Advances in Understanding Engineered Clay Barriers

r ' Advances in

UNDERS:t;ANDING

Engineered

Clay B arriers T \

A D V A N C E S IN U N D E R S T A N D IN G E N G IN E ER ED C LA Y B A R R IER S

PROCEEDINGS OF THE INTERNATIONAL SYMPOSIUM ON LARGE SCALE FIELD TESTS IN

GRANITE, SITGES, BARCELONA. SPAIN, 12-14TH NOVEMBER 2003

Advances in Understanding

Engineered Clay Barriers

Edited by

Eduardo E. A lonso & Alberto Ledesma

Technical University of Catalonia (UPC). Barcelona, Spain

A .A. BALKEM A PU BLISH ERS Leiden / London ; New York Philadelphia / Singapore

Cover; Picture based on the Febex Experiment. Courtesy o f ENRESA and Transedit, Spain

enresa

Svensk Karnbranslehantering AB

Copyright © 2005 Taylor & Francis G roup pic, London, UK

A ll rights reserved. N o p a rt o f this publication or the inform ation contained herein m ay be reproduced,

stored in a retrieval system, or transmitted in any fo rm o r by any means, electronic, m echanical, by

photocopying, recording or otherwise, without written p rio r perm ission fro m the publisher.

Although all care is taken to ensure the integrity and quality o f this publication and the inform ation

herein, no responsibility is assum ed by the publishers nor the author fo r any dam age to pro p erty or

persons as a result o f operation or use o f this publication and/or the information contained herein.

Published by: A.A. Balkema Publishers, Leiden, The N etherlands,

a m em ber o f Taylor & Francis G roup pic

w w w .balkem a.nl,www .tandf.co.uk

ISBN 04 1536 544 9

Printed in Great Britain

Advances in Understanding Engineered Clay Barriers - Alonso & Ledesma (eds)

© 2005 Taylor & Francis Group. London, ISBN 041536 544 9

Table of Contents

Preface

Opening

The FEBEX test as a benchmark case for THM modelling, historical

perspective and lessons learned

E.E Alonso & J. Alcoverro

Field emplacement and instrumentation techniques

Transducers and cable connections for measuring THM-processes in engineering

barriers - design and experiences

T. Sanden. R. Goudarzi & L. Borgesson

Components and processes affecting the recorded performance o f clay buffers

R. Pusch & G. Ramqvist

Examples of the performance o f tunnel plugs with and without recesses - theory and practice

R. Pusch <& L Borgesson

Emplacement o f bentonite blocks and canisters in a K.BS-3V

fashion - how feasible is the proposed technology?

C. Svemar

47

The Mock-Up-CZ project: a demonstration of engineered barriers

J. Pacovsky

53

The FEBEX “In Situ” test: lessons learned on the engineering aspects of

horizontal buffer construction and canister emplacement

J-L. Fuentes-Cantillana

Development o f backfilling techniques

D. Gunnarsson. L. Borgesson, L. Hallstedt <& M. Nord

Reliability o f THM instrumentation for underground research laboratories

J-L. Garcia-Siheriz <& I. Bdrcena

85

In situ saturated hydraulic conductivity of a backfill by means of pulse tests

C Mata & A. Ledesma

Measurements of water uptake in the buffer and the backfill of the

Aspo Prototype Repository using the geoelectric method

K. Wieczorek & T. Rothfuchs

Hydro geological and hydro mechanical monitoring o f the rock mass in

boreholes during the operation phase o f the Prototype Repository, Aspo HRL, Sweden

I. Rhen, T Forsmark & P. Aim

Identification o f “in situ” stresses from pore water pressure measurements

A. Marlinez-Parra, A. Ledesma & E.E. Alonso

Hydraulic characterisation o f the FEBEX granite: test performance and field interpretation

F. Ortuno, G. Carretero. L. Martinez-Landa & J. Carrera

115

125

133

Hydraulic surface packer tests on the EDZ in granite - comparison o f results

from PR-drift (Aspo) and FEBEX-drift (Grimsel)

L. Liedtke & H. Shao

143

EDZ in granite

B. Bazargan Sabet, H. Shao, J. Autio, Fco. J. Elorza, I. Cañamón,

J.C. Perez, R. Pusch & C. Svemar

Temperature Buffer Test instrumentation

T. Sandén, M. de Combarieu & H. Hokmark

163

Fundamental research material behaviour and laboratory testing

Thermo-mechanical and geochemical effects on the permeability of high-density clays

M.V. Villar. E. Romero & A. Lloret

177

Microstructural changes o f compacted bentonite induced by hydro-mechanical actions

E. Romero, C. Hoffmann. E. Castellanos, J. Suriol & A. Lloret

Hydro-chemical behaviour o f bentonite - based mixtures in engineered barriers

C. Mata & A. Ledesma

193

203

Hydro-mechanical properties o f backfill material

L~E. Johannesson, L. Borgesson & D. Gunnarsson

A constitutive model for compacted expansive clay

M Sánchez, A. Gens, L. do N. Guimaraes & S. Olivella

211

221

Short and long term stability o f selected bentonites in high saline solutions

J. Kasbohm, H. J. Herbert & K.-H. Henning

Bentonite swelling pressure in NaCl solutions - Experimentally determined data and

model calculations

O. Kamland, A. Muurinen & F. Karlsson

231

Analysis and distribution o f waters in the compacted FEBEX bentonite:

pore water chemistry and adsorbed water properties

A.M. Fernández & P. Rivas

Early chemical alteration of buffer smectite in repository environment

R. Pusch, H. Takase & T Noguchi

Chemo-mechanical behaviour o f high-density bentonites. Imbibition and diffusion tests

G. Musso & E. Romero

257

111

283

Fundamental properties o f bentonite pellet for Prototype Repository Project

y. Sugita, M. Chijimatsu & H. Suzuki

293

Adsorption site for Cs“^, and Sr^'*’ in bentonite by means o f EXAFS spectroscopy 301

M Nakano & K. Kawamura

ECOCLAY II. Laboratory experiment concerning compacted bentonite contacted to high pH solutions 307

O. Karnland, U. Nilsson & S. Olsson

Experimental and numerical study o f the T-H-M behaviour o f compacted FEBEX

bentonite in small-scale tests

M.V. Villar, M. Sánchez. A. Lloret, A. Gens & E. Romero

323

Barrier behaviour and THM modelling

T-H-M modelling of the Prototype Repository Experiment: comparison with current measurements

A. Ledesma & G.J. Chen

339

Simulation of the Prototype Repository

H.R. Thomas. P.J. Cleall & T.A. Melhuish

Hydraulic bentonite/rock interaction in FEBEX experiment

L. Borgesson & J. Hernelind

Temperature Buffer Test - comparison o f modelling results/experimental findings;

causes o f differences

H. Hökmark

353

373

Effects o f over-heating on the performance of the engineering clayed barriers o f the mock-up test 391

P.L. Martin & J.M. Barcala

Impact of canister corrosion on barrier behaviour 413

S. Olivella, J. Alcoverro & E.E. Alonso

Stresses and strains in a canister induced by corrosion 421

J. Alcoverro. S. Olivella & E.E. Alonso

Thermo-hydro-mechanical simulation of the concrete bulkhead during the heating phase

o f the runnel sealing experiment 4 3 1

R. Guo. B. Kjartanson <& J. Martino

Gas generation and transport measurements in the FEBEX project 443

N. Jockwer & K. Wieczorek

Coupled thermo-hydro-mechanical analysis o f the Prototype Repository Project using

numerical code THAMES 451

M. Chijimatsu, H. Kurikami & Y. Sugita

THM predictive modelling o f the Temperature Buffer Test - Clay Technology’s contribution 461

B. Faith, L. Borgesson. H. Hökmark & J. Hernelind

Mock-up for studying THM behaviour of swelling clay MX80 at temperature >100°C 483

C. Gatabin & J. C. Robinet

Chemical effects. HC and THMC modelling

Direct and inverse modelling of multicomponent reactive transport in single and dual porosity media 493

J. Samper. A.M. Fernández. L. Zheng. L. Montenegro, P. Rivas Z Dai

Pore water chemistry of saturated FEBEX bentonite compacted at different densities

A.M. Fernández & P. Rivas

505

Difftision coefficients and accessible porosity for HTO and chloride in

compacted FEBEX bentonite

M. García-Gutiérrez, J.L. Cormenzana, T Missana <& M. Mingarro

Reactive solute transport mechanisms in nonisothermal unsaturated compacted clays

J. Samper. L. Zheng. J. Molinero, A.M. Fernández, L Montenegro & P. Rivas

Predictive model of geochemical changes in porewater, buffer and backfill in

an engineered barrier system

A. Luukkonen

Thermo-hydraulic-chemical-mechanical behaviour o f unsaturated bentonite

H.R. Thomas. P .J Cleall & S.C. Seetharam

Inverse hydrochemical modelling o f aqueous extract experiments for the estimation

o f FEBEX bentonite pore water chemistry

J. Samper, A. Vázquez & L Montenegro

Testing coupled thermo-hydro-geochemical models with geochemical data

from FEBEX in situ test

J. Samper. A.M. Fernández. L. Zheng, L. Montenegro. P. Rivas & A. Vázquez

525

535

547

553

565

515

Closing conference

Nuclear repositories in granite. Research and development needs in the near future

C. Svemar. F. Huertas & R. Pusch

579

A uthor index 583

Advances in Understanding Engineered Clay Barriers - Alonso & Ledesma (eds)

© 2005 Taylor & Francis Group, London, ISBN 04 1536 544 9

Preface

Over the past few years the use o f clay barriers for waste isolating purposes has deserved much attention in

the geotechnical engineering community. This interest has been linked to a fundamental research activity

including the understanding of the behaviour of compacted materials and, particularly, compacted expansive

clays. The interaction between the barrier, the waste and the surrounding ground may involve several thermo￾hydro-mechanical and chemical coupled processes that have been analyzed by means o f “in situ” tests,

laboratory experiments and numerical modelling by many research groups. Within this context, some large

scale field tests have been developed in recent years by some European Agencies dealing with the management

o f radioactive waste. These experiments have provided an opportunity to calibrate and to validate computa￾tional models and to gain experience on instrumentation and installation techniques. This book constitutes a

state-of-the-art o f a number of relevant issues concerning engineered clay barriers. It includes about 50 papers

presented in a symposium held in Sitges {Barcelona, Spain) in November 12th-14th, 2003. Most of the

contributions refer to the following large scale tests being performed in Europe:

- Backfill and Plug Test & Prototype Repository, managed by SKB in Aspo (Sweden).

- Febex, managed by ENRESA (Spain) and carried out at Grimsel underground Laboratory (Switzerland).

- Temperature Buffer Test managed by ANDRA (France) and operating in Aspo Hard Rock Laboratory

(Sweden).

All o f them include bentonite barriers for radioactive waste isolation in granite environments. However, despite

of the specific characteristics of those tests, the majority o f the papers focus on the understanding o f the

processes involved.

Sessions of the symposium were organised according to four main topics:

- Field emplacement and instrumentation techniques (i.e. lessons learned in test installation and dismantling,

reliability o f instruments, “in situ” determination o f barrier and rock properties, advances in rock

characterization, site hydrogeology and its effect in test performance, impact o f test installation o f hydro￾geological conditions of host rock, EDZ in granite).

- Fundamental research. Material behaviour and laboratory testing (i.e. advanced constitutive modelling of

expansive compacted bentonite, permeability o f compacted bentonite, evolution of bentonite structure

during saturation and heating, effect of temperature on constitutive behaviour, properties and behaviour of

bentonite-based mixtures, role of small-medium scale cells in parameter/properties determination, advanced

laboratory techniques, geotechnical parameters and experimental determination, chemical and physical

properties o f interstitial water, illitization of bentonite, properties of bentonite-based materials).

- Barrier behaviour and Thermo-Hydro-Mechanical modelling (i.e. predicting long term saturation, interaction

between barrier and host rock, effect of rock heterogeneity and EDZ on buffer response, canister corrosion

effects, effects o f environmental conditions, effects o f overheating on barrier performance, long-term stress

state of the barrier).

- Chemical effects, Hydro-Chemical and Thermo-Hydro-Mechanical and Chemical modelling (i.e. advances

in hydro-chemical modelling of bentonite barriers, geochemical models o f water-bentonite interaction,

transport mechanisms through the barrier, advances in THMC modelling, geochemical modelling of

dismantling data).

The symposium was an opportunity to share experiences from those experiments and to identify areas of

uncertainties in order to guide future research activities. This book reflects this effort. Its publication has been

made possible thanks to the sponsorship of ENRESA and SKB, the Spanish and the Swedish National Agencies

for Radioactive Waste Disposal. Their support is gratefully acknowledged. The symposium provided also an

opportunity of meeting people from different countries with a common interest, which is nowadays essential in

an international and complex issue such as the safe disposal of nuclear waste. We hope that this publication will

contribute to a better understanding o f clay barriers and to an efficient and safe design of present day concepts

and engineering solutions for waste disposal. It is believed also that a large proportion o f the topics covered

by the different papers have a wide interest in the general field o f Geotechnical and Geo-€nvironmental

Engineering.

Barcelona, February 7th, 2005

Eduardo Alonso, Alberto Ledesma

Opening

Advances in Understanding Engineered Clay Barriers - Alonso & Ledesma (eds)

© 2005 Taylor & Francis Group, London, ISBN 04 1536 544 9

The FEBEX test as a benchmark case for THM modelling. Historical

perspective and lessons learned

Eduardo E. A lonso and J. A lcoverro

Departamento de Ingeniería del Terreno. Universidad Politécnica de Cataluña. Barcelona. Spain

ABSTRACT: The FEBEX (Full-scale Engineered Barriers Experiment in Crystalline Host Rock) “in situ”

test was installed at the Grimsel Test Site (Switzerland) to demonstrate the feasibility of the underground

storage o f high level nuclear waste in granite. The simulated canister is surrounded by compacted bentonite

whose thermo-hydro-mechanical properties are described in the paper. The measured performance o f the test

during three years constitutes the basis for a benchmark exercise involving various research teams and

computational models. The following aspects o f test performance are discussed in the paper: water pressure

changes induced in the host rock during the excavation o f the tunnel, the distribution and evolution of relative

humidity and swelling-induced stresses within the bentonite barrier, and the stresses and water pressures

recorded in the rock as a result o f the heating and swelling pressure development inside the excavated tunnel.

The paper discusses the merits of different formulations and the difficulties encountered to match some

aspects o f the observed behaviour.

1 INTRODUCTION

FEBEX was divided into three main parts, namely the

“in situ” test, a mock up test and an extensive

laboratory program on bentonite properties. The

FEBEX in situ test was designed as a demonstration

experiment of the Spanish reference concept for

disposal o f high level radioactive waste in granitic

rock. Nuclear canisters are deposited in deep

horizontal drifts. A bentonite buffer is provided

around the cylindrical canisters to isolate the waste

(Fig. 1 shows the general layout and the dimensions

of the experiment). In the FEBEX experiment,

compacted bentonite “bricks”, initially unsaturated

(Sr = 50%) were manually placed around cylindrical

steel heaters, which simulate the nuclear canisters.

These heaters maintained the temperature of the

bentonite-steel contact at a constant value of lOO'^C.

The bentonite buffer is subjected to two basic

phenomena: a progressive hydration induced by the

natural water provided by the granitic rock and a

desiccation action induced by the heaters. More than

600 sensors were emplaced in the rock, as well as a

bentonite buffer, during the installation. They have

provided time records o f several variables (tempera￾ture, rock pore water pressures, bentonite relative

humidity (RH), stresses in rock and bentonite,

displacements), which provide an accurate image of

the performance o f the near field rock, and, specially,

of the bentonite reaction.

The FEBEX drift was excavated in October 1995.

During excavation (using a TBM machine) and

immediately afterwards, pore water pressures were

recorded in boreholes located in the vicinity of the

gallery. Additional small diameter holes were

radially perforated from the FEBEX tunnel in order

to locate monitoring instruments. Water inflow into

the excavated tunnel was measured in February￾April 1996. Once the two heaters and the bentonite

buffer were in place, a concrete plug was built to seal

the test area (Fig. 1). Heating started on February 27th,

1997 (a date identified as Day 0 in the plots given in

this paper). The present paper discusses the data

recorded during the first 1000 days of test operation.

The FEBEX is described in detail in Enresa (2000).

This report provides data on the objectives o f the

experiment, on the geometry and construction

sequence, on the properties of the granitic rock and

the bentonite, on the interpretation o f monitoring

results and on the modelling work performed by the

research team participating in the FEBEX experiment.

One of the main concerns raised during the design

stage o f the experiment was the expected hydration

time of the bentonite buffer. In performance evalua￾tion studies it is assumed that a fully saturated

bentonite buffer is isolating the canisters. However,

no precise information on the time required for this

saturation was available. It was clear that the test

would provide data on the initial transient regime

o f the buffer hydration and on the progressive

' Bentonite blocks

Steel liner

' Heater (diameter 0.9)

• Granite Service zone, control and

data acquisition system

(Dimensions in meters)

Figure 1. General layout of the FEBEX “in situ” test.

development o f swelling pressures but probably not

on the time for full saturation.

The preliminary analyses carried out during the

design o f the test identified the key properties that

were expected to control the saturation o f the

bentonite barrier:

1 Water retention curves o f granite and backfill

2 Saturated permeability o f granite and backfill

3 Variation o f permeability with suction for granite

and backfill (relative permeability)

4 Water pressure boundary conditions o f the granite

in the vicinity o f the excavation.

Not all the properties or variables mentioned were

known in 1994, when the preliminary analysis was

performed. But some tests on rock and compacted

bentonite specimens were already available. For

instance, Figures 2 and 3 show the hydraulic con￾ductivity and the water retention properties of

compacted bentonite at various dry densities. It was

expected that the dry density of the blocks would reach

7d=1.65g/cm ^ in order to guarantee a swelling

pressure in excess of 5 MPa. Figure 2 indicates that

the saturated hydraulic conductivity o f the bentonite

was close to 2 — 6 x 10” '^ m/s for the expected dry

densities prevailing in the buffer. Data on relative

permeability was not available and the hydraulic

conditions of the granitic rock were roughly known.

Figure 2. Hydraulic conductivity of compacted specimens

of S-2 montmorillonite for two different dry densities and

varying temperature.

However, some thermo-hydro-mechanical anal￾yses were performed with the purpose o f establishing

the relative effects and importance o f the parameters

mentioned. The hydration time of the buffer was

defined as the time required for saturating a point

Figure 3. Suction water content relationship in tests on

unconfined samples, for FEBEX and S*2 bentonite

(Enresa, 2000).

Figure 4. Time to reach saturation in the reference point.

located at a radial distance o f 0.15 m from the heater.

The results o f calculations performed were repre￾sented in a dimensionless plot relating the dimension￾less time to saturation (T = K^entonite^^i. where L is

the thickness of the buffer) against the ratio of A finite

element program for hydro-mechanical analysis of

expansive soils was used to perform the calculations.

Several cases were run. characterised by different

combinations of granite and backfill permeability. A

bentonite buffer thickness L = 0,75 m was expected

at this time. The results are shovMi in Figure 4. The

plot shows a clear qualitative change in beha\iour

for a K bem onite^l^Tan.te tatio Of 8 X 10“". BclOW^ this

ratio, the time for saturation.

T L

( 1)

only depends on the saturated bentonite permeability.

Abov e it. the time for saturation increases and it is

also controlled by the granite permeability. It is

interpreted that, for K^niomii T ^m ie < 8 10"‘ the

granite provides all the necessary water to saturate the

buffer. Then, only the bentonite permeability controls

hydration times. When granite permeability decre￾ases and Kt«.ntonue^K.eranite > 8 X 10“^, the granite is

not able to supply all the water demanded by the

bentonite and both, the bentonite and the granite

permeabilities, control hydration times.

Grimsel granite has a matrix permeability suffi￾ciently high to provide all the saturation water

required by the bentonite. Time for saturation can

be easily calculated on the basis o f Figure 4. More

precise laboratory determinations o f the saturated

FEBEX bentonite lead to a figure Kbentonite^Kgranite =

3 X lO^'^’ m^s. Then, for Kgranue = and

L = 0.75m, a time for satination t = 31.7yr is

obtained. Saturated permeabilities (Kbeniomte^granite)-

It became clear that the expected time length o f the

experiment (3 to 6yr) was very short to allow the

saturation o f the buffer. The question of saturation

time has remained as a relevant discussion item as

new data coming from the laboratory, the “in situ”

and the “mock-up” tests, and the refined analyses

performed w'ere considered.

Therefore, the FEBEX in situ test data correspond￾ing to the first 1000 days covers the initial transient

response of the bentonite buffer, where thermo￾hydro-mechanical phenomena experience the fastest

rates o f change.

The main objective o f the FEBEX in situ test was

to demonstrate the feasibility o f the envisaged

emplacement procedure. This objective was largely

accomplished. The second objective was the study of

the THM processes in the near field and it was

reasonably well covered. The available data was

judged also o f interest to conduct a benchmark

exercise by interested research groups. This exercise

was developed within the context o f Decovalex III

Project. For the purposes o f conducting the bench￾mark case, in an ordered way. three parts or stages

were defined: the first one examines the granite

response to the excavation of the rock. The second

part refers to the buffer response and it is based on the

comprehensive data base obtained from sensors

located within the buffer. In the third part, attention

is turned again to the rock, affected now by the

temperature field generated by the heater and by the

swelling pressure of the bentonite against the tunnel

walls.

This paper provides some relevant data and

conclusions established during the performance of

this exercise. Some comparison plots between

predictions and actual measurements are also given.

A full report on the benchmark, which includes a

comprehensive evaluation o f the modelling exercise

is also available (Alonso & Alcoverro. 2004). Some

insight into some of the main phenomena observed is

given this paper.