May 24, 1991


MEMORANDUM

SUBJECT:   Review of Chemical Mass Balance (CMB) Analysis for the
           Communities of Polson and Ronan, Montana

FROM:      Tom Coulter, Environmental Scientist
           Source Receptor Analysis Branch (MD-14)

TO:        Mark Comp, Technical Representative
           Air Planning Section, Region VIII (8AT-AP)


     This is in response to the materials you sent to Edwin Meyer
(SRAB) asking for assistance on questions associated with the CMB
analysis for the two Montana communities.  This memorandum
summarizes the key points of our teleconference on May 2, 1991,
which included Bob Stevens, Teri Conner and Chuck Lewis of the
Source Apportionment Research Branch in AREAL.  The following
problems with the samples collected were identified:

           1.   There was considerable internal inconsistency
                between potassium (K) and the organic carbon (OC)
                components of the measured mass, assuming wood
                combustion to be a dominant source category  
                (which is likely in the winter samples).

           2.   Flagging K as a fitting species in the CMB
                analysis was theoretically a good idea but its
                assay on glass or quartz fiber filters was
                problematic for at least two reasons: 1) x-ray
                fluorescence analysis will only work well for 
                high atomic mass elements on such filters, and  
                2) high background concentrations and inter- and 
                intra-filter differences for K on filter blanks  
                were reported by the contractor Engineering      
                Science (Section 3: CMB METHODOLOGY AND RESULTS).

           3.   The roughly half of the samples which were
                collected as TSP were not particularly useful in
                apportioning woodsmoke as a potential contributor
                because there is no upper cut point in particle
                size.  The majority of the coarse fraction K is
                due to soil (road dust) and this would have had  
                to be carefully "corrected" to infer woodsmoke K.

           4.   The high background levels of key indicator
                materials like aluminum (Al), silicon (Si) and
                calcium (Ca) (important for implicating road dust
                contribution) in filter blanks precluded the
                proper resolution of the road dust component.

In general, there was much difficulty in applying CMB analysis to
such a small group of elements which were not sampled with this
analysis in mind, and thus were themselves poorly analyzed.

     We offered two specific recommendations:

           1.   Extract soluble K from the remainder of the      
                quartz filters (if still available) and reanalyze
                using ion chromatography or atomic absorbtion
                spectrometry.  The results may be applied to a
                literature equation to give a good answer for the
                OC component associated with woodsmoke.

           2.   Send a portion (e.g., 5" X 5") of the remaining
                quartz filters to R. J. Lee Co. so they can
                perform scanning electron microscopy analysis on
                individual particles to help better resolve the
                road dust component.

You said that you were going to check on the status of the filter
custody, and make a decision as to how to proceed at this point. 

      I hope that our comments are useful; please feel free to
call me at FTS 629-0832 or Bob Stevens at 629-3156 if you have
questions or need more assistance with this problem.

cc:   J. Dicke
      E. Meyer
      R. Stevens (MD-47)
      J. Tikvart
      D. Wilson