CALFED California Bay-Delta Authority Mercury Studies Quality Assurance Oversight

CALFED/California Bay-Delta Authority Mercury Studies Quality Assurance Oversight Program Results from Intercomparison Study 2 November 2005 Presented by: Beverly H. van Buuren February 22, 2006 San Francisco Estuary Institute Third Annual Mercury Coordination Meeting Regional Water Quality Board, Oakland, California

QA Oversight Program Goals Data that supports Decisions Develop Comparability Between Hg Projects Performance Requirements Linked to Program Goals Build Comparability with other Programs Data of Known and Documented Quality van buuren consulting, llc www. vanbuurenconsulting. com slide# QA/QC Tools & Systems to Improve Efficiency

How Intercomparison Studies Help • Comparability of Data (bias) – within current project schedule – year-to-year – other programs? • • Alerts program/project/method/lab issues Individual Lab Performance (esp. CRMs) Data of Known and Documented Quality CBDA since March 2000 (6 years!) van buuren consulting, llc www. vanbuurenconsulting. com slide#

Intercomparison Study Schedule Analyte and Matrix Date Samples Ship Date Final Report THg Water MMHg Sediment June 2005 April 2006 April 2007 Nov. 2005 June 2006 June 2007 THg Water MMHg Water THg Tissue MMHg Tissue THg Sediment MMHg Sediment Nov. 2005 Nov. 2006 July 2007 March 2006 Feb. 2007 Nov. 2007 van buuren consulting, llc www. vanbuurenconsulting. com slide#

Intercomparison Study 2 Analyte and Matrix Sample Concentration THg Water spiked lake water 5. 56 ng/L THg Sediment IAEA-SL-1 130 ng/g THg Tissue IAEA-407 222 ng/g MMHg Water spiked lake water 0. 173 ng/L MMHg Sediment ERM-CC 580 75 ng/g MMHg Tissue IAEA-407 200 ng/g • 4 laboratories plus the referee lab • samples shipped 11/17, results back 01/15, draft report 03/06 van buuren consulting, llc www. vanbuurenconsulting. com slide#

Evaluation of Laboratory Results Ιz-scoreΙ Versus Reference Value Rating ≤ 2 ± 10% Very Good >2 and ≤ 5 >10% and <25% Good >5 >25% Poor van buuren consulting, llc www. vanbuurenconsulting. com slide#

THg in Water Results QA group and lab E are investigating systematically low-bias results to assess significance and causes van buuren consulting, llc www. vanbuurenconsulting. com slide#

MMHg in Water Results Difference between reference value and all lab results was less than 3 times the MDL (0. 020 ng/L) van buuren consulting, llc www. vanbuurenconsulting. com slide#

THg in Sediment Results All laboratories employed different analytical methods, however results indicate good comparability of data van buuren consulting, llc www. vanbuurenconsulting. com slide#

MMHg in Sediment Results Significant high-bias in lab C results likely due to artifact resulting from distillation method. Other labs use extraction method. van buuren consulting, llc www. vanbuurenconsulting. com slide#

THg in Tissue Results Lab A “good” very close to “very good” Lab D and lab E do not perform THg tissue analysis for this project van buuren consulting, llc www. vanbuurenconsulting. com slide#

MMHg in Tissue Results Study results indicate good comparison of MMHg in tissue results across all participating labs van buuren consulting, llc www. vanbuurenconsulting. com slide#

Closing Remarks • The MMHg in sediment results submitted by lab C highlight the potential problem associated with the distillation of sediment samples…high-bias resulting from the methylmercury artifact. Projects using this method should have a specific plan for assessing if this method is appropriate to meeting project objectives (the QA group can help). • One intercomparison study with only one sample per analyte/matrix combination is not statistically significant. • Based on the results of this intercomparison study, comparability of data generated by laboratories is good across all matrix/analyte combinations. van buuren consulting, llc www. vanbuurenconsulting. com slide#

Contact Information Beverly H. van Buuren QA Manager beverly@vanbuurenconsulting. com Van Buuren Consulting, LLC 4320 Baker Avenue Northwest Seattle, WA 98107 (206) 781 -1692 Other projects we’re working on: Surface Water Ambient Monitoring Program Training classes for Monitoring Design and QA/QC Holding Time Study on Low-level Nutrients Appropriate QA/QC for Physical Habitat and Bioassessment Studies Intercomparison for Pyrethroids in Sediment QA/QC for Citizen Monitoring QAPP Review for the SWRCB DFA van buuren consulting, llc SWRCB www. vanbuurenconsulting. com OIT CIWQS Database slide# Expert Software System to draft QAPPs

Milestone DMA-80 Hg Analyzer • Commercially available automated mercury analyzer • Uses thermal decomposition of sample followed by amalgamation onto gold trap and atomic absorption spectrophotometry (AAS) • Used for total mercury analysis of tissues and sediments • EPA Method 7473 was developed to cover analysis of THg using this instrument van buuren consulting, llc www. vanbuurenconsulting. com slide#

QA Considerations for the DMA-80 • Currently 3 laboratories participating in the program use this analyzer • EPA Method 7473 is vague regarding important practical details (such as how to perform matrix spikes) • Other QC parameters such as instrument calibration and accounting for carryover contamination must be detailed so they conform to the QC requirements of the CBDA Mercury Program • Important to address QC issues without losing benefit of analysis with this instrument (savings in time/labor) van buuren consulting, llc www. vanbuurenconsulting. com slide#

Resolution of QC questions for DMA-80 • Spoke with Milestone applications chemist, Dr. Skip Kingston, and Wes Heim (MLML) to resolve questions and develop QA policies • Calibration will consist of 10 -point “primary calibration” performed every 2 weeks and 5 -point “daily calibration performed before an analytical run • Matrix spikes can be performed by addition of standard directly to sample boat van buuren consulting, llc www. vanbuurenconsulting. com slide#

Resolution of QC questions for DMA-80 • Carryover contamination can occur after running greater than 250 ng of Hg through the analytical system • Results will be reviewed after analysis to identify potential carryover van buuren consulting, llc www. vanbuurenconsulting. com slide#

Evaluation of Laboratory Results • Performance is evaluated through the use of zscores • lab = the mean of the three values reported by the participating laboratory • xref = the reference value established for the study • ref = 0. 05* xref to evaluate results 10% from the reference value van buuren consulting, llc www. vanbuurenconsulting. com slide#