and Assessment Methods for Underground Disposal of Radioactive Wstes, Sofia, Bulgaria, 6-10 February 1984, IAEA-SR-104/27. [85] SHIMOOKA al.t e , ,,K. 'Pilo t research project undergrounr fo s d disposal of radioactive waste in Japan', Radioactive Waste Management (Proc. Int. Conf. Seattle, 1983) Vol. 3, IAEA, Vienna (1984) 339.

[86] ANDRE JEHAN, R., et al. "Etudes de stockage géologique en France" Radioactive Waste Management (Proc. Int. Conf. Seattle, 1983), VoIAEA, .3 , Vienna (1984) 443.

[87] HODGKINSON, O.P., LEVER, D.A., Interpretatio Fiela f no d Experiment on the Transport of Sorbed and Non-sorbed Tracers throug Fractura h Crystallinn i e e Rock, Radioactive Waste Managemen Nucleae th d rtan Fuel Cycle , (1983,4 ) 129.

[88] MANFROY, P., "Construction d'un laboratoire souterrain pour un programme expérimental sur le rejet des déchets en formation argileuse", Radioactive Waste Management (Proc. Int. Conf. Seattle, 1983) Vol. 3, IAEA, Vienna (1984) 275.

[89] TASSONI sitn laboratord "I uan , ,E. y heating experiment clan i s y rocks", Seminar on the Site Investigation Techniques and Assessment Methods for Underground Disposal of Radioactive Wastes, Sodia, Bulgaria, 6-10 February 1984, IAEA-SR-104/23.

[90] HEIM, G., "Plans for characterization of salt sites in the United States of America", Seminar on the Site Investigation Techniques d Assessmenan t Method Undergrounr fo s d Disposa Radioactivf lo e Wastes, Sodia, Bulgaria, 6-10 February 1984, IAEA-SR-104/33.

77 [91] INTERNATIONAL ATOMIC ENERGY AGENCY, Radioactive Waste Management Glossary, IAEA-TECDOC-264, Vienna (1982).

[92] HEREMANS BONNE, , R. MANFROY, ,A. , Experimentatio,P. d an n no evaluation of nearfield phenomena in clay. The Belgian approach. Workshop on Near-field Phenomena in Geological Repositories for Radioactive Waste, Seattle, 1981, OECD/NEA, Paris (1981) 187.

[93] KÜHN, K., "Pilot research projects for underground disposal of radioactive wastes in the Federal Republic of Germany", Radioactive Waste Management (Proc. Int. Conf. Seattle, 1983) Vol. 4, IAEA, Vienna (1984) 351

[94] CARLSON et al., and PATRICK et al., Spent Fuel Test - Climax: Technical Measurements, Interim Report 1980r fo s , 1981, 1982", UCRL - 53064, 53294, 53294-82 (1980, 1982, 1983).

[95] MONTAN , PATRICK,D. Therma, ,W. l Calculation Designe th r ,sfo Construction, Operatio d Evaluationan Spene th tf no Fue l Test- Climax, Nevada Test Site, UCRL 53238 (1981).

[96] EATON, R.R., JOHNSTONE, J.K., NUNZIATO, J.W., KORBIN, C.M.n ,I Situ Tuff Water Migration/Heater Experiment: Post Test Thermal Analysis, SAND 81-0912 Sandia Natl. Lab., Albuquerque, New Mexico (1983).

[97] CEC, Nuclear Scienc Technologyd an e , Admissible kThermal loading Geologican i l Formations Consequence Radioactivn so e Waste Disposal Methods, Vol Synthesi, .1 s report 817R ,EU 9 EN/FR, CEC, Luxembourg (1982).

[98] COME, B., VENET, P., "Thermal loading effects on geological disposal", Radioactive Waste Management (Proc. Int. Conf. Seattle, 1983) Vol. 4, IAEA, Vienna (1984) 375.

78 LIST OF PARTICIPANTS

BELGIUM

Hereman. M . Mr s CEN/SCK ßoeretan0 20 g B-2400 Mol

FINLAND

Mr. J.-P. Salo TVO Power Company Fredrikinkatu 51-53 SF-00100 Helsinki, 100

FRANCE

Duri. M . nMr CEA/DEMT/DEDR 1 . B.PNo . F-91191 Gif-sur-Yvette

. RicharMr d ANDRA Division Technique a Fédératiol e d e ru n, 34 F-75015 Paris

INDIA

Mr. A. Kumar Babha Atomic Research Centre Trombay, Bombay 400 085

ITALY

Mr. E. Tassoni ENEA C.S.N. Casaccia S.MGalerii d . a Rome

JAPAN Muraok. S . Mr a Departmen Environmentaf o t l Safety Research Nuclear Safety Research Centre Tokai Research Establishment JAERI Tokai-mura, Ibaraki-ken

NETHERLANDS

Mr. J. Prij Netherlands Energy Research Foundation 1 x P.OBo . NL-175 PetteG 5Z n

79 SWEDEN

Mr. K. Anderson SKI Box 27106 Stockhol2 5 S-102 m

Mr. L.B. Nilsson Division KBS SKBF Box 5864 S-102 48 Stockholm

SWITZERLAND

Mr. F. van Dorp Société coopérative nationale pour l'entreposag déchete ed s radioactifs NAGRA Parkstrasse 23 CH-5401 Baden

UNITED KINGDOM

Savag. D . eMr Institut Geologicaf eo l Sciences Harwell, Didcot

UNITED STATES OF AMERICA

Mr Ballo. .L u Lawrence Liveraore National Laboratory 8 80 x Bo P.O. Livermore 9455A ,C 0

Mr. J.K. Johnstone Sandia National Laboratories 580x Bo 0P.O. Albuquerque 8718M ,N 5

INTERNATIONAL ORGANIZATIONS

COMMISSIO EUROPEAE TH F NO N COMMUNITIES

Mr. B. Come Office Directoth f eo r Generaf lo Science, Researc Developmend han t Commission of the European Communities 0 20 i Lo a l Rue d e B-1049 Brussels, Belgium

OECD/NUCLEAR ENERGY AGENCY

Mr. A. Muller Radiation Protection and Waste Management Division OECD/NEA boulevar, 38 d Suchet F-75015 Paris, France

80 CONSULTANT

SWEDEN

Werm. L . e Mr SKBF/KBS Box 5864 S-10 Stockhol8 4 2 m

81 ORDEO T W R IAEHO A PUBLICATIONS

An exclusive sales agent for IAEA publications, to whom all orders inquiried an s shoul addressede db bees ha ,n appointed in the following country:

UNITED STATE AMERICF SO A UNIPUB, P.O 433x .Bo , Murray Hill Station Yorkw 1015Y Ne ,N , 7

In the following countries IAEA publications may be purchased from the sales agent r booksellerso s liste r througdo h your major local booksellers. Payment can be made in local currency or with UNESCO coupons.

ARGENTINA Comision Nacional de Energi'a Atomica, Avenida'del Libertador 8250, R A-1429 Buenos Aires AUSTRALIA Hunter Publications GippA 8 5 ,s Street, Collingwood, Victoria 3066 BELGIUM Service Courrier UNESCO, 202, Avenue du Roi, B-1060 Brussels CZECHOSLOVAKIA S.N.T.L., Spâlené 51, CS-113 02 Prague 1 Alfa, Publishers, Hurbanovo nâmestie 6, CS-893 31 Bratislava FRANCE Office International de Documentation et Librairie, 48, rue Gay-Lussac, F-75240 Paris Cede5 x0 HUNGARY Kultura, Hungarian Foreign Trading Company P.O. Box 149, H-1389 Budapest 62 INDIA Oxford Boo Stationerd kan y Co. , Par,17 k Street, Calcutta-706 001 Oxford Book and Stationery Co., Scindia House, New Delhi-110 001 ISRAEL Heilige Co.d an r , Ltd., Scientifi Medicad can l Books , Natha3 , n Strauss Street, Jerusalem 94227 ITALY LibreriaScientifica, Dott. Lucio de Biasio "aeiou". Via Meravigli 16, I-20123 Milan JAPAN Maruzen Company, Ltd., P.O 5050x .Bo , 100-31 Tokyo International NETHERLANDS MartinusNijhoff B.V., Booksellers, Lange Voorhout 9-11, P.O 269x .Bo , NL-2501 The Hague PAKISTAN Mirza Book Agency, 65, Shahrah Quaid-e-Azam, P.O. Box 729, Lahore 3 POLAND Ars Polona-Ruch, Centrala Handlu Zagranicznego, Krakowskie Przedmiesci , PL-00-06e7 8 Warsaw ROMANIA llexim, P.O. Box 136-137, Bucarest SOUTH AFRICA Van Schaik's Bookstore (Pty) Ltd., Libri Building, Church Street, P.O. Box 724, Pretoria 0001 SPAIN Diaz deSantos, Lagasca 95, Madrid-6 Diaz de Santos, Balmes 417, Barcelona-6 SWEDEN AB C.E. Fritzes Kungl. Hovbokhandel, Fredsgatan 2, P.O. Box 16356, S-103 27 Stockholm UNITED KINGDOM Her Majesty's Stationery Office, Publications Centre 276x P.OBo .R , Londo5D 8 nSW U.S.S.R. Mezhdunarodnaya Kniga, Smolenskaya-Sennaya 32-34, Moscow G-200 YUGOSLAVIA Jugoslovenska Knjiga, Terazij , YU-1100 36 , P.O x e27 Bo . 1 Belgrade

Orders from countries where sales agents have not yet been appointed and requests for information should be addressed directly to: Division of Publications International Atomic Energy Agency Wagramerstrasse 5, P.O. Box 100, A-1400 Vienna, Austria