1 / 34

ASP 2012 Workshop September 17-20, 2012

Can You be Certain?. ASP 2012 Workshop September 17-20, 2012. Error and Decision Uncertainty. Major Contributions to Error. Sampling and Chain of Custody Method Selection Precision and Bias at Reporting Limit Impact of Matrix on Precision and Accuracy. Real World Examples.

yasir-hill
Download Presentation

ASP 2012 Workshop September 17-20, 2012

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Can You be Certain? ASP 2012 Workshop September 17-20, 2012

  2. Error and Decision Uncertainty Major Contributions to Error • Sampling and Chain of Custody • Method Selection • Precision and Bias at Reporting Limit • Impact of Matrix on Precision and Accuracy

  3. Real World Examples • Sampling and Chain of Custody I received 2 containers for samples C0053 and C0054. The station location has been written on all of the received containers. My issue is was the label place on the containers first, or was the station location written on the lids first. For C0053, I have 2 containers with the correct label, but one container was received with tag 341 which is for C0054. The lid shows station location RB-073012 collected at 16:45 which matches C0054. I have the same issue for C0054. I received 2 containers with the correct labels, however, tag 337 was received with one of these containers and has the station location FB-073012 collected at 16:40 which matches C0053. I have included pictures to try to help with any questions.

  4. Real World Examples • Sampling and Chain of Custody Does it basically look like they wrong caps were placed on the containers and the wrong tags was subsequently put with those containers?  Let’s just say there were no times/station locations written on the container caps.  Would it look like the tags were just attached to the wrong containers? Use the container labels.

  5. Real World Examples • Sampling and Chain of Custody • Testing required is not specific , (Gamma, RCRA+, Radium) • Bottles Labels not consistent with COC (Dup) (Sampling Times) • Lets try to make the font so small it fits in the box • Wrong Bottles used for sampling • Hand Written

  6. Real World Examples • Method Selection - White Phosphorus (P4) by 7580 using ISM with Dry Grind and Sieve White phosphorus is a toxic, synthetic substance that has been used in poisons, smoke-screens, matches, and fireworks, and has been used as a raw material in the production of phosphoric acid. It has been used in smoke-producing munitions since World War I. White phosphorus is thermodynamically unstable in the presence of atmospheric oxygen. As a result, until recently, the prospect of long-term environmental contamination from smoke munitions was considered unlikely. However, a catastrophic die-off of waterfowl at a US military facility has been traced to the presence of P4 in salt marsh sediments, and lead to the realization that P4 can persist in anoxic sedimentary environments.

  7. Real World Examples • Method Selection • Volatiles on cement powder from drilling through pads. • Pb210 on Elk Meat • Poor Performing Analytes (8270 for Phenol)

  8. Uncertainty and Decisions Mapping Perceived Environmental Uncertainty, Yanes-Estévez V., Oreja-Rodríguez J.R. & Alvarez P. … Rasch Measurement Transactions, 2006, 19:4 p. 1033-4

  9. Uncertainty and Decisions

  10. Uncertainty and Decisions

  11. Decisions

  12. Improved Decisions 7 Step Process

  13. Improved Decisions

  14. Improved Decisions

  15. Measurement Error Measurement Error = Sampling Error + Analytical Error UFP-QAPP Estimating Measurement Error Project Quality Objectives and Measurement Performance Criteria (Section 2.6) Field Documentation Procedures (Section 3.1.2.6) Quality Control Samples (Section 3.4) QA Management Reports (Section 4.2) Data Review (Section 5.2). UFP-QAPP Worksheet 37 Describe evaluative procedures to assess measurement error

  16. Measurement Error Uncertainty is not error – Error is the difference between the measured value and the true value.

  17. Evaluating Measurement Error

  18. Analytical Error Major Contributions to Analytical Error • Sampling, COC and Method Selection • Precision and Bias at Reporting Limit • Impact of Matrix on Precision and Accuracy

  19. Analytical Error Analytical Error starts with method selection Selection of Methods The laboratory shall use methods for environmental testing, including methods for sampling, which meet the needs of the client and which are appropriate for the environmental tests it undertakes. If the wrong method is chosen the decision error approaches ∞

  20. Limit of Detection UFP-QAPP: Minimum concentration of a substance that can be reported with 99 percent confidence that the analyte concentration is greater than zero.

  21. Limit of Quantitation UFP-QAPP: The minimum concentration of an analyte or category of analytes in a specific matrix that can be identified and quantified above the method detection limit and within specified limits of precision and bias during routine analytical operating conditions.

  22. Reporting Limit UFP-QAPP: The reporting limit is the quantitation limitachievable by the laboratory. The reporting limit must be at or below the project quantitation limit.

  23. Action Limit UFP-QAPP: Action limit/level — The numerical value that causes a decision maker to choose or accept one of the alternative actions. It may be a regulatory threshold standard, such as a maximum contaminant level for drinking water; a risk-based concentration level; a technology limitation; or a reference based standard.

  24. Decisions, Decisions, Decisions AL Critical Value RDL RL QL SQL LLOQ LOQ MDC PQL BDL MDA LOD MDL DL LPL

  25. Decisions, Decisions, Decisions

  26. Precision and Bias at the LOQ/RL/AL LOQs without Precision and Bias The LOQ and associated precision and bias must meet client requirements and must be reported. • What are the client requirements here? • How do we generate precision and bias at LOQ? • Do we estimate or calculate this? • How is this related to decisions? • How do we report this?

  27. Analytical Error Precision and Accuracy on Matrix of Interest Matrix specific QC samples indicate the effect of the sample matrix on the precision and accuracy of the results generated using the selected method. • Do we generate precision and accuracy for method on project sample matrix? • Do we generate precision and accuracy on project matrix at LOQ? • Do we estimate or calculate this? • How is this related to decisions? • How do we report this?

  28. Proposal for Evaluating Analytical Error in the Matrix of Interest Across the Range of Decisions Part 1: Alternating Matrix Spikes at 2 times LOQ and 10 times LOQ for the project duration. Part 2: Use the data to generate analytical error to ensure decisions are ____________.

  29. Benefits of Evaluating Analytical Error on the Matrix of Interest • This will give us analytical error at the LOQ, on the Matrix of Interest and identify possible method selection errors. • Can be easily used to calculate analytical error • for the matrix of interest across a range of decisions

  30. When, What and How • When ever there is enough data to calculate. • Mean • Standard Deviation

  31. What, When and How • Mean Recovery is used for Bias (Mean%-100%) • Coefficient of Variation (CV) is used for • Precision (SD/Mean) • ResultB ± (ResultB (CV*2)) (B) - May be adjusted for Bias This is Analytical Error (95% CI)

  32. Conclusions Any intelligent fool can make things bigger and more complex... It takes a touch of genius - and a lot of courage to move in the opposite direction. Albert Einstein

  33. Robert P. Di Rienzo • ALS Environmental Quality Assurance Manager Fort Collins, CO and Salt Lake City, UT Bob.DiRienzo@ALSGlobal.com

  34. Questions? Thanks for Listening

More Related