ML20040E173
| ML20040E173 | |
| Person / Time | |
|---|---|
| Site: | Fort Calhoun  |
| Issue date: | 05/19/1981 |
| From: | Craig K, Hall J, Rentler R ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY |
| To: | |
| Shared Package | |
| ML20040E171 | List: |
| References | |
| NUDOCS 8202030212 | |
| Download: ML20040E173 (24) | |
Text
..
EXA>1INATION OF CORRODED STUDS FROM FORT CALHOUN APRIL,1981 Prepared by:
Date: [ ((/
J. F. Hall, Principal Engineer
/
[.f!ff Reviewed by:
M/Tl -
Date:
/
R. M. Rentler Supr. Corrosion Engineering Technology 7!f/
Approved by:
/k no I
Date:
K. R. Craig Manager, Systems emistry and Corrosion 8202030212 820128
~
PDR ADOCK 05000285 P
d
~
Legal Notice "This report was prepared as an account of work sponsored by Combustion Engineering, Inc.
Neither Combustion Engineering nor any person acting ca its behalf:
Makes any warranty or representation, express or implied including a.
the warranties of fitness for a particular purpose or mercnantability,'
with respect to the accuracy, completeness, or usefulness of the information contained in this report, or that tne use of any information, apparatus, method, or process disclosed in this report may not infringe privately owned rights; or b.
Assumes any liabilities with respect to the use of, or for damages resulting from the use of, any information, apparatus, method or process disclosed in this report."
l l
l l
l t
i i
r TABLE OF CONTEVTS.
Page List of Tables i
List of. Figures ii l.
Abstract iii 2.
Introduction 1
3.-
Stud Examination A.
Visual Examination 2
4 B.
Scanning Electron Microscope Examination.
4 C.
X-Ray Diffraction Analysis 5
4.
Discussion 6
5.
Summary 7
6.
Reference 8
I.
(
I f
h 6
4 f
I
r i
~
LIST OF TABLES 4
i i
Table No..
Title Page' J
I Results. of EDS Analyses on =
9 i
Selected Areas of the Fracture Face-1
{
II-Results of X-Ray Diffraction Analyses 10 I
of Oxide Samples I
t 4
8 I
4 4
4 f
1 I-1 1!
l-e I
q i
5 t
4 t
i 1
1 i
1 I'
e i.
1 i
1 i
4 I
I
.m.,.-
..--,__,_.....-,,.-._..,.m,,,,.....-,.,.....,_
2 LIST OF FIGURES t
Figure No.
Title Page 1
Overall Views of Two Corroded RCP Studs 11 2
Close-Up Views of the Fracture 12 Area of Stud 3C 3
Area of Deposits and Pitting 13 Type Corrosion on Stud 3C 4
Close-Up View of the Reduced Area 13 in Stud 2B 5
Deposits in the Corroded Area of 14 Stud 2B 6
Areas on the Fracture Face of Stud 3C 15 Examined by the Scanning Electron Microscope 7
Scanning Electron Micrographs of Area 1 15 Shown in Figure 6 Showing the Shiny Deposit on the Fracture Surface and the Three Areas Examined by EDS 8
Scanning Electron Micrograph of Area 2 16 Shown in Figure 6 9
Scanning Electron Micrographs of Area 3 16 Shown in Figure 6 10 Scanning Electron Micrographs of Area 4 17
/
Shown in Figure 6 and the Location of Two Particles Examined by EDS 11 Scanning Electron Micrographs of Area 5 17 Shown in Figure 6 and the Two Locations Examined by EDS 12 Location of Samples Removed for X-Ray 18 Diffraction Analysis 11
1.
ABSTRACT Corroded studs from a reactor coolant pump were destructively examined after removal from service because of excessive corrosion.
During removal, one stud had fractured.
Tne examination determined that the most probable cause of the excessive corrosion, which l
-manifested itself as a localized attack of metal, was " boric acid attack".
Tnis type attack occurs on carbon and alloy steels exposed to borated : ater at relatively low temperatures under conditions that result in the concentration of toric acid.
The stud that fractured during removal from service failed in a ductile manner, indicating that the remaining cross-section was insufficient to support the de-tensioning load, and that the corrosion process produced no environmental degradation in the material.
l iii
2.
INIRODUCTION During testing prior to a-recent start-up at Ft. Calhoun, plant personnel detected primary system leakage coming from a reactor coolant pump (RCP).
Subsequent inspection showed that three of the four reactor coolant pumps were leaking.
In addition, corrosion damage was evident on the exposed shanks of a number of closure studs on the leaking RCPs. Of the 16 studs on each RCP, 4 studs on RC-3A, 8 studs on RC-3B and 6 studs on RC-3C were replaced because of excessive corrosion.
Seven of the studs were corroded to the extent that the nominal 3-1/2 inch diameter had been reduced to 1 to 1-1/2 inch.
Full hydraulic tensioner pressure of 4,500 psi (stud stress of 23,500 psi based on nominal cross-section) was requirad to remove all studs but one from the RCP's. However, the one remaining stud, which was severely corroded, fractured during removal at a tensioner pressure of less than 4500 psi.
The RCP closure studs were fabricated from ASTM A-193 Grade B7 carbon steel, chrome plated in the thread area, and phosphate coated in the shank area.
Tne studs were approximately 3-1/2 inches in diameter and 29 inches in length.
Reference 1 indicates that the RCP leakage was the result of leakage from the shaft seal on one RCP and from the gasketed joint between the pump casing and pump cover on all three RCPs.
Leakage from the gasketed joint was apparently caused by the deterioration of the Flexitallic gaskets.
Thus, there was an opportunity for the RCP closure studs to De exposed to borated water.
Previous experience at C-E witn similar situations in the laboratory has indicated that carbon and alloy steel fasteners exhibited accelerated corrosion when exposed to borated water solutions under conditions which promoted alternate wetting and drying.
Metal temperatures on these occasions were about 200 F.
Tnis type corrosion has typically been identified as "ooric acid attack" of
4 carbon steels.
When it has been observed on studs, they have been replaced and discarded without detailed examination.-
C-E submitted a proposal for a limited scope examination of a corroded RCP stud - from Fort Calhoun.
As accepted, the proposed 4
examination consisted of 'a visual examination, X-ray analysis of deposits removed from the ' stud and scanning electron microscopy examination of the fracture surface of the fractured stud.
Two studs were actually shipped to Windsor, including the stud that fractured during disassembly. @e non-fractured stud was originally shipped as a back-up if the fractured stud was unacceptable for examination.
Only limited work was done on this stud.
3.
STUD EXAMINATION A.
. Visual Examination The condition of the studs as-received at Windsor is documented in Figures 1-5.
Figure 1 shows overall views of both studs.
Figures 2 and 3 give close-up views of the fracture area, deposits and areas of corrosion on stud 3C (fractured stud).
Figures 4 and 5 show the l
heavily corroded area and deposits on stud 2B.
In stud 3C, the cross-section at the fracture area had been reduced by corrosion to approximately an oval with minimum and major diameters of 1.124 and 1.207 inches, respectively.
The fracture surface showed no clearly definable point of origin for the final-failure.
Further, the fracture surface was not the flat surface typically observed in brittle fracture.
Tne fracture surface had a fibrous appearance, an irregular contour, and extensive shear lips, all of which are typical of ductile fractures.
We side view in Figure 2 and the similar view in Figure 4 of stud 2B show numerous tears.in the stud in the areas where the diameter
reduction were greatest.
The tears were probably the result of the ductile rupturing of the metal as it was stressed near its ultimate strength.
%e infonnation contained in Reference 1 makes possible a crude calculation of the stresses applied to the remaining cross-sections of the studs.
Daring de-tensioning, the studs were subjected to a nominal stress of 23,500 psi, based on a 3-1/2 inch diameter stud.
Stud 3C failed at a stress of less than this value, but 23,500 psi was used in the calculation.
The failure stress was calculated by multiplying this stress by the ratio of initial cross-section area to final cross-section area.
Tne resulting stress for stud 3C was greater than 200,000 psi which is above the minimum specified ultimate strength for SA 193 grade B7 steel (115,000 psi).
- Thus, failure of the stud could be expected.
For stud 2B, a similar calculation showed the maximum stress in the stud to be about 180,000 psi during de-tensioning.
This value is also above the specified minimum ultimate strength of the material, but obviously was somewhat lower than the actual ultimate strength.
De numerous tears, however, suggest that the maximum stress during de-tensioning was very near the ultimate strength of the material.
Tne fracture surface ha,d.not been appreciably damaged by the failure and/or subsequent handling.
Shiny deposits were visible at several locations, as shown in Figure 2.
Tne origin of these deposits was not apparent. Part of the surface was also covered with a thin light orange-brown oxide layer tnat had formed since the fracture of the stud.
Figures 2 and 4 also show that the areas where the greatest corrosion occurred were devoid of deposits.
There were numerous large and small shallow crater-like pits in these areas. The figures snow that deposits were present above the most severely corroded areas. Stud
}
-4 2B had more extensive deposits than stud 3C.
Tho' deposits were predominantly orange, but there were-areas of yellow - or white
~ deposits (Figure 5).
Most of the deposits were, adherent, but could be removed by forcibly scraping with a spatula. Where deposits were scraped, the shallow pit-like areas described abse-were present.
On both studs, the area of significant corrosion began about 21 inches from the top of the stud.
On stud 3C, about 3-1/4 inches of the stud above the fracture face showed significant corrosion.
On stud 2B, the area of corrosion was about 5-1/4 inches in length. The bottom 3-3/4 inches, including the bottom 18 threads, of stud 2B were essentially free of corrosion.
However, at least seven threads were severely corroded.
'Ibe general contours of the individual - threads readily discernible although the depth of corrosion extended were below the original thread roots.
Tne thread contours were preserved by the relatively uniform manner in which corrosion proceeded in this area.
After the visual examination was completed, samples of the oxide were taken from two locations on each stud. Then, the fractured stud was sectioned about one-half inch below the fracture to provide a sample for scanning electron microscope examination.
B.
SCANNING ELEC' IRON MICROSCOPE EXAMINATION The fracture surface was examined prior to cleaning or descaling,to preserve any deposits or oxides on the surface.
Figure 6 shows the five areas on the fracture surface that were examined by the SBl.
Figures 7 to 11 show typical scanning electron micrographs of ear.h of the five areas.
The initial examinations were conducted at lw magnifications (10-50X) but the photomicrographs ootained at these magnifications did not show details of the fracture surface.
Accordingly, higher magnifications (up to 1500X) examinations were conducted on these areas.
1
. In some areas, the fracture surface was covered with debris or oxides which obscured surface details.
However, at all locations examined, there were at least some small areas where the fracture surface was visible.
In these areas, the surface consisted of the ductile dimples that are indicative of ductile fracture; i.e.,
the failure resulted from _ loading the material above its ultimate strength.
In Figure 9, the dimpled surface is especially well shown.
In Figures 10 and 11, the dimples are also visible, but surface debris and oxides at least partially obscure surface details.
There were no indications of cicavage or intergranular cracking at the areas examined.
Table I gives EDS analysis results obtained at the five areas on the fracture surface.
Included in Table I is the relative intensity of the energy peak for each element identified in each analysis.
Tne objective of the EDS analyses was to determine what, if any, contaminants were present on the fracture surface.
The sniny deposits present on the surface (Figure 7) were high in Zn, S, and Ti and were probably the result of surface contamination after fracture.
The deposits in area 5 (Figure 11) were rich in Si, as were the particles in area 4 (Figure 10).
The latter also included significant amounts of Mg.
Other elemental species identified included Ca, K, A1, Cr and Fe.
Most of these were probabiy present as the result of post-service contamination of the surface.
C.
X-RAY DIFFRACTION ANALYSIS Wo samples of the oxide deposits present were removed from each stud.
The deposit samples were taken from the reduced area and from the less severely corroded area above the reduced area on each stud.
These areas were sampled to determine the differences in the composition of the deposits at these areas.
Tne areas sampled are shown in the sketch in Figure 12.
. Table II gives the resuits of tne XRD analyses.
Boron compounds were identified in the sample removed from the reduced section of stud 3C and from both locations sampled in stud 28.
H B0 was identified 3 3 as a major ( >20% by weight) molecular species present in the deposit at location 2 in stud 2B and was identified as a minor ( 5 to 20% by weight) species in the deposit at location 1.
B0 was 23 identified as a minor species at location 1 and a trace ( < 5% by weight) species at location 2 on stud 2B.
Tne sample from the deposit nearest the fracture face on stud 3C contained compounds of Fe and B.
The C in the trace amounts of Fe (C, B)6 Probaoly 23 resulted from acetone used in sample preparation.
Fe (B0 )2 3
3 was also present in trace amounts in the deposits In addition to the B compounds, various oxides of Fe were also identified in the samples.
The major species present were Fe 023 and Fe00H, which were identified in each sample.
4.
DISCUSSION The condition of the studs was consistent with previous occurrences of corrosion on carbon or alloy steels which had been exposed to borated water.
The reduced cross-section, the presence of boron in deposits and the pit-like areas under deposits all point to so called i
boric acid attack as the mechanism of corrosion.
The specific l
mechanism tnat produces this attack is not known.
Of the possible mechanisms that might produce this type of corrosion, the most plausible appears to be a form of acid attack.
l l
Wnile it is true that boric acid is a relatively weak acid, it has a theoretical pH of about 3 in saturated solutions at 20C F, even when Li is present in concentrations of up to 0.1%
of B
concentration.
Li would be present if primary coolant was present.
Under conditions that promote wetting and drying, the borated solution will concentrate until satuarated and deposits of Doric acid will form.
l
. The effects 'of acidic solutions on iron and steel are well documented.
First, the thin, essentially protective, oxides that form on irons and steels exposed to 300-600 F air are dissolved by the acidic solutions.
Tne oxides tnat subsequently form on iron are soluble in acids and thus non-protective in nature.
In addition to being soluble in acidic solutions, the oxides also tend to be porous with numerous cracks, etc., so there are always routes by which the solution can reach unexposed metal.
The result of all this is that acidic conditions cause accelerated corrosion in carbon and alloy steels and at-a pH of about 3 this corrosion becomes greatly accelerated.
If acid attack is the mechanism that caused the damage in tne Fort Calhoun RCP studs, then it is possible to see how the corrosion process might cease when the temperature of the stud is increased.
When the water in the solution is driven off, dry deposits of H B0 salt will remain in place and the corrosion process will 3 3 presumably cease in the ausence of an electrolyte.
At the temperature of the stud is increased further, the boric acid salts will convert to meta-boric acid and finally boric oxide as follows:
2H 80 2
2HB02 + 2H O (+ )
Z B023+ H O (+)
2 3 3 2
These reactions are reversible.
Tnus, wnen B0 is rewetted, it 23 will convert back to H B0, thereby providing an acidic solution 3 3 and re-initiating the corrosion process.
This is the basis for considering boric acid Lttack to be intermittent.
5.
StM4ARY The destructive examination of the Fort Calhoun RCP studs resulted in the following findings and/or conclusions:
A.
Tne areas of corrosion were similar visually to areas of corrosion in carbon and alloy steels that were exposed to borated water under alternate wetting and drying conditions.
. B.
Boron, as boric acid (H M ),
was present in the deposits 3 3 adjacent to the areas of reduced section.
C.
The presence of boron in the deposits and the condition of the corroded areas indicate that boric acid induced corrosion of the carbon steel was the most probable cause of the excessive corrosion observed on the RCP studs.
4 D.
The fractured stud failed in a ductile manner.
The fracture surface contained only ductile dimples indicative of-ductile failure and no indications of cleavage or intergranular cracking such as would be present if brittle fracture or environmentally assisted cracking had occurred.
6.
REFERENCE 1
Fort Calhoun Licensee Event Report 80-010, dated June 17, 1980.
t 4
y
.~
Table I RESULTS OF EDS ANALYSES ON SELECTED AREAS OF THE FRAC 1URE FACE Area (1) Location Elements Present(2)
Fe Al Si S
K Ca Ti Zn Cr Mg 1.
1-Shiny Deposit E
B A
A
.2-Sniny Deposit E
E E
_A E
3-Fracture Surface A
E E
2.
1-Fracture Surface A
E E
E 3.
1-Fracture Surface A
E E
E 2-Particle on Surface A E
E 4.
1-Fracture Surface A
E E
E E
-- E 2-Particle on Surface A A
E E
E B
3-Particle on Surface A A
E C
5.
1-Deposit D
D A
E E
D E.
2-Fracture Surface A
E A
E E
E E
E Notes:
(1) See Figure 6 for locations of areas examined on the fracture face of stud 3C.
(2) A - The major peak, or any other peak greater than 75% of the major peak in intensity B 74% of the major peak in intensity C 49% of the major peak in intensity D 24% of the major peak in intensity E - Less than 10% of the major peak in intensity 1
i Taole II RESULTS OF X-RAY DIFFRACTION ANALYSES OF OXIDE SAMPLES Stud Location Major
- Minor
- Trace
- 3C' 1
Fe 0 F23(C B)6 23 1
Fe (B0 )2 Fe001 3
3 Fe 0 (01)*4fi 0 57 2
2 Fe 0 Fe 0 23 34 Fe001 2B 1
Fe 0 B0 23 23 Fe001 H B0 3 3 2
Fe 0 B0 23 23 Fe00H H B0 3 3 1
Major - >20% by weight j
Minor 20% by weight Trace - < 5% by weight l
1 i
i
- ~.
s
. 1 1
1
- r
- nr s. - --
i.
,;.y.,
l 's - - ' -- -, : r -'
- r '
c..':M..,,.,;,y-
.-M. m,
}
- s. _
-a 4
2 4
1 6
I j
t Stud 3C Neg. #48532-2 l
i i
WI s
y... ~ ~fM, w 24....,_.,,._...._.
h h
I mmw
- wb c
a s
$P57 E IE lS Y i* a w$ $ 9 $ $ 5fh f $l $u 6
Im$.5J '~ l7d tWWhb?hd??d$Nhi$i5GEF i;mh
-s.s2snrwelxe.cwuca%m-m;4.,1ps;m.;a.:s.)anm.&a b $n~4y A n:. t i W
..,m
- _ c
x w
.z g-r> w h.nr:w ~
v Stud 2B Neg. #48532-5 Figure 1.
Overall Views of Two Corroded RCP Studs.
1 1-I
m m
.r
.:-j
=,
y b
v w g
rz.
~d,* j;
- a.,,
+ ' " -
e
/
'.d q;
~.
r.
Neg. #48532-29
,m,,.m r -- _
- s.,. ;g
' QW;.
- s 1
- e
- ,a W
/
$1 * '.4'
} p' J.
f
. v.
.. - z 7,,.W.,,.
Y,kh'-
b
~
EY
%j
,k.#'
uf
+. '
- f+L
+4 yl gg.
. g g y; mne
- ~.
~
_.3. _,.y. :..
(
i
,1 Neg. #48532-30 Figure 7.
Close-up Views of the Fracture Area of Stud 3C.
l l
1 i
l
e *bf.$='W M 7'ShM A,
.g 3
i%KSf4 i q-f& % l
L JEgr.{J< fg;g,':f.g;l#
3N i
I.
g gg N, Qq: y. W',$ $:&&~1 B n~ h+y
'+Q.
r+
v ':
~-
r c
92h)f h'i#g. %,("
D. ?s
, q.
w s p,. x j;. R k
s; * '
- S,- p}f s:,
.f g:. ~ A 9
- . rn
, 3-
-?--
l z
i Neg. #48532-33 Figure 3.
Area of Deposits and Pitting Type Corrosion on Stud 3C.
U tj -.)
'C
- ?,
N
__ w.
J "n
ff
_3
_r.
a.x Rt 8
J'
, f,
~
' [i},
g k.
- }j y
9
~
Neg. #48532-17 Figure 4 Close-Up View of the Reduced Area in Stud 2B.
i
. t R$:h;b.,-
~~
f s$
aga.,m h/k 1
, kI, M
3p.Qg a
f' R
,y
(
q Idi.
g 1
I
?
l 1l' 4
l Neg. #48532-20 b;/?W.??a?f*y*%'
?.
af4
??*
.;~9: 4 p
^
g/
?. r,. p
~ p, e,
~.-
3,
- 4 '.,..- ~,+
);'
g',
- p
' y ' 'h h)'jffs.k 4
r...f
.a 4 -:.- Q v % '.-
y i
,.:[p,8
^gh7~.
[yz...A grI 'S '. 9s?f~Y.
..Q___
{
- A 4
s p
i Neg. #48533-4 Figure 5.
Deposits in the Corroded Area of Stud 2B.
f d
1 5 3
ei:
\\
\\ \\
(<
2 3
4 i
j 1
.r Figure 6.
Areas on the Fracture Face of Stud 3C Examined by the Scanning Electron Microscope.
5 UI w
g 13 m
m J
.>.j x
,j
),
~
M
.- :_.. ~
.1*.-~
a
- - h.;'
_}
l
$C['
. q
~ '
E
=
41 V*
1.i "'
j.}
h
=
3.,
~
4'
,i p
- Aa.
N 160X 10X Figure 7.
Scanning Electron Micrographs of Area 1 Shown in Figure 6 Showing Shiny Deposit on the Fracture Surface and the Three Areas Examined by EDS.
2
l f 1 s
j k:
b w i j' M'
;'Q
~
y.,
l ak r
50X Figure 8.
Scanning Electron Micrograph of Area 2 Shown in Figure 6.
q 9
wy
' 7,'.4
- fs p-2 2
f.
9 50X 1500X Figure 9.
Scanning Electron Micrographs of Area 3 Shown in Figure 6.
.. _..-., _ v -.,
__--m__-,,,,----,,,
--,.n.
- I Iwyh x
4..,.% M/
,;h.h, 5,
+
Y d - 11 y g;
1
~
rw t' y#{
6 b
y~
I,%,
w 3
.[
- a.
3 4.
g.
- p. $
d'
- G r
'f
~
.~
c
.~
34' 50X 1500X Figure 10.
Scanning Electron Micrographs of Area 4 Shown in Figure 6 and the Location of Two Particles Examined by EDS.
m-en v.~ w.1.gd.,A N5:
,,y.hyl
$$IAGNNiidh.y:y?
n~r up A
M9%ggfgl&t!9 y%Q[-[Mqrt?.
'k!
Nid Ts3 s
hhb
$N i 'N 94## "r?essi$
4
'$)
5 kIkik
'I ky-DF.7 Le$N
.h
' > ~ '
!'A*hYSi
' &}Fisp.1r1v~ m :." 2 m PM H a w$
~
g ;i M n P "'* * '*" C N " " C D D 7"""3
- ~:
^
t 50X 1500X Figure 11.
Scannirig Electron Micrographs of Area 5 Shown in Figure 6 and the 'Iko Locations F.xa:nined by EDS.
d
l 1 Lc:ation 1 Location 2 P
Stud 3C Location 1 Location 2 1
TTF r
n Stud 2B Figure 12. Location of Samples Removed for X-Ray Diffraction Analysis.