REMOVAL_1995 - Laserfiche WebLink

Home Browse Search

(MDL) . All terminology, procedures and frequency of determinations associated with the laboratory's establishment of the MDL and the reporting limit should be well-defined and well -documented. Documented precision, bias, and MDL for all methods performed in the laboratory. information should be maintained 4.4.2 Control Limits Procedures should be in place for establishing and updating control limits for analysis. Control limits should be established to evaluate laboratory precision and bias based on the analysis of control samples. Typically, control limits for bias are based on the historical mean recovery plus or minus three standard deviation units, and control limits for precision range from zero (no difference between duplicate control samples) to the historical mean relative percent difference plus three standard deviation units. Procedures should be in place for monitoring historical performance and should include graphical (control charts) and/or tabular presentations of the data. 4.4.3 Laboratory Control Procedures Procedures should be in place for demonstrating that the laboratory is in control during each data collection activity. Analytical data generated with laboratory control samples that fall within prescribed limits are judged to be generated while the laboratory was in control . Data generated with laboratory control samples that fall outside the established control limits are judged to be ssuspectranddshould9ben ,out-of-control"repe ted or reported twi h qualifation. hese data are considered iers Laboratory Control Samples - Laboratory control samples should be analyzed for each analytical method when appropriate for the method. A laboratoy control ts of analytesrrepresentative l of othe starget eanalytes or ither a control spiked matrix certified reference material . Laboratory control sample(s) should be analyzed with each batch of samples processed to verify that the precision and bias of the analytical process are within control limits. The results of the laboratory control sample(s) are compared to control limits established for both precision and bias to determine usability of the data. Method Blank -- When appropriate for the method, a method blank should be analyzed with each batch of samples processed to assess contamination levels in the laboratory. Guidelines should be in place for accepting or rejecting data based on the level of contamination in the blank. Procedures should be in place for documenting the effect of the matrix on method performance. When appropriate for the method, there should be at least one matrix spike and either one matrix duplicate or one matrix spike duplicate per analytical batch. Additional control samples may be necessary to assure data quality to meet the project-specific DQOs. Revision 1 ONE - 19 July 1992