ML19199A530
| ML19199A530 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 02/06/1978 |
| From: | Ross D Office of Nuclear Reactor Regulation |
| To: | Vassallo D Office of Nuclear Reactor Regulation |
| References | |
| TASK-TF, TASK-TMR NUDOCS 7905030178 | |
| Download: ML19199A530 (11) | |
Text
-_
.n
~,
FEB 0 6 '92 MEMORANDUM FOR:
D. B. Vassallo, Assistant Director for L AR's, DPM FRCH:
D. F. Ross, Jr., Assistant Director fcr Reactor Safety, DSS
SUBJECT:
THI-2, INPUT TO SER SUPPLEMENT H0. 2 Plant Ham:
IMI-2 Docket No.:
50-320 Milestone No.:
27-21 Licensing Stage:
OL Responsible Branch LWR-4 and Project Manager:
H. Silver Systems Safety Branch Involved:
Reactor Systems Description of Review:
Input to SER Supplement !b. 2 Review Status:
Complete The Reactor Systems Branch has prepared the attached ipput for SER supplement nieber 2.
This completes the RSB review of Three Mile Island Unit 2.
The following topics are addressed.
5.2.2 Overpressure Prctection during Startup and Shutdown
- 6.2 Net Positive Suction Head Assessment 6.3.3 Makeup Tank Isolation
- 15.2.2 Steamline Break Analysis
- 15.2.2.1 Secondan System Modification
- 15.2.2.2 Long Term Cooling following a steamline Break
- 15.2.8 Feedwater t.ine Ireak Note: Items =arked with
- involve limits on plant operation D. F. Ross, Jr., Assistant Director for Reactor Safety Division of Systems Safety
Enclosure:
SER Input 77343o3c4
... s qp s
.,i /
7905036 /7f
O r_.
s e
D. B. Vassallo cc:
S. Hanauer R. Mattson D. Ross S. Varga H. Silver T. Novak S. Israel J. Watt Distribution Docket File NRR Reading RSB Reading WATT Chron DSS:RSB DSS:RS M D_SS :RL9 _ -PSh[/
g Sa3trb[
SIsrael' Mk i
DFhIs f
_2I / 73.. _...-_2LtLZS.._._.
P/bl18. l 2h12d_.
o m.-
oc== m
<,.u,== =
.. z _ _
_,/
t.
5.2.2 Overaressure Protection Durine Startuo and Shutdown Several instances of reactor vessel overpressurization have occurred in pressurized water reactors in which the Technical Specifications implementing A;cendix G to 10 CFR 50 have been exceeded. Vessel stress limits as a function of pressure and vessel temperature decrease as the result of vessel irradiation through the life of the plant.
During the first fuel cycle, the applicant has administrative procedures and ecuicment to minimize the potential for excessive pressure transieats under startup and shutdown conditions. By procedure, either a steam or nitrogen bubble will be in the pressurizer with a high level alarm and a icw level-interlock to maintain specified level limits.
The presence of a bubble reduces the repressurization rate which results in more time for operator action. A single dual range relief valve will also be available.
The NRC staff has performed an evaluation of the Three Mile Island Unit No. 2 pressure vessel and determined that due to the small effects of radiation during the first fuel cycle, the allcwable stress limits are not reduced
- elow stresses resulting from overpressure evants limited by safety valve set points with the vessel at ambient temperature.
This evaluaticn provides the princioal basis for ccncluding that an overpressurization event during the first fuel cycle would not present an undue hazard relative to vessel failure.
'r I
i
=,
1
(/
t.
. The apolicant has provided a plant redesign incorporating a dual range setpoint for the pressurizer relief valve. During cooldown from hot shutdown, the NDTT mode for operation of the relief valve would be selected when the reactor fluid temperature is cooled to 275 F and the primary coolant pressure is below 450 psig. When in this mode, the relief valve would open should the pressure exceed 500 psig and the primary coolant temperatut es 0
remain below 275 F.
The applicant has evaluated this system considering scven different events representing the thirty events experienced in various PWRs.
The analyses were performed with code DYSID. Credit was taken for administra-tive procedures recuiring either a steam or nitrogen bubble in the pressurizer at all tinies. Credit was also taken for the pressurizer high level alarm and low level interlock to maintain the water between specified level limits. Tht '!sults of the analyses indicated that reactor system pressure would not exceed 500 psig during any of the events.
The staff has reviewed the dual set point design ad the results of the analyses to determine if adequate protection is provided through the life of the plant.
The design does not meet the single failure criteria because only a single relief valve has been provided. Also, the code DYSID has not been reviewed by the staff.
The long term solution, which must be imolemented prior to the second fuel cycle will recuire staff review and approval of the code OYSID and modifi-cations to the present design to meet all of the requirements identified below.
l D
\\le**
e.
1.
Credit for crerator action.
No credit can be taken for operator action until 10 minutes af ter the operator is made aware that a transient is in progress.
2.
Single f ailure criteria. The pressure protection system should be oesigned to protect the vessel, given any event initiating a pressure trnnsient. Redundant or diverse pressure protection systems will be considered as meeting the single failure criteria.
3.
Testabili*.z Provisions for periodic testing of the overpressure protecticn system (s) and components shall be provided. The program of tests and frequency or schedule thereof will be selected to assure functional cacability when required.
4 Seismic design and IEEE 279 criteria.
Ideally, the pressure protection system (s) snould meet botn seismic Category I and IEEE 279 criteria.
Tne basic objective, however, is that the system (s) should not be vulneraole to an event which both causes a pressure transient and causes a failure of equipment needed to terminate the transient.
5.
Reliability.
The system (s) provided must not reduce the reliability of tne emergency core cooling system or residual heat removal system.
~
r.
's
\\l L.
TMI-2 SER Supplement 2 6.2 Net Positive Suction Head Assessment In lieu of performing preoperational tests in the recirculation mode in conformance with Regulatory Guide 1.79, the applicant performed a separate scaled test series to demonstrate vortex control and pressure drop characteristics in the containment building sump area. The experimental pressure drop characteristics thus obtained were tnen ccmbined with system pressure drop characteristics obtained from preoperational tests drawing water frca the SWST to provide an experimental basis to assess the as-built system performance.
Losses associated with the screens and sump were higher than credicted in design.
Piping losses were generally less.
Table 6.2-11 of the FSAR presents a tabulation of available and required NPSH values for various operating modes and conditions.
The required NPSH values for the pumps are based on test values.
The available NPSH values are based on a conservative inter-pretation of experimental results.
The staff concludes that the 4.63 ft NPSH margin for the decay heat removal pumos will assure reliable operation in the recirculation mode.
T i/
<.,L I
TMI-2 SER SUPPLEMENT 2 5.3.3 Makeuo Tank Isolation In the event of an ESFAS signal, the operating makeup pump continues to function in the ECC mode as a high pressure injection pump.
Following procedure, the operator closes an isolation valve between makeup tank and makeup or HPI pump as a first step following ECCS actuation. The staff evaluation indicated that,1) if the operator closed the valve too quickly, it was possible for the pump to cavitate because the water supply frcm the SW1T had not been established, 2) if the operator failed to isolate the makeup tank, there was the possibility of hydrogen cover gas being drawn into the pump and reactor.
The apolicant has commited to automate isolation of the makeup tank prior to second operating cycle. He will provide redundant, safety grade va.ves with a closure time selected to assure that flow from the SWST has been established before isclation. Auto-
.ic isolation of the makeup puma recirculation lines will also be provided to assure full HPI flow.
Procedural and system changes have been made to assure safe operation during the interim period. Redundant safety grade valves have also been installed in the makeuo tank hydrogen cover gas charging line.
This will prevent overpressure or continued pressurization of the makeuo tank, should a LCCA occur while charging the makeup tank.
iIe er f
, L
_2 Analysis has been cresented which demonstrate that if the tank is procerly cnarged eacn time it is refilled, isolation would nc t be recuired to prevent hydrogen outflow.
The analysis spanned the period frcm ECCS actuation unt:: Af ter switchover to the recirculation mode (35 minutes or more after the break).
Procedures and 'imits for filling and pressurizing the makeup tank concitions are consistent with the assumptions used in the analysis.
Closure time for the makeup tank isolation valve and opening time for the SWST isolation valves are identical. Should the ccerator initiate closure of the makeup tank isolation valve concurrently with the ECCS actuation signal, there would be no interrupticn in the wa*a-tuacly to the pumo.
Based on system changes, acministrative procedures, and the time available for operator actions, the staff concludes that the isolation of the makeup tank following a LOCA is acceptable during the first fuel cycle. Modification of the isolation system to be incorporated prior to the second operating cycle is acceptable.
A g
h
\\t
- e..
TMI-2 SER SUPPLEMENT No. 2 15.2.2 Steam Line Break Analysis In order to justify plant operation until the first refueling, the acplicant has provided analyses of the steam line break applicable to the first fuel cycle when a 2".iK/" minimum shutcoun margin can be maintained.
No credit was taken for the operation of nonsafety-grade equipment in the secondary systen to raitigate the consequences.
Two periods during tne transler.t were identified as critical l
relative to fuel damage. During the first period, from 0,8 seconds to 5 seconds after the break, a portion of the core departs from nucleate boiling until reactor power is significantly reduced by control rod insertion. Later in the transient, the continued cool-down of the moderator causes return to subcritical power. Baron injection from the core flooding tanks terminates the loss of sub-critical margin.
Conditicns for departure frcm nucleate boiling early in the transient were maximized by assumotions regarding first fuel cycle reactivity feedback, initial steam generator inventory, break size and break location. Maximum runaut feedwater flow and 2 second initiation of emergency feedwater also act to maximize the rate of heat transfer from the primary system.
{/*M e
v
- c.,
s The reactor was assumed operating at 102% power at break initiation.
Selection of the most adverse (beginning of first cycle) reactivity feedback parameters permitted a slight power escalation to occur during the early portion of the transient.
Reactor coolant pumps were assumed to start coasting down at break initiation. Analyses performed with the assumptions outlined above led to less than 3% of the fuel experiencing DNB during the #irst five seconds.
The return to power period of the transient was analyzed for the case with the reactor coolant pumos continuing to operate as well as assuming a four-pump coast down coincident with the break.
End of first cycle and moderator temperature reactivity feedback parameters were assumed.
Control rod worth was based on the minimum shutdown rod worth of 2%
at hot shutdown conditions with the most reactive control rod stuck out.
Conservative assumptions were made relative to boron injection from the core flooding tanks and high pressure injection pumps.
For the period following reactor shutdown, the minimum DNSR was 2.09 occurring at 28 seconds for the case of continuing pump operation.
For the reactor coolant pump coastdown case a minimum DNBR of 1.93 occurred at 65 seconds. The staff concludes that no additional fuel failure would be expected during the return to subcritical power period.
An investigation was made to assure that fuel experiencing DNB during the initial cortion of the transient would not expcrience fur 2*r damage during the return to power.
It was determined that DNS would occur only on the upper portion of the rods during the early transient leading C
to peak clad temperatures of 1260 F at 8 seconds. The cladding temperature
.-,o
}/
c.
. decreased thereafter due to convective cooling. Rewet was not assumed. During the return to power, axial peaking occurred some six feet lower in the core with little or no effect on the previously heated regicn. No additional fuel damage would be expected during the return to power.
. It was determined that a ccolable geometry would be retained for the fuel rods entering DNS. The calculated peak cladding temperature of 1260 F coupled with the low differential pressure between the rod internal pressure and the primary coolant, precludes massive disruption or disintegration of the cladding due to oxidation embrittlement, bursting, or other potential failure mechanisms.
Three computer codes were used to accomplish the analyses. The PD007 code was used to determine reactivity feedback parameters.
Systems analysis was performed with TRAP 2.
Core thermal and hydraulic calculations were cerformed with RADAR. Several iterative interactions were reouired to converge upon the results presented.
The RADtR and PDC 07 codes hat + oeen reviewed and a0 proved by the staff.
The TRAP 2 code is currently urder review by the staff.
It is not expected that any method revisions required by the staff will dis-credit the overall conclusions derived from these analyses. However, any method revisions required by the staff review will be applied to this plant.
-. N
~
\\[
~ U*
I
I 1
Ine staf r.as :Oncluded that car.se.ative analyses nave cemonstrar.ec tnat less than three percent of tne fuel will experience damage.
The damage would not be expected to result in loss of ccolable geometry. An indeoendent evaluation by the staff concluded that radiological consequences of this accident are acceptable.
This determination applies to operation only until the first refueling.
'.5.2.2.1
,Se:oncar. Syste-5;p -
The applicant's program for modification of the secondary system prior to start of the second operating cycle is provided in reference 2.
The present design of the secondary system does not include safety-grade ecui; ment to mitigate the consequences of a steam line break.
In the previous section, it was justified that such equipment was not required during the first fuel cycle, pr'marily due to reactivity control available during the first fuel cycle, i
e 4
iI i,
e
- The plant cnanges as currently proposed by the applicant would provide redundant safety-grade feedwater isolation valves located in seismic Category I areas. Whether check valves currently in the feedwater lines will be replaced or retained has not been determined.
The staff concurs that the feedwater isolation valve change will satisfy staff requirements as identified relative to Issue 1 in NUREG-0138.
The final design must be in accorcance with applicable seismic and quality group requirements, and with IEEE-279 and IEEE-308 requirements.
New safety analyses will be required for both the steamline break and feedwater line break. These must demonstrate that the modified system will mitntain the consequences within acceptable limits under the most severe combinations of assumed conditions. The reanalyses will apply to secondary system breaks for the remainder of the life of the plant.
The staff concludes that the modification program will lead to an acceptable secondary system design. Staff review and approval will be required relative to the analyses and the classification of systems and components. We require that this program be completed prior to start of the second operating cycle.
-m O
h
. 15.2.2.2 Lono-Term Cooling Following A Steam Line Sreak The ability to achieve a long-term cooling condition following a steam line break was investigated during the review.
Detailed analysis of events through 75 seconds has indicated that conditions critical to fuel failure occur prior to 75 seconds. As noted in Section 15.2.2, the period from 0.8 to 5 seconds was the only time period for which fuel failure was predicted.
Conditions later in the accident when the minimum subcritical margin occurred did not produce additional fuel failure.
The applicant discussed conditions after 75 seconds, noting the following:
1.
High pressure injection would result in the reactor coolant i
system becoming water solid 10 to 15 minutes after tne break assuming no operator action is taken.
2.
Thert is the possibility that natural circulation flow would be lost due to voiding in the reactor coolant piping.
It would be expectcd that for the period between 75 seconds to 125 seconds the secondary system will continue to blow down and the primary system will continue to depressurize. With MSIV closure at 125 seconds, the unaffected steam generator is isolated and will begin to repressurize since auxiliary feedwater and heat rejection from the primary system would be available. The repressurization would continue slowly until the secondary system safety valves begin to open. After 125 seconds, the primary system slowly 0
0 0
reheats from about 400 F at the rate of 2 to 3 F per minute. When the system becomes water solid at 10 to 15 minutes, the pressure will
.j
,n-I/
6.
e
. rise quickly to 2500 psig (safety valve set point) and continue to heat up.
Both the primary and secondary systems will continue to release mass and energy frcm their respective safety valves until the operator terminates HPI flow and the heat release from the secondary system exceeds the decay heat.
The staff has considered the possible loss of natural circulation and agree that no ccditional fuel failures are likely as the core remains covered. Reactor coolant pump coastdown and circulation resulting from HPI injection would also contribute to maintaining adecuate cooling.
The pressure in the primary system reaching the safety valve set point value and the system pressure maintained at that value for a period of time poses two concerns which are not of immediate concern but will require additional studies.
Since the vessel is at a relatively low temperature when the system is repressurized, the reactor vessel material has less toughness at these lower temperatures and as such is more susceptible to failure at lower temperatures than at normal operating temperatures. The applicant has confirmed that the fracture toughness requiremer. s specified in Accendix G would be satisfied for any steam line break transient during the first five years of cperation.
The staff is proceeding with a generic review of reactor vessel pressure transient protection and a PWR main steam line break core and primary coolant boundary response.
It is planned to ccmplete these generic reviews by August 1979. On this basis, any addi donal requirements for plant modifications or changes in operating procedures m,-
l/
.L,
~
_8_
would be identified at least three years before the applicant's conclusions regarding reactor vessel fracture toughness analyses would have to be reconsider.c.
A second consideration is the large pressure gradients (s2500 psi) across the steam generator tubes than can result as a consequence of a steam line break. The staff has considered the effect of this pressure drop on tube leakage and integrity.
It has been determinec that increased pressure would not affect dose calculations and that a single through-wall crack which would permit a one gpm leak would not propagate to tube failure.
The staff concludes that long-term cooling can be achieved following a steam line break.
In addition, the applicant has agreed to reevaluate his conclusions regarding reactor vessel toughness and the staff expects to complete its generic review of vessel fracture toughness well in advance of any significant changes in fracture toughness of of the Three Mile Island Unit 2 vessel.
15 2.3 Feedwater Line Break The applicant has analyzed events assuming loss of a feedwater and the case of a feedwater line break inside of containment. A feedwater line break outside of containment was considered representative of the loss of feecwater event since check valves provided in each line would limit backflow to the break. The results of the analyses for both events indicate adequate margin relative to potential fuel damage or overpressurization of the primary system.
m7 ll ba t
k
_g.
A series of design changes have been made in the secondary system directed toward mitigating the consequences of a steam line break.
Although the feedwater line breaks have not been reanalyzed considering these changes, the staff and applic3at have reviewed the impact of such changes on the original analyses.
.: has seen determined that the changes would not make consequences more severe.
The mcdifications to the secon.:ary !
tem tr - implemented before the start of the second fuel cycle (15.2.2.1) could affect the plant response to feedwater line breaks, especially if check valves are removed. The timing of isola on valves could also affect the consequences.
The staff has concluded that the feedwater line break analyses provided in the FSAR are acceptable for the first fuel cycle.
Prior to the second fuel cycle, reanalyses of the modified system must be provided which demonstrate that feedwater line break consequences are acceptable relative to fuel damage and primary coolant boundary integrity.
O f
'N l/
V'u