ML19289E594
| ML19289E594 | |
| Person / Time | |
|---|---|
| Issue date: | 02/28/1979 |
| From: | Moseley N NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| To: | Thompson D NRC OFFICE OF INSPECTION & ENFORCEMENT (IE) |
| References | |
| REF-SSINS-0330, REF-SSINS-SSINS, TASK-AS, TASK-BN-79-10, TASK-TF, TASK-TMR BN--79-10, NUDOCS 7904210475 | |
| Download: ML19289E594 (10) | |
Text
~
A l'
DISTRIBUTION:
M ile i
Central file MAR 2 81973 IE Rdg.
ROI Rdg.
ADFC Chron.
N. C. Moseley E)ORMDUM FOR: Dudley Thompson, Executive Officer for Operations Support, IE i
FRON:
Norman C. Moseley, Director Division of i
Reactor Operations Inspection IE l
[
SUBJECT:
MOTIFICATION OF LICENSIHG BOARDS l
t i
On Febnrary 28,1979, six items concerning Babcock and Ifflcox f
designed nuclear plants were sent to you for forwarding to the appropriate licensing boards. At that time only a preliminary l
l evaluation had been done. We have completdd our evaluation of each of the items and that infomation is enclosed. This additional information should be forwarded to the licensing i
boards.
B/
i Morman C. Hoseley D1 rector
\\
Division of Reactor Operations Inspection IE
Enclosure:
Evaluations of Concerns cc:
S. E. Bryan E. L. Jordan R. F. lleishman, RIII J. C. Stone D. Kirkpatrick G. C. Gower V. D. Thomas D
O 790421047f 3
CONTACT:
J. C. Stone (528019) 1
/l'
/
T/
.:.IE--.D;R01.:TF_,,,,,,,,,,,,,,,,,,,,,,
OF FICE > -1
- I-- - -...--.: -
h8xyan
- A o d n'
%di$feTey JCStone:LD suRN4ue>
3/,],6/79 _,,,,j 3g,/79
,,, g {
3 9
,,,,,g,g,g,,,,,,.
f 3/
omrE m NPC FORM 318 t6 77)
W U. S. GOVERNMENT PRINTING OF FICE 1977-237-025
EXCERPT FROM MEMORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNITS 2 & 3 AND MIDLAND UNITS 1 & 2",
DATED JANUARY 8, 1979 FROM J.S. CRESWELL TO J.F. STREETER 1.
During a recent inspection at Davis-Besse Unit 1 information has been attained which indicates that at certain conditions of reacter coolant viscosity (as a function of temperature) core lif ting may occur. The licensee informed the inspector that this issue involves other B&W facilities. The Davis-Besse FSAR states in Section 4.4.2.7:
The hydraulic force on the fuel assembly receiving the most flow is shown as'a function of system flow in Figure 4-39.
Additional forces acting on the f uel assembly are the assembly weight and a hold down spring force, which resulted in a net downward force at all times during normal station operation.
The licensee states that there is a 500 F interlock for the start-ing of the fodrth reactor coolant pump. However, no Technical Specification requires that the pump be started at or above this temperature. A concern regarding this matter would be if assem-blies moved upward into a position such that control rod move-ment would be hindered.
4 DISCUSSION AND EVALUATION The potential for core lif ting in B&W plants is a concern which has been previously reviewed by NRR.
The concern was first raised in connection with the Oconee 2 and 3 reactors, where the primary coolant flow ~ rates were found to be in excess of the design flow rates.
For example, the Unit 2 flow rate was found to be 111.5% of the design flow rr.t e. Since this was very near the predicted core lift flow rate of 111.9%, an analysis was done by B&W to determine what ef fect core lifting would have on the previous safety analysis for these plants. This analysis-(dated May 2, 1975) indicated that the potential for core lifting did not result in an unreviewed safety question. A subsequent review of this B&W analysis by NRR also concluded that an unsafe condition did not exist (letter froc R. A. Purple to Duke Power, dated 9/24/75).
It should be noted that the potential vertical displacement of the core is limited to a very small distance by the upper core support structure.
Core lifting at power would result in a slight reduction in reactivity since the rising fuel would tend to engage the withdrawn control rods to a slightly greater extent than it would in the bottomed condition.
The amount of this change in reactivity is, of course, available for reinsertion should the fuel settle back to its original position. The potential reactivity increase caused by the settling of the 16 centrally located control rod assembly elements (assumed to have been subject to lifting in the Oconee 2 reactor) was calculated to be 0.1% a K/K. This value is insufficient to have much effect on the accident and transient safety analyses.
Section 1. An additional concern was the potential for damage to the fuel assembly end fittings which might be caused by fretting due to repetitive fuel movement.
Consequently, Duke Power was requested by NRR to make certain examinations of the Oconee 2 fuel during the first refueling to confirm that fuel element motion was not. occurring.
The results of this examination (letter from W. O. Parker to R. C. Rusche dated 7/21/76) showed that no fuel lifting or other type of motion had occurred during the first cycle of operation.
After the core lift concern was ide.rified, B&W developed never types of fuel holddown springs which prov;'a more margin against core lifting than the previous springs did. It is our understanding that the newer types of springs have been installed in a'll B&W reactors.
For these reasons, we believe that there is presently little likelihood that core lifting will-occur during normal power operation. At lower temperatures, there is an increased flow induced lifting force on the fuel due to the higher viscosity of the reactor coolant. Consequently, we view the restriction against 4 pump operation below 500 F as a prudent' precaution against fuel fretting. However, since the potential for core lif ting has little safety significance and because critical operation below 500 F is not permitted, we have no basis to recommend including this restriction in t:ie Technical Specifications.
e e
O
EXCERPT FROM MEMORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNITS 2 & 3 AND MIDIAND UNITS 1 & 2",
DATED JANUARY 8, 1979, FROM J.S. CRESWELL TO J.F. STREETER 2.
Inspection Report 50-346/78-06, paragraph 4, reported reactivity -
power oscillations in the_ Davis-Besse core. These oscillations have also occurred at Oconee and are attributed to steam genera-tor level oscillations. B&W report BAW-10027 states in A9.2:
The OTSG laboratory model test results indicated that periodic oscillations in steam pressure, steam flow, and steam generator primary outlet temperatures ev2ld occur under certain conditions.
It was shown that the oscillations were of the type associa-ted with the relationships between f eedwater heating chamber pressure drop and tube nest pressure drop, which are elimi-nated or reduced to levels of no consequence (no feedback to reactor system) by adjustment of the tube nest inlet resistance. As a result of the tests, an adjustable orifice has been installed in the downcomer section of the steam generators to provide for adjustment of the tube nest inlet resistance and to provide the means for eltsina-tion of oscillations if they should develop during the operating lifetime of the generators. The initial orifice setting is chosen conservatively to minimize the ueed for further adjustment during the startup test program.
We also note that the effect on the incere detector system for moultering core parameters during the oscillations is not clear.
DISCUSSION AND EVALUATION Power Oscillations of the order of 1.5% of full power have been observed at all of the Oconee plants and are considered normal.
In 1977 the power oscillations experienced by the Oconee 3 reactor increased to a maximum of 7.5% of full power. At that time the problem was reviewed by NRR with the conclusion that there was no significant safety considera-tion at that value (Note to B. C. Buckley from S. D. MacKay, dated January 27, 1978). It should be noted that the 7.5% power oscillations cause about a 1 F oscillation in core average temperature due to the short period of the oscillations. The important core safety parameters, which are, the departure from nucleate boiling ratio and the average maximum linear heat generation rate are affected very little by oscilla-tions of this amplitude. The primary cause of the power oscillations is believed to be a fluctuation of the secondary water level in the steam generators. This can be minimized by increasing the flow resistance e
SEction 2.
_2 in the downcomer region of the steam generators. The corrective effort at Oconee 3 was complicated by the fact that the orifice plate provided for this purpose could not be fully closed.
However, the oscillations at other B&W plants have been kept to about 1.5% of full power by appropriate adjustment of the downcomer flow resistance.
For these reasons, the power oscillations at B&W plants are not considered to be a significant safety concern.
EXCERPT FROM MDIORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNITS 2 & 3 AND MIDIAND UNITS 1 &
2",
DATED JANUARY 8, 1979, FROM J.S. CRESWELL TO J.F. STREETER 3.
Inspection and Enforcement Report 50-346/78-06 documented that pressurizer level had gone offscale for approximately five minutes during the November 29, 1977 loss of offsite power event.
There are some indications that other B&W plants may have prob-lems maintaining pressurizer level indications during transients.
In addition, under certain conditions such as loss of feedwater at 100% power with the reactor coolant pumps running the pres-surizer may void completely. A special analysis has been per-formed concerning this event.
This analysis is attached as.
Because of pressurizer level maintenance prob-less the sizing of the pressurizer may require further review.
Also noted during the event was the fact that Tcold went off-scale (less than 5200F). In addition, it was noted that the makeup flow monitoring is limited to less than 160 gpm and that makeup flow may be substantially greater than this value.
This information should be examined in light of the require-ments of GDC 13.
DISCUSSION AND EVALUATION The event at Davis Besse which resulted in loss of pressurizer level indication has been reviewed by NRR and the conclusion was reached that no unreviewed safety question existed.
The pressurizer, together with the reactor coolant makeup system, is designed to maintain the primary system pressure and water level within their operational limits only during normal operating conditions.
Cooldown transients, such as loss of of fsite powar and loss of feed-water, sometimes result in primary pressure and volume changes that are beyond the ability of this system to control. The analyses of and experience with such transients show, however, that they can be sustained without compromising the safety of the reactor. The principal concern caused by such transients is that they might cause voiding in the primary coolant system that would lead to loss of ability to ade-quately cool the reactor core. The safety evaluation of the loss of offsite power transient shows that, though level indication is lost, some water remains in the pressurizer and the pressure does not decrease below about 1600 psi.
In order for voiding to occur, the pressure must decrease below the saturation pressure corresponding to the system temgerature. 1600 psi is the saturation pressure corresponding to 605 F, which is also the maximum allowable core outlet temperature.
Voiding in the primary system (excepting the pressurizer) is precluded in this case, since pressure does not decrease to saturation.
Section 3.
. The safety analysis for more severe cooldown transients, such as the loss of feedwater event, indicates that the water volume could decrease to less than the system volume exclusive of the pressurizer. During such an event, the emptying of the pressurizer would be followed by a pressure reduction below the saturation point and the formation of small voids throughout much of the primary system. This would not result in the loss of core cooling because the voids would be dispersed over a large volume and forced flow would prevent them from coalescing sufficiently to prevent core cooling. The high pressure coolant injection pumps are started automatically when the primary pressure decreases below 1600 psi. Tnerefore, any pressure reduction which is sufficient to allow voiding will also result in water injection which will rapidly restore the primary water to normal levels.
For these reasons, we believe that the inability of the pressurizer and normal coolant makeup system to control some transients does not provide a basis for requiring more capacity in these systems.
General Design Criterion 13 of Appendix A to 10 CFR 50 requires instrumentation to monitor variables over their anticipated ranges for " anticipated operational occurrences". Such oceptrences are specifically defined to include loss of all offcite power. The fact that T cold goes off scale at 520 F is not considered to be a deviation
~
from this requirement because this indicator is backed up by wide range temperature indication that extends to a lov limit of 50 F.
Neither do we consider the makeup flow monitoring to deviate since the amount of makeup flow in excess of 160 gpm does not appear to be a significant factor in the course of these occurrences.
The loss of pressurizer water level indication could be considered to deviate from GDC 13, because this level indication provides the principal means of determining the primary coolant inventory. However, provision of a level indication that would cover all anticipated occurrences may not be practical. As discussed above, the loss of feedwater event can lead to a momentary condition wherein no meaningful level exists, because the entire primary system contains a steam water mixture.
It should be noted that the introduction to Appendix A (last paragraph) recognizes that fulfillment of some of the criteria may not always be appropriate. This introduction also states that departures from the Criteria must be identified and justified. The discussion of GDC 13 in the Davis Besse fSAR lists the water level instrumentation, but does not mention the possibility of loss of water level indication during transients. This apparent omission in the safety analysis will be subjected to further review.
EXCERPT FROM MDiORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNITS 2 & 3 AND MIDLAND UNITS 1 & 2",
DATED JANUARY 8, 1979, FROM J.S. CRESWELL TO J.F. STREETER 4.
A memo from B&W regarding control rod drive system trip breaker maintenance is attached as Enclosure 2.
This memo should be evaluated in terms of shutdown margin maintenance and ATWS considerations particularly in light of large positive moderator coefficients allowable with B&W facilities.
DISCUSSION AND EVALUATION Our investigation of the above circuit breaker problems has revealed that eight failures of reactor scram circuit breakers to trip during test have been reported from Babcock & Wilcox (B&W) type operating facilities since 1975. In each case, the faulty circuit breaker was identified as a GE type AK-2 series (i.e., AK-2A-15, 24, or 50). The causes for failure were attributed to either binding within the linkage mechanism of the undervoltage trip device (UV) and trip shaft assembly or an out-of-adjustment condition in the same linkage mechanism. B&W and GE' determined that the binding and the out-of-adjustment conditions resulted from inadequate preventive maintenance programs at the affected operating facilities.
In addition to the breaker problems experienced at the B&W facilities, three circuit breakers of the aforementioned GE type failed in similar fashion at the Oyster Creek operating facility on November 26, 30, and December 2, 1978. As in each case above, cleaning and relubricating of the UV/ trip shaf t assembly within the circuit breaker was required to correct the problem.
It is significant to note that during the November 30, 1978 event, both redundant service water pump circuit breakers failed to trip as required during the loss of off-site power test.
These failures in turn created a potential overload condition on the emergency busses during the sequential bus loading by each diesel generator.
However, both diesel generators successfully picked up their required bus loads without experiencing a unit shutdown from an overload condition.
With respect to the generic implications and safety significance of this issue, both B&W and GE are in the process of issuing alert letters to their customers. These letters are scheduled for issuance by late March and will describe the causes for failure and provide recommendations to resolve the problem.
Based on our study findings and on information obtained in discussions about the breaker problem with the knowledgeable people from B&W, GE and Region II, we plan to issue an IE Circular covering the matter. The thrust of the Circular will be directed toward the need for adequate preventive maintenance programs at all operating facilities.
Specific recommendations from GE to resolve the above breaker problem will also be mentioned in the Circular.
EXCERPT FROM MEMORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNIIS 2 & 3 AND MIDIAND UNITS 1 &
2",
DATED JANUARY 8, 1979, FROM J.S. CRESWELL TO J.F. STREETER 5.
Inspection and Enforcement Report 50-346/78-17, paragraph 6 refers to inspection findings regarding the capability of the incore detector system to determine worst case thermal conditions. The reactor can be operated per the Technical Specifications with the center incore string out of service. If the peak power locations is in the center of the core (this has been the case at Davis-Besse), factors are not applied to conservatively monitor values such as Fq and F delta H.
DISCUSSION AND EVALUATION We do not believe that there is a valid basis for requiring the center string of incore detectors to be always operable in B&W reactors.
The power distributions for various plant conditions, throughout the fuel cycle, are calculated prior 7.o the operation of the reactor.
The power distribution is verified at the beginning of operation, and periodically thereafter, by comparison with the available incore detectors. The power in fuel assemblies that lack detectors (including those with failed detectors) is derived by using the known power distribution to determine the power ratios between such an assembly and nearby assemblies that have detectors. These ratios can then be multiplied by the power in the measured assemblies to derive the power level in any specific unmeasured assembly.
The central assembly is not fundamentally different than any other assembly in this respect. Although this assembly is the highest powered assembly in the Davis Besse reactor at the beginning of the fuel cycle, this is not the case at all reactors. Nor does the central assembly have the highest power, in the Davis Besse reactor, at the end of the first fuel cycle. Since there is some variation between the calculated power distributions and the actual ones, an appropriate margin is assumed for this variation in establishing the allowable power peaking factors.
Fixed incore detectors must function in an extremely harsh environment and are subject to high failure rates.
In order to ensure that an ade-quate number will survive the fuel cycle, many more detectors are installed than are necessary for the power distributions determinations.
To require the central string to be always operable would likely result in unnecessary power restrictions. Neither the standard Technical Specifications (STS) for B&W plants nor the STS for CE plants (which also have fixed incore detectors) require the central detectors to be operable.
P EXCER?T FROM MEMORANDUM ENTITLED " CONVEYING NEW INFORMATION TO LICENSING BOARDS - DAVIS-BESSE UNITS 2 & 3 AND MIDLAND UNITS 1 &
2",
DATED JANUARY 8,1979, FROM J.S. CRESWELL TO J.F. STREETER
- 6. describes an event that occurred at a B&W facility which resulted in a severe thermal transient and extreme dif-ficulty in controlling the plant. The aforementioned facilities should be reviewed in light of this information for possible saf ety tsplications.
DISCUSSION AND EVALUATION Following the cooldown transient at Rancho Seco, NRR evaluated the event and concluded that no structural damage had occurred to the primary coolant system which would preclude future operation of Rancho Seco.
However, in their safety evaluations they concluded that positive steps should be taken to preclude similar transients and that the generic implications of this event should be reviewed. In addition, IE initiated a Transfer of Lead Responsibility, Serial No. IE-ROI 78-04, dated April 25, 1978, recommending that:
1.
NRR perform a generic review of the non-nuclear instrumentation power supplies for other B&W units, if design changes to the non-nuclear instrumentation (NNI) power supplies are required at Rancho Seco.
2.
NRR evaluate the susceptibility of B&W plants to other initiating events or failures which could cause similar significant cooldown transients.
This event is currently being evaluated by NRR.
O d