The result of first SF6 inter-comparison experiment (SICE) from 2016 to 2017

Transcription

1 The result of first SF6 inter-comparison experiment (SICE) from 2016 to 2017 Haeyoung LEE*, Sumin KIM, Hong-woo CHEO, Sang-Boom RYOO World Calibration Centre for SF6 National Institute of Meteorological Sciences Korea Meteorological Administration

2 World Calibration Centre for SF 6 has been run by Korea Meteorological Administration since 2012 Oct. under the MoU between WMO and KMA. We also have been in a cooperation with Korea Research Institute of Standards and Science (KRISS) for WCC-SF 6 under the MoU between two institutes with the agenda such as the analysis methods and calibration methods. Climate Change Monitoring Division

3 Maintenance of Scale Primary Standard BIPM/NMI GAW Report No.172 Central Calibration Lab Scientific Advisory Group < Traceability and compatibility> Inter-comparison Round Robin: Inter-comparison World Calibration Center <Support GAW Stations> Inter-Comparison Technical support Workshops Training and Education Quality Assurance /Science Activity Center GAW Stations/Labs World Data Center Measurement based on Link to the WMO scale Last GGMT meeting in La Jolla, we were encouraged to organize round-robin comparisons of SF 6 working standards between WMO participants (GAW No.229) Climate Change Monitoring Division

4 Cylinder preparation for travelling standards

5 Site description Anmyeondo GAW station (36.54 N, E, 46 m a.s.l) Seoul is 100 km away from AMY Air mass from Korea main land affects on SF 6 concentration In 2016, its annual mean was 9.5 ppt.

6 Cylinder preparations : filling Stainless steel filter D = 4.7 cm, pore size = 50 µm 1. Conditioning for 12 hours to almost 0 psi 2. Filling during 2.5 hours for a 29.5 L cylinder (Luxfer, Aluminum) up to 1500 psi (maximum 1700 psi) 3. Rolling during 3 hours for the homogenization inside of the cylinders Pressure gage Decabon tube, 50 m from tower to the lab (Nitta Moore ) Vent RIX SA-3E Fi lt e Filter Hy-Loc, rfi4-h-8t

7 Cylinder preparations : filling

8 Cylinder preparations : Adjusting target level Relief valve SF 6 SF 6 Zero air (The air with SF 6 almost zero concentration, Korea company) SF 6 (200 ppt from KRISS) Normally, we lower SF 6 concentration below our ambient level since the level of SF 6 in a cylinder is higher than other GAW stations. The formula to adjust lower level of SF 6 is following : Zero air pressure to put in the cylinder = (Amb SF 6 mole fraction target mole fraction of SF 6 )/Amb SF 6 mole fraction * 1500 psi (Final cylinder pressure)

9 Cylinder preparations : Adjust target level

10 Inter-comparison experiment between CCL and WCC - SF6

11 First inter-comparison experiment in 2013 Schematic Diagram of analysis system First version: conventional method Vent Sample loop (6 ml) MFC 100 ml /min EPC1 μ-ecd Regulator STD Restek Porapak-Q 80/100, 1/8 x12ft P-5 (CH 4 /Ar) GC system Agilent 7890A/G3440A, Agilent,USA Analysis condition Oven temp: 50 (or 60 ), isothermal Column: Activated Alumina F-1 (80/100), 4 m length, 1/8 OD (Restek, USA) Carrier gas: 20 ml/min of P-5 (CH4 5% in Ar) with makeup 30mL/min GC injection method: 6- port gas sampling valve

12 First inter-comparison experiment in 2013 Measurement result by peak area (A B A method *Ratio value of SP (or STD) = 2*B / (A+A )

13 Second inter-comparison experiment in 2015 Schematic Diagram of analysis system Recent version: Pre-concentration + Back flushing Analysis condition Oven temp: 50 (or 60 ), isothermal Column: Activated Alumina F-1 1/8 OD (Restek, USA) - Pre-column : 6 ft length, Main-column : 12 ft length Carrier gas: 60 psi of P-5 (CH 4 5% in Ar) with makeup 30 ml/min GC injection method: Pre-concentration - Trap material : Carboxen Absorption/Desorption: -60 /200 - Helium flow: 100 ml/min

14 Second inter-comparison experiment in 2015 Measurement result by peak area with one-point calibration (fdrift= 2*STD / (STD +STD )) Cylinder STD 6.72 Low STD 6.72 STD 8.87 High STD 8.87 Peak area fdrift Cylinder # SF6 From WCC 1 st SF6 from CCL Differences (WCC-CCL) Low (D282761) 6.23(±0.004) 6.23(±0.006) 0.00 High (D282765) 9.72(±0.006) 9.80(±0.009) The samples were injected 5 times in a set. STD 9.76 High STD 9.76 Cylinder # SF6 From WCC 2 nd SF6 from CCL Differences (WCC-CCL) High (D282765) 9.78(±0.023) 9.80 (±0.009) -0.02

15 Meanwhile.., Out of flying pan into fire Large differences between KMA and CCL from the SF 6 analysis result

16 Non-official inter-comparison experiment in Korea with WMO scale, 2016 Linearity check and decided having two point calibration (Technical note CCMD ) Non-official inter-comparison experiment for Korea network and WCC-SF 6 with WMO scale N 2 O is still tricky but SF 6 (8.98 ppt) is within the compatibility goal.

17 Very recent inter-comparison experiment in 2017 Measurement result by peak area with two-point calibration ( fdrift (i)= (3*STD )/(3*STD +(STD -STD )i) Cylinder STD 6.72 SA STD7.83 STD 6.72 Peak area fdrift Corrected ± 0.01 Cylinder STD 8.6 SB STD9.76 STD 8.6 One set Every sample was injected 5 times in a set Peak area fdrift Corrected Cylinder # 1 Set 2 Set 3 Set 4 Set Mean SF6 From WCC SA (D376125) 7.36 ± SB (D376130) 9.28 ± ± ± ± ± ± ± ± ± ± (±0.015) 9.29 (±0.018) SF6 from CCL 7.37 (±0.01) 9.29 (±0.01) Differences (WCC-CCL)

18 The Result first SF6 inter-comparison experiment

19 The first SF 6 inter-comparison experiment From May 2016 to Feb countries with 12 labs attended the SICE-2016 Two circuit consisted of Europe region and Asia- Pacific region NOAA (CCL) attended each circuit Two cylinders, high and low concentrations, were circulated in a circuit. Two labs used X2006 scale, seven labs used X2014 scale, and three labs used their own scale.

20 Result of WCC-SF 6 For high concentration CB10895 and CB10909 For low concentration CB11159 and CB10894 A analysis result example in circuit CRM1' D CRM2 CRM1" * * * Response (Averaged) Standard deviation ± 0.01 ppt Relative S.D. (%) f-drift R-corrected (Corrected peak area ) certified mole fractio n (pmol/mol) Sample mole fractio n 7.88 (pmol/mol) drift (%) Sample uncertainty 0.01 <KMA technical note, CCMD2016-6>

21 Result of WCC-SF 6 Circuit 1 High Low Mean Uncertain RDS(%) Mean Uncertain RDS(%) Before After Mean 9.21 ± ± Final 9.21 ± ± 0.02 Circuit 2 High Low Mean Uncertain RDS(%) Mean Uncertain RDS(%) Before After Mean 9.20 ± ± Final 9.20 ± ± 0.02

22 Result of Europe : 1 st Circuit : Their own primary standard scale : WMO-X2006, No mark: WMO-X2014 * UEA also submitted the value, which is corrected by WMO X2014 after final result submission and the differences decreased to around ± 0.03 ppt

23 Result of Asia-Pacific : 2 nd Circuit : Their own primary standard scale : WMO-X2006, No mark: WMO-X2014 * KRISS submitted the value, which is calibrated against WMO X2014 and its own scale

24 With all participants (including CCL) For high concentrations, 5 Labs are within the compatibility goal (± 0.02 ppt) and 9 labs are extended compatibility goal (± 0.05 ppt) among 12 labs. For low concentrations, 5 Labs are within the compatibility goal and 8 labs are in the extended compatibility goal among 12 labs.

25 Consider only WMO scale Excluding METAS, Heidelberg University, KRISS and NOAA For high concentrations, 4 Labs are within the compatibility goal (± 0.02 ppt) and 7 labs are extended compatibility goal (± 0.05 ppt) among 8 labs. For low concentrations, 4 Labs are within the compatibility goal and 6 labs are in the extended compatibility goal among 8 labs. Two labs used X2006 scale.

26 Summary - We have had inter-comparison experiment with CCL (NOAA) since 2013 biennially to keep the traceability and compatibility goal - It is getting better through few failure as showing that the differences from NOAA are within the compatibility goal in 2016(non official) and The 1 st inter-comparison experiment was held from 2016 to 2017 with 12 labs. - 2 labs used WMO-X2006 scale and 6 labs reported WMO X2014 scale. - And 50% of them are within the WMO target (± 0.02 ppt). - Just 1 or 2 labs are behind the extended compatibility goal (± 0.05 ppt). - We have a plan to have this inter-comparison experiment every two years regularly. Climate Change Monitoring Division

27 Thank you all, Brad Hall, NOAA, USA Grant Forster, UEA, UK Camille Yver Kwok, LSCE, France Frank Meinhardt, Schauinsland GAW station, Germany Ingeborg Levin, UHEI, Germany Ludwig Ries, Zugspitze GAW station, Germany Myriam Guillevic, METAS, Switzerland Simon A. Wyss, Empa, Switzerland Daniel Rzesanke, ICOS, EU Armin Jordan, MPI-BCG, Germany Rowena Moss, NIWA, New Zealand Jeongsoon Lee, KRISS, Republic of Korea Jegyu Yu, KMA/NIMS, Republic of Korea Climate Change Monitoring Division