Information Notice No. 94-62: Operational Experience on Steam Generator Tube Leaks and Tube Ruptures

UNITED STATES
NUCLEAR REGULATORY COMMISSION
OFFICE OF NUCLEAR REACTOR REGULATION
WASHINGTON, D.C.  20555

August 30, 1994


NRC INFORMATION NOTICE 94-62:  OPERATIONAL EXPERIENCE ON STEAM GENERATOR TUBE
                               LEAKS AND TUBE RUPTURES


Addressees

All holders of operating licenses or construction permits for pressurized
water reactors.

Purpose

The U.S. Nuclear Regulatory Commission (NRC) is issuing this information
notice (IN) to inform addressees of recent operational experience with steam
generator tube leaks and tube ruptures.  It is expected that recipients will
review the information for applicability to their facilities and consider
actions, as appropriate, to avoid similar problems.  However, suggestions
contained in this information notice are not NRC requirements; therefore, no
specific action or written response is required.

Background

The NRC staff reviewed events at pressurized-water reactors that involve tube
leaks or tube ruptures and determined that those events provide significant
operational experience regarding the handling of such events.  A summary of
the more significant events and the licensee actions taken in response to
those events follows.

Description of Circumstances

Braidwood Unit 1

On October 23, 1993, at 5:45 a.m., operators at Braidwood Nuclear Station
Unit 1, received indications of a primary-to-secondary coolant leak in the
1C steam generator.  The indications included:  (1) a 1C main steamline area
radiation monitor alert alarm and (2) increases in other secondary-side
radiation monitors such as the steam generator blowdown and the steam jet air
ejector exhaust radiation monitors.  At 6:45 a.m., chemistry samples also
showed increases in secondary-side radionuclide activity levels.  About
10:30 a.m., the leak rate was estimated to be about 863 liters [228 gallons]
per day.  At this time, the operations manager imposed an administrative leak
rate limit of 1136 liters [300 gallons] per day.  About 3:00 p.m., the leak
rate was determined to be between 1060 and 1211 liters [280 and 320 gallons]
per day and the licensee decided to shut down the reactor.

At 4:00 p.m., the licensee began shutting down the reactor at 5 percent per
hour, 1 megawatt per minute.  The main generator was taken off line at
10:52 a.m. the next day.  A subsequent inspection found that the leak was

9408240164.                                                              IN
94-62                                                               August
30, 1994                                                               Page 2
of 5


from a 3.3 centimeter [1.3 inch] long crack in a single tube located above the
top tube support plate near an anti-vibration bar.

Palo Verde Unit 2

On March 14, 1993, at Palo Verde Nuclear Generating Station Unit 2, a tube in
the No. 2 steam generator ruptured causing a primary-to-secondary leak of
approximately 900 liters [240 gallons] per minute.  Plant operators used the
emergency operating procedures to diagnose and mitigate the event but twice
failed to diagnose a tube rupture because the radiation monitors that would
have led to that diagnosis were not in an alarm status when the applicable
step in the procedure was reached.  As a consequence, the operators did not
isolate the affected steam generator until almost 3 hours after the rupture
occurred.  After the event, the licensee revised the procedures for diagnosing
a tube rupture and reviewed the circumstances preceding the tube rupture.

Prior to the March 14, 1993, event, the licensee used several methods to
estimate primary-to-secondary leakage.  The most commonly used method was
based on radionuclide activity in samples from the steam generator blowdown
lines.  On December 10, 1992, the manufacturer of the steam generators had
informed the licensee that, because of feedwater "spillover" in the steam
generator, the blowdown samples may be diluted by a factor of 5 to 10.
However, the licensee did not change the primary-to-secondary leak rate
procedures to reflect this information at that time.  After the tube rupture,
the licensee analyzed data from other radiation monitors and from chemistry
samples to better determine the actual leak rates that preceded the rupture.
Based on data from the condenser vacuum exhaust radiation monitor and a
xenon-133 gas grab sample, the licensee determined that the leak rate for the
faulted steam generator had spiked to approximately 400 liters [105 gallons]
per day on March 4, gradually decreased, and stabilized at approximately
76 liters [20 gallons] per day two days later.  The NRC staff reviewed the
data and concluded that, by estimating the leak rate based on samples from the
blowdown line, the licensee had significantly underestimated the leakage that
occurred before the tube rupture (see NRC IN 93-56 and 94-43).

Arkansas Unit 2

On March 9, 1992, at 12:30 p.m., operators at Arkansas Nuclear One Unit 2
received an alarm from the condenser pump vacuum discharge radiation monitor
indicating a primary-to-secondary leak.  The leak was estimated to be about
1360 liters [360 gallons] per day; approximately half the limit in the
technical specifications.  The estimated leak rate was confirmed by three
methods, argon and tritium sampling and the reactor coolant system inventory.
At 7:00 p.m., about 6 hours after the leak was initially detected, the
operators began to shut down the reactor and, at 8:21 p.m., the unit was taken
offline [see NRC IN 92-80 and Licensee Event Report (LER) 92-002].

On the morning of January 16, 1992, operators at the McGuire Nuclear Station
Unit 1 (McGuire) determined that the primary-to-secondary leak rate in steam
generator 1D was 83 liters [22 gallons] per day.  At approximately 4:00 p.m.,
the condenser air ejector and the steam generator blowdown radiation monitors
indicated that a Trip 2 setpoint had been reached.  That setpoint caused the
.                                                              IN 94-62
August 30, 1994
Page 3 of 5


operators to take actions regarding the leakage and to increase the sampling
frequency of the secondary coolant.  Samples taken at 5:47 p.m. and 6:35 p.m.
indicated that the leakage had increased to about 890 liters [235 gallons] per
day.  At 7:12 p.m., the licensee began a controlled shutdown of the reactor
and Unit 1 was taken off line at 6:49 a.m. January 17, 1992.  The licensee
found that the leakage in steam generator 1D was coming primarily from a
2.5 centimeter [1 inch] axial crack in tube 47-46 with a lesser amount from a
leaking sleeve-to-tube joint on tube 36-30.  The leakage from these sources
had increased from 83 liters [22 gallons] per day to about 890 liters
[235 gallons] per day in less than 19 hours.  McGuire Unit 1 has had
additional sleeve-to-tube joint leaks since the above event (see NRC IN 94-05
and LER 92-01).

Indian Point Unit 3

On October 19, 1988, at Indian Point Unit 3, the primary-to-secondary leak
rate increased rapidly from no indication of leakage to 7.6 liters [2 gallons]
per minute in 2 hours.  This amount of leakage was about 7 times greater than
the technical specification limit.  The leakage was attributed to a crack
found just above the uppermost tube support plate which extended about 250
around the tube circumference.  After this event, the licensee made several
improvements to the radiation monitoring equipment and the leak rate
monitoring procedures (see NRC IN 88-99).

North Anna Unit 1

On July 15, 1987, at North Anna Power Station Unit 1, a tube in the C steam
generator ruptured as a result of high cycle fatigue.  Denting of the tube at
the uppermost tube support plate was determined to be a contributing cause of
the tube failure.  Earlier, on July 14, the licensee had declared the steam
jet air ejector exhaust radiation monitor to be inoperable because of erratic
activity readings and had begun taking chemistry grab samples every 8 to 12
hours as required by the technical specifications.  Therefore, the operators
had no indication from the instrumentation normally used to quantify
primary-to-secondary leakage.  There were indications of increasing leakage
from other radiation monitors but these were not used to quantify leakage and
were set to alarm so as to ensure that technical specification release limits
would not be exceeded.  Consequently, the licensee was not fully aware of the
increasing leakage until minutes before the tube ruptured.

A subsequent review by the licensee of data from the air ejector radiation
monitor and from the chemistry grab samples showed that primary-to-secondary
leakage was present and increasing for a significant period of time prior to
the tube rupture.  The licensee calculated that the mean estimated leak rate
exceeded 380 liters [100 gallons] per day about 19 hours before the tube
rupture and was greater than 1900 liters [500 gallons] per day about 6 hours
before the tube rupture.  The licensee took several actions to ensure that, in
the future, similar precursor leakage would be detected and monitored so that
the plant could be shut down before a gross tube rupture could occur.  These
actions included:  (1) setting the air ejector monitor to alarm if a large
step increase in estimated leakage occurs, (2) increasing the frequency of
estimating primary-to-secondary leakage, and (3) installing N-16 radiation
.                                                              IN 94-62
August 30, 1994
Page 4 of 5


monitors to alarm consistent with the air ejector monitor and also at two
lower administrative levels to detect any initial crack propagation
(see NRC Bulletin 88-02).

Discussion

For some of these events, the response of the operators to shut down the
reactor and isolate the affected steam generator limited the contamination of
the secondary coolant and may have avoided actual tube ruptures.  In other
events, delays in detecting excessive tube leakage or in shutting down the
reactor and isolating the steam generator allowed increased contamination of
secondary coolant and increased the potential for the tube leak to become a
tube rupture.  Other events show how quickly very low leak rates can increase
well beyond technical specification limits.

Leak rate monitoring programs are important to minimizing the frequency of
steam generator tube ruptures.  The effectiveness of these programs depends,
in part, on their ability to detect, quantify, trend, and respond to the
primary-to-secondary leakage under various operating conditions.  Leak rate
monitoring programs are most effective when they provide, as close as
possible, real time information on leak rates and changes in leak rates.  At
some sites, data from the air ejector radiation monitors is continuously
displayed in the control room.  At other sites, main steamline radiation
monitors promptly detect increases in nitrogen-16 activity.  When combined
with appropriate alarm setpoints and operational limits, this information can
quickly alert operators to implement response procedures to monitor increases
in leak rates or to shut down the reactor and isolate the affected steam
generator.  Response procedures that provide clear guidance to operators
regarding rapidly increasing leak rates and leakage limits are important in
minimizing the potential for tube leaks to become tube ruptures.  The NRC has
issued several generic communications on primary-to-secondary leakage
monitoring.

Related Generic Communications

NRC Information Notice 94-43, "Determination of Primary-to-Secondary
Steam Generator Leak Rate," June 10, 1994.

NRC Information Notice 94-05, "Potential Failure of Steam Generator
Tubes Sleeved With Kinetically Welded Sleeves," January 19, 1994.

NRC Information Notice 93-56, "Weakness in Emergency Operating
Procedures Found as Result of Steam Generator Tube Rupture,"
July 22, 1993.

NRC Information Notice 93-52, "Draft NUREG-1477, 'Voltage-Based Interim
Plugging Criteria for Steam Generator Tubes'," July 14, 1993.

NRC Information Notice 92-80, "Operation With Steam Generator Tubes
Seriously Degraded," December 7, 1992.

.                                                              IN 94-62
                                                              August 30, 1994
                                                              Page 5 of 5


NRC Information Notice 91-43, "Recent Incidents Involving Rapid
Increases in Primary-to-Secondary Leak Rate," July 5, 1991.

NRC Information Notice 90-49, "Stress Corrosion Cracking in PWR Steam
Generator Tubes," August 6, 1990.

NRC Information Notice 88-99, "Detection and Monitoring of Sudden and/or
Rapidly Increasing Primary-to-Secondary Leakage," December 20, 1988.

NRC Bulletin 88-02, "Rapidly Propagating Cracks In Steam Generator
Tubes," February 5, 1988.

This information notice requires no specific action or written response.  If
you have any questions about the information in this notice, please contact
one of the technical contacts listed below or the appropriate Office of
Nuclear Reactor Regulation (NRR) project manager.

  /s/'d by CIGrimes/for


                                      Brian K. Grimes, Director
                                      Division of Operating Reactor Support
                                      Office of Nuclear Reactor Regulation

Technical contacts:  Ted Sullivan, NRR
                     (301) 504-3266

                     Joseph Birmingham, NRR
                     (301) 504-2829

Attachment:
List of Recently Issued NRC Information Notices