ML14041A179

Attachment 3 - Examination of Millstone Unit 3 Service Water Valves for Dealloying

ML14041A179
Person / Time
Site:	Millstone
Issue date:	02/06/2014
From:	Dominion Nuclear Connecticut
To:	Office of Nuclear Reactor Regulation
References
13-596
Download: ML14041A179 (50)

Text

Serial No.13-596 Docket No. 50-423 ATTACHMENT 3 EXAMINATION OF MILLSTONE UNIT 3 SERVICE WATER VALVES FOR DEALLOYING MILLSTONE POWER STATION UNIT 3 DOMINION NUCLEAR CONNECTICUT, INC.

Attachment 3, Page 1 of 39 Work Scope

• A total of 9 valves were submitted to the lab for examination. Those valves were:

3SWP*V222 3SWP*V223 3SWP*V018 3SWP*V019 3SWP*V696 3SWP*MOV24A 3SWP*MOV24B 3SWP*MOV24D 3SWP*V860 P A section of the 30 inch valve 3SWP*V005 was submitted to the lab after it was cut in the plant.

9 Each valve was sectioned lengthwise every 90 degrees and then inspected for evidence of dealloying. Photographs were taken of the cut valve in the area where the dealloying appeared to be heaviest.

• Metallographic cross sections were removed from each valve in two locations where the dealloying appeared to be the heaviest. In 3SWP*V005, cross sections were taken in four locations along the piece submitted to the lab. Measurements were made of the extent of the corrosion occurring in each sample in the unetched condition.

• Photographs were taken of the microstructure of each valve after etching in potassium dichromate, and a picture of each cross section was obtained after a macroetch was performed using acidified silver nitrate.

Attachment 3, Page 2 of 39 3SWP*V222 3/4"

Attachment 3, Page 3 of 39 Photos taken of valve 3SWP*V222. The left photo shows the inside of the valve before it was cut. Some sediment buildup was present inside the valve. The lower right photo shows a section of the valve after it was cut. Evidence of dealloying was visible in the saw cut face as indicated by the discolored zone (arrows).

Attachment 3, Page 4 of 39 4'-

12:00 IJ N.

9:00 38m

.7 Micrographs taken from two cross sections from 3SWP*V222 showing the extent of the dealloying. Most of the dealloying was discontinuous. Scale 1/32".Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 5 of 39 3SWP*V223 3/4" WE ýTC.OTT MOW j--

Attachment 3, Page 6 of 39 tIýI~

Photos taken of valve 3SWP*V223. Some minor sediment buildup was present inside the valve (left photo). The lower right photo shows a section of the valve after it was cut.

Evidence of spotty dealloying was visible in the saw cut (arrows), which appeared heavier on the right side as aligned in the photo.

Attachment 3, Page 7 of 39 9

a 12:00 30m" 17 a~wn I -

9:00 Micrographs taken from two cross sections from 3SWP*V223 showing the extent of the dealloying. Again, most of the dealloying was not continuous. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 8 of 39 3SWP*V018 211

Attachment 3, Page 9 of 39 Y

'I I I~II 1jul11 Photos taken of valve 3SWP*VO18. Some minor sediment buildup was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut.

Minor evidence of sporadic dealloying was visible in the saw cut (arrow).

6:00 9:00 P-'

Micrographs taken from two cross sections from 3SWP*VO18 showing the degree of the dealloying.

Some dealloying was also occurring behind the seat (arrows). All of the dealloying was spotty. Original magnification right photos 15X. Scale 1/32". Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 11 of 39 3SWP*V019 211

Photos taken of valve 3SWP*VO19. Some moderately heavy sediment buildup was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut. Evidence of dealloying was visible in the saw cut (arrows).

12:00

,,4-3 74)

.i f 9:00 Micrographs taken from two cross sections from 3SWP*V019 showing the degree of the dealloying.

Most of the dealloying was uniform. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 14 of 39 3SWP*MOV24A 311

Attachment 3, Page 15 of 39

!Ad I

I Photos taken of valve 3SWP*MOV24A. A light to moderate coating of sediment was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut. Light, sporadic evidence of dealloying was visible in the saw cut.

ttachment 3, Page 16 of 39 a2

.4- 'k-6:00 F

3:00 I 4

Micrographs taken from two cross sections from 3SWP*MOV24A showing the extent of the sporadic dealloying. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 17 of 39 3SWP*MOV24B 311 N

Attachment 3, Page 18 of 39 I

Photos taken of valve 3SWP*MOV24B. A light to moderate coating of sediment was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut. Evidence of dealloying was clearly visible in the saw cut as indicated by the arrows.

6 :00 4W .

ilip 9:00 Micrographs taken from two cross sections from 3SWP*MOV24B showing the amount of the dealloying. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 20 of 39 3SWP*MOV24D 311

Attachment 3, Page 21 of 39 A

Photos taken of valve 3SWP*MOV24D. A very light coating of sediment was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut.

Only faint evidence of dealloying was visible in the saw cut.

12:00 W-.- o4

• 3:00 Micrographs taken from two cross sections from 3SWP*MOV24D showing the extent of the dealloying. Some sporadic dealloying was occurring behind the seat area (arrows). Scale 1/32".

Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 23 of 39 3SWP*V696 211

Attachment 3, Page 24 of 39 4

~ ~.>

'1: ~

71' Photos taken of valve 3SWP*V696. A moderate amount of sediment was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut. A thin layer of dealloying was visible in the saw cut (arrows).

Attachment3ý Page 25 of 39 IM

-4 a 6:00 S

I.

.5.

Y 4

• h ~,

e. 2m 3:00

• 4* ,

Micrographs taken from two cross sections from 3SWP*V696 showing the extent of the dealloying. Most of the dealloying was discontinuous. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 26 of 39 3SWP*860 311

Photos taken of valve 3SWP*V860. A moderate amount of sediment was present inside the valve (upper photo). The lower right photo shows a section of the valve after it was cut. A uniform layer of dealloying was visible in the saw cut (arrows).

28 of 39 6:00

'8~ 4P

- kr 4. A 9:00 41' .7 Micrographs taken from two cross sections from 3SWP*V860 showing the extent of the dealloying. The dealloying was fairly uniform. Scale 1/32". Original magnification right photos 15X. Etch left photos- Silver nitrate/Nitric acid. Right photos unetched.

Attachment 3, Page 29 of 39 Section of 30" 3SWP*VO05

Attachment 3, Page 30 of 39 9

0 4.

Pictures obtained from the cross section removed from location 1 outlined in the upper photo. The dealloying was sporadic at this location. Complete dealloying from the ID surface was only 0.5mm deep. Paths of the dealloying extended almost 5mm deep. Some corrosion also occurred from the flange edge (arrow).

Attachment 3, Page 31 of 39 ItAN W

9 Pictures obtained from the cross section removed from location 2 (outlined area in upper photo). The dealloying was also sporadic at this location. The deepest measured was 5.3 mm. Additional corrosion was also occurring from the edge of the sample (arrow).

Attachment 3, Page 32 of 39 Pictures obtained from the cross section removed from location 3. The dealloying was more uniform at this location, with complete dealloying extending 3.2 mm from the ID surface.

Attachment 3, Page 33 of 39

"* .11 Pictures obtained from the cross section removed from location 4. The dealloying was fairly uniform at this location. Complete dealloying appeared to be a little over 1mm deep but pockets extending almost 5mm deep were noted.

Attachment 3, Page 34 of 39 Microstructures (Arrows Indicate Dealloyed Areas)

Attachment 3, Page 35 of 39 Microstructures (Arrows Indicate Dealloyed Areas)

Attachment 3, Page 36 of 39 Microstructures (Arrows Indicate Dealloyed Areas)

Attachment 3, Page 37 of 39 Summary Valve ID Internal Condition Depth of Dealloying 3SWP*V005 Unknown Location 1- Uniform to 0.5mm, total to 4.8 mm Location 2- Sporadic attack. Total depth to 5.3mm Location 3- Somewhat uniform to 3.2mm, total depth at 6.7mm Location 4- Uniform 1.3mm, total depth 4.7mm 3SWP*V018 Light to moderate 6:00- Discontinuous dealloying to 2.6mm, some dealloying deposit layer. behind seat 9:00- Discontinuous dealloying to 2.8mm, some dealloying behind seat 3SWP*V019 Heavy deposits 12:00- Fairly uniform attack to 3.8mm 9:00- Fairly uniform attack to 2.9mm 3SWP*V222 Light deposit build 12:00- Uniform attack to 1.5mm, total depth to 4.4mm up 9:00- Uniform attack to 0.9mm, total depth to 3.8mm 3SWP*V223 Light deposit build 12:00- Uniform attack to 1.7mm, total depth to 3.9mm up 9:00- Uniform attack to 1.4mm, total depth to 3.4mm 3SWP*MOV24A Moderate deposit 6:00- Discontinuous dealloying to 4.1mm total build up 3:00- Very discontinuous dealloying to 3.8mm total

Attachment 3, Page 38 of 39 Summary Continued Valve ID Internal Condition Depth of Dealloying 3SWP*MOV24B Light to moderate 6:00- Fairly uniform attack, total depth 5.3mm deposit build up 9:00- Discontinuous, spotty attack, total depth 4.3mm 3SWP*MOV24D Light deposit along 12:00- Discontinuous attack to 2.3mm ID 3:00- Continuous to 1.3mm, total to 3.8mm 3SWP*V696 Moderate deposits 6:00- Continuous to 1.6mm, total to 4.8mm along ID 3:00- Continuous to 1.2mm, total to 4.2mm 3SWP*V860 Moderate deposits 6:00- Continuous to 4.2mm, total dealloying to 5.5mm along ID 9:00- Continuous to 3.4mam, total to 4.5mm

Attachment 3, Page 39 of 39 Plot of Maximum Depth of Dealloying in mm 7

6A 5

4 3

" Full Dealloying Depth 2 " Total Dealloying Depth 1

0 MOV24D

Serial No.13-596 Docket No. 50-423 ATTACHMENT 4 BEND TEST REPORT FOR MILLSTONE UNIT 3 SERVICE WATER VALVES MILLSTONE POWER STATION UNIT 3 DOMINION NUCLEAR CONNECTICUT, INC.

Letter Attachment STRUCTURAL TESTING OF 3-INCH ALUMIUMN-BRONZE VALVE REMOVED FROM SERVICE

1.0 INTRODUCTION

Full scale component testing was performed on a cast aluminum-bronze (AI-Brz) valve. The valve was removed from service due to reported dealloying of the AI-Brz casting. Photographs of the as-received valve are shown in Figure 1. The valve was marked by hand with the number "V-659" on one of the flange faces. The body is a split casing design joined together with six bolts. The valve was devoid of any internals. There are visible surface discoloration and corrosion deposits at isolated locations in the flange-to-body region.

The objective of the testing was to establish the structural capacity of the valve to maintain pressure boundary integrity under applied bending loading. This was accomplished by bend testing the valve in a beam assembly test rig. The bend test was performed by Intertek AIM engineers using the laboratory facilities of Anamet, Inc., located in Hayward, CA. The test was conducted on September 23, 2013.

2.0 DESIGN DATA The design information provided by Dominion on the valve and piping system is listed below (Ref.1):

Piping System: 3-inch Sch. 40 Piping material SB 466 Alloy 706 (Cu-Ni)

Valve Casting: SB 148 Grade C95400 (AI-Brz)

The specified minimum yield for the AI-Brz cast valve is 30 ksi (Ref. 2). The specified tensile strength is 75 ksi. The specified minimum yield for the Cu-Ni pipe is 13 ksi. The specified tensile strength is 38 ksi. The pipe material is not as strong as the valve so that the structural capacity of the piping system is controlled by the pipe strength.

3.0 TEST ASSEMBLY AND PROCEDURE A schematic illustration of the test fixture and pipe/valve assembly is shown in Figure 2. The valve was tested in 3-point bending as a pipe/valve assembly as illustrated. The pipe/valve assembly was constructed from Al 06 carbon steel pipe segments and Al 05 flanges that were butt-welded together and then bolted to the valve to form a beam structure. The test fixture for loading was fabricated from I-beam, plate, and angle iron members to transmit the machine load to the pipe/valve test assembly.

The pipe/valve assembly was instrumented with strain gages to provide additional data on beam loading distribution. The strain gages were mounted on the top and bottom sides of the pipe Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-1

L (relative to the neutral axis for bending). Photographs of the testing fixture and pipe assembly just before testing are shown in Figure 3.

The valve was oriented such that the suspected area of dealloying was located at the position of maximum outer fiber tensile stress. This was accomplishment by visual examination of the outside surface of the valve where through-wall leakage appeared to have occurred in service.

The approach to determine valve orientation for the test was discussed with Dominion prior to testing. Figure 4 shows the side of the valve that was loaded in tension.

The bend tests were performed to a maximum machine load of around 6000 kips following the pre-test procedure. The 6000 lb maximum test load was established based on a load that would produce a stress that exceeds limit load conditions for the Cu-Ni service pipe. The test was terminated when the maximum test load was achieved or failure of the valve was observed, whichever occurred first.

The following is the general procedure that was followed in the bend test (Ref. 3):

1) The test load fixture is assembled and attached to the testing machine loading platform.

The structure was verified to be secure and stable under load.

2) The pipe/valve components are assembled and fitted to the support structure. Internal support blocks were inserted in each pipe end at the load contact points to prevent ovalization of the pipe during the test.

3) Measurements of the geometry are taken and recorded to include the lengths of the pipe assembly and load points, locations of any strain gages relative to the load points, etc.

The recorded dimensions are given in Figure 5.

4) Prior to testing, the four strain gages were checked and zeroed as required.

5) The maximum loading rate was set at 200 lbs per minute. Unloading of the test was set at the same rate limit. Full unloading and reloading of the assembly was avoided so that plastic strain ratcheting of the AI-Brz does not occur.

6) After the test was completed, the assembly was visually inspected to check for shifting/movement at the load points or permanent deflections.

The load versus crosshead deflection of the loading machine was digitally recorded. In addition, the strain gage data was recorded by a data logger. The machine test results and strain gage measurements were correlated via the test time after start for cross-plotting purposes after start.

4.0 TEST RESULTS The 3-inch valve was loaded to 6,000 lbs with a measured deflection (crosshead travel) of 0.397 inch. The valve did not fail under the maximum test loading conditions. No permanent distortion (plastic deformation) was visibly apparent.

Figure 6 shows the load versus deflection plot generated from this test. The valve assembly response to the loading was relatively linear over the full load range. After the initial loading, there are two distinct linear portions in the load-deflection plot. The change in slope is most likely due to change in loading transfer across the bolted joints. Also plotted in Figure 6 are the Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-2

strain gage data for the four measurement locations. The strain in the pipe legs was symmetric and linear over the full load range.

The bending moment acting on the valve body at the minimum section between the flange and body was computed from the information given in Figure 4 and the following equations (Ref. 4),

M- PL 2 x (0 < x < L1 )

L M= PL 1 (L- x) (L1< x < L)

L The maximum test load for the pipe/valve (6,000 Ibs) corresponds to a calculated bending moment of 72,240 and 75,270 in-lbs at the two flange regions (area between the pipe flange and valve body). The maximum bending moment of 75,270 in-lbs would produce an equivalent bending stress in the Cu-Ni pipe of 43.6 ksi. This stress exceeds the specified ultimate tensile strength of the system piping material.

5.0

SUMMARY

AND CONCLUSIONS The structural capacity of the valve to carrying pipe bending loads was the subject of the test program. The following findings and conclusions were drawn from this test:

1) The load-deflection curve for the test assembly showed linearly behavior to the maximum test load of 6000 lbs.

2) The valve did not fail or show visible distortion under the test loading conditions.

3) The stress at the valve section where visible surface discoloration was observed was calculated to be 43.6 ksi based on the nominal pipe size for the system.

4) The structural capacity of the valve, in its current degraded state, exceeded the structural (limit load) capacity of the Cu-Ni system piping.

REFERENCES

1. Email from R. Schonenberg (Dominion) to R. Cipolla (Intertek), "Valve Bend Test,"

(9/18/2013)

2. ASME Boiler and Pressure Vessel Code,Section II, Part B, "Non-Ferrous Materials,"

(1995 Edition)

3. "Procedure for Conducting Bend Tests of Dealloyed Piping Components," Intertek Project Procedure, (September 2013)

4. Roarke, R., J., Formulas for Stress and Strain, 4th Edition, McGraw Hill, (1965)

Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-3

Figure 1 - As-Received 3-inch Valve Body Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-4

BEND TEST CONFIGURATION 3" Sch 40 Pipe Not to Scale OD = 3.5" t = 0.216" Figure 2 - Valve Bend Test Schematic Configuration Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-5

SG 3 (C)

SG 4 (T)

F(C)

Figure 3 - Photographs of Valve Bend Test Assembly Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-6

Figure 4 - Tension Side of Valve during Bend Testing Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-7

p Lgl I

I I

-I I-Im Lg2 I

II l L1 L2 F-* x Parameter Dimension, (in)

Overall Length, L 55.76 Length to Load Point, L, 27.88 Length from Load Point, L2 27.88 x, = 24.08 Length to Valve Necks X2= 30.67 Lgi= 17.56 Length to Strain Gage, Lg Lg2 = 18.44 Figure 5 - Bend Test Dimensions Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-8

Bend Test inch Valve Load vs Displacement 8000 7000 6000 5000

.0 a: 4000 "O

0

.- I 3000 2000 1000 0 4 i 0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 Crosshead Displacement, (inches)

Bend Test inch Valve Load vs Microstrain 8000 7000 6000 5000 a:.0

-J 4000 3000 2000 1000 0

-1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 Gage Microstrain, (x 1,000)

Figure 6 - Bend Test Results for 3-inch Valve Intertek AIM Dominion Nuclear Connecticut AES 13058423-2-1, Rev. 0 January 2014 A-9