Protecting People and the EnvironmentUNITED STATES NUCLEAR REGULATORY COMMISSION
GL79055
UNITED STATES
NUCLEAR REGULATORY COMMISSION
WASHINGTON, D. C. 20555
TO ALL PRESSURIZED WATER REACTOR LICENSEES
SUBJECT: SUMMARY OF MEETING HELD ON OCTOBER 12, 1979 TO DISCUSS RESPONSES
TO IE BULLETINS 79-05C AND 79-06C AND HIGH PRESSURE INJECTION
(HPI) TERMINATION CRITERIA
On October 12, 1979, members of the NRC staff (the staff) met with the three
pressurized water reactor (PWR) vendors and representatives of the
corresponding Owners' Groups in Bethesda, Maryland. The purpose of the
meeting was to discuss the variations in analyses and operator guidelines
submitted in response to IE Bulletins 79-05C and 79-06C and to discuss
differences in proposed HPI termination criteria. A list of attendees is
provided as Enclosure 1.
BACKGROUND
The initial review of the preliminary chronology of events which occurred at
the Three Mile Island Nuclear Station, Unit 2 (TMI-2) led to the
identification of six potential human, design, and mechanical failures which
appeared to have led to core damage in the reactor. One of the identified
items was the tripping of the reactor coolant pumps (RCPs) during the course
of the accident. Once forced flow was terminated by securing the RCPs, core
damage occurred since the steam and water separated in the primary system
and caused the core to become uncovered. Instructions were issued to the
holders of operating licenses, in the form of IE bulletins, which required
operating procedures to specify that in the event of HPI initiation with
RCPs operating, forced flow would be maintained in the RCS.
Subsequent analyses by each of the Nuclear Steam Supply System (NSSS)
vendors indicated that for a certain spectrum of small breaks in the RCS,
continued operation of the RCPs would increase the mass lost through the
break and prolong or aggravate the uncovering of the reactor core if the
RCPs were subsequently tripped at certain times into the accident. While the
size and location of such a break and the time of tripping the RCPs, which
would produce unacceptable results, varied for each analysis, all three
vendors agreed that an acceptable action under loss-of-coolant-accident
(LOCA) symptoms would be to trip all operating RCPs immediately, before
significant voiding in the RCS occurred. As a result of this information, IE
Bulletins 79-05C and 79-06C were issued which required immediate tripping of
all operating RCPs upon reactor trip and initiation of HPI caused by system
low pressure. These bulletins also required the licensees to perform more
detailed analyses on the effect of RCP trip for both LOCA and non-LOCA
transients.
The conclusions reached by the PWR vendors in these analyses vary to a
considerable degree as summarized in Enclosure 2. It is the NCR staff's
present judgment that the major differences in results are attributable to
model differences, which are highlighted in Enclosure 3. The various RCP
trip criteria that have been proposed by the vendors are shown in Enclosure
4.
.
- 2 -
An additional concern relates to the criteria for HPI termination. Enclosure
5 lists the criteria proposed by the various vendors. The staff can see no
reason why the same set of safety considerations would lead to more than one
uniform criterion for termination of HPI.
This meeting was requested by the staff to discuss, in an administrative
sense, how the PWR regulated industry can reach agreement on these two
issues.
DISCUSSION
The NRC staff opened the meeting by stating that its goal for the
discussions was to attempt to reach agreement among the staff, the Owners'
Groups, and the NSSS vendors on several issues involving small break LOCA
analyses and RCP trip criteria, as well as the criteria which should be used
for termination of HPI. The staff pointed out that NUREG-0578 ("TMI-2
Lessons Learned Task Force Status Report and Short-Term Recommendations" -
July 1979) requires that the implementation of small break LOCA emergency
procedures and the associated operator training be completed by December 31,
1979. Therefore, the staff considers timely resolution of these issues to be
imperative.
1. SMALL BREAK LOCA AND RCP TRIP REQUIREMENTS
There was general agreement among the participants that: (1) there are
significant differences in the small break LOCA models used by the
vendors; (2) there are differences of opinion on what model features
are conservative; and (3) none of the models have been demonstrated to
be overall conservative by integral test comparison. The staff said it
has done some calculations with RELAP-4/Mod 7, as summarized in
Enclosure 6; however, these calculations are also not definitive enough
to enable the staff to determine which set of model assumptions is
suitably conservative. Based on the above conditions, the staff pointed
out that it is not surprising that the conclusions reached by each
vendor in their small break LOCA calculations vary with those reached
by the other vendors. The staff stated, it believes that the different
results obtained by the vendors cannot be attributed solely to physical
plant differences. The other participants, however, were not able to
reach the same conclusion.
All participants were able to agree that resolution of differences is
needed in order to proceed with the task of writing/rewriting the small
break LOCA operator guidelines. The staff pointed out that it is very
important for the licensed operators to be convinced that tripping of
RCPs is the proper course of action for a small break LOCA, and in
order to achieve this, there should be general agreement among the
vendors, supported by analyses, on the criteria to be used to formulate
the licensees' emergency procedures.
Westinghouse (W) stated that the resolution of this issue was on its
critical path for completing the requirements of NUREG-0578 (Section
2.1.9) on schedule. Combustion Engineering (CE) stated that it was not
a critical issue at this time; however, continued inability to reach a
common agreement would impact its schedule. Babcock & Wilcox (B&W) has
completed its small break LOCA guidelines, detailed emergency
procedures, and operator training. B&W's concern was centered around
not confusing the operator. B&W does not want to modify
.
- 3 -
its guidelines and subsequent licensees' emergency procedures and
training based upon another interim position. B&W would like to resolve
the issue in a timely fashion, and if changes are needed in the
guidelines, procedures, and training, the changes should not be made
until the issue is resolved.
The next item discussed was the need for an automatic RCP trip for
small break LOCAs. B&W desires that an automatic trip design be
approved by the staff based on its recent proposed design submittal.
The B&W design uses coincident signals from: (a) safety injection
initiation (1600 psig) and (b) RCP current. W desired to have a
detailed technical meeting in the near future to discuss its proposal
of using coincident signals from: (a) safety injection initiation and
(b) low primary system pressure of 1250 psig. CE has not decided about
the need for an automatic trip. CE's analysis shows that in most cases
it would require only two out of four RCPs to be tripped. In addition,
the CE analysis shows that no RCP trip would be required provided at
least two HPI pumps were used during the accident.
The NRC staff questioned whether the requirements of IE Bulletins
79-05C and 79-06C to "immediately" trip all operating RCPs upon a
reactor trip and HPI initiation caused by low pressure is too
restrictive. Agreement was reached by the participants that the
requirement to immediately trip the RCPs should be modified such that
additional time is allowed for the operator to determine if the
transient which causes the low pressure is a LOCA or non-LOCA event. It
was generally agreed that the reduced pressure for the RCP trip should
fall between the setpoint of the steam generator relief valves and the
safety injection actuation setpoint. B&W stated that while they agreed
that a lower pressure than the 1600 psig used for safety injection
actuation could be acceptable, the reasons it had chosen its coincident
signals was because the B&W plants do not have a safety-grade pressure
transmitter that would assure a reliable signal, at a lower pressure,
from which to trip the RCPs. CE modified its response by stating that
it also agreed with the concept of using a lower pressure; however, it
still felt that a certain time delay (after reaching the specified
pressure limit) should be afforded the operator prior to tripping the
RCPs to allow for operator actions to recover from the transient, if
appropriate.
This topic was closed out by a restatement by the staff that it was
critical for the industry to reach agreement on the subject to assure
that the licensed operators understood that tripping of the RCPs and
the criteria for doing so do not represent an arbitrary NRC decision,
but in fact are based on the recommendation of the NSSS vendors and
have the endorsement of the licensees.
2. HIGH PRESSURE INJECTION TERMINATION CRITERIA
The basic thrust of this portion of the meeting was to come to
agreement on the necessity to pursue a common set of criteria for the
manual termination of HPI following automatic initiation caused by
reactor coolant system (RCS) low pressure.
.
- 4 -
As a result of the investigation of the TMI-2 accident, one of the
prime contributors to the damage which occurred to the core was
attributed to the premature termination of HPI. Shortly after the
accident, IE Bulletins 79-05A, 79-06A, and 79-06B (applicable to B&W,
W, and CE operating plants respectively) were issued which directed
licensees to take certain actions which would minimize the chances of
such an accident. One of the action items of these bulletins dealt with
HPI termination criteria. In general, the bulletins required, as an
interim measure, that if HPI automatically initiated because of a low
pressure condition in the RCS, it must remain operating until one of
two criteria is met. These criteria are:
a. Both low pressure injection (LPI) pumps are in operation and
flowing for 20 minutes or longer at a flow rate which would assure
stable plant behavior or
b. The HPI system has been in operation for 20 minutes, and all hot
leg and cold leg temperatures are at least 50F below the
saturation temperature for the existing RCS pressure. If the
50F subcooling cannot be maintained after HPI cutoff, HPI
shall be reactivated. The degree of subcooling and the length of
time HPI is in operation shall be limited by the
pressure/temperature considerations for vessel integrity.
Subsequent to these bulletins and based on vendor analyses, vendor
guidelines were developed independently for HPI termination criteria.
The various criteria are summarized in Enclosure 5. The staff stated
its position that there should only be one set of criteria for HPI
termination and that it should be equally applicable to all PWR
facilities. All participants agreed on this position. The staff's goal
in this area is to get one set of reliable criteria which can be easily
monitored and will not cause operator confusion.
The mechanics of getting agreement on the criteria to be used was then
discussed. It was agreed that the simplest set of criteria which would
still satisfy all concerns would be the logical choice. The criteria
selected would have to hold for all cases; i.e., could not lead the
operator to erroneous conclusions concerning the status of the RCS
inventory and degree of subcooling. The criteria would also have to
assure that the influence, if any, on pressure vessel integrity would
be considered. All participants, including the staff, agreed that these
criteria do not need to include a requirement that the HPI system be
operated for any arbitrary length of time, such as the 20-minute
requirement specified in the IE bulletins.
CONCLUSIONS
All participants agreed that, based on the information presently available
to the staff, no clear solution to establishing a unified set of criteria
for RCP trip during small break LOCAs and HPI termination is readily
apparent. An indepth technical meeting with the W Owners' Group to discuss
its analysis and criteria associated with these issues will be scheduled
shortly. A similar meeting will be arranged with the CE Owners' Group if it
is necessary. A meeting with B&W does not appear necessary at this time.
Following the technical meeting(s), it may be
.
- 5 -
necessary to have another administrative meeting with the Chairmen of the
Owners' Groups and the NSSS vendors to obtain a decision on criteria
selection.
The criteria selected by the staff should have full vendor and licensee
endorsement. These criteria will most likely be transmitted to the Chairmen
of the Owners' Groups for comment followed by individual letters being sent
to the licensees.
R. A. Capra, B&W Project Manager
Project Management Group
Bulletins & Orders Task Force
Enclosures:
1. List of Attendees
2. Summary of Conclusions on IE Bulletins 79-05C and 79-06C
3. Model Differences for SBLOCA Analyses
4. RCP Trip Criteria
5. HPI Termination Criteria
6. Summary of Staff Calculations
