Resolution of TMI Action Item II.k.3.5, "Automatic Trip of Reactor Coolant Pumps" (Generic Letter No. 83-10f)

                               UNITED STATES 
                       NUCLEAR REGULATORY COMMISSION 
                          WASHINGTON, D. C. 20555 

                              February 8, 1983

TO ALL LICENSEES WITH BABCOCK & WILCOX (B&W) DESIGNED NUCLEAR STEAM SUPPLY 
     SYSTEMS (NSSSs) 

SUBJECT:  RESOLUTION OF TMI ACTION ITEM II.K.3.5, "AUTOMATIC TRIP OF REACTOR
          COOLANT PUMPS" (GENERIC LETTER NO. 83-10f) 

Gentlemen: 

The purpose of this letter is to inform you of (1) the staff's conclusions 
regarding your analysis of LOFT Test L3-6, (2) the continued acceptability 
of your ECCS evaluation model for predicting small break LOCAs with Reactor 
Coolant Pump (RCP) operation and (3) criteria for resolution of TMI Action 
Item II.K.3.5, "Automatic Trip of Reactor Coolant Pumps." 

We have completed our evaluation of your analyses of LOFT Test L3-6 and 
conclude that the evaluations acceptably predict the test results. 
Therefore, we find the currently approved B&W evaluation model for small 
break LOCAs in continued conformance with Appendix K to 10 CFR 50 for the 
case of limited RCP operation after reactor trip and for the range of 
licensed B&W reactor designs. 

We have reviewed industry analyses and performed our own analyses to 
determine whether RCP trip is necessary during LOCAs, and evaluated the 
desirability of continued RCP operation during non-LOCA transients and 
accidents, including steam generator tube ruptures. We have concluded that 
there is a wide range of transients and LOCAs where it is beneficial for the
operators to maintain forced circulation cooling and mixing through 
operation of the RCPs. However, some of the calculations show that for 
certain small break LOCAs, primarily those with only one of the two High 
Pressure Safety Injection (HPSI) Pumps assumed available, continued 
operation of the RCPs or continued operation of the RCPs followed by delayed 
RCP trip could lead to core damage. 

Some uncertainty in these conclusions remains. Specifically, there is a 
complex interrelationship among break size, break location, RCP trip delay 
time, available safety systems, and peak cladding temperature (PCT) for each
type of NSSS design. Moreover, although the staff's and each vendor's 
calculational models adequately predicted LOFT test L3-6, there appear to be
subtle differences embedded in the computer models which, when applied to 
large, commercial, PWR designs, yield differing results regarding the 
necessity for RCP trip during small LOCAs. 



8302090098 
.

                                     -2- 

Because of this, we place substantial weight on the views of the reactor 
designers and the utilities which are almost unanimous in asserting that for
some small LOCAs with less than the maximum available HPSI flow, delayed RCP
trip could lead to core damage. Some utilities indicated their preference to
keep the RCPs running for all events; however, this view appeared to be 
based solely on the desire to maintain forced circulation and did not 
consider the consequences of delayed RCP trip. 

While acknowledging the industry's general conclusion that the RCPs should 
be tripped for small LOCAs, both the staff and the industry recognized that 
there are other accident sequences of much higher probability than the small
LOCA where the absence of forced circulation makes the operator's job more 
difficult and can increase the likelihood of operator errors. For this 
reason, we believe that a balance should be struck between the competing 
risks associated with tripping the RCPs early and leaving them running 
following transient and accident events. 

Based on our discussions with both licensees and the reactor manufacturers, 
and our internal evaluations, we believe that appropriate pump trip 
setpoints can be developed by the industry that would not require RCP trip 
for those transients and accidents where forced circulation and pressurizer 
pressure control is a major aid to the operators, yet would alert the 
operators to trip the RCPs for those small LOCAs where continued operation 
or delayed trip might result in core damage. 

In summary, we have concluded that the need for RCP trip following a 
transient or accident should be determined by each licensee on a 
case-by-case basis, considering the Owners Group input. However, the staff 
must ensure that whatever decision is made regarding pump operation, it will 
result in safe, reliable operation of reactors and will not adversely affect 
the ability of licensees to comply with the Commission's rules and 
regulations. 

The enclosure to this letter provides guidance for the development of either
(1) satisfactory setpoints for RCP trip or (2) the technical bases for 
allowing continued RCP operation in the event of a small LOCA at a 
licensee's facility. As stated in the enclosure, manual tripping of the RCPs 
for a LOCA can be allowed under certain conditions. 

We recognize that possible differences exist between the requirements of 10 
CFR 50.46, which assure ample core cooling capacity, and the approaches 
described in the enclosure which are based upon assuring proper operator/ 
system response under conditions that may be faced during accidents and 
transients. Accordingly, in such cases, we will consider a request for 
exemption from specific requirements of 10 CFR SS50.46 pursuant to 10 CFR 
SS50.12. 
.

                                    - 3 - 

For the purpose of providing uniformity of setpoints and methods and for 
minimizing potential confusion that could arise because of diverse actions 
by individual licensees, we strongly urge that licensees work collectively 
with owners of similar plants (i.e., owners group) and propose setpoints and
methods consistent with other licensees. 

If a licensee elects to trip RCPs, when RCP trip setpoints are developed 
which are believed to substantially meet the guidance provided in the 
enclosure, we encourage licensees to begin implementation of these new 
setpoints at operating plant(s)*. We caution that careful judgment should be
used when developing proposed methods and setpoints in accordance with the 
guidance in the enclosure. If RCPs are to be tripped, we recommend that the 
licensees utilize event trees to systematically evaluate RCP trip setpoints 
to minimize the potential for undesirable consequences due to a misdiagnosed
event. 

Specifically, we recommend the setpoints be evaluated for events where the 
RCPs could be tripped when it is preferable they remain operational. We 
further recommend the setpoints also be evaluated for the case when the RCPs
are not tripped early in the event and for which a delayed trip may lead to 
undesirable consequences. 

We are not requiring a formal submittal of the analyses which support either
RCP trip setpoints or the decision to leave the RCPs operational for all 
events. However, once the technical bases for the decision are established, 
we intend to conduct inspections of individual licensees led by Regional 
personnel. During these inspections, we will examine the translation of the 
10 CFR 50, Appendix K, and RCP operation mode evaluations into plant 
procedures. We would expect the evaluations to include consideration of the 
guidance contained in the enclosure to this letter. Copies of these 
evaluations should be made available to the staff at these inspections. 

Alternatively, a licensee may choose to make either an individual submittal 
or reference a generic (i.e., owners group) submittal which provides the 
technical justification for treatment of RCPs during transients and 
accidents. In that case, an inspection would not be necessary. 

The requirements set forth in this letter supersede the actions required in 
IE Bulletins 79-05C and 79-06C. 

*Unless implementation entails a change to technical specifications or an 
unreviewed safety question, which require NRC approval prior to 
implementation. 
.

                                     -4- 

Accordingly, within 60 days following receipt of this letter, please provide
your plans and schedules for resolution of this issue for your facility. You
should also indicate whether you desire to make a submittal concerning this 
issue. If you cannot respond within 60 days, you should indicate within 30 
days when your schedule will be submitted. The information requested should 
be sent to Mr. D. G. Eisenhut, Director, Division of Licensing, Washington, 
D.C. 20555 pursuant to 10 CFR SS50.54(f). 

This request for information was approved by the Office of Management and 
Budget under clearance number 3150-0065 which expires May 31, 1983. Comments
on burden and duplication may be directed to the Office of Management and 
Budget, Reports Management, Room 3208, New Executive Office Building, 
Washington, D.C. 20503. 

If you believe further clarification regarding this issue is necessary or 
desirable, please contact Dr. B. Sheron (301-492-7460). 

                                   Sincerely, 


                                   Darrell G. Eisenhut, Director 
                                   Division of Licensing 

Enclosure: 
Resolution of TMI Action 
     Item II.K.3.5 

cc w/enclosure: 
Service Lists