Laserfiche WebLink
accompanied each sample. To minimize sampling exposure and potential <br /> loss of sample integrity, the number of custodians in the chain of <br /> possession was kept to a minimum. Appendix G contains all chain-of- <br /> custody records. <br /> 5.2 LABORATORY QUALITY CONTROL <br /> Laboratory QC for the air quality testing program was provided by <br /> analyzing QC samples and comparing the results with the actual test <br /> data. QC control samples consisted primarily of blank samples, and <br /> spiked samples. Duplicate analysis of some samples was also conduct- <br /> ed. The results of this QC testing were used to assess the data's <br /> reproducibility, accuracy, and completeness by the methods recommended <br /> in the Guidelines. These methods and the results of the assessment <br /> are briefly described below. <br /> 5.2. 1 Data Reoroducibility <br /> Data reproducibility was evaluated by calculating the relative percent <br /> difference (RPD) of the results of replicate analyses according to the <br /> formula given in EPA Handbook for Anaivy t'Cal Quality Control in Water <br /> and Wastewater Laboratories (June 1972, Section 6) : <br /> RPD = I + x <br /> x1x2 I x 100 <br /> 1 2 <br /> where <br /> x1 = first result <br /> X2 = second result <br /> Reproducibility was calculated for each analyte in each set of dupli- <br /> cate samples. The reproducibility for each duplicated set is present- <br /> ed in Tables 8a and 8b. <br /> Of 38 duplicate pairs, the concentrations differed by more than <br /> 20 percent in 13, or in 34 percent of the data pairs. Most of these <br /> PJ9 9390304.00D 5-6 Rev. 0 09/20/88 <br />