Sie sind auf Seite 1von 7

ELS EVI ER

Nuclear Engineering and Design 168 (1997) 183-189


N u c le a r
E n g in e e d .n g
a n d D esngn
Behaviour of reference electrodes in the monitoring of corrosion
potential at high temperature
M. Navas, M. D. Gomez Bricefio
CIEMA T. Avda Complutense-22, 28040 Madrid, Spain
Abstract
The main problem of corrosion potential measurements is the necessity to develop a reliable high temperature
reference electrode. The aim of this work was to study the behaviour of three types of high temperature reference
electrodes that have been developed for use in a BWR type reactor operating conditions. The reference electrodes
were copper/copper(I) oxide with yttria-stabilized zirconia (YSZ) membrane; a silver/silver chloride couple and
platinum. The application of these electrodes to corrosion potential monitoring is illustrated with several examples in
different environments. The thermodynamic behaviour and stability of these sensors have been studied. Correlations
between corrosion potential and water chemistry have been determined. 1997 Elsevier Science S.A.
1. I nt r oduc t i on
Cor r osi on damage of nucl ear mat eri al s in hi gh
t emper at ur e wat er envi r onment s can be fr equent l y
rel at ed t o wat er chemistry. The i mpor t ance of t he
specific envi r onment and t he associ at ed oxi di zi ng
condi t i ons on cor r osi on has l ong been recognized.
Cor r osi on pot ent i al of a met al or al l oy exposed t o
t hat aqueous sol ut i on is a key par amet er f or
det ermi ni ng t he mat eri al cor r osi on behavi our.
In maki ng el ect rochemi cal pot ent i al (ECP)
measur ement s at high t emperat ures, t he mai n
difficulty is in devel opi ng a reliable reference elec-
t rode. Design of high t emper at ur e reference elec-
t rodes was very i mpor t ant dur i ng '80s, and
reference el ect rodes of many t ypes and designs
were devel oped, but each of t hem still has limita-
tions in t hei r uses. So, great care is requi red t o
assure t hat valid and useful dat a are bei ng ob-
tained.
The requi rement s, t o obt ai n reliable dat a wi t h a
reference el ect rode, are: el ect rode pot ent i al inde-
pendent of t he envi ronment ; chemi cal and t her-
modynami c stability; reproduci bl e measurement s;
el ect rode pot ent i al knowl edge on a t her mody-
nami c scale such as st andar d hydr ogen el ect rode
(SHE) and so on. Par t i cul ar emphasi s has been
pl aced on t he i mpr ovement s of system integrity,
reliability and mat eri al s stability.
In this st udy t hree t ypes of reference electrodes
have been fabr i cat ed wi t h some modi fi cat i ons and
i mpr ovement s with respect t o t he el ect rodes f ound
in l i t erat ure dat a, as follows.
(1) Ag/AgC1 el ect rode is a silver wire with
el ect rodeposi t ed silver chl ori de, encl osed in a solid
Rul on chamber , filled wi t h demi neral i zed wat er
(Fig. 1). Some silver chl ori de crystals are added in
or der t o avoi d changes in silver chl ori de solubility
with t emperat ure. As l ong as t he solubility of
silver chl ori de is known, t he pot ent i al of the refer-
0029-5493/97/$17.00 1997 Elsevier Science S.A. All rights reserved.
PII S0029-5493(96)01 355-6
184 M. Navas, M.D. Gornez Brice~o / Nuclear Engineering and Design 168 (1997) 183 189
ence el ect rode is easily cal cul at ed in a SHE scale,
usi ng t he Ner ns t equat i on.
(2) Cu/ Cu20 el ect rode consists of a yt t ri a-st abi -
lized zi rconi a (YSZ) cl osed-end t ube, filled wi t h an
i nt ernal j unct i on of copper / cupr ous oxi de solid
mi xt ur e (Fig. 2). It can serve as a membr ane t ype
el ect rode because t here is a di rect rel at i onshi p
bet ween t he conduct i on of oxygen ions t hr ough t he
cerami c and t he pH in t he adj acent aqueous phase
(Ni edrach, 1982). In this case, it woul d wor k as a
membr ane- t ype pH sensor and when t he pH of t he
system is known and const ant , it woul d act as a
reference el ect rode.
(3) Pl at i num wire can be used as a reference, onl y
when t here is a st oi chi omet ri c excess of dissolved
hydr ogen concent r at i on. Then this sensor acts as a
reversible hydr ogen el ect rode and when pH and
TEFLON
. E$3
~OOY
STAINLESS
STEEL BOO~
AUTOCLAVE LID
RULQN
SILVER
' W I R E
i , , I ~ CI A41
ECTRO(
Fig. 1. Ag/AgCl reference electrode.
-'R
~LASS
'/WOOL
=- c~zo
POWDER
CERAMIC
MEMORANE
Fig. 2. Cu/Cu20 reference electrode.
hydr ogen part i al pressure are const ant s it can be
used as a reference electrode.
2. Experimental
The ECP measur ement s and vari ous cor r osi on
tests were per f or med in a testing facility, which
simulates BWR condi t i ons wi t h high pr e s s ur e - -
high t emper at ur e wat er (80 kg c m- 2 and 280C).
The test assembly includes a one-l i t t er stainless steel
aut ocl ave in which f our el ect rodes can be fitted.
M. Navas, M.D. Gomez Bricego /Nuclear Engineering and Design 168 (1997) 183-189 185
Tabl e l
Chemi s t r y resul t s
Gas concent r at i on Mat er i al s I mpur i t i es (ppb) a (f) 6t S c m- i ) (e) 6t S c m- t ) [02] (e) T (C)
200 ppb O~ l nconel 600 0.10 0.29 114 277
AISI 316NG
125 ppb H2+ 20 AISI 304
ppb 02 Inconel 600
20 Cu 2+ 0.36 130 277
500 C1- 1.7 2.3 96 279
20 SOl + 0.13 0.28 180 274
20 CI
100 SO4- 0.33 0.43 150 273
0.07 0.26 2 282
20 Cu 2+ + 0.34 279
20 C1
20 C1- 0.13 0.45 ...... 278
25 SO ] - 0.14 0.33 5 275
25 SO] 0.15 0.41 2 279
100 S O] - 0.35 0.40 5 275
(f) is experimental dat a in the feedwater; (e) is experimental dat a in the effluent water.
Labor at or y studies have been carri ed out under
t wo si mul at ed baseline envi ronment s: nor mal wa-
t er chemi st ry (NWC): 200 ppb O2; and hydr ogen
wat er chemi st ry (HWC): 125 ppb H 2 + 20 ppb O2.
Cor r osi on pot ent i al measur ement s have been
carri ed out f or di fferent materials: t ype AISI 304
austenitic stainless steel; t ype 316NG austenitic
stainless steel and nickel base alloy, Inconel 600.
The cor r osi on pot ent i al s were conver t ed t o SHE
scale by addi ng t he cor r ect i on f act or f or each
electrode. Compar i sons bet ween t he cor r osi on po-
tentials, obt ai ned with t he di fferent electrodes,
were made in or der t o st udy t hei r reliability and
t o confi rm t he effect of impurities on cor r osi on
pot ent i al s. In addi t i on, el ect rode pot ent i al s were
verified using a pl at i num el ect rode under HWC.
Such measur ement s were used t o assess t he viabil-
ity and behavi our of these sensors, compar i ng
t heoret i cal and experi ment al values.
Mor eover , di fferent types of impurities and
concent r at i ons were added t o eval uat e t he effect
on t he mat eri al cor r osi on based on cor r osi on tests
and ECP measurement s. Chemi cal additives used
in this st udy include Na2SO4, C1Na and CuC12.
Concent r at i ons and chemi cal condi t i ons are pre-
sented in Tabl e 1. Dur i ng a run, conduct i vi t y,
level of dissolved oxygen and t emper at ur e were
cont i nuousl y moni t or ed in t he feed and t he out l et
waters. It must be poi nt ed out t hat t he addi t i on of
impurities, within t he ppb range, di d not change
t he sol ut i on pH and t he var i at i on of wat er con-
duct i vi t y was f ound t o be no mor e t han 1 laS
1
cm
3. Re s ul t s and di s cus s i on
3.1. Platinum electrode
Thi s el ect rode showed a reversible behavi our in
chemi st ry with hydr ogen addi t i on. However ,
some irreversibility has been observed wi t h high
levels of oxygen or with el ect roact i ve species
which can be r educed mor e easily t han prot ons.
So it is necessary t o use anot her high t emper at ur e
reference el ect rode which is insensitive t o r edox
changes.
3.2. Silver electrode
In t he first el ect rode design t here was a final
chamber in or der t o avoi d possible chl ori de con-
t ami nat i ons of t he sol ut i on. However , it has been
pr oved t hat t he above i nt r oduced a liquid j unc-
t i on pot ent i al drop. So, in t he last measurement s
under NWC, this chamber was el i mi nat ed and the
186 M. Navas, M.D. Gomez Brice~o / Nuclear Engineering and Design 168 (1997) 183-189
2 0 0
100
-100
u J
I - 2 O O
C9
>
E -aoo
13_
( . ~ - 4 0 0
LU
~ E c o r r A I S I 3(~4 V S P t
i i i i i I ~ E c o r r A I S I 3 0 4 v s C u
~ - E c o r r A I S I 3 0 4 V $ A g
i
~ E A g e l e c t r o d e V S P t
" ,,,~, H~ i n c r e a s e
- 6 0 0 . . . . . . .
- 7 0 0 I , I ~ I , ) ~ I , I i [
0 10 20 ao 40 so 60
Ti me (Days)
Fig. 3. Cor r osi on pot ent i al of AISI 304 SS and el ect rode pot ent i al s under HWC + 100 ppb SO42 .
results were improved without any difference in
the potential measured with the copper electrode.
In addition, the electrolyte lost (C1-) from the
electrode has not been detected during the experi-
ments.
The Ag/AgC1 electrode has exhibited some
problems in the accuracy of the electrode poten-
tial. This potential can vary between 130 and 190
mVsHE, according to different authors (Leibovitz,
1982; Indig and Weber, 1983; Taylor, 1991), be-
cause of the uncertainty of AgC1 solubility data at
high temperature. In this study, the electrode po-
tential was 130 mVsH E in most cases, only de-
creasing in some measurements under HWC. This
may be due to moderate concentrations of hydro-
gen diffusing through the Rulon wall and causing
a secondary reaction to occur with the silver
chloride:
H2 + 2AgCl(s)~-,2Ag + 2C1- + 2H +
In the above case, there is a lower electrode
potential which generates an increase in the corro-
sion potential measured (Indig, 1990), as shown in
Fig. 3, in which there was an accidental increase
in hydrogen concentration which decreased the
electrode potential and increased the corrosion
potential.
3.3. Copper electrode
A reversible behaviour has been exhibited by
this electrode, because electrode potential mea-
sured with a platinum electrode has shown good
stability, accuracy and reproducibility. In addi-
tion, corrosion potential measurements made with
this electrode showed the same variations and the
same values as those obtained with the platinum
electrode, so this also suggested reversible be-
haviour, independent of chemistry impurities.
Moreover, reproducibility has been achieved
when the same material and chemical environ-
ment have been used, as, for example, is shown in
Fig. 4. The best characteristics of this electrode
were its easy and reproducible construction and
its high strength in the absence of contaminants.
With regard to chemical and thermodynamic
viability of this electrode, the thermodynamic of
YSZ membrane has been extensively studied by
different authors (Macdonald et al., 1990) and
(Niedrach and Stoddard, 1985). Macdonald et al.
concluded that this sensor did not exhibit a ther-
modynamic behaviour, because Cu(II) can be
formed from the internal Cu20 when a sensor is
cycled. Considering this, and because available
thermodynamic data were limited, the powder
M. Navas, M.D. Gomez Brice~o /Nuclear Engineering and Design 168 (1997) 183-189 187
300
2 0 0
%
> I O 0
g
0
W
- 100
i i i i
o - - - o - ~ o - - - - o - - O ~ o _ _ - o _ _ o - - - - O ~ o _ _ o ~ O ~ o
E c o r r l 6 0 0
- - m- - E vs Cu (run1)
- - e - - E v s C u (run2)
- - x - - E vs Ag (run1)
- - v - - E vs Ag (run2)
T e m p e r a t u r e I
- - a ~ T ( r u n l )
- - o - - T ( r u n 2 )
I I I I
100 200 300 400
T I M E ( h )
Fig. 4. Reproduci bi l i ty o f Inconel 600 corrosion
300
- - - - 250
m
" 0
m
200 ~ ,
"--I
C
m
150
100
500
potentials under NWC + 100 ppb SO42-.
mi xt ure was anal yzed by el ect ron spect roscopy f or
chemi cal anal ysi s ( ESCA) , bot h bef ore and after
el ect rode use. In all cases, Cu(II) was not f ound,
i ndi cat i ng a reversible behavi our, accordi ng t o the
results o f Ni edrach and St oddard (1985).
3.4. Corrosion potential measurements
Us ual l y, corros i on pot ent i al s were equal and
reproduci bl e, i ndependent o f the el ect rode used,
as it is pos s i bl e t o observe i n Fi g. 5 under HWC
and Fi g. 6 under NWC. Theref ore, this suggests
that the reference el ect rodes are reliable and have
chemi cal stability, apart f rom the except i ons men-
t i oned above. In addi t i on, the desi gns yi el ded
reproduci bl e results and were free f rom seal fail-
ure.
The reference el ect rodes seem t o provi de reli-
able data, t hus maki ng it pos s i bl e t o study the
effect o f wat er chemi stry o n t he corros i on pot en-
tial. Corros i on pot ent i al meas urement s f or differ-
ent materi al s, such as stai nl ess steel type AISI
304, AI SI 3 1 6 NG, and Inconel 600 were made
under di fferent chemi stri es. An i ncrease i n corro-
03
>
O
IJJ
3 0 0
2 0 0
1 0 0
0
- 1 0 0
- 2 0 0
- 3 0 0
- 4 0 0
' ' I ~
zx
- _ .
- - v - - ~ __~y- - v v - ~ v - - v - - v v - - v - - v - - y
- - v-
o - - o - - o - - o - - o - - o - - o - - O - - O - - O ~ o ~ o
f --x--Ecorr16oo vs Pt
--- ~--- Ecorq6oo v s Cu
zx Ecorrma o vs Ag
- - e - - E c u electrode VS Pt
- - V - - E A g el~trode VS Pt
- 6 0 0
- 7 0 0 I I I
0 1 0 0 2 0 0 3 0 0
T I M E ( h )
Fig. 5. Corrosion potenti al o f Inconel 600 and electrode potentials under HWC + 100 ppb SO~] - .
188 M. Navas, M.D. Gomez Brice~o / Nuclear Engineering and Design 168 (1997) 183-189
-1-
>
g
Q.
O
LU
200
150
100
50
0
- 50
- 1 0 0
- 150
- 200
I I I I I
. . . . - - . - - . - . . . - . . . . - - . . .
i i i i
. ~ 7 _ . . . ~ 7 . . - - V - - ' ?
W
, F ] . . O " [3
O . u.
............ O. . D" ~ t~
x " - - x - - ^ - ~ - - ~ - . . _ - -
- - 1 E " +
. , +
O . . .
I
[ 3 " + +, ' + "
+ - - O - - Ecorrl 600 v s Cu
- - x - - Ecor r 1600 v s Ag
(:3 ECrFAISI 316NG VS Cu
+ EcrrAI SI 316NG VS Ag
- - - v - - - E A g e l e c t r o d e M S C u
l I I I I I , I i I , I ,
50 100 150 2 0 0 250 3 0 0 3 5 0 4 0 0 4 5 0 500
T I M E ( h )
Fig. 6. Corrosion potentials o f AI SI 316NG SS, I nconel 600 and electrode potentials under NWC + 100 ppb SO42-.
sion potential, probably owing to oxide film for-
mation, was observed in the first days when the
electrode surface was fresh (Fig. 7). Occasionally,
the corrosion potential decreased in the first mea-
surements until desired gas concentrations were
achieved.
The results obtained in these measurements
were consistent with those reported in the litera-
ture. The corrosion potential showed a clear de-
pendence on the dissolved oxygen level. However,
the changes observed on the corrosion potential
owing to chemical additives, within the ppb range,
were small in agreement with other authors (Lin
et al., 1992), and in practice sometimes were
negligible because of the small conductivity varia-
tions.
4. Summary
The corrosion potential of a metal or alloy
exposed to aqueous solution is a key parameter
200
150
100
50
0
- 50
I
o~
>
g
o_
0
III
- 100
- 150
- 200
-
oi> / ~"
i i i i
w
. . . . . . . . . . . . . . . . . . . . . . . . v . . . . . . . . . . . v ~2 . . . . . . . . . . . w . . . . . . . . . . .
..:...'
- - o - - Ecorrl eoo v s Cu
- - x - - E c o r r 1 6 0 o v s Ag
o ECOITAIS1316NG MS C U
..... + EcrrAIs1316NG MS Ag
- - - v- - - EAg electrode MS Cu
I I I I
50 100 150 200
T I M E ( h )
Fig. 7. Corrosion potentials o f AI SI 316NG SS, I nconel 600 and electrode potentials under NWC + 20 ppb Cu 2+ .
M. Navas, M.D. Gomez Brice~o /Nuclear Engineering and Design 168 (1997) 183-189 189
for det ermi ni ng the corrosi on behavi our of ma-
terials. However, ECP measurement s are
difficult because it is necessary t o develop a high
t emperat ure reference electrode and great care is
required to assure t hat valid and useful dat a are
being obtained.
In this work, part i cul ar emphasis has been
placed on the i mprovement s of system integrity,
reliability and materials stability of the reference
electrodes. The accuracy and reliability of these
sensors have been studied in a variety of chemi-
cal condi t i ons by means of electrode pot ent i al
and corrosi on pot ent i al measurements. Most of
the electrochemical potentials, measured with
the different electrodes, were equal and repro-
ducible. Therefore, this indicated t hat the refer-
ence electrodes were reliable and had chemi-
cal stability. In addi t i on, the designs yielded re-
producible results and were free from seal fail-
ure.
Re f e r e nc e s
M. Indig, Technology transfer: aqueous electrochemical mea-
surements room temperature to 290C, Corrosion 46
(1990) 680.
M. Indig and J. Weber, Electrochemical Potential Measure-
ments in a Boiling Wat er Reactor, EPRI NP-3362 (1983).
J. Leibovitz, Improved Electrodes for BWR in Plant ECP
Monitoring, EPRI NP-2524 (1982).
C.C. Lin, F.R. Smith, N. Ichikawa and M. Itow, Electrochem-
ical potential measurements under simulated BWR water
chemistry conditions, Corrosion 48 (1992) 16.
D.D. Macdonald, S. Hettiarachchi and S.J. Lenhart, The
Thermodynami c Viability of YSZ pH Sensors for High
Temperature Aqueous Solutions, EPRI NP-6005 (1990).
L.W. Niedrach, Use of a high temperature pH sensor as a
' pseudo-reference electrode' in the moni t ori ng of corrosion
and redox potentials at 285C, J. Electrochem. Soc. 129
(1982) 1446.
L.W. Niedrach and W. Stoddard, Moni t ori ng pH and corro-
sion potential in high temperature aqueous environments,
Corrosion 41 (1985) 45.
D.F. Taylor, Response of electrochemical sensors to ionizing
radiation in high temperature aqueous environments, Cor-
rosion 47 (1991) 115.