USGS Sediment Laboratory Quality Assurance Project Gary P. Johnson

1. USGS Sediment Laboratory Quality Assurance Project Gary P. Johnson Sediment Data Collection Techniques March 25, 2014

2. Goals of the Sediment Laboratory Quality Assurance (SLQA) Project • Ensure the physical sediment data produced or used by the USGS-WRD are of a known quality • Evaluate laboratory bias and variability • Document long-term data comparability and consistency

3. Project Guidance The USGS SLQA Project is managed by: Water Resources Discipline Office of Water Quality Branch of Quality Systems (BQS) The Office of Surface Water and Office of Water Quality collaborate on issues pertaining to sediment and water quality.

4. SLQA Participants Since 1997, over 30 SLQA studies have been conducted, usually twice per year. The most recent SLQA study, conducted during Oct 2013, included laboratories from the following sectors: • 9 USGS • 6 Contract (4 State, 1 City, 1 University) • 8 Volunteer (3 State, 2 Federal, 2 Private, 1 University) Participation in the SLQA studies is required for USGS-WRD and contract laboratories.

5. Sample Preparation Two SLQA studies are conducted annually, during spring and fall. Each laboratory receives 9 sediment-water mixtures and is asked to determine sediment properties that they perform routinely: • Suspended sediment concentration • Sediment mass • Fine material mass (< 0.063 mm) • Sand material mass (0.063 - 0.250 mm) • Particle size distribution (< 0.063 mm)

6. SLQA Sample Properties

7. Material Used in SLQA Samples Fine material • Test dust collected from Salt River Valley, Arizona • Material is sieved to < 0.063 mm Coarse material • Salic material collected from South Dakota • Sieved to 0.063 – 0.250 mm All material is dried at 105o for at least 24 hours, prior to sample preparation

8. Evaluation of SLQA Results Percent Difference = [((Reported Value – Target Value)/Target Value) * 100] Percent difference for each sample, for each parameter is plotted on a control chart (Suggested) Acceptable Criteria Control limits: • + 3 f-pseudosigma from median percent difference ( f-pseudosigma = [(75th percentile – 25th percentile)/1.349] ) • + 5% from median percent difference whichever value is greater

9. Control Chart Example – Class 1 Sediment Laboratory Quality Assurance Program - Study 2, 2013 Sediment Concentration Percent Difference Results Class 1 Samples: target material mass = 66 mg, 66 mg, 72 mg; target SSC = 165 mg/L, 165 mg/L, 180 mg/L Sediment Concentration Percent Difference Percent difference = [((Reported value – Target value)/(Target value)) * 100]

10. Control Chart Example – Class 3 Sediment Laboratory Quality Assurance Program - Study 2, 2013 Sediment Concentration Percent Difference Results Class 3 Samples: target material mass = 2640 mg; target SSC = 8800 mg/L, 8800 mg/L, 6600 mg/L Sediment Concentration Percent Difference Percent difference = [((Reported value – Target value)/(Target value)) * 100]

11. Suspended Sediment Concentration Class 1 Class 2 Class 3 Suspended Sediment Concentration Percent Difference SLQA Studies conducted 1997 - 2009

12. Sediment Mass Class 1 Class 2 Class 3 Sediment Mass Percent Difference SLQA Studies conducted 1997 - 2009

13. Fine Material Mass Class 1 Class 2 Class 3 Fine Material Mass Percent Difference SLQA Studies conducted 1997 - 2009

14. Sand Material Mass Class 1 Class 2 Class 3 0 Sand Material Mass Percent Difference SLQA Studies conducted 1997 - 2009

15. Scanning electron microscope photograph of fine-size material aggregation and adherence to larger sand-size material

16. Follow up Evaluations Purpose • Identify and mitigate sources of error to improve data quality Procedure • Laboratories are asked to analyze additional quality assurance samples in a follow up evaluation to a SLQA study when more than 2 of the 9 determinations for each parameter do not meet the acceptable criteria

17. SLQA Double Blind Sample Program Purpose • Provide quality control sediment-water samples which can be submitted to laboratories for analysis, disguised as environmental samples • Laboratory is not cognizant of the sample’s properties and quality assurance status (double blind) Objective of using double blind samples • Identify magnitude of bias and variability in environmental data due to analytical method

18. Double Blind Sample Program Effects of Bias and Variability (from USGS Quality Control Sample Design and Interpretation Course) • Abundance of constituents is misrepresented • Bias that is not constant over time can be misinterpreted as environmental differences or changes • Variability impedes ability to detect differences in constituents and makes relation of variables more difficult to define

19. Double Blind Sample Program Ordering samples BQS web site: http://bqs.usgs.gov SLQA Project Project Contacts

20. Double Blind Sample Program Ordering samples • Contact Mark Woodworth • Specify target sediment concentration Target Concentration (mg/L) = Total sediment mass (mg) Total water volume (L) • Selectfine and sand mass • Choose number of samples

21. Double Blind Sample Program Submitting results • Once laboratory analyses have been completed, submit results using the online submission form or, e-mail project personnel with your results SLQA Project web site Submit Data (http://bqs.usgs.gov/slqa/frontpage_home.htm) • A data summary will be mailed to you

22. Double Blind Sample Program Sample preparation Items required for sample processing • Sample bottles and lids (identical to the ones used for submitting environmental samples to ensure the samplesare not identified as quality assurance samples) • Analytical balance accurate to the nearest 0.1 gram • Deionized water • Laboratory glassware: beakers, flasks • Squirt bottle

23. Double Blind Sample Program Sample preparation Sample Inspection Prior to preparing the double blind samples, inspect the sample vials provided by BQS: • Check each vial containing dry sediment for damage during transit • Check for sediment material on the outside of the vials and inside the plastic bags in which the samples were shipped • If sample integrity is questionable, return the sample to BQS and it will be remade

24. Double Blind Sample Program SamplePreparation 1)Record sample vial identifications (most double blind samples are comprised of 2 sample vials containing the sand and fine-size material portions) 2)Record tare weight of sample bottle without the lid, to the nearest 0.1 gram

25. Double Blind Sample Program SamplePreparation 3) Measure out required volume of deionized water to add to sediment to achieve the desired concentration [ Calculate water volume (L): Sediment mass (mg) Desired concentration (mg/L) ] 4) From this volume, pour approximately: • 50 mL into a squirt bottle • 50 mL into a clean sample bottle

26. Double Blind Sample Program SamplePreparation 5) Empty contents of both sediment vials into the sample bottle, which contains the 50 mL of water (the water in the sample bottle helps to keep the fine material from escaping as dust) 6) Rinse both vials and caps to remove additional material into the sample bottle

27. Double Blind Sample Program SamplePreparation 7) Add remaining water in squirt bottle and beaker to sample bottle (include the excess water from squirt bottle siphon) 8) Record total volume of water added to sample

28. Double Blind Sample Program SamplePreparation 9) Measure and record the total sample weight, without the lid, to the nearest 0.1 gram (bottle + water + sediment) 10) Seal cap on bottle and label sample with fictitious, but realistic, information Sample preparation is complete

29. Please visit the SLQA Project website: http://bqs.usgs.gov/slqa/ or contact: Mark Woodworth– (303)236-1875, woodwort@usgs.gov