Asbestos analysis methods for carpeting samples, 1993

NYC 9/11 Public Portal Document

fi

Toble I—Results of Carpet Anoiyses for Asbestos Samples Randomly Chosen in Cafeterio Carpet After Conventional (Dry) Vacuuming (ail values in asbestos structures per square centimeter)

Sample # LAB A (Sonication) LAB B (Sonication) LAB B (Microvac)

1 5,400,000 4,800,000 21,000 2 3,050,000 3,300,000 30,000 3 *68,000 *5,400 *<350 4 3,600,000 3,800,000 74,000 S 3,400,000 3,000,000 50,000 6 4,300,000 2,500,000 95,000 7 3,200,000 3,600,000 18,000 8 2,000,000 4,700,000 35,000 *ot detection limit

Dota on Precision and Percent Recovery containing I to 5% chrysotile. All furnishings had been removed from the cafeteria and the area had been vacuumed with a Studies on the precision and level of recovery of asbestos of the conventional dry vacuum cleaner twice before the samples were bulk-carpet piece ultrasonic shaking (sonication) technique and the collected. Previous use and traffic patterns were not taken into micro-vacuuming.technique were perfonned as pan of the 1988 account in collecting the samples. The samples were collected in EPA studySix samplc.s were collected using each method from a random manner and some samples may have been from area.s carpet artificially contaminated with approximately 1 billion (I x where an appliance such as a soft drink machine may have stood 1C’) iLsbcstos structures per square foot (s/ft) or 1.08 x 10'^su/cm-. previously. The data which include some interlaboratory compari- Tlie artificial contamination was accomplished by spraying a son.s are presented in Table I. Although there appears to be one 'known area of a carpet with a water .solution of known asbcsio.s outlier in the data set. the relative stands deviation was calculated concentration. Because there was no independent way to measure using all eight data points for each method. The S, for the the concentration of asbestos on the carpet, naaccuracy detenni- microvacuum technique was 77% and for the bulk-carpet .sonica­ nation could be made. However, the relative efficiency of recovery tion procedure was 51 % for Lab A and 47% for Lab B. The mean of one method to the other could be assessed. Tlic mean asbestos recovery of the microvacuuming technique compared to the bulk- recovery using the microvacuuming technique wa.s 2.3 x 10^ s/ft- carpet sonication procedure was 1%. (2.5 X IO*.s/cm-,. This was approximately 3% of the mean recovery of the hulk-carpet .sonication extraction technique which was 7.9 .x DiscuSsion IO** s/rt- (8.5 X 1(>5 s/cm-). Tile calculated coefficient of variation (CV) for the microvacuum technique was 166%. The CV for the The answer lo the question about which method is best for bulk-carpet sonication procedure was 43%. It should be noted that assessing the asbestos level in carpeting depends on the specific the values given in reference 2 for structures per square foot arc the question being asked. In the 1988 EPA carpet .study theauthors correct values.' * Tlterc wa.s an English/mctric conversion error in concluded that sonication of bulk-carpet samples provided a more the article which provided incorrect values forstructurcs per square precise and accurate estimate of asbestos concentrations in carpet meter throughout the pjipcr. than the micro-vacuuming sampling technique. Their conclusion A similar .set of tests using “real world” carpeting that had was based in part on data which showed that the microvacuuming been contaminated over a-15 -year period of normal use has been technique was shown torecoversignificantly less asbestos from the performed by the authors’ laboratories. The carpet samples were carpet than the ultrasonic extraction technique. collected from a cafeteria in the Social Security Administration in The fact that one technique is more efficient in recovering Baltimore. Maryland Thccafeieriahad an acoustical plasterceiling fibers than the othermay be important forsome studies investigat-

NYC-WTC_000144626