ML20042F354

Special Rept Re Failure to Prevent Performance Degradation of Reactor Bldg Cooling Units.Caused by Mgt Deficiency & Inadequate Program.Cooling Unit Declared Inoperable & Removed from Svc for Cleaning & Placed Back in Operation

ML20042F354
Person / Time
Site:	Oconee
Issue date:	04/30/1990
From:	Barron H DUKE POWER CO.
To:	NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
NUDOCS 9005080221
Download: ML20042F354 (10)

Text

.,:. 6i Dukeikuwr Compazy - (803) 8a25363

.= 0:ener Nudear Station .

PO Bar H39 '

Seneca, S C 296t9 DUKEPOWER April 30, 1990 i

U. S. Nuclear Regulatory Commission Document Control Desk  ;

Washington, DC 20555 E!

Subject:

Oconee Nuclear Station, Unit-3 g Docket No. 50-287 Species 1 Report Concerning Failure to Prevent Performance Degradation of

' Reactor Building Cooling Units o

Gentlemen This report is provided for information regarding failure to prevent performance degradation of. Reactor Building Cooling Units.

If you have any questions, please contact' Rick Matheson at (803) 885-3119..

Very truly yours, MM uw H. B. Barron i Station Manager

/ftr Attachment xc: Mr. S. B. Ebneter Mr. P. H. Skinner Regional Administrator, Region II NRC Resident Inspector i U.S. Nuclear Regulatory Commission Oconee. Nuclear Station 101 Marietta St., NW, Suite 2900 Atlanta, Georgia 30323 INPO Records' Center Suite'1500 Mr. L. A. Weins 1100 Circle'75' Parkway Office o?~ Nuclear Reactor Regulation Atlanta, Georgia 30339 U.S.. Nuclear Regulatory Commission  !

Washington, DC 20555 A$

F 9005080221 900430 PDR ADOCK 05000287

.g PDC -

t

o . ,

MANAGEMENT DEFICIENpY RESULTS IN FAILURE TO DETECT PERFORMAM"E i DEGRADATION OF REACTOR BUILDING COOLING UNITS ARSTRACT On February 20, 1990, the Design Engineering Safety Analysis group performed an evaluation of the cooling capacities of the Unit 3 Reactor' Building Cooling Units (9BCUs), based on test data gathered earlier that day. The evaluation' concluded that, assuming a single failure of the'RBCU-train with the highest capacity, service-induced fouling of the two worst' trains, ("A" and "C"), could have prevented-the RBCU system from performingLits intended safety functions.. Immediate corrective action was taken at 1452 hours0.0168 days <br />0.403 hours <br />0.0024 weeks <br />5.52486e-4 months <br />, as soon as the evaluation results were received at the station; RBCU train "C" was declared inoperable, placing the unit into a seven day ILO; the 3B train was placed in operation; and then the 3C RBCU was taken out of service for cleaning and testing. At 1334 hours0.0154 days <br />0.371 hours <br />0.00221 weeks <br />5.07587e-4 months <br />, on=

February 122, the 3C RBCU was declared operable following testing which indicated capacity was 61% of design capacity. Subsequent corrective actions vore to clean and test the "A" and "B" RBCU trains. Previous LERs have reported Oconce, Nuclear Station's recurring problems with RBCU

' degradation. Corrective actions from the earlier LERs were not adequately implemented therefore the Root Cause of this event is attributed to Management DefJclency, inadequate program. This event is being reported as a Special Report. ,

BACKGROUND The Reactor Building Cooling System [ Ells:BKl is an Engineered Safeguards (ES) System [ Ells:JE) which, in conjunction with other ES systems, provides heat removal capability for the containment atmosphere during normal and post-accident conditions. The system consists of three Reactor BuildinR Coolant Unit (RBCU) trains that function as' air-to-water heat l

exchangers, located entirely within the Reactor Building (RB) [EIIS:NG).

During. normal operation, trains "A" and "C" operate with fans in high speed while the "B" train f an is of f. Iow Pressure Service Water (LPSW)

[E!IS:BI] flow is also diverted from the "B"-RBCU to the non-safety j related Reactor Building Auxiliary Cooling (RBAC) [EIIS:BK) units, r ollowing an accident, such as a LOCA, an ES signal automatically aligns LPSW to the "B" RBCU and fans on all three RBCU trains'shif t to low speed.

The fans force containment air across RBCU cooling coils and reject RB

!. energy to the environment via the LPSV system. Following an accident, the l RBCUs help to ensure that post-IDCA conditions do not exceed the l

environmental qualifications (EQ) of RB equipment required to be operable to mitigate the consequences of a' LOCA. The station Final Safety Analysis Report Section 15.14.5 states that RB design pressures will not be exceeded following a worst-case LOCA even without any RBCU capacity. RBCU

.l operability criteria is therefore based solely upon the ability to maintain the RB environment within the EQ envelope following a LOCA and assuming a single failure of the RBCU train with the highest capacity.

I

n . .g. , %[ _

i Technical Specification 3.3.5 requires ti.at the station enter a seven day.

Limiting Condition of Operation (LCO) if less than 3 operable RBCU trains are available during power operations. While the station is in an LCO due to an inoperable RBCU train there is no requirement to consider a-single-f ailure of the worst train when determining operability of the RBC -

system.

EVENTS DESCRIPTION i

Service induced fouling of the Reactor Building Cooling Units (RBCUs) heat transfer surfaces has been identified as a~ problem at Oconee Nuclear "

Station since 1986, particularly in Unit 3. LERs 269/87-04, 287/88-03, and 287/89-01 reported events where cooler degradation caused the unit to be shutdown or resulted in the establishment of power level limitations.

The corrective cetions specified in these LERs, in part, required an increased frequency of performance testing. to ensure that fouling of the

• coolers was trended and the coolers cleaned prior to.the point where i degradation caused the RBCUs to be inoperable.

Several factors hindered station personnni 'in .their. attempts to anticipate the extent of cooler fouling and ensure that cleaning of the cooler coils was performed before degradation caused the coolers to be inoperable.

1) The rate of degradation is af fected by numerous variables- RB and LPSW temperatures, building humidity, boron concentration in the Reactor Coolant System (RCS) [EIIS:AE], and RCS' leakage rate, to name a'few. Unit 3's degradation rate is particularly difficult to characterize.

2) It is generally accepted that baron from RCS leaks is the main-

' contributor to the particulate which-is causing cooler surfaces to foul, however, the octual transport mechanism is not completely understood and therefore t he actions necessary to ,

prevent or limit the rate of fouling are not readily determined.  !

Specific corrective actions required by previous LERs are as follows:

LER 287/88-03 Perform test.ing of Unit 3 RBCUs " periodically with frequency '

adjusted as necessary."

LER 287/89-01

1) Establish a task force to determine the cause of, and provide recommendations to prevent, cooler fouling.

2) Install experimental instrumentation to allow continuous monitoring of Unit 3 RBCU status during unit operations.

3) Do performance tests of Unit 3 RBCUs "as required based on-l indication of the installed experimental instrumentation

l. and results of safety analysis performed af ter testing."

l t

.3 4

. q Also, In order to establish a broader base for cooler performance, the  :

RBCU task force decided to perform tests prior to shutdown for refueling as well as following Unit. start-up. The' data obtained from all performance testing is transmitted to the Design Engineering-(DE)

-Mechanical Oconce Systems Engineering (MOSE) group.

MOSE personnel use the data provided to calculate the heat removal i capacity of each RBCU train as well as- ;.he rate of degradation given the current lake water temperature. This information is provided to the DE Safety Analysis (DESA) group.

DESA prepares an Operability Evaluation (OE), also referred to as a safety ,

analysis in previous LERs, which is, in part, an estimate of how long the coolers will remain operable based on the current capacity and calculated degradation rate. The OE is sent to the station Compliance section for review and distribution at the station. Performance Engineer."A" (PE "A")~

and MOSE DE "A" (DE "A") agreed that the RBCUs were to be tested approximately half-way through the operable period established by DESA.

This was believed to be a suitably conservative testing frequency.

A task force was established in January,1989 with PE "A" acting as the c.hnirman. The group was comprised of members from various station organizations and DE. Between January and mid-1989 the task force met regularly and made significant contributions to the RBCU problem resolution effort.

}

The task force developed an effective cooler cleaning method and deduced 'l that the source of the fouling was baron, likely from RCS leakage, i precipitating out on cold cooler tubes. In addition, the task force was 1 able to determine that- tube side fouling was not a significant source of cooler performance degradation. Af ter June of 1989, however, the full task force no longer functioned as a viable organization; meetings were ]

inf requent and not well attended, and there was no consistency cf support  !

from the most station sections ~. A core group comprised of_ members from l the Performance, Operations, Maintennnce Engineering, and' Design _ l

. Engineering Sections still continued to work together to resolve critical l L issues. 1 i

In late February 1989, experimental Instruments (hygrometers) were in=talled in accordance with LER 287/89-01 and Temporary Modification 515 l to allow cont (nuous monitoring of RBCU performance. Results were  ;

encouraging and it nppeared that the use of the new instruments would 3 enhance the performance testing program but thn experimental Instruments  ;

failed due to the harsh service environment within approximately 8 weeks of installation. Attempts to find suitable replacement instruments l capable of withstanding the environment continued until November 1989, at l which time replacements were ordered.

On November 2 1989, prior to shutting down for refueling outage (RFO) 11 l on Unit 3, PE "A" coordinated the testing of the RBCUs. The test results

]

indicated that the coolers were quito clean, as evidenced by the heat j trans fer capacity available. (Subsequent visual inspection, performed  !

while RF0 11 cleaning of RBCUs took place, verified cooler cleanliness). Y DESA provided an Operability Evaluation (OE) to the station based on the l

i a

h

[

, , ~4~' '

test data. This evaluation. indicated that_the RBCUs would remain operable until November, 1990.

Although the performance tests indicated that the coolers were clean, Operations personnel entered the Reactor Building as part of a normal pre-shutdown tour, and attempted to identify potential sources of airborne boron. No now sources of RCS leakagn were identified.  !

During the Unit.3 RF0 the coolers were all cleaned using hot water. This-wns the cleaning method which had proven to provide the best results.

Unit 3 was started up on December 20, 1989 and, after allowing some time. l for RB and system temperature stabilization, on January.9, _1990, a  ;

performance test of Unit 3 RBCUs was performed.' The results of this test j indicated that there had been substantial fouling of all three RBCUs in j the relatively brief (approximately 20 days) period since unit startup, j l

PERCENT OF DESIGN CAPACITY- l (100 % = 80.tfillion BTU /Ur) l 1

TEST DATE "A" RBCU "B" RBCU "C" RBCU i i

11/2/89 98% 104% 92%

01/9/90 55% 71% 55 (64%) l PE "A" and DE "A" discussed the test results and determined that .!

conditions at the time of the test were sufficiently different from  !

conditions when the November 7 test was performed to have skewed the 'j results: The January test was performed with a relatively cool-and dry RB n which caused a much lower RB-to-LPSW delta temperature-(d/t). This lower d/t was believed to have the not effect of making the coolers appear to be ,!

less officient than they actually were. Another factor which caused the j test results to be questioned was the 33% degradation of the "B" cooler in l spite of the fact that it had not. been operated since it was last cicaned. ,

Current thinking presupposed that degradation occurred as a result of [ '

baron in the RB atmosphere plating out on the cold surfaces of an in-service cooler. A degraded "B" cooler implied to PE "A" and DE "A" {

that the test results were not representative of the actual status of the {

coolers. A second test performed later on January 9th, at slightly higher j RB temperat.ures, indicated a 64% capacity for the 3C RBCU. This 9%  !'

increase over the earlier test suggested that cooler performance was more sensitive to building heat load (and d/t) than was originally thought. {

PE "A" discussed the testing results and conclusions drawn with j Performance Supervisor "A" (PSA) and the Performance Manager (PM). j i

Although the OE provided to the station on November 7 indicated that Unit  !

3 RBCUs would be operable until November, 1990, and PE "A" and DE "A" both .;

believed that the January 9 test results were not representative of the  !

actual cooling capacities, it was decided to perform another set of RBCU tests in the near future to confirm that higher RB temperatures would  ;

result in more accurat.e test results. It was the consensus opinion that if the next test could be done at conditions closely approximating the November test the results would also mimic the earlier test results.

4

' 4 ,

. .5-It should-be noted that Operations (OPS) personnel were aware that a test of the RBCUs was performed in January. OPS is normally made aware of the test results'when the OE is distributed to them after being reviewed by the station Regulatory Compliance group. However, since the OE statement performed in November, 1989 indicated the coolers were expected to be operable until the following November, DE "A" and PE "A" determined that a new OE need not be performed. OPS management was not informed of the. test results and therefore had no inkling that the test indicated that degradation appeared to have occurred at a greater than expected rate. It.

should be noted that even if the degradation had occurred to the extent Indicated by the performance tests,-the coolers were still operable at this time.

On February 2, 1990, PE "A". noticed an increasing trend in RB temperatures and discussed this trend with the Unit 3 Operations Coordinator on ,

February 14, 1990. On February 16, Unit 3 Operations Manager (OM) roccived the Unit 3, EOC-11 RBCU Haintenance Technical Report from the station Maintenance Engineering Services group. This report included a

" Findings /Results" section which' stated, la reference to the January 9 r test results, that though the numbers did not indicate it, the cleaning of the-coolers during the RF0 "most likely increased the heat removal-capability of the coolers."

On February 17, 1990, Shift Supervisor "A"-(SS "A") was informed by a Senior Reactor ~0perator (SRO) that the 115 degree F. RB temperature limit-of PT/3/A/600/0, " Periodic Instrument Surveillance" had been reached and notification of OPS management was required. On Honday, the 19th,'SSA.

discussed the rising RB temperature with OM, expressed his' concerns that the coolers were fouled, and made plans to do a building entry the next day. That afternoon discussions between OPS and Performance personnel resulted in the decision being made to perform on RBCU performance test as soon as practical. A plot of RB temperatures continued to show.an increasing trend which caused unanimous concern.

On Tuesday morning, February 20, 1990, SSA and an OPS Nuclear Production Engineer (NPE) entered the RB to attempt to identify potential sources of I

RCS leaks and'to make a visual inspection of the Reactor Building Auxiliary Coolers (RBACs). No new sources were found but the inspection l

did reveal a significant amount of fouling on the RBAC coils-indicated by the almost complete lack of any air flow across the coils. Fouled RBAC coils can be on indicato of RBCU coil condition.

l On Tuesday afternoon, the 3A and 3C RBCUs were tested and-the results were verbally communicated to DE. DE "A" analyzes the test. data to determine cooling capacities. The analysis indicated that the coolers had' degraded i and had unacceptable capacities:

3A RBCU - 25.0% of Design 3C RBCU - 23.0% of Design i At 48% combined capacity, and assuming a single failure of the "B" train, cooling capacity was insufficient to allow operation at 100% reactor power. DESA determined that a minimum of 82% capacity was required from the two worst coolers to allow operation at 100% reactor power. It could not be determined when cooler degradation had progressed to the point

• =l

, 6' 4

where available capacity was insufficient to allow operation at 100%

power.

t Consequently, after discussion between OPS, Performance, and Compliance personnel, the decision was made to declare the 3C RBCU inoperable and '

take it.out of service for cleaning .and testing. Therefore, at 1452 hours0.0168 days <br />0.403 hours <br />0.0024 weeks <br />5.52486e-4 months <br />, on February 20, 1990, wit.h Unit 3 at 100 % Full Power, the "C" '

train was declared inoperable which caused Unit 3 to enter a seven day Limiting Condition of Operation (LCO) in accordance with Technical Specification 3.3.5. The "B" RBCU train was , aced in operation while cleaning and testing of the "C" train was pe:( aned. Although the "B" train capacity was not tested prior to placing it in service, it was  ;

assumed to be 71% based on the January 9th test results and the. fact thac the "B" train had not bann operated since that date. The 3B RBCU was tested later on the 20th and had an indicated capacity of 68L ,

While the station was in the LC0 there was no -' longer the requirement to assume a single failure of the best train. Instead, operation at 100%

power was dependent upon the combined capacities of the two operable trains, in this. instance, "A" and'"B". The combined cooling capacity was 25% + 68% = 93%, more than the 82% required to allow operation at 100%

reactor power. The "C" cooler was' cleaned and performance tests indicated a 61% cooling capacity and, therefore, _at 1334 hours0.0154 days <br />0.371 hours <br />0.00221 weeks <br />5.07587e-4 months <br />, on February 22, 1990, the 3C RBCU was declared operable and returned to service. Once the 3C train was cleaned and back in service the event was terminated since a combination of any two RBCU trains then had the necessary post-accident cooling capacity. Subsequent actions taken were to clean and test the "A" and "B" RBCU trains.

CONCLUSIONS The Root Cause of this event is determined to be flanagement Deficiency, inadequate program due to failure to adequately implement previously specified corrective actions. Rather than establishing formal and detailed guidelines for the accomplishment of corrective actions in order to assure that Reactor Building Cooling Units were maintained operable, the program was allowed to evolve ~ over time as testing provided new -

information. While such latitude was conducive to allowing the program to be enhanced and refined, it included limited controls or balances for

those responsible for implementation of the testing program. Given Unit ,

3's past problems and historically erratic fouling rate it is concluded that the program should have provided more guidance and should have been more closely scrutinized by management.

l' The Root Cause identified in previous LERs was Other, due to service induced fouling. The root cause of this event is flanagement Deficiency due to failure to prevent recurrence of a previously identified recurring problem. Because the problem was not corrected this event is considered to be recurring in spite of the fact that the root cause determinations are different. Corrective actions from previous reports have already been addressed in this report.

,.' "P f

?.h .

~ 5

+

,s There were no radioactive _ material releases,. radiation exposures, or personnel injuries as a result of this~ event. The event did not involve equipment failure and'is not.NPRDS reportable.

CQRRECTIVE ACTIONS i IMMEDIATE

1) . The 3C Reactor Building Cooling Unit (RBCU) was declared _

inoperable, removed from service for cleaning, and placed back j

i in operation following evaluation to ensure effectiveness of the cleaning.

2)- The "B" RBCU train was placed in service.

' SUBSEQUENT

1) The 3A and 3B RBCUs were cleaned and tested.

PhANNED '

1) The RBCU Honitoring Program will be formalized as a section in the Performance Manual to define the monitoring process. 'l

2) The Performance Section will revise the procedure used,to gather data for RBCU operability evaluations. Tho' procedure will be j used as the mechanism to control the Monitoring Program and will l include:

i

-n. The method for gathering data,

b. Controls for transmittal of test _ data to Design Engineering (DE).  :

I

c. Controls to ensure appropriate station personnel are notified of RBCU operability status.  ;

d. Limits and actions to be taken based on operability margin.

3) The RBCU Task Force will be reactivated to pursue a solution to fouling problems and seek methods to ' locate sources of the-fouling. The Task Force will disband af ter completing its 3 mission and will notify Station management, in writing, of its -1

]

Intent to disband. The Task Fa ce will oversee the following '

activities;

a. The Nuclear Performance Evaluation section will investigate  !

the feasibility of'an on-line spray system to prevent fouling of the air side of the cooling coils.

b. DE will investigate the feasibility of performing an i evaluation to show that the amount of service induced fouling which typically occurs during 1 refueling cycle would not' prevent the RBCUs from performing their intended safety function. ,

\

{

I l

4 f

.. .- . 8C WN r

.c.. The Task Force will investigate reducing the opportunity I for fouling by limiting low Pressure Service Water flow  :

through the RBCUs.

-i

~

d. The Maintenance group wil1 develop and lui ready to implement a plan to search for Reactor Coolant System leakage during unit trips and outages.

SAFETY ANALYSIS 1 f

Degraded Reactor Building Cooling Unit-(RBCU) capacity has little impact, on' peak Reactor Building (RB) temperatures' following the most severe-Design Basis (DB) accident but will slightly increase the cooldown time.

• The concern is that, immediately-following a DB LOCA, inadequate RBCU capacity might result;in environmental conditions which could adversely: ,

impact the ability of important safety systems to perform their intended function. The' operability criterion that the RB temperature profile remain within the environmental' envelope for a DB TOCA assures that safety systems remain operable. For operability determination purposes it is implicit that maintaining RB temperature.bolow the Environmental' Qualification (EQ) envelope will also ensure that design RB pressures.will ,

not be exceeded.

The Design Engineering and Performance personnel involved in performance testing and trending of cooler degradation rates have speculated that the capacities of the RBCUs would actually be much improved during a DB accident regardless.of their relative cicanliness prior to the start of the accident. This reasonirg is logically supported although as yet it.

has not been empirically proven.

The support stems f rom thn similarity between the cooler coll . cleaning, method used by the-station and RB' conditions following an. accident. The method of cooler cleaning employed by station Maintenance Support. . .

personnel is to pour approximately 200-300 gallons of hot.(almost boiling)

water over the coils .and let the water gravity drain to the RB basement.

L Experience has shown that the hot water dissolves the encrusted boron deposits and transports the soluble boron to.the building sump. .The -

effectiveness of this cleaning method has been proven by both visual examination and post-cleaning performance testing.

l E

It is believed that the RB environment will similarly act to clean the l' coolers following a DB LOCA. Design conditions following an accident result in a steam-air mixture at 286 degrees Fahrenheit (deg. F.), a RB pressure of 73 psia, a Inkn water inlet temperature of 75 dog. F., and a worst case baron concentration at start of core life.

L At 286 degrees, water is calculated to be able to hold (or dissolve) 95,883 ppm baron (which is greater than. worst case concentration by nearly a factor of 100) and it is estimated that,'following a DB LOCA, each RBCU

• l train will be condensing close to 130 gpm of RB steam-water mixture. This .

I volume of hot water pouring over the cooler tube surfaces will be

a cleaning method even more effective than the one which is currently used -

and has proven so effective. The engineering judgement of several Duke organizations supports the premise that a self-cleaning phenomena will occur _following an accident and the coolers will be able to provide the cooling necessary to ensure post-accident EQ limits are not violated, None the less, as has been stated, our current accident scenarios require

-adequate RBCU capacity be available in the event of a LOCA and,_ for some indeterminate period of time preceding February 20, 1990', the Unit 3A and 3C RBCUe had inadequate cooling capacity available. .llowever, no accidents requiring the mitigating function of the RBCUs occurred and there were no '

radiological releases associated with this event, therefore the health and safety of the public was not affected. -

?

1 l

l l

l