ML20198N019
| ML20198N019 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 01/09/1998 |
| From: | Casey Smith NRC (Affiliation Not Assigned) |
| To: | NRC (Affiliation Not Assigned) |
| References | |
| NUDOCS 9801200219 | |
| Download: ML20198N019 (104) | |
Text
L
.I N r
e a ct:u g*
1.
UNITED STATES
- r j
NUCLEAR REGULATORY COMMISSION t
WASHINGTON, D.C. 20666-4001 '
4,**..,/
January 9, 1998 MEMORANDUM TO: Docket File FROM:
Craig W. Smith, Project Manager _
j Project Directorate I-3 Division of reactor Projects - 1/II Office of Nuclear Reactor Regulation
SUBJECT:
TRANSMITTAL OF INFORMATION TO DOCKET NO. 50-443 The attached report, dated January 7,1998, is being transmitted to the docket for Seabrook Station, Unit 1. The draft report provides the licensee's preliminary root cause analysis for the leaks in the piping of the "B' train of the residual heat removal (RHR) system.
The leakage was reported by the licensee on December 5,1997.
Docket No. 50-443
Attachment:
As stated
\\
r 1RD 31E gg;c ggpy
@l 9001200219 900109 DR ADOCK 0500 3
I.llll ll llllll ll ll
oa l
EXECUTIVE
SUMMARY
On Ikcember 5,1997, system leakage of a drop every two to three minutes (weepage from four locations) was discovered on the inlet piping to the Residual Heat Removal pump 8B suction F
relief valve (RC-V-89). The weepage sites were in the metal adjacent to the welds (heat affected zone) that connected the piping spool piece to the relief valve flange nnd the 3" to 12" weld-o-let that connected the pipe to the 12" RIIR pump suction line. One weep was in the upper heat affected zone and three were in the lower heat affected zone. North Atlantic personnel performed a preliminary evaluation of the weepage sites and were unable to quantify the air leakage in scfh. Therefore, the Operations Manager conservatively determined that the Limiting Condition for Operation for Tecimical Specification 3.6.1.1, Containment Integrity, could not be satisfied. He declared the Containment inoperable and entered the referenced Action Statement.
in addition, the Residual Heat Removal (RHR) and Containment Building Spray (CBS) systems, which utilize the same section of pipe, were declared inoperable. A plant shutdown was commenced and completed in accordance with the requirements of the Specification. A one hour report was made to the NRC pursuant to 10CFR50.72(b)(i)(a).
The weepage sites were caused by chloride induced trans5ranular stress corrosion cracking (SCC). The source of the chlorides was a non standard insulating jacket that had been on this section of pipe since at hast 1988. The insulating jacket consisted of a piece of cloth firewrap type material wrapped m red duct tape. The insulating jacket was not the type ofinsulation designed to be in p ace on this section of pipe. It is unclear exactly how or v/hy this insulating l
jacket was placed on this pipe. Separate ehtoride analysis of the insulating jacket resulted in values of 189 ppm and 349 ppm. It is unknown what the original chloride value of the cloth or the duct tape was when it was installed.
Laboratory examinations consisting of microscopic and chemical analysis have determined that SCC is the most likely cause of the pipe weepage. The analysis determined that the SCC had started on the outside of the pipe and had migrated inward. It is suspected that repeated wettings and dryings of the insulating jacket, caused by wetting of the surfaces during draining of the pipe as part of maintenance performed in refueling outages, may have caused the chlorides to concentrate on the pipe over time.
Original radiographs of the two welds on the spool piece were retrieved and examined. The review showed no anomalies in the weld or the base metal. It was also noted that the two welds on the spool piece were performed by different welders in different companies. The flange to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o-let to pipe weld was a field weld performed at Seabrook. This indicates that the welds and welder qualification do not appear to be a contributor to the problem as evidenced by the fact that the radiographs showed no anomalies and the welds were performed and examined by different individuals yet the heat affected zones of both welds had the weepage sites. Also, there did not appear to be any anomalies in the base metal of the pipe which further indicates that the weepage was due to in service conditions rather than original installation.
1 Dr AFTRevMon 5 D7198 1
This jacket had been installed on the RC-V-89 inlet pipe since at least October of 1988. From this time until its removal on 12/5/97, it was not identified as inappropriate for this application.
The monitoring of the boron buildup on this pipe began in March of 1996 and continued until the weepage was discovered on 12/5/97. During this time it was believed that the source of the boron was related to water spillage on the jacket and on insulation on the 12" RHR pump suction line.
It was thought that the water absorbed by the insulation was drawn up the jacket in a wicking process and then evaporated leaving the boric acid residue behind. This idea was reinforced by the fact that at no time did anyone recoit finding moisture or wet boric acid when inspecting or cleaning the boron buildup. For these reasons thejacket was not removed and a positive determination of the source of the boron buildup determined until 12/5/97. After being informed for the first time that the boron buildup had been cleaned and then reappeared, the system engineer led a team into containment to remove the jacket and inspect the pipe. However, as an organization there were missed opportunities to identify, document, evaluate, and resolve the issues prior to December 1997.
The piping section containing the defects was removed and subjected to a leak rate test. The leak rate test pressurized the pipe to between 50 and 52 psig, This is greater than the containment design basis accident pressure of 49.6 psig. The makeup flow required to maintain the pressure at 50 - 52 psig wa: measured. The leak rate was determined to be 1.363 standard cubic feet per hour (scfh). This compares to an allowed penetration leak rate of 37 scfh or 0.05 La as described in the containment leakage rate testing program. In addition, the piping section was analyzed with the assumption of a bounding 0.5 inch diameter through wall hole. The evaluation determined that the piping section and the weld-o-let were fully capable of carrying all design loads. The structural integrity of tlm piping system under all normal and accident conditions would have been maintained. The evaluation is provided in Attachment E.
Based upon the evaluation it was determined that the piping spool was degraded ana nonconforming, but that the three systems, Containment, RHR and CBS were fully c%bie d performing their specified safety functions if they had been called upon to do so a vere
. Operable utilizing 'he guidance of Generic Letter 91-18, Revision 1. Therefore, it was recommended that Operations review the engineering evaluation and withdraw the one hour repor; of December 5,1997.
Based on a review of the events and causes of the conditions evaluated in this report the event evaluation team recommends to the SORC that they recommend to the Station Director that the plant be restarted after the completion of the applicable corrective actions.
l l
DMFTRedston 3 V788 2
a l.0 -
EVENT DESCRIPTION 1
Thru wall weepage sites were discovered on the five inch long 3 inch diameter line that connects RHR relief valve RC-V-89 to the main 12 inch RHR header. The relief valve and its inlet piping have had a history of observed boric acid buildup.
2.0 INITIATING CONDITIONS The installation ofinappropriate material on the inlet piping of RC-V-89.
3.0 REPORT CATEGORY l hour report per 10 CFR 50.72 (b)(i) (a) (Withdrawn) 4.0
SUMMARY
OF PLANT PERSONNEL ACTIONS ASSOCIATED WITH RC-V-89 INLET PIPING
4.1 BACKGROUND
System leakage of a drop every two to three minutes' was discovered on the surface of the 3" pipe spool that connects RC-V-89, RHR-P-8B suction relief valve, with the 12" pump suction line inside containment. The discovery of this condition resulted in the plant being shut down to repair the piping. The following narrative is a description of the events leading up to this event and the recollections of the individuals who have been interviewed by the members of the event evaluation team formed to investigate the causes of this condition.
Every refueling outage RC-V-89 is removed from the flanged connection at the end ofits 3" diameter,5"long inlet pipe which is welded into the pump suction line. This is done to support local leak rate testing. Removal of RC-V-89 requires that it be unbolted at the discharge flange. When this is done, the length of piping that extends from the valve's outlet to it's final discharge point, the PRT, drains backwards out of the flange.
- According to the ve.lve test engineer, when the water in the valve discharge line was drained in this manner during OR04, the top of the insulation on the suction line and the firewrap surrounding the RC-V-89 inlet pipe were wetted. When the valve test engineer tested RC-V-89 during OR05 the insulation in the area was wet.
DAAFT Acedsdes 5 1/7/98 3
i FIGURE 1 REllEF VALVE RC-V.89 5
{
']
a lE T
4
.4 b
/
x N
1 THRU WALL WEEPAGESITE y
IN HEAT g(
AFFECTED ZONE (SHOWED MORE
/
LEAKAGE THAN
/
1 OTHERS) 3" SS PIPE
\\
/
h
/
3 THRU WALL
\\
V WEEPAGE SITES
\\
/
INHEAT
\\
/
AFFECTED ZONE h
d 3' WELDOLET \\
l A
w
/
12' TRAIN B RHR PUMP SUCTION HEADER DRAFTRedston $ U7i98 4
1
4 i
The 3" valve inlet pipe tL is welded to the relief valve mounting flange was covered with a fiberglass fire protec on cloth wrapped in sed duct tape. The cloth looks similar to insulation, but apparently is material used to protect piping or other components from welding are strikes or burning debris during welding operations. The duct tape on the firewrap is wrapped from %" below the top to the bottom of the firewrap. The tape has taken on a seamless reddish appearance, looking like a painted on, thir, rough textured ceramic or plastic coating. Photographic evidence proves that this taped firewrap has been on this section of pipe since at least 1988.
4.1.1 March 1996 Observations On March 20,1996, during a routine containment walkdown at power an Operation's Department Nuclear Systems Operator (NSO) documented on Operation Department Instruction (ODI) 36, " Containment Tours" the fact that the RC-V-89 inlet piping firewrap had boron residue on it. Observations by different NSOs and HP technicians during subsequent containment walkdowns were documented on ODl-36 and on the RTS for routine cleanup of boric acid from various plant components. Operations and HP supenisors were informed of the condition in the area of RC-V-89 after these containment entries and observations of RC-V-89. RC-V-89 remained on the ODI 36 list of components with boron buildup until December of 1997.
4.1.2 November 1096 Observations On November 20,1996, after repeated observations of the boron buildup at RC-V-89 the RCS system engineer was asked by HP to investigate the build up of boron crystals that had formed in the area of RC-V-89. An HP Department supenisor awarded a spot recognition award to an HP Technician who raised a question about the condition at RC-V-89 including the boron deposits and the insulation on the inlet piping. The RCS system engineer went into containment with the HP technician and took a camera with him. The RCS system engineer observed boron deposits on the relief valve inlet pipe firewTap and on the bottom of the flange. The RCS system engineer suspected that the flange connection of the ielief valve to the riser pipe had leaked and that the water had flowed down rad accumulated on the inlet pipe firewrap. He removed enough of the boron from the flange to satisfy himself that the flange connection of the relief valve to the inlet pipe was not leaking at that time. The inlet pipe and flange was photographed and the individuals left the containment.
The RCS system engineer told the HP personnel at the HP checkpoint that the boron residue was dry. The RCS system engineer discussed the situation with the valve test engineer and was told about the water that was spilled on the insulation and firewrap during refueling outages. The valve test engineer explained that he thought when water that was spilled on the insulation evaporated, the boron residue was left behind; The RHR pump suction line and relief valve inlet pipe can get as hot as 350 degrees when RHR is in senice. This could cause evaporation and the subsequent formation of boron crystals. That fact combined with the understanding that the normal ambient temperature of containment can l
DMFTRevhlon 5 V7M8 5
l
~ _ - - - - - -
x i=
be as high as 110 degrees, made the explanation seem reasonable. The explanation was relayed to HP and they accepted that explanation.
- Adverse Condition Report (ACR) on this condition nor did they ensure the boron ~
buildup was documented in the RTS for routine cleanup of boric acid from
- various plant components. Both engineers assumed that HP would enter the condition at RC-V-89 on the boron cleanup RTS and that it would get cleaned.
Neither engineer considered the non standard insulation significant enough to warrant a WR or ACR being written. They did not recognize that the reddish color was from duct tape.- They did not conclude that there was a potential for negative consequences as a result of the non standard-looking insulation on this pipe. This same conclusion was reached by others including managers and supervisors who were shown pictures of this pipe with its boron buildup and reddish colored insulation.
4.1.3, February 1997 Observations No additional documented actions took place concerning the condition at RC-V-i 89 until February of 1997. In February, an HP Supervisor contacted the RCS syst_em engineer and asked him again to evaluate the boron buildup on RC-V-89.
The HP technicians felt that something was not right about this situation and they asked the supervisor to have it looked at again. No evidence of water or moisture was reported tn the system engineer at that time or any other time from when he had first become aware of this condition. The system engineer spoke to the valve test engineer about the new call from HP concerning the conditions at RC-V-89.
The valve test engineer was making regular entries into containment to modify the containment personnel hatch entry system. Since the valve test engineer was going into containment, the system engineer gave the valve test engineer the November 20,1996, picture of the RC-V-89 inlet pipe and its firewrap. He asked the valve test engineer to compare the current condition of the boron buildup with the condition that existed at the time the picture was taken.
l After checking the area at RC-V-89 the valve test engineer responded back to the RCS system engineer that there was no change in the status of the boron buildup on the RC-V-89 inlet pipe firewrap. The RCS system engineer concluded that since the condition of the boron buildup had not changed, the explanation of the water spillage and subsequent evaporation and boric acid buildup were still plausible. It did not appear that the RC-V-89 inlet pipe firewrap boron buildup
_ had ever been cleaned and so they thought that the condition was stable and acceptable.
~-
- Neither engineer asked HP if they had ever cleaned the boron buildup on the firewrap. They also did not review the RTS for routine cleanup of boric acid from -
various plant components. ( Note: The only documented evidence of RC-V-89/RC-V-89 inlet pipe being cleaned is logged on this RTS on June 21,1997.)
(Both still felt they completely understood the situation.Both of the engineers 2
stated clearly and decisively that if at any time they had received reports of
^
. anArtn&s wm 6
g y
,93. - ---
9y y
e..
va e 4 4%-g,.
e
+
=.,-y
,..,n-
- +.,,..-
,.w
,er.y
-.r--
%.-,e-%.-
m
.m-.-
-r,
moisture or water on the RC-V-89 inlet pipe or firewrap they would have immediately generated the paperwork necessary to investigate and repair any leaks. They both believed that they did not have any type ofleak.
Both engineers felt it was not necessary to write an ACR at this point because generally, boron buildup on piping or components in and ofitself did not meet the threshold for writing an ACR. ACRs were not written on every case of boron buildup in the plant, if a component had a minor leak at a mechanicaljoint or packing that was causing the boron to build up, the valve packing gland or pump shaft seal would be added to the RTS for routine cleanup of boric acid from various plant components. If the leak was more than minor, a WR would be written to repair the component or to perform cleaning that would involve removing ofinsulation. Neither engineer thought that a leak existed. They both thought that eventually the water would all evaporate and the boren buildup would be cleaned up by HP using t' e RTS for routine cleanup of boric acid from various plant components. However, neither engineer ensured that the RC-V-89 condition was added to the RTS.
4.1.4 March 1997 Obst sations On March 17,1997, Technical Specification Surveillance procedure EX1810.208 "RHR Train B ISI Functional Test" was performed. The inspector who pt:rformed the surveillan<. correctiv documented the condition of the RC-V-89 inlet pipe firewrap as a dis Nancy in the comments section of the RTS as required. He reported to the ISI test supervisor that, after close irm yction, he did not see any signs of water or moisture but noted the boron evident at the top and bottom of the firewrap. Procedure ES1807.001," Visual Examination Procedure" requires questionable or indeterminate conditions that are not going to be reworked or replaced to have an engineering evaluation performed by Technical Support personnel or by a NDE Level 111 qualified person.
ES1807.001 also requires an independent review of all evaluations for questionable or indeterminate conditions. Contrary to this procedural requirement, an independent review was not completed for the evaluation done by the ISI test supervisor for the boron buildup on the inlet to RC-V-89.
The acceptance criteria used throughout ESI 807.001 is the term " evidence of leakage". The ISI test supervisor knew there had been boron buildup in this location before and he spoke to the valve test engineer about it. He knew that the valve test engineer was familiar with the situation. The valve test engineer explained to the ISI test supervisor that he believed the boron buildup that the field inspector, was reporting, was the same boron buildup that the two engineers had discussed in November 1996 and February 1997.
The ISI test supervisor discussed the situation with the RCS system engineer and the RCS system engineer agreed with the conclusion reached by the valve test i
engineer. Both engineers were convinced that because there was no visible sign of moisture or water evident at the firewrap, then there was no leakage. This was the basis for the evaluation performed under ESI 807.001. The field inspector DMFTRevision s 1/798 7
discussed this situation with his supervisor.. The field insnactor's tupenisor accepted the conclusions of the engineers.
This ISI inspection was conducted 15 months after the water was spilled onto the insulation that was thought to be creating the boron buildup on the RC-V-89 inlet pipe firewrap. None of the individuals involved had an estimate for how long the spilled water would continue to evaporate, or how long the boron would continue to buildup on the firewrap. They did not think it was getting any worse. The boron buildup was not increasing, so they assumed they had a static, understood condition. The field inspector when shown the November 20,1996 picture also said he thought that it did not look like there had been any change from November of 1996 to March of 1997. He had first thought there might be a small leak under the firewrap, but when he heard the explanation from the engineers he decided that they were correct and that there was no leak.
The ISI test supenisor wrote WR 97W000833 that described the problem as boric acid deposits at the relief valve connection to 12" pipe. He suggested cleaning or replacing the insulation as required The RCS system engineer, who filled in the section of the WR that described the work that should be performed, wrote " clean boric acid residue and remove damaged insulation (reddish color)"
The field inspector who did the field walkdowns for the RHR ISI fimetional test wrote an ACR describing the poor work practices of the valve maintenance personnel who disassembled the relief valve and spilled the water on the insulation. The basis of this ACR was the information that he received from the valve test engineer who told him he had seen these poor work practices. This ACR did not address a potential leak in the pipe.
4.1.5 OROS Observations In early June of 1997 the RCS system engineer performed a containment walkdown in preparation for the containment close-out at the end of OR05. He investigated the status of the RC-V-89 inlet pipe firewrap. The system engineer discovered that the work that was to have been performed per work request WR 97W000833 was not complete. The engineer discovered that no work had been performed on the firewrap.
Some work had been completed using WR 97W000833 and the WR had been routed to the RCS system engineer for his review and close out. The system engineer retrieved the WR and examined the comments on the task performed.
The following comment was written on the WR: " Removed and replaced insulation. We pre-fabbed Nukon blanket and applied to valve. Metaljacket was de-conned by H.P. 5/17/97." The system engineer told his supervisor that he i
could not sign off and close out the WR because the work that was performed was not the work that was supposed to be performed.
1 he RCS system engia:er did not know at the time that the insulators did not l-understand exactly what insulation they were supposed to remove. The insulators did not consider the reddish colored firewrap to be insulation They noticed the DRAFTRedston 3 V7;98
-8 i
l insulation on the 12" RHR suction line below the RC-V-89 relief valve inlet pipe was wet, so they removed the old insulation, fabricated new insulation and replaced the insulation. They did not ask any questions of their (NAESCO) supervisors about the " reddish color" comment on the WR. None of the insulation that the insulators removed was red in color.
The RCS system engineer told his supervisor that no work had been performed on the RC-V-89 inlet pipe firewTap. His supervisor told him to write an ACR describing the adverse condition and to check with the NAESCO supervisor of the insulators to see if he could discover exactly what component was worked on as described in the WR. The system engineer did not write the ACR. He intended to write the ACR after he discovered what had occurred and had gotten the correct work performed. This did not occur before the RC-V-89 pipe weepage sites were discovered on December 5,1997.
The system engineer discussed the situation with the insulator's NAESCO supervisor who said he would investigate what happened and get back to the RCS system engineer. The RCS system engineer never closed out the WR, and received no information from the insulator's NAESCO supervisor. The RCS system engineer's supervisor did not follow up to see if the engineer wrote an ACR or check on the status of the open work request on the RC-V-89 inlet pipe firewrap. It was a busy time, refueling outage OROS had just concluded, everyone had a lot of paperwork to review and the priority of addressing this issue was low.
It was viewed as only a housekeeping and cleanup issue.
On June 21,1997,just prior to the close-out of containment following OR05, the boron buildup at the top and bottom of the RC-V-89 inlet pipe was cleaned. (note:
this is the first documented cleaning at RC-V-89) No water or moisture was evident during the cleanup of the boron buildup.
4.1.6 Third Ouarter 1997 Observations Sometime during the third quarter of 1997 HP technicians and supervisors tracked the status of the condition at RC-V-89 and they noted that boron was building up again on the top and bottom of the inlet pipe firewrap. An HP supervisor brought a picture of the RC-V-89 inlet pipe, taken by an HP technician, which showed boron buildup to the HP Department Manager.
The HP Manager had a conversation with the Maintenance Manager, Assistant Operations Manager, Oversight Mana;;er, Mechanical Maintenance Department Manager and the Operations Department Technical Supervisor. The HP Manager told the other managers durm~ g this discussion that his department personnel had identified the boron buildup that was reappearirig on the RC-V-89 inlet pipe firewrap. The HP Manaler showed the other managers the photograph given to him by his department supervisor. The HP manager believed that this picture represented the retum of boron buildup since the valve was cleaned on June 21, 1997. He told them that the boron was again building up in the same spot but did not make it clear that this represented a new buildup. This was the first time that they had heard about this condition.
DRATTRevhlon 3 V7M 9
.=
y The managers then discussed the condition with the valve test engineer. At this i
time none of the managers had heard the explanation of the boron buildup and for-l"
. all of them except the HP_ Manager, it was the first time they became aware of the boron buildup associated with RC-V-89.
-l When they showed the valve test engineer the picture, he told the managers his explanation of the cause and history of the boron buildup. He explained to them about the water spillage on the valve during the valve's removal for testing and about his theory of boron buildup. He told them that it occurred during every -
outage, that it was not a problem and that he and the system engineer were.
U monitoring it.
o When the HP manager told the valve test engineer that it had been cleaned on
. June 21,' 1997 the valve test engineer repeated his contention that the water spilled during OR05 could continue to evaporate out of the wet insulation, (firewrap) and cause new boron buildup even after it had been cleaned.. All of the managers
- accepted the engineer's explanation. They knew that the engineer had a lot of hands on experience with this particular valve and piping and felt that his.
l explanation was feasible. Therefore, none of the managers asked for any follow-up or pursued any other explanations for the boron buildup The engineer left and did not mention the meeting to his supersisor or to the RCS system engineer. The RCS system engineer when told of this meeting, after the plant shutdown, said
[
that he would have inspected the area at RC-V-89 for evidence of boron buildup within a few weeks of knowing that it had been cleaned. He was not informed -
that it had been cleaned on June 21,1997.
4.1.7 October 1997 Observations The HP Manager relayed the information he had received from the valve test engineer to his department personnel and the HP technicians continued to monitor conditions at RC-V-89. An HP technician was assigned the task of cleaning RC-V-8S a October 3,1997. HP documented the cleaning on June 21 but they did not document the cleaning on October 3,1997. Two pictures were taken after th?
October cleaning. An HP Supervisor called the RCS system engineer shortly after
' the RC-V-89 inlet pipe was cleaned and told him that they had cleaned the valve r-and notified the Shift Manager. The RCS system engineer asked him to monitor the valve's piping condition. The RCS system engineer agreed to wait a few weeks and then go into containment to monitor the status of the valve. This was the first tirne the RCS system engineer recalls being notified that the boron i
buildup at RC-V-89 had.been cleaned.
L 4.1.8 ' December 1997 Observations b
z On December 3,1997, as a follow up the Octo er conversation with HP, the RCS e
system engineer went out to investigate the status of conditions at RC-V-89.- He Ediscovered that the boron buildup had reappeared. On December 4 he sent an E-
^ Mail message to various department supervisors and managers advising them that
" DRMTneuen i IME 10 a-.
since the boron buildup had been cleaned and reappeared that either the water was
" wicking" up from the soaked insulation on the 12" RHR pump suction line, below the RC-V-89 inlet piping, or there was a leak under the firewrap.
The engineer brought the situation to the attention of his supervisor and manager during their morning department meeting on December 4,1997. The Technical Support Manager informed the management team on December 4,1997, at the Station Directors morning meeting, that another containment entry would be necessary to remove the firewrap and inspect the piping under it. There was an incore flux map scheduled for that day. That fact, together with the understanding that the piping was low pressure and isolated from the RCS in a standby condition, resulted in the containment entry being scheduled for December 5, 1997.
On December 5,1997, a containment entry was made and the firewrap was removed from the RC-V-89 inlet pipe. The pipe was wiped clean and after a few minutes 4 small drops of water appeared on the surface of the pipe. There were no noticeable cracks or pinholes or any other visual signs of uamage or irregularity on tlie pipe surface. The control room was immediately notified. A meeting of station management was held and it was decided that since there were no isolation valves between where this piping was and where it connected to the RL ST and RHR pump suction outside containment, that Technical Specification 3.6.1.1 Containment Integrity was no longer satisfied. The associated Action Statement was entered and within I hour plant power was being reduced. The plant was brought to Mode 3 ar,d then Mode 4 and Mode 5 within the time limits imposed by the Action ftatement.
5.0 CHRONOLOGY Datemme Description l
1988 or 3" Piping below, (upstream of),1-RC-V-89 was wrapped with a thin layer of before firewrap like material and covered with red duct tape. The 3" piping was captured in a photograph taken in 1988 with no boric acid present. Refer to Figure 2.
6/2/87 RC-V-89 discharge flanges had boric acid residue. Found during hot functional testmg. WR 87W005122 initiated.
10/19/87-RC-V-89 removed, inspected, cleaned, and reinstalled, per WR 87 WOO 5122.
I1/29/88 RC-V-89 inlet and discharge flanges had boric acid residue. Found during startup testing. WR 88 WOO 5991 initiated.
10/6/89 RC-V-89 removed, inspected, cleaned, and reinstalled per WR 88W005991.
DRAFTRevhion $ H248
}l
Datemme Description 12/16/90 WR 90 WOO 6166 initiated for boric acid leak at flanged connection on RC-V-89.
Boric acid was present on top of 3" inlet pipejust.below inlet flange but not present on lower portion of pipe at branch connection to 12" RHR suction header. Work transferred to 91W001930 which performed relief valve setpoint test.
7/26/91 OR01 Cooldown from Mode 4 to Mode 5 using B Train RHR 8/26/91 i RC-V-89 unbolted and rotated for OR01 Type C LLRT on discharge piping, (91RE0012601).
8/29/91 WR 91W001930 performs RC-V-89 setpoint verification, cleans boric acid residue and corrects flange leak reported earlier on 90W006166.
9/8/92 OR02 Cooldown from Mede 4 to Mode 5 using B Train RHR 9/29/92 RC-V-89 unbolted and rotated for OR02 Type C LLRT on discharge piping, (92RE0012601).
10/1/93 Forced Outage Cooldown from Mode 4 to Mode 5 1/30/94 Forced Outage Cooldown from Mode 4 to Mode 5 4/10/94 OR03 Cooldown from Mode 4 to Mode 5 using B Train RHR 4/18/94 RC-V-89 had heavy boric acid residue at side plug hole. WR 94001373 written and work was transferred to WR 93 WOO 4346 which performed Type C LLRT on RC-V-89 discharge piping in OR03.
5/12/94 RC-V-89 removed from system for relief valve setpoint testing,(94RE00116001).
5/17/94 OR03 Type C LLRT performed on RC-V-89 discharge piping,(94RE00126001),
prior to reinstalling valve.
7/20/94 ISI Functional Test on RHR Train B piping perfomled. Pi. ng below RC-V-89 t
visually inspected with RH/RCS pressure at 300 psig. No leakage or boric acid residue was noted in the area of RC-V-89.
I1/4/95 OR04 Cooldown from Mode 4 to Mode 5 using A Train RHR 11/23/95 RC-V-89 was removed for relief valve set pressure surveillance per RTS 95RE00116001. Piping and insulation below the valve became wetted with reactor coolant when the valve flange mechanicaljoint was unbolted.
DRAFTRevision $ 1/788 12
e-Date/ rime Description 11/23/95 OR04 Type C LLRT performed on RC-V-89 discharge piping,(95RE00126001),
prior to reinstalling valve.
3/20/96 Boron residue on RC-V-89 inlet pipe noted during Operations containment walkdown and noted on ODI-36 5/17/96 HP makes an entry on the boric acid RTS noting a boron buildup at RC-V 89 i1/96 RCS system engineer informed by HP of boric acid residue at RC-V-89. The system engineer and HP inspects the valve and finds boric acid residue on flange, fasteners arid pipe below valve. Several photographs,(Figure 3), of the RC-V-89 inlet pipe are taken. The boric acid residue is removed from flange and fasteners.
Flange was found dry with no evidence of any more leakage. The matter is discussed with the IST valve test engineer and the history of water from valve disassembly wetting insulation during OR04 is recounted. It is assumed that this was source of the boric acid residue.
2/97 RCS system engineer is told by HP of boron on valve. As IST valve test engineer is going in containment, the RCS system engineer gives him the picture from 11/96, (Figure 3), and asks him to check it out. The IST valve test engineer reports no change from picture.
3/18/97 ISI Functional Test on RHR Train B piping performed by VT-2 QC inspector. RC-V-89 and associated piping is visually inspected with RCS pressure at 240 psig.
Boric acid residue was noted at RC-V-89 inlet flange and below the valve at the weld-o-let connecting the RC-V-89 branch line to the 12" RHR pump suction header.
3/18/97 Engineering evaluation performed by ISI functional test supervisor to accept questionable conditions as is. Boric acid residue attributed to outage activities that drain the RC-V-89 associated piping when the valve is removed for testing. These activities had caused wet insulation below RC-V-89 in OR04, (11/95 - 12/95).
3/18/97-ACR was initiated by VT-2 QC inspector to document the poor work practices in OR04 that had caused the wetted insulation and had not reported or corrected the condition. The QC inspector was shown a 11/96 picture of RC-V-89 including the inlet piping and concluded that the condition did not appear to have significantly changed.
3/26/97 >
Work request 97W000833 was initiated by ISI functional test supervisor to address boric acid residue coming out of the RC-V-89 fire wrap on the inlet piping.
Durrneashms vzu
'13
Dateffime Description 3/28/97 RCS system engineer reviews WR 97W000833 and adds instructions to clean boric acid residse and remove damaged insulation,(reddish color).
5/10/97 ON5 Cooldown from Mode 4 to Mode 5 - B Train RHR 5/17/97 WR 97W000833 work performed by insulators. Material wrapped around 3" pipe below RC-V-89 was not considered by the insulators to be insulation and was not removed. Wetted insulation around 12" RHR suction header is replaced per 97W000833.
5/28/97 HP Turnover Sheet notation mr.. concerning boric acid re;idue on inlet pipe to RC-V-89. Technical Support notified.
6/2/97 Type C LLRT performed on RC-V-89 discharge piping. Valve Hange is unbolted and rotated in place to perform the test. Piping and insulation below the valve may have been wetted when the valve Dange mechanicaljoint was unbolted.
6/15/97 During containment close-out, RCS system engineer discovers fire wrap like approx.
material,(reddish wrapping), not removed and baron not cleaned. Informs supervisor who requests that an ACR be written. RCS system engineer talks to CS supervisor responsible for insulators but did not generate an ACR at that time. The CS supenisor states that he will follow up on the problem. RCS system engineer holds WR 97W000833 open pending further evaluation.
6/21/97 HP technician cleans boric acid residue off of pipe.
Third Qtr HP technician checks pipe and finds boron has retumed. Informs department 1997 supenisor and manager.
Third Qtr llP department manager discusses with maintenance, oversight and operations 1997 supervision to discuss issue. The IST valve test engineer is asked about RC-V-89.
The IST valve test engineer gives baron wicking / buildup theory and his explanation is accepted. (Refer to Figure 4).
10/3/97 An HP Technician finds boron on pipe, cleans it takes pictures and tells RCS system engineer and operations. (Refer to Figure 5).
I1/20/97 An HP Technician notices boric acid residue growth on pipe and thinks it should be cleaned on next rounds He does not tell anyone of this.
12/3/97 RCS system engine.:r inspects pipe as followup to HP Technician's earlier communication. Obterves the condition of the boric acid residue and alerts supervisien and management of his findings.
DRAFTRevision $ in98 14
l f
Datemme Description 12/4/97 Potential leak at RC-V-89 is discussed at Station Director's meeting and a containment entry is plarmed for 12/5/97.
12/5/97 Insulation sadiape removed from 3" pipe. Thru wall pipe weepage site is found.
12/5/97 Event Team formed 11:00 12/5/97 SORC meeting held to discues Opuability issues associated with the 3" piping 11:30 upstream of 1-RC-V-89 12/5/97 Entered T.S. 3.6.1.2.
I1:52 12/5/97 Commenced power decrease at 25% per hour.
12:00 12/5/97 Entered Mode 2 17:25 12/5/97 Entered Mode 3 17:56 12/5/97 Commenced cooldown using RHR Train A 22:48 12/6/97 Entered Mode 4 07:24 12/6/97 Entered Mode 5 15:10 DRATTRevision 3 inM 15
1 l
4 l
l FIGURE 2 l
l
+
q i
[k ~S
{(i s.
,+
)
,.. h' 4
- .M 4
g
.r e A
4 s.$
g _.
I p
- c..
. ' 4.g f
' ~ ~. r;, _
. :J,
- PC d 6-M EST. 19X1 N 6F SL![C DK4FT Revision $ In,93 16
. - ~.
1 i
i FIGURE 3 1
A I
i i
\\
.v.,<-
+ -+
z,..
~
I!
I l
4 i
v.
o q
l i
g.
'i
- ?
_. m l
g\\'
t
\\
fg
.~
~
l p
4
~
a.
DR 4FT Raision 5 Iliv8
}7 I
+-
-wyw--w-- -,
- w.v--
w w
r mcrw wv-w-
-w-ww-9 y-=-g-w
==vv-ew m
FIGURE 4 i
~s is m?
t,.,
I di 4
P...4 en
.e t.
- .a J;
vs.,y i
A..
.w
.et e
4hm t\\ u i
g 1
EV i
z:
A l
- Ng. ;w n '
- . tr* *. *; t.r3
- s:-
- ~
1
.; d.;p; ;Q yp J,\\-
.y
..a
<.= a;_~.;-_h. i.g f 2: x..g,
.g:\\
,c.
4
.~
.~
I jV
. 7 ',k k
8 ame.
..x es wa n - w.
a z u:a.
.us..,a =www::
1 FIGURE 5 l
rx~- yg.;ggs 3 9; y.yy, ;,. 3my we:
1 1
I p
.p l
4
, -G g
.i y
.x 5,'
- 5 s
f l
+
n m
a j
I i
l 1
i'
'y i
s 417
' s 3.
9 i
+
- i
..s k
c.
\\
m 2
+
f5M
\\
y
/
'. e,tLNQTs -. -;
- Nn, e
/'ft....' 'J
~
f h
. > :.u r'.
n ha e
+
, g,~
i 4
l 1
l l
DRAIT Revision.%
1598 18 i
....c
6.0 DESCRIPTION
OF EOUIPMENT RESPONSE
6.1 BACKGROUND
RC-V-89 is a relief valve off the 12 inch suction line to 'B' Train RHR Pump 8B. The spool piece, Line # RC-88-1-601-3, connecting the 12 inch line to the inlet of the relief valve had exhibited weepage from sites in two separate areas. Specifically, it was observed in the heat affected zone (HAZ) of the weld at the weld neck flange on the inlet to RC-V-89 and also on the HAZ of the field weld on the spool piece at the weld-o-let on the 12 inch suction line. The spool piece sits in the venical direction and is approximately 5 inches long. It is 3 inch diameter schedule 40S, type 304 stainless steel pipe. Per Specification 249-1, the spool piece was to be insulated with 1" fiberglass insulation for heat retention (Note: At this time the insulation is not installed for the spool piece to RC-V-89 nor the identical spool on Train 'A' at RC-V-24). The weepage was identified when the plant was at 100% power with RHR in its normal standby condition.
During lower operation, this line is essentially stagnant at a temperature of approximately 100 F and 50 psi (static head of the RWST). During startup and shutdown it may be exposed to a temperature in the range of 300-350 F and pr:ssures between 200 and 300 psig.
6.2 AS FOUND CONDITIONS On December 5,1997, system leakage of a drop every two to three minutes (weepage from four locations) was discovered on the inlet piping to the Residual Heat Removal pump 8B suction relief valve (RC-V-89). The weeps were in the metal adjacent to the welds (heat affected zone) that connected the piping spool piece to the relief valve flange and the 3" to 12" weld-o-let that connected the pipe to the 12" RHR pump suction line.
One weepage site was in the upper heat affected zone and three were in the lower heat affected zone.
Visual examinations of the spool piece indicated that it had been wrapped with a layer of a woven cloth fire wrap directly against the pipe surface and standard red duct tape around the wrap. The condition of the covering indicated that the materials were somewhat discolored and appeared to be fused or matted together. Seabrook Chemistry Department took samples of the material ( both the fire wrap and the tape collectively) and performed a chemical analysis which identified 349 ppm chloride and <24 ppm fluoride (minimum detectable).
During outages, relief valve RC-V-89 is removed from the system for testing. In the process the discharge line flange is unbolted. This releases water from the line some of which is not collected and wets the 3 inch spool piece, its covering and the 12 inch suction line and its insulation. Although the present chloride readings for the covering materials are 349 ppm, they may have been higher in the past. The covering material was on the pipe since at least 10/88. Removal and testing of RC-V-89 along with subsequent wetting of the area had occurred in the past.
DnAFTsoms ums 19
9 6.3 HISTORICAL REVIEW OF ASME MATERIALS AND WELD DOCUMENTATION Original radiographs of the two welds on the spool piece were retrieved and examined by NDE Level Ill personnel. The review showed no anomalies in the weld or the base metal. It was also noted that the two welds on the spool piece were performed by different welders in different companies. The flange to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o-let to pipe weld was a field weld performed at Seabrook. This indicates that the welds and welder qualification wt:re not a contributor to the problem as evidenced by the fact that the radiographs showed no anomalies, the welds were performed by different welders / welding operators, and examined by different NDE inspectors, yet the heat affected zones adjacent to both welds had weepage sites. Also, there did not appear to be any anomalies in the base metal of the pipe which further indicates that the weepage was due to in-service conditions rather than original installation.
A document search of the heat number fvr the spool piece was performed to determine where else piping of that heat may be installed in the plant. (Appendix B). A chemical analysis completed at Mass Materials Research confirmed that the material was SA312, type 304 the material specified and the correct material for the applications.
6.4 SERVICE LOAD CONDITIONS A review of the piping analysis shows that the stresses in the spool piece piping are minimal During RHR operation, the combined str,:sses have a safety factor of 4.75 on the normal design allowable. During power operation with RHR in standby, the safety factor on normal design allowable is about 28.0. The piping design in this area is dictated by the faulted condition with stresses induced by a Loss Of Coolant Accident. In addition, there have been no indications of any mechanical or flow induced vibration in this area, nor has there been any problem history with relief valve RC-V-89 that would have resulted in abnormally high thrust loads. To date, there does not appear to be any high service induced stresses that could initiate any failure mechanism which could result in the weepage found. The only area on the piping spool piece that has been exposed to high stress levels is in the heat affected zones of the welds due to residual stresses from welding.
6.5 CORROSION MECllANISMS While there does not appear to be any sersice induced mechanism which could have comributed to the weepage, a combination of the residual stresses in the pipe and environmental conditions (i.e. chloride contaminants) on the outside of the pipe are conditions that make the pipe susceptible to Stress Corrosion Cracking (SCC). Although the chloride level in the wrapper is relative low at 349 ppm, this in itself would be sufficient to initiate the stress corrosion process when coupled with the high residual stresses located in the heat affected zones adjacent to the weld. The higher the residual stress, the more sensitive the material is to smaller amounts of chloride. The chlorides would also tend to concentrate at crevi:es on the surface due to the washing of the DRAFTRevhlon 3 UM8 20
covering material by the water and subsequent drying off of this water when the line goes back up in temperature. This could result in potentially high levels of chloride concentrated in selective areas around the heat affected zone of the welds.
6.6 TESTS AND INSPECTIONS Tl
' piece was removed and a metallurgical analysis was performed to determine the uuse and extent of degradation that led to the weepage. The piece was removed by making a cut at the weld of the weld-o-let to the spool piece and unbolting the inlet flange at RC-V-89. The analysis involved one section of the spool piece being analyzed r.
Seabrook and the other section analyzed at ABB Laboratories in Windsor Locks, Connecticut. The piece was cut approximately 31/2 inches down from the top of the flange. The top piece was analyzed at Seabrook and the bottom was shipped to ABB Labs. Subsequently, the upper section was also sent to ABB for analysis.
Prior to sectioning the removed piece, a leak rate test was performed on the piece to determine what effect the degraded section would have had on containment integrity.
The test consisted of clamping a gasketed plate over each end of the piece. One end was fitted with a test connection which was pressurized to 52 psi. Th. ieak rate was determined to be 1.363 Standard Cubic Feet per Hour (SCFH) which is much less than the allowed value for a typical penetration which is on the order of 37 SCFH. It was also noted that the leakage appeared to be coming from the gasketed end plates. Results of the test indicate that the pipe spool had not degraded to the extent that containment integrity was challenged.
The following description depicts the work and examinations performed on the section analyzed at Seabrook and the conclusions reached from that work:
A Liquid Penetrant (LP) examination was performed at the weld of the flange to the 3" pipe on the inside and outside diameter of the piece. One linear indication was found that went through the wall. The crack ran at an approximate angle of 45 to the longitudinal axis of the weld. It progressed into the base metal of the 3" pipe. The piece was sectioned at a location perpendicular to the weld (See Figure 6). The sectioned surface displayed a branching type crack running perpendicular to the surface of the pipe at a distance approximately 3/8 of a inch from the weld in the heat affected zone. It appears to have originated on the outer surface and progressed to the inside diameter of the pipe, i
l DJtAITRMslon 3 1/7/98 2}
i 1
FIGURE 6 i
i W
\\
'9 s
i i
9 e
f
.x t
l P
+
l l
e A
124T7 l
Note: The branch cracking in the piping cross-section shown in Figure 6 above has been magnified by acid etching. Without etching the actual defects would not be visible in the photograph.
l i
l The failure appears to be a corrosion assisted degradation process such as chloride induced stress corrosion cracking that had initiated on the outside and migrated to the i
inside diameter. However, it was decided that more confirmatory information was
(
required from the analysis being done at ABB on the lower section of the spool piece.
)
Additional needed information includes: obtaining the chemical composition of the pipe, l
performing an SEM (scanning electron microscope) examination on the crack surface and l
performing a chemical aralysis of any contaminants that may be present on the crack I
surface.
l r
l i
I i
l i
i i
l l
DRMT Revision 5 1/7,98 22 r
-,,.,,-.-~,c...w.,r..
--,.-,c,----------,.
l l
After removal of the 3" spool piece, a LP examination was performed on the surface of the weld-o-let in preparation for its replacement. The examination revealed branching type indications similar to those seen on the sectioned piece. Considering the fact that the indications appear to be due to the same mechanism (i.e. chloride induced stress corrosion cracking), it was decided to also perform LP on the 12" suction line as this area may also have been exposed to leached chlorides. Some indications were found in the heat affected zone of the weld on the 12" line where it attaches the weld-o-let. The indications were repaired. However, the indications on the weld-o-let were more extensive and could not be repaired. The weld-o-let was removed and replaced with a new fitting.
Additional analysis and testing of the 3" pipe spoolincluding work performed at ABB Labs are included in Engineering Evaluation SS-EV-970031,(Attachment D).
7.0 CONCLUSION
S 7.1 The through wall weepage in the heat affected zones of the 3 inch pipe below RC-V-89 was determined to be chloride induced stress corrosion cracking. This cracking initiated on the outside and migrated to the inside diameter.
7.2 The source of the chlorides was the inappropriate material, (firewrap and duct tape) installed on the RC-V-89 inlet pipe. The combined effects of wetting the firewrap, the presence of chlorides, inservice temperatures, and the residual stresses in the piping heat affected zones were all necessary to produce the degraded condition.
7.3 The condition developed over a long period of time,(i.e., years), following the completion ofinitial plant construction. The welds and welder qualification were not a contributor to the problem as evidenced by the fact that the radiographs showed no anomalies, the welds were performed by different welders / welding operators, and examined by different NDE inspectors, yet the heat affected zones adjacent to both welds had weepage sites. Furthermore, similar piping configurations with similar inservice conditions, (but without the presence of inappropriate chloride bearing material and wet external insulation), did not show signs of stress corrosion cracking.
7.4 Results of the leak rate test, performed on the removed 3" spool piece, indicate that the condition ofit was not degraded to the extent that containment integrity was challenged.
7.5 Based on the chronology the best estimate for the development of a through-wall weepage site in the RC-V-89 inlet pipe was in the March 1996 time frame.
7.6 The bori: acid reduction program and the inservice inspection program failed to identify the RC-V-89 inlet pipe weeps in a timely manner.
I l
DRAFTRevhlon 5 Im98 23
7.7 Reaction time to problems identified by the Corrective Action Program are in need ofimprovement. 'lhete were missed oppoitunities to identify, document, evaluate, and resolve the issues prior to December 5,1997.
7.8 Many examples of a " questioning attitude" were observed during the course of this event evaluation, llowever, the followup validation and verification process was not used.
7.9 There is a wide spread lack of knowledge of the problems associated with the introduction of unauthorized material onto stainless steel piping & components.
7.10 No similar situations to those described in conclusions 7.1 and 7.2 were discovered based on the s ctions taken per recommendations 8.1.1 and 8.1.2.
7.11 The R11R and CBS systems were capable of performing the containment integrity, shutdown cooling, emergency core cooling safety functions and containment building spray with the piping section containing the identified defects.
8.0 RECOMMENDATIONS 8.1 PRIOR TO IIEATUP (MODE 4) 1 Conduct a point of origin boron leak detennination inside containment using ES 1801.006.
Include in this inspection a look for temporary insulation and /or fire Hanket material wrapped around stainless steel piping Responsible Individual: Peterson Category: P Due Date: Complete 2
Conduct a point of origin boron leak determination outside centainment using the station boron leak list. Include in this inspection a look for temporary insulation and /or fire blanket material wrapped around stainless steel piping Responsible Individual: Peterson Category: P Due Date: Complete 3
Evaluate the results of the above point of origin boron leak determinations. Ensure that all identified active leaks are repaired and all boron buildup is verified to be from leakage from mechanicaljoints. (no thru wall)
Responsible Individual: Peterson Category: P Due Date: Complete 4.
Conduct liquid penetrant testing on the heat affected zone of the weld-o-let to 12 inch pipe weld associated with the RC-V-89 spool.
Responsible Individual: Grillo Category: P Due Date: Complete l
DLUT Reshion 3 1/7,98 24
5 Canduct NDE on the heat affected zones of the welds on the RC V 24 relief valve branch header to include the 12 inch pipe to weld-o let, weld-o let to 3 inch pipe and 3 inch pipe to flange.
t Responsible individual: Grillo Category: P Due Date: Complete 6
Conduct c visual inspection, using NDE level 111 personal, of the removed section of the RC-V-89 reliefepool piece.
Responsible Individual: White Category: P Due Date: Complete 7
Perform a metallurgical failure analysis on the removed piece of the RC V 89 relief spool piece. Evaluate the results of the failure analysis to determine if the failure mechanism is applicable anywhere else in the plant.
Responsible Individual: White Category: P Due Date: Complete 8
Conduct a Chloride and Fluoride leachability test on the material that was wrapped around the RC-V 89 relief spool piece.
Responsible Individual: Linville Category: P Due Date: Complete I
9 Review the work control and corrective action data bases far similar cases ofinappropilate materials on stainless steel piping. Address any identified issues.
Responsible Individual: Peterson Category: P Due Date: Complete 10 Repair / Replace deg aded piping at the inlet flange to RC-V 89.
Responsible Individual: Peterson
- Category: P Due Date: Complete Il ' Complete an engineering evaluation of the degraded condition to determine if the RHR and CBS systems would have been capable of performing their safety function.
Responsible Individual: White Category: P Due Date: Complete 12 Complete the independent reviews of the VT-2 inspections.
Responsible Individual: Beuchel Category: P Due Date: Complete i-l; 1
l ORMTneunsInm 25 l
., ~ -,,,
-~.
8.2 PRIOR TO STARTUP (hiODE 2) 1.
Ensure the VT2 inspectors performing the inspections and test supervisors performing evaluations required by recommendation 8.2.2 have been briefed on this event.
Responsible Individual: Beuchel Category: P Due Date: Prior to Mode 2 2.
During mode 3 at NOP/NOT perform EX1810.001 RCS functional test Responsible Individual: Beuchel Category: P Due Date: Prior to Mode 2 3
Reinforce the expectation and reason why an ACR must be written for conditions that are outside of normally expected conditions. Communicate the importance of a questioning attitude when accepting explanations for unexpected conditions. Use this event as an example.
Responsible Individual: Feigenbaum Category: P Due Date: Complete 4
Complete an evaluation of the corrective action program to determine if any changes are necessary prior to startup.
Responsible Individual: Mal iH Category: P Due Date: Complete 8.3 SHORT TERM (< 90 DAYS) 1.
Complete a root cause on this event.
Responsible Individual: Makowicz Category: P Due Date: January 31,1998 2
Conduct confirmatory laboratory testing to validate the exact failure mechanism. (ABB final report)
Responsible Individual: White Category: P Due Date: January 31,1998 3
incorporate the infonnation recorded during the inspections performed as a result of 8.1.1 and 8.1.2. into the boric acid monitoring program.
Responsible Individup!: Beuchel Category: P Due Date: March 15,1998 4
Generate an EWR to add a drain line to the relief discharge lines for RC-V-24 & 89 Responsible Individual: Kline Categoiy: P Due Date: March 15,1998 5
Evaluate the roles and responsibilities in the work control close-out process to determine if it is appropriate to have non supervisory personnel sign the[ supervisor signature block] on work requests arJ RTSs.
Responsible Indi idual: Makowicz Category: P P
we: March 15.1998 DRUTRedston 5 M98 26
8.4 LONO TERM (> 90 DAYS) i Enhance the boric acid reduction program using the results of the root cause and j
benchmarking with other plants.
Responsible Individual: Beuchel Category: P Due Date: May 22,1998 1
2 Develop a program to increase the awareness of station and contractor personnel concerning the use of unapproved materials on stainless steel.
Responsible Individual: Linville Category: P Due Date: May 22,1998 '
i 3
Buy new photographic equipment for the lip department Responsible ir dual: 1,inville Category: P Due Date: April 10,1998 4
Evaluate the effectiveness of configuration controls that are applied to the removal and reinstallation of thermalinsulation.
Responsible Individual: White Category: P Due Date: June 30,1998 5
Develop a station policy for addressing wetted insulation.
Responsible Individual White Category: P Due Date: June 30,1998 6
include a discussion of this event in the OR06 briefing package. This package should include, as a minimum:
The executive summary of the event evaluation report The information from recommendations 8.2.4,8.3.1,8.4.2,8.4.4, and 8.4.5 Responsible Individual: Anderson Category: P Due Date: December 31,1998 7
Evd=t: the practices of draining systems that could result in wetting insulation.
Responsible Individual Peterson Category: P Due Date: May 22,1998 DRAFTRedelen 5 h7M -
27_-
y-w
,-e
,,,y W
y,. -..,... - >
.y
....,.*,w-.vr-,,y y,-,-7 r"T-
---v+v
+
.-. ~. - _.
-.. - ~ ~ -.. ~
.. -.--.. ~
+
i 9.0 EVENT EVALUATION TEAM j
Event Team Manager:
Jerry Peterson Event Team Leader:
Gregg Sessler l
Event Team:'
Brian Brown Michaelliarrington j
James liill Richard Julian -
I John Lavoie Romas Schulz -
10.0 ATTACliMENTS l
- A)
Corrective Action Program - Failure Scenario B)-
Documentation Illstory C)
Industry Experience D)
Engineering Evaluation SS EV-970031 l
E)
. Engineering Evaluation SS EV 970030 Evaluation of Safety System Function
+
t I
T t
t l
3
. D M FT a n M en s tit w 2g.
- m-9 4
y s
.-,--.-,,,._.-%.v,
--.,w--
s-,-,...,,-
y..-4,-,
w.....,.s,-,-,
ATTACHMENT A Failure Scenario - Corrective Action Program
' Note: Thepurpose ofthis evaluation is to determine ifchanges are neededto the Corrective Action Programprior to startup. This has set been determinedto be the root cause ofahis event. The root cau e evaluation uillcontinue. Issues relatedtoprocedures' rocesses.
l p
accountability, lateralintegration, etc. will be comideredin this evaluation. Also note that the expectation listedin this table is the current expectation. andnot necessarily the expectation Ihat existedat Ihe time. Thepurposefor using ahe cur c~sexpectation is to ensure that Ihe currentprogram is adequate.
4 Date Activity Current Expectation Current Expectation Met?
19/88 This insulation has been installed since at least 1988. No Unexpected conditions are reported and No. Multiple I
(or-one who saw it ever questioned the fact that it looked evaluated.
organizations did before) difTerent than normal insulation.
not question this until covering or that it 12/5/97 was only on 'B' train.
OR04 Test engineer knows of water drained from RC-V-89 Adverse cor.ditions shall be reported.
Yes.among some relief discharge flange onto 12" suction line wetting documented, analyzed, and/or trended people. There is not insulation. This condition was not viewed as unacceptable a cicar expectation at this time.
that wetted insul. be reported 3/96 Operations identified boron on RC-V-89 inlet piping and Follow normal work control process as No. Should place documented it on the containment walkdown list (ODI-directed in ODI-36.
on RTS or write a 36), which wss the expectation at the time.
work request.
Ili96 IIP technician identifies boron buildup at RC-V-89 and Per the requirements of the boron Yes, by existing notifies his supervisor. System engineer was notified of program, notify th<: program support boron program.
problem. (No ACR or WR generated, owneship of engineer of boron from non-mechanical His should bc l
problem transferred to system engineer) sources.
evaluated for rev.
._.,.m s.-.-.
O Date.
Activi:y Current Expectation Current Expectation Met?
System engineer inspects and finds no evidence of Adverse conditions shall be reported.
No. With the point leakage. Takes picture and discusses with test engineer. Is documented, analyzed, evaluated, and/or of origin unknown.
i told of spill in OR04. Origin of boron not established.
trended.
System Engineer
[
(Consistent with the program at the time no WR or ACR should document generated, supervision not informed) Wrong assumption on an ACR or %R.
(MJS) was made that boron was from spill theory by test Also, the non-t engineer and system engineer. Other pessible sources, standard insulation including through wall Icakage, were not evaluated. Once and the fact that f
it was determined that it was from the spill, no one put the only one train had pipe on the RTS as a housekeeping item.
insulation on the 3' pipe should have been pursued.
f 237 IIP notifies system engineer to again evaluate the boron Follow up on a presiously identified Yes. Actions were
[
buildup. Ilas seen no evidence that anything has been problem consistent with done.
boron p up.
267 System engineer discusses with test engineer, gives test Adverse conditions shall be eqv tc4 No. Again the engineer 11/20/96 picture and test engineer enters documented, analyzed, evaluated, and/or deficient condition
(
containment and compares current condition with picture.
trended.
is noted and not i
Test engineer already has mind set ( AU4) that boron is due reported cc L
to spill during OR04. Told system engineer no change.
evaluated.
j 367 VT2 inspector reports questionable or indeterminate Adverse conditions shall be reported.
Partially. The condition to own supersision and ISI supervisor. Checks documented analyzed, evaluated,and/or inspector mote an t
for moisture, finds none. Documents on RTS and writes trended.
ACR. but limited ACR 97-0583. VT2 inspector buys into spill / wicking the problem theory"and initiates ACR for OR04 poor work practice.
statement to poor i
Tunnel vision / group think (SK1) starts to become a factor.
work practice based
}
j m h tet i
engineer's theory,
[
and did not include suspect firewrap.
I i.
f t
DRAFTReshier 5 U7/98 3g i
t Date Activity Current Expectation Current Expectation Met?
. 367 ISI supervisor initiates work request 97WR000833 to Qintioning attitude is practiced in daily No. This
[
clean or replace insulation as required due to spill during decision making.
evaluation was OR04 theory. Evaluation written for ISI inspection based r the acceptancejustifying boron as non-leak related.
unevaluated theory.
Spill / wicking theory is basis. Tunnel sision/ group think (SK1)is present.
337 MRT reviews ACR 97-0583 (identifying OR04 poor work MRT will ensure that proper action is ParGally. Based on practices and failure to correct the deficient condition).
taken in response to a deficient the information ACR was dispositioned as trend only.
condition.
available to them I
this ar. fwd to be an isointed work t
performance issue.
However, the f
ii m 4 4 wiate
{
firewrap wm not questioned based on the photograph.
Insulators remove' and replaced insulation at base of Questioning attitude is practiced in daily No. The supersisor 537 d
relief valve connection to 12" pipe per WR 97-833.
decision making.
or system engineer Intended removal of reddish firewrap on 3" line not should have been I
performed due to vague work package which was not contacted.
questioned by the insulators. (DP) Document preparation issue. Did not question " reddish color" reference on work package.
I t
t DRAFTRevkimr 5 UMA 3l
Date Activity Current Expectation CurreC Espectation Met?
637 System engineer walks down containment and notices Adverse conditions shall be reported, No. Theengineer's reddish firewTap(insu'ation7) not removed. Told documented. analyzed. enluated, and/or supenisor directed i
supenisor and Construction Services supenisor about trended.
him (correctly) to perceived performance issue ofinsulators. Told by writean ACR.
f supenisor to initiate ACR. Did not write ACR, did not This was never close out 97W000833 because intended work not done. Also,the performed. Supervisor did not follow up on the ACR.
work could have i
tr:en performed at I
this time bitW on plant conditions.
& 97
, Boron cleaned by IIP on 6/21/97just prior to containment To clean boron and doc nnent.
Yes.
close-out.
737 Supervisor and ilP Manager notatied ofILD technician's Problems are identified. documeated.
Yes.
f identification of boron buildup on RC-V-89 inlet pipmg.
and corrected in a timely manner. If they j
can't be corrected promptly then interim measures are established l
737 IIP Manager meets with several station managers to Questioning attitude is practiced in daily No. De theory discuss the twon buildup on the firewTap that was decision making.
was not challenged.
l l
discover ed in July. Test engineer expiains history and An ACR or WR spill / wicking theory to managers pant. Managers accept Adverse conditions shall be reported, was not generated.
i test engineer's explanation. Tunnel vision / group think documented, analyzed, evaluated, and/or t
(SK1) is present. Further questioning or pursuit of follow-trended.
l up explanation did not occur.
1937 IIP technician again reports boron on the firewTap and Follows up on a previously identified Yes. (IIP) cleans it Supenisor and system engineer are notified. IIP problem supenision also notified Operations Shift Manager of the
[
situation. IIP doesn't wTite WR because aware that system l
engineer still has one open (97W000833).
i i
l i
L owTsen,s uns 32 E.
~
Date Activity Current Expectation Current Expectation Met?
12/3/97 System engineer inspects pipe as follow-up to hP report Adverse conditions shall be reported.
Yes. Informed of boron problem. Observes boron is back and alerts documented, analyzed, evaluated, and'or supervisor and management of findings. No ACR initiated by system trended.
drafled ACR.
engineer.
12/5/97 Firewrap and tape removed from 3" pipe. Thru wall weepage sites discovered. Entered T.S. 3.6.1.2 and comrnenced plant shutdown.
Conclusions His event represents a failure of the corrective action program to identify, document, evaluate, and resolve a deficient condition in a timely manner. Ilowever, this event does not represent a widespread breakdown in the program for the following reasons.
He fundamental problem conceming the corrective action program's failure to resolve this condition in a timely manner is that there wrre several missed opportunities to document prob!-ms using the corrective action program. These opportunities were missed based partly on a wrong assumption and misjudgment by the system engineer and a program support engineer. Evaluation of all possible sources of the boron (which would include through wall weepage) by the system engineer and a program support engineer would have identified the problem. A willingness to accept the opinion of the designated expert without challenge by other personnel, including managers, contributed to the organization missing opportunities to recognize a problem that required documentatior and management review in a timely manner.
He first missed opportunity was upon the discovery of boron on November 1996 with the point oforigin not established (a theory was proposed but there was no attempt to verify it). Again in February the condition was noted by IIP but the system engineer and the program support engineer continued to believe there was no Icak and that they knew the source of the boric acid buildup. In March an ACR was written, but its problem statement was limited based on the theory developed by the program support engineer. In June, when the WR written in March failed to result in the intended work being completed, an ACR was not initiated despite one being requested. This would have been appropriate since the work was performed on the wrong component (the insulation on the 12' pipe). Again in July and in October the condit on was again identified and communicated by IIP supervision. Given the explanation of the source of the boric acid buildup, no i
followup action was taken. Failure to document these problems removed an opportunity for the organization to ensure that the boron had been propenly evaluated.
DRAFTRerhime 5 U7M 33
A failure also occurred in the work control process when it was :ecognized that a work request had not been in ykiiwated as intended without any followup action.' The work request is still open.' This is an example where not ading the next question removed an qvoitwnity to resolve this ydkn; in a more timely manner.
In all cases there was evidence to suggest that there was not a leak present (no detectable moisture and no edlence, until July 1997, ofan increase in boron on the pipe). Ilowever, there us also no evidence of the point oforigin of the boron, only a theory that was never tested or verified (e.g., through cleaning and regular monitoring of the I' pipe). There was also no evidence that the reason unusual insulation" was on one train but not on the other was ever pursued.
Enhancing the organization's questioning attitude is an existing mission of the Corrective Action Propom. Past actions include information i
' passed in the Operating' Experience Summary and in Seabrook Today. Better mechanisms for transmitting Sd cok case studies (root cause evaluations performed on Seabrook events that show where a more rigorous questioning attitude could have prevented an event) are being investigated.
RecummendedAction Reinforce the expectation and reason why an ACR must be wntten for conditions that are outside of rmrmally expected conditions.
Communicate the importance of a questionic', attitude when accepting non-standard explanations for unexpected conditions. Use this event as a case study.
Action: Feigenbaum Due: Ccmplete
?
DRAFTRevMen 5 10M8 34
__. ~.
ATTACHMENT B Archived / Historical data retrieval for documentation associated with 1 RC V 89 i
The following documentation has been retrieved and retained for review associated with 1 RC-V 89 event. The documentation reflect the history of fabrication and installation of piping, including material traceability for comr 'ents utilized in their fabrication.
Dravo Spool Sketch E2936-22BD-Std WT. ShiLS Pipe, SA312 TP 304, HT# 04935.
e MTR# 411. Final approval of fabricatw. 'vas completed and signed off by Dravo, hiarietta, Ohio, on 4-7 82. This welded piping assembly consists of 3" spool, and a 3" RF WN FLG, with a weld designated as shop weld "C" to spool piece E2936-2280. Note: Radiography was performed on shop weld "C". Review of radiographs was performed 12/06/97 by NAESCO Level 111, no discrepancies were noted 'n the original radiogtaphs. Documentetion for Spool Sketch is located in SBYDCC on Roll 4113, Frame 310, ar.d in hard copy record.
Material Test Report # 411-3" Piping was supplied to Dravo, by hicJunki Corporation, from Al Tech Specialty Steel Corporation, Dunkirk, N.Y. hiTR# 411 indicates that two(2) heats of 3" piping were supplied, consisting of Ht. 04935 and lit. 04785, with a total quantity furnished of 408' 7/8". Document location in SBYDCC, CSN# 1862, Roll 768. Frame 1707.
NOTE, Additionally on 5/13/86 a 8' length of 3" Sch 40, Tp 304, Ht # 049' mMTR M11 was received by UE&C from Dravo on Receiving inspection Report, RIR#13029. Review of documentatinn it appears that this piping was part of the original purchase received by Dravo. Documentation for these records is located in SBYDCC, Roll 6774, Frame 1284.
Dravo Spool Sketch E2936-707-Item #4,3" on 12" STD. WT. Weld-O-Let, SA182-F304, llT# L342, hiTR.# 42. This is the 3" connection (1-RC-88-01, F0108) and Dravo Spool E2936 2280. Fabrication of spool assembly was completed Il-30-79, for Dravo shop weld "F"
Examination of final weld configuration was performed by PT examination on 10 79, with seceptable results. Documentation for Spool Sketch is located in SBYDCC hard copy files.
hinterial Test Report # 42-14-12 X 3" Weld o-let, HT# L342 was supplied to Dravo by e
G+W Bonne) Forge, Carlinville, Ill. Documentation of hiTR# 42 is located in SBYDCC, CSN# 726, Roll 330, Frame 411.
P/H Fiels! Weld Packnee-1 HC-88-01. F0108-Review of Pullman-Higgins field weld 1-RC 88 Ol,F0108, which consists of spool piece E2936-2280 and E2936-707, was completed 3-1 83. Radiography of this weld was performed originally 3-7-83, film discarded, due to unacceptable film (-15%) density. Weld was re-radiographed 715-85, and 7-20-83, with acceptable results. NOTE Radiographs were retrieved and a review performed on 12/06/97 by NAESCO Level 111, with no discrepancies noted.
Documentation for abcve weld package is located in SBYDCC, Roll 5333, Frame 1384.
P/H Welder Ounlifications. Stamp No. "RP" - Retrieved and reviewed welder o-
- qualifications "RP". Records were requested on 12/06/97. Documentation for these recerds
may be located in SBYDCC, Roll 4021/ Frame 1731, Roll 4197/ Frame 2355, and Roll 5825/ Frame 2075.
P/H As Bulle Verification Walk Down As built verification was performed and completed on 1-19 85, for 1-RC 88-01, BIP RC-M 3. Documentation for this package may be located in SBYDCC, Roll 5333/ Frame 1384.
Reggtitive Task Sheets for VT-2 Eumminations - The following RTS associated with the performance of VT-2 examinations conducted on piping system (s) related to 1-RC-V-89.
RTS# 94RE00185001 RTS# 95RE00185001 RTS# 97RE00185001-Identified Boric Acid residue in the area of 1-RC-V-89.
Work Requests Related to 1 RC-V-89 activities (Boric Acid Residue)
WRW 87W005122 RC-V 89 discharge flanges had boric acid residue. Found during hot functional testing. (SBYDCC. Roll 673, Frame 830)
WR# 88 WOO 5991 RC V-89 removed, inspected and reinstalled. (sBYDCC, Roll 1274, Frame 938)
WR# 90 WOO 6166 RC-V 89 flar.ge leak. (SBYDCC, Roll 2135, Frame 1130)
WR# 91W001930 RC-V 89 no lea' age identified. (SBYDCC,2190, Frame 1958)
WR# 94W001373 RC-V 89 leaking at side plug hole, heavy boric acid buildup.
(SBYDCC, Roll 3827, Frame 254)
WR# 97W000833 WR. initiated to address boric acid crystals on insulation and wrapping under valve. Instructions added to WR. to clean boric ac d residue and remove damaged insulation,(reddish color).
(WR. maintained in hard copy file)
Dravo Spool Retrieval by llT# 04935 and 11T# 04785.
Unit I(E2936) Dravo Spool Pieces I Spool Pe. #
Item #
Ht.#
Feet Inches Remarks E2936-406 1C 04785 1
E2936-462 1A 04935 18 9%
1B 04785 3
6%
IC 04785 1
95/8 E2936-466 lA
-04785 11 91/8 IC 04785 3
E2936-472 1B 04785 10 5/8 E2936-517 2A 04785 1
3 2B 04785 5
4 2C 04785 10 6%
E2936-523 1A 04785 6
8%
E2936-714 -
1 04785 13 715/16 E2936-715 IB 04785 3
11 DRAFTRedston $ inM8 36
v Spool Pc. #
Item #
Ht.#
Feet Inches Remarks 1C 04785 11 %
ID 04785 2
10 1/16 E2936-716 2
04785 5
10 15/16 E2936-719 2A 04785 4%
2B 04785 6
6%
2C 04785 9
2 9/16 E2936 720 1
04785 15 2 9/16 E2936 721 1A 04785 3
67/8 IB 04785 1
3 E2936-832 1A 047h5 4
1B 04785 9 9/16 1C 04785 8
~
1D 04785 5 7/8 E2936 833 2A 04935 2
8 3/16 E2936-930 1
04935 5
E2936-1019 1A 04785 11 15/16 1B 04935 20 57/8 E2936-1156 1
04785 10 2 15/16 E2936-l l57 1
04785 7
1 5/8 E2936-1158 1A 04785 2
4 3/16 l
1B 04785 9
3 E29361194 1A 04935 515/16 IB 04935 6
1C 04935 3
3 1D 04935 1
6 lE 04935 3
E2936-1204 1A 04935 3 11/16 E2936-1204 IB 04935 3
6 15/16 IC 04785 3
6 15/16 E2936-1522 1
04935 77/8 E2936-2147 1A 04935 3
IB 04935 4
E2936-2202 1A 04935 1
7 7/16 IB 04935 2
-10 7/16 E2936-2135 1A 04785 1
7 5/16 IB 04935 9
Total 205 11 13/16 Da.4rracwen,< vzSs 37 w
4 w
eew --
,., - - - - - ~ - ~
,.~,.,,,,-,,w,
,.----.,,,,.-~~w.,,--m
-y-,.-
.-,-e,-,-..,-3-,,,.-.---.,-,-,-.,.
.--y.
. = -. _
I ATTACHMENT C INDUSTRY EXPERIENCE ASSESSMENT Impet of Halogens on Siminless Steel Surfaces Surface contamination of stainless steel piping foreign material was the subject of a Westinghouse Technical Bulletin (NSD TB-80 2," Cutting Oil Used on Reactor Components")
reviewed by the ISEG in 1985. This resulted in recommendations (SSP #850025, dated 1/10/85 and SSP #850562, dated 7/25/85) regarding a consumable program, now the expendable products program. These recommendations were all implemented by November 7,1986.
The resulting expendable products program is largely based on the concem of halogenated material that could potentially lead to stress conosion cracking. It has been largely successful, but some incorrect materials have beca used. Some ACRs have been dispositioned on the use of improper tape. In addition, the event evaluation team identified several ACRs concerning the presence of other foreign materials on stainless steel, one example being melted plastic.
The fire wrap with red duct tape may have been placed on the pipe during construction days.
There are some instances that may be more recent. The ACRs previously written demonstrate that NAESCO personnel are aware of the requirements. Corrective action upon discovery of foreign substances however has not always been timely.
Events of Corrosion of Borated Water Systtau There have been a number of previous events in the industry of corrasion of borated water systems that have some pertinence to the subject event. Corrosion of borated water system components has resulted in a number instances of degradation or leakage. A search ofindustry experience has revealed four types of common events, as follows:
- l. Italogen (most often chloride)-induced, transgranular stress corrosion cracking (TGSCC) in heat traced systems containing high boric acid concentrations (typically 12 weight percent)
- 2. Boric acid corrosion of carbon steel components
- 3. Stress corrosion cracking caused by local stresses and external contamination
- 4. Primary water stress corrosion cracking (PWSCC) ofInconel 600 material lhere have also been several other related events. These are not as common as the other events, but also provide lessoas learned or insights, in order to determine if further lessons learned are applicable to the subject Seabrook Station event, each of the above type of events were reviewed and are analyzed below:
- liningen-Induced. Tranwranular Stress Corrosion Cracking (TCSCC) in Heat Traced Systems y
Containino Hich Boric Acid ConcentrafiQDS l
l DaArtsomes Ims 38
___._._______.._______._.____q 1.
-l This phenomena involves both intemal (from the inside out) and extemal (from the outside in) attack. The cases of external corrosion all involved wetted insulation. The concentration of boric acid in the system were high (typically 12 weight percent). These systems were all heat traced.
The problems occurred in stagnant piping regions. INPO and NRC database searches revealed the following events of this type:
Docesseet Event Title Casas (s)
Reasarks Date INPO SER 12'28/82 Cracking in Internal TGSCC; near A chemistry evaluation of the l
57 83 Stagnant Boric Acid welded regions; possible inner surface of the piping Piping sources of revealed contamination levels contat vation include a of up to 110 ppm chlorides,200 contaminated batch of ppm sulfates, and 84 ppm boric acid or recycled fluorides. These levels were contaminated boric acid.
high enough to cause cracking of tne type 304 stainless steel when combined with the-stresses associated with the weldir g process.
r OE 6536 by 2/17/94 Stress Corros' a Stress corrosion Very poor maintenance of -
Millstone 2 Cr,~ king of Boric cracking; boric acid system. A number of packing Acid System Piping residue collected in the leaks and other component and fitting-insulation, forming leaks soaked the insulation.
concentrated halogen compounds.
OE 7201 by 3/31/95 CVCS Boric Acid External and internal Possible source of pipe surface Turkey Transfer Piping TGSCC; presence of a contamination include poor Point 3 Through Wall halogens on the outside construction practices, Leakage and inside surfaces of contaminated insulation, and the piping the use of service wster to wash down piping components.
Possible riources ofinternal contaminatioriinclude poor construction practices and i
contaminated boric acid.
I e
i ORArtanden3-V19e f
39
Boric Acid Corrosion of Carbon Steel Comnonents l
Exposure of carbon steel to boric acid has resulted in a number of events of significant corrosion of material. INPO and NRC database searches revealed the following events of this type:
j Desemoet -
Event Date.
- Tiens(
- Cease (s)
Reeseks
- ~
SER 72 83 Point Beach:
Damage to Carbon Boric acid induced Body-to-bonnet valve -
4/28/83 Steel Bolts and Studs corrosion leaks. In the Yankee on Valves on Small Rowe case, a valve Yankee Rowe:
diameter Piping body defect resulted in a 4/5/83 Caused by Leakage high velocity spray onto of Borsted Water the studs.
NRC IN 86-3/14/87 Degradation of Boric acid induced Reactor vessel head 108,S.!
Reactor coolant corrosion of carbon multirle corrosion-l System Pressure steel; source of boric induced degradations.
Boundary acid was leak from a Demonstrates lower instrument tube imponance of small seal on one of the leakage onto nearby het instrument tubes.
surfaces.
Generic Based on 5 Boric Acid Corrosion Boric acid induced item 2 states that it is Letter 88 0$
cvents,4 in of Carbon steel corrosion of carbon steel imponant to establish the 1986 and 1 in Reactor Pressure potential path of the 1980 Boundary leaking coolant and the Components in PWR pressure boundary Phnts components it is likely to contact.
Generic Letter 88-05 has resulted in our boric acid reduction program The issue related to the prevention of spilling or drops onto carbon steel components is also applicable to cpillage onto insulation that covers hot stainless steel piping. Such insulation inside the containment building is covered wiQ metal. Only small areas are exposed to the problem. Nonetheias, to help preclude possible leaking into non metal covered portions, the event evaluation team recommends an assessment of addressing wetted insulation.
7 DR.4f7 AtWsden J 1/7,98 40
l Stress Corrosion Crackino CanceA by Local Stresses and External Contamination This phenomena results from very localized stresses. INPO and NRC database searches revealed i
the following events of this type:
Desumwat; l INest iTWe{
Casss(s) '
Ileasarksf.
o s Date f Yankee Rowe 1/24/84 Reactor Coolant Cold bending without This is the only documented LER 84 001 Pressure Boundary subsequent softening, failure of this type in the Leak abrasions, and history of this plant's -
concentration of operation. Insulation chlorides that had exhibited compliance with leached out ofinsulation Regulatory guide 1.36.
Water (source not explained in the report) soaked through the insulation, transporting leachable chlorides to the hot (520'F) pipe surface where subsequent evaporation concentrated the chlorides.
OE 3088 by 12/21/8*
Chloride Induced TGSCC; source of condensed steam vapor is Duane Arnold Transgranular Stress chlorides is decomposed believed to have run Corrosion cracking electrical cabling through the decomposed of CRD Piping insulation jacket.
cable jacket, resulting in leaching of chlorides.
Water and chlorides drained from a field option box located directly and dripping onto the affected i
P Pc Turkey Point 4 3/10/94 Unit 4 Shutdown Outside diameter RCS leakage was from an LER 94-002 Due to RCS chloride induced incore instrument guide Boundary Leakage transgranular stress
- tube, corrosion cracking; source of chloride contamination was improper cleaning techniques EAR MC W 8/26!95 Damage of Primary A net like population of Plant is Kalinin NPP Unit 2,95-044 Circuit Drain Pipe corrosion -induced which commenced under Wet cracks from chloride commercial operation in Jnsulation contamination from both 1986.
a weld leak a on loop drain line and from the thermalinsulation.
i DRMYResNeue $ W7M 4l.
^m 1
J
-y' zg g
v Prii-
==
wwTT m
fw d'--
m-ymy->w-w---
y
-- g w-*y y'-'+rw
=
t M
t'
- * ^
- e-n-w9y-'
dy r-
o Primarv Water Stress Corrosion Cracking (PWSCC) ofInconel 600 Material This phenomena is internal attack, i.e. from the inside out. It was first identified in 1987, and new events continue to occur. There have 1,cen a number of problems on pressurizer connections, exclusively of the Combustion Engineering design, which are constructed of this material. The PWSCC ofInconel 600 phenomena has more recently been detected on reactor vessel head penetrations (applicable to Westinghouse plants also). INPO and NRC database searches revealed the following events of this type:
Document ~
Event Date Title
, Cause(s)
Remarks -
SER 2 90 Pressurizer heater PWSCC Combussion Engineering Sleeve Cracking pressurizer NRC IN 90-Primary water PWSCC Combustion Engineering 10 Stress Corrosion prescrizer Cracking ofinconel 600 Calvert Cliffs 3/21/94 Pressurizer lleater Source of stress was Combustion Engineering i LER 94-003 Sleese Cracking axial scratches, pressurizer mechanism of corrosion was PWSCC San Onofre 3 7/22/95 Reactor Coolant Combustion Engineering LER 95 001 System Pressure pressurizer Boundary Weepage San Onofre 3/3/97 Reactor Coolant PWSCC Combustion Engineering Unit 2 LER System Leakage pressurizer 97 004 from Pressurizer Thermowcli Gentric Sescrai events of Degradation of PWSCC and The IGA attacx was at Letter 97 01 1991 to 1994 Control Rod Drive intergranular attack Zorita involved Mechanism Nozzle IGA contamination by resia and Other Vessel bead injection into the Closure Head RCS.
Penetrations DPAFTJtohlen 5 1/M8 42
I Other Related Eventt f
Docuaseet Esent Title Cause(s)
Rensarks Date INPO SER 12/21/83 Stress Corrosion Internal stress corrosion The halogenated 83 83 Cracking of Pilot cracking; halogenated hydrocarbon was suspected Line for Pressurizer hydrocarbons and to have hen a mixture of Safety Valve oxygen had appa ently trich! 7. ethane and been introduced into tetrachloroethene, a volatile the pressurizer from the substance routinely used for auxiliary spray line and cleaning components concentrated in the steam space Diablo 10/9/87 Accumulator Nozzle IGSCC; stress was Source of contamination Canyon 2 Cracking Due to improper welding could be from consumables LER 87 Intergranular Stress process; contaminativa used during the welding f
023 01 Corrosion was chlorides and process, or uncleaned off sulfur found in the contaminated hydrotest socket region of the water nozzles OE 6536 2/17/94 Through Wall Leak intergranular corrosion; Defects were in heat by.
on an ECCS Pump stresses :lue to weld affected zones of the Pallisades Suction Valve repairs made during castings I
manufacture Surry 2 9/12/95 Operation with Non.
Circumferential cracks The leak is believed to have LER 95 Isolable Leak on of pressurizer n'tztes initiated from the inside out 007 01 Pressurizer by Internal TGSLC and and subsequently created an Instrumentation external TGSCC and environment on the outside Nozzles IGSCC of the nozzle that led to the external cracking l
Surry 2 2/23!96 Loss of containment Local stresses and heat The affected section of l
LER 96-Integrity Requires affected zones of piping had been in service i
011 Plant Shutdown overlapping welds in for 23 years, ben; radius piping at a i
fixed saddle support 1
i I-f-
5
- mt4rrAcWsden 5 14M 43
Conc!usions These industry events, along with the Seabrook Station event, demonstrate the importance of avoiding contamination of piping systems (both externally and internally). Panicularly, this data demonstrates that extemally generated corrosion requires simultaneous and unique circumstances and conditions to create a problem. Generally, high temperature, halogen contamination, and a stress (during the weld processing and/ or operational) are required.
Wetted insulation can result in leachable chlorides of a magnitude (but must be in conjunction with the other simultaneous conditions) to result in a problem. As evidenced by the above data,
this is a occui..... Nonetheless, diligence is required to avoid a continuous type wetting con ('
Also, should a significant spill occur, as during disassembly evolution, insulation shoui e removed, the piping cleaned, and dry insulation be installed.
It is also imperative that affected workers are knowledgeable on the importance of not allowing unqualified foreign substances to remain on stainless steel piping.
l DMETRedston 3 UZ48 44
ATTACHMENT D ENGINEERING EVALUATION SS-EV 97-0031 RC-%89 Inlet Spool Piece Failure Analysis Engineering Evaluation Number SS-EV-97-0031
~
IJ furgest R( V 89 is a relief valve off the 12 inch suction line to 'B' Train RHR Pump 8B. On 12/5/97, se spool piece, on line # RC-88 1-oui 3, connecting the 12 inch line to the inlet of the relief valve exhibited weepage at each end. Specifically through wall weepage was observed in the heat affected zone (11AZ) of the weld at the weld neck flange on the inlet to RC-V-89 and the HAZ of the field weld on the spool piece at the weld-o let on the 12 inch suction line. The purpose of this evaluation was to determine the cause of the through wall weepage.
2.0 Background
The spool piece sits in the vertical direction and is approximately 5 inches long. It is 3 inch diameter schedule 40S, type 304 stainless steel pipe. Per Specification 249-1 the spool piece was to be insulated with 1" fiberglass insulation for heat retention (Note: The insulation was not installed for the spool piece to RC %89 nor the identical spool on Train 'A' at RC-V24).The weepage was identified when the plant was at 100% power with RHR in its normal standby condition. During power operations, this line is essentially stagnant at a temperature of approximately 100*F and 50 psi (static head of the RWST). During startup and shutdown the piping is heated to its normal startup/ shutdown temperature range of 300-350 F for about 3-5 hours. Over the life of the plant this would work out to a " time at temperature" of approximatelv 400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br />. A total of four small weepholes were observed on the spool piece weld heat affected zones. One at the top weld which affixe3 the spool piece to the flange, and three at the field weld which affixes the weld-o-let to the spool piece.
Visual examinations of the spool piece indicated that it had been wrapped in a two layer covering consisting of a woven cloth fire wrap against the pipe surface and standard red duct tape around the wrap. The condition of the covering indicated that the materials were somewhat charred and appeared to be melted or matted together. Seabrook Chemistry Department took chips of the material and performed a chemical analysis which identified 349 ppm chloride and 24 ppm fluoride. During outages, relief valve RC V 89 is removed from the system for testing. In the pmcess borated water drains from the line wetting adjacent material in the area including the 3 -
i.nch spool piece, its covering and the 12 inch suction line and its insulation. Although the present chloride readings for the covering materials are 349 ppm they may have been higher in the earlier stages priar to leaching, it appears that the covering material was on the pipe since original Durrnedans Inse 45
constmetion (10 years or more) and that the removal and testing of RC-V-89 along with subsequent wetting of the area had occurred approximately 6 times.
Original radiographs of the two welds on the spool pier.e were retrieved and examined by NDE Level ill personnel. The review showed no anomalies in the weld or the base metal. It was also noted that the two welds on the spool piece were performed by different welders in different companies. The flange to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o let to pipe weld was a field weld performed at Seabrook This indicates that the welds and welder qualification do not appear to be a contributor to the problem as the radiographs showed no anomalies and the welds were performed and examined by different individuals yet the heat affected zones of both welds had the weepholes. Also, there did not appear to be any anomalies in the base metal of the pipe which furtFer indicates that the weepage was due to in service conditions rather than original installation.
10 Discussigt; A document search of the heat number for the spool piece was performed to determine where else piping of that heat may be installed in the plant.This was done in anticipation that the problem was due to poor piping material. However, this failure mechanism has been eliminated based on chemical analysis completed at Mass Materials Research Labs which confirmed that the material met the specification for SA312 Type 304 Stainless Steel pipe.
A review of the piping analysis shows that the stresses in the spool piece piping are minimal.
During R11R operation, the combined stresses have a safety factor of 4.75 on the normm design allowable. During power operation with RHR in standby, the safety factor on normal design allowable is about 28.0. The piping design in this area is dictated by the faulted condition with stresses induced by a Loss Of Coolant Accident. In addition, there have been no indications of any mechanical or flow induced vibration in this area. Nor has there been any problem history with relief valve RC-V-89 that would have resulted in abnormally high thrust oads. In summarizing, there does not appear to be any high service induced stresses that could initiate a failure mechanism which could be attributed to the weepage. The only area on the piping spool piece that has been exposed to high stress levels is in the heat affected zones of the welds due to residual stresses from welding.
There does not appear to be any service induced mechanism which could be attributed to the weepage. However, a combination of the HAZ weld residual stresses, sensitized microstructure in the HAZ, exposure to intermittent temperatures greater than 200 F, and environmental conditions (i.e. chloride contaminants) on the cutside of the pipe are indicators that the pipe may be susceptible to Chloride induced Stress Corrosion Cracking (SCC). Although the chloride level in the wrapper is not excessive at 349 ppm, this in itself would be sufficient to initiate the stress l
corrosion process when coupled with the high residual stresses located in the weld's heat afTected l
zone. The higher the residual stress, the more sensitive the material is to smaller amounts of chloride. The chlorides would also tend to concentrate at crevices on the surface due to the wrNng of the covering material by the borated water and subsequent drying off of this water unen the line goes back up in temperature. This could result in potentially high levels of chloride concentrated in selective areas around the heat affected zone of the welds.
DRUTRedston 5 V7.98 46 i
C
\\
~A dxision was made to remove the spool piece and perform metallurgical analysis to determine the cause and extent of degradation that led to the weepage. The piece was semoved by making a cut at the weld of the weld-o let to the spool piece and unbolting the inlet flange at RC-V 89.
The analysis involv'ed one section of the spool piece being analyud at Seabrook and the other section analyzed at ABB Laboratories in Windsor Locks Connecticut. The piece was cut approximately 31/2 inches down from the top of the flange. The top piece was analynd at
}
Seabrook and the bottom was shipped to ABB Labs.
{
Prior to sectioning the removed piece, a leakrate test was perform-d on the piece to determine l
what effect the degraded section would have had on containment integdty. The test consisted of clamping a gasket.1 plate over each end of the piece. One end was fitted with a test connection which was pressurized to 52 psi. The leakrate was determined to be 1.363 Standard Cubic Feet 3
per Hour (SCFH) which is much less than the allowed value for a typical penetration which is on the order of 37.0 SCFH. It was also noted that the leakage appeared to be coming from the.
gasketed end plates. Results of the test indicate that the pipe spool was not degraded to the extent that containment integrity was challenged.
The following describes the work and examinations performed on the section analyzed at Seabrook and the conclusions reached from that work. A Liquid Penetrant (LP) examination was performed at the weld of the flange to the 3" pipe on the inside and outside diameter of the piece.
One linear indication was found that went through the wall. The crack ran at an approximate angle of 45' to the longitudinal axis of the weld. It progressed into the base metal of the 3" pipe.
The piece was sectioned at a location perpendicular to the weld. The sectioned surface displayed a branching type crack running perpendicular to the surface of the pipe at a distance approximately 3/8" from the weld in_the heat affected zone. The indication originated on the outer surface and progressed to the inside diameter of the pipe. This testing is documented in Reference 1.
The failure appears to be a corrosion assisted degradation process such as chloride induced stress corrosion cracking that had initiated on the outside and migrated to the inside diameter.
However,it was decided that more confinnatory information was required from the analysis
. being done at ABB on the lower section of the spool piece. Additional needed information
- included: obtaining chemical analysis of the pipe in the non-heat affected zone, performing an
- SEM examination on the crack surface, and performing a chemical analysis of any contaminants that may be present on the crack surface.
- Some indications were found in the heat affected zone of the weld on the 12" line where it attaches the weld-o-let. The indications were repaired. However, the indications on the weld-o-let
- were more extensive and could not be repaired. The weld-o-let was removed and replaced with a (ne' fitting.
w b
l Jaan7Jtadmiseflesi 17 4
m m,
++...w.-
.-c mo o
,,,-m m
=-
_,-m
..-e
Results of the analysis conducted by ABB Laboratories were consistent with those previously duermined at Seabrook. Upon receipt of the spool piece, a LP examination was performed on the outside and inside surface of the pipe. The revealed two through wall linear indications and one non through wall which suppons the theory of a failure mechanism originating frc n the outside and progressing inward. The methodology and conclusions of the failure investigation are contained in Attachments 2 & 3.
NAESCO provHed technical oversight to the ABB analytical work by assigning the Duke Engineering & Services Metallurgical Engineer to follow the work at ABB and provide his own assessment. Ifis report is contained in attachment 3.
4.0 Safety Significance The spool piece and connecting valve and piping are classified ANS Safety Class 2, seismic category I components. The piping components were designed, manufactured, and tested in accordance with the ASME Boller and Pressure Vessel Code, Section 111.
Subsection NC. These components are part of the Residual liest Removal System which provides for safety related shutdown cooling, and low head injection during ECCS operation.
Thc trough wall weepage resulted in the Unit entering Technical Specification 3.6.1, resulting in a plant shutdown. This evaluation provides the technical assessment of the cause for the through wall weepage, with the purpose of preventing reoccurrence in the future. Replacement piping components have been fabricated and installed in accordance with the ASME Code Requirements Other corrective actions are included in the Event Evaluation for this occurrence.
==
Conclusions:==
The repons conclude that the cause of the weepage sites was a chloride induced transgranular stress corrosion cracking mechanism initiated fiom the outside of the pipe working in. The Chlorides were leeched out of the tape /firewrap material affixed to the outside diameter of the pipe. This is not a typical failure mode for stainless steel in the benign PWR coolant chemistry environment, however the unique wrap material, inadvertently left on the subject pipe spool at construction, exposed the stainless pipe to a deleterious local environment which, when combined with the typical liAZ microstructure, led to the observed through wall weepage.
Attachments:
- 1. Duke Engineering & Senices Memorandum, File #RC-V-89 A, dated 12/8/97, from KR Willens to RE White.
1
- 2. ABB
- 3. Duke Engineering & Senices Memorandum, File # RC-V-89 B, dated 12/12/97, from KR Willens to RE White DK41T Revision 3 - 1/L48 48
1 l
Attachinent E ENGINEERING EVALUATION SS EV-97-0030 EVALUATION OF SYSTEM SAFETY FUNCTIONS WITli RESPECT TO WEEPAOE AT INLET PIPING TO RC-V89, B-TRAIN RilR PUMP SUCTION RELIEF VALVE Prepared by: R. J. Faix J
December 11,1997 1
Purpose:
During a recent shutdown to address the pipe wall defects in the inlet piping to RC.V89, an engineering evaluation was requested to detenaine if the R11R system would have bw i tapable of performing its safety function. The RilR pump suction and the containment building spray pump suction from the RWST is a common piping system. The CBS System will also be evaluated. The applicable safety functions are contaimnent integrity, shutdown cooling, emergency core cooling and contairunent depressurization via containment spray. The purpose
- of this evaluation is to detennine if the RiiR System and CBS were capable of performing the above safety functions.
Background:
On December 5,1997, Seabrook Station initiated a reactor shutdown from full power to repair weepage from a thmugh wall defect in the inlet piping to the B-train RiiR pump suction relief valve. The defect was identified by the presence of boric acid crystals on the pipe wall and on adjacent pipe insulation. Prior to the reactor shutdown, the pipe insulation was removed, the piping was cleaned and inspected.
Two welds exhibited weepage sites. One weld, the pipe to RilR relief valve inlet weld neck flange, had a single defect site. The second weld, the weld-o let to pipe weld, exhibited three weepage sites. The system leakage was characterized as a drop every two to three minutes. The system pressure at the welds was the elevation head of the RWST or less than 50 psig.
The relief valve and inlet piping a-r shown on UFSAR Figure 5.1-3, Sheet 4. The AllR System is shown on UFSAR Figure 5.4-11. The piping is ANS Safety Class 2, ASME Section III, Class 2, Seismic Category I.
. Discussion:
The applicable safety functions for the RiiR System are containment integrity via the containment penetration and piping, shutdown cooling and emergency core cooling. The safety function for the CBS system is containment depressurization.
Daninewa,s um 49
e Containment Integrity:
The RHR System cooldown piping forms a portion of the containment boundary. Tha containment penetration is described in UFSAR Section 6.2.4. The piping defects and weepage occurred in the containment boundary of the piping system.
The piping defects may challenge the containment integrity since the penetration piping would have a pre existing defect. The design basis of the ECCS is that it is capable of sustaining an active or a passive failure in the long term and continue to perform its safr function. With a pre-existing defect in the containment penetration piping and a postulateu passive failure in the ECCS piping system, a leakage path to the environment could occur.
ne piping section containing the deuis was subjected to a leak rate test after being removed from the piping system. The leak rate test was conducted by pressurizing the pipe ID to 52 psig, the containment design pressure, and measuring the makeup flow required to maintain the pressure ct 52 psig. The leak rate was determined to be 1.363 standard cuUc feet per hour (scfh).
This cornpares to an allowed penetration leak rate of 37 scfh or 0.05 L, as cescribed in the Containment Leakage Rate Testing Program.
Since the tested leakage through the degraded piping section and the defects wat much less than the allowed leakage for containment penetrations, the penetration was capable o performing its comainment integrity function.
Shutdown Cooling:
During shutdown cooling the piping is subjected to an operating pressure of 365 psig and operating temperature of 350 'F. The piping design pressure is 600 psig at 400 'F. The piping section is the inlet piping to the RHR pump suction relief valve set at 450 psig. So that realistically, the maximum pressure in the degraded pipe section is 450 psig.
A section through the defect 5 the pipe to inlet flange weld identified a small defect in the pipe wall below the weld. The base metal indicated no signs of pipe wall degradation other than the single defect. The pressure minimum wall in the 3 inch pipe run is 0.064 inches. The pipir.g section was fully capable of carrying all design loads and maintaining the structural integrity of the piping system as demonnted in Calculation C-S-1-45545, RC-V89 Inlet Pipe Flaw Evaluation. This calculation evaluated a bounding 0.5 inch diameter flaw that is sufficient to account for the cumulative flaw area and determined that the calculated stresses were below code allowable stresses and that the flaw was stable in accordance with GL-90-05 criteria.
Similarly, the defect in the pipe to weld-o-let weld identified three small defects in the pipe wall below the weld. The base metal indicated no signs of pipe wall degradation other than the three defects. The pressure minimum wall in the 3 inch pipe run is 0.064 inches. The piping section wa= fully capable of carrying all design loads and maintaining the sttuctural integrity of the piping system as demonstrated in Calculation C-S-1-45545, RC-V89 Inlet Pipe Flaw Evaluation.
mnanws inss 50
e u
1 Since the piping section is capable of maintaining its structural integrity, the RilR System is capable of performing its shutdown cooling function.
i Faergency Core Cooling:
During emergency core cooling the piping section forms part of the pressure boundary of the RilR pump suction. Leakage from the RWST into containment as a result of the piping defects 1
is infmitesimally small to be of any consequence. During ECCS injection the RiiR pumps take j
suction from the RWST and deliver flow to the RCS cold legs. The pressure at the location of the defect is the RWST elevation head as it exists during plant operation. During the recirculation phase of ECCS operation the pressure at the location of the defect is the containment pressure since the RiiR pump suction is transferred the containment ECCS recirculation sumps.
The piping section loading is bounded by the structural loads. With the minor defects at the two locations, the piping is fully capable of maintaining its structural integrity. Since the piping section is capable of maintaining its structural integrity, the RliR System is capable of performing its emergency core cooling function.
Containment Depressurization (CBS System):
The containment building spray pumps and the RilR pumps both take suction from the RWST following an accident. The CBS System is separated from the RIiR system by two check valves in series. Since the nature of the weepage sites could potentially affect the opern. ion of the CBS system, the CBS System was conservatively declared inoperable.
The CBS pumps wwre fully capable of taking suction from the RWST and delivering the required flow to the containment spray headers inside containment. The interaction of the CBS system with the RilR System was of no consequence since the ECCS was fully capable of performing its safety function as described above.
Piping Pressure Houndary Integrity:
The defects in the RilR suction relief valve inlet piping represent a deviation from the ASME code and, as such, may not conform to one or more applicable codes or standards specified in the UFSAR. The piping does not meet the minimum wall thickness at the location of the defect.
North Atlantic performed a code repair consisting of replacing the pipe sections. This activity necessitated a plant shutdown to complete the repairs.
==
Conclusion:==
The RliR System was capable of performing the containment integrity, shutdown cooling and emergency core cooling safety functions with the piping section containing the identified detects. The CBS system was fully capable of performing its safety function of containment
. depressurization.
aurrsewons ums 5:
l
- 7. g,h8 January 9, 1998 1
s MEMORANDUM TO: Docket File FROM:
Craig W. Smith, Project Manager Original signed by Project Directorate 1-3 Division of reactor Projects -I/11 Office of Nuclear Reactor Regulation
SUBJECT:
TRANSMITTAL OF INFORMATION TO DOCKET NO. 50-443 The attached report, dated January 7,1998, is being transmitted to the docket for Seabrook Station, Unit 1. The draft report provides the licencee's preliminary root cause analysis f(w the leaks in the piping of the "B' train of the residual heat removal (RHR) system.
The leakage was reported by the licensee on Detember 5,1997.
Docket No. 50-443
Attachment:
As stated DISTRIBUTION CPU.BLIC)
PDI-3 Rdg.
B. Boger (w/o attachment)
R. Eaton (w/o attachment)
M. Thadani(w/o attachment)
C. Smith (w/o attachment)
T. Clark (w/o attachment)
DOCUMENT NAME: G:\\ SMITH \\ SMITH \\MA03Eu.DKT To receive a copy of this document, indicate in the box:dre "C" = Copy without attachment / enclosure "E"=Copywithattachment/pr)c/bs "N" = No copy 0FFICE P0la3/PM //
l P01 3/LAJ M If (A)DPOI/JhfyV l
l f,*ME CSmith &
TCtark m D V REalon j i DATE 1/9 /98 1 M /96 1/ll/98I i
OFFICIAL AECORD COPY
~7 1
w I*
,7 EXECUTIVE
SUMMARY
f On December 5,1997, system leakage of a drop every two to three minutes (weepage from four
_ locations) was discovered on the inlet piping to the Residual Heat Removal pump 8B suction
~ relief valve (RC-V-89). The weepage sites were in the metal adjacent to the welds (beat affected i
- zone) that connected the piping spool piece to the relief valve flange and the 3" to 12" wcld-o-let that connected the pipe to the 12 ' RHR pump suction line. One weep was in the upper heat affected zone and three were in the lower heat affected zone. North Atlantic personnel n
f performed a preliminary evaluation of the weepage sites and were unable to quantify the air leakage in scfh. Therefore, the Operations Manager conservatively determined that the Limiting
- Condition for Operation for Technical Specification 3.6.1.1, Containment Integrity, could not be
- satisfied. He declared the Containment inoperable and entered the referenced Action Statement.
In addition, the Residual Heat Removal (RHR) and Containment Building Spray (CBS) systems, l
which utilize the same section of pipe, were declared inoperable. A plant shutdown was --
l commenced and completed in accordance with the requirements of the Specification. A one hour report was made to the NRC pursuant to 10CFR50.72(b)(i)(a).
The weepage sites were caused by chloride induced transgranular stress corrosion cracking (SCC).' The source of the chlorides was a non standard insulating jacket that had been on this section of pipe since at least 1988. The insulating jacket consisted of a piece of cloth firewrap type material wrapped in red duct tape. The insulatingjacket was not the type ofinsulation designed to be in place on this section of pipe, it is unclear exactly how or why this insulating
-jacket was placed on this pipe. Separate chloride analysis of the insulating jacket resulted in values of 189 ppm and 349 ppm. It is unknown what the original chloride value of the cloth or the duct tape was when it was installed.
Laboratory examinations consisting of microscopic and chemical analysis have determined that SCC is the most likely cause of the pipe weepage. The analysis determined that the SCC had started on the outside of the ripe and had migrated inward. It is suspected that repeated wettings and dryings of the insulating jacket, caused by wetting of the surfaces during draining of the pipe as part of maintenance performed in refueling outages, may have caused the chlorides to concentrate on the pipe over time.
- Original radiographs of the two welds on the spool piece were retrieved and examined. The review showed no anomalies in the weld or the base metal. It was also noted that the two welds on the spool piece were performed by different welders in different companies. The fhnge to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o.-let to pipe weld was a field _ weld performed at Seabrook. This indicates that the welds and welder -
_ qualification do not appear to be a contributor to the problem as evidenced by the fact that the radiographs showed no anomalies and the welds were performed and examined by different
- individuals yet the heat affected zones of both welds had the weepage sites. Also, there did not appear to be any anomalies in the base metal of the pipe whicu further indicates that the weepage was due to in-service conditions rather than original installation.
(
i p
4 I
Thisjacket had been installed on the RC-V-89 inlet pipe since at least October of 1988. From this time until its removal on 12/5/97, it was not identified as inappropriate for this application.
The monitoring of the boron buildup on this pipe began in March of 1996 and continued until the weepage was discovered on 12/5/97. During this time it was believed Gat the source of the boron was related to water spillage on the jacket and on insulation or the 12" RHR pump suction line.
it was thought that the water absorbed by the insulation was drawn up the jacket in a wicking process and then evaporated leaving the boric acid residue behind. This idea was reinforced by the fact that at no time did anyone report finding moisture or wet boric acid when inspecting or cleaning the boron buildup. For these reasons the jacket was not removed and a positive determination of the source of the boron buildup determined until 12/5/97. After being informed for the first time that the boron buildup had been cleaned and then reappeared, the system engineer led a team into containment to remove the jacket and inspect the pipe. However, as an organization there were missed opportunities to identify, document, evaluate, and resolve the issues prior to December 1997.
The piping section containing the defects was removed and subjected to a leak rate test. The leak rate test pressurized the pipe to between 50 and 52 psig. This is greater than the containment design basis accident pressure of 49.6 psig. The makeup flow required to maintain the pressure at 50 - 52 psig was measured. The leak rate was determined :o be 1363 standard cubic feet per hour (scfh). This compares to an allowed penetration leak rate of 37 sefh or 0.05 La as described in the containment leakage rate testing program. In addition, the piping section was analyzed with the assumption of a bounding 0.5 inch diameter through wall hole. Tha e zaluation determined that the piping section and the weld-o-let were fully capable of carrying all design loads. The structural integrity of the piping system under all normal and accident conditions would have been maintained. The evaluation is provided in Attachment E.
Based upon the evaluation it was determined that the piping spool was degraded and nonconforming, but that the three systems, Containment, RHR and CBS were fully capable of performing their specified safety functions if they had been called upon to do so and were Operable utilizing the guidance of Generic Letter 91-18, Revisie 1. Therefore, it was recommended that Operations review the engineering culuation and withdraw the one hour report of December 5,1997.
Based on a review of the events and causes of the conditions evaluated in this report the event evaluation team recommends to the SORC that they recommend to the Station Director that the plant be restarted after the completion of the applicable corrective actions.
DRAFTRedslen $ V7198 2
' 's l.0 EVENT DESCRIPTION Thru wall weepage shes were discovered on the five inch long 3 inch diameter line that connects RiiR relief valve RC-V-89 to the main 12 inch RilR header. The relief valve and its inlet piping have had a history of observed boric acid buildup.
2.0 INIDATING CONDITIONS The in.;tallation ofinappropriate material on the inlet piping of RC-V-89.
3.0 REPORT CATEGORY l hour report per 10 CFR 50.72 (b) (i) (a) (Withdrawn) 4.0
SUMMARY
OF PLANT PERSONNEL ACTIONS ASSGCIATED WITH RC-V-89 INLET PIPING
4.1 BACKGROUND
Sycem leakage of a drop every two to three minutes was discovered on the surface of the 3" pipe spool that connects RC-V-89, RHR P-8B suction relief valve, with the 12" pump suction line inside contain'nent. The discovery of this condition resulted in the plant being shut down to repair the piping. The following narrative is a description of the events leading up to this event and the recollections of the individuals who have been interviewed by the members of the event evaluatior. team formed to investigate the causes of this condition.
Every refueling outage RC-V-89 is removed from the flanged connection at the end ofits 3" diameter,5"long inlet pipe which is welded into the pump suction line. This is done to support local leak rate testing Removal of RC-V-89 requires that it be unbolted at the discharge flange. When this is done, the length of piping that extends from the valve's outlet to it's final discharge point, the PRT, drains backwards out of the flange.
According to the valve test engineer, when the water in the valve discharge line was drained in this manner during OR04, the top of the insulation on the suction line and the firewrap surrounding the RC-V-89 inlet pipe were wetted. When the valve test engineer tested RC-V-89 during OROS the insulation in the area was wet.
DMFTRolstonS V7/98 3
4 s
FIGURE 1 REJEJ VALVE RC.V-89 F
K
~
/
7-k
[
/ K V
' i l
4 i
N
/
N
/ N g
/
1 THRU WALL WEEPAGESITE IN HEAT AFFECTED ZONE (SHOWED MORE
/ \\
/
OTHERS)
LEAKAGE THAN k
/
3' SS PIPE h
/
3 THRU WALL
\\
/
WEEPAGESITES
\\
/
IN HEAT
\\
/
AFFECTED ZONE
\\
/
3' WELDOLET \\
J Wl 12' TRAIN B RHR PUMP SUCTION HEADER DRAFTKevision 3 in48 4
l The 3" valve inlet pipe that is welded to the relief valve mounting flange was covered with a fiberglass fire protection cloth wrapped in red duct tape. The cloth looks similar to insulation, but apparently is material used to protect piping or other components from welding are strikes or burning debris during welding operations. The duct tape on the firewTap is wrapped from %" below the top to the bottom of the firewrap. The tape has taken on a seamless reddish appearance, looking like a painted on, thin, rough textured ceramic or plastic coating. Photographic evidence proves that this taped firewrap has been on this section of pipe since at least 1988.
4.1.1 March 1996 Observations On March 20,1996, during a routine containment walkdown at power an Operation's Department Nuclear Systems Operator (NSO) documented on Operation Department Instruction (ODI) 36, " Containment Tours" the fact that the RC-V-89 inlet piping firewTap had boron residue on it. Observations by different NSOs and HP technicians during subsequent containment walkdowns were documented on ODI-36 and on the RTS for routine cleanup of boric acid from various plant components. Operations and HP supervisors were informed of the condition in the area of RC-V-89 after these containment entries rad observations of RC-V-89. RC-V-89 remained on the ODI 36 list of components with boron buildup until December of 1997.
4.1.2. November 1996 Observe?.Qns On November 20,1996, after repeated observations of the boron buildup at RC-V-89 the RCS system engineer was asked by HP to investigate the build up of boron crystals that had formed in the area of RC-V-89. An HP Department supervisor awarded a spot recognition award to an HP Technician who raised a question about the condition at RC-V-89 including the boron deposits and the insulation on the inlet piping. The RCS system engineer went into containment 4
with the HP technician and took a camera with him. The RCS system engineer observed boron deposits on the relief valve inlet pipe firewrap and on the bottom of the flange. The RCS system engineer suspected that the flange connection of the relief valve to the riser pipe had leaked and that the wa;er had flowed down and accumulated on the inlet pipe firewrap. }le removed enough of the boron from the flange to satisfy himself that the flange connection of the relief valve to the inlet pipe was not leaking at that time. The inlet pipe and flange was photogthphed and the individuals left the containment.
The RCS system engineer told the HP personnel at the HP checkpoint that the beton residue was dry. The RCS system engineer discussed the situation with the valve test engineer and was told about the water that was spilled on the insulation and firewTap during refueling outages. The valve test engineer explained that he thought when water that was spilled on the insulation evaporated, the boron residue was left behind. The RHR pump suction line and relief valve inlet pipe can get as hot as 350 degrees when RHR is in service. This could cause evaporation and the subsequent formation of boron crystals. That fact combined with the understanding that the normal ambient temperature of containment can mmen;aons ms 5
4 be as high as 110 degrees, made the explanation seem reasonable. The 1
explanation was relayed to HP and they accepted that explanation.
Neither the engineers nor anyone from HP wrote a Work Request (WR) or Adverse Condition Report (ACR) on this condition nor did they ensure the boron buildup was documented in the RTS for routine cleanup of boric acid from various plant components. Both engineers assumed that HP would enter the condition at RC-V-89 on the boron cleanup RTS and that it would get cleaned.
Neither engineer considered the non standard insulation significant enough to warrant a WR or ACR being written. They did not recognize that the reddish 4
color was from duct tape. They did not conclude that there was a potential for negative consequences as a result of the non standard-looking insulation on this pipe. This same conclusion was reached by others including managers and supervisors who were shown pictures of this pipe with its boron buildup and reddish colored insulation.
4.1.3. February 1997 Observations No additional documented actions took place concerning the condition at RC-V-89 until Februsry of 1997. In February, an HP Supervisor contacted the RCS system engineer and asked him again to evaluate the boron buildup on RC-V-89.
The HP technicians felt that something was not right about this situation and they asked the supervisor to have it looked at again. No evidence of water or moisture was reported to the system engineer at that time or any other time from when he had first become aware of this condition. The system engineer spoke to the valve test engineer about the new call from HP concerning the conditions at kC-V-89.
The valve test engineer was making regular entries into containment to modify the containment personnel hatch entry system. Since the valve test engineer was going into containment, the system engineer gave the valve test engineer the Noveraber 20,1996, picture of the RC-V-89 inlet pipe and its firewrap. He asked the valve test engineer to compare the current condition of the boron buildup with the condition that existed at the time the picture was taken.
After checking the area at RC-V-89 the valve test engineer responded back to the RCS system engineer that there was no change in the status of the boron buildup on the RC V-89 inlet pipe firewrap. The RCS system engineer concluded that since the condition of the boron buildup had not changed, the explanation of the water spillage and subsequent evaporation and boric acid buildup were still plausible. It did not appear that the RC-V-89 inlet pipe firewrap boron buildup had ever been cleaned and so they thought that the condition was stable and acceptable.
Neither engineer asked HP if they had ever cleaned the boron buildup on the firewrap. They also did not review the RTS for routine cleanup of boric acid from various plant components. ( Note: The only documented evidence of RC-V-89/RC-V-89 inlet pipe being cleaned is logged on t is RTS on June 21,1997.)
h Both still felt they completely understood the situation.Both of the engineers stated clearly and decisively that if at any time they had received reports of DRAFTRedston $ frins 6
l s
1 moisture or water on the RC-V-89 inlet pipe or firewrap they would have immediately generated the paperwork necessary to investigate and repair any leaks. They both believed that they did not have any type ofleak.
Both engineers felt it was not necessary to write an ACR at this point because generally, boron buildup on piping or components in and ofitself did not meet the threshold for writing an ACR. ACRs were not written on every case of boron buildup in the plant. If a component had a minor leak at a mechanicaljoint or packing that was causing the boron to build up, the valve packing gland or pump shaft seal would be addeh the RTS for routine cleanup of boric acid from various plant components. If the leak was mere than minor, a WR would be written to repair the component or to perform cleaning that would involve removing ofinsulation. Neither engineer thought that a leak existed. They both thought that eventually the water would all evaporate and the boron buildup would be cleaned up by HP using the RTS for routine cleanup of boric acid from various plant components. However, neither engine; ensured that the RC-V-89 condition was added to the RTS.
4.1.4 March 1997 Observations On March 17,1997, Technical Specification Surveillance procedure EX1810.208 "RHR Train B ISI Functional Test" was performed. The inspector who performed the surveillance correctly documented the condition of the RC-V-89 inlet pipe firewrap as a discrepancy in the comments section of the RTS as required. He reported to the ISI test supervisor that, after close inspection, he did not see any signs of water or moisture but noted the boron evident at the top and bottom of the firewrap. Procedure ES1807.001," Visual Examination Procedure" requires questionable or indeterminate conditions that are not going to be reworked or replaced to have ari engineering evaluation performed by Technical Support personnel or by a NDE Level 111 qualified person.
ESI 807.001 also requires an independent review of all evaluations for questionable or indeterminate conditions. Contrary to this procedural requirement, an independent review was not completed for the evaluation done by the ISI test supervisor for the boron buildup on the inlet to RC-V-89.
The acceptance criteria used throughout ESI807.001 is the term " evidence of leakage". The ISI test supervisor knew there had been boron buildup in this location before and he sp ke to the valve test engineer about it. He knew that the valve test engineer was familiar with the situation. The valve test engineer explained to the ISI test supervisor that he believed the boron buildup that the field inspector, was reporting, was the same boron buildup that the two engineers had discussed in November 1996 and February 1997.
The ISI test supervisor discussed the situation with the RCS system engineer and the RCS system engineer agreed with the conclusion reached by the vP :est engineer. Both engineers were convinced that because there was no i
'e sign of mois+ure or water evident at the firewrap, then there was no leakag als was the basis for the evaluation performed under ES1807.001. The field inspector DMPTReMn5 W8 7
- _ _ - - - _ _ _ _ = _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ ___ _________________ _ - _ _ _ _ - -
g r_
S.
4 1
~
. discussed this situation with his supervisor.. The field inspector's supervisor o
accepted the conclusions of the engineers.
This ISI inspection was conducted 15 months after the water was spilled onto the
- insulation that was thought to be creaticg the boron buildup on the RC-V-89 inlet
- pipe firewrap. None of the individuals involved had an es;imate for how long the spilled water would continue to evaporate, or how long the boron would co,tinue--
to buildup on the firewrap. They did not think it was getting any worse. The
- boron buildup was not increasing, so they assamed they had a static, understood condition. The field inspector when shown the November 20,1995 picture also saio he thought that it did not look like there had been any change from November -
of 1996 to March of 1997. He had first thought there might be a small leak under the firewrap, but when he heard the explanation from the engineers he decided _
that they were correct and that there was no leak.
The ISI test supervisor wrote WR 97W000833 that described the problem as boric acid deposits at the relief valve connection to 12" pipe.- He suggested cleaning or replacing the insulation as required. The RCS system engineer, who filled in the.
section of the WR that described the work that should be performed, wrote " clean boric acid residue cad remove damaged insulation (reddish color)"
The field inspector who did the field walkdowns for the RHR ISI functional test 4
wrote an ACR describing the poor work practices of the valve maintenance -
personnel who disasxmbled the relief valve and spilled the water on the insulation. The basis of this ACR was the information that he received from the valve test engineer who told him he had seen these poor work practices. This ACR did not address a potential leak in the pipe.
I 4.1.5 OROS Ohnervatiens In early.iune of 1997 the RCS system engineer performed a containment walkdown in preparation for the containment close-out at the end of OR05. He investigated the status of the RC-V-89 inlet pipe firewrap. The system engineer discovered that the work that was to have been performed per work request WR a
97W000833 was not complete. The engineer discovered that no work had been
- performed on the firewrap.
Some work had been comp;eted using WR 97W000833 and the WR had been routed to the RCS system engineer for his review and close out. The system -
engineer retrieved the WR and examined the comments on the task performed.
The following comment was written on the WR: " Removed and replaced insulation., We pre-fabbed Nukon blanket and applied to valve. Metal jacket was
' de-conned by H.P. 5/17/97." The system engineer told his supervisor that he
- could not sign off and close out the WR because the wo' k that was performed was r
not the work that was supposed to be performed.
i he RCS system engineer did not know at the time that the insulators did not understand exactly what insulation they were supposed to remove. The insulators
- did not consider the reddish colored firewrap to be insulation. They noticed the sammaw.=s was' s
6
.m
.._,_. u.
w
..,_-~_ _ _..
e e.,
j insulation on the 12" EHR suction line below the RC-V-89 relief valve inlet pipe was wet, so they removed the old insulation, fabricated new insulation and replaced the insulation. They did not ask any questions of their (NAESCO) supervisors about the " reddish color" comment on the WR. None of the insulation that the insulators removed was red in color.
The RCS system engineer told his supervisor that no work had been performed on.
the RC-V 89 inlet pipe firewrap. His supenisor told him to write an ACR describing the adverse condition and to check with the NAESCO supervisor of the insulators to see if he could discover exactly what component was worked on as described in the WR. The system engineer did not write the ACR. He intended to write the ACR after he discovered what had occurred and had gotten the correct work performed. This did not occur before the RC-V-89 pipe weepage sites were discovered on December 5,1997.
The system engineer discussed the situation with the insulator's NAESCO supenisor who said he would investigate what happened and get back to the RCS system engineer. The RCS system engineer never closed out the WR, and received no information from the insulator's NAESCO supenisor. The RCS system engineer's supervisor did not follow up to see if the engineer wrote an ACR or check on the status of the open work request on the RC-V-89 inlet pipe firewrap it was a busy time, refueling outage OR05 had just concluded, everyone had a lot of paperwork to review and the priority of addressing this issue was low.
It was viewed as only a housekeeping and cleanup issue.
On June 21,1997, just prior to the close-out of containment following OROS, the boron buildup at the top and bottom of the RC-V-89 inlet pipe was cleaned. (note:
this is the first documented cleaning at RC-V-89) No water or moisture was evident during the cleanup of the boron buildup.
4.1.6 Third Ouarter 1997 Observations Sometime during the third quarter of 1997 HP technicians and supeni:, ors tracked the status of the condition at RC-V-89 and they noted that boron was building up again on the top and bottom of the inlet pipe firewrap. An HP supervisor brought a picture of the RC-V-89 inlet pipe, taken by an HP technician, which showed boron buik'up to the HP Department Manager.
The HP Manager had a conversation with the Maintenance Manager, Assisumt Operations Manager, Oversight Manager, Mechanical Maintenance Department Manager and the Operations Department Technical Supervisor. The HP Manager told the other managers during this discussion that his department personnel had identified the boron buildup that was reappearing on the RC-V-89 inlet pipe
- firewrap. The HP Manager showed the other managers the photograph given to him by his department supervisor. The HP manager believed ' hat this picture represented the return of boron buildup since the valve was cleaned on June 21, 1997. He told them that the boron was again building up in the same spot but did not make it clear that this represented a new buildup. This was the first time that they had heard about this condition.
aumovwns ms 9
The managers then discussed the condition with the valve test engineer. At this time none of the managers had heard the explanation of the boron buildup and for all of them except the HP Manager,it was the first time they became aware of the boron buildup associated with RC-V-89.
When they showed the valve test engineer the picture, he told the managers his explanation of the cause and history of the boron buildup. He explained to them about the water spillage on the mve during the valve's removal for testing and about his theory of boron buildup. He told them that it occurred during every outage, that it was not a problem and that he and the system engineer were monitoring it.
When the HP manager told the valve test engineer that it had been cleaned on June 21,1997 the valve test engineer repeated his contention that the water spilled x
during OR05 could continue to evaporate out of the wet insulation, (firewrap) and cause new boron buildup even after it had been cleaned. All of the managers accepted the engineer's explanation. They knew that the engineer had a lot of hands on experience with this particular valve and piping and felt that his explanation was feasible. Therefore, none of the managers asked for any follow-up or pursued any other explanations for the boron buildup The engineer left and did not mention the meeting to his supersisor or to the RCS system engineer. The RCS system engineer when told of this meeting, after the plant shutdown, said that he would have inspected the area at RC-V-89 for evidence of boron buildup within a few weeks of knowing that it had been cleaned. He was not informed that it had been cleaned on June 21,1997.
4.1.7 Ostober 1997 Observations The HP Manager relayed the information he had received from the valve test engineer to his department personnel and the HP technicir.c.s continued to monitor conditions at RC-V-89. An HP technician was assigned the task of cleaning RC-V-89 on October 3,1997. HP documented the cleaning on June 21 but they did not document the cleaning on October 3,1997. Two pictures were taken after the October cleaning. An HP Supervisor called the RCS system engineer shortly after the RC-V 89 inlet pipe was cleaned and told him that they had cleaned the valve and notified the Shift Manager. The RCS system engineer asked him to monitor the valve's piping condition. The RCS system engineer agreed to wait a few weeks and then go into containment r ;onitor the status of the valve. This was the first time the RCS system engineer recalls being notified that the boron buildup at RC-V-89 had been cleaned.
4.1.8 Drcember 1997 Observatiorg On December 3,1997, as a follow up the October conversation with HP, the RCS system engineer went out to investigate the status of conditions at RC-V-89. He discovered that the boron buildup had reappeared. On December 4 he sent an E.
Mail message to various department supervisors and managers advising them that 1
l DMFTReddonS W8 10
e.,
since the boron buildup had been cleaned and reappeared that either the water was
" wicking" up from the soaked insulation on the 12" RHR pump suction line, below the RC-V-89 inlet piping, or there was a leak under the firewTap.
The engineer brought the situation to the attention of his supervisor and manager during their morning department meeting on December 4,1997. The Technical Support Manager informed the management tean on December 4,1997, at the Station Directors morning meeting, that another containment entry would be necessary to remove the firewrap and inspect the piping under it. There was an incore flux map scheduled for that day. That fact, together with the uMerstanding that the piping was low pressure and isolated from the RCS in a star,,y condition, resulted in the containment entry being scheduled for December 5, 1997.
On December 5,1997, a containment entry was made and the firewrap was removed from the RC-V-89 inlet pipe. The pipe was wiped clean and after a few minutes 4 small drops of water appeared on the sur ace of the pipe. There were r
no noticeable cracks or pinholes or any other visual signs of damage or iiregularity on the pipe surface. The control room was immediately notified. A meeting of station management was held and it was decided that since there were no isolation valves between where this piping was and where it connected to the RWST and RHR pump suction outside containment, that Technical Specificatic.r 3.6.1.1 Containment Integrity was no longer satisfied. The associated Action Statement was entered and within I hour plant power was being reduced. The q
plant was brought to Mode 3 and then Mode 4 and Mode 5 within the time limits imposed by the Action Statement.
5.0 CHRONOLOGY Date/ Time Description 1988 or 3" Piping below, (upstream of),1-RC-V-89 was wrapped with a thin layer of before firewrap like material and covered with red duct tape. The 3" piping was captured in a photograph taken in 1988 with no boric acid present. Refer to Figure 2.
t 6/2/87 RC-V-89 discharge flanges had boric atd residue. Found during hot functional testing. WR 87W005122 initiated.
10/19/87 RC-V-89 removed, inspected, cleaned, and reinstalled, per WR 87W005122.
11/29/88 RCTV-89 inlet and discharge flanges had boric acid residue. Found during startup testing. WR 88 WOO 5991 initiated.
10/6/89 RC-V 89 removed, inspected, cleaned, and reinstalled per WR 88W005991.
DAAf7Acwh,5 Jn38 11
i I
Date/Fim) i Description 12/16/90 WR 90W006166 initiated for boric acid leak at flanged connection on RC-V-89.
Boric acid was present on top of 3" inlet pipe just below inlet flange but not present on lower portion of pipe at branch connection to 12" RHR suction header. Work transferred to 91W001930 which performed relief valve setpoint test.
7/26/91 OR01 Cooldown from Mode 4 to Mode 5 using B Train RHR 8/26/91 RC-V-89 unbolted and rotated for OR01 Type C LLRT or. discharge piping, (91RE0012601).
8/29/91 WR 91 W001930 performs RC-V-89 setpoint verification, cleans boric acid residue and corrects flange leak reported earlier on 90W006166.
9/8/92 OR02 Cooldown from Mode 4 to Mode 5 using B t rain RHR 9/29/92 RC-V-89 unbolted and rotated for OR02 Type C LLRT on discharge piping, (92RE0012601).
10/1/93 Forced Outage Cooldown from Mode 4 to Mode 5 1/30/94 Forced Outage Cooldown from Mode 4 to Mode 5 4/10/94 OR03 Cooldown from Mode 4 to Mode 5 using B Train RHR 4/18/94 RC-V-89 had heavy boric acid residue at side plug hole. WR 94001373 written and work was transferred to WR 93W004346 which performed Type C LLRT on RC-V-89 discharge piping in OR03.
5/12/94 RC-V-89 removed from system f'r r,.ief valve setpoint testing,(94RE00116001).
5/17/94 OR03 Type C LLRT performed oa RC-V-89 discharge piping,(?4RE00126001),
prior to reinstalling valve.
7/20/94 ISI Functional Test on RHR Train B piping performed. Piping below RC-V-89 visually inspected with RH/RCS pressure at 300 psig. No leakage or boric acid residue was noted in the area of RC-V-89.
I1/4/95 OR04 Cooldown from Mode 4 to Mode 5 using A Train RHR 11/23/95 RC-V-89 was removed for relief valve set pressure surveillance per RTS 95RE00116001. Piping and insulation below the valve became wetted with reactor coolant when the valve flange mechanical joint was unbolted.
l l
ourrmans mm u
Date/ Time Description 11/23/95 OR04 Type C LLRT performed on RC-V-89 discharge piping,(95RE00126001),
prior to reinstalling va've.
3/20/96 Boror residue on RC-V 89 inlet pipe noted during Operations containment walkdown and noted on ODI 36 5/17/96 HP makes an entry on the boric acid RTS noting a boron buildup at RC-V-89 11/96 RCS system engineer informed by HP of boric acid residue at RC-V-89. The system engineer and HP inspects the valve and finds boric acid residue on flange, fasteners and pipe below valve. Several photographs, (Figure 3), of the RC-V-89 inlet pipe are taken. The boric acid residue is removed from flange and fasteners.
Flange was found dry with no evidence of any more leakage. The matter is discussed with the IST valve test engineer and the history of water from valve disassembly wetting insulation during OR0. is recounted. It is assumed that this was source of the boric acid residue.
2/97 RCS system engineer is told by HP of boron on valve. As IST valve test engineer is going in containment, the RCS system engineer gives him the picture from 11/96, (Figure 3), and asks him to check it out. The IST valve test engineer reports no change from picture.
3/18/97 ISI Functional Test on RHR Train B piping performed by VT-2 QC inspector. RC-V-89 and associated piping is visually inspected with RCS pressure at 240 psig.
Boric acid residue was noted at RC-V-89 inlet flange and below the valve at the weld-o-let connecting the RC-V-89 branch line to the 12" RHR pump suction header.
3/18/97 Engineering evaluation performed by ISI functional test supervisor to accept questionable conditions as is. Boric acid residue attributed to outage activities that drain the RC-V-89 associated piping when the valve is removed for testing. Thece activities had caused wet insulation below RC-V-89 in OR04, (11/95 - 12/95).
3/18/97 ACR was initiated by VT-2 QC inspector to document the poor work practices in OR04 that had caused the wetted insulat!on and had not reported or corrected the condition. The QC inspector was shown a 11/96 picture of RC-V-89 including the inlet piping and concluded that the condition did not appear to have significantly changed.
3/26/97 Work request 97W000833 was initiated by ISI functional test supervisor to address boric acid residue coming out of the RC-V-89 fire wrap on the inlet piping.
DRAFTReshion $ V788 13
l L
I Dateffime Description 3/28/97 RCS system engineer reviews WR 97W000833 and adds infructions to clean boric acid residue and remove damaged insulation,(reddish color).
5/10/97 OR05 Cooldown from Mcde 4 to Mode 5 - B Train RHR 5/17/97 WR 97W000833 work performed by insulators. Material wrapped around 3" pipe below RC-V-89 was not considered by the insulators to be insulation and was not removed. Wetted insulation around 12" RHR suction header is replaced per 97W000833, 5/28/97 HP Tumover Sheet nctation made concerning boric acid residue on inlet pipe to RC-V 89. Technical Support notified.
6/2/97 Type C LLRT performed on RC-V-89 discharge piping. Valve flange is unbolted and rotated in place to perform the test. Piping and insulation below the valve may have been wetted when the valve flange mechanicaljoint was unbolted.
6/15/97 During containment close-out, RCS system engine r discovers fire wrap like approx.
material, (reddish wrapping), not removed and boron not cleaned. Informs supervisor who requests that an A CR be written. RCS system engmeer talks to CS supervisor responsible for insulators but did not generate an ACR at that time. The CS supervisor states that he will follow up on the problem. RCS system engineer hcids WR 97W000833 open pending further evaluation.
6/21/97 HP technician cleans boric acid residue off of pipe.
Third Qtr HP technician checks pipe and finds boron has retarned. Informs department 1997 supervisor and manager.
Third Qtr HP department manager discusses with maintenance, oversight and operations 1997 supervision to discuss issue. The IST valve test engineer is asked about RC-V-89.
The IS T valve test engineer gives boron wicking / buildup theory and his explanation is accepted. (Refer to Figure 4).
10/3/91 An HP Technician finds boron on pipe, cleans it takes pictures and tells RCS system engineer and operations. (Refer to Figure 5).
11/20/97 An HP Technician notices boric acid residue growth on pipe and thinks it should be cleaned on next rounds. He does not tell anyone of this.
12/3/97 RCS system engineer inspects pipe as followup to HP Technician's earlier communication. Observes the condition of the boric acid residue and alerts supervision and management of his f'mdings.
DRATTRedston 5 UM6 14
e Datemme Description 12/4/97 Potential leak at RC-V 89 is discussed at Station Director's meeting and a containment entry is planned for 12/5/97.
12/5/97 Insulation and tape removed from 3" pipe. Thru wsil pipe weepage site is found.
12/5/97 Event Team formed 11:00 12/5/97 SORC meeting held to discuss Operability issues associated with the 3" piping 11:30 upstream of1 RC-V-89 12/5/97 Entered T.S. 3.6.1.2.
11:52 12/5/97 Commenced power decrease at 25% per hour, 12:00 12/5/97 Entered Mode 2 17:25 12/5/97 Entered Mode 3 17:56 12/5/97 Commenced cooldown using RHR Train A 22:48 12/6/97 Entered Mode 4 07:24 12/6/97 Entered Mode 5 15:10 DRArtanms viva
}$
4 4
FIGURE 2 l
if-s
- j g.
.m n
ne.
l pr.
~
.. k f '.
g' I
g
'y y
}.
l h
k,,-~
,a,
- .f k
1 -
i s
YA l
' ^ '
Q l
.. y ls l Y
J-
~
f s'
a K-U-99 EST.1900 rsot W 1LIK L
i f
i 4
4 a
DMFTRo%5 MM l6 i
k-.
. ~.,. _ _ _
m-
- _.... ~ - _ _ -.
.-~~ __ - -.-
J FIGURE 3 t
,; y.jp, p4 Ly'.
y.;,
)
1
..s s n..
..:Ty
'f g
' ' -==
2
^2
~
I t
i g
~
- JT'
.4
~
./
j l
9
-~
d l
)
.~.
=.. -
,o r.
I t
g 1
t Ay '
4.
^
'. y f.
s
- QR e%;,.
(
) cj., -'"..Wegh,"
. ', A
'. ' ~ ~ ~
i 3-
,?
Y
'^
j f
.s
.r:.
',\\.
, -., j,e 4 :,- -..
.s.
4, m.
s; g
x.
y.
e-t e.--
r
.r.
l DMFTR&slon3 l&98 gy
wr a
awh..-..mMwe.Aw.aesa:
e4mur.u' imps 4 mise so m u E sieAM.Je em. wa.s.w.Jo._6e.d.lpm.me.
W Meih h 4 ti.,eli.am e ease w --
&e4.m._
4 4
9 FIGURE 4
- s v.
o)
N l
I bb l
W
.y I
W
- m i
l e
A w
i ep
,*I
~
- f
- w.~,; ?,;;%f C4.? * '*. 4t*,M,Qp-j!;,02C2dWll%
- ?j %>$.*?EU[Iff* Q,.:*n Q
%,Y
.i 7;n:?.i;9:
o.* Q.
d.P..
u,s. k';&
- g;
,5. rd -
.tc nu.yhyM:::fM T.m. -(h, -,_" q-W
) vp
.:n m a. k. r-iv
-..f, age 2:
.o
.+2s m
c--
ac
.sw FIGURE 5
.4_.5 m.a-, uunnecemw-w-s.
't sp.
i
- .)
>y 4
w
.v l
.+;
. T.
ll
.. j I-
- 1
,e 2
3 Y
}
h i
j f.
4
)4 at f
((.
j i
J l
'?
y 4
2
.I W
.)
j
, m
,7 e ',
w J
(J, e
j e
a i
J
,p 4
e, oe q
. h3
.t i
. [f
/
1 2
1 4
Q,,.g '49.u - g g " > % f J an2<?
?:ippy,P"4
- w. * '.3., u.;C"ESpd 2
.%nvf
'p -A 2
- g i4 D kp.Mg
e 7g
,.y y
i
~
7h
.ja%1t.'wg :
-t.,o :. r...x w -
J':t T ayQ n
.>.J..3*.h i
k.
..:m: w : ms m r:s:.~wm.u f
1 i
a i
-DRAFTReiston 5 1/7AF8 18 d
l
~,.. _. _.. -.
l
6.0 DESCRIPTION
OF EOUIPMENT RESPOhLSE
6.1 BACKGROUND
RC-V-89 is a relief valve off the 12 inch suction line to 'B' Train RHR Pump 8B. The spool piece, Line # RC 88-1-601-3, connecting the 12 inch line to the inlet of the relief valve had exhibited weepage from sites in two separate areas. Specifically, it was observed in the heat affected zone (HAZ) of the weld at the weld neck flange on the inlet to RC-V-89 and also on the HAZ of the field weld on the spool piece at the weld-o-let on the 12 inch suction line. The spool piece sits in the vertical direction and is approximately 5 inches long. It is 3 inch diameter schedule 40S, type 304 stainless steel pipe, Per Specification 249-1, the spool piece was to be insulated with 1" fiberglass insulation for heat retention (Note: At this time the insulation is not installed for the spool piece to RC-V-89 nor the identical spool on Train 'A' at RC-V-24). The weepage was identified when the plant was at 100% power with RHR in its normal standby condition.
During power operation, this line is essentially stagnant at a temperature of approximately 100*F and 50 psi (static head of the RWST). During startup and shutdown it may be exposed to a temperature in the range of 300-350*F and pressures between 200 and 300 psig.
6.2 AS FOUND CONDITIONS On December 5,1997, system leakage of a drop every two to three minutes (weepage from four locations) was discovered on the inlet piping to the Residual Heat Removal pump 8B suction relief valve (RC-V-89). The weeps were in the metal adjacent to the welds (heat affected zone) that connected the piping spool piece to the relief valve flange and the 3" to 12" weld-o-let that connected the pipe to the 12" RHR pump suction line.
One weepage site was in the upper heat affected zone and three were in the lower heat affected zone.
Visual examinations of the spool piece indicated that it had been wrapped with a layer of a woven cloth fire wrap directly against the pipe surface and standard red duct tape around the wrap. The condition of the covering indicated that the materials were somewhat discolored and appeared to be fused or matted together. Seabrook Chemistry Department took samples of the material ( both the fire map and the tape collectively) and performed a chemical analysis which identified 349 ppm chloride and <24 ppm fluoride (minimum detectable).
During outages, relief valve RC-V-89 is removed from the system for testing. In the process the discharge line flange is unbolted. This releases water from the line some of which is not collected and wets the 3 inch spool piece, its covering and the 12 inch suction line and its insulation. Although the present chloride readings for the covering materials are 349 ppm, they may have been higher in the past. The covering material was on the pipe since at least 10/88. Removal md testing of RC-V-89 along with subsequent wetting of the area had occurred in the past.
1 1
aunneuens ms 1y
6.3 HISTORICAL REVIEW OF ASME MATERIALS AND WELD DOCUMENTATION Original radiograph: <. f the two welds on the spool piece were retrieved and examineo by NDE Level 111 personnel. The review showed no anomalies in the weld or the base metal. It.vas also noted that the two welds on the spool piece were performed by different velders in different companies. The flange to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o-let to pipe weld was a field weld perfon.,ed at Seabrook. This indicates that the welds and welder qualificatien were not a contribuar to the problem as evidenced by the fact that the radiographs showed no anomalies, the welds were performed by different welders / welding operators, and examined by different NDE inspectors, yet the heat affected zones adjacent to both welds had weepage sites. Also, there did not appear to be any anomalies in the base metal of the pipe which further indicates that the weepage was due to in-service conditions rather than original installation.
A document search of the heat number for the spool piece was performed to determine where else piping of that heat may be installed in the plant. (Appendix B). A chemical analysis completed at Mass Materials Research confirmed that the material was SA312, type 304 the material specified and the correct material for the applications.
6.4 SERVICE LOAD CONDITIONS A review of the piping analysis shows that the stresses in the spool piece piping are minimal. During RHR operation, the combined stresses have a safety factor of 4.75 on the normal design allowable. During power operation with RHR in standby, the safety factor on normal design allowable is about 28.0. The piping design in this area is dictated by the faulted condition with stresses induced by a Loss Of Coolant Accident. In addition, there have been no indications of any mechanical or flow induced vibration in this area, nor has there been any problem history with relief valve RC-V-89 that would have resulted in abnormal.y high thrust loads. To date, there does not appear to be any hl h service induced stresses that could initiate any failure mechanism which could result in the weepage found. The only area on the piping spool piece that has been exposed to high stress levels is b the heat affected zones of the welds due to residual stresses from welding.
6.5 CORROSION MECHANISMS While there does not appear to be any service induced mechanism which could have cc.tributed to the weepage, a combination of the residual stresses in the pipe and environmental conditions (i.e. chloride contaminants) on the outside of the pipe are conditions that make the pipe susceptible to Stress Corrosion Cracking (SCC). Although the chloride level in the wrapper is relative low at 349 ppm, this i'4 itself would be sufficient to initiate the : tress corrosion process when coupled w,th the high residual stresses located in the heat affected zones adjacent to the weld. The higher the residual stress, the more sensitive the material is to smaller amounts o' chloride. The chlorides would also tend to concentrate at crevices on the surface due to the washing of the mtArtanms mm 20
~
4 covering material by the water and subsequent drying off of this water when the line goes back up in temperature. This could result in potentially high levels of chloride concentrated in selective areas around the heat affected zone of the welds.
6.6 TESTS ANDINSPECTIONS The spool piece was removed and a metallurgical analysis was performed to determine,
the cause and extent of degradation that led to the weepage. The piece was removed by making a cut at the weld of the weld-o-let to the spool piece and unbolting the inlet flange at RC-V-89. The analysis involved one section of the spool piece being analyzed at Seabrook and the other section analyzed at ABB Laboratories in Windsor Locks, Connecticut. The piece was cut approximately 31/2 inches down from the top of the flange. The top piece was analyzed at Seabrook and the bottom was shipped to ABB -
Labs. Subsequently, the upper section was also sent to ABB for analysis.
Prior to sectioning the removed piece, a leak rate test was performed on the piece to determine what effect the degraded section would have had on containment integrity.
The test consisted of clamping a gasketed plate over each end of the piece. One end was fitted with a test connection which was pressurized to 52 psi. The leak rate was determined to be 1.363 Standard Cubic Feet per Hour (SCFH) which is much less than the allowed value for a typical penetration which is on the order of 37 SCFH. It was also noted that the leakage appeared to be coming from the gasketed end plates. Results of the test indicate that the pipe spool had not degraded to the extent that containment integrity was challenged.
The following description depicts the work and examinations performed on the section analyzed at Seabrook and the conclusions reached from that work:
A Liquid Penetrant (LP) examination was performed at the weld of the flange to the 3" pipe on the inside and outside diameter of the piece. One linear indication was found that went through the wall. The crack ran at an approximate angle of 45" to the longitudinal axis of the weld. It progressed into the base metal of the 3" pipe. The piece was sectioned at a location perpendicular to the weld (See Figure 6). The sectioned surface displayed a branching type crack running perpendicular to the surface of the pipe at a distance approximately 3/8 of a inch from the weld in the heat affected zone. It appears to have originated on the outer surface and progressed to the inside diameter of the pipe.
' DMTTJtch$ mnt 2}
FIGURE 6 e
1 k
._lM r
.~
, 5., e,
. - -. _-., (
e2 J. '
m
}
\\_.
a s
h
~. : '
~
A wnm Note: The branch cra: king in the piping cross-section shown in Figure 6 above has been magnified by acid etching. Without etching the actual defects would not be visible in the photograph.
The failure appears to be a corrosion assisted degradation process such as chloride induced stress corrosion cracking that had initiated on the outside and migrated to the inside diameter. However, it was decided that more confirmatory information was required from the analysis being done at ABB on the lower section of the spool piece.
Additional needed information includes: obtaining the chemical composition of the pipe, performing an SEM (scanning electron microscope) examination on the crack surface and performing a chemical analysis of any contaminants that may be present on the crack surface.-
PRAFTRedslen$ Irin8 22
\\
After removal of the 3" spool piece, a LP examination was performed on the surface of the weld-o-let in preparation for its replacement. The examination revealed branching type indications similar to those seen on the sectioned piece. Considering the fact that the indications appear to be due to the same mechanism (i.e. chloride induced stress corrosion cracking), it was decided to also perform LP on the 12" suction line as this area i
may also have been exposed to leached chlorides. Some indications were found in the heat affected zone of the weld on the 12" line where it attaches the weld-o-let. The indications were repaired. However, the indications on the weld-o-let were more l
extensive and could not be repaired. The weld-o-let was removed and replaced with a new : 'tting.
Additional analysis and testing of the 3" pipe spool including work performed at ABB
. Labs are included in Engineering Evaluation SS-EV-970031, (Attachment D).
7.0 CONCLUSION
S 7.1 The through wall weepage in the heat affected zones of the 3 inch pipe below RC-
. V-89 was determined to be chloride induced stress corrosion cracking. This cracking initiated on the outside and migrated to the inside diameter.
7.2 The source of the chlorides was the inappropriate material, (firewrap and duct tape) installed on the RC-V-89 inlet pipe. The combined effects of wetting the firewrap, the presence of chlorides, insersice temperatures, and the residual stresses in the piping heat affected zones were all necessary to produce the degraded condition.
7.3 The condition developed over a long period of time, (i.e., years), following the completion ofinitial plant construction. The welds and welder qualification were not a contributor to the problem as evidenced by the fact that the radiographs j-showed no anomalies, the welds were performed by different welders / welding operators, and examined by different NDE inspectors, yet the heat affected zones adjacent to both welds had weepage sites. Furthermore, similar piping configurations with similar inservice conditions, (but without the presence of inappropriate chloride bearing material and wet external insulation), did not show signs ofstress corrosion cracking.
7.4
'tesults of the leak rate test, performed on the removed 3" spool piece, indicate Jat the condition ofit was not degraded to the extent that containment integrity was challenged.
7.5 -
Bascion'the chronology the best estimate for the development of a through-wall weepage site in the RC-V-89 inlet pipe was in the March 19% time frame.
~ 7.6 The boric' acid reduction program and the inservice inspection program failed to identify the RC-V 89 inlet pipe weeps in a timely manner.
Dama,w.= s um '
23
7.7 Reaction' time to problems identified by the Corrective Action Program are in -
need ofimprovement. There were missed opportunities to identify, document, l
evaluate, and resolve the issues prior to December 5,1997..
7.8 Many examples of a " questioning attitude" were observed during the course of this event evaluation. However, the followup validation and verification process was not used.
7.9 There is a wide spread lack of knowledge of the problems associated with the introduction of unauthorized material onto stainless steel piping & components.
7.10 ' No similar situations to those described in conclusions 7.1 and 7.2 were discovered based on the actions taken per recommendations 8.1.1 and 8.1.2.
7.11 The RHR and CBS systems were capable of performing the containment integrity, shutdown cooling, emergency core cooling safety functions and containment -
building spray with the piping section containing the identified defects.
8.0 RECOMMENDATIONS 8.1 PRIOR TO HEATUP (MODE 4)
I Condt.ct a point of origin boron leak determination inside containment using ES 1801.006.
Include in this inspection a look for temporary insulation and /or fire blanket material wrapped around stainless steel piping Responsible Individual: Peterson Category: P Due Date: Complete 2
Conduct a point of origin boron leak determination outside containment using the station boron leak list. Include in this inspection a look for temporary insulation and /or fire blanket material wrapped around stainless steel piping
- Responsible Individual: Peterson
- Category: P Due Date: Complete
.32 Evaluate the results of the above point of origin boron leak determinations. Ensure that all
- identified active leaks are repaired and all boron buildup is verified to be from leakage from mechanicaljoints.- (no thru wall)
Responsible Individual: Peterson:
Category: P Due Date: Complete
- ~ " 4.-
Conduct liquid penetrant testing on the heat affected zone of the weld-o-let to 12 inch pipe
~ weld associated with the RC-V-89 spool.
Responsible Individual: Grillo Category: P Due Date: Complete DaET Rndsden s ' J/z9s 24 -
i l
(
5 Conduct NDE on the heat affected zones of the welds on the RC-V 24 relief valve branch header to include the 12 inch pipe to weld-o-let, weld-o-let to 3 inch pipe and 3 inch pipe to flange.
Responsible Individual: Grillo Category: P Due Date: Complete 6
Conduct a visual inspection, using NDE level Ill personal, of the removed section of the RC-V-89 relief spool piece.
Responsible Individual: White Category: P Due Date: Complete 7
Perform a metallurgical failure analysis on the removed piece of the RC-V-89 relief spool piece. Evaluate the results of the failure analysis to determine if the failure mechanism is applicable anywhere else in the plant.
Responsible Individual: White Category: P Due Date: Complete 8
Conduct a Chloride and Fluoride leachability test on the material that was wrapped around the RC-V-89 relief spool piece.
Responsible Individual: Linville Category: P Due Date: Complete 9
Review the work control and corrective action data bases for similar cases ofinappropriate materials on stainless steel piping. Address any identified issues.
Responsible Individual: Peterson Category: P Due Date: Complete 10 Repair / Replace degraded piping at the inlet flange to RC-V-89.
Responsible Individual: Peterson Category: P Due Date: Complete 11 Complete an engineering evaluation of the degraded condition to determine if the RHR and CBS systems would have been capable of performing their safety function.
Responsible Individual: White Category: P Due Date: Complete 12 Complete the independent reviews of the VT-2 inspections.
Responsible Individual: Beuchel Category: P Duc Date: Complete DMFTRedston 3 iM8 2$
} *.,;
\\
x 8.2 L PRIOR TO STARTUP (MODE 2) 1.
Ensure the VT2 inspectors performing the inspections and test supervisors performing evaluations required by recommendation 8.2.2 have been briefed on this event.-
Responsible Individual: Beuchel Category: P Due Date: Prior to Mode 2 1
- 2. - During mode 3 at NOP/NOT perform EX1810.001 RCS functional test Responsible Individual: Beuchel Category: P Due Date: Prior to Mode 2 3
Reinforce the expectation and reason why an ACR minst be wrinen for conditions that are outside of normally expected conditions. Communicate the imponance of a questioning attitude when accepting explanations for unexpected conditions. Use this event as an example.
Responsible Individual: Feigenbaum Category: P Due Date: Complete 4
Complete an evaluation of the corrective action program to determine if any changes are
]
necessary prior to startup.
Responsible Individual: Makowicz Category: P Due Date: Complete 8.3 SHORT TERM (< 90 DAYS) 1.
Complete a root cause on this event.
Responsible Individual: Makowicz Category: P Due Date: January 31,1998 2
Conduct confirmatory laboratory testing to validate the exact failure mechanism. (ABB final report)
Responsible Individual: White Category: P Due Date: January 31,1998 3
Incorporate the information recorded during the inspections performed as a result of 8.1.1 and 8.1.2. into the boric acid monitoring program.
' Responsible Individual: Beuchel Category: P Due Date: March 15,1998 4'
Generate an EWR to add a drain line to the relief discharge lines for RC-V-24 & 89 4
Responsible Individual: Kline
- Categoryi P Due Date: March 15,1998 c
5 Evaluate the roles and responsibilities in the work control close-out process to determine if
' it is appropriate to have non supervisory personnel sign the[ supervisor signature block] on 4
work requests and RTSs.
Responsible Individual: Makowicz Category: P '
Due Datei March 15,1998
.' D a A m en M an i I/7 d 26
...=
a-
.a.
8.4 LONG TEPM (> 90 DAYS) 1 Enhance the boric acid reduction program using the results of the root cause and benchmarking with other plants.
Responsible Individual: Beuchel Category: P Due Date: May 22,1998 2
Develop a program te increase the awareness of station and contractor personnel concerning the use of unapproved materials on stainless steel.
Responsible Individual: Linville Category: P Due Date: May 22,1998 3
Buy new photographic equipment for the HP department Responsible Individual: Linville Category: P Due Date: April 10,1998 4
Evaluate the effectiveness of configuration controls that are applied to the removal and reinstallation of thermal insulation.
Responsible Individual: White Category: P Due Date: June 30,1998 5
Develop a station policy for addressing wetted insulation.
Responsible Individual White Category: P Due Date: June 30,1998 6
Include a discussion of this event in the OR06 briefmg package. This package should irclude, as a minimum:
The executive summary of the event evaluation report The information from recommendations 8.2.4, 8.3.1, 8.4.2, 8.4.4, and 8.4.5 Responsible Individual: Anderson Category: P Due Date: December 31,1998 7
Evaluate the practices of draining systems that could result in wetting insulation.
Responsible Individual Peterson Category: P Due Date: May 22,1998
' DRAFTRidslon 3 1/788 27
j
)
9.0 EVENT EVAI,UATION TEAM i
Event Team Manager:
Jerry Peterson Event Team Leader:
Gregg Sessler Event Team:
Brian Brown Michael Harrington James Hill Richard Julian John Lavole Thomas Schulz 10.0 ATTACHMENTS
.A)
Corrective Action Program - Failure Scenario B)
Documentation History C)
Industry Experience D)
Engineering Evaluation SS EV 970031 E)
Engineering Evaluation SS EV-970030 Evaluation of Safety System Function 2
4 e
,;..x
.m,.
saarrandWons. Anm :
28
. ~
ATTACHMENT A Failure Scenario - Corrective Action Progmm Note 71se purpose ofthis evaluation is to determine ifchanges are needed to the Corrective Action Programprior tc rtartup. llris has set been determinedto be the root cause ofthis event. The root cause enluation willcontinue. Issues relatedtoprocedures(processes, accountability, lateralintegration, etc. will be consideredin this evaluation. Also note that the expectation listedin this table is the cur ent expectation, and not necessarily the expectation that existedat the time. l'he pumosefor using the current expectation is to ensure that the currentprogram is adequate.
Date Activity Current Espectation Current Expectation Met?
19/88 This insulaten has bem installed since at least 1988. No Unexpected conditions are reported and No. Multipic (or one who saw i: ever questioned the fact that h looked evaluated.
organizations did before) difTerent than normal insulation.
not question this until covering er that it 12/5/97 was cely on *B' train.
ORS 4 Test engineer keews of water drained from RC-V-89 Adverse conditions shall be reported.
Yes,among some reliefdischarge flange onto 12 suction line wetting documented, analyzed, and/or trended peopic. There is not insulation. This condition was not viewed as urmgable a clear expectation at this time.
that wetted insul. be reported.
3/96 Operations identified boron on RC-V-89 inlet piping and Follow normal work control pie as No. Should place documented it on the containment walkdown list (ODI-directed in ODI-36.
on RTS or write a 36), which was the expectrion at the time.
work request.
I1/96 HP technician identifies boron buildup at RC-V-89 and Per the requirements of the boron Yes, by existing notifies his supervisor. System engineer was notified of program, notify the program support boron program.
problem. (No ACR or WR generated, ownership of engineer of boron from non-mechanical This should be problem transferred to system engineer) sources.
evaluated for rev.
I Date Activity Current Expectation Current Expectation Met?
System engineer inspects and finds no evidence of Adverse conditions shall be reported, No. With the pomt leakage. Takes picture and discusses with test engineer. Is documented, analyzed, evaluated, and/or of origin unknown.
told of spill in OR04. Origin of boron not established.
trended.
System Engineer (Consistent with the pup.m at the time no WR or ACR should document gener.ted, supervision not informed) Wrong assumption on an ACR or WR.
( AUS) was made that boron was from spill theory by test Also, the non-engineer and system engineer. Other po:;sible sources, standard insulation including through wall leakage, were not evaluated. Once and the fact that it was determined that it was from the spill, no one put the only one train had pipe on the RTS as a housekeeping item.
insulation on the 3" pipe should have been pursued.
l 2/97 IIP notifies systerr. engmeer to again evaluate the boron Follow up on :: previously identified Yes. Actions were buildup. Ilas seen no evidence that anything has been problem consistent with done.
boron program 2/97 System engineer discusses with test engineer, gives test Adverse conditions shall be reported.
No. Again the engineer 11/20/96 picture and test engineer enters documented, analyzed, evaluated, and/or deficient condition containment and compares current condition with picture.
trended.
is noted and not Test engineer already has mind set (hU4) that boron is due repe-ted or to spill during OR04. Told system engineer no change.
evaluated.
3/97 VT2 inspector reports questionable or indeterminate Adverse conditions shall be reported, Partially. The condition to own supervision and ISI supervisor. Checlo documented, analyzed, evaluated, and/or inspector wrote an for moisture, finds none. Documents on RTS and writes trended.
ACR, but limited ACR 97-0583. VT2 insnector buys into spill / wicking the problem theory"
.d initiates ACR for OR04 poor work practice.
statement to poor Tunnel vision / group think (SKI) starts to become a factor.
work practice based on the test engineer's theory, and did not inci'ide suspect firewrap.
t DRAFTRevhloor3 inns 3g
.~
Este -
Activity Current Expectatiew Current Expectation Met?
3/97 ISI supervisor initiates work request 97WR000833 to Questioning attitude is practiced in daily No. 'Ihis clean or replace insulation as required due to spill during decision making.
evaluation wm OR04 theory. Evaluation written far ISI inspection base.1 on the acceptancejustifying boron as non-leak related.
unevaluated theory.
Spill / wicking theory is basis. Tunnel vision / group think (SKl)is present.
3/97 MRT reviews ACR 97-0583 (identifying OR04 poor work MRT will ensure that proper action is Partially. Based on i
practices and failure to correct the deficient condition).
taken in response to a deficient the information ACR was dispositioned as trend only.
condition.
available to them this 4 4 J to be an isolated wo k performance issue.
Ilowcver, the
- i. % wm ; ate firewrap was not questioned based on the photograph.
5/97 Insulators removed and replaced insulation at base of Questioning attitude is practiced in daily No. The supervisor relief valve connec; ion to 12" pipe per WR 97-833.
decision making.
or system engineer Intended renoval of reddish firewrap on 3" line not should have been performed due to vague work package which was not contacted.
questioned by the insulators. (DP) Document preparation j
issue. Did not question " reddish color" reference on work package.
I i
I 3
i f
DRAFTRnisien 3 1/7MR 31 i
,.r Date Activity Current Expectaties Current Espectaties Met?
667 System engineer walks down containment and notices Adverse conditions shall be reported, No. The enginecr's reddish firewrap(insulation?) not removed. Told documented, analyzed, evaluated, and/or sonenisor directed supervisor and Construction Services supersisor about trended.
him (correctly) to perceived performance issue ofinsulators. Told by write an ACR.
supervisor to initia'e ACR. Did not wTite ACR, did not This was never close out 97W000833 because intended work not done. Also, the performed. Supervisor did not follow up on the ACR.
work could have been performed at this time based on plant conditions.
637 Boron cleaned by IIP on 6/21/97just prior to containment To clean bron and document.
Yes.
close.out.
737 Supervisor and IIP Manager notified ofIIP technician's Problems are identified, documented, Yes.
identification of boron buildup on RC-V-89 inlet piping.
and corrected in a timely manner. If they can't be corrected promptly then interim meas'.es are established 737 IIP Manager meets with several station managers to Questioning attitude is practiced in daily No. The theo y dims the boron imildup on the firewrap that was decision making.
was not challenged.
discovered in July. Test engineer explains history and An ACR or WR spill / wicking theory to managers present. Managers accept Adverse conditions shall be veported, was not generated.
test engineer's explanation. Tunnel vision / group think documented, analyzed, evaluated, and/or (SK1) is present. Further questioning or pursuit of follow-trended.
up explanation did not occur.
1967 LIP technician again reports boron on the firewrap and Follows up on a previously identified Yes. (IIP) i cleans it. Supervisor and system engineer are notified. IIP problem supervision also notified Operations Shift Manager of the situation. IIP doesn't write WR because aware that system engineer still has one open (97W000833).
DRAfTKeMar3 UM8 32
t j
Dnee Activity Ce m etEn g i Corrent Espectatian Met?
- 12/3/97 Systern engineer.wis pipe as follow-up to HP report Adverse conditions shall be reported, Yes. Infonned of boren proldem. CL. #s boron is back and alerts '
documented, analyzed, evaluated, and/or supervisor and
.._.,,-- -4 of findings. No ACR initiated by system trended.
drafted ACR.
engmeer.
i I
- 12/5 &7 Firewrap and tape em.mvcd from 3" pipe. Thru wall weepage sites dks,.mmi. Entered T.S. 3.6.1.2 and i
4 commenced plant shurdown.
l Condusions i
\\
l This event.,y.m.e a failure of the corrective action program to identify, document, evaluate, and resolve a deficient condition in a timely l
manner. However, this event does not isy.wnt a widespread treakdown in the program for the following reasons.
The fundamemal problem concerning the corrective action program's failure to resolve this condition in a timely manner is that there were l
several missed opportunities to document put,kes using the conective action program. These opportunities were missed based partly ont a 9
w.ag assumption and misjuds..wn; by the system engineer and a program wpport engineer. Evalation of all possible sources of the boron l
(which would include through wall weepage) by the system engineer and a pqri n %,vit engineer would have Edentified the problem. A i
willingness to accept the opinion of the designated expert without challenge by other personnel, including managers, contributed so the
(
orgasNion missir:g opportunities to recognize a problem that required dca.....;.iion and management review in a timely manner.
4 The firW missed opportunity was upon the discovery of boron on November 1996 with the poim oforigin not established (a theory was
{
- ~
propo:M tmt there was no attempt to verify it). Again in Fst,.-y the condition was noted by HP but the system engineer and the prograrr.
l I
support engineer continued to believe there was no leak and that they knew the source of the boric acid buildup. In Marci. an ACR was written, but its problem statement was limited based on the theory dcvc'~psd by the evogram support engirieer. In June, when the WR l
written in March failed to result in the intended work being co...ekka, an ACR was not initiated despite one bemg r q=a*d_ This would j
have been' appropriate since the work was poifvi...cd on the wrong cc...y wat (the insulation on the 12" pipe). Again in July and in October the condition was again identified and communicated by IIP supervision. Given the explanation of the source of the boric acid buildup, no j
followup action was taken. Failure to dca.s...c..; these rvi,k.as removed an opportunity for the organization to ensure that the boron had l
j been properly evaluated.-
i i
3
-mrArraww =5 mms 33 i
i L
=..
=
b f A failure also L.J in the work control pwcess when it was recognized that a work request had not beer implemented as intended without any followup action.- The work request is still open. This is an exampi: where not asking the next question removed an opportunity to resob : this problem in a more timely manner.
In all cases there was cvaa to suggest that there was not a leak present (no detectable moisture and no evidence, until July 1997, of an increase in boron on the pipe). However, there was also no evidence of the point of origin of the boron, only a theory that was never tested or verified (e.g., through cleaning and regular monitoring of the 3" pipe). There was also no evidence that the reason unusual " insulation' I
was on'one train but not on the other was ever pursued.
i Enhancing the organization's questioning attitude is an existing mission of the Corrective Action. Program. Past actions include information passed in the OpNing Exg.cr.ce Summary and in Seabrook Today. Better mechanisms for transmitting Seabrook case studies (root cause 6
' evaluations performed on Seabmok events that show where a more rigorous questioning attitude could have prevented an event) are being investigated.
ReceenneendedAcdien
- Reinforce the expectation and reason why an ACR must be written for conditions that are outside of norally expected conditions.
Communicate the importance of a questioning attitude when accepting non-standard explanations for unexpected conditions. Use this event as a case study.
' Action: Feigenbaum Due: Complete i
I I
I b
i RNrAFTReeMon 5 1/7At 34 i
~ ~
I ATTACHMENTB Archived / Historical data retrieval for documentation associated with 1 RC V 89 The following documentation has been retrieved and retained for review associated with 1 RC.
V 89 event. The documentation reflects the history of fabrication and installation of piping, including material traceability for components utilized in their fabrication.
Dravo Spool Sketch E2936-2280- 3" Std WT. SMLS Pipe, SA312 TP 304,117# 04935.
e MTR# 411. Final approval of fabrication was completed and signed off by Dravo, Marietta, Ohio, on 4 7 82. This welded piping assembly consists of 3" spool, and a 3" RF WN FLG, with a weld designated as shop weld "C" to spovl piece E2936 2240. Note-Radiography 3
was performed on shop weld "C" Review of radiographs was performed 12/06/97 by NAESCO Level Ill, no discrepancies were noted in the original radiographs. Documentation for Spool Sketch is located in SBYDCC on Roll 4113, Frame 310, and in hard copy record.
Material Test Report # 411-3" Piping was supplied to Dravo, by McJunkin Corporation, e
from Al Tech Specialty Steel Corporation, Dunkirk, N.Y. MTR# 411 indicates that two(2) heats of 3" piping were supplied, consisting of Ht. 04935 and Ht. 04785, with a total quantity fumished of 408' 7/8" Document location in SBYDCC, CSN# 1862, Roll 768, Frame 1707.
NOTE, Additionally on 5/13/86 a 8' length of 3" Sch 40, Tp 304, Ht # 04935. MTR Hil was received by UE&C from Dravo on Receiving Inspection Report, RlR#13029. Review of documentation it appears that this piping was part of the original purchase received by Dravo. Documentation for these records is located in SBYDCC, Roll 6774, Frame 1284.
Ikavo Snool Sketo E2936 707-Item #4,3" on 12" STD. WT. Weld-O Let, sal 82-F304, llT# L342, MTR.# 42. This is the 3" connection (1 RC-88-01, F0108) and Dravo Spool E2936 2280. Fabrication of spool assembly was completed I l-30-79, for Dravo shop weld "F". Examination of final weld configuration was performed by PT examination on 1016-79, with acceptable results. Documentation for Spool Sketch is located in SBYDCC hard copy fiies.
Material Test Report # 42 1412 X 3" Weld-o let, HT# L342 was supplied to Dravo by e
- G+W Bonney Forge, Carlinville,111. Documentation of MTR# 42 is located in SBYDCC, CSN# 726, Roll 330, Frame 411.
P/H Fleid Weld Packmee.1-RC-88-01. F0108-Review of Pullman-Higgins field weld 1-e RC-88-01,F0108, which consists of spool piece E2936-2280 and E2936-707, was completed 3 1-83. Radiography of this weld was performed originally 3-7 83, film discarded, due to unacceptable film (-15%) density. Weld was re radiographed 7-15-85, and 7 20-83, with
, acceptable results. NOTE Radiographs were retrieved and a review performed on 12/06/97 by NAESCO Level 111, with no discrepancies noted.
Documentation for above weld package is located in SBYDCC, Roll 5333, Frame 1384.
P/H Welder Oumimcatlana. Stamp No. "RP"- Retrieved and reviewed welder e
qualificatior.s "RP". Records were requested on 12/06/97. Documentation for these records t
l
j i
may be located in SBYDCC, Roll 4021/ Frame 1731, Roll 4197/ Frame 2355, and Roll 5825/ Frame 2075.
P/H As Bulli Verl&stion Walk Down As built verification was performed and completed e
on 1 19 85, for 1 RC 88 01. BlP RC M 3. Documentation for this package may be located in SBYDCC, Roll 5333/ Frame 1384.
Renetitive Taak Sheets for VT-2 E===l==ttans - The following RTS associated with the e
performance of VT 2 examinations conducted on piping system (s) related to 1 RC-V 89.
RTS# 94RE00185001 i
RTS# 93REN183Hi RTS# 97RE0vl85001. Identified Boric Acid residue in the area of 1 RC-V-89.
Work Requests Related to 1 RC-V 89 activities (Boric Acid Residue) e WR# 87 WOO 5122 RC.V 89 discharge flanges had boric acid residue. Found during hot functional testing. _ (SBYDCC, Roll 673. Frame 830)
WR# RRWOO5991 RC-V 89 removed, inspected and reinstalled. (SBYDCC, Roll 1
1274, Frame 938) l WR# 9CWOO6166 RC V-89 flange leak. (SBYDCC, Rcll 2135, Frame 1130)
WR# 91W001930 RC-V-89 no leakage identified. (SBYDCC,2190, Frame 1953).
WR# 94W601373 RC-V-89 leaking at side plug hole, heavy boric acid buildup.
(SBYDCC, Roll 3827, Frame 254)
WR# 97W000833 WR. initiated to address boric acid crystals on insulation and wrapping under valve. Instructions added to WR, to clean boric acid residue and remove damaged insulation,(reddish color).
(WR. maintained in hard copy file)
Dravo Spool Retrieval by HT# 04935 and HT# 04785.
Unit I(E2936) Dravo Spool Pieces Spool Pc. #
Item #
Ht.#
Feet Inches Remarks l
E2936-4%
1C 04785
-1 E2936-462 1A 04935 18 9%
IB 04785 3
6%
+
IC 04785 1
95/8 E2936-466 1A 04785 11 91/8 IC 04785 3
E2936-472 1B
-04785 10 5/8 l E2936-517 2A 04785 1
3 2B 04785 5
4 2C-04785 10 6%
E2936 523 1A 04785 6
8%
E2936 714 1
04785-13 7 15/16 E2936-715 iB 04785 3
11 i
- SAMT Andslen J l/Mt -
36-m
, ~. - -,,
- --.,we,..-,.-
-,c,,
, -.i..,
Spool Pc. #
Item 8 H t.#
Feet Inches Remarks 1C 04785 11 %
1D 04785 2
10 1/16 E2936 716 2
04785 5
10 15/16 E2936 719 2A 04785 4%
2B 04785 6
6%
2C 04785 9
2 9/16
, E2936 720 1
04785 15 2 9/16
' E29?6 721 1A 04785 3
67/8 IB 04785 1
3 E2936 832 1A 04785 4
~
IB 04785 9 9/16 1C 04785 8
1D 04785 5 7/8 E2936 833 2A 04935 2
8 3/16 E2936 930 1
04935 5
E29361019 1A 04785 11 15/16
~
1B 0493S 20 57/8 li2936 1156 1
04785 10 215/16 E29361157 1
04785 7
15/8 E2936-1158 1A 04785 2
4 3/16 IB 04785 9
3 E2936-1194 1A 04935 5 15/16 IB 04935 6
1C 04935 3
3 ID 04935 1
6 IE 04935 3
E2936-1204 1A 04935 3 11/16 E2936-1204 1B 04935 3
6 15/16 1C 04785 3
6 15/16 E2936-1522 1
04935 77/8 E2936-2147 1A 04935 3
IB 04935 4
E2936 2202 1A 04935 1
7 7/16 1B 04935 2
10 7/16
~~
E2936 2135 1A 04785 1
7 5/16 IB 04935 9
j Total 205 11 13/16 c a 4 r r a,w w.a 5 m m 37 m
-m w
-4 m
m7 ev w-
4 ATTACIIMENT C INDUSTRY EXPERIENCE ASSESSMENT Impact of Halogens on Stainless Steel Surfaces Surface contamination of stainless steel piping foreign material was the subject of a Westinghouse Technical Bulletin (NSD 'IB 80-2, " Cutting Oil Used on Reacter Components")
reviewed by the ISEG in 1985. This resulted in recommendations (SSP #850025, dated 1/10/85 and SSP #850562, dated 7/25/85) regarding a consumable program, now the expendable products program. These recommendations were all implemented by November 7,1986.
The resulting expendable products program is largely based on the concem of halogenated material that could potentislly lead to stress corrosion cracking. It has been largely successful, but some incorrect materials have been used. Some ACRs have been dispositioned on the use of improper tape. In addition, the event evaluation team identified several ACRs conceming the presence of other foreign materials on stainless steel, one example being melted plastic.
The fire wrap with red duct ta,* may have been placed on the pipe during construction days.
There are some instances that may be more recent. The ACRs previously written demonstrate that NAESCO personnel are aware of the requirements. Corrective action upon discovery of foreign substances however has not always been timely.
Events of Corrosion of Borated Water Systems There have been a number of previous events in the industry of corrosion of borated water systems that have some pertinence to ttle subject event. Corrosion of borated wat:r system components has resulted in a number instances of degradation or leakage. A search ofindustry experience has revealed four types of common events, as follows:
- 1. Halogen (most often chloride)-induced, transgranular stress corrosion cracking (TGSCC) in heat traced systems containing high boric acid concentrations (typically 12 weight percent)
- 2. Boric acid corrosion of carbon steel components 3 Stress corrosion cracking caused by local strerses and external contamination
- 4. Primary water stress corrosion cracking (PWSCC) ofInconel 600 material There have also been several other related events. These are not as common as the other events, but also pmvide lessons leamed or insights, in omer to determine if further lessons learned are applicable to the subject Seabrook Station event, each of the above type of events were reviewed and are analyzed below:
Halogen-Induced Trangranular Stress Corrosion Cracking (TGSCC)in Heat Traced Systems Containing Mich Boric Acid Concent ations l
DMFThedslan 3 1/788 38
e *,
This phenomena involves both internal (from the inside out) and external (from _.: outside in) attack. The cases of external corrosion all involved wetted insulation. The concentration of boric acid in the system were high w, scally 12 weight percent). These rystems were all heat traced.
The problems occurred in stagnant piping regions. INPO and NRC database searches sevealed the following eve.o of this type:
Document Event Title Cause(s)
Remarks Date INPO SER 12/28/82 Cracking in Internal TOSCC; near A chemistry evaluation of the 57 83 Stagnant Boric Acid welded regions;possible inner surface of the piping Piping sources of revealed contamination levels contamination include a of up to 110 ppm chlorides,200 contaminated batch of ppra sulfates, and 84 ppm boric acid or recycled fluorides. These levels were contaminated boric acid.
high enough to cause cracking of the type 304 stainless steel when combined with the stresses associated with the welding process.
OE 65?5 by 2/17/94 Stress Corrosion Stress corrosion Very poor maintenance of Millstone 2 Cracking of Boric cracking; boric acid system. A number of packing Acid System Piping residue collected in the leaks and other component and Fittings insulation, forming leaks soaked the insulatio..
concentrated halogen compounds.
OE 7201 by 3/31/95 CVCS Boric Acid Extemal and intemal Possible source of pipe surface Turkey Transfer Piping TGSCC; presence of a contamination include poor Point 3 Through Wall halogens on the outside construction practices, Leakage and inside surfaces of contaminated insulation, and the piping the use of t r.vice water to wash down piping components.
Possible sources ofinternal contamination include pour construction practices and contan.inated boric acid.
DMIT Aewslen 3 M98 39
J3oric Acid Corrosion of Carbon Steel Comnontnis j
Exposure of carbon steel to buri: acid has resulted in a number of events of significant corrosion of material. INPO and NRC database searches revealed the following events of this type:
l 1 % aent Event Dale l'
~ Title -
' Cause(s)
- Remarks,
SE!.72 83 Point Beach:
Damage to Carbon Boric acid induced Body to-bonnet valve 4/28/83 Steel Bolts and Studs corrosion leaks. In the Yankee m Valves on Small Rowe case, a valve Yar '
Rowe:
diameter Piping body defect resulted in a 4/5/5s Caused by Leakage high velocity spray onto of Borated Water the studs.
NRC IN 86 1/14/87 Degradation of Boric acid induced Wactor vessel head 108.S.)
Reactor coolant corrosion of carbon multiple corrosion-System Pressure steel; source of boric induced degradations.
Boundary acid was leak from a Demonstrates lower instrument tube importance of small seal on one of the leakage onto nearby hot instrument tubes, surfaces.
Generic Based on 5 Boric Acid Corrosion Boric acid induced item 2 states that it is Letter 88-05 events,4 in of Larbon steel corrosion of carbon steel important to establish the 1986 and 1 in Reactor Pressure potential path of the 1980 Boundary leaking coolant and the Components in PWR pressure boundary Plants components it is likels o contact.
Generic Letter 88-05 has resulted in our boric acid reduction program. The issue related to the prevention of spilling or drops onto caibon steel components is also applicable to spillage onto insulation that covers hot stainless steel piping. Such insulation inside the containment building is covered with metal. Only small areas are exposed to the problem. Nonetheless, to help preclude possible ! aking into non metal covered portions, the event evaluation 1:am recommends an assessment of addressing wetted insulation.
DMFTRedston 5 UMi 40
4 Stress Corrosion Crackino Cauwd by Local Stresses and Fxternal Contamination This phenomena results from very localized stresses. INPO and NRC database searches revealed the following events of this type:
{Tited.
~
j Casse(s) L 14emat.;5 LDoeseeni Event l U
g
- Dete Yankee Rowe 1/24/84 Reactor Coolant Cold bending without This is the only documented LER 84 001 Pressure Boundary subsequent softening, failure of this type in the Leak abrasions, and history of this plant's concentration of operation. Insulation chlorides that had exhibited compliance with leached out ofinsulation Regulatory guide 1.36.
Water (source not explained in the report) soaked through the insulation, transporting leachable chlorides to the hot (520'F) pipe surface where subsequent evaporation concentrated the chlorides.
OE 3088 by 12/21/88 Chloride incucad TGSCC; source of condensed steam vapor is Duant Arnuld Transgranular Stres?
chlorides is decomposed believed to have run Corrosion cracking electrical cabling through ti.e decomposed of CRD Piping insulationjacket.
v.able Jacket, resulting in leaching of chlorides.
Water and chlorides drained from a field option box located directly and
- tripping onto the affected pipe Turkey Point 4 3/10/94 Unit 4 Shutdown Outside diameter RCS leakage was from an LER 94-002 Due to RCS chloride induced incore instrument guide Boundary Leakage transgranular stress tube.
corrosion cracking; 4
source of chloride contamination was improper cleaning techniques EAR MOW 8/26S5 Damage of Primary A net like population of Plant is Kalinin NPP Unit 2,95-044 Circuit Drain Pipe corrosion induced which commenced under Wet cracks from chloride commerciai operation in insulation contamination from both 1986.
a weld leak a on loop drain line and from the thermal insulation.
f DA4FTnoddu 5 V7ac
_4}
~ -..
~..
Primary Water Strent Corrosion Crackino (PWSCC) ofInconel 600 Material This phenomena is internal attack, i.e. from the inside out. It was first identified in 1987, and new events continue to occur. There have been a number of problems on pressurizer connections, exclusively of the Combustion Engineering design, which are constructed of this material. The PWSCC ofInconel 600 phenomena has more recently been detected on reactor vessel head penetrations (applicable to Westinghouse plants also). INPO and NRC database searches revealed the following everts of this type:
Deemassef
? Event Date (Titief ICmMs) 4 d R5enkrksl a.
SER 2 90 Pressuriur heater PWSCC -
Combustion Engineering Sleeve Cracking-pressuriur NRC IN 90-Primary water PWSCC Combustion Engineering 10 Stress Corrosion pressurizer Cracking ofinconel 600 Calvert Cliffs 3/21/94 Pressuriur Hester Source of stress was Combustion Engineering I LER 94-003 Sleeve Cracking axial scratches, pressuriur mechanism of corrosion was PWSCC San Onofre 3 7/22/95 Reactor Coolant Combustion Engineering LER 95 001 System Pressure pressuriur Boundary Weepage San Onofre 3/3/97 Reactor Coolant PWSCC Combustion Engineering Unit 2 LER System Leakage pressuriur 97 004 from Pressurizer Thermowell Generic Several events of Degradation of PWSCC and The IGA attack was at Letter 97-01 1991 to 1994 Control Rod Drive intergranular attack Zorita involved Mechaism Nonle IGA ccatamination by resin I
and Other Vessel bead injection into the Closure Head RCS.
Penetrations MWM 3 l/iM.
~ 42-
~
l __
., e J
Other Related Evente l
Document Esent Title Cause(s)
Remarks l
Date INPO SER 12/21/83 Stress CorTosion Internal stress corrosion The halogenated 83 83 Cracking of Pilot cracking;halogensted hydrocarbon was suspected Line for Pressuriur hydrocarbons and to have bem a mixture of Safety Valve oxygen had app,rently trichloroethane and I
been introduced into tetrachloroethene, a volatile the pressurizer from the substance routinely tud for auxiliary spray line and cleaning components concentrated in the s*eam space WW4 Diablo 10/9/87 Accumulator Nozr'c IGSCC; stress was Source of contamination Canyon 2 Cracking Due to improper welding could be from consumables 1.ER 87 Intergranular Stress pxess; contamination used during the welding 023 01 Corrosion was chlorides and process, or uncleaned off sulfur found in the contaminated hydrotest socket region of the water nozzles OE 6536 2/17/94 Through Wall Leak intergranular corrosion; Defects wcre in heat by on an ECCS Pump stresses due to weld affected zones of the Pallisades Suction Valve repairs made during castings i
manufactare Surry 2 9/12/95 Operation with Non-Circumferential cracks The leak is believed to have LER 95-Isolable Leak on of pressurizer nozzles initiated from the inside out 007-01 Pressurizer by internal TGSCC and and subsequently created an Instrumentation extemal TGSCC and environment on the outside Nozzles lGSCC of the nozzle that led to the extemal cracking Surry 2 2/23/96 Loss of containment Local stresses and heat The affected section of LER 96-Integrity Requires affected zones of p4ing had been in service 011 Plant Shutdmyn overlapping welds in for 23 years, bent radius piping at a fixed saddle support i
f DMFT AeWsdea $ MM 43
Conclusions Triese industry events, along with the Seabrook Station event, demonstrate the importance of avoiding contamination of piping systems (both extemally and internally) Particularly, this data demonstrater thi externally generated corrosion requires simultaneous and unique circumnances and conditions to create a problem. Generally, high temperature, halogen contamination, and a stress (during the weld processing and/ or operational) are requirca.
I Wetted insulation can result in leachable chlorides of a magnitude (but must be in conjunction with the other simultaneous conditions) to result in a problem. As evidenced by the above data,
this is a rare occurrence. Nonetheless, diligence is required to avoid a continuous type wetting condition. Also, should a significant spill occur, as during disassembly evolution, insulation should be removed, the piping cleaned, and dry insulation be installed.
' It is also imperative that affected workers are knowledgeable on the importance of not allowing unqualified foreign substances to remain on stainless steel piping.
a DMFTReMon 3 IMM -
44
-s -
s
.,3. c i -
ATTACHMENT D ENGINEERING EVALUATION SS EV-97-0031 RC-V-89 Inlet Spool Piece Failure Analysis Engineerlag Evaluation Nunnber SS-EV 97 0031 1.0 Purnon RC V-89 is a relief valve ofTthe 12 inch suction line to 'B' Train RHR Pump 8B. On 12/5/97, the spool piece, on line # RC-88 1-601-3, connecting the 12 inch lire to the inlet of the relief valve exhibited weepage at each end. Specifically through wall weepage was observed in the heat affected zone (HAZ) of the weld at the weld neck flange on the inlet to RC V-89 and the HAZ of the field weld on the spool piece at the weld-o let on the 12 inch suction line. The purpose of this evaluation was to determine the cause of the through wall weepage.
2.0 Background
The spool piece sits in the vertical direction and is approximately 5 inches long. It is 3 inch diameter schedule 40S, type 304 stainless steel pipe. Per Specification 249-1 the spool piece was to be insulated with 1" fiberglass insulation for heat retention (Note: The insulation was not installed for the spool piece to RC-V-89 nor the identical spool on Train 'A' at RC-V24).The weepage was identified when the plant was at 100% power with RHR in its normal standby condition. During power operations, this line is essentially stagnant at a temperature of approximately 100'F and 50 psi (static head of the RWST). During startup and shutdown the piping is heated to its normal startup/ shutdown temperature range of 300-350*F for about 3 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br />. Over the life of the plant this would work out to a " time at temperature" of approximately 400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br />. A total of four small weepholes were observed on the spool piece weld heat affected zones. One at the top weld which affixes the spool piece to the flange, and three at the field weld which affixes the weld-o let to the spool piece.
Visual examinations of the spool piece indicated that it had been wrapped in a two layer covering consisting of a woven cloth fire wrap against the pipe surface and standard red duct tape around the wrap. The condition of the covering indicated that the materials were somewhat charred and appeared to be melted or matted together. Seabrook Chemistry Department took chips of the material and perfonned a chemical analysis which identified 349 ppm chloride and 24 ppm L
fluoride. During outages, relief valve RC-V-89 is removed from the system for testing. In the l
process borated water drains from the line wetting adjacent material in the area including the 3 inch spool piece, its covering and the 12 inch suction line and its insulation. Although the present chloride readings for the covering materials are 349 ppm they may have been higher in the earlier stages prior to leaching. It appears that the covering material was on the pipe since original l
l DMFTRedstonS V7m 45 i
e
i i-i cor.struction (10 years or more) and that the removal and testing of RC-V-89 along with subsequent wetting of the area had occi.tred approximately 6 times.
t Original radiographs of the nyo welds on the spool piece were retrieved and examined by NDE Level 111 personnel. The review showed no anomalies in the weld or the 'mse mets). It was also 1
noted that the two welds on the spool piece were performed by different welders in different companies. The flange to pipe weld was performed as a shop weld by Dravo, supplier of the spool piece. The weld-o let to pipe weld was a field weld performed at Seabrook. This indicates' that the welds and welder qualification do not appear to be a contributor to the problem as the radiographs showed no anomalies and the welds were performed and examined by different
+
individuals yet the heat affected zones of both welds had the weepholes. Also, there did not appear to be any anomalies in the base metal of the pipe which further indicates that the weepage was due to in service conditions rather than original installation.
i le Disemmalan:
+
T A document search of the heat number for the spool piece was performed to detctmine where
[
else piping of that heat may be installed in the plam.This was done in anticipation that the i
problem was due to poor piping material. However, this failure mechanism nas been eliminated
{
based on chemical analysis completed at Mass Materials Research Labs which confirmed that the me.terial met the specification for SA312 Type 304 Stainless Steel pipe.
A review of the piping analysis shows that the stresses in the spool piece piping are minimal.
Du.ing RHR operation, the combined stresses have a safety factor of 4.75 on the normal design allowable. During power operation with RHR in standby, the safety factor on normal design allowable is about 28.0. The piping design in this area is dictated by the faulted condition with stresses induced by a Loss Of Coolant Accident In addition, there have been on indications of any mechanical or flow induced vibration in this area. Nor has there been any problem history with relief valve RC V-89 that would have resulted in abnormally high thrust loads. In summarizing, there does not appear to be any high service induced stresses that could initiate a failure mechanism which could be attributed to the weepage. The only area on the piping spool piece that has been exposed to high stress levels is in the heat affected zones of the welds due to residual stresses from welding.
There does not appear to be any service induced mechanism which could be attributed to the i
weepage. However, a combination of the HAZ weld residual stresses, sensitized microstructure in the HAZ, exposure to intermittent temperatures giratar than 200 F, and environmental
' conditions (i.e. chloride contaminants) on the outside of the pipe are indicators that the pipe may
~ be susceptible to Chloride induced Stress Corrosion Cracking (SCC). Although the chloride level
. in the wrapper is not excessive at 349 ppm, this in itself would be sufficient to initiate the stress corrosion process when coupled with the high residual stresses located in the oveld's heat affected zone'. The higher the residual stress, the more sensitive the material is to smaller amounts of chloride. The chlorides would also tend to concentrate at crevices on the curface due to the l
washing of the covering material by the borated _ water and subsequent drying off of this water-when the line goes back up in temperature. This could result in potentially high levels of chloride concentrated in selectivo areas around the heat affected zone of the welds.
nnerneuens mn 4
~~.-..-.m
.m
, _.. _. _. _ ~... - _ _.. _ _ _.. _ _ _
^**
A decision was made to remove the spool piece and perform metallurgical analysis to determine the cause and extent of degradation that led to the weepage. The piece was removed by making a cut at the weld of the wcid-o let to the spool piece and unbolting the inlet flange at RC-V-89.
]
The analysis involved one section of the spool piece being analyzed at Seabrook and the other.
section analyzed at ABB Laboratories in Windsor Locks Connecticut. The piece was cut -
1 approximately 31/2 inches down from the top of the flange. The top piece was analyzed at Seabrook and the bottom was shipped to ABB Labs.
i J
. Prior to sectioning the removed piece, a leakrate test was performed on the piece to determine i
'what effect the degraded section would have had on containment integrity. The test consisted of clamping a gasketed plate over each end of the piece. One end was fitted with a test connection j
which was pressurized to 52 psi. The leakrate was determined to be 1.363 Standard Cubic Feet 4
per Hour (SCFH) which is much less than the allowed value for a typical penetration which is on j
the order of 37.0 SCFH. It was also noted that the leakage appeared to be coming from the a
gasketed end plates. Results of the test indicate that the pipe spool was not degraded to the extent that containment integrity was challenged.
The following de:cribes the work and examinations performed on the section analyzed at Seabrook and the conclusions reached from that work. A Liquid Penetrant (LP) examination was performed at the weld of the flange to the 3" pipe on the inside and outside diameter of the piece.
One linear indication was found that went through the wall. The crack ran at an approximate angle of 45' to the longitudinal axis of the weld. It progressed into the base metal of the 3" pip..
The piece was sectioned at a location perpendicular to the weld. The sectioned surface displayed l
a branching type crack running perpendicular to the surface of the pipe at a distance i
approximately 3/8" from the weld in.the heat affected zone. The indication originated on the e
outer surface and progressed to the inside diameter of the pipe. This testing is documented in Reference 1.
2 The failure appears to be a corrosion assisted degradation process such as chloride induced stress corrosion cracking that had initiated on the outside and migrated to the inside diameter.
l.
However, it was decided that more confirmatory information was required from the analysis being done at ABB on the lower section of the spool piece. Additional needed information included: obtaining chemical analysis of the pipe in the non-heat affected zone, performing an SEM examination on the crack surface, and performing a chemical analysis of any contaminants that may be present on the crack surface.
Some indications were found in the heat affected zone of the weld on the 12" line where it L sttaches the weld +1et. The indications were repaired. However, the indications on the weld-o-let were more extensive and could not be repaired. The weld-o-let was removed and replaced with a new fitting.-
l u.
p a
u N
lBeQTBMagna $ kng,
4y
.I 4
~
--,_.-._..._..m~_..~.-,
.-..m
o,,
k*
Results of the analysis conducted by ABB Laboratories were consistent with those presiously determined at Seabrook. Upon receipt of the spool piece, a LP examination was performed on the outside and inside surface of the pipe. De revealed two through wall linear indications and one non through wall which supports the theory of a failure mechanism originating from the outside.
i-and progressing inward. De methodology and conclusions of the failure investigation are contained in Attachments 2 & 3.
NAESCO provided technical oversight to the ABB analytical work by assigning the Duke i
Engineering & Services Metallurgical Engineer to follow the work at ABB and provide his own i
assessment. His report is contained in attachment 3.
4.0 Safe:y SigalSence-1 1
The spool piece and connecting valve and piping are classified ANS Safety Class 2, seismic category I components. De piping components were designed, manufactured, and tested in accordance with the ASME Boiler and Pressure Vessel Code, Section Ill.
Subsection NC. These components are part of the Residual Heat Removal System which provides for safety related shutdown cooling, and low head injection during ECCS operation.
' The through w.ll weepage resulted in the Unit entering Technical Specification 3.6.1, resulting in a plant shutdown. This evaluation provides the technical assessment of the cause for the through wall weepage, with the purpose of preventing reoccurrence in the future Replacement piping components have been fabricated and installed in accordance with the ASME Code Requirements. Other corrective actions are included in the Event Evaluation for this occurrence.
Camelusions:
De ieports conclude that the cause of the weepage sites was a chloride induced transgranular stress corrosion cracking mechanism initiated from the outside of the pipe working in. The Chlorides were leeched out of the tape /firewtap material affixed to the outside diameter of the pipe. This is not itvpical failure mode for stainless steel in the benign PWR coolant chemistry environment, however the unique wrap material, inadvenutly left on the subject pipe spool at construction, exposed the stainless pipe to a deleterious local environment which, when combined with
- x typical HAZ microstructure, led to the observed through wall weepage.
Attachments:
- 1. Duke Engineering & Services Memorandum, File #RC-V-89 A, dated 12/8/97, from KR Willens to RE White.
~ 2i ABB
_ 3. Duke Engineering & Services Memorandum, File # RC-V-89-B, dated 12/12/97, from KR 3Willens to RE White l-l 7
.. =..
- ~.
h 1
g e*
Attachment E ENGINEERING EVALUATION SS-EV-97-0030 EVALUATION OF SYSTEM SAFETY FUNCTIONS WITH RESPECT TO WEEPAGE AT INLET PIPING TO RC V89, B TRAIN RHR PUMP SUCTION RELIEF VALVE Prepared by: R. J. Faix December 11,1997
Purpose:
During a recent shutdown to address the pipe wall defects in the inlet piping to RC-V89, an engineering evaluation was requested to determine if the RHR system would have been capable of performing its safety function. The RHR pump suction and the containment building spray pump suction from the RWST is a common piping system. The CBS System will also be evaluated. The applicable safety functions are containment integrity, shutdown cooling, emergency core cooling and containment depressurization via containment spray. The purpose of this evaluation is to determine if the RHR System and CBS were capable of performing the above safety functions.
Background:
On December 5,1997, Seabrook Station initiated a reactor shutdown from full power to repair weepage from a through wall defect in the inlet piping to the B train RHF-pump suction relief valve. The defect was identified by the presence of boric acid crystals on the pipe wall and on adjacent pipe insulation. Prior to the reactor shutdown, the pipe insulation was removed, the piping was cleaned and inspected.
Two welds exhibited weepage sites. One weld, the pipe to RHR relief valve inlet weld neck flange, had a single defect site. The second weld, the weld-o-let to pipe weld, exhibited three weepage sites. The system leakage was characterized as a drop every two to three minutes. The system pressure at the welds was the elevation head of the RWST or less than 50 psig.
The relief valve and inlet piping are shown on UFSAR Figure 5.1-3, Sheet 4. The RHR System is shown on UFSAR Figure 5.411. The piping is ANS Safety Class 2, ASME Section III, Class 2, Seismic Category I.
Discussion:
The applicable safety functions for the RHR System are containment integrity via the containment penetration and piping, shutdown cooling and emergency core cooling. The safety function for the CBS system is containment depressurization.
a DRAFTRedsion3 V7m 49 l-
- e.,,
l ?., ;s
,5*
pi Containment lategrity:
~
The RHR System cooldown piping forms a portion of the containment boundary. The containment penetration is described in UFSAR Section 6.2.4. The piping defects and weepage occurred in the containment boundary of the piping system.
The piping defects may challenge the containment integrity since the penetration piping would have a pre existing defect. The design basis of the ECCS is that it is capable of sustaining an i
active or a passive failure in the long term and continue to perform its safety function. With a pre-existing defect in the containment penetration piping and a postulated passive failure in the ECCS piping system, a leakate Path to the environment could occur.
o
- The piping section containing the defects was subjected to a leak rate test after being removed from the piping system. The leak rate test was conducted by pressurizing the pipe ID to 52 psig, the containment design pressure, and measuring the makeup flow required to maintain the pressure at 52 psig. The leak rate was detemiined to be 1.363 standard cubic feet per_ hour (sefh).
- This compares to ar. allowed penetration leak rate of 37 scfh or 0.05 L, as described in the Containment Leakage Rate Testing Program.
Since the tested leakage' through the degraded piping section and the defects was much less than the allowed leakage for containment penetrations, the penetration was capable of performing its containment integrity function.
Shutdows Cooling:
During shutdown cooling the piping is subjected to an operating pressure of 365 psig and operating temperature of 350 'F. The piping design pressure is 600 psig at 400 'F. The piping section is the inlet piping to the RHR pump suction relief valve set at 450 psig. So that realistically, the maximum pressure in the degraded pipe section is 450 psig.
A section thrr a i the defect in the pipe to inlet flange weld identified a small defect in the pipe wall below the weld. The base metal indicated no signs of pipe wall degradation other than the single defect. The pressure minimum wall in the 3 inch pipe run is 0.064 inches. _The piping section was fully capable of carrying all design loads and maintaining the structural integrity of the piping system as demonstrated in Calculation C S-1-45545, RC-V89 Inlet Pipe Flaw l
Evaluation. This calculation evaluated a bounding 0.5 inch diameter flaw that is sufficient to account for the cumulative flaw area and detennined that the calculated stresses were below code allowable stresses and that the flaw was stable in accordance with'GL-90-05 criteria.
Similarly, the defect in the pipe to weld-o let weld identified three small defects in the pipe wall
- below the weld. The base metal indicated no signs of pipe wall degradation other than the three
~
defects. The pressure minimum wall in the 3 inch pipe run is 0.064 inches. The piping section twas fully capable of carrying all design loads and maintaining the structural integrity of the
, piping system as demonstrated in Calculation C-S-1-45545, RC-V89 Inlet Pipe Flaw Evaluation.
~
NLUTaoMens Mse
_$0
., _ _ -- - -. - - -.. - - -.. - -. _ - ~. - -
- =,,
- @ T Since the piping section is capable of maintaining its stmetural integrity, the RHR System is capable of performing its shutdown cooling function.
Esmergency Core Cooling:
During emergency core cooling the piping section forms part of the pretsure boundary of the RHR pump suction. Leakage from the RWST into containment as a result of the piping defects is infinitesimally small to be of any consequence. During ECCS injection the RHR pumps take suction from the RWST and deliver flow to the RCS cold legs. The pressure at the location of the defect is the RWST clevation head as it exists du-ing plant operation. During the recirculation phase of ECCS operation the pressure at the location of the defect is the containment pressure since the RHR pump suction is transferred the containment ECCS recirculation sumps.
He piping section loading is bounded by the structural loads. With the minor defects at the two locations, the piping is fully capable of maintaining its stmetural integrity. Since the piping section is capable of maintaining its structural integrity, the RHR System is capable of performing its emergency core cooling function.
Containment Depressurization (CBS System):
The containment building s pray pumps and the RHR pumps both take suction from the RWST following an accident. The CBS System is separated from the RHR system by two check valves in series. Since '. nature of the weepage sites could potentially affect the operation of the CBS system, the CBS System was conservatively declared inoperable.
The CBS pumps were fully capable of taking suction from the RWST and delivering the required flow to the containment spray headers inside containment. The interaction of the CBS system with the RHR System was of no consequence since the ECCS was fully capable of performing its safety function as described above.
Piping Pressure Boundary Integrity:
The defeets in the RHR suction relief valve inlet piping represent a deviation from the ASME code and, as such, may not conform to one or more applicable codes or standards specified in the UFSAR. The piping does not meet the minimum wall thickness at the location of the defect.
North Atlantic performed a code repair consisting of replacing the pipe sections. This activity necessitated a plant shutdown to complete the repairs.
Coselusion:
l The RHR System was capable of performing the containment integrity, shutdown co. ling and l
emergency core cooling safety functions with the piping section containing the identified defects. The CBS system was fully capable of performing its cafety function of containment depressurization.
1 DA4FTAndslen 5 1/7,94 51 1
-