ML16341A314
| ML16341A314 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 10/05/1984 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML16341A313 | List: |
| References | |
| TASK-1.C.1, TASK-TM GL-81-21, GL-83-22, TAC-51994, NUDOCS 8410150532 | |
| Download: ML16341A314 (18) | |
Text
Diablo Canyon Units 1 and 2
Safety Evaluation Report Generic Issue 81-21, Natural Circulation Cooldown
~kd On June ll, 1980, St. Lucie Unit 1 experienced a natural circulation cooldown event which resulted in the formation of a steam bubble in the upper head region of the reactor vessel.
Consequently the NRC Generic Letter dated Nay 5, 1981 was sent to all PWR licensees (Reference 1).
Per that letter the licensees were asked to provide an assessment of he ability of their facility procedures and training program to properly manage similar events.
This assessment should include:
(1)
A demonstration (e.g., analysis and/or test) that controlled natural circulation cooldown form operating conditions to cold shutdown conditions, conducted in accordance with their pro-
- cedures, should not result in reactor vessel
- voiding, (2)
Yerification that supplies of condensate grade auxi liary feedwater.
are sufficient to support their cooldown method, and (3)
A description of their training program and the revisions to their procedures.
PG&E responded to this request by letter dated December 7,
1981 (Reference 2).
The following is our evaluation of the response.
(
8410150532 84l005 PDR ADOCK 05000275 PDR
S
'valuation In its submittal PG8E refers to a study performed by Westinghouse for the Westinghouse Owners Group (WOG).
This study evaluates the potential for steam formation in Westinghouse NSSS's and recommends modifications to the I'perator guidelines.
The results in the'estinghouse
- report, W-OG-57 (Reference 3), are bounding in that they are applicable to all 2, 3, and 4 loop Westinghouse plants.
The report concludes that in previous analyses for operating guidelines and safety analyses, void formation in the upper head is explicitly accounted for if it is calculated to occur.
These previous analyses indicate that voiding is not a safety concern because the voids will collapse when they come in contact with the subcooled region of the vessel.
The present analysis differentiates between T t and T
ld plants.
hot cold Tco 1 d p 1 a nts are th o s e wh ic h dur in g norma 1 re a ctor coo 1 a nt p u mp operation, have sufficient flow between the downcomer and the upper head such that the temperature of the upper head is approximately the same as the cold leg temperature.
Th t plants have an upper head hot temperature that is between the hot leg and cold leg temperatures.
This evaluation applies to the Th t analysis because the Diablo Canyon Units hot 1 and 2 are considered to be Th t plants.
.hot
The analysis is done using the WFLASH code with a best estimate model.
The WFLASH code has 2-phase capabi lity and can track void propagation.
The analysis assumes an inverted top hat upper support plate design since it results, in a large upper head volume and hence conservatively large total heat in the upper head region.
The initial upper head temperature is conservatively set equal to the hot leg temperature.
lp t<etal heat addition to the upper head area from the vessel and internals is taken into account.
It is assumed that the reactor coolant pumps are stopped at the beginning of the transient.
The analysis is done for two cooldown rates, 25 F/hr and 50'F/hr.
An analysis has also been done which accounts for the effect of the control rod drive mechanism (CRDN) cooling fans.
These fans blow containment air across the vessel'head and provide cooling of the upper head and the CRDHs.
One of the conditions that must be met during a cooldown is that the primary system pressure be 400 psia when the primary system temperature is 350'F.
These conditions will permit the residual heat removal system (RHRS) to be used to continue plant cooldown.
- However, RHR entry conditions vary somewhat from plant to plant.
The analysis without the CRDN fans shows that upper head voiding will occur unless the depressurization is halted at 1200 psia and cooldown continued
to a hot leg temperature of 350'F and the upper head is allowed time to cool off before depressurization to the RHRS point.
The reference report calculates this cool-off period to be approximately 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> for a 25'F/hr cooldown rate and approximatelg 27 hours3.125e-4 days <br />0.0075 hours <br />4.464286e-5 weeks <br />1.02735e-5 months <br /> for 50'F/hr cooldown rate.
An addition analysis includes the effect of the CRDH cooling fans and results in a significant increase in the rate of cooldown of the upper head.
Per the reference report the CRDM fan cooling system removes 780KW (12 Kw/drive train times 65 drive trains for the analyzed plant) at full power.
This energy removal is equal to an upper head cooldown rate of 32'F/hr when the upper head temperature is 600'F.
Assuming that the cooldown rate is proportional to the temperature difference between the upper head metal and the contain-ments atmosphere, the CRDM fans wou,ld,cool the upper head at a rate of 17'F/hr when the upper head fluid is 350'F.
Based on these analyses the Westinghouse report makes the following recommendation for operator guidelines:.
l.
If the CRDM cooling effect is available the operator can reach shutdown cooling entry conditions without void information if a 25'F/hr cooldown rate is used and a 50'F subcooling at the hot leg is maintained.
2.
If the CROM fans are not available the operator should commence a 25'F/hr cooldown and.depressurize at a rate which maintains 50'F subcooling until
the system reaches 1900 psia.
At this point the depressurization rate should be changed so that a 200'F subcooling margin is maintained unti 1 the system reaches 1200 psia.
At this time the depressurization should be stopped,,but the cooldown continued.
When the hot leg temperature reaches 350'F, a 20-hour holding period sho'uld be allowed before depressurization to RHRS entry conditions.
Although the above recommendations were based on best estimate
- analyses, these analyses were conducted for a worst case plant i.e.,
a 12" thick inverted top hat upper support plate with upper head region volume of 847 ft.
Recognizing that no plant fits the description, Westinghouse conducted another set of analyses that account for the variations in the upper head internal design, i.e., whether the upper support plate is of the top hat, flat, or the inverted top hat design.
The upper support plate design determines the rate of heat conduction and the upper head water volume which must be adequately cooled before depressurization to the RHRS conditions is attempted.
This additional set of analyses was presented by Westinghouse in the background information for the Westinghouse emergency i esponse guidelines ERGs
(.ES-0.2).
The Diablo Canyon Units have a 5" thick top hat upper support plate design with upper head region volume of 508 ft The Westinghouse ERGs recommend a
200'F subcooling margin and an 8 hour9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> cooloff period at 1200 psig for this type of 'plant.
The licensee has revised it's natural circulation cooldown procedures in accor dance with the Westinghouse ERGs.
The plant pr ocedures EPOP-23, natural circulation cooldown, Rev.
3, July 1983 specifies a cooldown rate of no more than 25'F/hr, a subcooling margin of 200'F (if CRDM cooling fans are not operating),
and a cooloff period of 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> at 1200 psig before the RCS can be depressurized to the RHRS entry pressure of 425 psig.
The licensee stated th'at the Diablo Canyon Technical Specification require that a
minimum of 178,000 gallons in one condensate storage tank per unit plus a
minimum of 270,000 gallons in a common fire water storage tank.
This source of water is backed up by the plant ultimate heat sink, the Pacific Ocean.
The staff emphasizes the importance of procedures and operator training.
The NRC staff has reviewed, as part of TMI. Action Plan Item I.C. 1, the proposed Westinghouse Emergency
Response
Guideline (ERG)
Program as described in Westinghouse Owners Group (WOG) letters of November 30, 1981, July 21, 1982'and January 4,
1983.
The results were presented in NRC Generic Letter 83-22, "Emergency
Response
Guidelines,"
(Reference 4).
The licensee advised the staff that the Westinghouse procedures will be appropriately implemented into the plant specific procedures by March 1985 (Reference 5).
The staff concludes that adequate procedures will be available for the operator to safety conduct a controlled natural circulation cooldown.
Regarding-the matter of training, the licensee has incorporated in the operator requalification training program a presentation of the St. Lucie Unit 1 event, including aspects of how voiding occurs and its consequences, signs that voiding is occurring, and procedures to prevent and mitigate voiding.
The licensee also informed the staff that past simulator training, conducted at the Zion Station, does include upper head voiding aspects.
Simulator training to be performed at'he DiaElo Canyon 'simulator likewise will include training for upper head voiding.
In addition, all operators have participated in at least one.and have observed two of the five natural circulation tests that were performed during the low power test program as required by License Condition 2.C.(8)f. for Diablo Canyon Unit 1 (Reference 6).
The staff concludes that the training program with respect to upper head voiding is adequate.
Conclusion Upper head voiding, in itself, does not present any safety concerns provided that the operator.
has adequate training and procedures to recognize and react to the situation.
Yoiding in the upper head makes RCS pressure control more difficult and therefore, if the situation warrants, natural circulation cooldown should be done without voiding.
The Westinghouse analyses provide'he length of the holding period necessary to cooldown the upper head region on natural circulation without void formation when the CRDM fans are not available.
Natural circulation cooldown tests are planned for Diablo Canyon.
These tests will provide experimental verification of the upper head cooling rate calculations.
The staff concludes that the licensee has verified it has sufficient condenate supplies.
The staff finds that upon acceptable implementation of the NRC-approved Westinghouse Owners Group Emergency
Response
Guidelines in accordance with Generic Letter 83-22 with appropriate plant specific modifications, the licensee's procedures will be adequate to perform a safe natural circulation cooldown.
The staff finds the operator training with respect to upper head voiding acceptable.
References
( 1)
Generic Letter 81-21, "Natural Circulation Cooldown",
May 5, 1981.
(2)
P.
A. Crane Jr., Pacific Gas and Electric Co., to F. J. Miraglia, Jr.,
NRC,'ated December 7,
1981.
(3)
Jurgensen, R.
W. to P.
S.
Check, "St. Lucie Cooldown Event Report,"
W-OG-57, April 20, 1981.
(4)
Generic Letter 83-22, "Emergency
Response
Guidelines,"
June 3, 1983.
(5)
J.
- 0. Schuyler, Pacific Gas and Electric Co., to D.
G. Eisenhut,
- NRC, dated April 30, 1984.
(6)
J.
- 0. Schuylei., Pacifsc Gas and Electric Co., to D.
G. Eisenhut,
- NRC, dated July 24, 1984.
1 1
~
'I pe' 4