IR 05000338/1987024

Augmented Insp Team Repts 50-338/87-24 & 50-339/87-24 on 870715-0814.Three Violations & One Unresolved Item Noted. Major Areas Inspected:Licensee Actions Re Steam Generator Tube Failure in 870715

ML20237K261
Person / Time
Site:	North Anna
Issue date:	08/21/1987
From:	Cantrell F NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
To:
Shared Package
ML20237K198	List: IR 05000338/1987024 ML20237K195
References
50-338-87-24, 50-339-87-24, NUDOCS 8709040277
Download: ML20237K261 (45)

Text

_ _____

[pa aeovy' UNITED STATES

!" -

t NUCLEAR REGULATORY COMMISSION I,

j :: E REGION 11 o ~8 101 MARIETTA ST., N.W., SUITE 3100 ATLANTA, GEORGI A 30303

Report Nos.: 50-338/87-24 and 50-339/87-24 Licensee: Virginia Electric & Power Company Richmond, VA 23261 Docket No and 50-339 Facility Name: North Anna 1 and 2 Inspection Conducted: July 15 - August 14, 1987 Team Leader: F. S. Cantrell Team Members: J. L. Caldwell L. P. King W. E. Holland L. E. Nicholson B. K. Revsin S. F. Gagner W. P. Kleinsorge L. B. Engle M. M. Glasman H. F. Conrad -

W. J. Ross C. Dodd, NRC Consultant from Oak Ridge was contracted to help review the condition of the steam generator evaluation and testin Approved y: ,

F. S. Cantre"li d 7-A/ O Date Signed 2B Section Chief, Division of Reactor Projects l SUMMARY The Augmented Inspection Team (AIT) was charged with determining whether the I licensee's actions in response to a Steam Generator (S/G) tube failure on

!

July 15, 1987, were adequate to protect the health and safety of the public and that appropriate action was being initiated to determine the cause of the event and to implement corrective action as necessary to minimize the potential for a recurrence of the S/G tube leak prior to restart of the unit. An additional objective was to determine the adequacy of symptom based emergency operating procedures mandated by the NRC and developed by the Westinghouse Owners Group to cope with serious events of this typ PDR ADOCK 0500

i

\

This inspection involved the following areas: Overview of the Event (paragraph 4), NRC Activities (paragraph 5), Sequence of Events (paragraph 6), 1 Unit 1 Steam Generator. Inspection and Repair History (paragraph 7), Inservice Inspection of Unit 1 S/G Following the Event (paragraph 8), Metallurgical

,

Aspects (paragraph 9), Chemistry Aspects (paragraph 10), Radiological Aspects

'

(paragraph 11), Licensee Response to the Event (paragraph 12), Consideration of Shutdown of Unit 2 (paragraph 13), Investigation and Corrective Action (paragraph 14), and Consideration for Restart (paragraph 15).

Results: The overall results achieved were outstanding in that the operator tripped the plant, isolated the leak and brought the plant to cold shutdown in I

seven hours without using the S/G power operated relief valves. This contributed to a negligible release to the environment. Three violations and one unresolved item were identifie t I

i l

l l

.

i REPORT DETAILS Licensee Employees Contacted

• L. Stewart, Vice President - Nuclear Operations J. L. Wilson, Manager - Nuclear Operations Support

• J. W. Ogren, Director, Operations and Maintenance Support l R. J. Harwick, Manager - Quality Assurance

• N. L. Clark, Manager - Nuclear Programs and Licensing

• E. W. Harrell, Station Manager R. C. Driscoll, Quality Control (QC) Manager l

• G. E. Kane, Assistant Station Manager

• M. L. Bowling, Assistant Station Manager

• R. O. Enfinger, Superintendent, Operations M. R. Kansler, Superintendent, Maintenance [

• A. H. Stafford, Superintendent, Health Physics

• J. A. Stall, Superintendent, Technical Services '

J. R. Hayes, Operations Coordinator D. A. Heacock, Engineering Supervisor G. G. Harkness, Licensing Coordinator

• • J. N. Esposito, Westinghouse (W) Service Technology Division - Pittsburgh (STDP)

• • Kuchirka, W STDP

• • E. Sessions W STDP

• • W. Hirst, W Nuclear Technology Services, Division Pittsburgh, (NTSDP)

• • E. Gold, W STDP

• • J. Allen, W Nuclear, Components Division (NCD) Tampa

• • R. C. Bilyeu, VEPCO Licensing

• • L. F. Becker, Jr. , W STD Other licensee employees contacted include technicians, operators, mechanics, and Westinghouse personnel. The licensee also used an independent consultant, NPR Ass 3ciates, to maintain an overview of the recovery effor i

• Attended exit interview j

• • Persons contacted at Forest Hills Site, Pittsburgh, July 29-30, 1987. Exit Interview (30703)

Interim inspection findings were discussed with the Plant Manager and other licensee representatives July 18, July 24, and July 30, 198 l The inspection scope and findings were summarized on August 14, 1987, with those persons indicated in paragraph I above. Violations were identified involving failure to have approved safety evaluations for operating with j leaking explosive plugs in the steam generator and foreign objects in the  !

Steam Generator (VIO 338/87-24-01) (parcgraph 7); failure s

i

. _ _ _ _ - - _ _ _ _ - _ - . _ _ ___-

- _ _ - _ _ _ _ _

to promptly notify HP when a radiation alarm was received on the main steam line monitor (VIO 338/87-24-06) (paragraph 12); and failure to perform adequate operability channel checks on radiation monitors as required by Technical Specification (VIO 338/87-24-03) (paragraph 11 &

12). One unresolved item involving verbatim compliance with emergency i procedures was identified (URI 338/87-24-04) (paragraph 12). Other concerns identified involved the cause of the tube rupture, plans to examine other tubes and to reduce potential for recurrence; lessons to be learned from the way this event was handled and a human factors evaluation to determine any additional changes needed to improve performance; advisability of calling day supervision while .i the urgent evaluation steps of the emergency procedures; failure of the Steam Generator blow down monitor to respond to the tube rupture; adequacy of cautions in alarm and abnormal procedures when. equipment becomes inoperable; frequency of calibrating and resetting radiation monitor; and potential radiation levels in area that require manual manipulations by operator In addition the following Inspector Followup Items (IFI) were identified:

-

IFI 338/87-24-02, Evaluation of the loose parts monitor alarm by the licensee (paragraph 7).

-

IFI 338/87-24-05, Review of dif ferences in Emergency Plan-3 Rev. O, and Rev. 1 (paragraph 12).

-

IFI 338/87-24-07, Licensee reevaluation of the adequacy of the design and setpoints associated with steam jet air ejector isolation on high radiation levels (paragraph 12).

-

IFI 338/87-24-08, Securing of low head safety injection early due to pump concerns (paragraph 12).

-

IFI 338/87-24-09, Hydrogen gas control in the steam generator following a tube rupture including procedural changes (paragraph 10).

The licensee acknowledged the inspectors findings. The licensee did identify as proprietary some of the material provided to and reviewed by the inspectors during this inspection; however, this material is not discussed in this report.

3. Unresolved Items *

One unresolved item was identified involving verbatim compliance with Emergency Procedures (URI 338/87-24-04) (paragraph 12).

An Unresolved Item is a matter about which more information is required to determine whether it is acceptable or may involve a violation or deviatio [:

4. Overview of Event At approximately 6:35 a.m., on July 15, 1987, North Anna Unit I was manually tripped from 100?s power due to indications of a steam generator tube rupture. Approximately 20 seconds later, the safety injection system automatically initiated. All safety systems performed as designe A " Notification of Unusual Event" was initially declared; however, this

_

condition was upgraded to an " Alert" in accordance with emergency procedures at the direction of the Interim Station Emergency Manager. All notifications to NRC and offsite agencies were made in a timely manner and the unit was stabilized in accordance with emergency procedures. Station personnel accountability was accomplished and station emergency facilities (Technical Support Center and Local Emergency Operations Facility) were manned and activated. After stabilization, the unit was taken to cold shutdown and arrived at this condition at 1:33 p.m. Plant conditions were stable and the emergency was terminated at 1:36 p.m., July 15, 198 During the course of the event, effluents were released from the faulted )

("C") steam generator prior to its isolation via the condenser air ejector and turbine auxiliary feedwater pump exhaust; however, the release was less that 1?6 of Technical Specification limits. This low release level was partly due to the fact that the unit had just returned to full power operation on the previous day from a two month refueling outag During that outage, all fuel assemblies that were to be reused were leak tested and defective fuel was not reused for the reloa )

5. NRC Activities NRC Headquarters was notified by North Anna personnel at approximately 6:51 a.m., that a steam generator tube rupture event had occurred on Unit 1. Region II was notified of the event by HQ and immediately proceeded to staff the Region II emergency response cente The NRC l resident inspector (RI) who was on site responded to the control room l prior to the manual reactor trip and monitored the licensee's actions ;

during the event. An open line between the North Anna site and the NRC l operations center was maintained until the event was terminated early that afternoon. About 7:45 a.m., a decision was made by regional management to send an inspection team to the site. This initial team, headed by the North Anna region based section chief, consisted of the North Anna and Surry se r.i or resident and resident inspectors, a region radiation specialist, and a headquarters public affairs perso This team was augmented by other specialist in eddy current testing, metallurgy, and water chemistry from NRR and Region I The ' team arrived on site approximately 2:00 p.m., on July 15, 1987. The team met with licensee management at approximately 3:00 p.m. , to discuss the status of the unit and to outline Licensee /NRC interface for recovery action Virginia Power of fered to supply any assistance required by the tea During the j

__ _ _ _ _ _ _ _ _ _ - _

next day's review of the event, the team was upgraded to an Augmented Inspection Team (AIT). The team conducted inspections for the remainder of the week ending July 17, 1987, to ascertain the circumstances involved in the event. The AIT did not conclude its inspection at that time due to the ongoing activities by the licensee to develop a root cause analysis, which required subsequent inspection activitie AIT inspections continued during the weeks of July 20 and July 27, 1987. Region II issued a confirmation of action letter to Virginia Power on July 22, 1987, requiring NRC concurrence prior to restart of the Unit. The licensee's

'

report entitled North Anna Unit 1, July 15,1987, Steam Generator Tube Rupture Event Report, Revision 0, dated July 29, 1987 was reviewed, however, the root cause of the event had not been determined when this report was issue An AIT exit with VEPCO management was held on August 14, 198 . Sequence of Events Initial Plant Conditions Prior to the Unit 1 Event Unit 1 achieved 100% power from the 1987 refueling outage July 14, 1987. Unit 2 was operating at 82% power with a refueling outage scheduled to begin July 31, 198 (1) On June 17, 1987, while Unit 1 was heating up in preparation to enter Mode 4 the "A" reactor coolant pump tripped due to an electrical ground in the moto The five year inspection on this motor was completed during the outag (2) On June 21, 1987, Unit 1 experienced a vacuum on the reactor coolant system, while replacing the 1A Reactor Coolant Pump motor and a loss of 17,000 gallons of reactor coolant system inventory during the cooldown and depressurization process (Inspection Report 338,339/87-21).

(3) On June 29, 1987, Unit I returned to power operation but tripped from approximately 18 percent power due to a high level fifth point feedwater heater caused by the divert valve being improperly isolate (4) On June 30, Unit I was returned to 16 percent powe (5) On July 1, at 24 percent power, the manual isolation valve upstream of the "C" feed regulating valve would not open. The stem was broken attempting to open i The plant was shutdown to mode 4 (205 F) to repair the valve. The plant was restarted on July (6) On July 7, the unit was ramped up to 80% but was ramped down to 50% in order to maintain level in the moisture separator reheater (MSR). It was discovered that IB MSR stop valve was only partially ope l

!

(7) On July 10, the unit was shutdown to repair the 18 MSR stop valve. The valve was replaced on July 11, and'the unit was made critical on July 1 !

(8) On July 13, 1987, the steam jet air ejector radiation monitor {

failed low. An action statement was entered in accordance with 1 Tech. Spec. 3.3.3.11 and a deviation report was submitted,- A work request was written; however, the lines were drained and

the condenser air ejector monitor was returned to servic I Subsequently, the work request was cancelled. On July 14, 1987, the monitor failed low and was declared inoperable again. Work :

request 453809 was issued and the monitor remained in an inoperable status until after the even (See paragraph 12.) ]

b. Plant Conditions and Personnel Actions During the Event The sequence of events detailed in this report was reconstructed from the printouts of the alarm typewriter attached to the control room P-250_ process computer, the Sequence of Events Recorder (SER) driven by the Hathaway annunciator system, the data printouts extracted from the Emergency Response Facility Computer (ERFC) in the Technical ;

Support Center, and the strip charts obtained from control room recorder The licensee supplemented this information with interviews with the shift supervisor and control room operators (CRO)

involved in the event. The inspectors reviewed the above data and ,

compared the chronology with the notes taken by the resident inspector during the even A small discrepancy in the indicated thaes was noted between the j P-250, SER and ERFC. An analysis of the data revealed that the P-250 l led, the SER was within 20 seconds of the P-250 and the ERFC was about one minute behind the P-250. For the purpose of this chronology, times will be annotated based on the P-250 response. The radiation monitor alarm printer times were approximately one hour behind the P-250 time The radiation monitor printer indicated a time of 05:34:37 for the initial alar July 15, 1987 Steam Line Time From Rad Radiation Monitors Monitor Alarm Normalized to (Read. X .1 = MR/HR)

Printer 6:30 A B C 04:57 1.1 .4 05:34:09 1.1 .0 05:34:37 06:30 8.2 :34:40 06:30:03 Alarm Reset 05:35:59 06:30:22 9.0 :36:24 06:30:47 Alarm Reset 05:36:35 06:30:58 7.9 _ _ _ _ _ - _ _ _ _ _ _ _ _ _ -

\

6 -

0631 Unit 1 CR0 observes the pressurizer level decreasing rapidl Initial level decline appears to be at a rate of approximately 1% every 40 second :37:25 06:31:52 Alarm Reset 05:37:25 06:31:52 8.2 Unit 2 SRO paged the shift supervisor and the Unit 1 SR0 to return to the control roo .

Unit 1 CR0 took manual control and increased charging rat Pressurizer pressure had dropped to appx. 2135 psig, and pressurizer level to less than 59%. - -

05:37:58 06:32:43 Alarm Reset 05:38:07 06:32:52 7.0 :42:09 06:36:54 Alarm Reset 0633 The shift supervisor entered the control room and directed '

that letdown be isolated, and took direct control of directing activities on Unit Charging pump suction was manually realigned to the refueling water storage tank (RWST) and a 2% per minute turbine ramp down was initiate A Volume Control Tank (VCT) low level annunciator alarmed with an indicated level of 20.3%.

0634 Shift Technical Advisor (STA) and NRC Resident Inspector '2_3 arrived in the control room and began monitoring ,-

activitie "'

Pressurizer pressure and level continue to decrease with pressure approx. 2109 psig and level approx. 49%. S/G "C" level begins slight increase and feedwater flow to that S/G compensates by decreasing. Superintendent of Operations was notified and directed the unit to be manually trippe :35:24 CR0 manually tripped the reactor and turbine and initiated Emergency Procedure 1-EP-0 " Reactor Trip or Safety Injection". CR0 noted pressurizer level at approx. 45% and e pressurizer pressure at approx. 2100 psi :35:41 Automatic initiation of safety injection (SI) on low pressurizer pressure (1765 psig).

0636 Main feedwater pumps tripped and auxiliary feedwater pumps runnin S/G 1evels at or near 0% (narrow range) on all S/G The CR0 noted that pressurizer pressure was less "

than 1700 psig and pressurizer level less than 5%. The -

a'larm typewriter noted that a second charging pump started and both low head safety injection pumps runnin .

- - _ _ _ _ . . _ _ _ _ . . - - - - -

7 I

0637 Main generator output breakers (G-12) opens. Pressurizer level drop appears to have been turned around from a level of approximately 2.7?;. The level gradually increases and j levels off at approximately 5? '

0639 A Notification of Unusual Event was declared. The Unit 2 1 SR0 was assigned the duties as the Interim Station Emergency Manager and initiated the Emergency Plan Implementing Procedure (EPIP). i The "C" S/G 1evel begins to recover with "A" and "B" S/Gs-following.

0640 Entered Emergency Procedure 1-EP-3 " Steam Generator Tube Rupture" (per operator interviews)

0641 The alarm typewriter notes that Tavg is less than 543 l degrees '

Interlock P-12 set prohibiting steam dump actuation.

0642 The pressurizer level is holding relatively steady at i approximately 5?; with the S/G 1evels trending upwar l 0644 Operators reset safety injection and phase A signal. Low head safety injection pumps "A" and "B" secured.

0646 Auxiliary feedwater to "C" S/G isolate The shift supervisor determined at this point that the tube rupture was in the "C" S/G based on the continual rise in level for that generato The "C" main steam trip valve was closed.

0647 The alarm typewriter noted that "A" S/G level was at 23?4 (narrow range) and increasing. The' level of "C" S/G was noted to be 25?4 (narrow range) and increasing approximately 2 minutes earlie This was indication that the tube rupture was in the C steam generater since the "A" S/G is fed by the tt -bine driven auxiliary feedpump that delivers significantly more flow than being delivered to the remaining two S/Gs by the motor driven auxiliary feedwater pump The steam supply from "C" S/G to the turbine driven auxiliary feedwater pump 1-FW-P-2 was isolate Steam dump valves were activated and operator commenced'a rapid cooldown using the "A" and "B" steam dump valves to the condenser. The alarm typewriter noted the "B" S/G level at 25?s (narrow range) and increasing and the

!

_ __ _

8 l 1'

pressurizer level at 8.5?; and decreasing. Reactor coolant i

system (RCS) wide range temperature was at approximately 540 degrees F. at the start of cooldown.

l 0650 Unit 1 CR0 noted pressurizer level off scale lo Initial notification of the Unusual Event made to the state / local governments and the NRC via the emergency l notification telephon l l

0652 RCS cooldown in progress with cold leg temperatures l recorded as 509.5 degrees F. on "A" and "B" loops and 52 degrees F. on "C" loo l Interim Station Emergency Manager upgraded the event

'

0654 classification to " Alert".

0655 Initiated station callout, accountability and access control in accordance with the appropriate Emergency Plan j Implementing Procedures (EPIP). 1 0657 RCS temperature is being maintained at approximately 480 degrees The pressurizer spray controller was set at 100?; demand and spray valves verified open to bring j pressurizer level back to scal The CR0 noticed that pressurizer level was back on scale trending up. RCS pressure is approximately 1,000 psig and trending down from the spray valve operatio The source range nuclear instruments were required to be manually re-energize The pressurizer low level heater cut off interlock cleared with pressurizer level at 1 574 and increasin The pressurizer heaters are energize '

0701 The CR0 manually de-energizes the pressurizer heaters with pressurizer level approximately 48?6 and increasing. RCS pressure at this point has leveled at approximately 900 i psi The notifications were made to the state / local governments and the NRC of the'uograded " Alert" classificatio ,

I 0704 The CR0 briefly opened one pressurizer PORV to reduce RCS j pressure below the affected S/G pressure. The SRO observed a primary pressure reduction of approxit.;ately 40 psig and instructed the CR0 to close the PORV and spray valves. The alarm typewriter noted a pressurizer relief tank pressure j l of 15 psig indicating operation of the PORV.

l

,

l

pressurizer level at 8.5% and decreasing. Reactor coolant j system (RCS) wide range temperature- was at approximately i 540 degrees F. at the' start of cooldow Unit 1 CR0 noted pressurizer level off scale lo Initial notification of the Unusual Event made to the l l

l state / local governments and the NRC via the emergency notification telephon RCS cooldown in progress with cold leg temperatures recorded as 509.5 degrees F. on "A" and "B" loops and 52 degrees F. on "C" loo Interim Station Emergency Manager upgraded the event classification to " Alert".

0655 Initiated station callout, accountability and access ,

control in accordance with the appropriate Emergency Plan j Impleme.' ting Procedures (EPIP). l 0657 RCS temperature is being maintained at approximately 480 degrees The pressurizer spray controller was set at 100% demand and spray valves verified open to bring pressurizer level back to scal The CR0 noticed that pressurizer level was back on scale trending up. RCS pressure is approximately 1,000 psig and !

trending down from the spray valve operatio The source range nuclear instruments were required to be manually re-energize l 0700 The pressurizer low level heater cut off interlock cleared with pressurizer level at 15% and increasin The pressurizer heaters are energize The CR0 manually de-energizes the pressurizer heaters with pressurizer level approximately 48% and increasin RCS pressure at thi s point has leveled at approximately 900 psi The notifications were made to the state / local governments and the NRC of the upgraded " Alert" classificatio The CR0 briefly opened one pressurizer PORV to reduce RCS pressure below the affected S/G pressure. The SR0 observed a primary pressure reduction of approximately 40 psig and instructed the CR0 to close the PORV and spray valves. The alarm typewriter noted a pressurizer relief tank pressure of 15 psig indicating operation of the POR l l

9 l I

The SR0 not.ed the "C" S/G 1evel increase had stoppe l The safety injection reauction criteria was determined to be satisfied and the "B" charging pump was secure Initiated the isolation of the boron injection -tank (BIT)

flowpath and established the normal charging flowpath.

0706 The "A" and "B" pressurizer spray valves are closed. The pressurizer level was at approximately 8096 and increasing.

0709 The normal letdown path has been re-established as indicated by a 47 gpm outlet flow from the non regenerative heat exchanger.

0710 The pressurizer heaters were energized based on a decreasing level in the ruptured S/ The CR0 closed the RWST suction valves.

0711 The pressurizer level peaks at approximately 88?s.

0713 CR0 secures the "B" and "C" reactor coolant pumps.

0714 The alarm typewriter notes the spray demand at 76?s The RCS pressure is now being maintained by manual control of spray and heaters.

0715 The Superintendent of Operations and the SRO on call arrived in the control room.

0718 Operators transition into the Emergency Procedure, ES " Post-Steam Generator Tube Rupture Cooldown using Backfill".

0720 The Station Manager arrived in the control room.

0721 The motor driven auxiliary feedwater ( AFW) pump 1-FW-P-3A is secured. The AFW flow had been secured to "C" S/G at 0646.

0722 The alarm typewriter notes that pressurizer level is at 73?;

and decreasing.

0723 The motor driven AFW pump 1-FW-P-3B to the "B" S/G was secured. The AFW pumps are subsequently run intermittently to support S/G feed requirement The NRC Region II Incident Response Center was activated and placed in standb i 10 ,

I 0725 The CR0 started the "B" condensate pump. Both "A" and "B" condensate pumps were then runnin l 0726 The NRC Headquarters Response Center was activated and placed in standb The CR0 begins a RCS cooldown in accordance with EP l This method depressurizes the ruptured S/G by draining i through the ruptured tube into the RC j 0730 The emergency diesel generators are secured as documented

in the appropriate operating procedure.

l l The Assistant Station Manager arrives in the Control Room and initiates transition of the EPIPs and communications from the Control Room to the Technical Support Center (TSC). l 0739 The Station Manager assumes the position of Station ,

Emergency Manage I 0745 The unit turbine is placed on the turning gea The NRC response site team is formed and dispatched from Region I The fuses were removed from the steam jet air ejector (SJAE) radiation monitor to fail the instrument high and  :

I allow the discharge to be diverted to containmen l The discharge was then manually diverted to containmen The site TSC was activate The CR0 secured the "B" condensate pum The licensee's Corporate Emergency Response Center was activate The CR0 noticed a brief alarm or the Loose Parts Monitoring System for "C" S/ The CR0 started to use auxiliary spray to supplement the RCS depressurizatio The Local Emergency Offsite Facility was activate The containment partial pressure exceeded the allowable setpoint due to the condenser air ejector exhaust diversion to the containment.

l l

l l

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

i l

1049 The pressurizer power operated over pressure relief valve key switches were placed in " auto" for ocold over pressure protectio Entered Mode 4 (less than 350 degrees F.)

l 1153 The CR0 cycled the reactor trip breakers to re-enable automatic safety injectio The CR0 placed "A" and "B" charging pumps and the "B" low head safety injection pump in " pull-to-lock".

1219 The RHR system is placed in service to continue RCS cooldow The "A" reactor coolant pump is secure The main steam system is secure l 1312 The SJAE exhaust was restored to normal alignmen Entered Mode 5 (cold shutdown).

l 1335 The Station Emergency Manager terminated the emergenc l

1 1326 The licensee notified the NRC, state and local governments of the termination of the emergenc j The licensee implemented a recovery organization (Attachment 1). j 7. Unit 1 Steam Generator Inspection and Repair History The following is a history of the inspection and repair of the Unit 1 steam generators that was provided by the licensee after the even Unit Start-up in 1978

-

1979 Refueling Outage Tubes Inspected Leakage Tubes Plugged S/G A - 440 None S/G A - 94 S/G B - 133 None S/G B - 94 S/G C - 480 2 leaks (Row 1) S/G C - 96

.

__- _ _ _ _ - _

Comments: Resin intrusion during cycle. Row 1 tubes preventively plugged based on vendor recommendation. 2 other tubes plugged due to denting. Denting first observed. (Denting is the formation of magnetite between the outside diameter of the S/G tube and the tube suppurt plate resulting in a measurable restriction at that point.)

Boric acid treatment initiated based on vendor recommendatio Leakage rate barely detectable.

-

1980 Refueling Outage Tubes Inspected Leakage Tubes Plugged S/G A - 485 None None S/G B - 476 None None S/G C - 478 None None Comments: None

-

1982 Refueling Outage Tubes Inspected Leakage Tubes Plugged S/G A - 107 None None S/G B - 1165 None None S/G C - 243 None None Comments: Partial tube end repair due to split pin damage in S/Gs A and C.

-

January 1984 Forced Outage (Leakage Rate 396 GPD)

Tubes Inspected Leakage Tubes plugged S/G B - 579 3 leaks in S/G B S/G B - 4 S/G C - 552 2 leaks in S/G C S/G C - 5 Comments: No progression in tube denting observed. Row 1 leaking explosive plugs repaire Partial tube end repair performe Distorted indications at support plates first notice Distorted indications are signals from the eddy current inspection that cannot be categorized as a defect or lack of a defec i I

.l l

l

-

May 1984 Refueling Outage Tubes Inspected Leakagee Tubes Plugged S/G A-100?; unplugged None S/G A - 10 tubes S/G B-100?; unplugged S/G B - 1 tubes S/G C-100?; unplugged S/G C - 5 1 items Comments: Profilometry on selected tubes in all 3 S/G Partial tube end repair performe Attempted tube removal. Distorted indications observed. Foreign object located and removed in S/G j 2 tubes plugged preventively. One hundred percent inspection means i all of the tubes were inspected partially but all cold legs are no l inspecte August 1985 Forced Outage (Leakage Rate 213 GPD)

Tubes inspected Leakage Tubes Plugged S/G A - 830 3 leaking tubes S/G A - 13 .

Comments: Distorted indications observe ' November 1985 Refueling Outage (Leakage Rate 90 GPD)

Tubes Inspected Leakage Tubes Plugged S/G A-100?; unplugged None S/G A - 9 tubes S/G B-100?; unplugged 2 leaks in B S/G B - 17 tubes S/G C-100?; unplugged 4 leaks in C S/G C - 47 tubes

Comments

Two tubes removed with 4 support plate intersections. 30 l

tubes from the three steam generators were plugged due to " strong" distorted indications. Sample specialized NDE applied in S/G Other Events During 1986 - March 1987

-

Extensive examination of tubing and materials with EPRI and Westinghous I I

I

I l

!

-

Preparation and. submission of WCAP (a Westinghouse Report) to l NR Meeting with NRC staff in March 198 Developed eddy current rule base for April 1987 Refuelin )

-

1987 Refueling Outage (Leakage Rate - 11.5 GPD B; 14.6 GPD C)

Tubes Inspected Leakage Tubes Plugged i

S/G A-100?6 unplugged None S/G A - 83 l tubes  !

S/G B-100?; unplugged 2 tubes in B S/G B - 62 ;-

tubes j S/G C-100?; unplugged 4 tubes in C S/G C - 118 i tubes j Comments: Extensive additional NDE performed included: f

-

Profilometry of more than 100 tubes in each S/ x1 probing of nearly 100?s of available tubes. (Hot Leg)

i

-

Rotating pancake probing of all identified tubesheet ,

indications and a sample of support plate intersection l Tube end repair complete U-bend stress relief performed on all .;

available Row 2 tubes in all 3 steam generator Support plate !

stress relief demonstration performed in S/G B. Two tubes removed i from S/G A containing 2 tubesheet indications and one support plate f intersectio I

.

i

!

!

>

l

i

1 I i

)

-

- .

--____

NORTH ANNA UNIT'l TUBE PLUGGING SUMMARY OUTAGE DATE STEAM GENERATOR TOTAL TUBES A B C SEPTEMBER '79 94 94 96 284 JANUARY '84 0 4 5 9 i

!

l MAY '84 10 1 5 16 AUGUST '85 13 0 0 13 l

l NOVEMBER '85 9 17 47' 73 APRIL '87 83 62 118 263 TOTAL 209 178 271 658 j (6. 2%) (5. 3*') (8.0%) (6.5%) 1 Af ter reviewing historical data, the inspector was provided a S/G safety l evaluation performed by Westinghouse for Unit 1 dated June 30, 1987. In l that report, two abnormal conditions were evaluated:

l

-

SECL-87-245 - Steam Generator Secondary Side Foreign Objects

-

SECL-897-230 - Plant Operation with Leaking Explosive Steam Generator 1 Tube Plug i The evaluation concluded that operation of North Anna Unit 1 during ;

Cycle 7 with foreign objects does not constitute an unreviewed safety question. That evaluation also concluded that operation of North Anna Unit 1 for one complete heatup and cooldown cycle with leaking explosive plugs af ter the 1987 outage does not constitute an unreviewed safety question. Af ter reviewing the above evaluation, the inspector asked the licensee if a safety committee review had been conducted on the two abnormal conditions prior to restart of the uni The answer was that a safety committee (SNSOC) review had not been conducted prior to restar Technical Specification 6.5.1.6.c requires that the SNSOC shall review all proposed changes or modifications to plant systems or equipment that affect nuclear safet Technical Specification 6.5.1.7.b requires the SNSOC to render determinations in writing with regard to whether or not the above item constitutes an unreviewed safety question. As of July .18, 1987, the SNSOC had not reviewed the abnormal. steam generator conditions as required by Technical Specifications. Failure to conduct an evaluation to determine if a condition constitutes an unreviewed safety question is identified as a violation (338/87-24-01) for Unit 1 only.

l

16 Loose Parts Accountab.ility At 0900 hours0.0104 days <br />0.25 hours <br />0.00149 weeks <br />3.4245e-4 months <br /> (2-1/2 hours into the subject event) a 'ioose parts alarm was indicated in S/G "C". This alarm is thought to be associated with possible debri s from the R9C51 tube ruptur Subsequent to the event, the AIT team requested the licensee to reverify the NA-1 procedures and control records for loose parts accountability for the 1987 refueling outag The licensee's procedures included SSS2.4.2VP-3 (Fiberoptic Examination and Removal of Objects From Steam Generator Secondary Side), SSS2.2.2.VP-3 l (Post-Activity Sign-Off for Area Cleanliness Secondary Side Activities), SSS2.4.2.VP-4 (J-Nozzle Thickness Inspection)

SSS2.2.2.VP-4 (Steam Generator Tubesheet Cleaning Full Recirculation System), SSS2.7.2.VP-5 (Removal and Reinstallation of Handhole Covers and Bottled Tubelane Blocking Device Sleeves and Split Plates),

SP2.4.2.VP-5 (Flow Slot. Photography) and SSS2.4.2.VP-2 (Inspection of S/G Steam Drum, Moisture Separators, Feedwater Ring, Support Brackets, Welds and Other Components). Based on the licensee's verification of tho above procedures, the licensee concluded loose parts could be ruled out as a possible precursor to the subject tube ruptur In addition, the above procedures will be followed to verify loose parts accountability after the current S/G "A", "B" and

"C" inspection and prior to restart for NA-1. The evaluation of the loose parts monitor alarm by the licensee is identified as an Inspector Followup Item (338/87-24-02).

8. Inservice Inspection of the Unit 1 S/G Following the Event Eddy Current Testing (ECT)

l NRC guidance on eddy current testing of steam generator tubes is

'

conta'ned within Regulatory Guide 1.83 " Inservice Inspection of PWR Steen Generator Tubes". The guidance is general in nature, for example paragraph C.2.a. of Regulatory Guide 1.83 states:

" Inservice inspection should include nondestructive examination by eddy current testing or equivalent techniques. The equipment should be capable of locating and identifying stress corrosion cracks and tube wall thinning by chemical usage, mechanical damage, or other causes."

l

"The inspection

'

The only guidance for overall sensitivity is C. equipment should be sensitive enough to detect imperfections 20% or more through wall."

l In practice, the NRC has endorsed the American Society of Mechanical j Engineers Boiler and Pressure Vessel (ASME) Code as the minimum ,

'

requirements for performing eddy current testin Accuracy and .l precision requirements are specified in the ASME Code,Section V, '

Article 8 - Appendix 1 (Summer 1983 edition). These requirements incorporate a calibration standard consisting of a tube with round,

- - - - - - - _ _ _ - - - _ _ .

i

flat bottom holes of different depths. The ASME calibration standard is designed to verify consistent eddy current system response. It does not establish optimum working sensitivity in all cases nor does it provide for the best estimation of the ' depth of tube wall discontinuities under all circumstances. Use of the ASME standard in estimating flaw depth may result in conservative or nonconservative estimates depending on characteristics of the particular flaw. For this reason, paragraph C.2.e of Regulatory Guide 1.83 reads as follows:

" Standards consisting of similar as-manufactured steam generator tubing with known imperfections should be used to establish sensitivity and to calibrate the equipment. Where practical, these l

standards should include reference flaws that simulate the length, depth, and shape of actual imperfections that are characteristic of past experience."

The reliability of ECT to detect and size defects is a ' strong function of the geometry, orientation, and volume of the defect; the defect's location relative to extraneous sources of noise (affecting signal to noise ratio); the type of eddy current equipment employed  ;

i (e.g., data analysis equipment, test probes, etc.); and the training, l

'

experience and alertness of the data evaluator Reliable detection j

'

and sizing of defects cannot be assured for all defect types for all possible test conditions and equipment. New damage mechanisms may be encountered during inservice inspection which are not reliably I i detected or sized because of non-optimal inspection procedures for l the specific flaw type, location, and orientatio In these situations, alternative inspection procedures and equipment will need to be employed to ensure a reliable inspectio It is important to note that adherence to minimum requirements as specified in the ASME Code does not ensure that ECT inspection programs will reliably detect and accurately size many of the kinds l of defects occurring today. The NRC staff relies on the utility's awareness of the limitations of various eddy current test techniques I and being alert for signs that alternative, more sensitive techniques should be employed. Apart from safety considerations, utilities have a strong economic incentive to employ ECT techniques as appropriate to ensure the reliable detection and sizing of flaw Multi-frequency ECT techniques (not required by NRC) are in universal use by virtually all U.S. utilities. Utilities are utilizing state-of-the-art digital data acquisition and analysis systems (e.g. , Zetec's MIZ-18 system)

which provide a significant increase in dynamic range and signal to l noise ratio over analog system Specialized eddy current probes l such as various types of surface riding pancake probes are in frequent use when dealing with the detection and sizing of ,

intergranular stress corrosion cracking (IGSCC) and intergranular

'

attack (IGA) flaws.

!

l j

_ _ _ - _ _ _ _ _

l l

As stated above, the. difficulties of eddy current testing are such that an eddy current test cannot be designed for unknown types of degradation, that is., the testing system, probes and instrumentation must be chosen on the basis of prior experience in the sam > or sister units. For example, bobbin differential coils are used to detect localized discontinuities with abrupt boundaries such as axial cracks, pits, and vibration damage. Circumferential discontinuities are not usually adequately detected by bobbin coils and, therefore a

" pancake array" probe must be used. Alth ugh very slow in operation, the rotating pancake coil can be used to detect axial or circumferential discontinuities occurring at dimensional or geometric 3 changes in tubes such as dented or expanded regior s. The following

'

describes the probes used in the eddy-current inspection Probe Types Used For Steam Generator Tests Differential Bobbin Probe: Coils are coaxial with the tube, about 0.050 in. long and about 0.0E0 in. apart. They are usually 0.720 i in diameter and are operated in an absolute and differential bridge mod With the MIZ18 eddy-current instrument, they are driven at four multiplexed frequencies (10, 200, 400, and 600 kHz). Tha-eddy-current pattern in the tube is also coaxial to the tube, and any tube property that interrupts or changes the flow of eddy currents will cause a change in the coil impedance. These tube property variations include tubesheets, tube supports, dents, magnetite on the tube or in the crevica, defects in the tube, and intergranular attac Only the axial component of defects will interrupt the circumferential flow of eddy currents produced by the bobbin coil so that circumferential defects, with very little axial component, produce very low amplitude signals. These signals can be easily lost among signals from other property variation Pancake Array Probe (8x1): This probe consists of eight independent pancake coils operated in an absolute mode, being driven at 200 and 400 kHz. These probes ara typically 3/16 in. in diameter, 0.010 i long, and contoured to fit the curvature of the tub The eight coils are arranged in two rings of four coils each, and overlapped in a manner such that every point on the tube passes under at least one coi Each coil is individually spring loaded against the tube to minimize distance between the coil and tube wall, or " lift-off". The eddy-current flow pattern from these coils is circular, around the coil axis, and a crack of any orientation will interrupt tha main flow of eddy currents. The coil is smaller than the bobbin coil and has a more concentrated field, so a small defect causes a larger change in signa The coil is, however, more sensitive to the variations in coil-to-conductor spacing or lift-off than the larger bobbin coil. While the spring loading against the tube wall helps, irregular and sharp dents will give a substantial lift-off signa Since information at only two frequencies are recorded (200 and 400 kHz), this coil type does not have as much data available as the

.

bcbbin or rotating pancake ; oil.

. . . . . . . . . _ . . . . . . . _ . . - _ . . . . . . . __ . . . . . . . . . . . . _ _ . . . . . . . . . . . . . . . . . . . . . _ _ . . . _

_

,

Rotating Pancak, Coil (RPC): This probe is similar to the individual 8x1 coils, but is smaller (typically 0.125 in.). It has a still smaller focus, which gives better resolution of small defects, sees less of the tube outer diameter artifacts, and is more sensitive to lift-of The probe head, containing the coil, is rotated and the coil is sprung against the tube wal Data are recorded at three frequencies (at least), and a very fine and time-consuming scan is made of a " suspected area" of a tube. The spring loading and size of l this probe are such that it rides the surface fairly well, and a three-dimensional plot of the data gives a good contour of any defect Profilometry: Several types of profilometers are used in S/Gs, including optical, strain gauge and eddy current; the latter is being used now at North Anna. There are two types of eddy current profilometers, both of which measure the distance between a fixed eddy current coil and a conductor, or " lift-off" One type uses a spring that is constrained by the tube wall and the distance between the coil and the spring is actually measured. The spring can ride on small point on the tube, but any oxide coating on the tube wall will hold the spring away from the wall and give a false readin The probe may be a single coil and spring, and be rotated, or it can be an array of coils and springs and make the measurements on a single l pass. The rotating, single spring probe has been used to inspect and profile the dents at earlier outage The other type of eddy current profilometer measures the distance I from the fixed coil to the conducting tube directly. The probe is l held near the tube center by " whiskers" or centering spring Reading of the tube surface profile is not a point but rather an

'ntegrated surface response, with portions of the surface nearer the i

i, coil giving more signal than the surface further from the coil. The

" focal spot size" of the coil is approximately equal to the coil diameter, which is 0.165 in. for the probe being use As long as the coil to tube wall distance does not vary within this focal spot size, the probe can make an accurate measurement, limited by the accuracy of the calibration standard. Resolutions of 10 micro-inches are easily obtainable, but the calibration accuracy is more on the order of 100 microinches. The probe used (HC720DT/PS) is an eight coil array and is driven by the MIZ18 (the same instrument used for the other eddy current test) at frequencies of 200kHz and 400 kHz. A common curve of eddy current response is plotted against tube to coil distance (the response is quite non-linear) for the

,

'

general type of coil used. Calibration measurements are then made by inserting the probe in the center of a standard that contains two different internal diameters. Measurements are made on each coil, I e .d a gain and off-set are calculated that best fit each individual coil to a common calibration curv :

As the tube is scanne.d, the coil-to-tube distance is measured for all eight coils. The profile of the tube is then determined from these eight measured value Distorted Indications in Tube Support Plate Intersections With the preceding discussion of eddy current coils and the practical difficulties of eddy current testing in hand, further discussion of the specific problems involved in detecting cracks in the vicinity of i denting, as is the present case, is necessary. The presence of geometric changes, such as those caused by denting severely distorts the eddy current signal. In the past, it has been Industry practice to not report or evaluate signals that do not have a signal-to-noise ratio greater than 3 to 1. In fact, past practice sometimes has dictated that in the case of denting, profilometry rather than eddy current testing has been the oniv inspection metho Unless there was some prior evidence of cracking either due to leaking or l metallographic examination of pulled tubes, it has been standard practice to examine historical data for changes, and if none have occurred, to report the indications as " distorted signals". In the presence of dents, copper or magnetite, ad hoc reporting criteria was developed since standards did not exis ;

I Inspection during the 1980 refueling outage for the first time indicated that denting had occurred in the SG hot legs of North Anna Unit 1 as a consequence of the 1979 resin intrusion events. It was 3 af ter this inspection that boric acid treatment was initiate Subsequent inspections have all confirmed denting to some degree in virtually al' of the seven hot leg support plate intersections, i During the January 1984 forced outage caused by leaking tubes and i plugs in S/G "B", distorted bobbin probe eddy current indications I were first noted. S/G "B" had 13 tubes and S/G "C" had 2 tohes with distorted indications at either the first, second or third hot leg support plate intersections. None of the leakers was identified as having significantly distorted indications although four tubes in "B" and four tubes in "C" were plugged as a consequence of support plate eddy current indications which could be read as being greater than 40% through-wall defects. The distorted signals were reported as

" distorted indications" During the May 1984 refueling outage when all 3 S/Gs were inspected, the distorted signals were not reported, due to the belief that the distortions were due to complex signals arising from the presence of denting, copper and magnetite. This was industry practice at the tim In the August 1985 forced outage due to prirr.ary to secondary leakage in S/G "A", distorted indications were identified, but t,hese tubes were not separated from a larger class of 95 tubes which contained some degree of distortio Consequently, only thirteen tubes were removed from service. In this inspection, support plate eddy current

_ _ _ _ __ _____ ___ _ _ _ _ _ _ _ - . _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __

______________________;

i signals reported as " distorted indications" in the November 1985 inspection, were considered as unquantifiable and were reported as

"possible degradation". A total of 140 of these signals were observed in 95 tubes. Ten of these signals were observed at the sixth or seventh tube support plate elevation on the cold leg side of tubes that were inspected through the U-ben l The unit was returned to service in mid-August 1985 with trace leakage in S/G "B" and "C". In five days, the leak rate suddenly increased to 90 gpd and remained v udy until the November refueling, l During this outage, bobbin coil fady current inspection revealed ,

clear and strongly distorted indications as well as dent signals at i

'

support plate crevice locations. The distorted indications were now classified as signals visible at 400 kHz with characteristics representative of tube degradation, but with an unquantifiable phase angl Examination of these signals at mixed frequencies did not provide sufficient resolution to define the distorted indications at the support plates in terms of tube degradatio Thirty tubes were ;

plugged on the basis of the distorted signal During the spring 1987 refueling outage, more accurate inspections I performed with three-frequency instruments using digital data reduction and analysis techniques permitted the classification of what were previously referred to as " distorted tubesheet indications". These were first observed in 1984, and attempts to resolve these indications led to the use of the 8x1 probe and the rotating pancake coil (RPC). Inspections with these probes resolved l the distorted tubesheet signals into axial cracks for some of the tubes, with the others confirmed as having no cracks. In addition to these, circumferential defects were located in the tubesheet expansion region. These defects were detected by the 8x1 probe and verified and mapped by the RPC. About 150 tube support junctions in steam generators "A" and "B" wera also inspected with the RPC. These intersections had not shown any indications when inspected with the bobbin coil, and they did not show any indications in the RPC ;

inspection. Tube pull data from the 1985 and 1987 outages revealed l l that there was intergranular corrosion (IGC) at the top of the first i tube support plate, on the outer diameter of the tube, up to 28'4 i l through wall. In addition, there were circumferential cracks on the 1 l tube inner diameter at the top of the tube supports, associated with the dent l Eddy-Current Inspection Plan J After the failure of tube R9C51 in a circumferential manner at the top of the seventh tube support in the cold leg, an extensive eddy-current testing program was planned with emphasis on detecting circumferential defect This progran includes the inspection of every tube support junction (and the straight tube sections in between) in all three S/Gs with an 8x1 pancake array probe. This is the most extensive and sensitive inspection program attempted to date

,

1

l l

for S/G inspections,.and will strain the availability of probes and j data analysts. This probe has the sensitivity to detect all inner !

diameter defects, either axial or circumferential, 20% or deeper, with a length of 3/16 in or longer. In addition, it should also be able to detect outer diameter cracks and intergranular attack on either the inner or outer diameter. All indications detected by the 8x1 probe will also be confirmed using the RPC prob The NRC consultant observed the calibration of the 8x1 array probe

'

with the standard and concluded that it was done in the best manner.

l The tubing standard used for the pancake coils has a range of outer i diameter circumferential electro-discharged machined notches ranging from 20 to 100%. The standard scans showed good depth separation between the outer diameter notches of different depths at 400 kHz, and a good separation between the tube support signals and defect signals at 200 kHz, although some of the depth measurement ability was lost at this lower frequency. Although no notch standard was j available for inner diameter defects, they certainly could be detected and estimated from an interpretation between no defect and 100% defec e. Staff and Consultant Conclusions It is concluded that equipment and techniques being used for the present North Anna Unit 1 steam generator inspection are the best available at the present time for commercial eddy-current S/G inspection and that data analysis is being performed in a competent and professional manne The NRC's consultant while on site reviewed and observed the eddy l

!

current data analysis and provided the following evaluation: I

,

"The written data analysis methods are clear and detailed, with more

'

than adequate examples for all three types of eddy current inspections. The senior data analysts are very experienced with the facility, the equipment, and the general types of tube degradation that have occured at other Westinghouse facilities and with the ,

methods of detecting tube degradation. The Westinghouse Intelligent Eddy Current Data Analysis Program (IEDA) which recognizes defect

, signatures is being used as an aid in flagging suspect bobbin coil indications which are then dispositioned by the data analys The ,

data from each tube is independently reviewed by two different analysts, with one using the Westinghouse IEDA system and the other using a Zetec Digital Data Analysis System (DDAS).

All the data analysts are certified at least Level II in accordance with Amer'can Society of Nondestructive Testing ( ASNT) requirement Certification required Industry experience, class room training, technical education, and testing on both general eddy current knowledge and specific eddy current knowledge for S/G inspectio The analysts are given additional training by Westinghouse and are

.- _ _ _ _ _ - _ - _ _ - _ _ _ _ _ _

required to ptss a test that covers the specific data analysis used for the type of eddy current tests used at North Anna Unit 1. These tests are designed and administered by a certified ASNT level III person in eddy current testin In addition, each senior analyst on l two shif ts is also certified as ASNT 1.evel III."

In conclusion, the staff and the NRC's consultant agree that the overall testing program, the extent of the examinations, the execution of the program, and the analysis techniques are acceptable and responsive to NRC recommendations. Further, since the staff and its consultant believe this S/G inspection program is more extensive and sensitive than any other inspection program conducted on any domestic facility to date they find the licensee's inspection program to be acceptabl . Metallurgical Aspects Background Primary to secondary leakage was first detected in February 1985 in trace amounts. Noticeable step changes occurred in April 1985. The leakage gradually increased to a maximum value of 213 GPD in Steam Generator A (S/G A) in late Jul The unit was taken off-line in August 1985 and a video inspection of the tubesheet was performed, while the S/G was filled and pressurized. A total of 3 tubes were identified as leaking. Subsequently, a total of R30 tubes were eddy current tested, twelve (12) of which had ;;iuggble indications (eleven of these were greater than 90% thru wall).

The Jnit returned to service in mid-August with trace leakage in S/Gs B and C. The leakage suddenly increased, after approximately five days on-line, to approximately 90 gpd where it remained until the November 1985 refueling outage. S/G activities performed during this outage included complete eddy current examination of all three S/Gs and the removal of two tubes containing a total of four support plate intersections. The Eddy Curre.nt Test (ECT) program resulted in the plugging of 43 tubes for indications greater than Tech. Spec. limits, plus 30 additional tubes containing strong distorted indications (for preventive purposes). The unit returned to service in January 1986 with no primary to secondary leakage. Trace, intermittent leakage was detected in S/G A, beginning in February 198 The two tubes renoved from the S/G C, during the November 1985 outage contained eddy current signals at the Tube Support Plate (TSP); Row 3, Cc,lumn 41, hot leg (R3C41 HL) included the lower three support plate intersections and the section from the tubesheet region, Tube R9C58 HL included the 1st support plate intersection and the section I

i l

l

from the tubesheet regio These tube sections were evaluated in l detail. This evaluation included detailed physical measurement !

characterization of the metallurgical structure, analysis of deposits, and evaluation of the corrosion resistance'of the tubing in an accelerated cracking environment of steam plus hydrogen at 750 degrees F.

Corrosion and Metallographic Reports - The two tubes pulled in the i November 1985 outage, (R3C41HL and R9C58HL) both contained axially-oriented stress corrosion cracks which initiated from both the ID and OD surfaces. These cracks ranged from 10?; of wall thickness to 100?J through-wall . The cracks were in the vicinity of the t4be-support plate intersections and resulted from increased stress levels due to voluminous magnetite formation between the support plate and the OD tube wall surface. These cracks were above and below tube support plate surfaces and not between support plate q surface Results of corrosion tests indicated the specimens were subject to Primary Water Stress Corrosion Cracking (PWSCC) P Reverse-U-Bend Testing (RUB). Modified Huey-Tests and SCC Tests of ovalized tube samples in steam were normal and within known limits for this material.

The microstructure was considered typical for mill-annealed I-600 material; the material was evaluated using the EPRI Dual-Etch procedure.

During the spring 1987 cefuelling outage, six tubes were identified as leaking (S/G B-2 and S/G C-4) and two tubes were removed from S/G A (R17C31-HL and R18C36-HL) containing two tubeshee+ indications and one support plate indication. These tube samples were submitted to 1 Westinghouse (W) Service Technology Division, Pittsburgh, for l Metallurgical Evaluation, which was planned to include the following: characterization of the degradation; definition of the degradation mechanism; correlation of Eddy Current data to examined degradation; burst testing, and stress-corrosion crack testin A ,

'

series of short, circumferential ID cracks were found at tubesheet top location These results were discussed in a June 17, 1987 meeting with Virginia Power in Richmond. In addition, burst testing has been performed on one of the tube segments.

On Wednesday, July 15, 1987, the reactor was shutdown because of a tube leak in "C" S/G of approximately 600 gpm.

On Monday, July 20, 1987, af ter the removal of the hot leg primary manway covers, water was observed flowing from tube R9C51 ' wi th secondary side water level at approximately the level of the 7th support plate. On July 21, the location of the rupture was fixed at the intersection of tube R9C51 and the cold leg side of the 7th support plate by remote visual examinatio b. Inspection (1) Visual Examination The licensee performed a remote visual examination of the of tube R9C51 using an endoscope electronic imaging probe. The probe of the endoscope was introduced into both the hot and the cold leg tube side The inspectors viewed video tape recordings of the examination. The following observations were made from the video taped examination from the cold leg sid The tube failed over 360 degrees of the circumference, and the fractured ends displaced in the axial direction approximately one-half inch. The tube failed just above support plate #7 on the cold leg sid As viewed from the cold leg side upward at the break location, an area of 60 degrees or less is noted to be angled to the tube This may represent a final failure location in tensile overload on cyclic bendin The fracture surface is generally rough and granular in appearanc As viewed from the side, from the tube ID, the fracture edge is irregular and appears to be circumferential in orientation with little or no axial orientation of the elements of the crack.

l Where several small axial cracks or tears do appear, they seem to be associated with a small thin zone of final ruptur They do not appear to be individual axial crack . The fracture surface does not appear to show a zone of flat fracture which might be associated in an initial fatigue crack. Although some bending may have been associated with the final rupture, no indication of fatigue is obvious as a ,

possible. crack initiation poin I The rough irregular nature of the edge of the fracture is similar to the edge features produced by stress corrosion c ra c ki n g .

There is no clear indication that' the fracture initiated from the ID rather than the OD. The outside OD edge of the fracture cannot be viewed by the fiber optics video prob There are indications from the video tape that a small parallel zone of irregular circumferential cracking is visible in the 90 degree angle view. This small zone of

_ _ _ _ _ _ _ . _ . - _

i

cracking appears to be just below the primary fracture which would place it at the level of the top of the No. 7 support plat The use of video tapes, without further laboratory work is not sufficient to clearly identify the cause and nature of the fractur (2) Document Review (a) The inspectors reviewed the below listed documents to determine whether: the S/G tubes were ordered and specified by VEPCO, W consistent with the FSAR and the VEPC0PurchaseOrder[P.O.);theS/Gtubeswereorderedand  !

specified by W consistent with North Anna FSAR and VEPCO P.O.; and the S/G tubes were t, manufactured, inspected and I tested consistant with North Anna FSAR, VEPC0 P.O. and W -

! Identification Title W-G-677164, Rev. 1 Equipment Specification for  ;

l dtd 12/18/69 Reactor Coolant System "51" l

!

with Addendum 677307 Steam Generator j Re O, dtd 8/6/69 Re , dtd 3/12/71 Re , dtd 1/21/74 Re , dtd 3/8/76 Re , dtd 10/15/84 North Anna Final Safety Analysis Report W-2656A95 Material - Nickel Chromium Iron dtd 12/5/69 Tubing ASME SB-163 Specification for Seamless Nickel and Nickel Alloy Condenser and Heat Exchanger Tubes W Purchase Order Purchase Order from W Tampa No. 54-5-1441-400 Division to W Blairsville for S/G Tubes Westro Alloy Tubing Test Results for Set 44

_ _ _ _ _ _ _ _ _ _ _ - _ - _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ - _ _ _ _ - _ - . - _ _ --. . . _ _ -

l

i W Drawing No. 632C566, 51 Series Vert. Stm. Gen. Tube Rev. 7 Schedule Westro Alloy Tubing Tcst Results for Set 45 l Tubing Log for Shop Order 1263 Form N1 Manufacturer's Data Report for Nuclear Vessels for Steam Generator 1263 (b) The inspectors reviewed the below listed documents to evalute the metallurgical examination and testing performed by W at the licensee direction of S/G tubing removed from Unit 1 S/G ,

Documents Reviewed 1 Identification Title WCAP-11311, " North Anna Unit 1 Steam dated October 86 Generator Tube Integrity Safety Evaluation" WCAP, 11310 " North Anna Unit 1 Steam dated October 86 Generator Tube Integrity .

Safety Evaluation" SG 86-11-017, " Examination of Tubes !

dated November 86 R3C41HL and R9C59HL Steam Generator C, North Anna Unit 1" With respect to the inspection above, the licensee and W l Pittsburgh were unable, at the time of this inspection, to provide the inspectors copies of the below listed documents. It should be noted that there is no regulatory 4 basis requiring the retention of their documents; ,

therefore, no further action will be taken in this matte Unavailable Documents P.O. From VEPC0 to W for S/G P0 from W to W Tampa for S/G Specification for S/G

_ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ .

_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _ _ _ _ _ _ - _ _ _ _ _

(c) Conclusions The inspectors determined that the licensee with their contractor W have pursued an aggressive and conservative metallurgical monitoring and evaluation program. W has i

identified both Primary Water Stress Corrosion Cracking I

(PWSCC) from the ID of the tubes and Intergranular Attack (IGA) on the OD of the tubes. The W studies indicate that the general corrosion rate (not including localized PWSCC  ;

'

and IGA) is consistent with the FSA . Chemistry Aspects Chemistry Control Separation of the ruptured steam generator tube initiated a series of events that drastically changed chemistry control of both the primary and secondary coolant systems and resulted in the development of a potentially explosive gaseous environment in "C" steam generato The licensee was able to cope with these changes, maintain reactivity control through the use of the chemical shim, boric acid, and safely eliminate the hydrogen-oxygen gas mixtur At the time of the tube break Unit I was operating at 100% power and maintaining chemistry control as recommended by the Steam Generators Owners Group (SG0G) and the Electric Power Research Institute (EPRI).

The most recent analysis of the reactor coolant had been made at 0033 hours3.819444e-4 days <br />0.00917 hours <br />5.456349e-5 weeks <br />1.25565e-5 months <br /> on July 15, 1987, and showed the boron concentration in the reactor coolant to have been 1320 parts per million (ppm) and the concentration of hydrogen dissolved in the reactor coolant to have been 36.9 cc/kg. Likewise., an analysis of steam generator blowdown

, from "C" steam generator had been made at 0515 hours0.00596 days <br />0.143 hours <br />8.515212e-4 weeks <br />1.959575e-4 months <br /> on July 15 and

!

had revealed the key control parameters were within acceptable ranges

! (pH = 7.5, boron = 9 ppm, conductivity = 1.85 umho/cm).

At approximately 0624 on July 15, in-line monitors for the steam generator blowdown system indicated a sudden decrease in pH and .an

,

increase in conductivity. The rate of the pH change increased at I about the time, 0630 hours0.00729 days <br />0.175 hours <br />0.00104 weeks <br />2.39715e-4 months <br />, that the main steam radiation monitor alarmed. These changes reflected the mixing of reactor coolant (pH of ~7.0, ~2 ppm lithium hydroxide, and ~1300 ppm boric acid) and the steam generator wate (1) Boron Control of Reactivity At 0635 on July 15, safety injection was initiated for approximately 34 minutes. During this time highly borated water (~2300 ppm) was injected into the primary coolant, and, even though borated water was being lost from the primary system to

"C" steam generator, the concentration of boron in the reactor coolant increased from 1320 to ~1700 ppm at 0807 hours0.00934 days <br />0.224 hours <br />0.00133 weeks <br />3.070635e-4 months <br /> and to

i

i

~1900 ppm when the emergency was terminated at 1335 hours0.0155 days <br />0.371 hours <br />0.00221 weeks <br />5.079675e-4 months <br /> on l July 1 (The boric acid concentration continued to rise to

~2000 ppm throughout the next six hours and remained at that level.) These changes were monitored by the chemistry staff on a thirty minute interval from 0830 to 1138 hours0.0132 days <br />0.316 hours <br />0.00188 weeks <br />4.33009e-4 months <br /> and then at hourly intervals at the request of the Control Room operators.

l l (2) Effect of Accident on Secondary Water Chemistry Control The chemistry staff monitored changes in the pH and conductivity in the condensate, feedwater, and "C" main steam line from approximately 0630 hours0.00729 days <br />0.175 hours <br />0.00104 weeks <br />2.39715e-4 months <br /> until the "C" main steam line was isolated at 0646 hours0.00748 days <br />0.179 hours <br />0.00107 weeks <br />2.45803e-4 months <br />. Chenges in these parameters reflected the increase in acidity (decrease in pH) caused by influx of the more acidic reactor coolant and volatile boric acid throughout the secondary coolant. Monitoring was terminated at 0646 hours0.00748 days <br />0.179 hours <br />0.00107 weeks <br />2.45803e-4 months <br /> when the main trip valve from "C" steam generator was closed, and the unit entered into an emergency mode of operation. At 1310 hours0.0152 days <br />0.364 hours <br />0.00217 weeks <br />4.98455e-4 months <br />, near the end of the emergency, condenser vacuum was broken and AVT (all volatile treatment) chemistry control of the secondary coolant was terminate A final analysis of the blowdown from all three steam generators was performed at 2305 hours0.0267 days <br />0.64 hours <br />0.00381 weeks <br />8.770525e-4 months <br /> on July 1 The values for all control parameters except pH, such as )

conductivity, boron (for denting control), and ammonia were higher than their pre-accident values by a factor of 2 to 10 in

"A" and "B" steam generators. The concentration of boron in "C" j steam generator water had increased from a pre-accident value of 9 ppm to 564 ppm as the result of dilution by the reactor coolant.

l Although the tube rupture caused the conductivity of the water ,

'

i in all three steam generators to exceed action level limits associated with corrosion control, these levels were not considered to have had any safety significance, nor to have caused additional tube corrosion because the unit was rapidly l cooled to Mode 5 within seven hours, l (3) iransfer of Radioisotopic Species from the Reactor Coolant As part of the licensee's efforts to establish the magnitude of I radioactive releases from the primary coolant subsequent to the tube failure, samples of water in "C" steam generator were i rep 9atedly analyzed after the steam generator had been' isolate Th - highest values for iodine isotopes (Iodine-131, Iodine-132, Icdine-133, Iodine-134, and Iodine-135) as well as the l

_____-____ _ __ -

1 I

l 30 l l

activation produ.ct Sodium-24 were observed ir.the initial sample i at 0800 hours0.00926 days <br />0.222 hours <br />0.00132 weeks <br />3.044e-4 months <br /> on July 1 When compared with pre-accident analyses made on July 6 and July 13, the post-accident values were observed to have increased by an approximate factor of 106, i.e., from 10 7 - 10 8 uci/g to 10 3 - 10-' uci/g, an iodine dose equivalent of 6.79 x 10 3 uC1/ (Increases in the concentrations of these radionuclides in "A" steam generator were ~102. ) By 2305 hours0.0267 days <br />0.64 hours <br />0.00381 weeks <br />8.770525e-4 months <br /> of July 15, the concentrations of soluble radio nuclides in "C" steam generator water had decreased by factors of 3 to 102 Since there was no loss of water except through letdown of the primary coolant the licensee ,

was able to control these radioisotopes through ncemal radwaste l cleanu (4) Control of Hydrogen ,

As discussed earlier, Unit 1 was operating with a hydrogen overpressure in the reactor coolant to ensure that no dissolved oxygen remained in the water. At the time of the tube break the equilibrium concentration of hydrogen in the primary coolant was l

'

! 36.9 cc/kg. During the early stages of the incident sufficient l loss of reactor coolant into the "C" steam generator occurred to I

decrease the equilibrium concentration to 13 cc/k Because of l the rapid de pressurization of the primary coolant in the hot steam generator water the equilibrium concentration of dissolved hydrogen in the steam generator decreased as the excess hydrogen volatized to the steam space in the upper volume of the steam generator. The first gas analysis of this space was performed at 2250 hours0.026 days <br />0.625 hours <br />0.00372 weeks <br />8.56125e-4 months <br /> on July 16, and concentrations of gases were determined to be 4.4?s Hydrogen, 16?s Oxygen, and 76?s Nitroge {

As a result of concerns by the AIT, a second analysis was j performed at 1425 hours0.0165 days <br />0.396 hours <br />0.00236 weeks <br />5.422125e-4 months <br /> on July 17 at which time the concentrations of the gases was reported to be 15.3?s Hydrogen, 13.9?s Oxygen, and 8ois Nitrogen. The results were discussed with NRC representative The presence of gaseous oxygen has been attributed by the licensee to the use of oxygenated water from the condensate storage tank (CST) to achieve backfilling of the reactor coolant system via the tube break in the steam generator. During normal operation demineralized and deaerated water is transferred from the water treatment plant to the CST through a pipe at the base of the tank. The tank is vented to air, and, thus the water at the top is saturated with air (~8 ppm of dissolved oxygen). I However, condensate makeup water and suction for the auxiliary j feedwater pumps is also transferred through a pipe at the base !

of the CST. Consequently, the concentration of dissolved oxygen l in water in the lowest part of the CST was thought to be very l

_ _ _ _ _ _ _ _ _ _ _ _ - _ .

low (i .e. , ~100 ppb); however, the licensee was not able to

~

provide an exact valu Therefore, the amount of oxygen transferred via the auxiliary feedwater pumps to "C" steam generator was not know During this event, the licensee did not add hydrazine to this water to reduce the concentration of oxyge A generally accepted guideline is ohat a mixture of hydrogen and oxygen gases in ratios in excess of 4?s hydrogen and 5?s oxygen !

represents a flammable condition. Mixtures of these gases in ratios of 18 to 59?; hydrogen and >5?; oxygen will detonate in the presence of an ignition sourc The inspector was informed by l the licensee that this information was known and taken into consideration during the development of hydrogen removal procedure l The NRC review of the results of the July 17 analysis, discussed the normal procedure for purging gas from the steam generator, and their plans to purge the "C" S/G gas space. The concentra-tion of hydrogen in the upper volume of the steam generators was subsequently reduced to a non-detectable level by means of emergency procedure, EWR 87-482, approved on July 16, 1987, by the chairman of the licensee's safety committee (SNSOC) which I referenced Engineering Drawing 11715-FK-12A. Because a cap had been welded on the vent at the top of S/G '"C", a ' jumper' for purging gases through an existing rei?rence leg penetration in

, the steam generator was attached at 0400 hours0.00463 days <br />0.111 hours <br />6.613757e-4 weeks <br />1.522e-4 months <br /> on July 18.

'

Purging via the process vent was accomplished by a vacuum pump while replacing the gas in the steam generator with nitrogen. A j gas sample taken at 0420 on July 19 was analyzed to contain 094 '

Hydrogen, 22?s Oxygen, and 86?s Nitroge Although the removal of hydrogen gas from the "C" steam generator was performed in a controlled manner, the inspector discussed the potential for an explosion during such an operatio The licensee agreed to analyze this operation in greater detail and to consider incorporating the results into a permanent procedur This is an Inspector Followup Item (338/87-24-09).

(5) Summary A review of the licensee's actions to control the plant's chemistry during the af termath of the tube break shows that proper and adequate measures were taken to maintain safe environments for the recovery perio Layup of Unit 1 The inspector reviewed the Control Room logs to identify the steps taken, and the chronology of these actions, to remove the

.

.

.. . .. .. . .

. . . . . .

.

. ..

I i

contaminated primary.and secondary coolants, to flush the systems to remove potential corrosive, and to layup the steam generators and the carbon steel components of the secondary system. "A" and "B" steam generators were cleaned by drain and fill techniques and, after AVT chemicals were added, layed up wet with the chemically controlled water cycling through the steam generator, while a nitrogen atmosphere was maintained above the water. This is the recommended method for short term layup of steam generator "C" steam generator and the associated loop of the reactor coolant system were also cleaned by drain and fill techniques after hydrogen gas had been eliminated. A nitrogen atmosphere was always maintained to prevent entry of air. After the "C" steam generator was tested for leaks on July 26, it was drained and layed up in this condition under an atmosphere of nitroge The inspector considered the licensee's layup program for the steam

generators to be acceptable. At the time of this inspection, the remainder of the secondary cooling water system (e.g. , condenser hotwell and condensate /feedwater lines) were filled with AVT-controlled water but were in a stagnant condition. The inspector discussed with the licensee the desirability of continuing to cycle water in these systems to minimize general corrosion of the carbon steel pip { Role of Chemistry in the Tube Break Although the cause of the tube break cannot be established for certain until metallurgical examinations are made, the licensee's preliminary analysis indicated that the circumferential crack was ;

initiated by primary and secondary stress corrosion. Most of the cracking is thought to have initiated from the inner surface of the tube as the result of the denting phenomenon that was exhibited at many PWRs before increasingly stringent chemistry control, based on AVT chemistry, was recommended by the Steam Generator Owners Group (SG0G) in 198 As the result of tube degradation in several PWRs in NRC Region II, an inspection program was developed in 1983 to use the information being developed by EPRI and the SG0G to encourage licensees to improve chemistry control of the primary and the secondary coolant Four inspections have been made in this area at North Anna since June 1984. During the initial inspection, the inspector noted that denting of steam generator tubes had been observed in Unit 1 since the first refueling outage in September 1979. During the outage the licensee had also removed a 'small amount' of metal oxide sludge (409s by weight copper and 18?; by weight iron) from the three steam generator The licensee had attributed this evidence of general corrosion to inleakage of ion exchange resin particles from the

i l

33 l condensate polisher to the steam generators and subsequent thermal degradation of the resins to corrosive acidic sulfate specie In the second fuel cycle the licensee, on the advice of Westinghouse,.

began to add small concentrations (5 to 10 ppm) of boric acid to the secondary system in an effort to stop the denting proces No j additional dents were observed during the second refueling outage and only a small amount of sludge was remove However, during the seconi fuel cycle the steam generators were subjected again to an i acidic environment for a limited period of time as the result of a leak of sulfuric acid from the water treatment plan Subsequent to these chemistry inspections at North Anna it was evident that the conditions for denting continued to exist; i.e.,

,

hundreds of pounds of metal oxide were being transported to each

! steam generator during each fuel cycle; large amounts of copper were present in this sludge; and transport of chloride ions were also present in the steam generators as the result of condenser leaks.

I During the last two years significant progress has been made both in eliminating ingress of corrosive contaminants into the secondary water system as well as in the operation of the condensate cleanup system. Also chemistry control has significantly improved as the result of the licensee's endorsement and implementation of the guidelines developed for PWR water chemistry by the SGOG. However, I these improvements have not slowed the general wastage of both carbon  !

steel pipe and copper alloy feedwater tube During the last refueling outage a total of approximately 3500 pounds of copper-iron oxide sludge was removed from the tube sheet regions of the three  ;

steam generators, and 118 additional tubes were plugged because of l indications attributed to dentin It is evident that the licensee

'

needs to take additional actions to slow down degradation that can be attributed to many factors over the operating life of the unit. The licensee has already delegated the responsibility for evaluating such actions to the North Anna Recovery organizatio . Radiological Aspects Environmental Two points were identified by the licensee as pathways for release of gaseous radionuclides to the environment during and subsequent to the S/G tube rupture event. These were the Condenser Air Ejector (CAE)

and the steam driven Auxiliary Feedwater Pump Turbine (Terry Turbine). For the CAE, a release pathway existed from 0630 to 0646 at which time S/G "C" was isolate Small amounts of activity continued to be discharged via this pathway ' from S/Gs "A" and "B" until 0756. At 0756, CAE exhaust was diverted to containment and release via this pathway ceased. The pathway from the Terry Turbine was determined to exist from 0636 hours0.00736 days <br />0.177 hours <br />0.00105 weeks <br />2.41998e-4 months <br /> (initiation of the auxiliary feedwater pumps) to 0745 when the pump was shutdow The steam

_ _ _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ - _

-__- _

34 l l

l

'

supplying the Terry T.urbine originated from all three S/Gs from 0630 to 0648 at which time the "C" S/G steam supply to the turbine was )

manually isolate S/G "A" and "B" continued to release small amounts of radioactivity until 074 Activity released was determined from the Kamen Vent-Vent "A" stack l monitor strip chart recording taken during the event (the exhaust path for the CAE), and an isotopic analysis of a grab sample taken j from the CAE exhaust at 0650 hours0.00752 days <br />0.181 hours <br />0.00107 weeks <br />2.47325e-4 months <br />. Duration of the release was 86 minutes from the CAE and contained a total of 9.90E-2 curies of noble gase For the Terry Turbine, a release of 6.04E-2 curies of noble gases was calculated. It was assumed that at the time in question (0636 to 0648 hours0.0075 days <br />0.18 hours <br />0.00107 weeks <br />2.46564e-4 months <br />), the release recorded by the Kamen monitor in the stack from the CAE immediately prior to turbine trip was representative of the release rate from the Terry Turbin Since both CAE and the Terry Turbine take steam from the S/G main steam header, and since these two routes represented the majority of steam flow for the plant at that time, this assumption appeared conservative in that when flow decreased in one pathway, an increase should be observable in the other. This was confirmed by observation of radiation monitor chart recordings from the Kamen Vent-Vent "A" and the Terry Turbine exhaus Using the isotopic distribution as determined from the CAE grab sample, the release rate determined by the Kamen monitor, the flow rate for the Terry Turbine, and assuming a release duration of 12 minutes, release from all three steam generators via this pathway was determine i Tritium, particulate and halogen ccmponents of the Terry Turbine !

release were also considered. These parameters were determined using the radionuclides distribution determined from S/G blowdown samples taken at 0943 hours0.0109 days <br />0.262 hours <br />0.00156 weeks <br />3.588115e-4 months <br />. Release of tritium, particulate, and halogen l from S/G "C" was determined to be 0.024 curie For all release pathways a total activity of 0.159 curies was release Eighty-four percent of this activity was noble gase Radionuclides of highest activity released were argon-41, xenon-133, 135m, 135 and 138, and krypton-85m, 37 and 88. Activity released was determined to be equivalent to a whole body dose rate of 4.04E-4 mrem / hour or a dose of 5.7E-4 mren for the event at the site boundary. This value is significantly below the Technical Specifi-cation 3.11.2.1 limit of 500 mrem whole body dos Inplant and Offsite During the event, environmental monitoring teams were dispatched both offsite and onsite to provide data for determining the magnitude of the release and for performance of dose projection Radiation

levels, removable contamination levels and sample results were obtained; however no detectable activity above background was .

observed from any of these medi j i

Few radiological consequences were observed inplant during and l subsequent to the event. When releases were known to be occurring ,

inplant, air samples were obtained in the vicinities of the CAE and !

the Terry Turbin Examination of counting records showed inplant ;

activity to be less than 25 percent of the Maximum Permissible Concentration (MPC). Smears taken in these same areas showed no detectable removable activit Additionally, surveys for fixed and j removable radiation in the Technical Support Center and the d Operations Support Center during the event confirmed that radiation above background levels was not detectabl c. Radiation Monitors One of the first indications of a problem at the facility was sounding of an annunciator alarm on the main steam radiation monitors. "A" and "B" steamline monitors were in alert and the "C" steamline monitor was in high alar High alarm setpoint is 0.26 millirem (mr) per hour. Other radiation monitors of concern for this event were effluent monitors (process vent, vent-vent stack A, and vent-vent stack B), CAE monitor and the Terry Turbine monitor As described above, the CAE monitor had been declared out of service during the time of the event and the licensee was taking periodic grab samples of this effluent as required by Technical  ;

Specifications. Consequently, the inspector reviewed the calibration '

records for the Unit 1 main steamline monitors, the Terry Turbine and the vent and process monitors. Records indicated that these monitors had been calibrated on the required frequency (18 or 24 months).

However, examination of the data from the last calibration showed that, when "as found" readings for steamline "A" and "C" monitors (RI-MS-170 and RI-MS-172, respectively) had been taken during return for routine calibration, the monitors responded nonconservatively for radiation levels greater than 645 mr/ hour by more than 50 and 60 percent, respectively. Similarly, for the Terry Turbine radiation I monitor, upon return for routine calibration, the "as found" readings showed an under response in that, below 2900 mr/ hour, the monitor

'

showed no respons Technical Specification (TS) 4.3.3.11 states that each radioactive gaseous effluent monitoring instrumentation channel shall be-demonstrated operable by performance of the channel check, source check, channel calibration, and channel functional tes TS 1.17 states that a system, subsystem, train, component, or device shall be operable or have operability when it is capable of performing its specified functio )

<

.

l

. .. .. ... . . . . . . . . . . . . .. . . _ --

l

l

)

The inspector reviewed the circumstances surrounding the out of '

service declaration for the CAE radiation monitor (1-RM-SV-121). The licensee stated that verification of operability of SV-121 (channel check) was performed three times daily in accordance with Procedures PT-37, Radiation Monitoring Equipment Check, November 14, 198 Section 5.1 of PT-37 stated that the test shall be considered acceptable if, for each radiation monitor, there is freedom of needle movemen A review of SV-121 readings taken between July 10 and July 15, 1987, were as follows:

Counts Per Minute 0000 to 0800 to 1600 to Date 0800 Hours 1600 Hours 2400 Ho g 07/10/87 90 100 10 /11/87 60 25 100 07/12/87 90 15 <10 07/13/87 <10 Out of service Out of service 07/14/87 450 1000 Out of service 07/15/87 Out of service Out of service Out of service During discussions with the licensee, the inspector was informed that SV-121 had been declared out of service on July 13, 1987, when the monitor failed low, i.e., zero activity and no free needle movement, ,

On July 14, 1987, the line to the monitor was suspected of containing I moisture. The line was drained, needle movement was again observed, j and the monitor declared back in service. Later that same day, the l monitor again failed low and was declared out of servic In discussions with licensee representatives, the inspector was  !

informed that PT-37 did not define any criteria for acceptability of the channel check or source check other than that of free needle s

movemen Consequently, an inoperability declaration ~ for any radiation monitor was left to operator discretio The inspector stated that the consistently declining counts per minute (cpm)

observed on SV-121 monitor on July 12 and 13,1987, when the reactor was increasing in power level suggests that inoperability of the monitor may have occurred as much as 24 hours2.777778e-4 days <br />0.00667 hours <br />3.968254e-5 weeks <br />9.132e-6 months <br /> or more prior to the monitor being declared inoperabl Further, monitor readings obtained on July 14, 1987, when the monitor was declared operable again were not consistent with previous data. The inspector' stated that free needle movement of the monitor by itself may not be adequate to declare a monitor operable in that it provided no reasonable assurance that the needle movement represented monitor eccuracy which then verified that the monitor was capable of fulfilling its required function, i.e., was operabl i Failure to verify 1-RM-SV-121 operability through a daily channel check was identified as a violation of TS 4.3.3.11 (50-338/87-24-03).

12. Licensee's Response to Event

- __ -

37 The resident inspector reported to the control room after hearing the

"Gaitronics" announcement: " Shift supervisor, report to the control room, now."

The inspector observed pressurizer pressure and level were dropping at a _,

significant rate. A decision was made by the shift supervisor to manually trip the reacto Prior to the inspector's entrance, letdown had been isolated and makeup had been maximize The immediate action steps of EP-0 " Reactor Trip or Safety Injection" were implemente The safety injection initiated automatically on low pressurizer pressure. The licensee used an extra control room operator to read the Emergency Procedures while the Unit 1 control room operator and backboards operator performed the step The inspector was standing behind the procedure reader and was not able 'o hear all of the steps or the response to these steps. He did, howevet hear the directions from the shift supervisor and logged the times that these were given. Later, interviews by the licensee with the operators indicated that they understood the steps or the procedure reader verified the This was not evident to the inspector who had previously witnessed the same accident on the simulator where positive acknowledgements were made to each step of the procedur When it was identified that there was a steam generator tube rupture, the licensee entered Procedure EP-3 " Steam Generator Tube Rupture". Step 3

" Identify Ruptured Steam Generators" requires the operators to proceed to step 4 which isolates the ruptured S/G if any of these four conditions are met:

(1) Unexpected increases in S/G narrow range level ,

OR  !

(2) High radiation from any S/G Blow down monitor OR (3) High radiation from any S/G steamline 0R (4) High radiation from any S/G sampl At 0630 and for the next several minutes, alarms were reset for high radiation f rom "C" S/G steamline monitor. Radiation alert alarm were being received on S/Gs "A" and "B" steamline monitors. At 0640, contrary to step 3 of EP-3, S/G "C" was not isolated until completion of steps 6 thru 13. This constituted an additional six minutes in which the S/G was not isolated. Following the event, the licensee interviewed each person directly involved in the event. The shift supervisor stated that at step 3 of EP-3, he felt he was not able to make the decision as to which steam generator ruptured; therefore, he followed "the response not obtained" portion of the procedure. This allowed him to follow steps 6 through 13 l

t

_

4 before securing the stearg generator with the ruptured tube in step (Prior to this event, a response on the S/G blowdown monitors was )

expected.) This appears to be a failure to follow the emergency procedure but is not cited as a violation at this time because of the shift l supervisor's stated concern that he was not sure that "C" was the-leaking S/G until he confirmed the radiation alarm by an increase in S/G leve The concerns about verbatim compliance were discussed with licensee management and they stated that the Westinghouse Owners Group would be requested to review the event in detail and to compare the operator l actions and results with the method recommended in the guidelines. The 1 results of this study will be presented to the NPsC. This is identified as an Unresolved Item (338/87-24-04), pending NRC review of the results of Westinghouse Owners Group stud The reactor was then depressurized and cooled down in accordance with ES 3.1 " Post S/G Cooldown Using Backfill" I

'

A subsequent evaluation by the NRC determined that the operators had been trained on Revision 0 and more recently on Revision 1 which was implemented in June 1987. There are significant changes between Revision j

' 0 and Revision 1 specifically in regards to EP-3 " Steam Generator Tube '

Rupture" The inspectors are examining the differences at this tim This will be identified as IFI 338/87-24-05. A S/G tube leak is an event l in which the operators have been trained extensively on the simulato Prior to the initiation of the trip, there had been several alarms on the S/G radiation monitor These alarms were silenced by the backboards i operator, but Health Physics was not notified immediately. The alarm I response procedure requires the operator to notify Health Physic Failure to notify Health Physics as required by alarm response at the time

.

i that the alarms were received was identified as a violation of Technical Specification 6.8.1 - Failure to follow alarm response procedures (338/87-24-06).

EPIP 1.01 " Emergency Manager's Controlling Procedure" was entered at 0639 and step 2.b.1 requests Health Physics to initiate EPIP 4.0 This was done at 6:50 a.m. Step 2 of EP-4.01 requires the senior health physics individual on t ite to report to the control room and step 5 of the procedure requires initiation of EPIP 4.08 " Initial Offsite Release Assessment". Health Physics reported to the control room and was told that the air ejector had diverted to the containment. At this time the air ejector discharge had not been diverted to the containmen This could have been significant because EPIP 4.01 " Radiological Assessment Director Controlling Procedure", Step 3, allows Health Physics to proceed to step 15 if no release has occurred; however, Health Physics appears to have implemented the necessary step EPIP 4.08 was entered at 7:08 a.m. The samples taken for determining the release were taken af ter the affected S/G was isolated. The inspector was not aware of any discussion that operations or Health Physics tright have l

-

. -

had concerning the inoperability of the divert to the containment with~

RMS-SV-121 faile The Unit 2 assistant shift supervisor, was required by his duties for Unit 2 to read the action status log for Unit 2 but not required to read that for Unit 1 and therefore may not have been aware of the inoperable Unit I radiation monito None of the abnormal procedures address the air ejector radiation monitor being failed low. The technical specifications do not address the automatic functions of failed monitors and only require grab samples every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> when the monitor is inoperabl l l

The inspector noted that 1-RM-SV-121 high and high high setpoint had been I increased several times prior to the unit 1 outage, but had not been reset l following the outage. The radiation background had not been e;tablishe I The significance of this is the alarm would have come on at a much higher i setpoint since the procedures determine the setpoint using a formula based j on the present count rate. The alarm setpoints should have been reset j after the outag '

! 10 CFR 50 Appendix B, XVI, Corrective Action, requires that measures shall

! be established to assure that conditions adverse to quality, and nonconformances are promptly identified and correcte In the case of l

'

significant conditions adverse to quality, the measures shall assure that the cause of this condition is determined and corrective action taken to preclude repetitio Contrary to the above, on July 1987 at 0833,1-RMS-SV-121 was removed from serv.ce and work request 453089 was written. The monitor lines were drained, the monitor was declared operable and the work request was cancelled. The monitor was restored to service before assuring that it was operable. No functional or calibration test was run and on July 14, 1987, at 2238, the monitor failed and the work request was reinstate The significance of this event is that the monitor was inoperable during the tube rupture and if the hdgh-high setpoint had been reached, it would not have automatically diverted its exhaust to containmen Failure to verify operabilty of the CAE radiation monitor prior to returning to service is another example of violation 338/87-24-0 When the work request was worked after the incident, it was determined that the voltage i supply had faile The failed monitor would also have closed off steam to the air ejectors if the hi-hi setpoint had been reached in conjunction with a safety injection. The licensee stated that occasionally when the SJAE exhaust is switched, the SJAE will trip which in turn would cause the condenser to begin losing vacuu Loss of condenser vacuum would result in loss of ability to bring the plant to cold shutdown using the condenser, and would require use of the S/G Power Operated Relief Valve (PORV) which discharges unmonitored directly to the atmospher In their review of this event, the licensee agreed to reevaluate the adequacy of the design and to determine if the setpoints of this equipment should be changed to assure that the condenser would be available for plant cooldown rather than

_

,

having to use the S/G F.0RV s . This is an Inspector Followup Item (338/87-24-07).

l The low head safety injection pump was secured early in the procedure and l because the " response not obtained" portion of EP-3 was used, it was i secured before the affected steam generator was isolated. There is no

! advantage to securing the low head safety injection early because it remains on recirculation and is readily available if the pressure l

decreases. When the licensee was questioned about this, the inspector was shown recommendations Liy the Westinghouse Owners Group that stated that-the purpose for stopping the pumps is to avoid pump damage. The recommendation is apparently not applicable to North Anna, since North Anna has full recirculation flow back to the RWST instead of the pump suction. This is identified as Inspector Followup Item (338/87-24-08).

In summary, the following areas are identified as weaknesses and need to be resolve ) Obtain resolution on whether the emergency procedures need to be followed verbati ) Communications between health physics and operations needs improvements, i.e., timeliness and accurac ) Operability of radiation mon.itors, including resolution of how to handle defeated automatic actions when the monitors have failed low.

l This also includes the containment purge monitor and the process vent I l monito ) The present location of the Kamen and NRC monitors require an operator on the backboards until he can be reliaved by nealth physic ) Assurance that the monitor setpoints are reset after a refueling outag ) Re-emphasis on procedure reading and response to the procedure reade ) The licensee needs to resolve the fact that control room positive pressure is lost when people constantly enter and exit the control room during an emergency. In particular in this event, the turbine building could have been considered an airborne are ) The automatic features of securing the steam to the air ejectors and diverting to the containment needs to be reexamined against the possibility of not having the condenser available for the steam dump ) Clearly define when the low head safety injection should be secure It appears that the Owners Group is in favor of securing it early in

_ _ _ _ - _ _ _ _ _ _ _

particular because some plants might not recirculate back to the RWST. This should be considered a generic issue and possibility of damage to the pump assessed if it can not run on recirculation for extended periods.

13. Consideration of Shutdown of Unit 2 During the steam generator tube rupture event on Unit 1, the licensee continued to operate Unit Discussions between the inspectors and licensee management determined that consideration for shutdown of Unit 2 was continually being evaluated by licensee management during and after the even The licensee conducted a formal evaluation for potential unreviewed safety question for continued operation of Unit 2 and had their safety committee review and approve the evaluation on July 29, 1987. The licensee provided a copy of the evaluation to the inspectors. The inspectors reviewed the Safety Evaluation for continued operation of Unit The inspectors consider that the licensee's conclusion for continued operation of Unit 2 through September 1987, based on steam generator material condition, operational time, and inspection is conservative.

14. Investigation and Corrective Action The results of the investigation of the cause of the failure and the corrective action will be documented in a subsequent report or NUREG.

15. Consideration for Restart NRC, Region II issued a confirmation of action letter to the licensee dated July 22, 198 In that letter, the requirement to obtain NRC concurrences prior to Unit I restart (Mode 2) was confirme H S T C I

L S A Y DN E H EO H P C TIC N

REM f E G

N U"*i li Y E

- D RC

- NR T A N A E )F S O

- M NETO U I

D T

- M E (L A OC N Z

- C I i I

- N

- A

-

G R

O 1 Y R &G T

N R E R N Y E E A I R

7 V G EYS E h

C A LTN V A O CEE O T N UF C T

C C E A A N AI S L E RM R I

- I

-

NR E L N E O A O I G I R T A C T I U A N IIIiIl{;, NA G T A HU L I

F S M C A E

T V E E

C R N I A A N P E E T R N

I A&

M

%.

, ll 1