ML19276H479
| ML19276H479 | |
| Person / Time | |
|---|---|
| Issue date: | 11/02/1979 |
| From: | Hendrie J NRC COMMISSION (OCM) |
| To: | Udall M HOUSE OF REP., INTERIOR & INSULAR AFFAIRS |
| Shared Package | |
| ML19276H480 | List: |
| References | |
| NUDOCS 7911290006 | |
| Download: ML19276H479 (4) | |
Text
\\Q f
.. s
~ / November 2, 1979 OF FICE* OF TH E CHAtR:.1AN CofK The Honorable Morris K. Udall, Chairman Subcommittee on Enercy and the Environment
% $ []'Tal\\ gl[g Cocinittee on Interior and Insular Affairs L UJli Ug, g United States House of Representatives Washington, D.C.
20515
Dear Mr. Chairman:
This is in response to your letter dated September 4,1979 concerning the in-formation in IE Bulletin 79-05C and 79-06C. That bulletin instructed all PWR licensees to trip all reactor coolant pumps following reactor trip and indi-cation of hich pressure inje: tion. Your review of the bases for these in-structions ir.dicated that core uncovery would occur for certain small breaks irrespective of whether or not the reactor coolant pumps are tripped.
As you know, a spectrum of p:stulated large and small break loss of coolant calculations is perfarmed for each cperating reactor.
The break size range ccvered in the small break calculations is approximately 0.5 sq. ft. down to 0.001 sq. #t.
The large break spectrum covers break areas greater than 0.5 sc. ft.
The lower er.d of the small break spectrum is defined as the break size for '..hich there is no r.st loss of fluid from the primary system due to
- he injection of makeup fluid from the charging system.
Fcr the postulated 'reaks censidered in these analyses, it should be n:ted c
that core un:overy has been :redicted for break areas equivalent to abcut a 4-inch cianeter pipe (.09 sq. ft.) or greater.
For these breaks, the prirary system depressurizes, whether the reactor coolant pumps are on or c##, and the core is recovered by fluid injected by the safety injection
- anks.
In a:ditien, for breaks equivalent to approximately a 2-inch di arete (.C1 sc. f:.) er s aller, the core remains covered (reactor
- colant pumps en or Off), assu ning a single f ailure in the high ;.* essure ir je: tion D:1) system.
(its area of a typical pilot operated relief valve is ap;-oximately C.C1 sq. ft.)
For breaks in the range frca a; proxima:ely 2 to 4 inches in diameter, a significant difference in the duration and decree of core uncovery is indicated depending upon the assumption of reactor ecolant pump operation.
Thus, the action of the reactor coolant pumps has a significant effect on core level for only a narrow portion of the small break spectrum.
It was our concern for postulated breats in this raage that provided the basis for the instructions in IE Bulletin 79-05C and 79-06C.
The above discussion is intended to provide the sensitivity of assumed Uith pum: operation over the rance of postulated small break accidents.
this backgrcund, the following response is provided to the questions in These answers are generic in nature and could vary somewhat your letter.
cepending upon plant arrapgement and safety injection configuration.
1424 266 7911290 OM
=
e-
.J
/
l.
Q. Are there small breaks in the~ieactor coolant system for which the core will be uncovered irrespective of operator actions?
A.
Tne rea: tor core will be uncovered for all breaks approximately 4-inches in dia:ieter and greater irrespective of operator actions.
For breaks in the range from 2 to 4 inches, it is essential that the operator quickly turn off the reactor coolant ptnps.
In addition, the degree of core uncovery can be reduced or eliminated if the operator quickly restarts a second high pressure injection punp, if one pump failed to start automatically upon HPI initiation.
If the operator fails to promptly turn off the reactor coolant punps for breaks in this range, and the pumps are subsequently tripped, the degree and time period of core uncovery will be more extensive.
The are is defined to be uncovered in this context whenever the vessel rixture level is below the top of the fuel.
Follcwing a s.all b-eak loss of coolant accident (LOCA), a two-phase mixture of 'luid will exist in the core region.
This bubbly mix ure of steam and water results from the flashing of the primary system fi uf d due t: system decompression and the heat addition from the c rs and me:al surfaces'within the vessel. The mixture initially swe:ls above the top of the core into the region of the reactor vessel upper plenum. As fluid is lost from the system through the assmed break, the mixture level can drop below the top of the core.
As the blowdov.n process continues, the reduction in sys:ea :ressure results in a decreased rate of fluid loss, and ar increase: rate of fluid injection from the HPI system. Tnere-f:rs, tr.e c:re level reaches a minimum and subsequently increases as the c
' low exceeds the break flow.
Tne core is rec:vered whe: the c.ixture again rises above the top of the fuel.
2.
O.
If :he :c e will be uncovered, how long is the period during which tre water livel will be bel:w the top of the fuel? What would be re peai clad temoerature curing the ancovery period?
T: what e n sn:.ill fuel be damaged during tne uncovery period?
A.
Tr e durzt'cr cf core uncovery is dependent upon the postulated break size and the number of systens assumed to be operable.
For breaks ati:h depressurize the primary system such that the safety injection tanks are ir.itiated, the uncovery time will be less than for smaller breaks which are recovered by fluid injected by the HPI system. The following uncovery times were detennined assuming that the reactor coolant ptr:s are quickly tripped and one HPI system is r.ot available.
Tnese calculations were performed with conservative models intended to beurd the anticipated syste: behavior. The degree of uncertainty in the analyses has not been determined, but is believed to be less than the degree :f conservatism provided in the LOCA models.
'P00R ORIR y
,_ y
J
./
Pioe Diameter Lencth of Uncovery Time Peak Cladding Temoerature 2 inch 100 seconds 1000* F 3
1500 1700 4
1000 1540 6
200 1375 The mixture level for these cases would not be expected to totally uncover the reactor core.
Consequently, the steam generated in the lower portion of the core would provide cooling to the fuel above the mixture level during the uncovery period.
The resulting calculated peak cladding temperatures for the above postu. lated breaks are all below the LOCA limit of 2200*F.
For cladding temperatures below 1450*F, no fuel damage is expected.
For temperatures in the range from 1450 to 2200*F, limited cladding ballooning and perforation would be expected for the hottest fuel rods, but the overall integrity and coolability of the reactor core would be maintained.
If the reactor coolant pnps are tripped later in the transient with a large amount of steam in the ry: tem, the peak cladding temperature is calculated to exceed 2200 F for some breaks in the 2 to 4 inch diameter range.
In this case, extensive core damage could result.
3.
Q.
If cperator acticns can prevent uncovering of the core, will the operator have sufficient information to take the appropriate actions to prevent uncovering?
A.
As indicated above, the degree of core uncovery can be reduced or prevented by operator action for breaks in the range from two to four incnes in diameter.
For breaks larger than four inches in diameter, operator action would not prevent uncovery but could reduce the severity of the event.
The initial operator action would be to trip the reactor coolant pumps and check to see if the HPI pump had started.
If not, the operator wou.r. start the second HPI pump and inject water into the primary system through the charging system.
For small breaks in this range, the primary system will quickly depressurize to saturated fluid conditions.
This reduction in pressure will trip the reactor and activate the ECCS signal.
The reactor trip and ECCS activation signals are the two signals required for manual tripping of the reactor coolant pumps by the operators.
It is tnerefore concluded that the operator has sufficient information to take appropriate action.
P00RORGlE.
1424 %8 y
J It is noted that other. events, such as a stuck open secondary system valve, could also have similar symptoms.
The utility owner's groups have considered these events and concluded that no significant adverse effects would result from tripping the pumps.
In order to accomplish the desired effect, the operators must trip the reactor coolant pumps within a few minutes; thus, it is recommended that the action be performed promptly.
In this regard, NUREG-0578 recomended that analyses of inadequate core cooling be performed by the reactor vendors. A portion of these anlayses is intended to p ovide the symptoms of impending core uncovery and the anticipated response of various system pararoters.
This study wili be used to generate improved guidelines for operator training and an assessment of the need for improved plant in:trumentation to detect and assess potential core uncovery. The NUREG-0578 schedule requires that the analyses and emergency procedure guidelines be provided in October 1979 and that retraining of operators be ccmpleted in January 1980.
Please contact me if I can be of further assistance in this matter.
Sincerely,
.. M W h b-
~u Joseph M. Hendrie Chairman c::
Rep. Steven Syms 1424
?69 f
..