Laserfiche WebLink
<br />Statistical Sampling for Environmental Monitoring Sntdies: OSM Office of Technology Transfer, 1998 Ch. Xl Pg. 2 , <br />For an v set ojk measurements, the maximum oJ[he k measurements is bounded below by the <br />mean of the k measurements and bounded above by k times the mean <br />Xk S maX,(X) i5 kXk. <br />! ti ik <br />Using this inequality, we can calculate the mean of each of the composite samples, compaze the <br />upper lwund on the maximum value to some specified threshold (the action limits) and determine <br />if there is any possibility that the set of ksub-samples contains a value above threshold. In this <br />fashion, we can identify all composite samples which may have individual measurements above <br />threshold and restrict further investigations to only those sub-samples. The next step is then to <br />assay the individual sub-samples from identified composites and identify those locations which aze <br />above threshold. In the instance where k times the composite mean is below threshold for all sub- <br />samplc:s, there is no need for further laboratory analysis and the mine operator will realize a 75% <br />reduction in laboratory analysis costs. In the other extreme, if all samples have k times composite <br />mean above threshold, then all samples will require assay and there is no cost savings. Inmost <br />instances we expect that the majority of material is in compliance with a small proportion of <br />material over threshold. Following is an example using SAR measurements from a systematic <br />sampl: of regraded spoils in a reclaimed mine. The data aze supplied by Bob Postle, Office of <br />Surface Mining, Reclamation and Enforcement. <br />Table ]gives the actual data and the composite sample means if compositing had been conducted. <br />To de:terntine if action should be taken, we simply locate those composite means such that k times <br />the composite mean is above threshold, in this case all but the third composite mean is above <br />threshold, so action would be required on all but 4 of the original sub-samples. In this instance <br />most of the material is above the OSM criteria of 18%. In such a scenario, we may still realize <br />substantial savings on assay costs because it is appazent that some material is above threshold in <br />even group of 4 samples except one. To locate the individual samples that are out of compliance <br />further lab work is needed, however, it may make better sense to remediate at a scale of of 4 sites <br />at a time rather than do further lab analysis to verify that most samples are above threshold. <br />~P~~ <br />