SDSB  79-22
                           Technical Report


                              July 1979
                      Particulate Measurement  -
              Efficiency of Pallflex T60A20 Filter Media
                                  by
                            Kevin D.  Stulp
                                NOTICE

Technical Reports do not necessarily  represent final EPA decisions
or positions.   They  are  intended to present technical analysis of
issues using  data which  are currently available.   The purpose in
the  release  of such reports is  to  facilitate the  exchange of
technical information and to inform the public of technical devel-
opments which may form the  basis for a final EPA  decision, position
or regulatory action.
               Standards Development and Support  Branch
                 Emission Control Technology Division
             Office of Mobile Source Air Pollution  Control
                  Office of Air,  Noise and Radiation
                 U.S. Environmental Protection Agency

-------
Abstract

     An  experiment  designed  to measure  the efficiencies of  dif-
ferent  lots  of  Pallflex T60A20 teflon-coated  glass  fiber  filters
has been  completed.   Results of these tests  indicated  that:   (1)
batch-to-batch variation of  filters is  generally less than  2
percent, and (2) this media will not satisfy the 98 percent filter
efficiency requirement on all vehicles.

Introduction

     The test procedure specified in the light-duty diesel particu-
late NPRM  (Federal  Register. Vol.  44, No.  23, February  1,  1979)
requires:    (1) a  filter  efficiency of  98 percent,  or greater
efficiency  and  (2)   that  fluorocarbon-coated  glass  fiber  filters
must be used for testing.   Currently, Pallflex Corporation  is  the
only manufacturer  of the required  filter media,  with the best
candidate being type T60A20.   Some manufacturers have  expressed
concern that the 98 percent  filter  efficiency requirement will be
difficult to  meet,  and  that there  is significant  batch-to-batch
variability with this media.  Hence,  a brief test program  was
conducted to  evaluate  the  significance  of these  concerns.    Two
different production vehicles were  used to evaluate batch-to-batch
variation of  the  above filters,  and  also  to  investigate  diffi-
culties  in  meeting  the 98 percent  filter  efficiency requirement.

Objective

     This study was  conducted  in order  to evaluate batch-to-batch
variation of Pallflex T60A20 filter media and  to  evaluate  diffi-
culties in satisfying the 98  percent filter efficiency requirement.

Test Procedure

     In  this  investigation,  the  procedure specified  for  filter
efficiency  evaluation in  the light-duty  diesel  NPRM (Federal
Register, Vol.  44,  No. 23, February  1,  1979) was followed.   There-
fore,  only highlights will be noted.

     This procedure  requires that  efficiency be  determined by
collecting  particulate  on a back-up filter while the vehicle is
operated over a cold start LA-4 cycle.  Filter efficiency  corre-
sponding to each bag  must  be determined.  The back-up  filter  must
be three  to four  inches behind the sample  filter as shown  below:

-------
                                     -2-
                                      3-4"
Flow
                     7
       7
              Sample Filter
Back-Up Filter
          For  this study,  two  diesel production vehicles were  used to
     evaluate vehicle-to-vehicle effects on filter efficiency as well as
     the batch-to-batch variabilty.

     Results

          The  results of  this study  (summarized in Table  1)  indicate
     that filter efficiency is vehicle  dependent,  with generally small
     batch-to-batch variation.

          Specifically,  none of  the batches  of Pallflex T60A20 media
     satisfied  the filter  efficiency requirement when  the Peugeot 504D
     was tested.  The means of filter efficiency ranged  from  a low of
     90.7 percent to  a  high of 93.7  percent with an  overall  mean and
     standard  deviation  of 92.1  and  0.9  percent respectively  for all
     batches combined.

          For  the 1978 Oldsmobile  Delta  88 350D, the Pallflex media met
     the 98 percent filter  efficiency standard  with  each  bag and batch
     combination,  the low mean being 98.0 percent,  the high mean being
     99.2 percent, and the  overall mean of 98.8  percent  (with a standard
     deviation  of  0.4 percent).

          These results  indicate  relatively  consistent batch-to-batch
     efficiency with Pallflex filters on  multiple tests on  a given
     vehicle.

-------
                                                  -3-






                                                Table 1




                                    T60A20 Pallflex Filter Test Media
Peugeot 504D


Batches
3991E2
1092B
4009E2
Filter
^*
Bag 1
91.99
91.42
90.71
Efficiency
%
Bag 2
91.70
93.74
93.49
Standard
~ %
Bag 1
1.26
3.22
0.73
Oldsmobile


Batches
3991E2
1092B
4009E2
Note: Filter
Filter
***
Bag 1
98.01
99.06
98.96
Efficiency
%
Baj> 2
98.44
99.13
99.16
efficiencies calculated according
Standard
~ %
Bag 1
0.12
0.15
0.12
Deviation

Bag 2
2.55
2.32
0.45
350D
Deviation

Bag 2
0.49
0.09
0.49
to Section 86. 110-81 (c) (iv)

Number
Bag 1
5
2
2


Number
Bag 1
2
2
2
Federal

of Tests
Bag 2
5
2
2


of Tests
Bag 2
2
2
2
Register
Vol. 44, No. 23, February 1," 1979.

-------
                                 -4-
Discussion and Conclusion

     The results indicate that while batch-to-batch variability is
not expected  to  be a problem, some manufacturers  may  have diffi-
culty in achieving a  98  percent  filter efficiency  using the Pall-
flex T60A20 media.  This means that it is possible that the filter
requirements,  as  specified  in  the NPRM,  may be  inequitable  for
those  manufacturers that  can  achieve  98 percent  efficiency.
Therefore,   some  modifications to  the test  procedure  may  be  re-
quired.

     Possible options are:

     1.    Allow each manufacturer  to select  a filter of their
choice and  require a demonstration of minimum  performance.  How-
ever,  there is a  major  difficulty with  this modification.   EPA
would have  to either use several different  filter media  types in
order to maintain lab-to-lab correlation,  or  choose one filter type
and let  lab-to-lab  correlation  suffer.  Neither are satisfactory.

     2.   Require a test-by-test  efficiency  correction based on one
or more back-up filters.   There  may  be inherent difficulties with
this modification,  also, as  some  researchers feel efficiency is
load  dependent,  and hence the back-up  filter may  be far  less
efficient than the sample filter.  There  is then some uncertainty
associated  with  the  efficiency  of the back-up  filter  which would
have to be  reconciled since its  measurement cannot be ascertained
precisely in day-to-day certification  testing.

     3.   Lower the minimum acceptable  filter efficiency.  However,
even with  a reduction to as  low as  95 percent,  the Peugeot would
still  have trouble  when Pallflex T60A20  filter  media is used.
While helpful,  further  modification  would still have  to  be made.

     4.   Require  a  back-up filter  during the certification test.
The combined  weight  of  both filters  would  be used in determining
particulate emissions  for those  tests  in  which  the calculated
efficiency  is  below  the  specified minimum.   This modification has
the advantage  of being relatively easy to use in day-to-day test-
ing.

Recommendation

     On the basis  of  the above  discussion,  it is recommended that
the test procedure be modified  in  accordance with  options  3 and 4
above.  That  is,  the following  changes to  the  test procedure are
recommended:

     1.     Require a  back-up filter  during actual  certification
testing.

-------
                                 -5-
     2.   Calculate the ratio of filter weights as indicated in the
light-duty diesel NPRM.

     3.   If  the  ratio is  greater than 0.95,  calculate particulate
emissions on the basis of the first filter net weight.

     4.    If  the ratio is  less than 0.95,  calculate particulate
emissions on the basis  of  combined  net weights of the first filter
and the back-up filter.

-------