Intercomparison of Mobility Particle Size Spectrometers

Transcription

1 Intercomparison of Mobility Particle Size Spectrometers Project No.: MPSS Basic information: Principal Investigator: Home Institution: Participant: Instrument No.1: Eija Asmi Finnish Meteorological Institute Erik Palmenin aukio Helsinki FINLAND David Brus FI- FMI DMPSS TSI CPC Model 30, SN: 10 Location of the quality assurance: Leipzig, lab 11 Comparison period: May 30, 201 June 03, 201 Last Intercomparison (with Project No.): Summary of Intercomparison: Pre- Status: The FI- FMI Homemade DMPS was not in a good condition. The system is out of the +/- % range of the MPSS No.1 from. Final Status: The FI- FMI Homemade DMPS did not pass the quality standards of ACTRIS and GAW. 1

2 List of Components: Specification MPSS No.1 FI- FMI DMPSS Position (Line) Company FMI Software.1 Home made CPC TSI CPC, Model 3772 Model 30 Flow ratio 1.0 : :.0 Source Kr5 Ni 3 HV cassette positive positive DMA Hauke medium Vienna type medium Flow meas. ü ü Dryer aerosol ü RH sensor ü ü T sensor ü ü RH sensor T sensor Dryer p sensor Sheath air ü ü ü ü : System arrived in Leipzig. Setup in lab 11 including checks and calibrations : Pre- status overnight run from , pm to am : CPC Workshop in lab 130. More details in the ACTRIS CPC Workshop report. It was necessary to clean the CPC. Back to lab 11 for setup, checks, and calibrations. Information regarding the PSL measurements: The pre- status on Monday showed a large peak at 21 nm compared to the DMA. We cleaned the DMA and changed the electrode inside due of scratches. We still had the same problem of the large peak at 21 nm. After changing the flow ratio in the software to 1: and calibrating the sheath air flow against PSL 203, we got a peak at nm. But we still had the large size distribution compared to the DMA. We changed the different parts (head, cylinder, and bottom parts) of the DMA one at a time, still there were no changes in the size distribution. But when we used a DMA, we achieved good results. David will try other DMAs in Finland. Concentration Problem: In the pre- status run, the concentration was too high compared to the Instrument. We did several tests to check how the software is working. David wants to check the software when he is back in Finland. The raw- concentration is in the range of +/- %. It looks like there is a problem with the inversion process. The DMPS concentrations for smaller particle (<30nm) is too high for both raw and inverted data : Overnight run from , pm to am : Checking and cleaning of the DMA : Overnight run from , pm to am : Final- status overnight run from , pm to am. 2

3 : Data check and backup. Finalize all for shipping. 3

4 Laboratory setup: total flow 20l/min atmosphere PSL generator Latex 203nm) zero line lab 11 LINE 1 2x Nafion dryers rel. humidity: < 30% LINE FI-FMI DMPSS SI-NLZOH MPSS CZ-CAS TSI MPSS GR-IPTOS TSI MPSS modified by FI-FMI Total CPC 30 MPSS No MPSS No.1 TSI CPC 30 #2353 MPSS No. TSI CPC 30 #2337 Additional Equipment: vacuum flow Bubble flow meter Gilibrator, Gilian (Sensidyne) Thermo Scientific Nanosphere Size Standard PSL 203nm (±nm) Aerosol nebulizer for PSL (homemade ) Voltcraft multimeter (0-00V), Keysight Technologies

5 Total CPC Status Instrument Variable Status Power good Laser good Total CPC 30, Flow good #2337 Liquid level full Institute Variable Status Power good Laser good Total CPC 30, Flow good #2353 Liquid level full Instrument Status Instrument Variable Status Saturator Temp 39.0 C Instrument No.1, TSI CPC 3772 # Condenser Temp 22.0 C Optics Temp 0.0 C Cabinet Temp 32. C Ambient Pressure 9.7 kpa Orifice Pressure 9. kpa Nozzle Pressure 2. kpa Laser Current 50 ma Instrument Variable Status Saturator Temp 3.9 C Instrument No.5, TSI CPC 3772 #710 Condenser Temp 22.0 C Optics Temp 0.0 C Cabinet Temp 3.3 C Ambient Pressure 99.0 kpa Orifice Pressure 73. kpa Nozzle Pressure 2.5 kpa Laser Current 52 ma Instrument Variable Status Instrument No., TSI CPC 3772 #7109 Saturator Temp Condenser Temp Optics Temp Cabinet Temp Ambient Pressure Orifice Pressure Nozzle Pressure Laser Current 39.0 C 22.0 C 39.9 C 33.3 C 99.2 kpa 7.7 kpa 2. kpa ma 5

6 Time Series Systems during the pre- status May st Figure 01: Time series (May 30, 201 0:00 pm May 31, 201 0:00 am) of the integrated particle number concentration (N - 00nm ) of the MPSS and total number concentration (N total ) of the TSI CPC 30. The inversion was performed using software. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. Components and zero check Pre- Status of the Candidate (May th ) Instrument Line Flow Zero MPSS No l/min 0 # cm - 3 Total CPC 30, # l/min 0 # cm - 3 FI- FMI DMPS l/min 0 # cm - 3 FI- FMI total CPC l/min 0 # cm - 3 High voltage calibration Instrument [V] 0 V mv 0 mv 00 mv MPSS No.1 final [V] 0 V 5 mv 200 mv 00 mv FI- FMI final - Check by David Latex 203nm ±nm (pressure 9 hpa, 23.0 C) Instrument Latex 203 [nm] slope MPSS No FI- FMI

7 50 5 FI-FMI DMPSS, Latex 203 nm +/- nm Peak at Dp = 21.0 nm 'FI-FMI DMPSS' 'fit_fi-fmi DMPSS' Tropos MPSS No.1, Latex 203 nm +/- nm Peak at Dp = nm ' MPSS No.1' 'fit_ MPSS No.1' 0 1 raw concentration in #/cm³ raw concentration in #/cm³ particle diameter Dp in nm particle diameter Dp in nm Figure 02: Measurement of latex 203 nm: Particle size distribution (raw concentration) for latex 203 nm on May 30 th, 201. Time Series Figure 03: Time series (May 30, 201 0:00 pm May 31, 201 0:00 am) of the integrated particle number concentration (N - 00nm ) of the MPSS and total number concentration (N total ) of the reference TSI- CPC Model 30. The inversion was performed using software. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 7

8 Particle Number Size Distribution Figure 0: Comparison of mean particle number size distribution of MPSS No.1 against FI- FMI from May 30, 201 0:00 pm until May 31, :00 am. Multiple charge correction, internal diffusion losses and CPC efficiency are included for both of the MPSS. Correlation 1x 3 MPSS No.1 N -00nm [#/cm³] 12 2 Curve Fit Results ; pm ; 5am Coefficient values ± one standard deviation a =0 ± 0 b = ± R² = TSI CPC 30 SN 2353 N total [#/cm³] 12 1x 3 Figure 05: Linear regression between the number concentrations of the TSI CPC Model 30 SN: 2353 and MPSS No.1. Multiple charge correction, internal diffusion losses and CPC flow corrections are included.

9 1x 3 1 FI-FMI MPSS N -00nm [#/cm³] 12 Curve Fit Results ; pm ; am Coefficient values ± one standard deviation a =0 ± 0 b =1.123 ± R² =0.9 DE- RR total TSI CPC 30 SN 2353 N total [#/cm³] x 3 Figure 0: Linear regression between the number concentrations of the TSI CPC Model 30 SN: 2353 and FI- FMI. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 1x 3 1 FI-FMI DMPSS N -00nm [#/cm³] 12 Curve Fit Results ; pm ; 5am Coefficient values ± one standard deviation a =0 ± 0 b =1.203 ± 0.0 R² = DE- RR MPSS No.1 N -00nm [#/cm³] 1 1x 3 Figure 07: Linear regression between the number concentrations of the MPSS No.1 and FI- FMI. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 9

10 Final Status of the Candidate June st Components and zero check Instrument Line Flow Zero MPSS No l/min 0 # cm - 3 Total CPC 30, # l/min 0 # cm - 3 FI- FMI DMPS l/min 0 # cm - 3 FI- FMI total CPC l/min 0 # cm - 3 High voltage calibration Instrument [V] 0 V mv 0 mv 00 mv MPSS No.1 final [V] 0 V 5 mv 200 mv 00 mv FI- FMI final Check by David Latex 203nm ±nm (pressure 9 hpa, 23.0 C) Instrument Latex 203 [nm] slope MPSS No FI- FMI FI-FMI DMPSS, Latex 203 nm +/- nm Peak at Dp = nm 'FI-FMI DMPSS' 'fit_fi-fmi DMPSS' Tropos MPSS No.1, Latex 203 nm +/- nm Peak at Dp = nm ' MPSS No.1' 'fit_ MPSS No.1' 0 1 raw concentration in #/cm³ raw concentration in #/cm³ particle diameter Dp in nm particle diameter Dp in nm Figure 0: Measurement of latex 203 nm: Particle size distribution (raw concentration) for latex 203 nm on June 02 th, 201. FI- FMI showed the correct size at 203 nm, but there is still the same problem with the duple peak.

11 Time Series Figure 09: Time series (June 02, 201 0:00 pm June 03, 201 0:00 am) of the integrated particle number concentration (N - 00nm ) of the MPSS and total number concentration (N total ) of the reference TSI- CPC Model 30. The inversion was performed using software. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. Particle Number Size Distribution Figure : Comparison of mean particle number size distribution of MPSS No.1 against FI- FMI from June 02, 201 0:00 pm until June 03, :00 am. The system was working with the original DMA from FI. The sheath air was calibrated to PSL 203 nm. David used the original FI- inversion. 11

12 Figure 11: Comparison of mean particle number size distribution of MPSS No.1 against FI- FMI from June 02, 201 0:00 pm until June 03, :00 am. The system was working with the original DMA from FI. The sheath air was calibrated to PSL 203 nm. We used the inversion from FI with the flow ratio 1:. Figure 12: Comparison of mean particle number size distribution of MPSS No.1 against FI- FMI from June 02, 201 0:00 pm until June 03, :00 am. The system was working with the original DMA from FI. The sheath air was calibrated to PSL 203 nm. Inversion with software. 12

13 Figure 13: Comparison of mean particle number size distribution (raw concentration) of MPSS No.1 against FI- FMI from June 02, 201 0:00 pm until June 03, :00 am. The system was working with the original DMA from FI. The sheath air was calibrated to PSL 203 nm. Correlation 000 MPSS No.1 N -00nm [#/cm³] Curve Fit Results ; pm ; 5am Coefficient values ± one standard deviation a =0 ± 0 b = ± R² = TSI CPC 30 SN 2337 N total [#/cm³] Figure 1: Linear regression between the number concentrations of the TSI CPC Model 30 SN: 2337 and MPSS No.1. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 13

14 12x 3 FI-FMI MPSS N -00nm [#/cm³] Curve Fit Results ; pm ; 5am Coefficient values ± one standard deviation a =0 ± 0 b =1.21 ± R² =0.0 DE- RR total TSI CPC 30 SN 2337 N total [#/cm³] 12x 3 Figure 12: Linear regression between the number concentrations of the TSI CPC Model 30 SN: 2337 and FI- FMI. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 12x 3 FI-FMI DMPSS N -00nm [#/cm³] Curve Fit Results ; pm ; 5am Coefficient values ± one standard deviation a =0 ± 0 b =1.207 ± R² =0.0 DE- RR MPSS No.1 N -00nm [#/cm³] 12x 3 Figure 13: Linear regression between the number concentrations of the MPSS No.1 and FI- FMI. Multiple charge correction, internal diffusion losses and CPC flow corrections are included. 1