ML19274D605
| ML19274D605 | |
| Person / Time | |
|---|---|
| Issue date: | 06/13/1978 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19274D604 | List: |
| References | |
| REF-GTECI-A-16, REF-GTECI-SY, TASK-A-16, TASK-OR NUDOCS 7902140038 | |
| Download: ML19274D605 (8) | |
Text
.
Interim Evaluation of Amendment 3 Summary of Topical Report During 1974 ASEA-ATOM, an overseas manufacturer of BWR's, advised GE that preliminary % formation from tests with single spray nozzles in pressurized steam environments indicated that the nozzle spray patterns were affected by several test parameters, including the steam pressure and nozzle flow rate.
More specifically, the characteristic cone angles of nozzles tested by ASEA narrowed in a steam environment by comparison to the cone angles observed in air.
Based on this information, GE began a program to quantify the effects on the spray distribution.
This report describes a series of single nozzle tests in steam using different types of nozzles typical of those used in BWR/2 through BWR/5 plants. These tests quantified the amount of single nozzle cone collapse and spray axis shift due to tne steam environment that would be expected in the upper plenum of a BWR following a LOCA. A particular nozzle (designated the "VNC" nozzle) that was designed to produce a hollow cone spray pattern was found to be most adversely affected by the presence of steam.
These effects on the "VNC" nozzle were then simulated in a full scale testing (air only) facility. That is, each "VNC" nozzle in the air testing facility was modified so that it would reproduce, in air, the spray pattern that the single nozzle steam tests showed would be produced by an identical "VNC" nozzle in a steam environment.
Full scale air tests, with the "VNC" nozzles so modified, were then performed for a typical BWR/4 or BWR/5 plant (core spray sparEen are identical for plants of the same size with these two designs).
Results of those 7902140038
9
. tests indicate that, even allowing a considerable margin to cover uncer-tainties in the test methodology and assumptions, minimum spray flow to any channel following a LOCA will not be less than half of the design flow that was previously demonstrated by tests and calculations which did not include steam effects (measurements from simulated VNC nozzles with collapsed spray cones show the minimum bundle flow to be 60.3". of design flow).
Since core spray distributions for other plants (with nozzles that are less severely affected by steam) should be less severely degraded, this factor of two is believed to.be a bounding or maximum degradation factor for all plants.
Summary of Staff Evaluation In response to staff questions regarding interim acceptability of CS flows, GE provided (1) measurements for each BWR size and classification plant of minimum bundle spray flows for one sparger only, in air, and calculated values of a figure of merit (FOM) where FOM = minimum measured channel spray flow from one sparger only without simulation of steam effects / flow necessary to remove decay heat by vaporization.
The minimum FOM for BWR/2 through BWR/5 was 1.3 calculated as described above or 0.65 considering the worst effect of cone collapse (i.e., with the factor of two reduction discussed aboves.
BWR FLECHT data (2) show little degradation in heat transfer for a FOM as low as 0.38.
Therefore, the above referenced data indicates that for the worst size and break locations with the worst attendant single failure, the spray flow from a single spray sparger is adequate to justify credit for cooling assumed in ECCS analyses. Also, considering the fact that the limiting
~
break with the worst single failure leaves two core spray systems available (plus one or more flooding systems in certain plants), it is een that tne minimum F0M would be approximately 1.3, i.e., twice C.65 quoted above.
3 (Flow typical of the minimum flow to any group of fuel bundles from only one-spray-system-operation wds present in tests previously run to measure spray heat transfer coef ficients (Full Length Emergency Core Heat Transfer tests)* - Thus, the air tests described above are evidence that adequate spray flow exists even with conservative consideration of steam effects, i.e., that a comfortable margin above a FOM of 1.0 exists with both spray systems operatirg for the worst break size - break location-single failure combination. For breaks where only one CS sparger is available (CS line break, for example), calculated PCT's are sufficiently far below that calculated for the limiting break so that any slight spray cooling degradation would not cause such a break to become limiting.
Although not specifically included in the tests described above, the same type of interim " flow margin" information has been provided on an individual plant-by-plant basis for each currently operating BWR/l plant, with similar acceptable margins indicated in all cases.
In addition, to provide final evaluation of the exact spray flow margins present, a series of tests and calculations will be performed in the evaluation of each size BWR/6 plant (218, 238 and 251 inch inside vessel diameter). These tests include:
1.)
Single nozzle, full scale tests will be perfcrmed in steam for each spray nozzle type. The horizontal spray facility (HSF) will be used for these tests, allowing the single nozzle to spray over a full scale representation of the re3ctor upper plenum region thac would be covere'd by a single nozzle.
2.) Data from the HSF will be used for three purposes: a) a " simulator" nozzle will be developed which will simulate, in air, the spray
- FLECHT tests were performed for bundle spray flow rates ? 0.5 gpm. Therefore 0.5 g::r represents the lower limit of acceptable flows for any FOM.
. pattern produced by the actual nozzle in the HSF in steam; b) the data will be used to calibrate a single nozzle calculational model which can extend the HSF data base; c) the data will be used directly as input to a multiple-nozzle rodel.
3.) Multiple nozzle full scale tests will be conducted in air to determine nozzle-to-nozzle interaction effects on overall spray distribution patterns. These tests will be conducted using the " simulator" nozzles described in 2-a above. The data obtained will be used with the single nozzle model described in'2-b above so that the model will predict spray distributions from multiple nozzles in steam.
4.) A full-reactor representation of core spray distribution in steam will then be obtained by using the multi-nozzle model described in 3 above in conjunction with a full scale, 360 spray test conducted in air using the " simulator" nozzles.
This process will be confimed by a representative 218 inch BWR/6 multi-nozzle steam test. This test will be conducted at the new facility in Lynn, Massachusetts. The Lynn facility is a fuil scale mockup of a 30 sector of a BWR/6 upper plenum.
The above methodology will be app;ied to each class of BWR. Once the initial BWR/6 results are known, it will be decided by mutual di2cussion between the NRC staff, GE and the licensees involved exactly how this will be accomplished.
In the unlikely event that the confimatory tests at Lynn do not support the margins presently believed to exist, two relatively simple alternatives exist for operating plants:
1.) LOCA re-analyses could be performed, using lower spray cooling coefficients for which an adeouate spray flow margin can be demonstrated. This could result in operating restrictions on some plants. Such future LOCA re-analyses might be allowed to take credit for other compensating effects pending confirmatory experimental results, such as less severe reflooding delays due to counter-current-flow limiting effects.
2.)
It is anticipated that relatively simple hardware modifications could be developed which would produce a better spray distribution in the post-LOCA steam environment.
Such modifications would probably involve re-aiming presently installed nozzles, replacing some or al s of the presently installed nozzles with one of the currently available nozzle designs which have been shown to be more effective in the post-LOCA steam environment, or replacing presently installed nozzles with an improved nozzle design which may be developed.
Such modifications would be feasible for operating plants.
In view of the spray flow margin presently believed to exist for operating plants and in view of the alternatives available in the unlikely event that the confirmatory tests do not confirm the margins presently believed to exist, we conclude that continued operation of licensed GE-BWRs, during the interim period pending final resolution of this Task, does not present undue risk to the health and safety of the public.
For BWRs currently under review for an Operating License (0L), all of the above statements, alternatives and conclusions are equally appiicable.
In particular, alternative 2) immediately above v ill be viable because results of further tests will be available for ;cnsideration in determining desirable design changes, and those design changes could be incorporated into a (non-radioactive) core that has not been operated at power.
In view of the
M margin in spray flow presently believed to exist for plants applying for an OL (as previously described) and in view of the alternatives available in the unlikely event that the confirmatory tests do not confirm the margins presently believed to exist, the staff has concluded that Operating Licenses can be granted with reasonable assurance that operation will not present undue risk to the health and safety of the public.
For BWRs currently applying for a Construction Permit, again all of the above statements, alternatives and conclur, ions are equally applicable.
Since necessary modifications, if any, tc the plant design can be accomplished while the plant is being constructed, the staff has concluded that, pending completion of this test program, Construction Permits can be granted with reasonable assurance that (1) there will be a satisfactory resolution of this concern prior to operation, and (2) operation will not present undue risk to the health and safety of the public.
Staff Cosition BWR FLECHT data show little degradation in heat transfer for a figure of merit, FOM, (FOM = minimum measured channel spray flow / flow necessary to remove decay heat by vaporization) as low as 0.38.
These FLECHT tests covered bundle spray flow rates as low as 0.5 GPM. Therefore, a FOM that is confortably above 1.0, after inclusion of steam effects (i.e., division by 2 in lieu uf a more realistic determination of steam effects) is considered sJfficient to allow for uncertainties involved in the steam effects and measurement techniques. and to justify use of Appendix ri specified heat transfer coefficients.
Amendment 3 to NED0-20566 and information provided separately by General
. Electric indicate that adequate spray flow exists to justify use of minimum core spray heat transfer coefficients specified by Appendix K.
A confirmatory experimental and analytical program has been initiated to verify this con-clusion.
Even if that program should not confirm the adequacy of the core spray, suitable alternatives exist to compensate for inadequacies of the core Therefore, continued operation of licensed BWR's and the licensing spray.
of pending applications does not present undue risk to the health and safety of the public pending final resolution of this concern.
1 REFERENCES
- 1) Letter and attachments, NFN 093-78, A. J. Levine to Darrell G.
Eisenhut, " Core Spray Distribution Proaram", March 1, 1978.
- 2) APED-5529, " Core Spray and Core Floodina Heat Transfer Effectiveness in a Full-Scale Boiling Water Reactor Bundle", June, 1978, F. A.
Schraub and J. E. Leonard.
.