Met One Instruments. Met One Instruments BAM-1020 Improvements

Transcription

1 Met One Instruments BAM-1020 Improvements Herbert Schloesser,, Ph.D. David Gobeli, Ph.D. Met One Instruments 3/31/05 1

2 List of Improvements Temperature sensor for compensating Temperature gradient during 1h Measurement New Filter tape 8 min Sample without loosing sample time Geometry Improvements Electronic Improvements How to upgrade conclusion 3/31/05 2

3 Additional Temperature Sensor What happens if air temperature or barometric pressure changes?? Air mass between detector and source changes. Small effect because only the delta T and delta P during 1 hour counts. But Measurement relies on stable temperature during 1 h Solution: measure delta T with additional sensor (delta P is already measured) Only needed if setup suffers from higher temp gradients 3/31/05 3

4 Additional Temperature Sensor h Differential Temperature Drift - Bam1020 Mass corrected Mass uncorrected 1h Delta T 5 mg/m h Temp diff time in hours Effect is about 2 ug/m 3 per degree C Typical temperature drifts seen in the order of 2 degrees C -> > Effect within the detection limit but systematic -5 3/31/05 4

5 New Filter tape The standard glass fiber filter tape is about 5 mg/cm 2 New fiber backed Teflon membrane filter tape is about 1 mg/cm 2 Beta absorption measurements with the BAM 1020 can be improved by use of the thinner filter material Tests show considerable improvement 3/31/05 5

6 New Filter tape Problems to solve Does calibration change? (no) Does thin tape work with current transport mechanism? (yes) Does tape seal reliably well with current transport mechanism? (no) Does Tape work with high concentration levels? (no) Max load is about 400 ug -> > 1h ambient ok 3/31/05 6

7 New Filter tape Dirk Felton Data Field testing Tests show considerable improvement 3/31/05 7

8 New Filter tape Issue with Air Flow Rate Tape has a potential to produce leakage using current design That is why Met One worked to improve nozzle geometry More down force needed to seal reliably Solution requires mechanical change of nozzle. Best done as factory upgrade 3/31/05 8

9 Geometry Improvements Reduce gap between detector and source less delta T effect More efficient use of beta counts Better detection limit Easy to do But changes calibration slightly!! 3/31/05 9

10 Electronic Improvements Flow control motor upgrade available Problems with newer motors (some newer control motors have had changed Spec s with higher needed current) Current was limited for older motor thus unit would not move motor with higher resistance. New Upgrade fixes this Problem Photomultiplier module shows drifts in dark counts with bigger temperature changes Upgrade fixes this behavior 3/31/05 10

11 Increasing Count time without reducing sampling time. Idea: improve count statistics by counting 8 minutes instead of 4 per measurement Increase of Count rate by factor of 2 delivers a gain in detection limit by (2) = 1.41 But counting 2 times 8 min would reduce Sampling time to less than 45 min Idea: using sampling time to measure next Zero filter value 3/31/05 11

12 Increasing Count time without reducing sampling time. Synchronised Count Rate I0 Prev I1x /5/ :24 1/5/ :12 1/6/ :00 1/6/ :48 1/6/ :36 Bam 1020 measures I1X stability count 1/6/ :24 Time 1/6/ :12 1/7/ :00 1/7/ :48 1/7/ :36 1/7/ :24 4 h time delay gives good match in count rates between I0 and I1X. Proves accurate positioning of the Bam 1020 and gives possibility y for further improvement 3/31/05 12

13 Increasing Count time without reducing sampling time. Further investigations on statistics are on the way Method would allow to use previous Count as zero measurement and use longer time to measure loaded filter. Need stability over 4 hours additional Temperature sensor needed to account for T and P changes Little issues Additional filter spot at midnight must go The first 4 hours after power up have to be conventional Software change needed 3/31/05 13

14 How to Upgrade?? Field upgrade Factory Upgrades 3/31/05 14

15 Field upgrades New Temperature sensor can be done if hole in front panel is there. Flow Controller upgrade for Bam1020 with serial# < D4488 can be done if BX961 is installed Geometry upgrades can be done in field with appropriate kit (overall Calibration adjustment must be made to the k factor of unit) - Met One Field Service 8 min Software upgrade upon availability 3/31/05 15

16 Factory Upgrades Mechanics for new Filter tape. Factory Upgrades will be available in the later part of this year Photomultiplier module All other mentioned improvements can also be done together in a one shot Performance upgrade. Turn around in 10 business days 3/31/05 16

17 Why Upgrade?? New Filter tape is big step in improvement Together with the mentioned performance upgrade Package the detection limit of the Bam 1020 will be improved by more than a factor of two Stability of Bam 1020 will be further enhanced. 3/31/05 17

18 CARB BAM vs. FRM Best Met One BAM 1020 vs. PM2.5 FRM 6/ / Slope: 1.02 Intercept: 3.01 r 2 : Met One BAM 1020 (ug /31/05 FRM (ug/m3) 18

19 CARB BAM vs. FRM Worst Met One BAM 1020 vs. PM2.5 FRM 6/ / Slope: 1.17 Intercept: 2.41 r2: Met One BAM FRM (ug/m3) 3/31/05 19

20 Met One BAM 1020 vs. PM2.5 FRM 6/ / Slope: 1.03 Intercept: 1.64 r 2 : Met One BAM 1020 (ug/ FRM (ug/m3) 3/31/05 20

21 Met One BAM 1020 vs. PM2.5 FRM 6/ / Slope: 1.07 Intercept: 1.58 r 2 : Met One BAM 1020 (ug/ FRM (ug/m3) 3/31/05 21