CONDITIONS - LONG CORROSION

OFHC-COPPERS IN SIMULATED REPOSITORY CONDITIONS - FINAL REPORT

Pertti Aaltonen Paivi Varis

Technical Research Centre of Finland Metals Laboratory

March 1993 Helsinki ISSN 0359-548X

IMATRAN VOIMA OY (IVO) TEOLLISUUDEN VOIMA OY (TVO)

Address 010191VO FINLAND

Telephone Nat. (90) 5081 lnt. +358 0 5081

Telex

124608 voima sf

Address Telephone Telex

Annankatu 42 C Nat. (90) 618 01 122065 tvo sf SF-00100 HELSINKI lnt. +358061801

FINLAND

Tekija(t) - Author(s)

Pertti Aaltonen, Pili. vi V aris

Technical Research Centre of Finland Metals Laboratory

Nimeke - Title

Toimeksiantaja(t)- Commissioned by

Teollisuuden Voima Oy

LONG TERM CORROSION TESTS OF OFHC-COPPERS IN SIMULATED REPOSITORY CONDffiONS - FINAL REPORT

Tiivistelma - Abstract

This research program "Long term corrosion tests of OFHC-coppers in simulated repository conditions" was planned to provide an experimental evaluation with respect to the theoretical calculations and forecasts made for the corrosion behaviour of OFHC-coppers in bentonite ground water environments at temperatures between 20 - 80 °C. The aim of this study in the first place was to evaluate the effects of ground water composition, bentonite and temperature on the equilibrium and possible corrosion reactions between OFHC coppers and the simulated repository environment. The tests were started in 1987 and this final report includes the results obtained after 72 months exposure time.

ISSN

ISSN-0359-548X

Sivumaara - Number of pages

32

Jakaja - Distributed by

Teollisuuden Voima Oy Nuclear Waste Office Annankatu 42 C 00100 HELSINKI Tel. 90-61801

Kieli - Language

English

Tekija(t) - Author(s)

Pertti Aaltonen, Paivi V aris

V altion teknillinen tutkimuskeskus (VTT) Metallilaboratorio

Nimeke - Title

Toimeksiantaja(t)- Commissioned by

Teollisuuden Voima Oy

HAPETTOMIEN OFHC-KUPARIEN KORROOSIO LOPPUSIJOITUSTA VAST AA VISSA OLOSUHTEISSA - LOPPURAPORTTI

Tiivistelma - Abstract

Tassa tutkimuksessa selvitetaan kokeellisesti hapettomien OFHC-kuparien korroosioilmiot ja -nopeus loppusijoitusta vastaavissa olosuhteissa ja verrataan saatuja tuloksia teoreettisesti laadittuihin korroosionopeusarvioihin. Ensisijaisesti pyritaan selvittamaan pohjaveden laadun, bentoniitin ja lampotilan vaikutusta havaittaviin korroosioreaktioihin. Tassa loppuyhteenvedossa esitetaan tulokset, jotka on saavutettu 72 kuukauden pituisen koeajan jalkeen.

ISSN

ISSN-0359-548X

Sivumaara - Number of pages

32

Jakaja - Distributed by

Teollisuuden Voima Oy Y dinjatetoimisto

Annankatu 42 C 00100 HELSINKI Tel. 90-61801

Kieli - Language

Englanti

PREFACE 1

1 INTRODUCTION 2

2 BACKGROUND 2

2.1 Disposal of spent fuel 3

2.2 The chemical environment in the sealed repository 4

3 EXPERIMENTAL METHODS 6

3.1 Immersion test environments 7

3.2 Canister materials 8

3.3 Evaluation of chemical reactions during immersion periods 8

4 TEST RESULTS 11

4.1 Ground water analysis 11

4.2 Bentonite analysis 18

4.3 Surface analysis 23

5 DISCUSSION 25

5.1 Ground water in the containers 25

5.2 Ground water wetted bentonite inside copper canisters 27

5.3 Corrosion products 28

6 CONCLUSIONS 30

7 REFERENCES 31

PREFACE

This study is a part of the research work coordinated by the Nuclear Waste Commission of Finnish Power Companies (YJT). The work has been carried out at the Technical Research Centre of Finland (VTT) and it was financed by the Teollisuuden Voima Oy (TVO). This is a final report of the test program started in 1987.

Contact persons during this program at TVO have been Jukka-Pekka Salo and Esko Peltonen and at VTT Paivi Varis and Pertti Aaltonen. The chemical analysis and bentonite pH measurements have been carried out by the Chemistry Labora- tory and Reactor Laboratory at VTT.

1. INTRODUCTION

The spent nuclear fuel is planned to be disposed in copper canisters in the Finnish crystalline bedrock at a depth of about 500 meters. This research program was planned to provide an experimental evaluation with respect to the theoretical calculations and forecasts for the corrosion behaviour of OFHC coppers in bentonite ground water environments at temperatures between 20- 80

oc

^(Aalto-

nen et al. 1984, KBS 1983). The aim of this study was to evaluate the effects of ground water composition, bentonite and chemical composition of OFHC coppers on the equilibrium and possible corrosion reactions between copper canister and the simulated repository environment at 80 °C. Also transport of corrosion products in the bentonite layer was measured. This final report includes the results obtained after 72 months exposure time.

2. BACKGROUND

The corrosion resistance of copper canisters in the expected repository environ- ment results from the protective corrosion product layer formed on the copper surface. The dissolution rate of this corrosion product layer is expected to be very slow if the pH is high enough (pH > 6) and the amount of oxidizing elements entering to the surface of copper canister remains small. Even the high chloride and sulphur concentrations in the ground water should not increase the solubility of copper and its protective corrosion product layers significantly. It has been assumed that the corrosion rate of OFHC coppers would mainly depend on the rate of transport of the oxygen, sulphide and chloride ions to the canister surface (KBS 1983).

The possibility of localized corrosion reactions, i.e., pitting and stress corrosion cracking of pure copper in the repository environment are usually studied using electrochemical methods or by a slow strain rate testing technique, respectively, and are not of primary concern in this test program.

2.1 Disposal of spent fuel

The planned spent fuel repository will be situated in the bedrock at about 500 meters below the surface. Encapsulation canisters are placed in holes drilled in the floors of tunnels at intervals of 6 meters, Fig. 1. Highly compacted bentonite is used in the hole around the canister. After the disposal operation the open tunnels and shafts of the repository are filled and closed with backfill material, a mixture of quartz sand and bentonite.

3300

Fig. 1. Sealed disposal hole and tunnel.

SAND-BENTONITE BACKFILL

HIGHLY COMPACTED BENTONITE BUFFER

CANISTER

2.2 The chemical environment in the sealed repository

The reactions between copper canister surface and ground water are influenced mainly by the redox-potential (~), pH, dissolved ions and the temperature. The chemical conditions expected in the granitic bedrock at great depths are listed in Table 1.

For the most deep ground waters the natural expected pH values are in the range of 6 -10 corresponding to the carbonate concentration in the range of 100 - 400 mg/1 as well as the alkalizing effects of clay minerals like bentonite (Snellman 1985). In order to maintain low corrosion rate and to avoid local corrosion reactions of copper reducing environments is necessary. In the ground water Fe³+ /Fe²+ redox couple is thought to be the prime redox potential determining reaction. This redox couple has been used to calculate the theoretical ~-values in the water and is expressed by the formula (1) presented by Brotzen (Snellman 1984)

Eh = (0.26 - 0.06 pH) + 0.1 . (1)

The effect of bentonite on the corrosion rate of copper is very important because of its low permeability. Low permeability means low flow rate of water passing the canister and thus also low rate of transportation of oxidizing elements as well as corrosion products to and from the canister surface. In addition highly compac- ted bentonite should maintain alkaline pH which is favourable for the low copper corrosion rate.

Radiation effects on the corrosion are not included in this study, because the wall thickness of the planned copper canister is high enough to provide sufficient self- shielding (Peltonen et al. 1985, KBS 1983). Premature mechanical failure of one canister can, however, produce radiolysis, increase redox-potential and also

corrosion rate by cathodic depolarizers, which will be formed as a result of recombination reactions of radiolysis products.

Table 1. Ground water conditions in granitic bedrock at depths before disposal (Snellman 1984 and 1985).

Natural expected range Saline waters (Snellman 1985) (Snellman 1984)

pH 6- 10

Eh, V -0.44- 0

Cond. m S/m 30- 100 50- 17 000

0, mg/1 < 1

HC0₃-¹¹ 100- 20 000

so~- ¹¹ 100- 400 1-1400

Hs- ¹¹ 0.03- 35

er

^{11 1- 50}

F ¹¹ 0.1 - 1 5- 20 000

HPO~- ¹¹ < 0.1

PO!- ¹¹ 0.01- 0.05

No₃- ¹¹ 0.01 - 0.6

N02- ¹¹ 0.5 - 7.5 < 0.01

Ca2+ ¹¹ 10-40 10- 1 800

Mg2+ ¹¹ 1 - 10 1- 900

K+ ¹¹ 0.02- 5 0.2- 300

Na+ ¹¹ 0.02- 0.2 10- 25 000

Fe2+ ¹¹ 1 - 10 0.02-40

Mn2+ ¹¹ 10- 100 0.02- 7

Si02 ¹¹ < 0.02 2- 300

3. EXPERIMENTAL METHODS

The aim of this work was to evaluate the interactions between the canister material, the bentonite filler material and the near field ground water in the repository.

The geochemical and hydrothermal conditions expected to exist in the repository were simulated by using an immersion test, where copper canisters filled with bentonite were exposed to various ground waters. The tests were designed to include the effects of the bentonite chemical composition, the bentonite swelling pressure, the temperature and the chemical composition of the ground water.

The swelling pressure against copper canister surface corresponding to the repository conditions was achieved by filling the copper canisters with bentonite clay, Fig. 2. Filling was made using mechanical pressing to original clay dry density of about 1.8 - 1.9 g/cm³•

Water bath 80 °C

PTFE-container --+---+- Ground water

Copper canister

Bentonite

Fig. 2. Test facility used to study corrosion of OFHC coppers in simulated repository conditions.

The stable chemical environment around the test canisters was obtained by closing the copper canisters inside containers filled with a specified ground water. On the cover of each copper canister there was a porous plug through which ground water wetted the bentonite. The containers immersed to a water bath at the temperature of 80

oc.

The container material was in the beginning high density polyethylene (hd-PE) which was changed to polytetrafluoroethylene (PTFE) after 9 months exposure because of the embrittlement of the hd-PE container covers.

3.1 Immersion test environments

The ground waters used in this study were based on the simulated ground water called MS-water. The MS-water was modified with chloride and sulphate addi- tions, Table 2.

Table 2. Contents (mg/1) of the most important species in the three used variations of simulated ground waters (Snellman 1986).

Environment so²₄ - No₃- N0₂-

er

^Hco-₃ ^cu+ Conduc- pH tivity

(mS/cm)

MS-water 59 60 570 0.00 1 8.4

MS + Cl- water 62 7600 600 0.00 19.7 7.9

MS+ So~-water 6200 62 590 0.00 9.2 8.2

The amount of dissolved oxygen in the ground water was not controlled during the tests. Due to the test arrangements the ground water was oxygen saturated in the beginning. However, the limited access of oxygen through container walls and the reduced solubility of oxygen in the water bath at 80

oc

reduced oxygen

content in the ground waters. Additionally the oxygen consumption by corrosion reactions decreased oxygen content inside test containers.

3.2 Canister materials

The immersion tests were started using only an OFHC copper as the canister material. However, later additional tests with canisters made of silver alloyed (0.1

% Ag) copper in MS-water+Ct were also initiated. The silver alloyed copper was considered as an alternative canister material for pure OFHC coppers, providing higher strength and thus better long term creep resistance at temperatures below 100 °C.

3.3 Evaluation of chemical reactions during immersion periods

Test containers with copper canisters were removed from the bath5 according to the test schedule presented in Table 3. Mter opening the containers, the pH, the conductivity and the chemical compositions of the ground water inside the con- tainers were analyzed.

The chemical compositions of ground waters were analyzed by an ion chromatographic technique. The dissolved copper in the ground water after the immersion tests was analyzed by an atom absorption spectrometric technique.

The pH and dissolved iron and copper contents in the bentonite were analyzed after sawing copper canisters into five sections, Fig. 3. The contents of dissolved Fe²+ and Cu²+ ions in the bentonite were analyzed by chemical methods. Cu²+

content in bentonite was also analyzed in some samples taken at the distance of 2.5, 7.5 and 12.5 mm from the canister wall.

The corrosion products formed in the MS-water+Cr ground water observed in the containers after the exposure of 12 months have been identified using ESCA

method. Based on the visual inspections the amount of localized corrosion reaction such as pitting corrosion on the canister surface was estimated.

Table 3. The immersion test matrix and related specimen codes. Tests were conducted for OFHC copper in various ground waters and for silver alloyed (0.1%

Ag) copper in MS-water+Cr environment.

OFHC copper

Test period (months)

Environment 3 6 9 12 24 36 72

MS-water Al A2 A3 A4 A5 A6 A7

MS-water + er Bl B2 B3 B4 B5 B6 B7

MS-water + SO/- Cl C2 C3 C4 C5 C6 C7

Silver alloyed (0.1 % Ag) copper in MS-water+C[ environment.

Test period (months)

Test material 6 12 24

OFHC copper Dl D2 D3

0.1 % Ag copper El E2 E3

10 % deformed

0.1 % Ag copper Fl F2 F3

50 % deformed

Fig. 3. Sectioning of the copper canisters for analysis.

The redox-potential (B.t) is very difficult to measure electrochemically after the immersion tests since atmospheric oxygen or carbon dioxide may significantly affect the electrochemically measured

&..

However, there are two alternative indirect methods for estimation of

E.t

values inside canisters after the immersion tests:

the measurement of the dissolved redox couple ratios such as SO/-ms- or Fe³+ /Fe²+ and

the estimates based on the analyzed mineral assemblages.

The former indirect method was applied in this study to evaluate the changes taking place in the redox-environment. In addition to the indirect redox indica- tions by using Fe3+/Fe²⁺ ratio

E..

was estimated by using experimental formula presented by Brotzen (1), which gives the

E..

as a function of pH.

4. TEST RESULTS

After 9 months exposure cracks were observed on the covers of some hd-PE containers. All containers were changed to more durable containers made of PTFE. Since both containers had the same volume, it was not necessary to add any water. However, the air leakage and the exposure of ground water to open air was affecting mainly the redox-potentials measured after 9 and 12 months expo- sure.

4.1 Ground water analysis

The pH values of all ground waters have remained within the +0.5 pH unit, during the first 9 months period, Fig. 4a. Between 9 and 12 months period the air leakage has increased redox-potential and thus decreased the measured pH values slightly, but after 24 months pH values have stabilized again. In the MS +

er

ground water the initial pH increased during the first 3 months and decreased after that obtaining the pH value of 7.5 after 12 months. Later the pH in the MS +

er

ground water increased slowly and stabilized after 24 months. The pH in all ground waters increased slowly and obtained after 72 months the value between 9 and 9.5.

The conductivity is a measure of dissolved ions in the ground water. Formation of solid species will change the conductivity of the ground water during the test.

The initial high conductivity of the MS +

er

ground water has first decreased during the 12 months test period obtaining about the same conductivity value, which was measured in the MS + S0₄^2- ground water, Fig. 4b. After 24 months the conductivity of MS+

er

ground water started to increase again. In all ground waters conductivity has increased slightly during the 72 months test period. The originally measured differences in the conductivity values of different ground waters has remained.

10

a)

- 16 E u

-

^(/) _-^E ₁₂

>-

1-

>

1- 8

u

:::l 0 :z

8

^'+

0

0 MS+water A MS+Cl--water 0 MS+so3--water

TIME (months)

0 MS-water A MS+CI--water 0 MS+SC 3--water .

---MS---0~

--- ~---:/

. MS+so3- ~/

0~--~--~--~--~--~--~--~~~~--~--~--~~~~

b) TIME (months)

Fig. 4. a) The pH of the MS ground waters during 72 months test period.

b) The conductivity of the MS ground waters during 72 months test period.

The pH and conductivity of

er

modified MS-water after 24 months exposure in contact with the silver alloyed coppers are presented in Fig. 5. The variation in these parameters corresponds to that observed for the OFHC copper. The pH increased slowly in all containers and obtained after 24 months exposure the values between 10 and 12. The measured conductivity values varied in different containers during the test and all measured values started to increase after 12 months exposure.

a)

b)

E u

-

^(/)

11

J: 0..

24

20

E 16

>- ..._

> 12 ..._

u :J 0 z u 0

8

3

0 OFHC-copper

~ 0. 1% Ag-copper;

1 O% deformation 0 0. 1% Ag-copper;

50% deformation

TIME (months)

0 OFHC-copper ll. 0.1% Ag-copper;

1 o% deformation

0 0.1% · Ag-copper;

50% deformation

3

TIME (months)

Fig. 5. a) The pH of the MS+Cr ground water during 24 months immersion test period in contact with the OFHC and silver alloyed coppers. b) The conductivity of the MS+Cr ground water during 24 months immersion test period in contact with the OFHC and silver alloyed coppers.

The chemical compositions of the ground waters during 72 months exposure are presented in Fig. 6. During the first months an increase in the dissolved copper (eu+) and nitrate (N0_3-) dissolving from the bentonite concentrations was observed which, however, levelled of with the time. The observed increase in the concentrations of these ions was highest in the MS + er ground water. The analyzed concentrations of dissolved copper (eu+) and nitrate (N0₃-) were in all analyzed ground water samples after 24 months exceptionally high. There exists no reasonable explanation for these systematically high measured concentrations.

The sulphate (SO/-) contents of the MS and MS +er ground waters increased from the initial value of 50 mg/1 up to about 500 mg/1. However, the sulphate content in the MS + SO/- water which originally was high, 6200 mg/1, has decreased in the beginning of the test obtaining after 12 months exposure a value of about 5000 mg/1. The chloride (Cr) contents in the MS and MS+ S0₄^2-ground waters have remained at about the initial level during the whole 72 months test.

The initial high er content in the MS + er ground water varied during the 9 first months obtaining the lowest value of 4300 mg/1. This reduction is obviously in connection to the air leakage to the test environment and following precipitation reactions consuming dissolved chloride ions. Mter that the chloride content has remained near to the initial value which was about 7 600 mg/1.

a)

10000 I I I I I 1 1 1 1 1 I A

I 1-y I I

~ -

=

^1000~ ....,o-

---0--o---o---o~

o~O

~

^{100,_ / cl- /o_}

~ 9--o---o--o--o---o---o~

z UJ

u z

0 u

0 3 6 9 12

(OJCu+

(OJN03

I I I

15 18 21 24 27 30

TIME (months)

/~

-

33 36 ~ ii.

Fig. 6. a) The chemical composition of the MS ground water during 72 months exposure at 80 oe in contact with a OFHe copper canister containing bentonite.

-

^0" E

z 0 1-<

a:: 1-

z UJ

u z u 0

b)

z 0 1-<

~

1-z

UJ u z 0

u

c)

0

(.6 I cu+

( .6) N03

3 6 9 12 15 18 21 ~4 27 30 33

TIME (months)

1---o---

(O) N03

( o J cu+

Fig. 6. b) The chemical composition of the MS +

er

ground water and c) the chemical composition of the MS + S0₄ ²- water during 72 months exposure at 80

oc

in contact with a OFHC copper canister containing bentonite.

The chemical compositions of the MS+Cr ground water during 24 months immer- sion test in contact with the OFHC and silver alloyed coppers are presented in Fig. 7. The chemical composition in all containers independent of the silver alloying or the degree of deformation changed in the same way during the immersion tests. Chloride concentrations decreased slightly but sulphate and nitrate concentrations dissolving from the bentonite clay increased during 24 months test. In general the amount of dissolved copper in the ground water was below the detection limit except in the container with OFHC canister which was taken out after 6 months immersion.

a)

- 1000

0 z

._ 100

c:(

0::::

._

z

UJ u

0 z

u ¹⁰

OFHC-copper

TIME (months)

Fig. 7. a) The chemical composition of the MS +

er

water during 24 months exposure at 80

oc

in contact with a OFHC copper canister containing bentonite.

b)

=

¹⁰⁰⁰

-

^O'l _E

z 0 .... 100 _~

0:::

....

z w u z 0 u 10

-

^{O'l E}

z 0

....

~ 0:::

....

z w u z 0 u

1000

100

10

0.1% Ag-copper, 1 O% deformation

TIME (months)

0. 1% Ag-copper, 50% deformation

c) TIME

Fig. 7. b) The chemical composition of the MS + Ct water during 24 months exposure at 80

oc

in contact with a 10% deformed silver alloyed copper canister containing bentonite and c) in contact with a 50 % deformed silver alloyed copper canister containing bentonite.

pH 11

10

9

8

4.2 Bentonite analysis

The pH and iron contents in the bentonite were measured in order to evaluate redox-equilibrium obtained in the interactions of copper, bentonite and ground water. The measured Fe³+/Fe²+ ratios, the measured pH of bentonite and the calculated E,. values during the 72 months exposure are presented in Fig. 8.

During the first 9 months the oxidation of copper consumed the available oxygen and the measured Fe3+ /Fe²+ ratio in all ground water environments decreased from the initial value of about 2 down to the value below about 1. The aeration of the test environment after 9 months exposure increased the Fe³+ /Fe²+ ratio in the MS and MS +

er

ground water environments up above 1 which was not, however, observed in the MS + SO/- environment. Mter 24 months test period Fe³+/Fe²+ ratios in all ground waters have increased and exceeded the initial values after 72 months.

EH (mV)

-310

-320

-330

-340 -350

-360 -370

a) Fig. 8.

0 - 0 ·

pH

~---o---~

Fe3+JFeZ+

36 72

a) The measured pH and Fe3+/Fe²+ ratio in the bentonite and the calcu- lated E,. during 72 months exposure of OFHC copper in the MS water.

6

5

pH

11

pH

11

8 EH

(m V)

-310 ⁶

-320 5

-330

6~6

⁴

-340 ⁶ 3

6'--6/

-350 2

-360

-370 0

b) (months)

EH

(mV)

-310 6

-320 5

-330

---0---

pH -340

-350

-360

-370

W-~~_j~~~~~~~~~~~~L-~0

_{15 18 21}

c) TIME (months)

Fig. 8. b) The measured pH and Fe3+ /Fe²+ ratio in the bentonite and the calcu- lated

E..

during 72 months exposure of OFHC copper in the MS+Cr water and c) in the MS+ SO ₄ ²- water.

The measured pH of bentonite in all environments increased with about one pH unit obtaining after 12 months the pH value higher than 10. Correspondingly the calculated

E..

value decreased in all environments and the redox-equilibrium potential obtained the value of about -360 mVsHE· The measured pH values in bentonite have decreased slowly after initial fast changes and correspondingly the calculated

E..

values have increased as a consequence of pH changes.

The measured pH and Fe3+/Fe²⁺ ratio in the bentonite which have been in contact with the silver alloyed copper canisters repeated the same changes as has been measured with OFHC copper canisters. Initially the bentonite pH increased and correspondingly the calculated E,. decreased, Fig. 9. Also the measured ratio of Fe3+ and Fe²⁺ indicated the same trends in the redox-potential. After 12 months immersion the measured pH started to decrease and correspondingly the calculated E,. increased. Also iron ratio indicates the increase of the redox-potential.

12 -340

> _E -350 ..r:.

11 ^LL.J -360

I _a. -370

10 -380

-390

9

a)

OFHC-copper

D ^D

X ^o..._

^D

TIME (months)

9 8 7 6

Fig. 9. a) The measured pH and Fe3+ /Fe²+ ratio in the bentonite and the calcu- lated E,. during 24 months exposure of OFHC copper in the MS+Cr water.

b)

I a.

:I: a.

11

12

10

9

-320

-330

-320

-330

-340

> _E

-350

-370

-380

-390 0

0. 1% Ag-copper; 10% deformation

3

TIME (months)

0.1% Ag-copper, 50% deformation

0---0

c) TIME (months)

7

6 +

N

5 ~

-

4t,

3 2

9 8 7 6

3 2

u.. (1)

Fig. 9. b) The measured pH and Fe3+/Fe²+ ratio in the bentonite and the calcu- lated ~ during 24 months exposure of silver alloyed copper with 10 % deformation in the MS+Cr water and c) of silver alloyed copper with 50

% deformation in the MS+Cr water.

The distribution of dissolved copper in the bentonite showed initially after 6 months immersion a clear decreasing concentration gradient, Fig. 10. However, after 24 months immersion the observed gradient has disappeared and the concen- tration in samples taken close the canister wall have the same copper concentra- tion as the specimens taken from the middle of canister, Fig. 10.

o\0

c.. I

z 0

~

<(

a:= ~

z w u z u 0

I :::::1

u

a)

ciP

c.. I

0 z

~

<(

a:= ~

z w u z u 0

I :::::1

u

b)

0,6 0,4 0,2 0,0

1, 2 1,0

0,8

0,6 0,4

0,2 0,0

OFHC-copper

o ^{6 months}

A 12 ¹¹ 0 24 ¹¹

2,5 7,5 12,5

DISTANCE FROM THE CANISTER WALL (mm)

0.1% Ag-copper 10% deformation

o 6 months .6. 12 ¹¹ 0 24 ¹¹

H~

0 0

2,5 7,5 12,5

DISTANCE FROM THE CANISTER WALL (mm)

Fig. 10. a) The content of dissolved copper in the bentonite as a function of the distance from the canister wall during 24 months exposure of OFHC copper in the MS+Cr ground water at 80 °C. b) The content of dis- solved copper in the bentonite as a function of the distance from the canister wall during 24 months exposure of 10 % deformed silver alloyed copper in the MS+Cr ground water at 80 °C.

I I T

o\O 1, 2 .... 0,1% Ag-copper

I 50% deformation -

0.

o 6 months

z _{1,0 f- 6. 12} -

0 ¹¹

1- ⁰ 24 ¹¹

<( 0,8- -

0::

1-z UJ _0,6- -

u 0

z

~~,~~

0 0,4- -

u I :::J

u ^{0, 2-} -

9 9- ^'T'

0,0 2,5 7,5 12,5

DISTANCE FROM THE CANISTER WALL (mm)

Fig. 10. c) The content of dissolved copper in the bentonite as a function of the distance from the canister wall during 24 months exposure of 50 % deformed silver alloyed copper in the MS+Cr ground water at 80 °C.

4.3 Surface analysis

A thin, light brown oxide layer was visible on the surface of all canisters. There are some small differences visible in the colour of surfaces depending on the environment. Especially in the MS+Cr ground water on the canister surface there were deposited green precipitates which obviously are copper carbonate (CuC0₃₎ compounds. In general the oxide layers were adherent both on the outer surface as well as on the inner surface of canisters and no pitting or intergranular cor- rosion was observed.

The corrosion product compounds on OFHC and silver alloyed coppers exposed to MS+Cr ground water were identified using ESCA. The corrosion products con- tained equal amounts of CuO, Cu₂0 and CuC0_3• Also some minor amounts of CuF₂ was observed. In Fig. 11 is the Pourbaix equilibrium diagram for the system including Cu-Cl-C0₂-S0₃-H₂0 at 25 °C. The shaded area indicates the corrosion potential-pH field experienced by copper surface during immersion tests.

2.8

2.4

Cu203

2.0

1. 6

CuSO,

1. 2

@_

-

^..._

cu++ .._

- _>

0.8

^CuCl

s.. 0.4

s.. 0

0-

u 0

w

--

^..._

_-

^..._

-0.4 ^..._

-

^..._

_- --

-0.8 -1.2

Cu

-1.6

-2 0 2 4 -6 _{8 10} 12 14 16

pH

Fig. 11. Potential-pH equilibrium diagram for the system including Cu-Cl-C02-

S03-H20 at 25 °C (Pourbaix 1973).

5. DISCUSSION

The corrosion rate of OFHC coppers in the repository conditions depends on the rate of transport of reacting ions to the canister surface and the transport of corrosion products from the canister surface (KBS 1983).

The results obtained during 72 months immersion tests showed that the test set up used in this study contained three different barriers controlling oxygen access to the canister surfaces. The first barrier was the water bath kept at 80 °C; water dissolves only about 3 ppm oxygen at 80

oc

instead of the 8 ppm at room temperature. The second barrier was formed by the container walls which decreased the ingress of oxygen from the water bath to the container. The oxygen content in the containers was smaller than that in the water bath because copper corrosion reactions consumed oxygen. The third oxygen barrier in the test set up was the porous sinter in the cover of each copper canister. Based on the estimated redox-potentials the oxygen content inside the canisters has been lower than in the containers.

In the beginning of the test metal surfaces were not oxidized, ground waters were air saturated and the bentonite was not wetted. This experimental study was planned to reveal the kinetics of reactions taking place near the canister surface after disposal. Thus the variations in reactions between different ground waters and different canister materials were the primary interests.

5.1 Ground water in the containers

pH and conductivity in the ground waters

The pH increased slowly during the test in all ground waters and obtained after 72 months test period the value of about 9. The observed shift in the pH to

alkaline direction was caused by the interaction between the ground water and the bentonite inside the canisters.

The conductivity of MS-water increased continuously during the test. The conductivities of the MS+Cr and MS+SO/- ground waters were changing as a function of the redox-potential. Decreasing redox-potential decreases the solubility of er and

so/-

containing substances and thus the conductivity of the ground water.

Dissolved ions in the ground waters

The amount of dissolved copper measured during the test was highest in the container filled with the MS+Cr water after 12 months exposure. This is in agreement with the data presented in the literature predicting high solubility of copper-chloride compounds. The total concentration of dissolved copper in equilibrium with solid copper should be in the order of 10-³ mg/1 when the redox- potential, ~' is about -350 mVsHE (KBS 1983). This value is about three orders of magnitude smaller than the values measured after 12 months in this investiga- tion in samples taken from the ground waters that had been inside containers.

After 36 months exposure the dissolved copper concentration in the MS+Cr ground water was about 0.3 mg/1 and the amount of dissolved copper remained about the same after 72 months exposure. The copper concentration in the range of 1···10 mg/1 corresponds to the redox potential between -50··· -100 mVsHE(KBS 1983). These values give a rough estimation about the redox-potential in con- tainers during the tests. The observed variation in the measured copper concentra- tion shows that initially there was high oxygen content in the ground water and also a unoxidized, "active" copper surface. However, the dissolution of copper and formation of passivating corrosion films consuming the oxygen available inside containers proceeds simultaneously. After 12 months the amount of dissolved copper starts to decrease because of the increased pH and retarded diffusion of additional oxygen through the container walls and decreased redox-potential.

Later, after 24 months exposure, the redox-potential inside containers starts to increase as a consequence of the reduced corrosion rate and the decreased oxygen consumption. Apparently the oxygen diffusion through the container walls remained constant.

5.2 Ground water wetted bentonite inside copper canisters

The test results covering 72 months exposure confirmed the anticipated increase in the pH of ground water in contact with the bentonite and the corresponding decrease in the redox-potential, Eh, as a result of corrosion reactions between the ground water and copper.

The measured pH in the ground water wetted bentonite was stable during the whole test period in all studied canisters within ± 0.5 pH unit. The measured pH values were between 9.0 and 10.5 depending on the ground water composition.

The calculated redox-potential in the wetted bentonite inside canisters varied as a function of the measured pH values between -310 ... -370 m V sHE· The interac- tion of pH and redox-potential was also confirmed by the measured Fe3+/Fe²⁺ ratio in the wetted bentonite.

The disturbance in the test environment oxygen equilibrium, produced by the change of containers after 9 months exposure, was also visible in the analysis concerning wetted bentonite as an increase in the ~ values. This increase demon- strates how sensitively ~ reactions can be observed through the measured pH values of bentonite and through the Fe3+/Fe²+ ratio. This ratio showed similar trends as the calculated redox-potential although the changes in the redox-poten- tial became visible more slowly. The direction and the magnitude of changes in the measured pH, calculated E,. and iron ratios were in accordance with each others. The pH in wetted bentonite was in these tests more sensitive indicator of changes taking place in the redox-environment than the Fe3+/Fe²+ ratio. The reason

for this is the used test arrangement, where copper surface area/bentonite volume was much larger than will be in real repository, emphasizing the importance of copper oxidation reactions in the control of redox-potential.

5.3 Corrosion products

The solid compounds found inside containers after the tests were identified by ESCA analysis. Corrosion products contained CuO, Cu₂0 and CuC0_3• The compositions of the corrosion products reflect the pH/redox-potential range experienced by the canister surfaces during the tests. The initial air saturated ground water and near neutral pH produced CuC0_3• Later the increasing pH made the formation of CuO possible and the equilibrium between CuO and C~O was obtained because of the reducing environment, i.e., shortage of oxygen.

The relative content of copper in the bentonite, expressed in weight-%, indicates high dissolution rate of copper in the beginning of the exposures. However, after 24 months the relative copper content has increased throughout in the bentonite and no concentration peaks were measurable near the canister surface. The decreasing corrosion and the drift of dissolved ions in the wetted bentonite levelled off the observed concentration peak.

The redox-potential after 72 months in the containers (E..::; -50 m Vs~ as well as in the canisters (E..::; -310 m V sHE) is low enough in order to avoid localized cor- rosion types of copper, i.e., pitting corrosion induced by the presence of chloride ions as well as stress corrosion cracking.

The lacking of localized corrosion during the tests can be explained by comparing the estimated corrosion potential of copper and the equilibrium pitting corrosion potential for copper in the chloride containing solutions. The corrosion potential of copper can be estimated on the basis of the pH and the corrosion products, CuO and Cu₂0, observed in the containers after tests, Fig. 12. The equilibrium

potential between CuO and C11zO is lower than +210 m V ^sHE always if the pH~ 8.

Also the redox-potentials in containers (E..~ -50 mV8^~ and inside the canisters

(E..~ -300 m V sHE) indicated that the possible changes in the corrosion potential of copper will be toward more negative potentials. At the bottom of a copper pit the phases Cu, C11zO and CuCl coexist in contact with the

er

-solution. The calculated coordinates for this point in the Pourbaix diagram are E = +270 m V ^sHE and pH=

3.5 representing the electrochemical equilibrium inside a copper pit (Pourbaix 1973). Pitting corrosion can take place only if the environment is oxidizing enough to polarize the potential inside the pit to more oxidizing direction and thus increase the copper dissolution as Cu²+ in the pit. However, if the corrosion potential of copper outside the pits is below the equilibrium pit potential, pitting is not possible. As a consequence there is a "protection potential" below which existing pits do not grow. In practice this "protection potential" is even higher than the equilibrium potential inside pit due to effects of concentration polariz- ation and the ohmic drop that must be overcome to polarize the pit internals. The corrosion potential of copper in all test environments during the 72 months test has remained well below the equilibrium potential of copper pits and thus no pitting corrosion was observed.

6. CONCLUSIONS

The used range of variations in test parameters such as in the chemical composi- tion of the ground water, chemical composition of OFHC coppers or the degree of deformation of canister material did not change markedly the reactions of copper surfaces with the environment in the simulated repository environments.

The most important parameters regarding the corrosion resistance of canister materials seems to be the rate of reagent transport to the canister surface and the transport of corrosion products from the canister surface. The changes observed in the pH or in the redox-potential during the immersion tests did not change the corrosion reactions.

7. REFERENCES

Aaltonen, P., Hanninen, H., Klemetti, K. & Muttilainen, E. 1984. Corrosion of copper in nuclear waste· repository conditions - A literature review (in Finnish).

Helsinki, Ydinjatetoimikunta, Report YJT-84-17. 62 p.

Asklund, A. M., Grundulis, V. & Ronnholm, B. 1966. Vatkemisk analys av silikatbergarter. Stockholm.

Holopainen, P., Pirhonen, V. & Snellman, M. 1984. Crushed aggregate-bentonite mixtures as backfill material for the Finnish repositories of low- and intermediate- level radioactive wastes. Helsinki, Y dinjatetoimikunta, Report YJT -84-07. 64 p.

KBS 1983. The Swedish Corrosion Research Institute and its reference group.

Corrosion resistance of a copper canister for spent nuclear fuel. Stockholm. KBS TR 83-24.

Peltonen, E., Vuori, S., Anttila, M., Hillebrand, K., Meling, K., Rasilainen, K., Salminen, P., Suolanen, V., Winberg, M. 1985. Safety analysis of disposal of spent nuclear fuel - normal and disturbed evolution scenarios (in Finnish).

Helsinki, Ydinjatetoimikunta, Report YJT-85-22. 309 p.

Peltonen, E. et al. 1986. Concept and safety assessment for final disposal of spent nuclear fuel in Finland. International Symposium on Siting, Design and Construc- tion of Underground Repositories for Radioactive Wastes. Hannover, 3- 7 March,

1986. IAEA-SM-289/28.

Pourbaix, M. 1973. Lectures on Electrochemical Corrosion. Plenum Press, New York.

Snellman, M. 1982. Ground water chemistry in bedrock. Helsinki, October 1982, Ydinjatetoimikunta, Report YJT-82-42. 65 p.

Snellman, M. 1984. Chemical conditions in a repository for spent fuel. Helsinki, January 1984, Ydinjatetoimikunta, Report YJT-84-08. 98 p.

Snellman, M. 1985. Kaytetyista polttoaineesta vapautuvien aktinidien (Pu, Np, Th), fissiotuotteiden (Cs, I) seka kapselin korroosiotuotteiden liuokoisuudet loppusijoitustilan kemiallisissa olosuhteissa. Helsinki, Y dinjatetoimikunta, Raportti YJT-85-15.

Snellman M. 1986. Chemical conditions in a repository, Finnish - German Seminar on Nuclear Waste Management. Espoo, September 23 - 25, 1986.

metaa/rap2/538k.paa

LIST OF YJT REPORTS PUBLISHED IN 1993 YJT-93-01

YJT-93-02

YJT-93-03

YJT-93-04

YJT-93-05

TVO-flowmeter Pekka Rouhiainen PRG-Tec Oy January 1993

Ore potential of basic rocks in Finland

Jouni Reino, Markus Ekberg, Pertti Heinonen, Tapio Karppanen, Antero Hakapaa, Esa Sandberg

Outokumpu Mining Services Oy February 1993

(in Finnish)

Long-term durability experiments with concrete-based waste packages in simulated repository conditions

Ari Ipatti

Imatran Voima Oy March 1993

The dissolution of unirradiated U0₂ fuel pellets under simulated disposal conditions

Kaija Ollila, Hilkka Leino-Forsman Technical Research Centre of Finland Reactor Laboratory

March 1993

Long term corrosion tests of OFHC-coppers in simulated repository conditions - final report

Pertti Aaltonen, Paivi V aris

Technical Research Centre of Finland Metals Laboratory

March 1993