ML20028A221
| ML20028A221 | |
| Person / Time | |
|---|---|
| Site: | Sequoyah, San Onofre, Yankee Rowe, La Crosse, 05000000 |
| Issue date: | 10/08/1982 |
| From: | Holahan G Office of Nuclear Reactor Regulation |
| To: | Phillips L Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML20027A691 | List: |
| References | |
| FOIA-82-411, RTR-NUREG-0578, RTR-NUREG-578 NUDOCS 8211170045 | |
| Download: ML20028A221 (7) | |
Text
e D (Lprs:T
\\,
i,.m
- v' UNITED STATES
[',g' j
NUCLEAR REGULATORY COMMISSION g
. b j.a wash rucTON, D. C. 20555 MEMORANDUM FOR:
L. Phillips, Section Leader. Thermal Hydraulic Section, Core Performance Branch, Division of Systems Integration, NRR FROM:
G. M. Holahan, Safety Program Evaluation Branch, Division of Safety Technolgy, NRR
SUBJECT:
REVIEW OF ADDITIONAL INSTRUMENTATION FOR DETECTION OF INADEQUATE CORE COOLING (NUREG-0578, ITEM 2.1.3(b)) AND RECOMMENDATIONS FOR LICENSING ACTION INTRODUCTION The purpose of this memorandum is to summarize the Analysis Branch review of the utility company responses to the short term lessons learned Item 2.1.3(b), and to provide recommendations for follow up action by groups in the new NRR organi-zation. The lessons learned requirement calls for an unambiguous indication of inadequate core cooling. This function can be provided by existing instrumenta-tion; through modifications to existing instrumentation; or through the addition of new instrumentation. The Westinghouse and Combustion Engineering Owners Group have proposed addi*ional instrumentation to measure water level in the reactor
(
vessel.
The General Electric and Babcock and Wilcox Owners Group have indicated that existing instrumentation (including the newly added subcooling meter for PWR's) is sufficient. The following section presents a review of each of the owners groups positions and staff recomendations for follow up actions.'
Westinghouse Plant Owners Group Proposal The Westinghouse Owners Group has proposed to install a system of reactor vessel pressure drop measurement to be used in combination with the existing core exit thermocouples and the subcooling meter. Differential pressure would be measured between the top of the reactor vessel and the bottom of the reactor vessel on two narrow range.and two wide range instruments. The system is intended to function as follows: with the reactor coolant pumps off, the pressure drop between the top and the bottom of the vessel would indicate the collapsed liquid level (the equivalent liquid level without voids in the two-phase region) in the vessel. This would be read on the narrow range instrument in terms of feet of liquid. With the reactor coolant pumps running, the pressure drop from the bottom to the top of the vessel would provide an approximate indication of the void fraction in the vessel. This would be read on the wide range instrument as percent of full flow AP with the vessel filled with water.
In order to determine the acceptability of this system, as a means of indicating ~
inadequate core cooling, we have reviewed LOFT and Semiscale data on vessel-(
differential pressure during large and small loss of coolant accidents (L-1-4,
' g i g 45 82'100s CONNOR B2-411 PDR
2-L..Phillips L-3-1 L-3-2, S-07-10, S-06-4 S-SB-Plc. 'S-SB-P7, S-SB-2A) and have also reviewed' calculations of vessel differential pressure during these type events. All of the tests and calculations indicate that vessel differential pressure is very sensitive to void fonnation in the vessel. This is true of the cases with the reactor coolant pumps running as well as those cases with the reactor coolant pumps tripped.
In each case the vessel differential pressure provided a more sensitive indication of a loss of coolant than, system pressure, core tegera-ture or pump current. This is, in part, due to the fact that there were all cold leg breaks and a core flow reversal occurred early in the event, for the larger break cases.
In each of these cases the core uncovery and heatup occurred during a time of mimimum vessel differential pressure, and the value of differential pressure at the time of recovery was approximately the same as that at the time of uncovery. However, the value of differential pressure at the time of core uncovery varied from case to case. This is because core uncovery and heatup is associated with the.two-phase mixture level, not the col -
lapsed liquid level (which corresponds to the measured different~ial pressure).
The extent.of the two-phase mixture region depends on the core inlet tempera.'
tures, system pressure, and core heat flux. Semiscale test S-07-10 (a 10%
cold leg break) provided an example. At 450 seconds into the event, accunulator injection occurred and the heater rods were quenched. At that time the collapsed liquid level was just above the core mid-plane but the two-phase mixture was clearly covering the core. In order to correlate vessel differential pressure to inadequate core cooling some estimate of the mixture level is required. 'An acceptable estimate of mixture level can be made using a sigle model, such as the Wilson Bubble rise model, with pressure, temperature and power as input.
(
The tests and the calculations also indicate that core uncovery and heat-up could last for a significant period of time (several minutes) for Loss-of-Coolant Acci-dents which are proceeding normally, that is, those cases in which the safety systems will mitigate the consequences of the event without additional operator action. A71 of the findings lead to the following conclusions relative to the proposed level measurement system:
1.
The system can provide an indication of void formation in the reactor vessel (with or without the reactor coolant pumps operating).
2.
The correlation of vessel differential pressure to inadequate core cooling,
requires the addition of an estimate of two-phase mixture level.
3.
A method of recording vessel differential pressure is needed to assist in event diagnosis (a one inch per minute strip chart, for exagle). :_-
4.
The proposed display (one narrow range and one wide range meter p r channel) is not sufficient to provide the unambiguous indication which is called for in NUREG-0578.
5.
The core exit thermocouples are needed to complement the vessel differential pressure and therefore should be designed and qualified for use during post accident conditions.
3
- s
~ b ^ '
'U
(
7_.
m-i:r. :_.f. : m k h*f.h $::
_.. ?.D ~.
__..m e
L. Phil;ips,
[ 6.
The core exit thermocouple temperatures', the collapsed liquid level and the estimated mixture level should be integrated into an easy-to-interpret dis-Play.
~
7.
We strongly encourage the development of a standard display and standard terminology to be used for instrumentation to detect inadequate core cooling.
8.
Procedures and training in the use of this system are required.
_ Yankee-Rowe and San Onofre 1 These two Westinghouse designed plants cannot easily incorporate the proposed system for measuring vessel differential pressure since the reactor vessels for I
these plants do not have penetrations in the lower plenum.
These plants are l
considering the use of the source range neutron detectors or dif.ferential pres-3 sure at other locations in the reactor cooTant system. These licensees should.
be asked to consider the use of the Combusti~on Engineering proposed system (dis-cussed later in this report).
The staff should then meet with these licensees and establish a reasonable compromise position in which the best available system not requiring major, structural changes will be implemented.
Sequoyah TVA has proposed to install the Westinghouse differential pressure system in addition to a measurement of differential pressure between the top of the reactor
( vessel head and the bottom of the hot legs.
This additional instrumentation will increase the accuracy of the level measurement in the upper plenum region and is therefore considered a desirable addition to the Westinghouse system.
Combustion Engineering Plant Owners Group Proposal Combustion Engineering evaluated the advantages and disadvantages of eight dif-ferent concepts for measuring level in the reactor vessel. The following systems were evaluated: Heated Junction Thermocouples; Floating Source; Fixed Neutron Source and Detector; Floating Dip Stick-Floating Spheres; Ultrasonic Probe; Buoyant Force Transmitter; and External Standpipes with Float Sensor, or DP Cell The CE recommended design is based on the use of heated jenction thermocouples.
The heated junction thermocouples measure the change in thermocouple output vol-tage as a result of the difference in the thermal conductive properties between steam and water. A series of heated junction thermocouples will be located at different axial positions above the core. Each discrete location would have two thermocouples connected in series but in electrical opposition to each other.
One of these thermocouples will be heated, thereby establishing a reference differential voltage output which is a function of the heater input. When the water surrounding the heated junction thermocouple is replaced by steam or air the voltage generated by the heated thermocouple will change because the heat generated by the heated thermocouple all changa haranca tha haat canaratari will b
f no longer be removed from the area, by the water. Thus the temperature of the I
T I
L. Phillips '
heated thermocouple will increase relative to the unheated thermocouple. This
( change will be used to determine the steam-water interface in the reactor vessel.
The concept of using heated thermocouples to measure liquid level is well estab-lished 'and has been successfully used in research facilities and non-nuclear -
industrial applications. However, the description of the proposed system does not indicate what the sensitivity of the system will be.
It is not clear whether the system will distinguish between water and a two-phase mixture of steam and water or between a two-phase mixture and single phase steam. The level separating the two-phase region from the single phase steam region is important since it is this transition which generally determines the adequately cooled region from the inadequately cooled region.
If it was accurately monitored, the transition from single phase water to two-phase steam and watar would provide an early warning of possible inadequate core cooling. The following conclusions are based on our review of the proposed system as described in the December 21, 1979 document "CE Post TMI Evaluation, Task 2. Conceptual Design for a Reactor Vessel Level Monitoring System."
1.
The system sensitivity to changes in the coolant condition needs to be addressed; and the expected response to Small Break and Large Break LOCA's needs to be evaluated by CE.
2.
The range of the instrumentation is described as, "from the top of the reactor vessel head to the bottom of the reactor vessel outlet nozzle." This range is not adequate.
We prefer a range extending down to the bottom of the core, but recognize that this is beyond the present state-of-the-art for heated
(
junction thermocouples.
However, the range must be extended down to the ele-vation of the upper core alignment plate.- This is extremely important since the egion above the bottom of the outlet nozzles is expected to remain uncovered for all large break LOCA's and some small break LOCA's, while the region between the top of the core and the bottom of the outlet nozzle will be recovered when the system stabilizes. This would be an important indica-tion that the accident has been successfully terminated.
3.
The core exit thermocouples are needed to complement the heated thermocouples and should therefore be qualified for use during post accident conditions.
4.
The proposed display (continuous monitoring and recording of the signal from l
each level sensor) is not sufficient.
5.
The core exit thermocouples and the level sensors should be integrated into an easy-to-interpret display.
6.
We strongly encourage the development of a standard display and standard terminology to be used for instrumentation to detect inadequate core required.
7.
Procedures ~ and training in the use of this system are required.
~
", t.,Phillips.
8.
All instrumentation should be qualified to the same standard used for the
(
Reactor Protection System at the time the plant was licensed.
Babcock and Wilcox Plant Owners Group Proposal The Babcock and Wilcox Plant Owners Group has submitted information intended to demonstrate that the core exit thermocouples and the system pressure taken together are sufficient to meet our requirements for an unambiguous indication of inadequate core cooling. The B&W document, " Analysis Surniary in Support of Inadequate Core Cooling Guidelines for a Loss of RCS Inventory" presents calculated core exit thermocouple temperatures corresponding to various clad temperatures.
We agree that in general the core exit thermocouples will provide an indication of the existence of inadequate core cooling. However, the core exit thermocouples will provide virtually no advanced warning of the potential for inadequate core cooling, as required in our October 30, 1979 clarification letter. The inten-tion of the instrumentation for detection inadequate core cooling is to provide the operators with a continuous indication of system condition 'and of the effec-tiveness o.f the safety systems.
In addition, the early warning of the approach of inadequate core cooling is necessary for proper planning of operator actiori to respond to inadequate core cooling. The following conclusions apply to the Babcock and Wilcox Plant Nners Group proposal.
V 1.
The proposal to' rely only on core exit temperatures and systems pressure is unacceptable.
2.
The B&W plant owners should be ordered, if necessary, to provide a system for (C
monitoring reactor coolant level in the reactor vessel or an alternate system for providing an early indication of the potential for inadequate core cooling.
3.
Since the core exit thermocouples are likely to be included as a portion of the final system, they should be qualified for post accident conditions.
4.
All instrumentation should be qualified to the same standards used for the Reactor Protection System at the time the plant was licensed.
General Electric Plant Owners Group Proposal The General Electric Plant Owners Group has proposed to use the presently installed instrumentation on BWR's to detect inadequate core cooling. The subject of instru-mentation to be used during small break accidents was reviewed extensively by the Bulletins and Orders Task Force and is discussed in great detail in Section 2.17 of NUREG-0626. " General Evaluation of Feedwater Transients and Smallbreak Loss-of-Coolant Accidents in GE-Designed Operating Plants and Near-Term Operating License Applications" (Section 2.17 is attached for your convenience).
The following coelusions are based on the Bulletin and Orders Task Force review and reconsnendations plus an additional review to assure consistency among vendors and consistency with the on-going work on Regulatory Guide 1.97, " Instrumentation j
for Light-Water-Cooled Nuclear Power Plants to Assess Plant Conditions During g
and Following an Accident."
--r--
7
L. Phillips
. ~
l.'
All BWR plants must have the capability to display and record vessel water level over a range extending from the top of the steam dome to the elevation
(
of the lowest pressure taps.
qualified for pmt accident conditions.This full range of measurements mfut be N
All instrumentation sholild be the time the plant was licensed. qualified to the same standards used for 2.
All level instruments should be referenced to the same point.
that a standard practice be established for all BWR's.
We recommend 3.
For those BWR's which do not have any level measurements below the top of the core Point 1),(additional measures are required to monitor i Big Rock Point, Humboldt Bay. Dresden 1, Oyster, Creek, Nine Mile This could require the installation of pressure taps on piping which enters the reactor at an elevation below the core.
Alternatively, a correlation between incore neutron detector. signals and core cooling may be established 4.
For the non-jet pump plants (other than the Humboldt Bay plant), interlocks must be installed to assure that at least two recirculation loops are open.
for recirculation flow, for modes other than cold shutdown.
5.
The relationship between core cooling (or core mixture level) and downcomer level must be established so that an operator can correctly interpret the
, level measurement provided.
l pressure measurement, the system pressure and the power level be j to provide an estimate of the core' mixture level.
d 6.
has been reviewed here and as part of the developme 1.97.
to the existing instrumentation and this addition would be c the stated objectives of Regulatory Guide 1.97, "(to) provide information a breach of the barriers to radioactivity release (i.e., fu reactor coolant pressure boundary and containment) and if a barrier has been breached..." (Reg. Guide 1.97, Draft 2, p.1.97-2). However, there remain several technical difficulties (location..for example) which make it inpractical to require the installation of ther,mocouples by January 1,1981.
Therefore, th6 addition of core exit thermocouples to BWR's should not be required as part of the Lessons Learned Requirements (NUREG-0578 fitting Regulatory Guide 1.97 to operating reactors.be included i but should Lacrosse BWR tation on the Lacross plant to monitor inadequate core coo presented in the previous section on GE BWR's also apply to this plant.The conclusions n
N
, L.,Phillips (
Sumary There are unresolved questions in the functional design area, hardware design' and qualification area and human factors area. The Core Performance Branch (Thermal Hydraulic Section). Instrumentation and Control Systems Branch, and the Human Factors Engineering Branch should therefore continue the review of the proposed approaches to providing ambiguous indication of inadequate core cooling. After the systems have been modified and found acceptable, the Core Performance Branch (Thermal Hydraulic Section), the Reactor Systems Branch and the Procedures and Test Review Branch should review the procedures for using these systems. The above conclusions should be reviewed by the above branches; and the owners group should be informed of the result as soon as possible.
The major items to be resolved are in the following categories:
B&W plant owners group failure to propose additional instrumentation; correlation of measured values to inadequate core cooling; development of adequate displays; upgrading of core exit thermocouples to post accident qualifications, and training and procedures for use of the new information.
G. M. Holahan Safety Program Evaluation Branch Divisf or, of Safety Technology Office of Nuclear Reactor Regulation
Enclosure:
S:ction 2.17
~
G 4
1
U
- b 4"s. n.n*1, umtro starts I
[,$,,,, }',;
reucLE AH REGULATORY Co!.**.*isslofJ
- iM /, j r.wssmcTou. o. c. rosss t,q, or Decenber 17, 1980
/7 1
).
Note to: Harley Silver From:
D. Ross, Jr,
Subject:
NEED FOR P,EACTOR VESSEL LEVEL INSTRUMENTATION The Met Ed/B&W procedures for detect'on of Inadequate Core Cooling rely primarily on the saturation meter and core exit thermocouples.
The saturation meter, while providing a basis for initial actions, does not distinguish betesen anomalous transients which can drain the pressurizer and cause primary loop saturation due to cooling and shrinkage of primary coolant versus loss of coolant inventory which could lead to inadequate core cooling if it continues.
Core exit thermocouples, do not detect the advent of inadequate core cooling until the level has dropped into the core and fuel heatup has begun.
If level instrumentation were available, the effectiveness of HPI in recovering the system and the trend of level indication (con.'inuing to lose coolant or refilling the system) would provide valuable diagnostic information on the nature of the transient before the level drops into the core.
In event of flow blockage conditions such
(
as occurred in THI-2, superheat can exist at the core exit thermocouples
'(
even though the core has recovered; level instrumentation is needed to j
diagnose the effectiveness of core recovery actions.
A recent event at St. Lucie (June 11,1980) provides a c1assic example of an anomalous event leading to steam bubble formation in the reactor vessel head. The condition was not recognized by plant operators since instrumentation was not available to detect the low level condition. A anst-event evaluation concluded that unsafe operator actions could h===
- m.
)een taken,.
Level instrumentation would have provTded means for prompt
' recognition and response to the situation.
Finally, vessel level information is important and possibly essential to proper emergency pn)cedures relating to use of the reactor vessel head vent required by the TMI Action Plan.
The staff believes that reactor vessel level infonnation will enhance the operating safety of PWRs.
Our position is clear in the September 24, 1980 letter and SER from D. Eisenhut to R. C. Arnold of Metropolitan Edison.
\\
l D. F. Ross, Jr.
k
[/t
] ;y _
,c d5
/
n, n--
--