ML20141E970
| ML20141E970 | |
| Person / Time | |
|---|---|
| Site: | San Onofre  |
| Issue date: | 05/16/1997 |
| From: | Rainsberry J SOUTHERN CALIFORNIA EDISON CO. |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9705210159 | |
| Download: ML20141E970 (13) | |
Text
p.4
'l
@fA)[ 'sof)IHf RN CAtsf ouMA
[* h l C Q
- j. L Rainsberry
.N L.L/s J tr Manager, Plant Licensing An (DISCV INTLk%AHOAAl* Carnpany i
May 16, 1997 U. S. Nuclear Regulatory Commission Attention: Document Control Desk Washington, D. C.
20555 Gentlemen:
Subject:
Docket No. 50-362 Steam Generator Tube Eggerate Supports San Onofre Nuclear Generating Station Unit 2
Reference:
Letter, J. L. Rainsberry (Edison) to Document Control Desk (NRC),
dated April 30, 1997: Steam Generator Tube Eggcrate Supports As requested by the NRC Project Manager for San Onofre Units 2 and 3,.in a May 9, 1997 meeting between Southern California Edison and the NRC Staff,
- e enclosed is an updated evaluation of the ability of the San Onofre Unit 2 steam generator tube eggcrate supports to perform their design function. This updated evaluation incorporates information on cause assessment and differences in operating histories between Units 2 and 3, which have been identified since transmittal of the original Unit 2 evaluation (referenced letter).
If you have any questions or would like additional information, please contact me, f
Sincerely, I
9705210159 970516 g
P9R ADOCK 05000362 s
p PDR Enclosure cc:
E. W. Merschoff, Regional Administrator, NRC Region IV K. E. Perkins, Jr., Director, Walnut Creek Field Office, NRC Region IV J. A. Sloan, NRC Senior Resident Inspector, San Onofre Units 2 & 3 M. B. Fields, NRC Project Manager, S&n Onofre Units 2 and 3 E. J. Sullivan, Jr., NRC Division of Engineering, EMC Branch Chief 209138 San Onofn Naclear Generating Station
-P.O,Ikn 128 ll l
l San Clemente, CA 92674 0128 714 R8-7420
Enclosure STEAM GENERATOR TUBE EGGCRATE SUPPORTS EVALVATION San Onofre Unit 2 1.
Summary This operability evaluation reflects Southern California Edison's (Edison's) best judgment based on the latest information available at this time. This updated evaluation reflects additional information on cause assessment and differences in operating histories between Units 2 and 3, which have been identified since completion of the original Unit 2 evaluation. As more information becomes available, or as additional insight into the physical mechanisms involved in Steam Generator (SG) tube eggcrate support degradation is gained, this evaluation will be reviewed for any potential impacts.
It is expected that another updated evaluation will be issued incorporating the results of the final cause assessment of the Unit 3 condition.
A degraded condition was noted on portions of SG tube eggcrate supports in Unit 3.
The degradation was detected by remote visual examination and was located at some peripheral locations on some of the eggcrates.
The degradation ranged from minor wastage of the eggcrate material to severe thinning in localized areas.
In some cases, small portions of the eggcrates were separated from the structure.
Assessment of this condition continues and will be completed in conjunction with returning Unit 3 SGs to service.
Inspections were performed to support chemical cleaning of the Unit 2 SGs.
The thinning was not observed in Unit 2.
The only degraded condition observed in Unit 2 eggcrates was one small area characterized as wastage and pitting.
This degradation has been evaluated as insignificant.
During the chemical cleaning outage, eggcrate and vertical strip thickness measurements were taken in both Unit 2 SGs to confirm the process monitoring of eggcrate conditions during cleaning.
Those measurements indicate the eggcrates are within the acceptance criteria established by the chemical cleaning specifications. Thus, operational corrosion of the eggcrates has been minimal, as was corrosion during chemical cleaning.
Eddy Current Test (ECT) results from the Unit 2, Cycle 9 refueling outage indicate minimal SG tube fretting indications at eggcrate locations, further substantiating that the eggcrates are intact in these areas.
The minor tube wear noted is consistent with wear phenomena that have been previously investigated and evaluated. This provides additional assurance that, at the time of inspection, the Unit 2 SG eggcrates were capable of controlling flow induced vibration.
Enclosure Several mechanisms are being evaluated for the observed degrr.dat' ion of eggcrate material on Unit 3.
Currently, a formal cause determination has not been completed.
The cause assessment to date indicates t0at the principal degradation mechanism appears to be Flow Accelerated Corrosion (FAC), with the predominant form probably erosion-corrosion. While this conclu; ton has not been finalized, it is clear that chemical cleaning has improved the overall tube bundle thermal hydraulics and helped restore the Unit 2 SGs to their nominal performance design conditions.
Considering the results of visual and ECT examinations performed on Unit 2, and the effectiveness of the chemical cleaning that was performed on Unit 2, there is no indication that the Unit 2 SG eggcrates are structurally degraded or are likely to degrade in the current operating period.
Therefore, there is reasonable assurance that the SG eggcrates meet the design requirements and are able to perform their designed safety functions.
II.
Background
The design function of the SG eggcrates is to provide a support structure for SG tubes to (1) minimize flow-induced vibration during normal operating conditions, and (2) maintain acceptable stress levels under accident conditions.
Action Request (AR) 970400975 documents that a pre-chemical cleaning video inspection of the secondary side of Unit 3 SG 3E088 revealed erosion, thinning, and pitting of some eggcrate material. AR 970401129 was created to document the operability of Unit 3 SG 3E089 in the event of findings similar to 3E088 (similar conditions were later substantiated).
Since Unit 3 is currently shutdown for refueling, the initial operability of the Unit 3 SGs was established based on providing only a " containment closure" function in accordance with Technical Specification 3.9.3.
Since Unit 2 is currently operating at full power, AR 970401646 was created to assess and document the continued operability of the Unit 2 SG eggcrates in Modes 1 through 6.
This AR will also be utilized to identify and track any additional actions associated with maintaining the Unit 2 SG eggcrates capable of performing their designed safety functions.
III. Discussion i
A.
Unit 2 Pre-and Post-Chemical Cleaning Video Inspections To support Unit 2's Steam Generator Chemical Cleaning (SGCC), video inspections were recorded to check and characterize the level uf deposits 2-
r Enclosure within the SGs.
The post-SGCC inspection was also intended to provide a visual record of SG cleanliness.
Three types of cameras were used for these inspections: one for the top of bundle overview, one for periphery bundle areas, and one for inner bundle areas..A brief description of the methods used for these inspections is provided below.
1.
Too of Bundle Overview This inspection consisted of lowering a large camera f rom above the tube bundle to provide a general area view of the upper bundle region.
This allowed observation of the batwings, vertical support straps and bars, and overall condition of the tubing in the upper bundle area.
Then, using a six-millimeter videoprobe, the upper portions of the batwings, batwing hoop, and batwing to outmost tube interfaces were examined.
2.
Outer Peripherv Inspection A six-millimeter videoprobe was inserted in the hot leg near the apex of the bundle. The videoprobe allowed observation of the tubes and eggcrates from the top of the bundle to the sixth eggcrate (counted from the top of the tube sheet). The areas of interest for this type of SGCC inspection were the diverter plate welds, eggcrate support rings, and eggerate intersection and eggcrate line contact areas to check for deposit remains.
3.
in-Bundle Inspection A small 1.5-millimeter fiber optic camera was utilized to provide an in-bundle observation of tube-to-tube line contacts and eggcrate intersections.
The insertion of this camera was generally limited to the tenth eggcrate and the batwing supports, but one time, in the stay cylinder area, the camera was able to penetrate to the seventh eggcrate.
4.
Lower Hand Hole Inspection A miniature Charge Couple Device (CCO) camera and the 1.5-millimeter camera were used for this inspection. The miniature CCD camera was used to inspect the underside of the first eggcrate support in the divider lane, stay cylinder, and a portion of the annulus. The 1.5-millimeter camera was also used to look at the underside of the first eggcrate near the periphery, within the bundle, and in the stay cylinder area.
Usually observations could be made up to the second eggcrate..
Enclosure B.
Unit 2 Pre-and Post-Chemical Cleaning Video Inspection Results The Unit 2 periphery inspections were performed on the hot leg at the apex of the bundle. This represents the area with the highest expected flow velocity.
The assessment performed at the time concluded that the eggcrates were in very good to excellent overall condition. A re-review of these videos was conducted for the areas where thinning of eggcrates was found in the Unit 3 SGs and is documented in an April 23, 1997 Memorandum For File by Mark Mihalik.
These inspections indicate the uppermost eggcrates (i.e., sixth, seventh, eighth, ninth, and tenth) in 2E088 are in very good to excellent material condition. The general appearance is sharp, well-defined edges on the support bars.
In contrast, the Unit 3 SG hot leg periphery regions (near the seventh, eighth, and ninth eggcrates) exhibited considerable eggcrate degradation.
One area, on the sixth eggcrate in 2E088, was noted with minor pitting and limited metal wastage. Theareaaffectedisapproximately1/4inchinlength.
Unlike the conditions noted on Unit 3, however, the pitting was very limited and judged to be insignificant.
The lowest two eggcrates (numbers one and two) in 2E088 were also inspected and were in very good condition. The inspection of 2E089 was limited to the lowest two eggcrates. These were also found to be in very good condition.
Due to limitations of the video equipment, eggcrates three, four, and five in either Unit 2 SG were not inspected.
The lowest eggcrate and portions of the vertical strips in both Unit 2 steam generators were gaged for thickness.
All thicknesses were within predicted values.
The SGCC video inspection was one of three elements used to assess the material condition of the eggcrates for chemical cleaning.
Care was exercised before relying on any one element, since all measurements were made remotely.
The other two methods used to assess eggcrate structure condition were physical gauge measurements and visual (unaided eye) inspections.
The results of these other two inspection techniques and measurements were consistent with one ar.cther, and support the conclusions obtained from the videos.
C.
Similarity of Each Unit's Steam Generators Visual inspection of the observed erosion of Unit 3 eggcrates indicates that both Unit 3 SGs are similarly affected.
This is to be expected, since erosion of the eggcrate material is considered to be sensitive to secondary side water chemistry and impurities.
Enclosure A review of macroscopic parameters which could affect each SG revealed that both Unit 2 SGs should be similar. This is based on the following: (1) both SGs have been fed with common feedwater treated with ammonia and hydrazine, since the beginning of their operation; (2) both SGs share one condenser - any sea water or air in-leakage experienced has affected each SG equally; (3) both SGs share one full flow condensate polisher demineralizer system (put in service in Cycle 3); (4) the make up water source, fed directly to the condenser, has been the same for each SG throughout operating life; (5) chemistry parameters, including types and levels of impurities, have been similar for both SGs throughout their operating life; (6) blowdown practices have been similar for both SGs; and (7) sludge removal via lancing has historically been similar.
A recent indication of the similarity is the deposit removal experienced during the Unit 2 chemical cleaning activity (completed January 1997).
The difference in deposit removal between both SGs was only about 800 pounds. The various individual metal oxides removed were also similar in weight.
This is significant considering that each SG had approximately 16,000 lbs of deposits.
Based on the above factors, the physical condition of 2E089 is judged to be similar to the observed conditions in 2E088 and, therefore, would have negligible eggcrate degradation.
D.
Unit 2, Cycle 9 Refueling Outage SG Tube Eddy Current Test Results A 100'4 bobbin probe eddy current test (ECT) of the Unit 2 SG tubes was performed during the Unit 2, Cycle 9 refueling outage.
These ECT results indicated minimal SG tube fretting indications at eggcrate locations.
Some minor tube wear was noted at batwings and vertical strips, but this is consistent with wear phenomena that have been previously investigated and evaluated.
No tube-to-tube fretting was identified. Thus, at the time of inspection, the Unit 2 SG eggcrates were capable of controlling flow-induced vibration.
E.
Potential Cause Assessment Based on the prevalence of eggcrate degradation in the peripheral regions of Unit 3 and the absence of similar eggcrate degradation in Unit 2, a review of macroscopic differences between the two Units was conducted to identify and evaluate possible causes. This review considered parameters such as secondary water chemistry operating and control methods, SG deposits transport mechanisms, and the resultant degree of SG fouling.
Based on a review of these parameters, the visual evidence, the affected material (low alloy carbon steel), and the hydraulic environment, the principal degradation mechanism -
4 Enclosure appears to be Flow Accelerated Corrosion (FAC), with the predominant form probably erosion-corrosion.
1.
Plant Systems and SG Deposit Comparison a)
Secondary Side Systems and Chemistry Control Both Units' feedwater and condensate system materials are similar and include titanium condeeser tubes, an equal number of copper alloy feedwater heaters (90-10 Cu-N1.and admiralty), and carbon steel tubed moisture sepa'rator reheaters.
Since commercial operation, the secondary side SG chemistry control philosophy has generally been the same for both Units 2 and 3.
EPRI guidelines were incorporated into Units 2 and 3 station procedures to I
establish a comprehensive and consistent chemical control program.
Both Units operated for the first two cycles without Full Flow Condensate Polishing Demineralizers (FFCPDs).
FFCPDs of equal design were installed and put into service at the start of each Unit's Cycle 3 operation.
While use of specific polishers can vary at any particular point in time, the operating i
i practices for each Unit's FFCPD are similar, The long-term chemical control ranges (i.e., pH) have been similar for both Units. A common water source provides condensate makeup. Ammonia and hydrazine have been used in both Units from initial operation through most of Cycle 8 operation.
SG blowdown practices have been similar on both Units.
During outages, SG protection practices (i.e., layup) have been similar.-
l There are a few differences in the chemical control of Units 2 and 3, but, in general, these differences have been either short in duration or do not impact steam generator chemistry. Unit 3 operated for the last 5 months of Cycle 8 operation (prior to the current refueling outage) with ethanolamine (ETA) for pH control in lieu of ammonia.
This change was implemented in an effort to reduce corrosion product transport to the Sgs. A reduction of approximately 4 parts per billion in the feedwater was accomplished.
This change is not significantforUnit3'sSGinternalssinceETAhasbeenshown(byindustry experience) to improve pH control. A detailed review is in progress, as part of the cause assessment, to determine if a more subtle chemical cause or operational difference may have contributed to the Unit 3 eggcrate thinning.
While the above discussion leads to the conclusion that the two Units have been operated in a similar fashion, the deposits produced, as discussed below, are different.
The reasons for this are not yet explained and are the focus of the cause investigation.._.
Enclosure b)
Deposit Characterization And Impact Evaluation Corrosion products and other impurities generated in the feedwater, condensate, and drain systems at pressurized water reactors are continually transported to the steam generators.
If not removed by blowdown or carried out in the steam system, these materials form deposits on the top of the tube sheet (sludge piles), on tube surfaces (scale), and, to a lesser extent, on the surfaces of the steam generator secondary side pressure boundary and structural components. At San Onofre Units 2 and 3, detailed evaluaticns of deposit loading were completed in support of planning for chemical cleaning.
The analyses.are primarily based on an integration of feedwater corrosion product transport data. These predictive analyses took into account, among other factors, estimates of blowdown efficiencies and the results of tube sheet sludge lancing performed at each refueling outage.
The deposit loading calculations completed prior to SGCC resulted in estimates of 16,700 pounds per SG for Unit 2 and 17,800 pounds per SG for Unit 3.
Chemical cleaning and subsequent sludge lancing at Unit 2 resulted in removal of an average of 16,300 pounds of deposits from each of the two steam generators, meaning that the deposit loading prediction was accurate to about 3%. Although sludge lancing r,as not yet been completed at Unit 3, chemical cleaning results and estimates of the size of the sludge pile from eddy current data suggest the amount of material that will be removed from Unit 3 will be between 19,000 and 20,000 pounds per SG, (about 17-23% more than at Unit 2).
The prediction that the Unit 3 steam generators were more heavily fouled than those at Unit 2 appears to have been confirmed.
c)
Deposit Characteristics in addition to calculating deposit loadings for the steam generators at Units 2 and 3, Edison performed detailed physical and chemical characterization of tube scale deposit samples from each unit prior to chemical cleaning.
Tube scale was estimated to comprise approximately 90% to 95% of the deposit inventory at Units 2 and 3, and was suspected to be the cause of thermal performance degradation that had been experienced over time at both Units. The structure and composition of the deposits at Unit 2 were i
consistent with the observed thermal performance degradation, which manifests itself as a decrease in main steam pressure.
For Unit 3, the structure of the tube scale suggested even greater thermal resistance, owing to higher density, slightlygreaterthickness(consistentwithhigherdepositloading),anda higher degree of consolidation. Consolidation is a process that increases density, clogs and fills pores in the deposits, and supports growth of thicker scale layers.
1 Enclosure d)
Deposit Impact Edison has compiled a detailed history of steam pressure trends at both Units 2 and 3.
As discussed earlier, the accumulation of deposits on tube surfaces was believed to be the predominant cause of steam pressure decline (performance degradation) at Units 2 and 3.
Removal of the tube scale was one motivation for the recent chemical cleanings.
In addition to tracking steam pressure, Edison routinely calculates global fouling factors for each steam generator at both Units.
The global fouling factor is an indicator of steam generator heat transfer performance, which adjusts for tube plugging, changes in primary plant temperature, and variations in core thermal power.
Steam pressure and fouling faciar trends for Units 2 and 3 prior to SGCC are presented in Figures 1 and 2, respectively. A review of this information reveals that:
The steam pressure degradation at Unit 3 prior to SGCC was approximately 75 psi. The corresponding steam pressure degradation for the same number of effective full power years of operation at Unit 2 was about 55 psi, suggesting that Unit 3 was the more heavily fouled unit.
Based on evaluation of extensive performance data, the best estimate fouling factor for Unit 3 was calculated to be 17% higher than at Unit 2 prior to SGCC.
2.
Flow Accelerateu Corrosion and Thermal Hydraulic Effects Video records of the Unit 3 eggcrates were reviewed in detail to attempt to establish the cause of the degradation.
Based on the visual evidence, the affected material (low alloy carbon steel), and the hydraulic environment, the princip:1 degradation mechanism appears to be FAC, with the predominant form probably erosion-corrosion.
FAC is a general term for processes that rely on mechanical assistance from a fluid environment for removal of a protective corrosion product (oxide) layer from base metal.
Elevated fluid velocities and/or droplet impingement removes the protective oxide layer thereby exposing new base material for repetition of the corrosion process.
The defining characteristics of FAC are general thinning and flow-induced scalloping of surfaces in an active flow region.
FAC is common throughout steam plant systems, and there are numerous parametric studies supporting development of practical FAC evaluation models.
An example of one popular FAC model is the EPRI "CHECKWORKS" program.
Common Enclosure elements of these FAC models are 1) affected material characteristics, 2) fluid chemistry, 3) the dynamics of the environment, and 4) initiation thresholds.
Each of these parameters are interrelated such that, in any specific application, precise determination of corrosion rates is not practical; however, relative corrosion assessment and general expectations are achievable.
In a steam generator secondary s.ide environment, carbon steel forms an oxide layer, which provides a barrier between the base metal and the feedwater system fluid.
Under normal conditions, steam generator fluid velocities are too low to cause significant removal of the protective layer. As a result, corrosion processes on internal carbon steel components are inhibited.
The post-SGCC inspection results for Unit 2 were consistent with the expectation that no significant FAC condition was present.
Close inspection of the videos depicting degraded Unit 3 eggcrate lattice bars revealed surfaces where aggressive corrosion exists in very close proximity to regions of unaffected material.
In many instances, different conditions exist on the same carbon steel lattice bar. This observation suggests that neither materialproperties(e.g., chromium /molybdenumcontent)norfluidchemistry are likely to be the primary causal factor.
Rather, this observation suggests that detailed, local hydraulic conditions are more likely to be the dominant influence.
Recognizing that the fluid dynamics at the eggcrate may be a dominant factor in the degradation process, the differences in steam generator performance between Unit 2 and Unit 3 were reevaluated. The significant result of this evaluation is that the deposits on the eggcrate bars can occupy a considerable fraction of the available flow area.
It is believed that the level of l
fouling, which er.isted before SGCC, approached a magnitude where flow through the tube bundle would be severely restricted.
Furthermore, since the restriction to inner bundle flow increases exponentially with fouling as the remaining flow area is choked off, a relatively small difference in eggcrate fouling could introduce a large effect on overall steam generator dynamics.
Therestrictionofflowthroughthetube/eggcratelatticeinterfacescould dramatically redirect flow through open lattice locations in the periphery and stay cylinder regions. Consequently, a relatively minor difference in deposit burden between the two Units could result in substantially amplified changes l
in hydro-dynamics in the areas where degradation has been observed.
Preliminary results from EPRI's steam generator thermodynamic modeling code
)
ATH0S substantiate the general hypothesis of increased fluid velocity in the j
periphery and stay cylinder regions under fouled conditions.
.g.
I l
Enclosure As' discussed previously, chemical cleaning has been successful at removing most, if. not essentially all, of the accumulated deposits within the Unit 2 SG tube bundle region. Although a final cause determination on the Unit 3 eggcrate condition has not been completed, it is clear that chemical cleaning.
Lj has improved the overall tube bundle fluid dynamics and thermal hydraulics.
l Thus, SGCC has helped restore the Unit 2 SGs to their nominal performance design conditions.
[
F.
Unit 2, Cycle 10 Mid-Cycle Outage Currently, a mid-cycle outage is planned to allow ECT inspection of Unit 2 SG tubes.
Video inspection of the Unit 2 SG eggcrates will be performed if warranted by the final cause assessment of the Unit 3 condition.
.IV.
Conclusion Based on the above evaluation, the Unit 2 SG eggcrate supports are able to i-perform their designed safety functions, and are operable.
i 1
l 1
J L
l
's t
1..
Steam Generator Pressure, SONGS 2 & 3 y
980 o.
E 960 l-940 E
SONGS 2 e
920-5g L
/
O g 900 j-3, g
p a
s 880-I@-\\.m I
g m q c
( g
[
a 860 g
g ep l
840 SONGS 3 820 0
1 2
3 4
5 6
7 8
9 10 Approximate EFPYs 72 Figure 1
1 1
Steam Generator Fouling Factor, SONGS 2 & 3 l
250 SONGS 3 6
N M,-#
o 200 -
- y yY
$3 150 -
I I
h
' f C
t
]
Q b s
2,u*
100 r-I
H I
w4 Y
Y *V SONGS 2
.o.c I
T 50 --
1 f
3 cr:
0-i i j
.1 r
._8
/
~50 -
-100 0
1 2
3 4
5 6
7 8
9 10 Approximate EFPYs 2
C a
Figure 2
=
-