ML20141M287

From kanterella
Jump to navigation Jump to search
Responds to NRC Re Insp Rept 50-309/92-04 & Forwards Root Cause Evaluation of SG Separator Can & J Nozzle Damage Discovered During 1992 Refueling Outage. Chemical Additions Control Ph
ML20141M287
Person / Time
Site: Maine Yankee
Issue date: 08/03/1992
From: Hebert J
Maine Yankee
To:
NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM)
References
JRH-92-63, MN-92-77, NUDOCS 9208110043
Download: ML20141M287 (43)
v • d • e


Text

{{#Wiki_filter:_ MaineYankee HtuAstt electarcif v eOH MAINt $31NCE 1977 EDISON DRIVE. AlGUSTA M AINE 04330. (207) f ?? 48f 8 August 3, 1992 MN-92-77 JRH-92-63 UNITED STATES NUCLEAR REGULATORY COMMISSION Attention: Document Control Desk Washington, DC 20555

References:

(a) License No. DPR-36 (Docket No. 50-309) (b) USNRC Letter to MYAPCo dated May 18, 1992 Inspection 3 50-309/92-04 W

Subject:

Inspection 92-04: Inservice Inspection Program and Steam Generator-Inspection at Maine Yankee Nuclear Power Station Gentlemen: 1.Ms letter responds to your request that a final report of Maine Yankee steam generator separator can and j. nozzle damage discovered during the 1992 refueling outage with regard to Reference (b) be submitted to you upon completion. If you have any questions concerning this report, please feel free to contact me. Very truly yours, }f8 li h p r V James R. Hebert, Acting Manager Licensing & Engineering Support Department PJP/sjb Enclosure c: Mr. Thnmas T. Martin Mr. Charles S. Marschall Mr. E. H. Trottier Mr. Patrick J. Dostie Mr. H. Kaplan Mr. Clifford J. Anderson (f 100059 L y 9200110043 920803 7 DR ADOCK 05000309 'l .a PDR \\

MEMORANDUM l CORPORATE ENGINEERING DEPARTMENT TO: C. HUNTER GILPATRICK DATE: JULY 16, 1992 4 FROM: P. PLANTE FILE: PJP-92-018

SUBJECT:

ROOT CAUSE EVALUATION OF MAINE YANKEE STEAM GENERATOR SEPARATOR CAN AND J. N0ZZLE DAMAGE DI' COVERED DURING THE 1992 RFFUELING OUTAGE EXECUTIVE

SUMMARY

This report documents the technical issues concerning the damage observed as a result of a visual inspection of the secondary side of' steam generator (S/G) #3 during the 1992 refueling. The following list summarizes the conclusions and recommendations from this investigation. Additional background details on how I arrived at these conclusions and recommendations can be found in the sections that follow. A CE critique of this report inay be found in Attachment 2, in their J"ly 6,1992, letter. I have modified this report slightly as a result of their review and Haine Yankee's internal reviews.

1. Single phase erosion-corrosion caused the observed J-tube damage.

2. Two phase erosion-corrosion caused the observed separator can damage. 3. Both cases of erosion-corrosion resulted from a combination of susceptible materials and a secondary side chemistry ph lcsophy which protected the copper alloys and the Inconel-600 S/G tubes in the syster. preferentially to the carbon steel. 4. Maine Yankee can expect a significant benefit by increasing the secondary operating pH and adding morpholine to the secoriary system. This will_ reduce corrosion product transport to the S/G's (decrease ialing and sludge), improve preservation of steam generator structures (egg-crates, dryers, tube support plates, feed rings, etc.), and improve preservation of the feed train (piping,-turbine / moisture-separator reheater components, etc.). 5. Maine Yankee should prepare to inspect and repair original J-tubes at the next refueling. 'l PJP:EAG c: R. Blackmore (For plant staff) C. Eames 3 T. Gifford I 7 S. LeClerc R._Liscomb _ J.l Hebert? { forMiibsi ttialf tM NRC[WhKipl ani throg^h' thE.Residentl's office)3 l S. Nichols P. Radsky L. Speed l G. Whittier Tech File 3.7 D. Zehler (ABB-Combustion Engineering)- Ra\\ PED 4\\PJP\\PJP92018 MEM

July 16,1992 PJP-92-018 Page 2 BACKGROUND In February,1983, Maine Yankee added J-tube discharge nozzles to the S/G feedring per Plant Design Change Request 2-83 in response to a water hammer incident (see Figure 1 for the Maino Yankee S/G configuration). The original bottom mounted discharge nipples were capped at this time. The J-tubes were constructed of 3", XS carbon steel to ASME SA-234, Grade WPB welded to the feedring as shown in Figure 2. In May,1984, Maine Yankee replaced all of the separator cans per Engineering Design Change Request 84-30 because we we'e experiencing 1.5-4% moisture carryover at 2630 megawatts thermal (MWT). By design it should be about 0.2%. We estimated that we could gain 15-33 megawatts electric (MWE) by reducing moisture carryover to 0.2%. The replacement cans installed at tlist time are known as System-80 separators and are somewhat different than the originally installed standard cans. Figure 3 compares both designs. During the 1992 secondary side visual inspection of S/G #3, conducted by L. Speed and C. Eames per Procedure 17-217-1, they noted damage to some of the steam separator cans and J-tubes. As a result, Maine Yankee inspected the other two steam generators and noted similar damage there as well. Foisow up inspections included ultrasonic characterization of the J-tubes and feed rirgs and removal and visual checks on the separator cans. Photographs of the anst severe J-tube damage encountered can be found in Figure 4. Ultrasonic data is too voluminous to reproduce hero, but Attachment 1 summarizes those findings for the J-tubes. The feed ring itself was also ultrasonically inspected and in general was found to meet nominal wall thickness, indicative of very little erosion, except for isolated cases in the immediate vicinity of the J-tubes. These areas were reinforced by the J-tube repair methodology discussed later on. Figure 5 illustrates some examples' of the more severe separator can damage. Separator can damage as a function of position on the can deck was examined. It was determined to be random, even though we know that flow varies depending on relative can deck position (see Enclosure 1 of ). Upon discovering the damaged J-tubes and separator cans, Corporate Engineering (CED) began an investigation to identify the root cause for the damage and what factors influenced its occurrence. -This report documents the findings of this investigation although most of the conclusions and remdial actions were reached very early on to support our startup schedule. UTILITY EXPERIGHS One of the first measures we undertook was to call other Combustion Engineering -(CF) plants and interview them on their operational experiences in this area. The results of this effort are summarized in Table 1. Some of the pertinent conclusions I've drawn from this survey are: 1. Maine Yankee is the only 3-loop CE pitat. 2. Maine Yankee and Palo Verde are the only plants with System-80 separator cans. R \\ PED 4\\PJP\\PJP92015.NEM

July 16, 1992 PJP-92-018 Page 3 3. Palo Verde 1 inspected their separator cans in May, 1992 with no signs of erosion. Their cans went into service about Lji years after Maine Yankee's cans. 4. Erosion of S/G feedtrain components is common, but unpredictable in CE S/G's. 5. No utility with standard separator can designs has experienced noticeable through wall erosion, including Maine Yankee. Some minor erosion was noted on Maine Yankee's standard cans; however, they have been in service since 1972, twice as long as the Systems 80 cans. 6. One CE atility had experienced thinning of bottom mounted discharge nipples which was the feedring discharge configuration previous to installation of the J-tubes in the early 1980's (Reference 3). This failure was attributed to cavitation. 7. Westinghouse plants have also experienced many instances of J-tube erosion (Reference 1, 4, and 5).- The generalities drawn from this survey and associated plant experience literature search are: - J-tube erosion is common in PWRs and is not surprising in our case. - I could not identify any other documented cases of separator can erosion in our industry. Our problem seems to be an isolated case. ABB's assessment of the damage encountered at Maine _?ankee has been included in Enclosures 2 and 3 of Attachment 2. FAILURE ANALYSIS Upon inspection of the failed J-tubes and separator cans, I concluded that both failures resulted from erosion-corrosion. It was necessary to arrive at the cause of failure so that appropriate remedial actions could be taken in a timely fashion. Why we were experiencing erosion-corrosion was also an important question;-however, it was not as crucial to selecting replacement materials as determining the mode of wall thinning. Several pieces of evidence led to the erosion-corrosion conclusion and are summarized below:

  • 'J-tubes History of J-tube failures at other utilities cited erosion-corrosion as the mechanism'of wall thinning.

Corrosion occurred predominately on the ID'of the J-tubes even though the external surface of the J-tube also showed pockmarks. Observations include: _ (1) that manufacturing information stamped into the elbow is still plainly visible in all cases; (2) pockmarks on the OD are large (1-3 mm) and somewhat isolated; and (3) pockmarks on the ID are small (0.1-0.5 mm) and. uniform throughout the surface. The above described morphology is consistent with erosion-corrosion' occurring under-single-phase conditions, which would be expected for the normally submerged J-tubes (Reference 7, page 952). R \\ PED 4\\PJP\\PJP92018.MEM - ___________________.____________________J

July 16, 1992 PJP-92-018 Page 4 While overall, the damage was more seven the closer the 1-tube was to the feedwater distribution box, the patt en was random enough to suggest erosion-corrosion. This is bec a one elbow would have sufficient wall thickness remaining, wh 1 'he next J-tube 'ver would be severely eroded (see Attachment 1 for Sund configuration data). This has been observed time and again in crosion-corrosion of carbon steel. This could be the result of variations in velocity of the feedwater along the feedring through the J-tubes or because of small variation in the residual element content of the base metal (particularly chromium, copper, and nickel). The velocity affect is illustrated in the maps in Attachment I where J-tube damage occurs more frequently near the feedring inlet where flow velocities are higher. With regard to chemistry, three different makes of elbows where observed from the re11 aced eroded elbows. There does not appear to be any correlation between a severely eroded elbow and a particular heat /make of elbow. The J-tube base metal (ASTM A106, Grade B) has a history of susceptibility to erosion-corrosion.

  • Separator Cans

~ The environment associated with the separator cans is wet steam and the damage observed, particularly in the spinner blades, is characteristic of the " Tiger Striping" damage normally associated with two phase erosion corrosion (see Figure 5b). In this case as well, small differences in residual alloying in the base metals also significantly affect erosion-corrosion resistance. This is particularly well illustrated in Figure 5a where the vertical strip between the eroded sections is the seam weld associated with the can's manufacture. In the case of carbon steel, weldments generally contain higher-amounts of alloying elements than the base metals they join together. In general, through wall erosion of the outside wall of de separator cans was associated with assemblies which experienced significant degradation of the spinner blade assemblies. This changes the angle of impact between wet ster.m and the wall of the can to one more apt to cause two-phase erosion-corrosion. I believe' that spinner blade erosion always precedes separator can' wall erosion. The apparent absence of this phenomenon at other utilities would suggest erosion-corrosior rather than simple water impingement erosion. The reason for making this statement is that all spinner blades see water impingenent and, therefore, all would be expected to exhibit some level of-erosion if this alone were the mechanism. However, erosion-corrosion can occur in this service if the proper combination of variables are present. R \\ PED 4\\PJP\\PJP92018.ME4

July 16, 1992 PJP-92-018 Page 5 Four "short" or original separator cans with orifice plates were aiso inspected. They had been in service since original construction. As shown in Figure 6, the condition of these spinnor blade assemblies was significantly improved over the System-80 cans even the'.gh they had been in service much longer. As a result of this observation, I believe the following statements are pertinent:

  • These short cans were orificed to reduce flow through them by 25%.

This indicates that velocity is certainly an important consideration as to what erodes and what doesn't.

  • Steam / liquid flow varies from one location to another on the can deck and may explain, along with minor variations in chemistry residuals, the widespread, but selective, nature of the separator can damage (see Letter 1 in Attachment 2 for details).

The fact that this type of can erosion is apparently unique to Maine Yankee is due, in part, to the fact that Maine Yankee has more total flow per separator than any other CE plant as shown in Table 2 (no information was available on Palo Verde or Waterford at the time this information was compiled). EROSION-CORROSION AND KEY INFLUENCING FACTORS s As I stated previously, the evidence-indicated that single-phase and two-phase erosion-corrosion was responsible for the J-tube and separator can damage at Maine Yankee. The only other mechanism which was considered at length was corrosion by excess amounts of carbon dioxide. However, Maine Yankee's condenser air-in-leakage (the primary source of carbon dioxide) is g!nerally less than 4 SCFM, which is very low for this industry. This theory was not t considered any further. Since this would not contribute appreciable amounts of carbon dioxide. With regard to erosion-corrosion, several variables are known to affect the occurrence and rate of this phenomenon in carbon steels. Each of these variables in turn will be discussed below with regard to Maine Yankes J-tube and separator can failures. The purpose for this is to determine which factors may have caused this damage (including our recent uprate to 2700 MWT) and over what time frame it may have occurred. I will not go into: great detail on the erosion-corrosion mechanism itself, but refer the reader to References 1, 7, and 10 for excellent summary articles of this phenomenon in PWRs.

  • Fluid Velocity:

The influence of fluid velocity on erosion-corrosion of carbon and-chrome-moly steels has been included in Figure 7, reproduced from Reference 10. While the conditions for the tests which resulted in this figure are not identical to Maine Yankee's service conditions, it does serve to illustrate a basic trend. For the case of Maine Yankee's J-tubes, feedwater velocity has been calculated at 16'05 ft/sec at.2630 MWT and 16.68 ft/sec at 2700 MWT. Upon looking at the effect of this l Ra\\ PED 4\\PJo\\PJP92018.MEM __a

July 16, 1992 PJP-92-018 Page 6 difference in Figure 7, two facts become cbvious. The 16 ft/sec or so velocity is at the low end of the velocity magnitude which would seriously increase erosion-corrosion. And the use of 2% Cr-1 Mo steel as replacement material significantly improves erosion-corrosion resistance (by about 15 times).

  • Fluid Temnerature:

The influence of fluid temperature and it's effect on erosion-corrosion of carbon and chrome-moly steels has been included in Figure 8, reproduced from Reference 10. Temperature of Maine Yankee's feedwater does not change appreciably from 2630 to 2700 MWth. Our feedwater temperature (about 440'F) has been superimposed on Figure 8 to show the relative effectiveness of the replacement materials..Although this figure was generated at higher velocities than occur in Maine Yankee J-tubes, it does illustrate some interesting general principles about the temperature influence, which includes a maxima at approximate 300'F regardless of the alloy used. This is well below the temperatures associated with the fluids in contact with the J-tubes and the separator Cans. a p3: The effect of changing pH on erosion-corrosion rctistance of carbon and chrome-moly steels can be found in Figure 9. % produced from Reference 10. When examining this figure, note that it does not compare alloy performance under the same conditions, it merely illustrates the beneficial effect of decreasing pH (increasing the pH number). Maine-Yankee procedures in the recent past have limited condensate pH to <9.2. This has generally been the trend since 1986 as shown in Figure 10. Prior to that, secondary system pH control (with Boric Acid) focused on maintaining the S/G pH, condensate " occasionally" exceeded 9.2. Since that was felt to be a contributor to copper transport to the S/G's, Maine Yankee made the change in 1986 to limit it. Now that copper has essentially been removed from the feedtrain and we have operated for 1 cycle after copper removal, Maine Yankee is planning-to increase pH once again. Essentially, pH is controlled through chemical additions. At Maine Yankee, pH is increased by increasing ammonia, a decomposition product of hydrazine. Maine Yankee will be considering the use of morpholine in conjunction with the existing additives. This will provide improved protection of. carbon steel because of two factors. The first is that morpholine portions itseW to the liquid phase and, therefore, helps to control corrosion in the immersed part of the steam generator (whereas-conventionally used ammonia tends to portion itself to the steam, protecting equipment away from the steam generator-] (Reference 7, page 958-959, and References 11 and 12). The second concerns the fact that morpholine results in a higher pH at-temperature than ammonia, which in turn reduces the corrosion rate in carbon steel. This-basic-fact has' been illustrated in Figure 11. The French have corroborated these results by concluding that erosion-corrosion of carbon steel in the . presence of ammonia is approximately-5 times higher than in the presance of torpholine, all other conditions being-equal. R e\\ PED 4\\PJPWM2018.MEM

I July 16, 1992 - PJP-92-018 Page 7 Other Variables: Other variables such as dissolved oxygen, can also affect erosion-corrosion resistance. Generally, low oxygen promotes erosion-corrosion. It is entmally low by design in PWRs and Maine Yankee has traditionally been below 5 parts per billion. However, since increasing dissolved oxygen increases the risk of other forms of daiaage, it's not a consideration for improving resistance to erosion-corrosion observed at Maine Yankee. Alloy composition has a strong ir?luence in erosion-corrosion, as shown in Figures 7-9. The original materials of construction, ASTM A106 carbon

teel in the case of the J-tubes and ASTM A570 sheet steel in the e.!se of the separator cans are hig'iy susceptible to erosion-corrosion under the proper conditions,. Howevt.c, even small amounts of residual alloying elements, such as chromium, copper, and molybdenum in these alloys can affect its erosion-corrosion resistance (Reference 8),

This fact generally explains the apparent -randoianess of erosion-corrosion, i.e., why certain J-tubes and separator cans eroded so severely and others adjacent to-them did not. The replacement material 2% Cr - 1 Mo steel, c is significantly more resistant to erosion-corrosion as shown in Figures 7-9. See the section entitled "Remediation" following this section for a discussion on the repair methods. Some discussion has been conducted concerning the role of sludge on the-damage observed at Maine Yankee. I feel it has a minor impact on the damage observed. My reasons are as follows:

1. The more severe cases of damage shown in Figure 4 can only be explained by chentical dissolution mechanism as opposed to a mechanical abrasion mechanism.
2. Corrosive type damage was observed on both the inside and outside surfcces of the J-tube. The outside surface would not presumably see abrasive damage.
3. Significant damage was also observed in the separator cans which would 3

not be affected by sludge carryover.

4. High sludge carryover may be seen prior to or at the same time as significant erosion-corrosion,- however, it is a symptom rather than a cause. High sludge would indicate a chemistry program which does not minimize carbon steel corrosion.

Moistiure carryover was checked both before and after these repairs were made. Improvement was noted'in all three steam generators with significant improvement in S/G 1 and 2. These values do not provide direct information about the failure itself, but does indicate a definite ost in terms of moieture carryover and consequently MWE. R:\\ PED 4\\PJP\\PJP92018.MM-n

July'10, 1992 P'P-92-018-Page 8 RfgDIATION Maina Yankee opted to address the damage through'the use of improved materials, whenever possible. In'accordance with a repair philosophy-established in 1985 (Reference 13):and an excellent service record since installation, Maine Yankee-selected 2%Cr - 1_Mo steel to repair J-tubes (as -- noted in Attachment 1)-and'all separator cans regardless of condition.--The-effectiveness of 2% Cr - 1 Mo has been well documented in-both this~ memo and in the. open literature- (see Reference -7). This section~will describe the-repair methods utilized to address this damage. J-tubes: J-tubes were replaced if the e.1 thickness o -.)und to be.less-than 0.18" in any location, via ultrasonic inspection t o.iques. The replacement design can be fouad in Figure 12. The repu.ement design was_such that all interior wetted surfaces are either made:of or-surfaced with. 2% Cr-1Mo steel. Additionally, all-chrome-moly welding was conducted on: a work bench..The saddle was designed from-P-1 carbon steel-'so that the field weld would be a simpler P-). to P-1 weld,. saving dose-and improving the chances 1 r of-making-a sound weld!the--first time.- Seoarator Cars: Separator can damage was confined to the spinner blade -assembly area as shown in Figure 3. The repair called for removalLof the spinner blade asseablytand any-damaged separator-shell :2nd ; replace with new chrome-moly ir.serts (shown ln Figure 13) welded to the sound remaining portion of the separator can as shown in Figure 14. All surfaces formerly eroded by the wet steam are now composed of chrome-moly steel.. . chemistry: Maine Yankee will be increasing the pH and morpholine in'the secondary system. This will reduce carbon steel. corrosion in both-the secondary side'of the steam generator and;the_feedwater/ condensate / extraction r steam system. CONCLUSIONS: 1. The cause of the S/G J-tube degradation observe'd at-Maine Yankee.in-March,L1992, was single phase-(liquid) erosion-corros_1on. I 2. -The cause of.the-S/G separator-can-degradation-observed at Maine -Yankee-in: March, 1992 was two phase (liquid'and vapor) erosion-corrosion. 3. A combination of susceptible materials and secondary ' side chemistry . philosophy which favors copper and S/G tube integrity-to carbon steel ~ integrity was responsible for creating-a condition.. conducive to _ erosion-corrosion;of the S/G. secondary ' side. 4. The total. flow per separator at Maine Yankee-is 'much higher than other older CE units _and_could explain-the isolated nature of Maine ~ Yankee's separator can damage; Rj\\PE04\\PJP\\PJP92015.MEM

July 16, 1992 PJP-?2-018 Page 9-s 5. While increasing power from 2630 MWT to 2700 HWT does increase some parameters which-favor erosion-corrosion, such as feedwater flow, the change was so small as to have an insignificant effect. 6. The repair methodology of using 2%Cr-1Mo steel in replacement hardware should prevent significant erosion-corrosion damage of these components in the future. RECOMMENDATION},: 1. Modify secondary side chemistry philosophy to significantly increase 1 morpholine concentrations [ specific target concentration to be determined] and increase pH as high as possible per currently accepted practices. 2. Inspect remaining J-tubes during the-next refueling outage and replace J-tubes ind* ating significant additional erosion from the previous cycle [th., t.old necessitate having spare chrome-moly elbow-assemblies on hand prior to the next outage). R

REFERENCES:

1. Woolsey, I.S. " Erosion-Corrosion in PWR Secondary Circuits." Central Electricity Research Laboratories, TPRD/L/3114/R87, March 1987. 2. Plante, P. " Trip Report: CE0G Materials and Chemistry Subcommittee Meeting - May 2 an 3, 1991." Maine Yankee, PJP-91-009, June 4, 1991. 3. CE Info Bulletin 81-10. " Steam Generator Feed Ring Nipples ' December 9,.1981. 4. EPRI. Steam Generator Reference Book." May 5, 1985, Pages 13-14, 14-15, 5. Westinghouse Technical Bulletin 82-07,. Revision 2, " Steam Generator J-Tube Wall Thinning." October 16, 1985. 6. INP0.0peration Experience 4144. " Steam Generator Thermal Sleeve and Feed Ring Erosion."- 18-Sep-90. 7. Cohen, P..(ed.). "The ASME Handbook on Water Technology for Thermal Power Systems." The American Society of Mechanical Engineers,1989. 8. Huijbregts, W. " Erosion-Corrosion of Carbon Steel in Wet Steam." Materials Performance, October 1984. 1 9. Maine Yankee Engineering Design Change Requesti84-30, " Modifications to the S/G to improve Quality."

10. Jonas, 0. " Control Erosion / Corrosion of Steels in Wet steam."
POWER, March 1985, Pages 102 and 103.

i Ri\\ PED 4\\PJP\\PJP92018.MEM - ~.

I July-16, 1992 PJP-92-018 Page 10 11. Beacher, J. "Uso Amines to Control System Corrosion." POWER, July 1981, Pages 74-77.

12. NALCO Utilities Newsletter.

October, 1988, 13. Plante, P. " Material Consideration for Erosion-Corrosion Resistance." Maine Yankee Internal Memorandum PJP-85-30, November 28, 1985. o { l Re\\ PED 4\\PJP\\PJP92018.MEM

>=>m. xm ga Q s a a a a a "5 5 5 5 5 5 5 5 3a "s s

= = = "s = w= = w. w.. w w w w 80 x x $ S x $ $ M8 x 2 2 z x E x x x x m ,,. =M a w p E 25 .=. o= ~ m o m., E. E E .~ w ~ w. w.m w - m w= w..m w w e w =. m. wo m m-Z s o o=. wxmoo o .o. .d >*oo O .o. m3.J m sw w e 6 w W W =w w >C

u. - W" g5

.s. $25 w ww a m a. .- m Z eh I m =W w D= 3r E d

e. oW"

= Se e u-n m E 8 a.= BW B" W g>m B = o. - W W m z sW e =ob b W a w .- w s-o s -=s -s a s-s a o om, w=o w> = m x.c m m = won

z .o sm om a r =w o 2" e o o e o o o o o o u mE = = = w =

=

o m 5e w w + a o m EC o w <c e. -o x = w w ~ ~. m m u - mm o o a w e o - w o o-o- g m=w = = = > =

>

w ed- -ew CL. e w e x.

q -

o u. zw wm m. w o> = w .e Q g 3 os a m" Es ~m E EU A e =w w w wo w w o - o e. o w= > =. > > = =

g Ct g C i.0 m ~ o. w u ma w wx m .m.A m m m m m m 13 % w o tad w w e. w ha w. e Pw = D*

  1. =

D= p e. NW g e 7' > W s,, c o es m c m. m ..o o = . g b O-E .k t I. Q. om g .bt "2 5 0 ~ ~ ~ b &2-2 m ~ ~ ~ ~ ~ ~ os= s s s s s - s s s y =mQ

== N .=s

  • =e M

.N N e 3h +s - gw m L %M L -Q 'W 4C El C O. 2.;. 2 m ~.m w = w. .N w g u "OE *h g W O O 3 8' O w 4m S.. w v w m o ,- a e 5 8 a 5 8

2. e. x a. -

M = v w s w w m m.o m .=x= w 3 .== a - - o6 .Z .J O m 4 m x-xse g( o .e* 64 o = ame = Q eC .a a E

== (

  • s A

E m m 3 .* N e') v I ____mm_.-._ _ ~ _. - - - - - - - - --- - - - ' - - - ~ ^ - "

u TABLE 2 DESIGN FLOW RATES FOR COM80STIDN ENGINEERING SEPARATOR CAN5' POWER STEAM STEAM CIRCULATION VATER STEAM NUMBER STEAM FLOW TOTAL FLOW OUTPUT FLOW PRESSURE RATIO LEVEL DRUM IN OF PER SEPARATOR PER SEPARATOR SITE' f x10'*8TU/tR )' - (X1078/HR) (PSIA) (V EVAP/W STEAM) (+/-) DIAM. FIN.) SEPARATOR fLB/HRI f tB/tR) 1 4.181 5.281 770 4.4 12/96 230 166 31.800 140,000 MAINE YANKEE (2440) 2.787 3.575 815 5.6 12/84 180 98 36 "N 204,300 MAINE YEKEE (2630) 3.005 3.684 860 5.5 12/84 180 98 39,650 222,000 MAINE YANKEE (2700) 3.085 3.920 854*> N/A* 12/84 '180 98 40,000 N/A* 2 2.437 3.112 770 5.2 12/18 180 98 31,750 165,100 3 4.386 5.603 815 4.3 12/64 230.13 166 33,750 145.150 4 4.386 5.576 850 4.2 12/64 230.13 166 33,600 141,100 5 4.386 5.603 815 ~ 4.3 12/64 230.13 166 33,750 145,150 6 4.727 6.200 900 3.7 12/64 230.25 166 37,350 138,200 i j 7 5.819 7.565 -900. 3.4 12/63 253.13 212 35,700 121,325 I l ' Adapted from EDCR 84-30 ' Includes Reactor Coolant Pump Power (12 HWT), "This number is desi9n. Actual steam pressure was approximately 825 PSIA on January 22, 1991. i l % t available. However, trends established above are expected to hold true for these parameters. l I R:\\ PED 45PJP\\PJP92018.MEM l

FIGURE 1 SCHEMATIC '0F MINE YANKEE'S STEAM GENERATORS Ol ~ W ) I f 4 16 ' \\. \\ \\. \\ \\" ~ f DRYERS g[ s I n u G .'f j g' .$S Y C N' DEC,K-i i %;C _.] MA NWAy _q Q" / 1 A i M EBUNME = pg , g1 l A^ \\ L Lf FEED RING b tc e R \\ PED 4\\PJP\\PJP92018.NEM

FIGURE.2 ORIGINAL J-TUBE CONFIGURATION AT MAINE YANKEE-3" X-STC (Carbon Steel) Fillet Weld ~ + Feed Ring E Abandoned .and Capped-Discharge Nozzle-Ra\\ PED 4\\PJP\\PJP92018.MEM ' l

FIGURE-3 1 l STANDARD AND SYSTEM 80 SEPARATOR CAN CC'.cIGURATIONS a wm _ (~ '( ell's.o.taan g .,ean %. p= ay~~;s,

,p.

~~' "E p' - m g - -- -

  • .J Ag g-

-- o- . g.G-m .~ ) .t N r. 2 i I g' G p _l y .g .1 I l. 4 c -., L)- e e. 2 ?D*

  • O fo.4 3

3 Inw; L l. 2. 5 'F' p g l-g u..-a e .- q - 1 %...ee.,', = a N -- t - ' L c,* STANDARD SEPARATOR CAN 3 SYSTEM 80 SEPARATOR CAN i R \\PE04\\PJP\\PJP92018.E n - - ' ~

FIGURE 4 EXAMPLES OF SEVERE J-TUBE DAMAGE gg I g '; g -.~, g3 y a f NL ' N4 ?s m , be- ~'

4..

) R:\\ PED 4\\PJP\\PJP92018.MEM i

FIGURE 5 EXAMPLES 0F SEVERE SEPARATOR CAN DAMAGE ]'7 .9 f *:., 5 ~

p;.....?

.z. , *j '. 4 s e,'e re ;', I

  • g i

' L * ;.d. f.1; ',, J i.$$~ 1fs ;'*B o e e <,3 e. e * ,s), i e ee *

,3 9 9 O O t

y ;_ (a) 7

f h-W I

-:.<j "i - J .Wdk..;;; slg[f&^# ' Weds;C /J' m f-%g a o e _g. _k e e e y . k e o e g (b) R:\\ PED 4\\PJP\\PJP92018.MEM

FIGURE 6 CONDITION OF SPINNER BLADE ASSEMBLIES ASSOCIATED WITH THE FOUR ORIGINAL INSTALLED SEPARATOR CANS STILL IN SERVICE IN STEAM GENERATOR 3. j 3 h R:\\ PED 4\\PJP\\ pip 92018.MEM ____w__ __a_--mm-:- - - - - - = - - - ""~ ~"""~~""~~~~ ~~"~ ""~ ^^ ' ' ~ ~ ~ ~ ' ^

FIGURE 7 FLOWING WATER INCREASES MATERIAL-LOSS RATE EXPONENTIALLY WITH FLOW VELOCITY.- CONDITIONS: 580 PSIG, 356F, pH = 7, 0, < 5 pg/kg, EXPOSURE TIME - 200 HR.- ^ 5000 1000 iQ _" 500 g\\ je e R O Bi V CW W a ,i 100 PpSO e. ~ //$ age T12) ~h~ 10 -4 5 21/4Cr-1Mo (A213 Grade T22) l ~ 3c E.- 1 = 0.5 i i r i 10 20 -30 40 m/sec lN32.8 65.6 98.4 131.2i ft/sec.: Flow: velocity, m/sec or ft/sec. Ra\\ PED 4\\PJP\\PJP92318. mot-

FIGURE 8 TEMPERATURE EFFECT ON EROSION / CORROSION. CONDITIONS: 580 PSIG, 115 FT/SEC, pH = 7, 0, < 40 pg/gk, EXPOSURE TIME - 200 HR 3000 Carbon steel 1000 1/2Mo 300 Grade TI) ~ } 4 %I~LAJ~ a Mo-Nb 100 ,f,,f A m. g 30 - 5 1Cr-1/2 Mo ki (A213 Grade T12) 39 E h3 \\ 2 1/4Cr-Mo E (A213 Grade T22) o 1 0.3 0.1 50 100 150 200 250 C 122 212 302 392 482 F Temperature, C or F Ri\\ PED 4\\PJP\\PJP92018.MEM

FIGURE 9 DECREASING pH (HIGHER pH NUMBER) REDUCES MATERIAL WEAR, PARTICULARLY BELOW pH 9.2. 1000 1/2Mo 5% ~ (A 161 drade T1). p= 580 ps ' T = 356 - 100 -. V = 128 ft /sec E SD ICr-1/2Mo Q

~~.,L, f (T22) '

A213 Grade cn -a N i d 10 1 1 5 \\ e t { Carbon steel. > \\ \\. t}k j P ambient E 1 ~ t q T= 167F E-V-5.25 ft/sec I [ 0,5 - Da = 20 pg/kg - \\, - t _{ \\ 2 \\l. m 0.1 i i i i 6-7 8 9 10 11. pH: At\\ PED 4\\PJP\\PJP92015.MEM _____________._____1._.___ ______..____..u.;

FIGURE 10 SURVEY OF CONDENSATE AVERAGE pH VS TIME AT-MAINE YANKEE 10.0,, ~ ~. ~ -~ ~ gp kkS.if.k j [_.g.: ... j..E. ' ?. 2 '.h' ...'..s N,. a.- 7p q R:. f,.. : ; 9 g ;..v .ye., -+ c.+. ...v g s%gg'jypt%sa;gs g :p e M. s g# 3 A l 9'0 ^ $" 5sf 1 _ [s s n i>4fl. I [' " [ /' E-d L L i w, n \\ aar ,_ _ f: _______4_ w e%. ) oc e ar me gy e-

  • 2 -

-- _ - -{ C

  • e.o CE & 4 PR RE V. ' Gilde ina i fo ' R To sKapper L yst im: )

'1914 m. W .y C. . g 7.0 C-. O L3 6.0 5.0 -1985' A J - 0 1986 A. J. 0 -l1967 A - J 0 1988 A J O 1989 A J 0 1990 A J O 1901 A J 0 1992 A J

Time R \\ PED 4\\PJPSPJP92018.MEM L

FIGURE 11 COMPARISON OF THE EFFECTIVENESS OF AMMONIA AND MORPH 0LINE IN REDUCING CARBON STEEL CORROSION (a) GRAPH OF EFFECT OF AMMONIA AND MORPHOLINE ON LIQUID PHASE pH VERSUS TEMPERATURE (FROM REF. 12) AND (b) PLOT OF EFFECT OF AMMONIA AND MORPH 0LINE ON CORROSION OF CARBON STEEL. 11 3 10-\\ Feedwater: E. \\ T = 25'C j \\ pH = 9.0 g_ lii Steam quality =90% 8-N = s 04 ~86-Neutralpy ~ 5 i 0 100 200 300 Temperature, *C (a) + 500 + $ATED ~ g C PAEuyA ~ x NON-PASEtVATED 5!H- + 41 1 I O 100 - 200 300 TIM E. h (b) R \\ PEDI.\\PJP\\PJP92018.NEM ____________________.-____o

FIGURE 12 REPLACEMENT J-TUBE CONFIGURATION Chrome - Moly 3" - 90* X-STC - Surfaced (Chrome - Moly Steel) Full Penetration Weld g SADDLE D Fillet Weld (Carbon Steel P-1 Mati) + Feed Ring Abandoned and Capped I Discharge Nozzle 1 Ra\\ PED 4\\PJP\\PJP92018.MEM

FIGURE'13 REPLACEMENT SPINNER BLADE CONFIGURATION-10.94 :$:- NOTE: BLEND RA.DIUS-8.00" 2.06" To i o q i-9- + 6 6 o. I O N At A puoclectos v "( *f "" 11 '"Dc

n MM l

[25 01.41s.03" n.as ui GRIND TO # M ^ 10 GAGE 16.0 0" z.0 6 ':' I. _.y N ' LEND-RADIUS-EQUIVALENT: ~ B 7.44":s: O SP INNER-- BL-ADE ASSEMBLY-- ~ ' MATERIAL: A 3871GRADEL22 x 3qs" lTHK.--. LPL Rt\\ PED 4\\PJP\\PJP92018.MEM ___________________________2

FIGURE 14 SEPARATOR CAN/ SPINNER BLADE FIELD ASSEMBLY DETAILS STEAM SEPARATOR CAN NOTE: BLEND RADIUS 1.50"=.50" OVERLp( 45* M e, FIELD INSTALLATION e R:\\ PED 4\\PJP\\PJP92018.NEM

ATTACHMENT 1 J-TUBE INSPECTION RESULTS AND REPAIR

SUMMARY

(3 PAGES) a \\. R:\\ PED 4\\PJP19JP92018.MEM l

Steam Generator #1 Feed Ring "J" Tubes (March 1992) gy) N 542'7 14 14 (oao> p2x2n13 12 o*# 0 12 p2242m ,0 *- 0 ( 2 " 2'7 0 I1 (.ua2m N2"2011 - 0 0 10 p214237 l < 2:42q 10 g p2042i-) \\ O\\ (o.1241s-) 9 g [0, l l c 8M M (o.10412")8 O D 7 pa n 227 (0217 7 9D6O O (o.101 5 5 [Hmw3nWallErosion) O Gry .gsh A O O 4 (a1241m p.m20s-) 4 - o Q D 3 M 1'4"T- '.(o.11-0.12") 3 O 2 (.18-) (Through Wall Erosion) 2 jy '(0.115-o.12") Nominal Wall = 0.3 (. ) - As found Ultrasonic Wall Thickness _$ Replaced (with Chrome-Moly) ( *: Replaced (with Carton Steel)

aa + 3 m I i r n -{. N n ~ e e o e E N Q N e g g 4 8 n s a c Q 1 E. ^ ~ o O h, c R G* O Q ~ i 9, O ,6 o g i Q d c[ M gm o 2 v-g $f 0 O = l M Q' a ~=( E"o O-1 i s* E 8 Q o is. .g N 1... Emi Q O

  • o-E s-~

4mp g v o - ag. O E[o f . Ens m -p g,g-O D h AL ~ W g w D c . g-

  1. g C

~~ a" O G. y n .g a jf l Q g,, a 'O y-gf -r e .o m g;fa c g a M. N O ^ ' C. C k b-20 n .c a 1 e 9*g ./ ,4 ka E s 4 o3m M' Q$ _d' 4U C ve .D o e .k k C 13 E - ky j a$uf ee an e-= bb--l k h h w 1

Steam Generator #3 Feed Ring "J" Tubes (March 1992) g,,g,g) <2m 14 14 (o28-) ( '8~ #0 13 iO*: O 12 (o22soan O i ( 22 man 12 o o 11 (024m (02H21711 O O 10( '6*2O O O (1010 g to.204.303 p . (o.2tn 9 0 o 8(021420 (o.os-o.127 8 .O-e tO a *D 7 (0.12-) (0207 7

  • * *
  • 6 (010 o

O (Through WaB Erosion) 5 5 (0 147 N . (Through Wall Erosion) 4 D 4 (0.12") .l t o. G 3 (010 (o.087 3 O Q {73,,,g3,,,, y,,,,,,) 2 1 (7,,,93,,,, g,,,,,,) ("77 l ' (o.1o") 1 J L Nominal Wall = 0.3 - I (. -) As found Ultrasonic Wall Thickness at-Replaced (with Chrome-Moly) (*,*: Replaced (with Carbon Steel) 1 1 l L L._____

m - ATTACHMENT 2- - 1. CE Letter MYC-84-078, "Mainef fankee Steam Separator Installation."- Dated-April 23, 1984, (4 pages).

2. 'ABB Letter CSE 92-136, " Flow Damage.- Maine. Yankee l Steam Generators."-

Dated March 24, 1992,~ (3 pages). 3. CE Infobulletin 92-01, " Steam Generator Component: Erosion - Corrosion Discovered.at Maine Yankee." Dated April 1,:1992,(3.pages).,

4. -- ABB Letter CSE 92-265,. Maine Yankee: Report " Root.Cause Evaluation of Maine-Yankee Steam Generator Separator Can and "J" Nozzle Damage Discovered L-During)the 1992: Refueling: Outage." Dated July.6, :1992 (2lpages).

y l 4 r i 6 s l

.j

.. u- -s s .. I, . As\\ PED 4\\PJP\\PJP92018.MEM : .-l ,r N.. a --____=_:_==-___ = -1

C-E P w:r Sy:t;ms Tot 2o3/683-1911 Combustion Engineenng. Inc. Teter 99297 1000 Prospect Hill Road Windsor. Connecticut 06095 - /93 POWER e, da H SYSTEMS p. /:: b/ v.. ~[ MyC-84-078 l April 23,1984 y f'.- Mr. S. LeClerc Maine Yankee Atomic Power Company RFD #2, PO Box 3270 Bailey Point, Ferry Road Wiscasset, Maine 04578

Subject:

Maine Yankee Steam Separator Installation

Dear Mr. LeClerc:

The ATH0S flow distribution computer code was used to model the steam flow through the replacement System 80 separators to be installed in the Maine Yankee steam generators in April /May 1984. The results of this analytical work are attached as Figure 1. Based upon these results, separator per-formance data, and previous analytical work, Combustion Engineering makes the following recommendations and-comments: (- l. .Since the total flow (steam plus liquid) distribution through the separators varies no more than 10% from the average (see Figure 1), orificing the separators would do little to help the overall carryever o performance of the new separators. Also, steam / liquid' distribution to the separators is such that a balanced loading occurs (i.e., the highest steam flows correspond to the lowest water flows and. vice versa); this balanced loading is beneficial to separator performs. ice. -2. To permit access to the tube bundle, it is recommended that'four of the old separators be left in place as shown on Figure-2. (A C-E drawing showing the new configuration and hardware will be issued within a week). Four is the minimum number of separators that must be left in place due to dryer drain pipe locations.and this location has lower carryover than the comparable set on the hot side. 3. The four old separators should be installed with a 5" orifice. This-5" orifice will reduce the flow thru these separators by approximately-25%. With orifices installed, these four standard separators will perform as well as the System 80 separators and will pose no moisture-carryover problem. 4. Since the pressure drop is essentially the'same in the new separators, they will have no affect on circulation or performance of the steam -generators. ka

Please do not hesitate to call if you have any questions. Very truly ycurs, C0'#;]STION ENGINEERING, INC. W ee_v w 0 R. C Jacques RCJ/REW/sms Project Manager cc: J. S. Randazza C. D. Frizzle E. C. Wood ( l l

c C Gil2[ / F1AlflE YANKEE STEAM FL0il DISTRIBUTI0fl Key Top Humber - Steam Flow per Separator x 10-3 lb/llr 100% atretch Power Level llottom flumber - Ilater Flow per Separator x 10-3 lb/llr Syneetry about 00-1600 Line Circled flumbers - flaximum flow rate per separator Average Steam Flow - 39,E)0 lb/hr/sep Boxed Numbers - flinimum flow rate per separator / seP 1 900 l 35.S i i3 o l 'lI. \\ .. i g 9o.h 3 3. (= s3 80 35.1 \\ t;5,9 \\ i t'l isa x \\ 39 /> 359 \\ ilanway 151 s-43 3Y7 ll? ~ K y I l9 IN l'/o t1anway l34.J l \\ efA 35.7 l'Is 3'l 'I 'f 5.9 fig 33,, N 38 7 ill i a 29 i s a. /03 4V 3 '#'h N:[ \\ , f <, III 43 9 T5 Dx 37 I 13 7 1I3 t/3.7 '[i , 'l* 7 '- 3r,.3 118/ i 3,9 ya i30 '00 ~ i ~~ 1800 llot Side Cold S;_,a L 1__e4-. _

.-u~..~-- ~-- - +. w.-_ ~~, _~- _n.-- - _ +.- ~.~-.-+x.,.,- -.n- ~. I .I - .~ a i ~ ) 4 k i M g' !.y 1 1 E) s W C.Z w a t:G-s ~ a c wu o t ,: n

o. ( M l

~ y "N. f o m t-MJ3 f vi 1,.] g,prirl ( ljli 'sl' +, { m's O O e'/ 9 s 1 vC l - a cco ,t p 8 1 's l t. i J2l og Q -i jj sliL alp ~ j { y L -rr-oj j Ji j [ E#r y - y - ylT9thr] l ?{'j(8 N l[k/. 1 L$g B-r Q@- r f y / y1 5 i .a y ~ M..! io, jgi t ,,J e T1 q '!q~* L ji \\ r_ i rw { .l l -- l ] .i)- . e, ..)

  • \\-

' 93 n yy ,1.) h- .* $ 2 [. k !- f-s l-u gg i .c s t

0 b c

e e w [- t ~-

ASEA BACWN BOVERI . March'24,1992 So,utheasthugar Service Center a ff.ICSE-dE'136 g; mig......l Mili Maine Yankee Atomic Power Company P.O. Box 408 Wiseasset, ME 04578 Attention: Mr. Steve LeClerc

SUBJECT:

FLOW DAMAGE - MAINE YANKEE STEAM GENERATORS

Reference:

(1) Fax data 70 pages, dated 3/20/92 (2) Fax data 32 pages, dated 3/21/92 (3) CE Calculation SS-26 " Evaluation ofIncreased Weight Steam Separator", dated 5/21/84 (4) Maine Yankee Drawing SK4100-1, Rev. O, dated 3/23/92 (5) MAP of 11 bladed versus 12 blad5d modified steam separating in S/G #1, dated 3/23/92 Feed Train Damage The referenced fax data includes UT measurements giving the actual conditions of the subject feedring as well as Maine Yankee (MY) evaluation and proposed fix. The data. reveals extensive flow damage to the 3" discharge elbows and very little damage to the remainder of the feed ring, feedwater distribution base and the thermal liner. ABB/CE is currently working with the CEOG to collect pertinent data from plant with damage to the feedring, feedwater distribution box or the thermal liner We suggest that the MY data be iceluded in the data base. ABB/CE concurs with your discharge elbow design change (MY sketch 92-1735-1). ABB/CE also believes that it is technically prudent to replace the damaged elbows defmed 3 on page 1 of reference (2). All replacement designated "necessary", " optional" and " optional / optional" should be replaced. The only exception could be elbow BB in steam generator 3. The UT results from the feed ring, the feedwater distribution box and thermal liner were reviewed and deemed acceptable i.e. no repairs are necessary. ABB Combustion Engineering Nuclear Power ' O d 7 """"

  • Sifd4."Y.L., s=2 Q{ej'JE o

Our recommendation obviously also includes an inspection at the next refueling outage. i This will not only confirm the status of the feed ring, the feedwater distribution box and thermal liner, but will also help in understanding the failure mechanism. Steam Separator Damage Repair of SteamSeparators As previously conveyed via telecons ABB/CE's understanding is that the lower part (spinner assembly) is being replaced with CR-MO steel. The new lower part v>ill be welded to the existing upper assembly in accordance with Reference 4. ABB/CE concurs that this repair technique is a technically sound and a prudent way to restore the integrity of the separators and return the plant to power operation. When connecting the new lower assembly to the old upper assembly, two or three rows of discharge holes may be lost. The separators will still function as intended, although some of die discharge margin has been removed. The new base will be fabricated from 3/16" stock rather than the original 12/10 GA steel. The number of spinner blades will be 11 or 12 versus the original ?2 blades only. The 11 blade spinner blade willimprove the fabrication process and snould have negligible affect on the steam separator performance. Eight (8) vane assemblies had one or more spinner blades protruding slightly (1/8" maximum) beyond the nominal location. There were a total of ten blades out oflocation in the eight (8) vane assemblies. The protruding material was rec.oved by mechanical means. The difference in blade location and contour are considered small and as such l the unbalance should be negligible. The installed locations of the 11 versus 12 blade assemblies are not considered critical l although ABB/CE suggests that accurate records be kept relative to their installed location. Locations similar to those defined in Reference (5) are acceptable. The modifications to the steam separators will have negligible affect on their ability to l remove moisture from tb steam. MY steam generators also have an additional margin of safety in the high capacity peerless dryers. These~ dryers are more efficient that the original cornigated plate dryers. l The modified cans (weight shift) will have no affect on the structural integrity of the can deck. The reference (3) calculation evaluates the structural ~ consequences from increasing the steam separators. The calculation clearly shows that sufficient margin exists, and minor weight differences (positive or negative) will have negligible affect on the results presented in the reference (3) calculation.

The modified steam separators if subjected to the same environment as the one causing the previous damage, could sustain erosion damage above the CR-MO steel base. The performance of the modified steam separators should be inspected at the next refueling outage. Although we concur that the repair technique is tecnnically prudent, no evaluation of the longevity of the fix was undertaken. From a technical standpoint, assuming identical operating conditions, it can be concluded that the replacement design will have equivalent or longer life than the origim.1 design. Sincerely, A J. H. Sodergren JHS:Jmj i l

NNN i M...~Ey g5 No.92-01 m .w ( h April 1,1992 Page 1 of 2 W J J Steam Generator Comoonent Erosion-Corrosion Dbeovered at Maine Yankee

== Introduction:== On March 14, a secondary side inspection of the No. 3 steam generator at Maine Yankee revealed damage to the steam separator cans as well as the feedwater J-tube nozzles. Based on this discovery, Maine O Yankee opened up the secondary sides of the No. I and No. 2 steam generators and confirtned that a similar 1 condition was present in all three steam generators. With support from ABB Combustion Engineering (C E), Maine Yankee implemented a comprehensive recovery effort to inspect and repair the damaged separator cans and J-tube nozzles. At this time, the degndation mechanism is suspected to be erosion corrosion. Steam separator cans and J-tube nozzles like those in Maine Yanken are in operation in other plants with C-E NSSSs. Discussion: At Maine Yankee, feedwater enters the steam generators through 28 J-tube nozzles welded to the feedwater ring. The feedwater mixes with the recirculating water from the dryers, and flows into the downcomer. au ( Upon exit at the bottom of the downcomer, the secondary water is directed upward over the vertical U tubes. g After leaving the U-tube heat transfer surface, the steam water mixture enters centrifugal type separaton. Each g generator has 98 separator cans. These cans have a fixed spinner blade assembly that imparts a centrifugal motion to the mixture which separates the water particles from the steam. The water exits from the perforated separator (2 housing and mixes with the feedwater. Final drying of the steam is accomplished by passage through eight banks of 5 inch book vane steam dryers which helps to limit the moisture content of the outlet steam to a maximum of mas G 0.2 % at the design flow. Visualinspection of the secondary side revealed erosion corrosion damage to the steam separator cans. Physically, W the damaged cans have eroded spinner blades, and in some cases the cans have through wall erosion of the can shellitself around most of the circumference in the vicinity of the top of the spinner blade assembly. Rese { separator cans an identical in design to those in the Palo Verde System 80 steam generators and similar in design to those found in other steam generators. They were installed in Maine Yankee in 1984 as part of an effon to O enhanc. secondarr sistem efficienc7 ameh Soon after the extent of the can damage was known, Maine Yankee made the decision to replace all of the spinner bhde assemblies of the separator cans in the three steam generators - a total of 294. The new spiacer blade ung assemblies are made of 21/4 Chrome 1 Moly and were wclded to the old top portion of the separator cans without omf any problems. The design of the separator was modified slightly from a 12-blade usembly to an 11-blade assembly. This design change is expected to have a negligible effect on performance. De number of blades in the spinner blade assembly was reduced to provide more room for the manufacturer to work int thereby, speed ng up the assembly process significantly. 0 Vimal inspections also revealed that several J-tube nozzles (90* elbows welded to the top of the feedwater ring) h were badly worn and some were completely missing. Hese J tube nozzles were installed in the feedwater ring as an upgrade in 1983 following a water hammer incident at Maine Yankee. He J-tube nozzles help to reduce the likelihood of damage in the event of another water hammer incident. After ultrasonic characterization, 38 out ( w a o. w ~ we.:. -.,am s=. .a.*- M.. SE. E. _.ur NJ.O.NYIv."N.n.5*t. m..SWMW., E"$tN."m7.nYdN YE* E s.ai E .eu.no C E octo290 mev. 9/853 I

N2. 92 01 April 1.1992 Pag) 2 ef 2 [ of a total of 84 J-tubes wers replaced. no replacement J-tube consists of a 21/4 Chrome-1 Moly elbow that is bench welded to a carbon steel saddle which is then field welded to the feedwater ring. no 21/4 Chrome 1 Moly Jebe nozz.les have significantly greate.r resis:ance to erosion-corrosion whilo retaining the same thermal expansion coefficient as the carbon steel saddles. ne carbon steel saddles were overlayed with 21/4 Chrome 1 Moly at the entrance to the elbows to increase its resistance to erosion-conosion degradation. Maine Yankee has also ultrasonically inspected the feedwater ring. ne resulu of that ultrasonic examin=tton d=eested that wear of the feedwater ring itself was n.inimA Geners!!y, the feedring pipe wall exceeded the minirnum required wall thickness except in areas near the damaged J-tube nozzles. Rese areas were subsequently reinforced by the new saddles that wens welded to the feedwater ring. Recommendation: CE recommends that utilities assess their plant's susceptibility to the erosion <orrosion damage observed at Maine Yankee by performing secondary Ide inspection of the separator cans and J tube nozzles during a refueling outage. Atelienbility: All C-E NSSS plants.

Contact:

Jan Sodergren (615n52 2833) ~ L c

a..

s[r evie c:n scavices : 7-7-92 eim, etsiszz et. cc:TT ssia 1 AEE ASEA BnoWN 8;lNERI Inter Office Correspondence To: Dan Zehler July 6,1992 Southeast Nuclear Service Center cc: J. H all CSE-92-265 G. Singh H. Harris

SUBJECT:

MAINE YANKEE REPORT " ROOT CAUSE EVALUATION OF MAINE YANKEE STEAM GENERATOR SEPARATOR CAN AND 'J' NOZZLE DAMAGE DISCOVERED DURING THE 1992 REFUELING OUTAGE" The subject report was reviewed by J. Hall, C. Singh and the writer. We concluded that the report is well written and properly approaches the techn!ct1 issues. The following comments are offered: 1. Velocity is a variable in the erosion phenomenon and as such did not receive the proper emphasis. Top of page 4 when discussing the J-tube erosion, the velocity contribution could have been discussed. The velocity is greater in the discharge elbows adjacent to the feedwater distribudon box. Also, local velocities can vary inside the elbows due to local disturbances. The depth ofinsertion of the elbow prior to welding to the.u:dring could affect the local disturbances and hence the local velocities. The random pattern of erosion may not entirely be the result of variations in ' tramp' chemical elements. All elbows were probably from the same heat and as such our experience is that the chemical composition would be rather consistent. 2. With regard to erosion of the cans, the flow velocity is an even more significant variable since we are dealing with a two phase flow. Single phase flow is subjected only to " dissolution wear" while two-phase flow is subject to both " dissolution wear" and " droplet impact wear". -The influence of velocity on " dissolution wear" is linear while " droplet impact-wear" varies with the fourth power of drop velocity. l Post it* brand fax transmittal memo 7571

  • *e posee. *)_

i

  • N 3e::.gJ f CJ2 P" L M _De eM c.,

on. G y Phone e Was f Paa a

svi e c:re se wiers s ?- 1 sim a s*57522:e1+ ectri asia a I i e Therefom it might not be surprising that Maine Yankee (with highly loaded l separators) is the only CE plant with eroded separators. It is not inconceivable that Maine Yankee's separator prior to the load increase ware not anbetad by erosion. Conclusion number 5 states that the change'in flow was so small, as a result of the load increase, that the effect was insigniht. Although we do not l necessarily disagree with the statement, the inauence of the flow increase may be'more significant than it seems to a casual observer. f 3. Reference 3 - The date of the bulletin should' read 1981. 4. Table I -- Fort Calhoun's discharge nipple flow damage was concluded to be the result of cavitation. 5. Table ! - The original Palisades steam generators did not have J-tubes. The replacement steam genemtors do. 6. Table 1 - Footnote 2 - All af ANO Unit 2 feodring flow damage occurred in or,e steam geserator. 7. We concur that replacing the eroded cans with 24 CR 1 Mo is a very good choice. Q J. H.~ Sodcrgren 1 JHS:Jmj I ^ V _}}

Retrieved from "https://kanterella.com/w/index.php?title=ML20141M287&oldid=4183697"