Information Notice No. 88-36:Possible Sudden Loss Of RCS Inventory During Low Coolant Level Operation
UNITED STATES
NUCLEAR REGULATORY COMMISSION
OFFICE OF NUCLEAR REACTOR REGULATION
WASHINGTON, D.C. 20555
June 8, 1988
Information Notice No. 88-36: POSSIBLE SUDDEN LOSS OF RCS INVENTORY
DURING LOW COOLANT LEVEL OPERATION
Addressees:
All holders of operating licenses or construction permits for pressurized
water reactors (PWRs).
Purpose:
This information notice is being provided to alert addressees to the potential
for a sudden loss of reactor coolant system inventory while conducting steam
generator tube inspections and modifications with hot leg nozzle dams in
place. 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 do not constitute NRC requirements; therefore, no specific action or
written response is required.
Description of Circumstances:
During the second refueling of Diablo Canyon Unit 1, in the spring of 1988,
deficiencies in the procedures to be used during the steam generator tube
inspections were identified that could significantly increase the probability
of a sudden ejection of reactor coolant followed by core uncovery.
In order for the steam generator tubes to be inspected at Diablo Canyon, they
were drained, by drawing air through reactor and pressurizer vents, until the
reactor coolant inventory was drained down to the mid-level of the hot leg
piping (see Figure 1). Lowering the reactor coolant to this level also un-
covers the steam generator primary side manways so that they can be removed to
gain access to the steam generator hot and cold leg plenums and their re-
spective hot and cold leg nozzles. Nozzle dams are then placed in these steam
generator plenum nozzles so that the reactor coolant level can be raised to
increase the net positive suction head to the decay heat removal pumps without
refilling the steam generators.
If the hot leg nozzle dams were all installed before all of the cold leg
nozzle dams were in place, a small increase in reactor vessel pressure would
cause reactor coolant to be rapidly expelled from the open cold leg manways.
This would occur because the increased pressure, unable to vent through the
dammed-up hot legs, would force the coolant down in the vessel, through the
cold legs, and out of the manways. A pressure increase of only 2-1/2 psig in
the vessel
8806020047
. IN 88-36
June 8, 1988
Page 2 of 4
would lower the coolant level to the point where the top of the fuel would
begin to be uncovered, with the level of the remaining coolant in the open
steam generator located at the bottom of the cold leg plenum manway.
Similar mechanisms have been identified at San Onofre Units 2 and 3 in their
response to Generic Letter 87-12 (Reference 1), and by the Westinghouse Owners
Group in an ongoing analysis of reactor behavior during the shutdown
condition.
The possibility of ejecting coolant by this mechanism can be eliminated by
ensuring that a steam generator hot leg plenum manway and its associated hot
leg pipe are kept open to provide an adequate vent path whenever any cold leg
openings are made. This can be accomplished by ensuring that a hot leg manway
is the first manway to be opened, and a hot leg nozzle dam is the last dam to
be installed. In addition, not installing the last hot leg nozzle dam until a
sufficient vent path is established in the reactor vessel or pressurizer will
reduce the possibility of developing a pressure differential which could eject
a dam.
Discussion:
On April 10, 1987, the Diablo Canyon Unit 2 reactor vessel became pressurized
to approximately 7 to 10 psig when the residual heat removal flow was lost for
a period of 1-1/2 hours (Reference 2). Fortunately, during this event the
man-ways, although loosened, were still in place and the nozzle dams had not
yet been installed. Operating a reactor coolant system that has been drained
to a low level often involves unusual problems that have a significant
probability of causing a loss of residual heat removal unless special care is
taken. NUREG-1269 (Reference 3), the report of the NRC investigation into the
Diablo Canyon event, discusses a number of these problems. These include the
following:
The level which is established for draining the steam generator tubes is
frequently only slightly above the level which will provide an adequate
suction head for the residual heat removal pumps. This marginal suction
head can lead to air entrainment due to vortexing at the suction point,
which may cause a loss of pump suction.
The temporary reactor vessel level measurement system necessary for this
type of operation tends to be inaccurate because of the long lengths of
tubing normally used. The possible air entrainment and the surface level
variations due to fluid flow at this low level provide additional
mechanisms that cause error in the level measurement.
The NRC has documented many instances where residual heat removal has been
lost, because of loss of pump suction, while the plant was being operated at
reduced reactor coolant water levels. Generic Letter 87-12 (Reference 1)
lists 37 loss-of-decay-heat-removal events, occurring from 1977 to 1987, that
were attributed to inadequate reactor coolant system level. In four cases,
including the 1987 Diablo Canyon event, boiling is known to have occurred
before residual heat removal could be reestablished.
Although small vents are normally established in the reactor vessel head and
in the pressurizer before the coolant level is drained down, these are far too
. IN 88-36
June 8, 1988
Page 3 of 4
small to prevent pressurization of the reactor coolant system after the
boiling point is reached. For the recent steam generator inspection at Diablo
Canyon, which was initiated 10 days after shutdown, the reactor was producing
5 MW of decay heat. This is sufficient to produce 5 lb of steam per second,
which would require a vent area greater than 12 square inches in order to hold
the pressure rise to less than 25 psi. During the 1987 Diablo Canyon event,
the reactor, which had been shut down for seven days, reached the boiling
point about 1/2 hour after decay heat removal capability was lost. The
pressure increased to the 7-to-10-psig maximum value a short time later even
though small vents were available in the vessel head and pressurizer.
With the hot leg nozzle dams in place the pressure rise would be quite rapid.
Generation of a small amount of steam would be sufficient to produce the
partial pressure of 2-1/2 psi necessary to uncover the core by ejecting the
coolant through the open cold leg plenum manway. This amount of steam could
be produced in less than a minute. However, the actual time to produce this
pressure would depend on the time to heat the reactor coolant to the higher
boiling point and on the rate of energy deposition in the cold materials in
the upper part of the reactor vessel and, to a lesser extent, in the
pressurizer. The time required for this to occur would likely be only a few
minutes.
Loss of residual heat removal capability after the nozzle dams are installed
and before the vessel level is raised would still result in a hazardous situ-
ation, however, more time would be available for operator action before loss
of coolant occurred. The nozzle dams used at Diablo Canyon are designed to
withstand about 50 psi of differential pressure. Approximately 1/2 hour of
additional time would be available before the reactor coolant heated up to the
approximately 300� F necessary to boil at this higher pressure. However, if a
cold leg dam were to be expelled at this point, coolant ejection through the
affected steam generator manway followed by core uncovery would be very rapid.
For this reason, it is prudent to provide a means of venting the vessel with
the dams installed. At Diablo Canyon, the schedule for detensioning the
reactor vessel head was advanced so that this would be done before the reactor
was drained for the steam generator inspection. Although the pressure neces-
sary to lift the detensioned vessel head, in order to vent the vessel, is less
than the pressure required to eject the nozzle dams, this pressure is greater
than that which would be required to uncover the top of the fuel by expelling
coolant through an undammed steam generator cold leg nozzle and the associated
manway. Therefore, even with the head detensioned, the hot leg nozzles should
be left open until all cold leg openings are closed.
Generic Letter 87-12 also identified a comparable mechanism for uncovering the
core by pressurization during low coolant level operation. An opening in a
cold leg, such as one caused by the opening of a reactor coolant system pump
or a loop isolation valve (in some plants), would vent the space of the af-
fected cold leg, maintaining this space at atmospheric pressure. Any pressure
increase, such as would be caused by boiling in the reactor vessel, would be
propagated throughout the remainder of the reactor coolant system, including
both hot and cold sides of steam generator primary spaces. This differential
pressure would force the coolant levels in the vessel down while the displaced
coolant would be forced up and out of the affected cold leg opening. As with
the mechanism already discussed, only about 2 1/2 psi would be required to
. IN 88-36
June 8, 1988
Page 4 of 4
expel the water down to the top of the core with the coolant in the affected
cold leg at the level of a pump opening. Although in this case some steam
condensation may occur in the steam generators, as the 1987 Diablo Canyon
event showed, this will not prevent pressurization. Note that this mechanism,
involving coolant expulsion through a cold leg opening, does not require
plugging the steam generator nozzles. As with the previous mechanism, this
hazard might be eliminated by venting the reactor vessel through a large
opening, such as a hot leg steam generator plenum manway or pressurizer
opening, before opening the cold leg.
The loss of residual heat removal capability during low reactor coolant level
operation has proven to be a frequent occurrence; leading in several cases to
boiling in the reactor vessel. If this should occur, pressurization of the
reactor vessel can lead to sudden core uncovery by the expulsion of coolant
through any opening in the cold leg side of the reactor coolant system. This
hazard can be eliminated by providing a large vent for the reactor vessel
space before opening the cold leg.
No specific action or written response is required by this information notice.
If you have any questions about this matter, please contact one of the
technical contacts listed below or the Regional Administrator of the
appropriate regional office.
Charles E. Rossi, Director
Division of Operational Events Assessment
Office of Nuclear Reactor Regulation
Technical Contacts: Paul P. Narbut, RV
(805) 595-2354
Donald C. Kirkpatrick, NRR
(301) 492-1152
Warren Lyon, NRR
(301) 492-0891
Attachments: 1. Figure 1 - Reactor Coolant System
2. List of Recently Issued NRC Information Notices
References:
1. Generic Letter 87-12, "Loss of Residual Heat Removal While the Reactor
Coolant System is Partially Filled," July 9, 1987.
2. IN 87-23, "Loss of Decay Heat Removal During Low Reactor Coolant Level
Operation."
3. NUREG-1269, "Loss of Residual Heat Removal System, Diablo Canyon,
Unit 2," April 10, 1987.
. Attachment 2
IN 88-36
June 8, 1988
Page 1 of 1
LIST OF RECENTLY ISSUED
NRC INFORMATION NOTICES
_____________________________________________________________________________
Information Date of
Notice No._____Subject_______________________Issuance_______Issued to________
88-35 Inadequate Licensee Performed 6/3/88 All holders of OLs
Vendor Audits or CPs for nuclear
power reactors.
88-34 Nuclear Material Control 5/31/88 All holders of OLs
and Accountability of or CPs for nuclear
Non-Fuel Special Nuclear power reactors.
Material at Power Reactors
87-61, Failure of Westinghouse 5/31/88 All holders of OLs
Supplement 1 W-2-Type Circuit Breaker or CPs for nuclear
Cell Switches power reactors.
88-33 Recent Problems Involving 5/27/88 All Agreement
the Model Spec 2-T States and NRC
Radiographic Exposure licensees
Device authorized to
manufacture,
distribute or
operate radio-
graphic exposure
devices and source
changers.
88-32 Promptly Reporting to 5/25/88 All NRC material
NRC of Significant licensees.
Incidents Involving
Radioactive Material
88-31 Steam Generator Tube 5/25/88 All holders of OLs
Rupture Analysis or CPs for
Deficiency Westinghouse and
Combustion
Engineering
designed nuclear
power plants.
88-30 Target Rock Two-Stage 5/25/88 All holders of OLs
SRV Setpoint Drift or CPs for nuclear
Update power reactors.
88-29 Deficiencies in Primary 5/24/88 All holders of OLs
Containment Low-Voltage or CPs for nuclear
Electrical Penetration power reactors.
Assemblies
_____________________________________________________________________________
OL = Operating License
CP = Construction Permit
Page Last Reviewed/Updated Tuesday, March 09, 2021