ML20215L231
| ML20215L231 | |
| Person / Time | |
|---|---|
| Site: | Grand Gulf  |
| Issue date: | 09/30/1986 |
| From: | Bost D, David Jones, Mashburn W MISSISSIPPI POWER & LIGHT CO. |
| To: | |
| Shared Package | |
| ML20215L208 | List: |
| References | |
| NUDOCS 8610280538 | |
| Download: ML20215L231 (13) | |
Text
- _ _ _ _ _ - . - _ - - _ - . - - . - - - _ _ . . _ - _ - - . - - - - _ _ _ - . _ -
GGSS SRV FATIGUE EVALUATION Prepared by !!ississippi Power and Light Nuclear Plant Engineering Revision 0 September, 1986 s ')
Prepared by: /4 2g D. PT Jones,. W / /
Reviewed by: Mw [/ /o/3/g/ if'
- N D.G/Bost, PE / Hi toswami, PE Reviewed by: L .3 I ok[g W. F. >1ashburn, PE Principal Civil Engineer Approved by: /0 y F.'W. Titus, PE Director, Nuclear Plant Engineering D
$ N"EUEEER Res8gg,,
PDn
i a 11 INTRODUCTION Condition 2.c (10)(a) of Facility Operating License No. NPF-29 for Grand Gulf Nuclear Station, Unit I, states in whole:
" Prior to start up following the first refueling outage, MP&L shall complete any modifications or replacement of equipment found necessary as a result of the fatigue evaluation. In the interim, MP&L shall document the occurrence of every safety relief valve (SRV) actuation into the suppression pool; the associated cumulative damage factors shall be calculated for typical representative equipment and kept up-to-date; and MP&L shall report to NRC any malfunction of equipment that occurs due to any safety relief valve discharge".
This report outlines the manner in which this condition has been satisfied and draws conclusions on the effects of SRV-induced fatigue loads on safety related equipment during the design life of Grand Gulf Nuclear Station, Unit I.
i l
m
. s-1 1.0 MONITORING 0F SRV LIFTS PRIOR TO RF01 Pursuant to the requirements of GGNS SSER2, Section 3.10, (the ori0 in of License Condition 2.c (10)(a)) which requires the documentation of every SRV discharge into the suppression pool, PMI 83/8729 requested plant operational personnel to ensure prompt NPE notification of all discharges.
In response, NPE periodically received notification whenever SRV discharges occurred. NPE logged each discharge into the SRV Actuation Log and attached the correspondence through which notification was made. The SRV Actuation Log is reproduced in Table 1.
2.0 FATIGUE EVALUATION APPROACH Three types of evaluations were performed to address the concern of SRV imposed fatigue on GGNS safety related equipment as listed below:
2.1 A large percentage of dynamically tested equipment underwent preliminary testing to obtain information on equipment resonances.
Analyses were perfomed to take credit for fatigue that was naturally imposed by these tests and comparisons were made with the expected level of SRV plant accelerations.
2.2 An in-plant SRV test was conducted at GGNS to determine the amount of conservatism built in to design spectra relating to SRV actuations. Response accelerometers were located at representative design spectra building model nodes and SRV's were actuated. Response spectra were measured at these locations and compared to the corresponding calculated design Required Response Spectra (RRS). In addition, actual stresses were measured during the SRV transient at critical points on representative equipment and the actual cumulative damage factors were calculated.
2.3 Calculations generated to qualify equipment by analysis were researched to determine the most critical dynamic stresses expected. In most cases, the load combination which included seismic stresses was most critical. This stress was conservatively assumed to be imposed by an SRV actuation and a 40 year projected cumulative damage factor was calculated.
3.0 BASIS FOR SELECTION OF EQUIPMENT TO BE EVALUATED In order to develop a list of equipnent that would be subject to the evaluation processes, a population consisting of all safety related containment /drywell equipment listed in FSAR Seismic Qualification Summary Tables 3.10-3 and 3.10-5 was reduced based on the following elimination process.
3.1 Equipment qualified by analysis was eliminated from the list.
This equipment was evaluated on a case by case basis as noted in section 3.4 for critical stresses that could conservatively be assumed to be induced by an SRV actuation.
i 2
3.2 A significant number of valves were static load tested assuming 69 horizontal and Sg vertical while nozzle forces and moments were applied. These items were eliminateu from consideration since no dynamic testing was performed.
3.3 Line-mounted equipment which had been dynamically tested was eliminated from consideration due to the considerable effort required to develop actual single valve RRS for each equipment class at the various locations.
3.4 Evaluation of equipment qualified by analysis was performed for items where critical dynamic stresses were readily determined.
These stresses were assumed to occur 1820 times, the expected number of SRV actuations, and 40 year projected cumulative damage factors were calculated.
4.0
SUMMARY
OF EQUIPMENT EVALUATED Research of safety related equipment listings indicated that approximately 944 individual safety related components were located in the containment /drywell area. It was further identified that 630 components were qualified by dynamic test, 90 components were qualified by static test and 224 components were qualified by analysis. This fatigue evaluation considered 480 of the 630 dynamically tested components and 55 of the 224 analyzed components. Figure 1 summarizes the percentage of equipment in the containment /drywell area that was evaluated.
Figure 2 shows the distribution of equipment by zone in the containment and drywell that was evaluated. The percentage value indicated for a particular area is the percentage of the total number of evaluated components (535).
Table 2 is a comprehensive list of evaluated equipment.
5.0 EVALUATION PROCEDURES AND RESULTS 5.1 Fatigue Imposed by Preliminary Testing: A method was developed in Reference 2 which makes possible the quantification of fatigue effects imposed by resonance testing (Reference 1). Such tests are typically performed prior to proof testing for the design f
basis seismic / hydrodynamic event. Resonance searches usually consist of sine sweep testing using a continuous sine wave.
Occasionally, sine beats or continuous sine waves were used to dwell on certain resonant frequencies prior to multifrequency tests. In a few cases, some additional random motion multifrequency testing was also performed. Since all three types of preliminary or additional testing have been performed, equations were derived for sine sweep tests, ;ine dwell tests, or random motion multifrequency testing.
For any of these type of tests, 3 variables associated with the test motion must be known, i.e., the peak acceleration level for
- sine tests or the equivalent white noise PSD level for random
! motion tests, any equipment resonant frequencies and the duration GGNS SRV FATIGUE EVAL
- 1 3
of the excitation at the frequency of interest. The procedure is to calculate an equivalent spectral acceleration level for each equipment resonant frequency given the level and duration of the input excitation at that frequency. Comparison can then be made between the calculated equivalent spectral acceleration and the corresponding design SRV spectral acceleration. If the ratio of the former to the latter is greater than or equal to unity, it can be stated that expected SRV fatigue effects were enveloped during the preliminary or additional testing and that tracking of cumulative damage factors for 40 years would not be necessary.
If the ratio is less than one, this does not indicate the equipment is inadequate to withstand the fatigue effects that will be imposed. It simply shows that preliminary testing was not adequate to simulate 40 years of SRV actuations and that some additional means must be utilized to directly demonstrate equipment adequacy. No multivalve actuations were used in the comparison since relatively few multivalve actuations are expected as compared with a large number of single valve actuations.
For the 2851 pieces of equipment considered, equivalent spectral accelerations were calculated at a total of 1180 natural frequencies. Of these, only the 2 calculated ratios for the Rod Position Multipl, exer Cabinets were less than one. Figure 3 plots a histogram of safety factors. All other equipment was shown to be adequate solely on the basis of preliminary testing.
5.2 SRV In-Plant Testing: Reference 3 discusses the significance of data gathered from the GGNS SRV in-plant test conducted to verify adequacy of plant design spectra and to determine fatigue effects on selected equipment.
In general, the test program showed that under 60 Hz, the frequency range of most concern for fatigue stresses, the measured response was typically an order of magnitude or more below predicted values.
As stated above, the preliminary testing performed during qualification of the multiplexer cabinets was not of sufficient duration to demonstrate adequacy of the equipment when subject to the postulated fatigue. This in no way identifies equipment inadequacy, but rather precludes formulation of a positive statement concerning fatigue adequacy. However, the SRV in-plant test demonstrated the presence of significant conservatism in applicable design spectra as discussed below. When an envelope spectra based on actual measured spectra in the area around the elevation 135'-4" floor was used in the evaluation, a safety factor of 1.8 was calculated. At this location in the plant, it was shown that design response values were about twice the measured response and that design peak acceleration values (ZPA) were at least 1.2 to 2.0 times greater than measured response.
1 480 total - (193 HCU's + 2 Hydrogen Recombiners): see section 5.2, paragraph 4.
7
= I 4
The range of magnitude of over prediction extended from this lower bound to a factor of 45 with an average over prediction factor of about 10. In addition, the equivalent spectral acceleration calculated for the Multiplexer Cabinets above was based on plant spectral accelerations at elevation 120.83 which are significantly larger in magnitude than elevation 147.58 spectra. These RRS locations are the closest to the elevation 135'-4" floor. Given these facts, along with the inherent design SRV RRS conservatisms noted above, it is judged that preliminary testing done on Multiplexer Cabinets adequately simulates actual expected SRV imposed fatigue.
The SRV in-plant test program also included the measurement of strains on some representative equipment to determine the levels of actual fatigue stresses. One HCU module at elevation 135'-4" and one Hydrogen Recombiner at elevation 208'-10" were chosen to be instrumented for the SRV test because they represented critical pieces of equipment and because of their vertical location in the plant. In all, 193 HCU's and 2 Hydrogen Recombiners are considered to be evaluated by virtue of the fatigue analysis based on SRV in-plant test data. Peak measured stresses were used to calculate the worst case cumulative damage factor for the HCU's and the Hydrogen Recembiners. A CDF of .036 was projected for 40 years of operation based on conservative assumptions made in selecting the maximum number of permissible stress cycles from the S/N curve, in establishing the expected number of significant stress cycles per SRV actuation, and in the use of fatigue strength reduction factors to account for potential material flaws. This demonstrates the adequacy of this equipment to resist SRV-induced fatigue stresses imposed at GGNS without failure through plant life.
5.3 CDF Calculation for Equipment Qualified by Analysis: A total of 55 components which have been qualified by analysis were evaluated by researching the applicable calculations for critical stresses.
It was found that this normally occurred for earthquake loads combined with operational loads. Only in the case of the MSRV Discharge Line Quenchers was SRV induced stress critical, as would be expected in this case. Critical loads as identified in the calculations were conservatively assumed to occur 1820 times and a CDF was calculated based on the resultant stress. No calculated CDF was greater than or equal to 1.0. Therefore, the evaluated equipnient can be expected to adequately withstand SRV transient induced fatigue.
6.0
SUMMARY
6.1 As required in Licensing Condition 2.c (10)(a), all SRV discharges into the suppression pool were documented in the event it became necessary to evaluate the ongoing acceptability of any equipment found critically sensitive to SRV-imposed fatigue failure. An SRV actuation log was maintained for this purpose.
5 6.2 A large percentage of safety related equipment that was subject to proof tests was also subjected to preliminary resonance testing or other additional testing. The extent to which this additional
~
testing could be considered to simulate GGNS SRV fatigue levels was evaluated. Less than 1% of equipment evaluated in this manner could not be completely justified on this basis alone. In this case,~ supplemental justification relating to the conservative application of building spectra and inherent spectral developmental conservatisms provided assurance that preliminary testing adequately simulates expected SRV-induced fatigue.
6.3 The GGNS SRV in-plant test provided information which demonstrates the conservative nature of plant design spectra. Building model node points were instrumented and SRV RRS were measured at these points and compared to calculated design spectra. In general, design spectra was shown to be conservative by an order of magnitude or more.- During the above test, an HCU and hydrogen recombiner were instrumented at known critical stress points to measure actual SRV stresses. Results of this effort showed that a design life CDF no higher than .036 could be expected on this equipment despite some very conservative assumptions regarding the number of significant stress cycles and in the use of material S/N curves.
6.4 In addition to the above, some equipment dynamically qualified by analysis was evaluated by calculation of CDF's based on the load combinations which produced the most critical stresses in the equipment. In general, seismic load cases controlled the designs.
Conservatively assuming that seismically induced stresses would act'with the expected SRV actuation frequency, no CDF's greater than 1.0 were calculated.
6.5 In no case did this study indicate a need for GGNS' hardware modifications due to fatigue imposed by SRV actuations.
7.0 CONCLUSION
S:
7.1 It is concluded that GGNS safety related equipment located in the containment will not be adversely affected when subject to repetitive SRV-induced dynamic loadings. This conclusion is based on results of the evaluation of the effects of preliminary qualification testing, the results of the SRV in-plant test, and calculation'of C0F's based on the controlling load combination for equipment qualified by analysis.
7.2 Positive results of all evaluations conducted show that tracking of CDF's at GGNS beyond the first refueling outage is not required ds a precautionary measure to preclude failure of safety related equipment in the containment. The terms of Licensing ' Condition 2.c (10)(a) and the requirements of SSER 2, Section 3.10, are considered to be satisfied by this report.
6 REFERENCES
- 1. NPE Calculation CC-Q1111-86009, Rev. O.
- 2. " Equipment Usage Factor for Equipment Qualification Testing", Nutech Report MPL-02-032.
- 3. " Grand Gulf In-Plant Safety Relief Valve Test Fatigue Evaluation Report", Nutech Report MPL-09-100
- 4. United States Nuclear Regulatory Commission Docket No. 50-416, Grand Gulf Nuclear Station, Unit 1, Facility Operating License, License No.
- 5. NUREG 0831, Safety Evaluation Report, Supplement 2.
TABLE 1 SRV ACTUATION LOG IR NO. DATE TIME VALVES 1 N/A 9/8/84 230 All 2 84-10-4 10/13/84 1349 51F 3 N/A 1/2/85 - 1/3/85 -
All valves once 51B twice 4 N/A 4/7/85 0958 51B, 51D 5 N/A 4/23/85 1814 41E 6 N/A 4/23/85 2157 51D 7 N/A 4/24/85 0323 47A 8 N/A 4/24/85 0615 47A 9 N/A 4/25/85 1204 47D, 41B, 41E, 47G 10 N/A 5/17/85 1955 47D, 47G, 51D, SIF, 51B, 51A l
11 85-8-1 8/7/85 1402 51A, 47G, 47D, 51B, 51F, 51D 12 85-8-5 8/16/85 1233 41C, 41A i
- 13 85-8-7 8/18/85 1200 51D 14 85-8-11 8/26/85 1800 51B 15 85-10-2 10/12/85 0900 41A, 41C, three times !
i 16 86-4-2 4/7/86 1912 51B twice 17 86-7-10 7/25/86 1520 51B, 51D l
l I
I NCIRPT SRV ACTUATION LOG
, 7 TABLE 2 EVALUATED EQUIPMENT Number of Components Spec./ System No. Evaluated Description E-035.0 48 Containment Electrical Penetration Assemblies J-301.0A 19 Rosemount 1153B Transmitters J-363.0 4 Containment Radiation Monitors J-561.0 58 Temperature Elements M-101.1 20 MSRV Discharge Line Quencher M-102.0 32 SRV, MSIV, ADS Accumulators M-190.0 2 Hydrogen Recombiners M-190.0 2 Drywell Purge Compressor Gearbox Housing and Base E61 2 Drywell Purge Compressors H22 12 Local Instrument Racks Multiple 28 Rosemount 1153D Transmitters Multiple 104 Rosemount 1151 & 1152 Transmitters C41 4 CR-2940 Switches B21 2 Robert Shaw Pressure Indicators H22 2 Rod Position Multiplexer Cabinets C41 2 Standby Liquid Control Pump C11 193 Hydraulic Control Units C41 1 Standby Liquid Control Storage Tank GGNS SRV FATIGUE EVAL
.- ,r-
SUMMARY
OF QUALIFICATION METHOD 8 EQUIPMENT EVALUATED 67 %
51 % OF TOTAL,
[ 77% OF CLASS A 8 g(vo i
, y a jl_ -
!b .
l
.g '9_
'5h . . . .
gf
- . . . * ? :- .
I
~+ .
24 %
i ,
f 6% OF TOTAL,
[
i 24% OF CLASS 9%
90/944 = 9% STATIC TEST I s
>21 224/944 = 24% ANALYSIS IJ ::1 630/944 = 67% DYNAMIC TEST l l FIGURE 1
.' . y
( ) NO. OF COMPONENTS IN ZONE
. % PERCENTAGE OF TOTAL EVALUATED POPULATION (12) 2%
208'
+ 7-??E - - _5 "~"^M" - -E-E - - - - - -- - - -Ex- "
,,gaua--- {- - - - -
="P C _ _
b b_
v '
"8 spi:
- _I85 Em ,,. ,:,:.
i>
emmmmm: 166' 2%
(448) biY f ffw . l47' 84'/o 13 5'
'(10) l
!* (24) 4%
l 93' l ZONE I IN-PLANT DISTRIBUTION l
ZONE 2 l EQUIPMENT
@ll[Si*i@i! ZONE 3 M ZONE 4 ZONE 5
[ ZONE 6 FIGURE 2 l__-__ - - . _ _ . _ . _ - - - - -
1 t
CALCULATED EQUIVALENT SPECTRAL ACCELERATION, i i
S. F. :
260 DESIGN SRV SPECTRAL ACCELERATION i
240 1
l' 4
y
~'
SAFETY FACTOR DISTRIBUTION
, 22o M w FOR 285 COMPONENTS WITH S
w 1180 NATURAL FREQUENCIES 200 g m
<r 180 $ FIGURE 3 U
z
! 160 u.
.i O I O 14 0 s
12 0 _
e i 10 0 80 60
]
40 ao "
l 1 "
r : -
.,- l:! S. F.
8 9 9 9 9 m
9 m
9 m
9 e
9 e
9 9 9 9 9 9 9 9 9 9 9 9--
m o e s e m o e o g o o g i
i l
i