This memo compares the methods and results from FLDOH and NAREL, and reviews the approach, results, and conclusions of the EPA NAREL report.
Both FLDOH and NAREL used high-purity germanium detector gamma spectroscopy to analyze an identical set of 21 drywall samples for 226Ra, 40K, 232Th, and 238U (greater- than-background concentrations of these naturally-occurring radionuclides may indicate potential phosphogypsum contamination). In addition to performing confirmatory measurements and a quality control/quality assurance audit, NAREL performed two types of statistical tests on the paired results from FLDOH and NAREL to test for a systematic difference between the two laboratories. NAREL also identified and discussed variations in the Florida procedures that differed from the procedures used by NAREL.
The Technical Team performed a data review of the results. In general there were no major differences between the averages of the radioisotopes detected by FLDOH or by NAREL as determined by the ANOVA statistical test comparing the domestic samples between the two laboratories. Similarly, there was no significant difference between the imported drywall samples. The ANOVA test determines if the procedures used by the two labs are statistically different.
The Technical Team also compared the imported drywall samples to the domestic drywall samples. The analyses were run for 226Ra content, 238U content, and 232Th content using an F-test which helps determine if two sets of data are similar. In this series of tests, Technical Team evaluations suggested that there were no significant differences in the two types of samples, the domestic drywall as compared to the imported drywall samples.
The analytical results from the FLDOH and NAREL laboratories were consistent. The few exceptions were explainable and do not change this conclusion. Notable differences are as follows:
The 1 standard deviation (1 ) was calculated differently by the two laboratories. FLDOH included only the counting error while NAREL calculated the total error. The reported errors compared favorably except for samples #40, 53, 56, and 58 (for 226Ra); #55 (for
40K), #56 and #58 (for 40K); #44, #56, and #57 (for
228Ra); and #42 (238
These observations do not impact the validity of the results and the inter-comparison of concentration values.
There are some distinct differences in concentration results reported by the two laboratories that were considered to be the result of using non-standard factors for gamma-ray intensity and differences in sample homogeneity and calibration source geometry.