ML20214F386

Submits Addl Info in Response to Deviations Noted in Insp Repts 50-424/86-11 & 50-425/86-06 on 860424 & . Metallurgical Bases for Final Evaluation Plan Used to Investigate Support Failure Encl

ML20214F386
Person / Time
Site:	Vogtle
Issue date:	11/06/1986
From:	Conway R GEORGIA POWER CO.
To:	Grace J NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION II)
References
GN-1167, NUDOCS 8611250286
Download: ML20214F386 (7)

Text

F'

^

S mB

==r AM;ra Gempa 30308 Te4 pum 404 SM 6724 f.% rgAd;e e m ?C N t Bc 4545

.. w a G << p;i 3rrs ?

3G !!0V 17 A 7 : 33 R.E.Conway

s. r a ', & em w,-

November 6, 1986 United States Nuclear Regulatory Commission File: X7BG10 Region II Log: GN- ll67 Suite 2000 101 Marietta Street, Northwest Atlanta, GA 30323

Reference:

50-424/86-11 and 50-425/86-06 Attention: Mr. J. Nelson Grace Georgia Power Company wishes to submit the following additional information in response to the deviation in your inspection report 50-424/86-11 and 50-425/86-06 dated April 24, 1986, and your correspondence dated October 8, 1986.

In response to Concern #2 of your letter, it was not Georgia Power Company's intention not to clearly communicate to the USNRC the modifications in the field inspection program by our letter of December 19, 1983 (Log: GN-294).

GPC, with good intent, believes that USNRC was kept informed of the investigation on support failure from the date it was determined to be potentially reportable to the date of the final evaluation report. Due to the technical bases described in tha attachments, the original program reported to NRC in the July 11, 1983 letter was modified. This modified program was reported to NRC in the update provided by the licensee in the December 19, 1983 letter. The final report provided by the licensee in the February 15, 1984 letter is consistent with the December 19, 1983 letter.

While our letters did not specifically state that the original proposed plan to inspect samples of weldments fabricated from heats 7417461 and 7419919 had been modified, the final evaluation that was conducted was adequately described. Since our earlier evaluation had concluded that the support failure is not related to a heat-specific condition, specific mention was not made of the changes in the inspection plan relating to the samples from the above two heats.

Georgia Power Company will continue to make every ef for t to maintain clear and open communication with the USNRC on all matters concerning Plant Vogtle.

'I 8611250286 861106 PDR ADOCK 05000424 G

V Tf PDR

(

a

. o Page Two In response to Concern #1 of your letter, Georgia Power Company (in our letter dated May 23, 1986, GN-920) has provided further technical explanation as to why the modification to initial field inspection program was justified. We are prcviding further technical justification in Attachment 1.

To further confirm the technical adequacy of the modified plan, GPC has since performed a documentation review and field walkdown to specifically identify embed plates fabricated from heat numbers 7417461 and 7419919 with welded attachments. A total of thirteen (13) plates were located - ten (10) from heat no.

7417461 and three (3) from heat no. 7419919. In addition, one weldment from the original statistical random sample plan was subsequently identified by document review to be from heat no.

7419919.

The resultant total of fourteen (14) embed plates (four with two welded attachments) exhibited varied weldments including tube steel attached by flare bevel welds with up to 3/16 inch fillet reinforcement (total of four) and tube steel, angles and W-shapes attached by fillet weld sizes in the range of 3/8 - 3/4 inch (total of fourteen). Further details are provided in Attachment 2.

The subject weldments were ultrasonically examined using equipment and methods outlined in ASME Section V, Article 5.

Since lamellar tearing was the suspect discontinuity, the area of interest was limited to the embed base material to a depth of 1/4 inch below the weld and to a point 1/4 inch beyond the root and toe of the weld. No recordable indications were identified in the fourteen weldments.

Based on the review and evaluations conducted, Georgia Power Company has concluded that the investigation performed by the Georgia Power Company of the support failure is adequate. The technical adequacy of the investigation has been further confirmed by the ultrasonic examination of a sample of support weldments fabricated from heat number 7417461 and 7419919.

In summary, the investigation performed earlier by GPC does adequately address the support failure issue as documented by our February 15, 1984 letter (GN-318). Further investigations and additional technical justifications, discussed above, supports our conclusions stated in this letter. GPC shall continue to make every effort to maintain clear and open communication with i USNRC on all matters concerning Plant Vogtle.

F ]

Page Three These responses contain no propietary information and may be placed in the NRC Public Document Room.

Yours truly, R. E. Conway DLC/ REC /sl xc: U. S. Nuclear Regulatory Commission Document Control Desk Washington, D.C. 20555 H. G. Baker D. R. Altman L. T. Gucwa J. P. O'Reilly P. R. Bemis C. W. Hayes G. F. Head J. A. Bailey G. A. McCarley O. Batum R. W. McManus R. H. Pinson G. Bockhold, Jr. Sr. Resident (NRC) P. D. Rice C. E. Belflower C. C. Garrett (OPC) B. M. Guthrie J. F. D'Amico J. E. Joiner (TSLA) D. E. Dutton W. D. Drinkard D. Feig (GANE) R. A. Thomas E. D. Groover NORMS 4

HQSil86023/SL

i ATTACHMENT 1 l

METALLURGICAL BASES FOR THE FINAL EVALUATION PLAN This attachment describes the metallurgical bases for the final evaluation plan used to investigate the support failure.

Specifically, the bases for the following statement are discussed:

"It is very unusual to see inclusions and banding close to plate surfaces. When these conditions do occur, it is usually in the center one-third of a plate. When inclusions in plate steel are close to the surface, the potential for lamellar tearing increases."

There is ample technical data demonstrating that the poorest short transverse tensile results and propensity for lamellar tearing occur near the plate thickness center. Poor short transverse performance is primarily attributed to nonmetallic inclusions and is a measure of susceptibility to lamellar tearing.

Numerous authors have cited variable results when evaluating the propensity for lamellar tearing, indicating that it varies not only from ingot to ingot, but with the plate's relative position in an ingot, and, also, with the position in the plate.

Kaufman, Pense and Stout (Welding Journal, March 1981) clearly demonstrated that lamellar tearing susceptibility increases with increasing depth below the plate surface layer, being greatest at the midthickness. In other words, the nonmetallic inclusions that contribute to lamellar tearing are more common in the center of the plate thickness than at the surface of near surface regions.

When any steel solidifies, the process starts at the outside and progresses from the ingot mold walls toward the center and from the bottom upward. Silicates tend to float upward due to buoyancy and convection. Sulfur also tends to segregate to the center in ingots. Thus, because plate midthickness corresponds to ingot center and plate surface corresponds to the ingot outer surfaces, fewer inclusions are expected at the outer surfaces.

In the USS Technical report by L. F. Porter, reference is made to Spaeder's work, which showed that the poorest short transverse results occurred with steel plate made from the top cuts at the ingot axis.

Metallurgical fundamentals, Stout's data, and Spaeder's data corroborate the Bechtel-M&QS report statements, that extensive inclusions near the surface are unusual.

MQS1186023/ ATTACHMENT /SL

Page Two The steel of concern at Vogtle is a slightly semikilled steel, having 0.03 and 0.04 percent silicon. For steels of this type, it is expected that the harmful inclusions would tend to be concentrated in the top center of the ingot; therefore, fewer harmful nonmetallics, silicates, and sulfides would be expectad at the plate surfaces. With such low silicon content, relatively few silicates would be expected. For this reason and because of the nature of the manganese sulfides which form in semikilled steel, the USS report stated:

"Semikilled steels would be expected to have better resistance to lamellar tearing than fully killed steels of similar sulfur content."

Farrar (Welding Journal, August 1974) reported on two steels which are very similar to the steels in question except that his steels were killed and should be more uniform than the semikilled steel at Vogtle. The short transverse test results, which provide some correlation to lamellar tearing propensity, vary with location across a plate width as well as location through the thickness direction and vary with the plate's original location within the ingot. With so much variation within a heat of killed steel, it is reasonable to have concluded that recurrence would not be uniform with the heat of semikilled steel.

Stout (Welding Journal, March 1981) stated that:

"The mechanics by which lamellar tearing occurs during welded fabrication of steel structures is now well understood. Fortunately, lamellar tearing is generally detected during fabrication and has rarely been held responsible for service failures."

Banded steel structures (ferrite bands are mixed with pearlite bands) are related to inclusions. Quoting the USS Technical Report:

"Because the inclusions tend to lie in or along the bands it is difficult to separate the influence of inclusions from the influence of banding".

The only study directly related to lamellar tearing was that of Farrar & Dolby and they concluded that pearlite banding and microsegregation of alloy elements adjacent to inclusions have little or no influence on susceptibility to lamellar tearing.

MOSil86023/ ATTACHMENT /SL

~

L

(-

Page Three In summation, it can be concluded from the report and technical articles that:

1) Poor short transverse performance is primarily attributed to nonmetallic inclusions and is a measure of susceptibility to lamellar tearing.

2) Banding exerts only a secondary influence on the susceptibility of the plate material to lamellar tearing.

3) Since nonmetallic inclusions are most likely to occur in the center of the plate thickness, which is in agreement with the previous project conclusion that "it is unusual to have inclusions near the plate surface".

4) When the variability of the nonmetallic inclusion problem is coupled with only one found failure and the knowledge that such failures are found during construction, a modified inspection plan based on statistical methods is prudent and adequate.

/

MQSil86023/ ATTACHMENT /SL

ATTACH!EllT 2  :

EMBED HEAT NO. SUPPORT NO. APPLIED LOADS / ALLOWABLE LOADS ATTACHMENT / WELD SIZE & TYPE  :

Pz(4) Vx(4) Vy(4) Mx(5) My(5) Mzi5) .

I 7317461 V1-1210-122-H606 0.18/1.0 0.3/1.0 0.3/1.0 13.11/25.6 9.15/25.6 8.1/200 Tube Steel / 1/2 inch fillet 2 7417461 VI-1210-122-H605 0.3/.1.0 0.3/1.0 0.18/1.0 3.06/16.2 9.3/16.2 5.94/200 Tube Steel / flare bevel. 1/8 fillet 3 7417461 VI-1210-057-H001 0.047/1.0 0.030/1.0 0.045/1.0 0.720/9.9 0.480/9.9 ----/200 Tube steel / 3/8 inch fillet 4 7417461 VA-1217-219-H002 0.3/1.0 0.3/0.5 0.3/0.5 5.08/10.0 5.08/10.0 3.3/200 Tube Steel / 3/8 inch fillet 5 7417461 VI-1210-057-H002 0.028/1.0 0.028/1.0 0.041/1.0 0.656/9.9 0.446/9.9 ----/200 Angle / 3/8 inch fillet V1-1210-122-H604 0.3/1.0 0.18/0.5 0.3/1.0 11.25/27.1 5.94/13.55 5.13/200 Angle / 3/8 inch fillet 6 7317461 VA-1314-010-H007 0.278/1.0 0.0/0.5 0.925/1.0 ---- ----

0.0/200 W-shape / 1/2 inch fillet 7 7417461 VA-1314-010-H008 1.722/2.0 0.798/1.0 0.830/2.0 26.09/31.6 12.73/15.8 17.28/200 Tube steel / 3/4 inch fillet 8 7417461 VI-1202-057-H003 -----/1.0 0.055/1.0 0.160/1.0 2.688/9.9 0.880/9.9 ----/200 Angle / 3/8 inch fillet VA-1217-219-H003 -----/1.0 0.030/1.0 0.045/1.0 0.720/9.9 0.480/9.9 ----/200 Angle / 3/8 inch fillet 9 7417461 VA-1217-211-H002 0.3/1.0 0.0/0.5 0.3/0.5 6.08/20.45 0.0/20.45 28.56/200 Tube steel / 3/8 inch fillet 10 7417461 VI-1217-196-H006 0.245/1.0 0.280/1.0 0.378/1.0 6.46/11.7 4.56/11.7 7.10/100 W-shape / 3/8 inch fillet V1-1314-029-H008 0.069/1.0 0.104/1.0 0.098/1.0 2.19/7.1 1.66/7.1 2.83/200 Tube steel / 3/8 inch fillet

!! 7419919 VI-2419-010-H601 0.266/1.0 0.263/0.5 0.352/0.5 2.84/7.2 5.51/7.2 4.98/100 Tube steel / flare bevel V1-2419-010-H602 0.266/1.0 0.263/0.5 0.352/0.5 2.84/7.2 5.51/7.2 4.98/100 Tube steel / 1/2 Inch fillet 12 7419919 V1-2301-155-H002 0.5/1.0 0.0/0.5 0.5/0.5 ---- . .-

0.0/100 W-shape / 3/8 inch fillet 13 7419919 X4CP-R5180 0.268/1.0 3.418/4.0 3.235/4.0 4.82/16.4 4.37/16.4 60.63/100 Tube steel / flare bevel 14 7419919 DS-1118103-270 0.3/---- 0.6/---- ----

18.0/---- ---- ----

Tube steel / flare bevel. 3/16 fillet NOTES: 1. Embeds 1 through 10 are shown on Dwg. 8AX2D08A023 Section F of AX2D08CO23

2. Embeds it. 12 and 13 are shown on Dwg. #AX2D08A025 Section G of AX2D08CO23

3. Embed 14 is shown on Dwg. #AX2D11A002 Section K of AX2D11C020

4. All values are in Kips

5. All values are in Inch-Kips s