ML19210F035

Interim Deficiency Rept Re Discovery of Extremely Brittle Rebar Produced by Fl Steel Corp.Caused by High Manganese Content.Defective Rebar Heats Will Be Returned by 791116. Other Heats Found Acceptable

ML19210F035
Person / Time
Site:	Phipps Bend
Issue date:	11/15/1979
From:	TENNESSEE VALLEY AUTHORITY
To:
Shared Package
ML19210F034	List: ML19210F033 ML19210F035
References
NCR-PBN-044, NCR-PBN-44, NUDOCS 7912130431
Download: ML19210F035 (2)

Text

'

ENCLOSURE PHIPPS BEND NUCLEAR PLANT DEFECTIVE FLORIDA STEEL REBAR 10CFR50.55(e) REPORT NO. 1 (FINAL)

NCR PBN-044 On October 17, 1979, TVA notified NRC-0IE Region II, Inspector R. W.

Wright, of a potentially reportable condition under 10CFR50.55(e) regarding defective (extremely brittle) No. 6 rebar produced by the Florida Steel Corporation, Charlotte, North Carolina (hereinafter referred to as Florida Steel) for use at the Phipps Bend Nuclear Plant.

This is the final report on the subject reportable condition.

Description of Deficiency No. 6 rebar (bought by TVA to meet ASTM A-615, Grade 60) of heat lot C0-0006 was produced by Florida Steel on October 1, 1979, at their plant in Charlotte, North Carolina. The entire heat (39 tons total) was shipped to the Phipps Bend Neelear Plant and arrived on October 12, 1979.

This heat, C0-0006, of rebar was rejected and placed in "over, short, or defective status" upon arrival because it was so brittle that when bundles of the rebar were lif ted, some of the rebar in the bundles broke.

Following the rejection of the material, Phipps Bend employees took six samples from unbroken bars in that heat and four samples of rebar of the other heats, two each of heats C0-0005 and C9-3834, received on the same rail shipment as the defective heat. The samples were tested and analyzed at Phipps Bend and the TVA Singleton Test Laboratory with the following results:

1. All samples tested of heats C0-0005 and C9-3834 passed tensile testing performed both at Phipps Bend and at Singleton Laboratory.

2. a. Five of the six rebar samples of heat C0-0006 broke well below acceptable limits (90,000 psi required ultimate strength)in testing at Phipps Bend.

b. The other sample of C0-0006 passed tensile testing at Phipps Bend with a 94,000 psi ultimate strength.

c. Samples of fractured rebar sent to Singleton Laboratory were verified to have failed by brittle fracture.

d. Singleton analyzed samples of the rebar of heat C0-0006 which had broken by brittle failure and found a manganese content of 4.7 percent, a level which would cause the rebar to be extremely brittle and which does not appear to be a normal manganese level for any type of normally used steel.

e. Singleton analyzed a sample of " good" rebar from the C0-0006 heat and found normal rebar chemistry. The manganese level in it was 0.89 percent.

1552 331 7 9 nn so 93 /

3. Mill certifications provided with the CO-0006 heat from Florida Steel show normal rebar chemistry. The maganese level shown therein is 0.77 percent.

IVA concludes from the above data that rebar heat lot CO-0006 was extremely nonhomogeneous.

Cause of Deficiency The cause of the deficiency is the presence of the high manganese content of part of the rebar causing it tc be extremely brittle.

The high manganese content in part of the heat evidently escaped detection due to the nonhomogeneous of the batch of material and because the sample taken was from a.section of the heat which had acceptable properties. The Florida Steel QA manager indicated verbally to TVA that the heat was properly sampled with one bar taken from the middle one-third of the heat of a batch of less than 50 tons as required by the purchase specification.

Safety Implications Because of the obvious nature of the deficiency and the fact that it was discovered before the rebar being approved for receipt, TVA does not believe that this rebar deficiency would or could have gone undetected.

However, this occurrence caused TVA to question the Florida Steel QA program which allowed it to happen and bring to question the possibility of similar occurrences of a less obvious nature.

C_orrective Actions llalf of the defective heat of rebar was returned to Florida Steel f rom the Phipps Bend Nuclear Plant on November 9, 1979. The other half will be shipped from Phipps Bend on or before November 16, 1979. TVA also tested other rebar heats shipped to Phipps Bend on the same rail car as the defective heat (see previous discussion) and found the other heats to be acceptable.

Means Taken to Prevent Recurrence TVA has been informed by the Florida Steel QA manager that they will review melting procedures with individuals (melters) working in the melt shop. They will discuss the problem and provide training to shearing shop supervisors and crcws to alert them to methods of discovering deficient material before shipment. Florida Steel will document internally the retraining and review discussed above.

TVA Quality Engineering Branch (QEB) engineers have had numerous discussions concerning this situation with Florida Steel. The TVA Charlotte inspection office has increased surveillance of Florida Steel operations. Also, TVA (QEB) has scheduled an audit of Florida Steel's Charlotte, North Carolina, office for the first quarter of 1980. .

The actions discussed above should prevent the chance of future recurrences of a similar problem.

1552 332