U

U.S. Department of Labor

Mine Safety and Health Administration
Industrial Park Road
RR1, Box 251
Triadelphia, West Virginia 26059

November 17, 1999

MEMORANDUM FOR LEE D. RATLIFF
                                  Assistant District Manager, North Central District
                                  Metal and NonMetal Mine Safety and Health

FROM:                       STEVEN J. LUZIK
                                  Chief, approval and Certification Center

SUBJECT:                  Executive Summary Analysis of Metal Samples from

                                  Kaiser Aluminum & Chemical Company's, Gramercy
                                  Plant Explosion

Metallurgical tests of evidence samples obtained from the Kaiser Gramercy Plant explosion were performed at Touchstone Research Laboratory Ltd., Triadelphia, West Virginia. Seven samples were tested as follows:

1) RJD001 Flash Tank #7 wall (thinned fracture)

Examination of fracture surface in the scanning electron microscope (SEM). Hardness measurements (4-5 rockwell measurements).

2) RJD002 Flash Tank #7 wall (thicker portion of fracture)

Examination of fracture surface in SEM.

3) RJD003 Dome of Flash Tank #6

Examination of fracture surface in SEM. Hardness measurements (4-5 rockwell measurements) Tensile samples (2 - Longitudinal orientation)

4) RJD004 Flash Tank #8 Wall

Examination of fracture surface in SEM. Hardness measurements (4-5 rockwell measurements)

5) RJD005 Flash Tank #9 Dome

Examination of fracture surface in SEM. Hardness measurements (4-5 rockwell measurements).

6) RJD006 Flash Tank #8 Cone

Examination of fracture surface in SEM. Hardness measurements (4-5 rockwell measurements). Metallographic cross-section.

7) RJD007 Blow Off Tank Dome

Examination of fracture surface in SEM. Hardness measurements (4-5 rockwell measurements).

The following summary conclusions are based on our review and analysis of the results from the Touchstone Laboratory tests: 1) All of the samples tested exhibited indications of a ductile fracture mode.

2) None of the samples that were tested exhibited any indications of metal embrittlement.

3) None of the samples that were tested exhibited any indications of metal fatigue.

4) The mechanical properties of each of the samples were consistent with their material specifications. Verification was made by preparation of an approximate conversion from Rockwell hardness to tensile strength for steel alloys. This conversion is referenced in the ASM Handbook, Volume 8: Mechanical Testing, 1985. The table below shows the hardness values (average HRB )obtained by Touchstone for each sample and the approximate tensile strength values obtained by conversion:

Sample ID	Average HR_B	Approximate Tensile Strength (psi)
RJD001	83	78,000
RJD003	81	74,000
RJD004	84	80,000
RJD005	85.5	83,000
RJD006	84	80,000
RJD007	81	74,000

5) The metallographic cross-section through the sample from the Flash Tank #8 Cone revealed that a portion of the wall thickness was as thin as 0.025.inches (the wall thickness just 6 inches away from this site was approximately 0.16 inches). It appears from the cross-sectional analysis that this portion of the cone wall had been thinned from the inside - most likely as a result of internal erosion.

In summary, no metallurgical defects were found in any of the samples examined. There were no indications of excessive fatigue or embrittlement of the steel used for the tanks. However, the cone section of the #8 Flash Tank exhibited excessive localized erosion. Therefore, although the steel itself had not been weakened, the tank(s) would have been weakened by the thinning of the wall.