ML20209G246
| ML20209G246 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 01/29/1987 |
| From: | Gallagher J PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | Butler W Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 8702050314 | |
| Download: ML20209G246 (13) | |
Text
r-4 PHILADELPHIA ELECTRIC COMPANY 2301 M ARKET STREET P.O. BOX 8699 PHILADELPHIA. PA.19101 1215l 8415001
""" ".f.ttt0?".'"
January 29, 1987 Docket No. 50-352 Mr. W. R. Butler, Director Project Directorate No. 4 Division of Boiling Water Reactor Licensing Attn: Document Control Desk U.S. Nuclear Regulatory Commission Washington, D.C.
20555
Dear Mr. Butler:
The purpose of this letter is to provide additional information for Philadelphin Electric Company's November 17, 1986 Application for Amendment of Facility Operating License NPF-39 which requests changes to the Limerick Operating License and Technical Specifications relating to operation with partial feedwater heating and increased core flow.
The additional information was requested by the NRC staff during a telephone conversation on January 7, 1987.
The questions are restated, followed by our response.
Question 1:
NEDC-31323, Section 5, Tables 5-1 through 5-4, detail feedwater nozzle and feedwater sparger fatigue usage for FFWTR and FWHOS for a 32-year thermal sleeve seal refurbishment.
What are the fatigue usage figures for a 28-year refurbishment?
\\
bg G702050314 B70129 PDR ADOCK 05000352 P
January 29, 1987 Mr. W. R. Butler Page 2 Response 1:
The fatigue usages factors for 28-year seal refurbishment are:
FWHOS FFWTR Case #1 Case #2 Nozzle 0.8094 0.6374 0.7040 Sparger
<0.8094
<0.6374
<0.7040 Question 2:
NEDC-31323, Section 5, Tables 5-1 through 5-4, are based upon How are the 26.6 cycles calculated?
26.6 cycles.
L Response 2:
On page 5-2 of NEDC-31323, it is stated that an 18-month fuel Thus, a 40-year plant life cycle is used in the analysis.
corresponds to 26.6 cycles.
Question 3:
consider feedwater NEDC-31323, Section 5, Tables 5-1 through 5-4, Why nozzle and sparger usage for FFWTR and FWHOS individually.
aren't they combined?
Response 3:
The format of the report is such that the effects of FFWTR However, their combined and FWHOS are determined separately.
effects can be determined by adding the per cycle fatigue For example, usages from the appropriate tables together.the damage due to one cy for the nozzle is 0.0270 + 0.0055 = 0.0325 additional usage for this cycle.
Mr. W. R. Butler January 29, 1987 Page 3 Question 4:
What type of feedwater nozzle thermal sleeve arrangement does LGS have?
Response 4:
The thermal sleeve arrangement is the triple thermal sleeve.
Question 5:
How does PECo track the 28-year seal refurbishment?
Response 5:
The feedwater nozzle thermal sleeve refurbishment is a function of nozzle fatigue usage factors as a result of RPV thermal cycles.
The 28-year seal refurbishment is based upon a postulated number of thermal cycles occurring during the 28-year period.
LGS monitors ECCS system injections into the reactor coolant and RPV thermal cycles pursuant to Technical Specifications 3.5.1.f and 5.6.1.
Surveillance tests are completed monthly and quarterly to monitor and report ECCS vessel injections and RPV thermal cycles, respectively.
Question 6:
NEDC-31323, Section 5, paragraph 5.3, discusses feedwater sparger fatigue.
Paragraph 5.3 states in part that the fatigue results are based on assumed conservative leak rates.
What are the leakage rates?
Response 6 The leakage rates vary with several factors including flow rate and gap size.
The leakage is calculated based on a model developed by GE based on tests done on the feedwater nozzle /sparger final design.
The documentation for these tests and their results are given in NEDE 21821-02, Supplement 2, " Boiling Water Reactor Feedwater Nozzle /Sparger Final Report," March 1979.
The method for calculating and using the leakage rates is given in Section 4.7.2.4, "High Cycle Fatigue Damage Evaluation" of NEDE 21821-02.
This j
report is included in the Limerick FSAR Section 3.9 Reference list.
Mr. W. R.
ButlGr January 29, 1987 Page 4 Question 7:
NEDC-31323, Section 4, page 4-1, states that ICF vibration analysis was performed by analyzing the startup test vibration data for the valid prototype plant (Browns Ferry 1).
The LGS FSAR states that Browns Ferry 3 was the LGS prototype plant.
Please comment.
Response 7:
In NRC letter from Tedesco to Sherwood, " Acceptance for Referencing Topical Report NEDE-24057P Assessment of Reactor Internals Vibration in BWR/4 and BWR/5 Plants," October 28, 1980, Browns Ferry 1 is listed as the valid prototype plant for Limerick 1 (Table 3, Page 7).
Browns Ferry 1 is also listed as the valid prototype for Browns Ferry 2 and 3 (Table 2, Page 6).
The LGS FSAR will be revised to reflect this information.
Question 8:
NEDC-31323, Section 4, discusses the flow-induced vibration response of the reactor internals.
Please provide additional detail on the subject.
Response 8:
Since the LGS Unit 1 plant did not have an instrumented vibration measurement program, the prototype (Browns Ferry 1) plant test data were used as the bases for this vibration assessment for plant operation in the ICF region.
The Browns Ferry 1 test results are described in NEDE-24057-P-A.
Results from the Browns Ferry 1 test program include data up to 113% core flow.
The balance flow test measurements of all sensors in all instrumented reactor internal components were reviewed, evaluated and compared with the acceptance criteria.
The absolute sum of the peak alternating stresses of all the vibration modes was obtained.
This is a very conservative method to evaluate vibration impact because it assumes sustained vibration at an alternating stress with an amplitude corresponding to the absolute sum of all vibration modes while the actual vibration is random and the amplitude seldom reaches this absolute sum value.
The maximum stress calculated following this method is 61% of the acceptance criteria of 10,000 psi peak stress.
This was determined for a jet pump strain gage at 113% of rated flow.
Hence, the data show that reactor internals response to flow induced vibration is within acceptable limits up to 113% core flow.
Mr. W. R. Butler January 29, 1987 Page 5 Fuel channels were not instrumented at the Browns Ferry plants.
The Limerick rated flow per fuel bundle is less than the Fitzpatrick rated flow per fuel bundle.
Hence, the Fitzpatrick fuel channel test results can be applied to Limerick.
An assessment based on Fitzpatrick test data, described in NEDE-24057-P-A, shows that the operation of LGS Unit 1 at 105% of rated core flow will not result in unacceptable fuel rod or fuel channel vibration for GE fuel.
This is based on Fitzpatrick data which shows that the maximum recorded vibration of the fuel channels was less than 2% of the allowable for conditions corresponding to at least 128% of rated flow.
Question 9:
Are the design limits for stresses in the reactor internals equivalent to the ASME Code,Section III, Subsection NG as stated in FSAR Question 210.86?
Response 9:
As stated in the response to FSAR Question 210.86, the design basis for LGS reactor internals is an earlier version of the ASME Code which was shown to be equivalent to the limits of Subsection NG of ASME Section III.
This response was reviewed and approved by the NRC staff in March 1983.
Subsequently, in late 1983, GE performed a New Loads (hydrodynamic loads) update of the LGS mechanical equipment and components including the internals.
The allowables of Subsection NG were used as the criteria for the load adequacy evaluation of the limiting reactor internals.
As shown in Table 3.9-6(b) of the FSAR, the selected internals with limiting loads do satisfy the criteria of Subsection NG.
Despite this New Loads update, the statement of Q/R 210.86 is valid for the design basis.
In 1986, GE performed ICF and PFH studies which included re-evaluation of the load adequacy of the reactor internals under the consideration of additional loads imposed by the ICF and PFH operations.
The conclusion, as stated in NEDC-31323, is that the Subsection NG design allowables in the FSAR are satisfied.
4 1
Mr. W. R. Butler January 29, 1987 Page 6 4
Question 10:
Are the stresses produced on the reactor internals by ICF and/or PFH operation within the design basis?
4 Response 10:
The stresses produced on the reactor internals by ICF and/or PFH operation are below the Subsection NG allowable limits.
Question 11:
Provide additional information regarding the effect of increased i
core flow and reduced feedwater temperature on Limerick LOCA analyses.
i Response 11:
Figure 6.3-10 of the Limerick FSAR shows the calculated results for Peak Cladding Temperature (PCT) versus break area at the bounding conditions of 105%' steam flow /100% core flow.
The DBA recirculation suction line break was found to be the limiting break.
The DBA suction break experiences the most severe early blowdown and also has the longest period of core uncovery (FSAR Figure 6.3-74).
For operation at Increased Core Flow (ICF), a plant specific Limerick calculation shows no change in the high power node dryout time for the DBA.
For other break sizes, the change in dryout time will also be small (less than 0.2 sec), with a corresponding PCT increase of less than 5 degrees F.
Figure 3
1 shows the effect of ICF on the calculated core uncovery duration as a function of break area for several plants.
In the range of 60-100%'DBA, it shows a maximum impact of less than 2 seconds.
For the Limerick DBA case, the increase in core uncovery duration was 1.6 seconds.
From the Limerick i
FSAR analysis, the high power node heatup rate at the time of
{'
reflooding was calculated to be 5.4 degrees F/sec.
This results in a PCT increase of about 9 degrees F for the limiting break.
The primary effect of increased core flow on LOCA analysis is to alter the flow coastdown response following the assumed trip of the recirculation pumps at the time of the break.
j Increased core flow also results in a small decrease in the downcomer and lower plenum subcooling, resulting in a net i
decrease of 0.1-0.2% in the initial system mass and in the l
initial flow rate out of the break.
The flow coastdown is i
i
. --,-,,-- -- ~
.-..n,.--.
--,-,---~-nw.
,n,-,-,--
- -,, -~
- - - - - - - ~ ~ -
Mr. W. R. Butler January 29, 1987 Page 7 important only during the early blowdown period.
The higher initial core flow makes the early boiling transition less severe in the top portion of the bundle.
At the high power node, where-the PCT is calculated to occur, the dryout time occurs shortly after jet pump uncovery.
This node generally shows a small (<0.2 sec) change in dryout due to ICF which tends to be slightly earlier.
The earlier dryout results from the lower initial system mass and the pumping of more water from the downcomer into the lower plenum during the recirculation coastdown period.
Changes to the duration of core uncovery for ICP are also small, typically 1-2 seconds for large breaks and calculated to be 1.6 seconds for Limerick DBA (see previous paragraph).
The core uncovery duration primarily depends on total vessel inventory and ECCS makeup capability.
The small decrease in initial vessel inventory with ICF leads to the slightly longer core uncovery duration.
This small change is also consistent with single loop operation LOCA analyses which show relatively small changes in core uncovery duration despite a much larger change in initial core flow.
Figure 2 presents the effect of reduced feedwat9r temperature on calculated PCT.
This figure is based on calculations performed for four plants and all cases show a PCT decrease with reduced feedwater temperatures.
Two of the comparisons at 65 degrees F reduced feedwater temperature were performed at increased core flow conditions, while the other points were done at rated conditions.
All of the comparisons at 65 degrees F reduced feedwater temperature were performed for plants (2 BWR/4s and a BWR/5) which are similar to Limerick (BWR/4-251).
Thus, these results are applicable to Limerick and it is concluded that operation with reduced feedwater temperature is bounded by operation at rated feedwater conditions.
Operation at reduced feedwater temperatures results in increased subcooling in the vessel which increases the mass flow rate out of the break for a given vessel pressure.
l However, the initial vessel pressure is lower (1015 psia i
versus 1055 psia for FSAR and ICF) when operating at reduced
~
feedwater temperature due to the reduced steam production.
i This low initial vessel pressure results in a decrease in the j:
calculated break flow rate relative to operation at rated feedwater temperature.
Operation with reduced feedwater temperature also increases the total system mass by about 1%
(due to increased core inlet subcooling) and delays the time of lower plenum flashing.
The decreased break flow tends to make the calculated LOCA response less severe, and the increased total system mass and delay in lower plenum flashing are both beneficial.
The increased system mass and the decreased break flow combine to result in later jet pump, break, and core uncovery times.
i i
I Mr. W. R. Cutler January 29, 1987 3
Page 8
.]
i
-Operation with reduced feedwater temperatures has several advantages over rated feedwater temperature operation in the removal of energy following a LOCA.
First, the dryout time occurs later (by about 0.2 to 0.4 seconds for a'65 degree F feedwater temperature reduction) due to the delay in jet pump 4
l uncovery caused by the increase in total system mass.
This allows more stored energy to be removed from the fuel prior to dryout.
The' calculated PCT is more sensitive to changes in dryout time than to any other single parameter; so this is i
the most significant effect of reduced feedwater temperature operation.
l l'
Second, the time delay between dryout and lower plenum flashing also increases slightly.
Lower plenum flashing occurs very early (about 8-10 seconds) into the event when 1
stored energy levels are still relatively high and the fuel l
has only been heating up for 1-2 seconds.
The delay in lower plenum flashing relative to time of dryout is beneficial because it allows the hot bundle to heat up more before flashing occurs.
This creates a larger temperature 4
i difference between the fuel rod and the flashing fluid which t
}
makes the heat removal process more efficient.
The amount of fluid which flashes is slightly higher for the case with reduced feedwater' temperature because of the increased j
initial mass in the lower plenum.
The flashing fluid is also j
at a lower temperature which further improves the heat i
removal capability. -Thus, more heat is removed from the fuel l
by the flashing process'for operation at reduced feedwater j
temperatures.
The delay in lower plenum flashing is relatively small and is not expected to have a significant l
effect on the calculated PCT.
I 1
i A third area where energy removal is improved by reduced feedwater temperatures is in the time of core uncovery.
The i
core uncovery occurs late (by about 1.5 seconds for a 65 r
degree F feedwater temperature reduction) due to the increase in total system mass.
The delay in core uncovery again allows more energy to be removed prior to the start of the final core heatup phase.
Decay heat levels will be slightly lower during the core uncovery period with reduced feedwater I
temperature operation resulting in a slower core heatup.
As a result of these combined effects, the peak cladding l
i temperature for operation with reduced feedwater temperatures
[
i is lower than when operating at rated feedwater temperatures.
l l
Figure 3 shows the calculated high power node uncovery i
duration versus break area for Limerick under FSAR and ICF j
j conditions.
The DBA has the longest uncovery duration in both cases.
Since the DBA also experiences the most severe early' blowdown, it is clearly the limiting case for ICF 3
operation.
As stated earlier, the calculated PCT increase l
I associated with Limerick ICF is 9 degrees F for the limiting
[
i
Mr. W. R. Cutlor January 29, 1987 Page 9 break.
The PCT curves for ICF and ICF plus PFH are shown in Figure 4.
If you have any questions or require additional information, please do not hesitate to contact us.
Very truly yours, Attachments cc:
Dr. T. E. Murley, Administrator, Region I, USNRC E. M. Kelly, Senior Resident Site Inspector Mr. Robert Martin, LGS Project Manager See Attached Service List
m e.
k 0
j 0
4 r
K 1
C RI E
M IL A
3/
0 R
8 W
B 9
WO LF
)o A
E B
R D
1 O C F
O D
EE S
(
E A
UR E
C R
Gf A
1 I
K IFF A
1 O
E 5 R 2 T
0 D-C D b 6
5 EF
/
F R
E W
B e
+
0 1
4 52 2
5 i
5 0
5 5
1 1i 0
0 4
1
/RW B
mO bv ZO-t>O >5yO9_3 Z hS<I O
l
O O
9 0
O M
N I
O e
O N
N eg c2 O
_ q 4
v-4 O
- rD v-
_ O o
C'}
H l
0 m._
c.
L..'
O N
Z
-N VE
.- 5 C
v-c.
N3 U
0 O
c0 L.L] C
- O w
e yy c_
]4 2
C y
u.
_ O U
a e
b h
o
<S HU e
o Owe
- h wW mN u.
I w
+
~
\\c3
- O c2 y
+
O
_ N h
l h*
h l
h h
h l
h h
h l
h C
N
+
to c3 O
CJ 4
(O cfJ O
e4
+
(D C2 O
v-e-
e-e-
CJ N
N N
CJ M
l l
l l
i i
i l
i I
I I
I I
I m
N I
sn (J ) 3DNYHO 10d D3 to D
I e
F (M
~-
a
~~
S.
L3 74 140
.g 123 2"
E
$ sco u
W g
s 3a w
30 l
1.0 M s,..k Ee E
[.'
2 so
!s _,
s
,I x
5 C %CY Ef 60
~
~ FSAR AMA L.t'S e s E
g U
30
~
p f
i i
t i
i I
O I
O 10 23 33 as 33 to 70 to to SCO e
8 ALkK ARtA 1%of f BM 3 < sue LIMERICX GENERATING STATION UNITS 1 AND 2 IINAL SAFETY ANALYSIS REPORT TOTAL TIME FOR WHICH HIGHEST Figure 3 POWERED NODE REMAINS UNCOVERED VS. BREAK AREA (FAILURE OF DIVISION 2 DC SOURCE)
FIGURE 6.3.74 R EV. 30,03/84
-I
~
4 e.
., S a.3 te 4
suaLLsasaa'usTMoo I unca saman ustwoo I
g
- c; I
i I
i l
- 7 + 79 tame r
g 7
L anaximuu c=mt i
3paar uma smaag E
i' l
E I
a
- i o
I h
pasimum sTramuwe sataa tcursion couransmamrt j!
I g
8 l
pani =uu sTramuna saux g saa.
g onsioscomrai==amTi g
pas:maran satAs l
(~..
l n
i l
is C
em em as em si sa sa u sassa n's 22aso.ot e
LIMERICK GENERATING STATION UNITS 1 AND 2 FINAL SAFETY ANALYSIS REPORT Figure 4 PEAK CLADDING TEMPERATURE AND MAXIMUM LOCAL OXIDATION VS BREAK AREA o
(FAILURE OF DIVISION 2 DC SOURCE)
FIGUR E 6.310 R EV. 30,03/84
..