Filter Sampler

From Jimenez Group Wiki
Revision as of 19:33, 22 January 2014 by Bpalm (talk | contribs) (Filter Baking)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

Electronic versions of the instruction manuals for the filter sampler and the sampling head are here and here

MCV HiVol Filter Sampler

  • Basic instructions for routine operation of the sampler are here.
  • Here is an image of the print-out summary here.

Filter Baking

  • The PSI group usually bakes their filters in the lab, and then ships/brings them to the field site.
  • The filters we have used at past campaigns are 2500QAT-UP 150MM from PALL
  • Filter Baking Procedure (from Peter Zotter at PSI): A whole pack of filters (~30) is put in a quartz petri dish and put in the oven (8h at 800°C) and on the next morning after the filters have cooled down, they are wrapped in aluminum foil. So the whole procedure takes about 24h (8h prebaking and a long time cooling down) even though the only workstep which takes some time for me is wrapping the filters in foil (~30min up to 1h). Barend van Drooge suggests putting the filters in foil before baking, so the foil is cleaned as well and handling of filters is minimized. WARNING: at 800°C the foil will probably melt!!!
  • Baking can be done in the Durning Lab (a teaching lab?) in the Engineering dept. Contact Greg Potts for help with baking (google his name for contact info). He prefers you bake during the day instead of overnight, so the ovens can be monitored.
  • Best to bake filters in quartz dish such as this: http://fusedquartz.qsiquartz.com/viewitems/quartz-dishes-watch-glasses/round-quartz-dishes?
  • Pyrex ("softening point" = 820C) is probably not a good idea, use quartz (melting point ~1700C). The Jimenez lab has a quartz dish, wrapped in foil in the drawer with the MCV materials

Blanks

  • Field and lab blanks must be made.
  • Lab blanks: PSI recommends 2 lab blanks be taken just after pre-baking and 2 more lab blanks be taken after all the preparation steps (i.e., prebaking -> wrapping in foil and storing the filters in the freezer -> defrosting and unpacking if you stored the pre-baked filters in the freezer before using them -> putting the filter in the filter holder -> remove filter from filter holder again and then put it in the freezer). There is no need to create lab blanks after the campaign.
  • Field blanks, the more the better. Taking a blank every three days is recommended, or two times a week if one wants to conserve filters and time.
 Steps for field blanks
(1) load clean filter into cartridge, 
(2) put cartridge into filter sampler 
(3) immediately remove the cartridge from the sampler (do not activate sampling)
(4) re-wrap the filter and place in freezer.

PM1 Sampling Head

  • The VOLUMETRIC (i.e., "real") flow rate must be 30 m3/hr for the size cut to be accurate. NOTE: At high elevations (Boulder), the sampler can achieve this flow without difficulty (moving less mass for given volume of air), but at sea level the pump/motor can barely achieve this flow rate.


Other Info

  • There are 15 filter holders, but not all need to be in the filter sampler when collecting. Thus, one can remove a few holders so that they can be pre-loaded with filters. This practice should minimize downtime when changing filters, and hopefully promote filter cleanliness.
  • There is a large rotameter that can be used to verify the flow out of the sampler. The flow read off the rotameter must be corrected for temperature and pressure (see the bottom of this webpage for a good summary of correction equation: http://www.skcinc.com/prod/320-100.asp)


BEACHON

  • Sampling Times and Limits of Detection:

1. Chris Geron (U.S. E.P.A) measured OC levels are 0.5-2 ugC m-3 for 48-hr samples (July & August 2008)

2. Barend L. van Drooge (Molecular Tracers Analysis) needs approximately 100m3 sample. Doing the math, that means 13 1/3 hours sampling time, if we divide filters into quarters.

3. Andre Prevot (14C analysis) needs 12 hrs sampling time (Patrick Hayes has a spreadsheet from Peter Zotter with calculations). This is figured with 0.5 ug/m3 concentrations, and two (16.3 mm diameter) punches taken from a filter. The 14C analysis will require 1/2 a filter given the area of the punches (2 punches for TC, 2 for OC, and 2 for Sunsent ECOC, plus extra material as a precaution). (From Peter Zotter on April 2012: need at least a CARBON loading of about 5 ug/cm2, and the sampling time needs to be set accordingly. If I would have more filter material than 47mm the filter loading can be lower but not lower than 3ug/cm2).

CONCLUSION: Sampling time should be a minimum of 12 hours, or 24 hours if one wants to be very cautious. That would correspond to 64 filter samples taken for BEACHON (32 days/12hours).


  • START: 6pm on 19th, if power capacity allows.


  • Tentative Filter Sampler Schedule:

See spreadsheet here: https://spreadsheets.google.com/spreadsheet/ccc?key=0AlKzqypIvZJtdDFILUFtMG9hY0M1WVZYb1ZrV3NWWHc&hl=en_US&authkey=CNGlwpQK


Jul 20th - Jul 26th: 24 hr filter collection (7 d, 7 filters)

Jul 27th - Aug 3rd: 12 hr diurnal average, 24 hr total collection time per filter (8 d, 8 filters, 2 day oversampling)

Aug 4th - Aug 11th: 6 hr diurnal average, 12 hr total collection time per filter (8 d, 16 filters, 2 day oversampling)

Aug. 12th - Aug 17th: 12 hr filter collection (6 d, 12 filters)

Aug 18th - Aug 19th: 48 hr filter collection (2 d, 1 filter)

Therefore, 44 filter samples will be collected. Also, 4 lab blanks and 9 field blanks will be collected. In total, blanks and samples will be 57 filters.