EPA/AA/CTAB/88-09
       Haa*-        ReP°rt:  Development of Resistively
       Heated Diesel Particulate Trap/Regeneration Systems
                                by
                        Robert  I.  Bruetsch
                           August 1988
decisions or
analysis  of
?he purDOS6
exchfnS  of
?echnicll
decision
aecision,
                  NOTICE

          do   not  necessarily  represent  final  EPA
  positions.  They  are  intended to present technical
  issues  using  data  which are  currently available.
      0 release of  such reports  is to facilitate the
            information and  to inform the  public of
            which may form the basis  for  a "
        or regulatory action.

  U.  S. Environmental Protection Agency
       Office of Air  and Radiation
         Office of Mobile Sources
   Emission Control Technology Division
Control Technology and Applications Branch
            2565 Plymouth Road
        Ann Arbor,  Michigan  48105

-------
                        Table of Contents
                                                          Page
                                                         Number
I.     Technical Approach and Current Industry Efforts .  .   1
II.   Techical Progress 	   4
      A.   Coloroll (Low Density) 	   4
      B.   Coloroll (Bypass Regeneration) 	   4
      C.   Sohio	                 5
      D.   Duriron	7
      E.   Hi-Tech Ceramics 	   7
      F.   Aker Industries	9
      G.   NGK	n
      H.   Corning	12
III.  Future Work	             13
      A.   Selee	13
      B.   Hi-Tech Ceramics 	  14
      C.   Other Material Suppliers 	  16
IV.   References	             17
V.     Appendix	                 A-1

-------
I•    Technical Approach and Current Industry Efforts

     New vehicle and engine emission standards  for  future model
years  have  stimulated  the  development  of  advanced  control
systems for  Diesel  particulate emissions  from  light  and heavy
duty  Diesel  vehicles.[1]    The   standards  which  new  Diesel
vehicle and engine manufacturers are now  challenged to meet are
listed in Table 1.

     In  past  years,  development  has  dealt   largely with  an
aftertreatment  control  technology commonly  known  as  the "trap
oxidizer"  or the  "catalyzed  trap oxidizer"  (CTO).   The common
CTO  is  a  means  of  trapping or  filtering  the particulate matter
in  Diesel  vehicle  exhaust  and  eliminating  it   in  place  by
thermal oxidation, commonly referred  to as regeneration.  These
devices  first appeared four  years  ago on one passenger vehicle
model  in the  United States.    Although production  of  CTOs  is
presently   suspended,   there    are  approximately   30,000  trap
oxidizer-controlled Diesel vehicles in  customer use.[2]

     Heavy-duty  trap  system   development  programs   are  in
progress  throughout  the world  sponsored by almost  every Diesel
vehicle  and  engine  manufacturer.   Many  of these  programs are
targeted  to meet  1991  and 1994 truck  and bus standards, while
others  are aimed  at  retrofit  system  development  for  buses and
some  off-highway vehicles.   Some  heavy-duty manufacturers will
conform  to the  1991  heavy-duty truck  standards  without traps,
especially  if  the  sulfur  content in Diesel  fuel  is  reduced.
However,  trap  oxidizers  may be  used  in  limited  instances  in
 1991.   While almost  every manufacturer  has a  trap development
program,  they  are  also  focusing  a share of  their efforts on
engine  modifications,  advanced   turbocharging and  even  flow
through catalytic  converters.[3]

      Durability and  packaging problems led to  the  suspension of
 the  only  production  trap systems  this  past  model  year.   As
 such,  alternative control  systems and  improved  trap  oxidizer
 systems  are needed  for Diesels to meet the current  light-duty
 and  future heavy-duty particulate emission standards.   Problems
with the developing  traps,  which are  generally either  wallflow
 or  porous  foam ceramic substrates,  include inadequate  surface
 area,   quick   backpressure  buildup   and/or   nonuniform   soot
 distribution and  regeneration.   Proper  catalytic  coating can
 alleviate the  regeneration  and  surface  area  problems  to some
 extent,  but  trapped ashes  from  lubricating  oil  can  decrease
 catalytic activity drastically.[4]

-------
                                       -2-


                                     Table  1

                    New Vehicle  and  Enoine  Emission Standards
               LDV (qpm)
         HC     NOx
                LPT (qpm)
              HD (g/BHP-hr)
 Part.   HC     NOx
Part.   HC     NOx     Part.
1988-9  0.41    1.0
0.2     0.8  1.2/1.7[1]  0.26[2]  1.3   10.7    0.6
1990    0.41    1.0
0.2     0.8  1.2/1.7     0.26    1.3   6.0     0.6
1991-3  0.41    1.0
0.2     0.8  1.2/1.7     0.26    1.3   5.0[3]  0.25[3]/
                                               0.10[4]
1994*   0.41    1.0
0.2     0.8  1.2/1.7     0.26    1.3   5.0     0.10(5]
[1]  Vehicles over  6,000  Ibs GVW remain at 2.3 gpm NOx until 1990.  Standards of
     1.2 gpm apply to LDTs up to and including 3,750 Ibs.  loaded  weight;  1.7 gpm
     standard  applies  to LDTs  equal  to  and  over  3,751  Ibs.   loaded  vehicle
     weight.  Emissions  averaging may  be used  to meet  this  standard  provided
     that  trucks produced  for  sale  in  California or  designated high-altitude
     areas may be averaged only within each of those area.  Diesel and gasoline-
     fueled engine families may not be averaged together.

[2]  Emissions averaging  may be  used to meet  this standard provided that trucks
     produced for  sale in California or  designated  high-altitude areas  may be
     averaged only within each of those areas.

[3]  Emissions averaging  may be  used to meet  this standard,  but these emissions
     may  not be averaged   with  HD  gasoline  engine  emissions.   Averaging  is
     restricted  to  within useful  life  subclasses  (see  below).   Also,  averaging
     is restricted  regionally—the  two  regions are California  and the other 49
     states.

[4]  For urban bus  engines,  the  standard is 0.10 g/BHP-hr—particulate averaging
     is not allowed with  this standard,  but emissions from these  engines  may be
     used in NOx averaging.

[5]  Emissions averaging  may be  used to meet  this standard.   However,  averaging
     is  restricted  to  within  useful  life   subclasses   (see  below).    Also,
     averaging is restricted regionally—the two  regions  are California  and the
     other  49  States.  Emissions  from  engines  used  in  urban  buses  may not  be
     included in the averaging program.

     Full useful life is established for 1985 and later defined as:
     Light heavy-duty (under  19,500 Ibs. GVWR) = 8 yrs/110,000 miles
     Medium heavy-duty (19.500-33,000 Ibs. GVWR) = 8 yrs/185,000 miles
     Heavy heavy-duty (over 33.000 Ibs.  GVWR) = Syrs/290,000 miles

-------
                               -3-


     The  concept  of resistively heated Diesel particulate traps
               regeneration  efficiency and  uniformity  has  been
              EPA  over the past two years. [5]  It was  found that
          *•  •  reear even with  a  particulate trap  reduced in size by  50
             heavv-duty  side,  the General  Motors  Corporation
           ^oQrfPOrt  ?f  traP  development  aimed  at meeting  the
           1994  P^ticulate  emission  standards. [8]    The  CMC
system  is  a   conventional   cordierite   ceramic  trap  with  a

-------
                              -4-


burner-bypass   system.    This   system   shows  promise   toward
achieving   290,000-mile   durability  though   only  100K   mile
durability  was simulated,  and  the  system does  not meet  both
1991 particulate and NOx  emission  standards  simultaneously.   If
the  1991  NOx  standard   of  5.0  g/BHP-hr  is  met,  the  lowest
particulate  emission  rate  achieved  is   roughly  0.3  g/BHP-hr,
i.e.,  20 percent  over the 1991  standard and  300  percent  over
the  1994  standard.   Among other heavy-duty efforts, Cummins  is
currently evaluating a resistively heated trap aimed at meeting
the  1991 and  1994 particulate  standards,  a  modified  Coloroll
system,  and  have  yet  to  report  their experience  with  it.
Southwest  Research Institute  has  also  proposed   an  evaluation
(see  Appendix) of a  modified Sohio silicon carbide  or  other
high-temperature resistant trap as part  of work to be  performed
by the Consortium  of Heavy-Duty Diesel Manufacturers.[9]

II.  Technical  Progress

     The  resistively heated trap  material  suppliers dealt with
can  be  sorted  into two  categories,  those who  have either sent
EPA   material  samples   for   evaluation  or   have  done  some
evaluation  themselves, and  those  who   have  either not  shared
with EPA the results of their  development work or have not done
much  development   recently.    The  "haves"   include  Coloroll,
Sohio,  Duriron, Hi-Tech  Ceramics  and Aker  Industries,  and the
"have  nots" include NGK,  Corning, Matsushita, most heavy-duty
manufacturers   and    others.    Suppliers    who    plan   either
development or  evaluation in  cooperation with EPA in the future
include  Selee, Hi-Tech  Ceramics and possibly  Sohio,  Garnet and
Duriron.   These latter programs  will be discussed  in  the next
section.

     A.     Coloroll  (Low  Density)

     A complete description  of the first and  second generation
Coloroll trap system evaluation is  contained in  reference [5].
Efficiency tests  were run on the  new Coloroll  filters which are
two-thirds   the weight  of  the  original filters   (i.e.,  more
porous).    Load-up  time  is about  the same  since these filters
are  still  surface  loaders (rather than  bulk loaders)  and more
soot  passes  through.   Baseline  0.45  g/mi  particulate  was
reduced  to 0.23 g/mi  for an efficiency  of  roughly 50 percent.
Regenerations  were  performed efficiently with similar results
as the previous samples.[10]

     B.     Coloroll  (Bypass Regeneration)

     Tests were run with the  original Fogarty  can  equipped with
a baffle plate and  dividers  to bypass  exhaust flow through two
unpowered  filters  while  the  third   (isolated)   filter   is
regenerated.    No  improvements  in  regeneration   efficiency   or
regeneration time were  obtained.   At low speed  and idle, very
 little heat  is available to  enhance the burn.  At high speeds,
backpressure increases in spite of the  regenerating filter, and

-------
 the  higher  amount  of  exhaust
 cools   the   burning    filter
 temperatures.   A  baffle plate
 allow  only  1  or  2  cfm
 regeneration.    The   basic
 problems  in   not  having   adequate   surface  area   to   lower
 regeneration  frequency  and  in   unknown  durability  of   the
 ceramic-to-metal   interface  even   when  intermediates  such  as
 steel wool  matting are used.
                      flow through  the  baffle  plate
                       down   below   soot    ignition
                    with  smaller holes,  which  would
                 maximum  exhaust  flow,   may   improve
                   Coloroll   system   still    exhibits
      C.
Sohio
      Sohio  Engineered  Materials,  Inc.  (formerly  Carborundum)
 previously  sent  EPA  several  2-inch   thick  5.25-inch  diameter
 silicon  carbide  foam  "doughnuts"  (1.75-inch  inside  diameter)
 for  evaluation.   These  hollow  cylinders were quite  heavy,  over
 600  g each  (0.95 g/cm*1),  and as such  required  too much  energy
 (over the  2.5  KW power  supply  capacity)  to heat  to  regeneration
 temperatures.  Three  doughnuts  were sandwiched  together  in  our
 test  rig  for  initial EPA  evaluation  (see Figure  1), to get  the
 filtering  capacity that  these  Sohio  doughnuts  exhibited  when
 tested previously  by  Southwest Research  Institute.[11]  EPA was
 at first more concerned  with  the resistance  problems which  were
 anticipated with this design.   It  was  believed that, even  when
 steel wool  intermediates were used between doughnuts,  it would
 be  better  to  have just  one  6-inch thick doughnut  to minimize
 the  number  of  material  interfaces at  two.    However,  in   the
 interest of  also reducing energy requirements,  EPA asked Sohio
 to significantly increase  the  inside diameter  such that outside
 surface area  is preserved but  the  filter mass  is  reduced.  A
drawing of the filter material  configuration  requested is shown
 in Figure 2.   Sohio has  yet  to  produce  such  a  sample  for  EPA
evaluation but is still committed to doing so.

                            Figure 1

           Proposed Sohio  Diesel  Trap Configuration
SOI.lt> CM4M/C

-------
                               -6-


                             Figure  2

        Sohio Resistive Diesel Particulate Filter Element
                         OPERATING RESISTANCE - 0.3 - 0.5 ohns
                             KMDS11Y - 80 - 100 ppi
                                                          5.25"
     Particulate  collection  efficiency  and  loading tests  have
been  carried  out on  the  Sohio  silicon  carbide  foam  samples.
Silicon  carbide  foam samples with filter  porosities of 45,  60,
80 pores  per linear inch  (ppi) have  been  received  from Sohio at
Ricardo   Consulting   Engineers,    pic.,   EPA's   Diesel    trap
evaluation contractor.                                          F

     The   Sohio   silicon   carbide   foam   examples  exhibit   a
C®:a He ly,  poor  maximum  collection  efficiency  of  30  percent,
with the  larger  pore-sized samples showing  signs of particulate
 blow  off.    Tnermogravimetric   analysis   of   the   particulate
samples  shows  that the  Sohio Diesel  particulate filter element
content [12]         makeuP of the  filtered particulate  carbon
          o    u   a  tTP1"1  vehicle  12-volt  battery  and  the
        \ 3-inch  x 1-inch  Sohio  80  ppi  silicon  carbide  foam
      .  °WS uthat a. material  surface temperature rise of  135°C
      i  K°a? in9  achieved ^  about 10  seconds.   Wire  wool  and
thin  (about 10mm) stainless steel plates  were  used to make  the
electrical  terminations.  Current  measurement was attempted  but
the  levels  involved  exceeded the full-scale range of  75 A  of
        K'..  Urtuh6f tests  Usin9 a Partially  discharged  6-volt
     ai  batter7 had  al lowed current  measurements to be made  and
n  ^ electrical  resistance  of the sample  to  be determined  as
0.055  ohms  with  electrical  contacts  on  the  2-inch  x  1-inch

-------
                              -7-
faces,   i.e.,  oriented for  maximum resistance.   This  level  of
resistance  was   considered  to  be   too   low   for   practical
particulate trap-application  (i.e.,  current requirements  for  a
given voltage  are  beyond  the  capacity of  conventional  power
supply sources),  but  because  of the briefness and simplicity of
the  tests,  the  resistance  level  measured may   not  accurately
reflect the inherent material  characteristics.

     D.     Duriron

     EPA's   initial   contact   at  Duriron   retired   and  his
successors do  not appear  to be aggressively pursuing  any major
development  in resistively  heating as  yet.   Initial evaluation
of  their  material   showed  a   need   for   improved  electrical
properties.   Duriron  is  currently supplying Ricardo  with some
porous   (nonconductive)   cordierite  samples   that  have  solid
conductive silicon  carbide  coatings.

     A  sample  of the  Duriron non-conductive  foam with integral
downstream membrane was  tested.   Results  showed  a particulate
collection  efficiency in the  region  of  40 percent  and  a back
pressure  rise  rate   between   that  of  the  standard  Coloroll
element   and   the  Sohio  80  ppi   Sic  foam.   This  collection
efficiency may be good for  low  ppi  samples, but  not so good for
high  ppi samples.  The Duriron material  is  believed to be low
ppi  (e.g.,   about 50  ppi)  but  the effect of  the  membrane  on
overall  porosity and  soot collection is unknown at  this time.

     E.    Hi-Tech  Ceramics

     The Hi-Tech Ceramics material was initially  evaluated in  a
l-inch  thick  by 2-inch  diameter  sample.   Their  material  was
subsequently evaluated for resistive  heatability by clamping  a
6-inch  long sample  between two  1/2-inch  thick  solid  stainless
steel   disc-shaped   electrical   leads,   which   are   in  turn
sandwiched  between two solid ceramic insulators  such  that all
current   is  dissipated through the porous SiC sample.   Current
and  voltage  were measured  as power was supplied to the  heater
assembly.   Initially, the  sample  was  weighted  to  determine
material density (equal  to  0.285  g/cm3)   which   is  roughly  20
percent  less dense  than the initial smaller sample.

      The  evaluation  showed   that  this  material   displays   a
negative  temperature  coefficient since   resistance  decreased
with   increasing   temperature.    Exact   filter   temperature
measurements  were  not   obtained  due  to   lack   of  an  optical
 instrument  which operates  in  the  range  of filter  temperatures
experienced  during thermal cycling of  the sample.   Resistance
 approached  1.0  ohm at very low power settings and  decreased  to
 roughly  0.5  ohms  at  power  settings  over 1500  watts  (maximum
current  loadings).   Current  applied to  the  heater  did not
 remain   stable at  a  given  setting  and decreased  1 to  2 amps
before  stabilizing.   Voltage drifted  down 20  to  25 percent from
 the  preset  loading as did the initial  sample.    The  electrical
properties  observed through  the  thermal cycling  of the material
 are  shown in Table 2.

-------
                               -8-
                             Table  2

                 Electrical  Properties  of Heated
              Hi-Tech Ceramics  Silicon  Carbide Foam
Test Run

    1
    2
    3
    4
    5
    6
    7
    8
    9
   10
Initial
Voltage
(volts)

  21.8
  25.5
  26.5
  28.3
  29.6
  18.0
  10.3
  15.0
  14.4
  12.0
Stabilized
 Voltage
 (volts)

   17.7
   20.4
   22.8
   24.4
   26.8
Current
(amps)
  20
  28
  38
  48
  58
  33
  28
  27.0
  24.5
  19.5
5
7
0
0
5
5
5
 Power
(watts)

   363
   585
   866
  1171
  1568
   603
   465
   405
   353
   234
  Heated
Resistance
  (ohms)

   0.86
   0.71
   0.60
   0.51
   0.46
   0.51
   0.54
   0.56
   0.58
   0.60
     Contrary to  the  initial  sample results, the  current drift
did  not  increase with  increasing  temperature.   Also,  perhaps
because  the  sample was  deformed  as-received,  further  material
deterioration was  observed  with continued thermal  cycling.  The
material  glowed  orange-hot  at  current  loadings above  30  amps,
but  the  glow  was  not  uniform and tended  to be concentrated
along the cracks  in the  material.   Length-wise  cracks developed
on  the  outer surface  of the  material,  but  the sample remained
intact   as   one   piece.   The  cracking  may   also   have  been
exacerbated  by the  use  of a  rigid  clamp  to  hold  the  heater
assembly  together  rather than  a  spring loading assembly which
would  compensate  for  the  different   expansion   rates  of  the
ceramic   and  metal   components.    The   negative  temperature
coefficient  (NTC) characteristic  of the material  is seen  by
comparing test 1  with test  10  in Table 2.  After  the  material
had  been heated  in tests 1  through 9,  the low  power resistance
was  0.60 ohms  compared  to  0.86  ohms  when the  material  was
 cold.

     A tubular cracked sample of  the Hi-Tech SiC ceramic  45 ppi
foam was  shipped  to  Ricardo  for  further   evaluation.   Although
fractured, a sample 75 x 50 x 25 mm was machined from the piece
and  characterization   tests  were carried out.[13]   Graphs  of
backpressure rise versus  loading  rate for the Hi-Tech and other
samples,  as  tested by Ricardo,  are contained  in  the Appendix.
The particulate loading  tests  showed  a  collection  efficiency of
5-20 percent at the standard engine condition  (medium speed and
load)  used  throughout   the   Ricardo   characterization  tests.
Tests at  a  lower engine  speed to  reduce  the  exhaust gas  flow
rate through the  sample  resulted in collection efficiencies  of
10  to  30 percent.    Backpressure  rise  rate was  insignificant
indicating  that   particulate  was  regularly being "blown  off"

-------
                              -9-
during the test  period.   This was  readily  apparent when  a  jet
of air directed  at  the  loaded foam sample removed from the test
installation  resulted  in  clouds  of  particulate  dust  being
emitted  from  the sample.   Subsequent electrical  tests en  the
Hi-Tech   sample   using   a   typical   12-volt  vehicle   battery
indicated an electrical  resistance  of about 6 ohms resulting in
little resistive heating of the sample.

     EPA  now  proposes   to   build  a  container   for   further
evaluation  of   the.  Hi-Tech  Ceramics  material  as  a  Diesel
particle  trap/regenerator.   The  proposed testing  apparatus  is
similar  to the  container  shown  in  Figure  1.    Testing  shall
include  determinations of  filter efficiency,  backpressure rates
and  regeneration efficiency on  the   EPA's  Mercedes  300D test
vehicle.    If    feasible,   our   Diesel   evaluation  contractor
(Ricardo)  shall   fabricate  the  proposed  container  and perform
initial  evaluations prior  to more detailed evaluation  of  the
final  system configuration  at  the EPA  Motor Vehicle  Emission
Laboratory.

     F.    Aker  Industries

     Aker  Industries  had   recently   been   formed  to  exploit
commercial markets  for  advanced ceramic  materials.  Dr. Benson,
the  CEO, had previously  been associated with Energy,  Research
and  Generation,  Inc.  (ERG) who  supply predominantly defense
establishments with specialized materials.

     ERG  had  been  involved with   particulate   traps  in  the
mid-1980s  and  had  supplied  electrically   regenerable silicon
carbide   foam   traps  to   various   engine   manufacturers  for
evaluation.   Their  trap  design,  shown in  Figure  3,  included  a
foam with a  variable density  so  that  electrical terminations
could  be  made  by  brazing  metal  contacts  onto  solid silicon
carbide.[14]    This  avoided  differential   thermal   expansion
problems  between  the  metal  contact  and  ceramic,   which  was
another  problem experienced  in  the previous evaluation  of the
Coloroll trap,   since the solid silicon  carbide  and the  brazed
metal   contact   presented  insignificant  electrical   resistance
compared to the main part  of the silicon carbide element and
were thus not resistively heated.  Tests with  this trap  design
showed that,  w(ith  a 100  ppi  foam about 40  mm thick, particulate
collection efficiencies  were only 30-35 percent.     ,

     Following  this work  a  non-rigid pleated fiber  mat  filter
design was developed.   A tubular construction was used with  an
outside diameter of 6 inches, a 0.02-inch  wall  thickness and  a

-------
                               -10-


                             Figure 3

             Aker  Industries Diesel Trap Configuration
                                 TURBO EXHAUST
    MARMON
   V EEC LAMP
                                                  FLANGE

                                             CERAMIC LINER
  ,  ELECTRE
   TERMINAL.
CERAMIC FELT
:NSULATKJH
                                                          CERAMIC
                                                         SPACED RING
                                                             EXHAUST
REMOVABLE
 END PLATE
 CONNECTOR
 STRAP

  GROUND
  TERMINAL

  CERAMIC
  SPACER
  KINGS
  -STEEL SHELL
                                                            CERAMIC
                                                            LINCR
                                                           •FLANGE
                       DUOCEL SILICON CARBIDE FILTER
                                                    •04	CERAMIC TELT
                                                        IWSULATIOK
 length of 8  inches.   A 10mm diameter  silicon carbide  fiber was
 used  in  the  matted  material.   Tests  were   conducted  on  a
 Mercedes-Benz 190D vehicle over  492  miles.   An  average trapping
 efficiency  of  91  percent  was  achieved  and  roughly  50g  of
 particulate  was  collected  before each  regeneration  which  was
 carried  out   be  placing  the filter  in a  kiln  and heating  to
 980°C  (1800°F).    Typically  pressure  drop  across  the  filter
 increased from 55  mbar for the cleaned condition to 90 mbar for
 the  loaded  condition.   The   ratio   of   mass   of   particulate
 collected at  a given  backpressure to  the mass of the  filter was
 approximately  1.00 compared  with only 0.10  for  the  Coloroll
 material.

      Aker Industries  are  currently  working  on  manufacturing  a
 silicon carbide  non-rigid fiber  mat  filter of  pleated  tubular
 construction.  The filter,  18  inches  long, will have an  inner
 diameter  of   5  inches  with  1-1/2  inch  deep pleats  giving  a
 surface  area enhancement  factor  of   5.2.   The   filter will  be
 corrugated in the axial  direction   and  will  be  supported  by
 metal  rings   at  various  intervals along   its  length.   It  was
 claimed  that the  pleated  structure   would minimize   radiative

-------
                              -11-


heat loss during regeneration  as  heat  would mainly  be  radiated
from  one  pleat  to  another.    in  addition,   Aker  Industries
explained that  a  non-rigid  fiber  mat  construction would  not
catastrophically  fail  by  cracking  as   is  the case in  a  foam
structure which behaves as a monolith in its failure mechanisms.

     Electrical terminations will  be similar  to those  used for
the  roam design.   Ricardo  currently lacks enough  EPA  contract
funding   for    purchasing   additional    material    samples   for
materials characterization  tests.   Aker Industries  stated that
it would be manufacturing filters  for clients later this year
and that  it may be possible to supply Ricardo with  some sort of
free sample at that time.

     Aker  Industries  described the manufacturing  process they
employed  for  making  silicon carbide.    A  textile  precursor  is
used  to  manufacture   a  vitreous   carbon  copy  which  is  then
infused  with  silicon  to  make   a  100  percent dense  slightly
silicon  rich  of  stoichiometric  beta  form of  silicon  carbide.
This  results  in  a material  with  a fine  grain size  giving it
improved  durability   and   a  positive   elecrrical  resistance
temperature  coefficient.   By  making   the fibers   100  percent
dense,  oxidation  of  the material under  high temperatures  is
reduced  to  surface oxidation  only.  Aker  Industries noted that
for  durability,  the  temperature  of  the   material during use
should  be  maintained  below  1200°C (2200°F)  and  that contact
with  vanadium  and alkali  metals  should  be  avoided  as  these
metals  react  with the  silicon  carbide by ion exchange.  Aker
also  claims  that  their manufacturing process  is much  less
costly  than the Coloroll process,  which utilizes expensive raw
materials.    Aker   Industries  explained   that  it  also  has
expertise   in   manufacturing  alumina,   silicon-nitride,  and
mullite.   It  has been agreed that  Ricardo  will keep in  contact
with Aker Industries to  monitor developments.

     G.    NGK

     Representatives  of NGK-Locke,   Inc.  met  with   EPA last
February and discussed the  possibility of  working together on  a
cooperative  trap  development  and  evaluation  program.   NGK has
considerable   experience  in  developing   wallflow  cordierite
honeycomb   trap  material   but  has  never  resistively   heated
conductive   ceramics.    Their   cordierite   traps   have  been
evaluated   in  the  EPA  laboratory  in  the  past  with  wire
face-heaters.    NGK  is   also  capable  of  producing  variable
porosity traps to  improve  uniformity  of   soot   loading and
regeneration.   NGK indicated at  this meeting  that they had some
experience  working with silicon-carbide  and  other conductive
ceramics,  but that they had not  as yet produced  a  trap made of
these materials.

-------
                                 -12-
              ^	 that  NGK produce  a  wal If low-type trap  made
  of  silicon carbide  or some  other durable  conductive  ceramic.
  The  configuration, which  we believe  lends  itself  to  improved
  electrically heating,  is  a rectangular  cmhs+-rai-o  ^«^-i~~ ;.i*-u  -
  ji	     -^    orientation

                                       is shown in Figure  4.
                                      ilong  each cell wall between
  —  	*^«*a  wnxuii  ire  placed  on  opposite  sides  of  the
  rectangular substrate.


                              Figure 4

                   Proposed NGK Wallflow Monolith
                 Si-C Resistive Trap Configuration
H.
            Corning
 over   hto             had  invested  $2  million P«  year
 (wU)  for Diesel ^ p^rtYcuUte  f n?P^ th\ wal1  f low monolith
 that  this rate  nf  fmrof?      filtration,  but  it  was unlikely
 years. [14]   cSrrentlv  !S?nt W<^Uld  continue °v^ the next  two
 Mines  who were sucCL^ir     6S  °f  WFMs  included Canadian
 achieving   particulat-p ffully, ,US1^ WFMs in  mining machinery  and
 percent   These mach^-  collectlon    efficiencies    of   90-95
 fuch ?hat auto  regenerationUofa^a  ?ufficiently high duty cycle
 and   reliably.   ^£5 ^   are   ,1-n?  fi^er  tak6S ?lace re^larly
manufacturers7  for particulatf  i in-r^9   rPP}ied  tO  6ngine
was reported that manv mTi *  «. flltratlon  development,   and  it
*wne^££ertti0^£^¥Kt™*r*  were conducting tests with
control syl?lms  and  t^e  reUa?A7tvth%high, C°St  °f  burners and
                     tne  reliabili   and safety question marks.

-------
                              -13-


Corning   reported   that   a   typical  burner   used   in   these
applications,  a  Webasto,  cost  around  $5,000  each.   It  was
Coming's opinion that in a bypass system, two  filters would be
needed since raw exhaust  would not want to be allowed to bypass
straight to  the  atmosphere.   It v.'as mentioned  that development
work was still continuing with the Mercedes-Benz 300 TD filter.

     Corning   claimed  that   sufficient   studies   have   been
performed   developing   regeneration   strategies   such   that
durability  problems  with  Corning  WFMs have been  overcome.    It
was  reported  that  a  15-inch  diameter  by   15-inch   long  unit
supplied by  Corning  has completed 4,000 hours of operation with
no problems.   Corning also added  that they had had  no vehicle
recall claims  on three-way catalyst substrates since 1975.

     Corning  was asked  whether any  of  their  current programs
involved  the  development  of  electrically   regenerable  Diesel
particulate   filters   (DPFs)  using   electrically   conductive
ceramic  filter  materials.    Corning  reported  that   they  had
worked  with Sohio 10 years  ago  and  had successfully  extruded  a
silicon  carbide  WFM;  however,   sufficient  porosity for  DPF
applications  had not  been  achieved.   There were  also thermal
shock  resistance  concerns  with  this  type  of  DPF.   It  was
reported  that  Corning had discussed the possible application  of
a  silicon  carbide WFM DPF  with Ontario  Research  during  the
proposal  stage  of the current EPA/Ricardo  contract.  Corning
also  reported  that  they  had made  some  new  developments  with
electrically conductive materials which may be of interest for
DPF applications, but would not discuss these until outstanding
patent  issues  had been  resolved, probably  later this year.

Ill. Future  Work

     A.     Selee

     John Howitt (formerly  of Corning)  visited EPA in May  1988
and  is  fabricating samples of  reticulated  foam which he claims
can  be,, but may not  need to be,  resistively heated. [15]  Selee
will   soon   be  delivering  samples   of  these  foams for  EPA
evaluation.    A picture   of   the  trap  configuration  initially
envisioned  is  shown  in Figure 5.   This trap configuration is  in
a  "top-hat" shape such that  all  particulate must flow  through
an enhanced surface  area axially and escape through  the sides
and outlet  end both  radially and axially.

      Selee  has  since  decided to change  this  configuration  to
the  one  shown in Figure 6  for  ease  in  fabrication.   EPA  is
currently  fabricating  the  container shown  in Figure  6  to  be

-------
                          -14-
                        Figure 5

               Proposed Selee Diesel Trap
                Evaluation Configuration
                   Hi
 to  other  trap  material
 evaluating  in t'he near future
conductive ceramics i
          o,
B-     Hi-Tech Ceramics
some  preliminary  evaluatnn
found to be  resLtTvely
filtering  efficient
samples and hiJr
the material  are  inadequat
through the substrate
                                        for
                                       be eas"y  adaptable
                                        h"EPA   may   be
                                                     other
                                      Hi
                             -   mplea'  EPA  has
                               - mat/rial  and  ^ has  been
                                cardo also performed  some
                                bui1duP>  tests   on  ^ese
                                        Characteristics  of

-------
                                                 -15-
                                               Figure 6

                              PROPOSED HTESEL TRAP TORmATION CONTAINFP
                   SPRING ASSEMBLY
                   WITH SET SCREWS
       9-0"  (22.9 on)
HI-TECH DIESEL TRAP SUBSTRATE

                                                                            SOLID CERAMIC
                                                                           INSULATING DISC
                                                                       ELECTRICAL TERMINALS
                                                                         V8"  (0.3 on)
                                                             SOLID CERAMIC INSULATING RING
                                         6.5"  (16.5  cm)

-------
                              -16-


backpressure like  a  conventional trap.  The reason  for  this is
the  low  material  density  at  45  ppi.    EPA  advised  Hi-Tech
Ceramics  to  configure  the  next  samples  in 80  to  100  ppi  for
improved  filtering.     It   is   recognized,   however,   that  the
increase  in  density  will also  require an  increase  in external
energy  required to  regenerate.   Hi-Tech is  currently retooling
for these design  changes and should also  be providing the more
optimized  trap  samples for EPA evaluation sometime  later this
Fall.

     C.    Equipment Status and Other Material Suppliers

     The  EPA laboratory is  currently functional for all aspects
of   Diesel    particulate   control    technology    evaluation.
Particulate  sampling  and  analysis  systems have  been repaired
and improved over the past  year and test rigs  have been and are
being  fabricated  to handle  a  wide  variety  of  test material
configurations.  The Mercedes  300D  test  vehicle has  been  kept
in  proper maintenance  and  all  other  instrumentation  is  set up
and operational.   Other material suppliers  have been contacted
and may  also  be  providing samples for EPA evaluation.  Among
these   companies   are   Brunswick  Technetics,   Garnet,    Nippon
Ceramics,  Union Carbide,  Norton Industrial Ceramics,  Cheltenham
Induction Heating, and  Mohawk Sintered Alloys,  Inc.

-------
                              -17-


IV.   References

     1.     "Mobile Source  Emission  Summary,"  Office  of  Air and
Radiation, U.S. Environmental Protection Agency,  March 15, 1987.

     2.     Automotive News,  1988  Market Data Book  Issue,  April
1988.

     3.     Developments  in Diesel Particulate Control,  SP-735,
Society  of  Automotive  Engineers,   International  Congress  and
Exposition, Detroit, MI, February 29-March 4,  1988.

     4.     "Fuel   and   Lubricant   Effect   on  Durability  of
Catalytic  Trap Oxidizer (CTO)  for  Heavy-Duty Diesel Engines,"
Saito,  Koichi,   Yasuo   Ikeda,   and  Shoichi  Ichihara,  Nippon
Shokubai Kagaku Kogyo Co., SAE Paper 880010, March 1984.

     5.     "Evaluation   of  a  Resistively   Heated  Conductive
Ceramic Fiber  Diesel Particulate Filter,"  Bruetsch,  Robert I.,
EPA/AA/CTAB-87-04,  U.S. Environmental  Protection  Agency,  June
1987.

     6.    "An Exhaust Gas  Aftertreatment  System  to  Reduce
Particulates for  Full-size Passenger Cars," Lepperhoff, George,
Jan  widdershaven,  and George Lutkemeyer, FEV Motorentechnik and
Ivan Hedin, Volvo  Car Corp.,  SAE Paper  880003, March  1988.

     7.    "New   Results  of  Passenger   Car   Diesel   Engines
Pressure   Ware Supercharged with   and without   a   Particulate
Trap,"  Hiereth,  Hermann and  Gert Withalm,  Daimler-Benz AG, SAE
Paper 880005,  March  1988.

     8.    "Development of   a  Particulate  Trap  System  for   a
Heavy-Duty Diesel Engine,"  MacDonald,  J.  Scott  and Gerald M.
Simon,  General Motors  Corp.,  SAE  Paper  880006, March  1988.

     9.    "Addendum   No.    1  to   SwRI   Proposal  08-2857A-:
Development  of a  High-Temperature  Resistant Diesel  Particulate
Trap,"  Bykowski,   Bruce  B.   and  Charles  T.   Hare,   Southwest
Research  Institute,  San Antonio,  TX, February 5,  1988.

      10.   "Status of  In-House Electrically  Regenerable  Diesel
Trap Evaluation," Bruetsch,  Robert I., EPA  memo  to Charles L.
Gray, Jr., June  15,  1988.

      11.   Formulation    Evaluation    of    Alternate     Diesel
Particulate  Trap Media,"  Bykowski,  Bruce B.,  Southwest Research
 Institute,  San  Antonio,  TX,  American  Society  of  Mechanical
Engieners Paper  No. 87-ICE-36,  New York,  NY,  February  15-20,
 1987.

      12.    "Technical  Progress  Report,"   Sown,   Nick,   Ricardo
Consulting Engineers,  pic.,  EPA Contract No.  68-03-3519,  August
 8,  1988.

-------
                              -18-
IV.   References (cont'd)

     13.    "Technical  Progress  Report,"   Bown,   Nick,   Ricardo
Consulting Engineers, pic. ,.EPA Contract No.  68-03-3519,  August
2, 1988.

     14.    "Technical  Progress  Report,"   Bown,   Nick,   Ricardo
Consulting Engineers,  pic.,  EPA Contract  No. 68-03-3519,  July
18,  1988.

     15.    "Selee  Diesel  Particulate  Filter,"   Howitt,  John,
Selee  Corporation,  Hendersonville,   NC,  presented  to  EPA  by
Selee  at the  Motor  Vehicle  Emission Laboratory, February 29,
1988.

-------
V. Appendix

-------
     « .........
                                                       - 2S
   K F
c. .......... „.?.',.'""'  B'"""t
                                                                  "*«
- 60
   40
 u
 «  0
u

;*-20
o
J
           A
                                         t.
                                        a.
                                         V
                                         e
                                        O9
                                  ^v	    „^..«5«'—»•••
                                       \.

-------
02/00 cWl la: 13
        OFF MM.,,. i  Ch.r.M.r,.M,., T.,t.
                 F...  -
   9



 = 8
 « 5
      I

      I

      I
      I
      I
 c.
0.
o
e
03
                I
                t
                I
                I
p.r«v..,.0t;8L.^l,;-(«,f-o
                                       2.4

-------

              •- I5W
  16
 ^ 10
 O
 oc



 r 8
 flC
9
fi
r»

: 4
a.
 i

 <
 i
 i
 i
 i

 i
i

i
i
i

-------
  SOUTHWEST   RESEARCH   INSTITUTE

  POST OFFICE DRAWER 28S10 • 6220 CULEBRA ROAO • SAN ANTONIO. TEXAS. USA 78284 • 15121 884-S111 • TELEX 24484*


             t
                               February 5, 1988
TO:

ATTN:

SUBJECT:   Addendum No.  1 to SwRI Proposal  08-2857A-, "Development  of  a
             High-Temperature Resistant Diesel Particulate Trap"
                               L  BACKGROUND

      This addendum to the subject proposal was prepared as a result of encouraging
experimental data and favorable responses from the consortium participants.  The
initial project resulted in the acquisition of  significant generic data pertaining to
the regeneration  of trapped paniculate.   Using  these  data, several approaches to
improve current regeneration techniques, and suggest new ones,   were identified.
One of the  new approaches  identified was to experiment with alternate trap media
intended  to  withstand  the  highly  exothermic  combustion  of  trapped  diesel
p articulate.

      The alternate trap medium  briefly  experimented with  was a silicon  carbide
foam.  This foam was evaluated using a heavy-duty diesel engine in a  transient test
cell.   Based  on preliminary  tests, trap efficiency was good  and the resistance to
regeneration  excellent.   Several areas  require further development.   These  areas
include trap  design considerations,  trap  efficiency determinations,  application of
regeneration  techniques,  and  long-term  durability.   This  addendum outlines  a
proposed work effort to investigate the aforementioned  areas.

      Because of the successful  approach of using multiclient  sponsorship, this
addendum proposes to continue the previous consortium format.  New participants
are able to join this work effort under the same conditions  as new participants were
able  to join the original consortium.  Improvements in  the overall management and
direction of the consortium will be implemented based on the suggestions received
from the current participants.
                    SAN  ANTONIO.  TEXAS
                    DALLAS. TEXAS • DETROIT. MICHIGAN • HOUSTON. TEXAS • WASHINGTON. OC

-------
                            IL  STATEMENT OF WORK


       This section describes the objectives, approach, and scope of work proposed.
  The scope o£ work will remain flexible because of the exploratory research nature of
  the project.


  A.   Objective



       The primary objective of this addendum  will be to design and evaluate a high-
  temperature  resistant diesel paniculate trap.   To  meet this primary  objective,
  several tasks are required, each with its own objective. During the first part of the
  past  consortium, the objective was to  obtain  generic data  to better understand
  regeneration.  This addendum's activity will be more specific to hardware design and
  development of an actual system.
  B.    Approach
                                                          ,
  consortium  activity' will be n» J      •           "*  coac^<"»  from  the put

             — —
 C.   Scope of Work
                                               °f E^°~  Research proposes to

 exploratory na ur  of «^Bro~ H° "^ *" ?rinCi?ai obie«ive.  Because of the

 for each area oTL^f                   M PiaM ^ be
 -ill be prepared «*
      1.    Task 1 - Trap Design, Engine Selection, and Baseline Tests
 the  past            .^

 Moughnuts." The number of Mougm^ and L. S-""^7  °f  ^^  Capbide
 selected arbitrarily.  After additXT,.             Slgn ot the traP container were

 container design^S bt made  T?e ^je^s l^T ^TT""141 " ^ ^
 evaluate several trap desisns usinJ rhl a ?      ?" phase o£ the Pr°Ject wiil be to

 be  performed  on ? eS  th^        ^ C*Wd' f0am'  TlieSe e*alu«io'* ^
                                  emits
evaluations  will  be  cr              °*     *****  ?«*««»«..   The
                     crned             i
backpressure effects, and effects Len^Ionl,P  etflcienc7'  traP  capacity,
trap location in the  •xh«J  M "«rSi ?     °Pera"°= and emissions.  The effect of
determined.                '     *" " en?lne tra^ie« cycle effects, also will be
will be repe                                         "^^ -P P«— ««
                         «                  tirst of these additional engines will

-------
                                        -                            selection
  trap ef ectivenw   -e sH     S1?amcail£ paniculate character differences on
           ~™
  medium andaBonerthe -T^' ^^ desl?n °£ a P«ticul.t. era,
  durability on  a^SitTS ±Le? ^f"' accePwlU mv°ive  a s<«ionary
  Catena!. For example ^eU^ T^'  " a  ^^""P-"™" "^S'1  of  the
  exhaust temperatures  waTlTnSo^^    '^ e:XCUrsions to aciieve ^S^ and low
  term  thermal resistance  The SS^Sf h°B .an11SOi,at;d tfap t0 ^""a-  "s long-
  the thermal aspect 5? trkn intJS?   p    U?latad fr°m  vibration £o  ««»«. only
  trap efficiency!  aid winL^o^ I:  ?artlc*ia'
-------
   litigated to ^aLte their  ene^7StS  !f d ^  US6  °f fuel additi-*  will  be
   regeneration. A*dapt£fon o ^stat wf r!f '   ?** *' tem?era*ure requirement for
   if the.  consortium participant ? relul,  ,"•      ' SyStSmS ^ be P*«°««°00  °n a cost
required  flexibility  to redirect  the  Tecii ° -*h" Specific  experiments and the
reimbursement contract  aDDron«aro    T     ®Ifort' i:  necessary, make  a cost
                          .popnate.   In  order  to  begin   this program,  ten

-------
participants, each funding 515,000 a year for two years,  are  required.  Additional
participants will result in an increase in the level of effort  at project initiation.  Tie
cost estimate assumes that engines will be supplied at no charge by the participants
to serve as a particuiate  generators.  All labor, fuels,  traps,  and  iny incidental
expenses are included in the  proposed level of effort.  Attached is a copy of  the
contract addendum for your review and approval. It is essential that the contract be
approved as is unless there is a substantial problem.  In this way,  all participants will
have a common agreement with SwRI in the shortest possible time.

     The project  will require about Z4 months to complete, exclusive of the final
report.  This estimate is based on obtaining the necessary engine and traps from  the
suppliers rapidly.  The  first 1Z to 14 months  are allocated to Task 1 - Trap  Design,
Engine Selection and Baseline  Tests, and Task Z - Trap Durability. The  remainder of
the project (10 to 1Z months)  will be devoted to Task 3 -  Regeneration Techniques,
and one  month for Task 4 - Trap/Regeneration System Demonstration.  Target date
for project  initiation  is about  June 1,  1988.  The  project  can begin earlier, if
approved before that date.

                         V.  PROJECT MANAGEMENT

     Article Xn of  the Consortium Agreement defines the Advisory Committee as
one person from  each  participating organization and one person from SwRI.  The
defined  purpose   of   this   committee   is   to   provide  liaison,    review,   and
recommendations  related to the project.

     One  of  the  first  requirements of the Advisory Committee will be  to  approve
engine selections and assist in their supply to SwRL If no agreement  can  be  reached
easily  on engine  selection, SwRI will make such decision and seek to acquire  the
engine.

     Early in the  project, a detailed draft plan of performance will be prepared  and
mailed  to  the Advisory Committee.  The Advisory  Committee  will be asked to
respond  quickly  with  comments,  suggestions and  recommendations  on the plan.
SwRI  will  attempt  to incorporate as  much of  the  input  from  the Advisory
Committee as possible within time and budget constraints, thereby finalizing  the
plan. It is anticipated that due  to  the wide scope of interests by the participants,
not all ideas for research will be accommodated. In this case, modifications to  the
test plan will be at  the discretion of SwRI.  Final decisions  on research  will be  the
authority and responsibility of S wRL

     The Emissions Research Department of SwRI will be in charge  of this  project,
and the  project manager will  be  Bruce Bykowski.  The project  leader will be Robert
Fanick.  Others lending support to this project will be Charles Hare,  Director of  the
Emissions  Research Department, and Tarry Ullman, Senior Research Engineer.
Several  other  Institute staff  members from  other divisions  will  be  used as
consultants in the area of ceramics and electronic controls.

                                VL SUMMARY

     It  has been a pleasure to prepare this addendum to the SwRI research  activity
pertaining to particulate traps.  We have attempted to address the important points
requested by our past consortium members in  the Dreoaration of the follow-on study.

-------
The proposed activity will truly expand our knowledge  in participate  trapping and
regeneration  by  applying  past  results  to  the  development  of   a  full  size-
trap/regeneration system.  We ire now confident that we are in  a better position to
develop a workable system that can benefit ail the participants.

      Quite frankly, the level of effort to achieve total  success  is unknown because
the degree of required research is unknown.  '.Ve do  feel,  however, that the effort
proposed  will  contribute significantly to the development  of  a particulate  trap
system that incorporates the findings of our past consortium, and should result in a
prototype device. The more participants joining, the better the chances for success.
We encourage  all past consortium participants to support our continued work, and
will actively seek new support.

      As mentioned earlier, the target  date for starting  the project is June 1, 1988,
or earlier.  In order for this date to  be met, it is essential that  we  receive your
response to this addendum and associated contract by May 1, 1988, or earlier.  3y
offering this additional follow-on study as an "addendum1  to  a current project, we
hope  that your company's internal review process will be simplified. Those potential
participants who  were not members  of the  original consortium  will be sent the
original proposal and contract (08-E857) along with this addendum (OS-ZS57A).

      Please feel free to contact Mr. Bruce Bykowski regarding items of a technical
nature, or Ms. Sharon Rowe  for business  and contract  items.   We look forward  to
your  favorable review and early authorization.
 Submitted by:
Approved by:
 Bruce B. Bykowski
 Manager, Advanced Technology
 Department of Emissions Research
Charles T. Hare
Director
Department of Emissions Research
 cc:  Robert Fanick

-------