Metrology Infrastructure in Gas Analysis

1. Metrology Infrastructurein Gas Analysis Ed W.B. de Leer NMi Van Swinden Laboratorium MiC Symposium, Beijing 18-22 October 2004

2. Importance of Gas Analysis • Economic Value • Natural Gas (calorific value) • Chemical process industry (ethylene,etc.) • Electronic industry (reactive gases, H2O) • Food industry and health care (purity, composition) • Environmental regulations • Industrial stack emissions (NOx, SO2, HCl, etc.) • Automotive gases (CO, CO2, C3H8, NO) • Occupational hygiene (VOC, particles) • Ambient air quality (Benzene, O3, CH4, N2O, VOC, etc.) • Legislation • Breath Analysis (ethanol) • Odor Control (n-butanol)

3. Infrastructure for Gas Analysis ISO TC 158 Traceability to SI Instrument Manufacturers Reference Gas Producers Laboratory Accreditation

4. Start of Traceability Chain (I) For static gas mixtures, the traceability chain starts with a primary realisation of the definition of the mole fraction or amount of substance fraction for a specified chemical entity. nB = amount of substance of entity B

5. Primary Realisation of mol/mol • A primary realisation (PSM) of the amount fraction in a static gas mixture involves (ISO 6142:2001): • accurate weighing (incl. buoyancy correction) of pure parent gases (or liquids) in a clean high pressure cylinder (uc ≈ 0.001-0.1%) • Impurity analysis of the parent gases/liquids • Analytical verification of the composition against a coherent set of primary standards (uc ≈ 0.05-1%) • Stability testing Problems: low amount fractions (nmol/mol) for reactive gases, sorption on cylinder wall, isotopic variation, phase separation

6. Stability of PSMs av= +0.004 s.d. = 0.105

7. Start of Traceability Chain (II) For dynamically generated gas mixtures, the traceability chain starts with a primary realisation for amount concentration (mol/m3) for a specified chemical entity. N.B. Pressure and temperature need to be defined as well! Usually, permeation, diffusion or injection of a pure gas in a flowing gas stream produces the primary realisation.

8. Key Comparison of National Standards

9. Experimental Set-up KCs (I) PilotNMI Participating NMI Pure gases Pure gases* gravimetry gravimetry comparison PSM* PRM secondary methods (NDIR, GC, FTIR, ….)

10. CCQM-K3 – Automotive gases

11. CCQM-K4 Ethanol in air

12. CCQM GAWG Key Comparisons

13. CCQM GAWG Key comparisons (II)

14. New Set-up CCQM-P23 PilotNMI Participating NMIs Pure gases Analytical comparison under repeatabilty conditions Gravimetry Series of PRMs

15. “High Accuracy” comparisons Author: Martin Milton All laboratories send a standard to the pilot laboratory • Preparation errors can now be “randomised”

16. Gravimetry: 1000 mmol/mol CO

17. CCQM-P23 • CCQM-P23 used mixtures of CO/N2 at three amount fractions 50,000 1,000 and 10 mmol/mol. • In the first phase these were all analysed by NDIR at NMi (The Netherlands) • Evidence was found for 13C/12C variations in pure CO and for an influence of the 13C/12C ratio on the NDIR signal. Extreme outliers from KRISS (Korea) and NIST (USA) • In the second phase two sets of mixtures were re-analysed: • 50,000 mmol/mol by NIST (USA) using GC/TCD • 1,000 mmol/mol by Metas (Switzerland) using GC/FID

18. GC-TCD results GC-TCD Results for 50,000 umol/mol CO in nitrogen 1,03 1,02 1,01 1,00 0,99 Ratio to Control 0,98 0,97 0,96 0,95 0,94 0,93 47000 47500 48000 48500 49000 49500 50000 50500 51000 51500 52000 52500 Concentration (umol/mol)

19. GLS for GC-TCD versus Gravimetry Author: Martin Milton • Conclusions for 50,000 mmol/mol • Std dev of residuals is 1mmol/mol (0.002% rel)- after excluding two outliers • At 50,000 mmol/mol, the gravimetric values are very consistent within their stated uncertainties. k=2

20. GLS for 1,000 mmol/mol • Conclusions • Std dev of residuals is 1.7 mmol/mol (0.17%relative) • At 1,000 mmol/mol , five (out of nine) of the gravimetric values are not consistent within their stated uncertainties. Author: Martin Milton k=2

21. Greenhouse gases: CH4 Author: Adriaan van der Veen

22. GLS for CH4 Greenhouse gas Conclusion: The std dev of residuals in x-direction is 0.012 μmol/mol (0.6% rel)The uncertainty of two gravimetric standards is not compliant with the regression

23. CCQM GAWG Pilot Studies

24. Gas CMCs in KCDB August 9, 2004 18 Countries 554 CMC entries 146 Measurands 12Gas Matrices

25. Function of NMI gas Standards Primary realization of gas composition standards Production of traceable gas transfer standards Calibration of secondary gas standards, instruments, etc NMI Production of pure gases Cylinder treatment Production of traceable calibration gases for field applications Industry Workfloor applications in gas analysis e.g. testing automotive exhaust gases, natural gas composition SI traceable measurement results Workfloor

26. Accreditation of Gas RM Producers • Accredited as Calibration Laboratory (ISO/IEC 17025) • Belgium (1) France (2, incl. LNE) • Germany (2) Italy (2) • Netherlands (2, incl. NMi) Spain (6) • Sweden (1) Switzerland (3) • United Kingdom (7, incl. NPL) • Accredited as Test Laboratory (ISO/IEC 17025) • Mexico (1) • Accredited as Reference Material Producer (ISO Guide 34) • Australia (2) Japan (2, incl. NMIJ) • Accredited by Jcss Japan (3) • Legal Metrology Systems e.g. China, Russia

27. Standardisation • ISO/TC 146 Air Quality • ISO/TC 158 Gas Analysis • ISO/TC 193 Natural Gas • ISO/TAG 4 Metrological Issues • OIML Recommendations

28. Conclusion • The metrological concepts of traceability and measurement uncertainty are widely accepted in the field of gas analysis • Cooperation between NMI’s, Standardization bodies, and Industry is essential for continued development