EPA/AA/CTAB-88-02
                        Technical Report
           Resistively  Heated  Methanol  Dissociator for
                   Engine Cold Start  Assist  -
                          Interim Report
                               by
                      Gregory  K.  Piotrowski
                           March 1988
                             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 developments which may form the basis for a final  EPA
decision, position 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

-------
       UNITED STATES ENVIRONMENTAL PROTECTION AGENCY

                   ANN ARBOR. MICHIGAN  48105
                                                       OFFICE OF
                                                     AIR AND F1ADIATION
May 18, 1988

MEMORANDUM


SUBJECT:   Exemption From  Peer and Administrative  Review
FROM
TO:
        Karl H. Hellman, Chief
        Control Technology and Applications Branch

        Charles L. Gray, Jr., Director
        Emission Control Technoloav Division
     The attached  report entitled, "Resistively  Heated  Methanol
Dissociator  for  Engine  Cold  Start  Assist  -  Interim  Report,"
(EPA/AA/CTAB/88-02)  describes  a  methanol  dissociation  system
that  provides  a  cold  start assist  for  a light-duty  methanol
engine.    Methanol  was   boiled  in   a   steam-heated   vessel,
superheated  and  passed  to  a   dissociator which  made  use  of
resistively  heated  porous   silicon   carbide   technology.    The
product gas  from  this  system was  used  to start and  idle  a  test
engine which had been  cooled to  43°F.

     Since  this report  is  concerned  only with the  presentation
of data and  its analysis and does not  involve  matters  of  policy
or   regulations,   your   concurrence   is   requested  to   waive
administrative  review  according to  the policy outlined  in  your
directive of April 22,  1982.
Concurrence: /  -•-"'
                                                Date
               Charles  L.  Gray ,/J/-,  Dir.,  ECTD
Nonconcurrence:
                                             Date
               Charles  L.  Gray,  Jr.,  Dir.,  ECTD

cc:  E. Burger,  ECTD

-------
                       Table  of Contents
                                                          Page
                                                         Number
I .    Summary	   1
II.  Introduction  	   1
III. Dissociator Operation 	   2
     A.    Boi ler Unit	   2
     B.    Superheater 	   4
     C.    Catalytic Dissociator 	   4
IV.  Supporting Equipment  	   4
V.    Program Design  	   5
VI.  Discussion                  	   5
     A.    Boiler Flow Rate	   5
     B.    Superheater Operation 	   6
     C.    Cold Start Experiments  	   7
     D.    Reactor Power Requirements  	   8
     E.    Emission Levels	12
VII. Conclusions	12
VI11.Future Effort	13
IX.  Acknowledgments	13
X.    References	14

APPENDIX A - Dissociator Element Specifications 	 A-1
APPENDIX B - Hydrogen Content Determination 	 B-1
APPENDIX C - Test Engine Specifications   	 C-1

-------
I.    Summary

     A methanol  dissociation system was  constructed to provide
a  cold  start  assist  for  a methanol-fueled  light-duty engine.
The system  consisted  of  a  steam-heated methanol  boiler,  a gas
superheater,  and  a  platinum-catalyzed  dissociation  element.
The dissociation element  substrate  was  resistively  heated and
constructed primarily from  fibrous silicon carbide.

     The  objective  of  this experimentation  was  to  start  and
idle a  4-cylinder engine on the  product  gas  from this methanol
dissociator.

     H2,  CO,  non-dissociated   methanol,   and  possibly  other
reactor products  from the dissociator  were piped  to the exhaust
gas recirculat ion port of  the  test engine.   This  product  gas
mixture served as the fuel  supply during the start experiments;
the stock  methanol  injectors on  the test engine  were disabled
prior to the start testing.

     The  engine  was  allowed  to  receive product  gas   for  5
seconds   before   cranking   was   attempted.     At  an   engine
temperature of 73°F  the  product  gas mixture was  sufficient to
start and idle the engine.  The engine  was  then  cooled to 43°F;
the product  gas  mixture  was  sufficient to start  the  engine on
the second crank attempt and idle without laboring.

11.  Introduct ion

     Light-duty M100  neat  methanol-fueled  engines are difficult
to  start  and  run in  cold  weather  because of the high boiling
point of  methanol, methanol's  high heat  of  vaporization  (5.5
percent  of  the   heat of  combustion compared   to  less  than  1
percent for  gasoline),  and  the  increased fuel  flow  needed for
methanol  (about   double  that  of  gasoline).   Gasoline-fueled
engines start  with  less  difficulty under  the  same  conditions
partly because of the easily ignitable  light ends of  this  fuel
such  as  butanes,  which  are   vaporized  at   relatively  low
temperatures.

     Some   state-of-the-art  methanol    engines  require   the
addition   of  gasoline   to  the   fuel   to   improve   their
startabiIity. [1]   Other  methanol  engines utilize  separate cold
start systems  relying  on gasoline  or  propane  for  cold  start
assist.[2,3]   Finally,   some researchers  have  suggested   that
stratified-charge combustion  will  produce reliable  cold  starts
of  a neat  methanol-fueled engine  at  relatively  low  ambient
condi t ions.[4]

     Methanol may  be  catalytically  decomposed  to hydrogen  and
carbon monoxide  gases.  Hydrogens'  higher  flame  speed  and lower
boiling point may make it an ideal cold start  fuel.

-------
                               -2-

     The  goal  of  this  project  was  to  construct  a  methanol
dissociator  that  could  provide H2  and CO  fuels  in quantities
that  could  be  used  as  a  cold  start  assist   for  a  methanol
engine.   This  dissociator  would  utilize  resistively  heated
ceramic technology to provide  the  energy necessary to bring the
catalyst   to  operating  temperature   quickly.    A   methanol
dissociator  using  this  technology  has  been  tested  by  EPA for
its  cold  start  assist  potential.[5]   Though   that  particular
dissociator  was  unable  to  provide  enough  dissociated  product
gas  to  start  a  4-cyUnder  engine,   the  resistively  heated
ceramic was able to quickly heat  methanol  vapor to temperatures
necessary  for  the  dissociation  reaction  to  occur.   EPA used
this  resistively  heated  ceramic   together  with equipment  that
would  increase  the  flow  of   methanol   vapor feed  in  order  to
improve the reactor's output and efficiency.

III. Dissoeiator Operation

     Methanol  may  be   dissociated   to   hydrogen  and  carbon
monoxide via the reaction:

     CH3OH< ,  ,	* 2H2 (g , + CO, 9 ,

     The dissociator described below accomplished  this reaction
by  a  three-step  process  (Figure  1).   First,  methanol  was
vaporized  in  a  boiler  and  moved  by  an inert  carrier  gas  to a
superheater.    The   methanol/carrier   gas  mixture   was   then
superheated  to approximately  600°F  and the superheated mixture
passed to a catalyzed and heated  dissociator.   The hot methanol
vapor  was  dissociated  in this final  stage and passed  into  an
engine intake manifold.  A detailed  description  of each part  of
the process  is given below.

     A.    BoiIer Unit

     The  boiler  was  a  5-gallon   capacity,  type  316  stainless
steel  pressure  vessel.   The  vessel   was  fitted  with a  steam
heating coil, pressure  relief  valve,   liquid temperature monitor
and  pressure gauge.   The boiler  was  filled with  approximately
three  gallons  of methanol  prior  to each  test  and  sealed;  no
provision was made  for  adding  fuel to  the  vessel  following the
commencement of  a test.

     Liquid  temperature  and   pressure  inside  the boiler  were
maintained at approximately 190-195°F and 30 psig respectively
during testing.  Fuel  feed rate  was  determined indirectly; the
filled vessel was carefully weighed prior to and  at  the end  of
testing,  and the test process timed.

     The flow of vaporized  methanol was assisted by the passage
of a pressurized carrier gas through  the boiler.   Nitrogen  at  6
standard cubic  feet per hour  was added  to the  boiler  during
testing and  the vessel  was  purged at  the  same  flow  rate  for  10
minutes prior  to the  start of  heating.   This  inert gas  flow
served three functions.

-------
                -3-
             Figure 1

Methanol Dissociation System
Nitrogen


Boiler
    Dissociator
 Superheater
      Flame
     Arrestor
To Engine

-------
                              -4-

     1.     The  N2  acted  as  a  carrier  gas  to  assist  product
flow throughout the system;

     2.     The  inert   N2  provided  an  atmosphere  safe  against
rapid oxidation and fires in the dissociator;  and

     3.     The  absence  of  air  in  the  dissociator  hindered
undesirable,   competing  chemical   reactions  such  as  complete
oxidation of  the methanol feed.

     B.     Superheater

     The superheater was constructed  from  a 3-foot-long section
of 1-inch diameter 304 stainless  steel  pipe.   This pipe section
was  heated  by  wrapping   it  with  a  ceramic-bead-insulated
nichrome  wire   heater  that   utilized   120-volt   alternating
current.  Power to the heater was controlled by a thyristor.

     C.     Catalytic Dissociator

     The hot  methanol  vapor and  N2  carrier gas  passed through
a  porous silicon  carbide  filter which  had been  coated  with
platinum.   This  filter  was   resistively  heated   by  passing
120-volt alternating current  through  it.   This  catalyzed heater
acted as the methanol  dissociator.

     The filter  was  shaped  in  the  form of a hollow cylinder.
Methanol vapor  flowed  radially  through  the porous heater walls;
the dissociation  reaction  occurred as  the vapor contacted the
hot  catalyzed  walls.   Element  specifications  are  given   in
Appendix A.

IV.  Supporting Equipment

     H2   composition  in the product  gas mixture was  determined
by gas chromatography.  A GOW-MAC Model  69-550 gas chromatograph
was used, and its operation is detailed  in Appendix B.

     The test engine used for this experimentation was a Nissan
CA18E,   single-overhead    camshaft,    1.8-liter   displacement
engine.    Details  and  engine   specifications  are  provided   in
Appendix B.

     Exhaust unburned  fuel  (UBF)  emissions were  measured with a
Beckman  Model  400  flame ionization  detector.   A FID response
factor of 76 percent was used.[6]

     Exhaust formaldehyde was measured using a dinitrophenyI-
hydrazine  (DNPH)  technique.[7]   Exhaust  carbonyls  including
formaldehyde  are drawn  through DNPH-coated cartridges  forming
hydrazone derivatives.  These  derivatives  are  separated  from
the   remaining   unreacted   DNPH  by  high  performance  liquid
chromatography    (HPLC).    A    spectrophotometer    in    the
chromatograph   effluent   stream  drives   an   integrator   which
determines formaldehyde derivative concentration.

-------
                               -5-

V.   Program Design

     The  goal  of  this  program  was   to   start   and  idle  a
light-duty  vehicle  engine  using  the  product  gases  from  the
methanol  dissociator  as  a   fuel   supply.    The   program  was
conducted in two phases:

     1.    Construction    of    the    dissociator    and    flow
           measurement; and
     2.    Engine testing of the dissociator.

     The  first  phase  of  the  program   involved  building  the
dissociator  and  measuring  the  methanol  feed  rate  from  the
reservoir.  We  also attempted to  determine  product composition
and  thereby establish  reactor  efficiency.    The  second  phase
involved piping  the  product  to a warmed engine and attempting a
start.   Immediately  prior to  the  attempted  start  the methanol
engine main  injectors  were disabled, in order to limit the fuel
to the product gas supplied from the dissociator.

VI.   Discussion

     A.    Boiler Flow Rate

     The   low   flow  rate   provided  by  the   boiler  in  the
dissociator  previously   tested   [5]  may   have  substantially
contributed to  that dissociator's inability  to start  the test
engine.  The  dissociated  product  flow  rate  to  start  an  engine
similar  in displacement to our test  engine has  been calculated
[8,9]  as .3  to  .6  grams per second  or .45  grams  per second.
The  product  generated  by the  lower flow dissociator,[5]  .013
grams   per   second   H2/CO   and   .053   grams   per   second
non-dissociated  methanol,  was   low when   compared  to   these
calculated requirements.

     A steam-heated  boiler was  used  in  the  system  presented in
this report to  increase  the feed rate to the  dissociator.   The
temperature of  the  liquid  in  the boiler  is a  function  of  the
pressure above  the  liquid charge  in the boiler, the  amount  of
fuel  charge present,  and  the condition of  the steam flowing
through  the heating  coils.   The  flow of vaporized  methanol  out
of  the boiler  is a  function  of boiler  pressure,  temperature,
and  the  flow  rate  of  carrier gas  through  the boiler.   These
variables   were   difficult    to   simultaneously   control.
Measurement of vaporized methanol  flow  rate  was accomplished by
weighing  the  boiler  prior  to testing, obtaining  steady-state
conditions and operating over  a  period of  time,  then reweighing
the  boiler  after the cessation  of testing.    The difference  in
weight was  divided  by  the  time of  operation  to   determine  a
rough flow rate.

-------
                               -6-

     At the  conditions of 30  psig in  the  vapor  dome above the
liquid methanol,  195-200°F  liquid temperature  and  6  SCFH  N2
into  the  boiler,  approximately 1900 grams  per  hour of methanol
feed was delivered  to  the superheater.   This was  a significant
improvement  over  the  methanol  feed   rate  of  the  lower  flow
dissociator mentioned  previously,  and  within the  guidelines  of
the  rates necessary  to  ensure  .3 to  .6  grams  per  second  of
dissociated product gas to the engine.

     A  higher   methanol   flow  rate   could  be   obtained  by
increasing   the   flow   of  N2   carrier  gas   to   the  boiler.
Although  this  technique   increases  the  amount  of   inert  N2 fed
to the engine, a  vaporized methanol flow rate  in excess of 2500
grams per  hour was made  possible  by increasing  N2 flow  to  20
SCFH.  The  use of  helium at  6 SCFH,  rather than  N2 ,  also had
the  effect  of raising the vaporized  methanol   flow  rate  over
2500  grams  per hour.  The use of  these  higher  flow  rates was
not necessary, however, as a rate of 1900 grams per hour proved
sufficient to start and idle the test  engine.

     B.     Superheater  Operation

     The superheater was  capable of heating the  methanol  vapor
(1900  g/hr)  and  N2  carrier  gas   (6   SCFH) stream  to  600°F.
Higher gas  temperatures  were  made possible by  lengthening the
pipe  and   adding  additional   nichrome   wire  heaters   (gas
temperatures  out  of the  superheater   in  excess of  1200°F were
recorded).

     A problem with coking of the methanol  fuel was observed  at
elevated superheater temperatures.  This  coking led to plugging
of the superheater  and coating of  the  catalyst with coke.   The
coking,  or  decomposition  of  the  methanol  to elemental  carbon,
appeared  to  occur  mainly on  the   inner  wall   surfaces  of the
superheater.   Coking   on   the  catalyst   substrate   surface  also
appeared to have  occurred,  although some of the carbon  in the
catalyst can may have flowed through from the superheater.

     Coke   formation   appeared  to  positively   correlate  with
operating   temperature  in  the  superheater  and  dissociator.
Temperatures  in  excess  of 900°F   and  a vaporized methanol/N2
gas  mixture  invariably   led   to  sooting  of  the  superheater.
However,   some carbon  formation  was   noted  at  a   gas  mixture
temperature  of  525°F  and an  outside  wall   temperature  of the
superheater of 700°F.

     Other researchers working with methanol dissociation  have
noted this coking problem.[10,11]   The  optimum  solution  to this
problem appears  to be  the  use of   a catalyst that  promotes the
dissociation  reaction  at   very  low temperatures  (less  than  or
equal to  570°F).   For  the present  effort,  we attempted  to hold
the  gas mixture  temperature below  600°F  at all  points  in the
system.

-------
                               -7-

     C.    Cold Start Experiments

     A  cold  start  of  the  73°F  overnight-soaked  engine  was
attempted using  only the  product  gas  from  the  dissociator as
the  fuel  source.   (Prior  to  testing  the  engine  main   fuel
injectors were disconnected.)   Test  conditions are  given in
Table 1.

                             Table 1

            Test Conditions For 73°F Cold Start Test

            Variable                         Condition
Engine temperature                       73°F (overnight soak)
Boiler liquid temperature                192°F
Boiler pressure                          30 psig
Methanol  flow rate from boiler           1900 grams per hour
Gas temperature out of superheater       532°F
Gas temperature out of dissociator       300°F
Gas temperature into EGR port            149°F
N2 carrier gas flow into boiler          6 SCFH

     Product gas  from  the  dissociator was  fed  to  the  EGR port
for 5  seconds  prior  to a crank attempt.  The engine started on
the first  3-second crank attempt.   It  idled without   laboring;
at  no  time was  there any  indication of stalling.  We allowed
the engine to idle for 1 minute, then shut it off.

     Approximately  12  feet   of  plastic  tubing  connected  the
dissociator  to  the  EGR  port.    A   valve  to  allow  emissions
sampling and a  flame  arrester were  also  located  in  this  line.
These  restrictions,  however,  did not  combine  to  reduce fuel
flow  to   the  point   that   engine  performance   at   idle  was
noticeably affected.  Although  the fuel  entry passageway  to the
combustion  chambers  was  not  standard  (via  EGR  chamber  to
cylinders)  it  proved  sufficient  to   allow  a  start and  idle at
the conditions in Table 1.

     It was very difficult  to calculate a dissociator  yield for
this experiment.   The  Tedlar sample  bags  used  to determine H2
content  of  the  product  gas  also  collected   a  substantial
quantity  of  liquid which  could not  be  passed  through  the gas
chrooatograph.    The  product  that remained   in   the  gas  phase
however  was  analyzed;  the  H2   content   of  the  gas  was  7
percent.   This indicates a  yield of  less than 10  percent.   This
experiment was repeated with  essentially  the  same conditions as
described  in  Table 1.   A  6  percent  H2  concentration in  the
sample bag  was  obtained, and the engine started  and  idled with
no observable difficulty.

-------
                               -8-

     The  experiment  was  then  repeated   at   a   lower  engine
temperature.  The  test  cell  was chilled  and the  engine soaked
to a  temperature  of  43°F.   Lower  temperatures  were  impractical
because of  the  possibility of freezing water  lines  in the test
cell.  The  higher  flow  rate  of methanol was probably  due  to an
improved  quality   of  steam  (less  water)  allowing  for a  more
efficient heat  transfer to  the  methanol.   Attempts  to control
the  boiler  and  superheater  to the same operating  conditions as
in the  previous experiment  were made,  but  the quality of  the
steam  in  the   lines  is a function  of demand  on the building
steam system on a particular day.  The  conditions  for this test
are given in Table 2.

                             Table  2

            Test Conditions For 43°F Cold  Start Test

            Variable                         Condition
Engine temperature                        43°F
Boiler liquid temperature                 193°F
Boiler pressure                           30 psig
Methanol  flow rate from boiler            2600 grams per hour
Gas temperature out of superheater        600°F
Gas temperature out of dissociator        300°F
Gas temperature into EGR port             139°F
N2 carrier gas flow into boiler           6 SCFH

     Product gas was  fed from  the  dissociator to  the  EGR port
for 10  seconds prior  to the  first crank attempt.   The engine
fired but  did not start  on the  first  3-second crank.  On  the
second crank attempt,  the  engine  started  and  idled  smoothly.
We  allowed  the  engine  to   idle  for  30 seconds;   it  did  not
exhibit any  tendency to stall.  We then shut the engine off.

     During  this  experiment  a  considerable quantity  of  liquid
again accumulated  in  the sampling bag.  The gaseous  portion of
the  sample   was   tested   by   gas   chromatograph,  and   a   H2
concentration  of  12  percent was  obtained.    This  indicates  a
dissociator  efficiency yield of less than 15 percent.

     D.    Reactor Power Requirements

     The  gas  temperatures   in  the  superheater and dissociator
were kept below 600°F during the cold  start  testing in order to
prevent  the  coking   mentioned  previously.    However,    it   is
necessary  to  provide  a  significant   amount  of  energy  to  the
methanol   at  the  catalyst  surface  in  order to  further  the
dissociation reaction.

     Karpuk[12] in a private communication to  EPA,  calculated  a
power  requirement  of  2256  watts necessary  to dissociate 2000
g/hr of methanol at 25°C.   Our  calculations  of this requirement
show  it  to  be 2230  watts,  essentially  the  same.   Therefore,

-------
                               -9-

increasing amounts of  power  were applied to a  catalyzed filter
while  flowing 1900  g/hr  vaporized methanol  through  it.   The
objective  of  this  test  was  to  determine  how  power  level
affected  both  gas  temperature  out  of  the  dissociator  and
product yield.

     Gas  temperatures   into  and  out  of  the  dissociator  at
various  power  levels  were  recorded   after  the  reaction  had
reached  steady  state.   Voltage drop  was  measured  across  the
dissociator  while  current  was  measured with  an  ammeter  in
series.   H2  concentration   in  the   product   gas   stream  was
determined at higher power levels.

     The power supply  used  for this experiment was a Power Mate
AC to  DC  converter  allowing  a maximum DC output of  1000 watts.
Resistance across  the ceramic  filter  was measured at  .65 ohms.
Details from this experimentation are given in Table  3.

                             Table 3

       Dissociator Power/Gas Temperature Characterization

                          Gas Temperature     Gas Temperature
Volts    Amps    Watts    Into Dissociator   Out of Dissociator

   7       5       35          550°F                195°F
   9       8       72          550°F                225°F
  11      10      110          550°F                256°F
  13      11      143          540°F                304°F
  14      13      182          540°F               340°F
  15      14      210          540°F                355°F
  16      15      240          540°F                385°F
  17      16      272          546°F               410°F
  18      18      324          547°F               430°F
  20      20      400          547°F               470°F
  22      23      506          570°F               571°F
  23      23      529          570°F               571°F

     Total power dissipated was  limited  to 529  watts.   Attempts
were made, but  failed to increase the current above 23  amps at
23  volts  with  this  equipment.   H2   concentrations  in  the
product  gas  of  6  and  10  percent  were determined   for  power
levels of 400 and  529 watts,  respectively.  Again,  significant
amounts of  liquid  collected  in the sample bags;  this  indicates
that the conversion  rate of methanol  was less than 10  percent
even at the higher power level.

     The 5KVA  power  supply  that Coloroll  provided  to  EPA  was
next utilized  to  increase the  power  to  the dissociator.   Gas
temperature  into  the   dissociator  was  also   increased  over
previous  levels  in  order to  determine whether  this would  aid
the dissociation reaction.  The results  of  this experimentation
are presented in Table 4.

-------
                              -10-

                            Table 4

       Dissociator Power/Gas Temperature Characterization

                          Gas Temperature       Gas Temperature
Vol ts    Amps    Watts    Into Dissociator    Out of Dissociator
                                                     138°F
                                                     150°F
                                                     265°F
                                                     363 °F
                                                     401 °F
                                                     428°F
                                                     440 °F
                                                     460°F
28.40    35.5   1008.2**        755°F                448°F
*    This power level maintained only 60 seconds.
**   This  power  level  maintained  less than  10  seconds;  unit
     failed during this test.

     The highest power  levels attained  here  were not sustained;
the  information  presented in Table 4 was  obtained from several
experiments.   At  gas  temperatures out  of  the dissociator   in
excess  of   450°F  numerous  leaks  were   experienced  in  the
dissociator  housing.   Several  methods  of  improving  the  seams
were tried;  the  concentrated methanol  vapor attacks most gasket
materials   at   temperatures   in   excess   of   450°F,   however.
Overheating  of the electrical  connections was  also  a problem:
several measurement attempts at  higher  power were  halted due  to
melted instrument connections.

     Typically 5  to 10 minutes  was  necessary  for  the reaction
to achieve steady-state status  (constant gas  temperature out  of
the  dissociator)  as  the level  of  power  applied changed   in
increments of  20  watts.   This is due  to  heat  loss  through the
insulated  dissociator  housing   as well   as   from  gas  passing
through the  dissociator.   At the  last  data point  presented  in
Table  4,  1008 watts,  only a 448°F gas temperature out  of the
dissociator  was   measured.     The   dissociator   failed   after
operating  for  less  than  10  seconds at this  power  level.   The
unit therefore never  achieved steady-state operating conditions
at this power  level.

     The  dissociator  failed  shortly after  the application   of
1008 watts.   The  stainless   steel  wool gaskets and  stainless
steel   contacts   were   unaffected;    instead   the   nichrome
flaae-sprayed  ceramic  ends physically  separated from  the  rest
of  the hollow  ceramic cylinder.  This indicates  that  failure
was  due  to a difference   in  thermal   expansion  coefficients
between the silicon carbide and  the nichrome flame spray.

-------
                              -11-

     H2  concentration   in  the  product  gas  was  measured  at
approximately 40 percent at a power  input  of  611 watts.  Again,
a  substantial  portion of  liquid was  present  in  the sample bag,
making  it  difficult to  calculate  a H2  yield.   The  higher gas
temperature   into   the   dissociator,   because   of   greater
superheating  during   this  testing,  may   have  substantially
contributed  to  the higher H2  bag concentrations.  A thin film
of carbon  due to  coking of  the  fuel  was evident,  however,  on
the surface  of  the catalyst when  the  unit  was dismantled  after
testing.

     A sample of  the  liquid which collected   in  the  Tedlar bag
at  611  watts  was  taken  and  analyzed  for  its  distillation
curve.   The  equipment  used   for   this  analysis  was  a   Model
TS-74645 AR-2  Automatic Distillation  Apparatus  (ADA Ml), made
by  the  Precision   Group,  GCA   Corporation.    This  analysis
indicated  that  the  liquid  in  the  sample  bag  was 99  volume
percent methanol and I percent water.

     The data presented here  suggests   that  the  dissociation
reaction may  be more  limited  by  the  amount  of  energy supplied
to the feed  gas  and  catalyst  surface  than  by  the amount  of
catalyst present  on  the substrate.   H2 concentrations  in the
product  gas  increased   slightly,  from  6  to  approximately  10
percent, as  the product  gas  temperature  increased from 300°F to
571°F  respectively.   According  to  an  unpublished  memorandum
from  Nissan,[13]   a fresh Pt  catalyst  could  have  generated  a
H2  concentration  of  approximately  30  percent  in  the   571°F
product  gas,   a   substantial   increase  over   the   level  we
experimentally  obtained.  Increasing  the power  output  to 611
watts  and  the  feed gas  temperature  to  755°F increased  the  H2
concentration  to  40  percent  (cooling of  the feed  gas  in the
reactor  from  755°F to 571°F would make  available an additional
140 watts  for  the dissociation   reaction).   However,  even  at
this  increased  power  level   a  substantial  amount  of  methanol
condensed  in  the  sample  bag,   indicating  a  much   lower  H2
concentration,  had  the bag  contents been entirely  vaporized
prior to chromatographic analysis.

     Compounding  the  problem  of  an   insufficient  amount  of
energy  necessary  to completely  dissociate 2000 grams  per hour
of feed gas is the energy  loss from the  dissociator  due to heat
transfer to  its environment.   Heat losses occur convectively to
the  air   from   the  ceramic-wool   insulation   blanketing  the
dissociator housing and conductive Iy  to the metal  fittings and
polymer  tubing  which  carry  the  product  gas   to  the  engine.
These  losses, though  unquantified  here,  would  lessen the amount
of energy  available  to  initiate  the endothermic  dissociation
reaction.

     The product  yield,  therefore,  increases   as  the rate  of
energy supplied to  the  dissociator is increased, at  conditions
of approximately  600 watts and  a  2000  grams  per  hour  methanol
feed  rate  to the  dissociator.  This  indicates   that  at  these
conditions   the  rate  of  energy  supplied  to  the  dissociator
limits the  product yield.

-------
                              -12-

     E.    Emission Levels

     Levels of  emissions characterized  as  unburned  fuel  (UBF)
and  formaldehyde  were  measured  in  order  to  determine  whether
the  choice  of   fuel   systems   influenced   pollutant  emission
levels.  The test  cycle was a cold start followed by a 5-minute
idle.  UBF and  formaldehyde emissions are expressed  in  average
rates  of  g/min  and  mg/min  respectively;  this  information is
presented in Table  5.   NOx  and  CO emissions  were  not measured
due  to problems  with   the  analyzers.   Each  test  was conducted
following an overnight  soak of the engine at 73°F.

     UBF emissions  for  the  dissociator fuel  are  reported  as if
they are all methanol.   While this neglects the effect that the
CO  and H2  have  on  the  result,   the  nature of  the  exhaust
emission measurement capability  in the test cell  at the time of
the tests make this an  appropriate approach.

     UBF emissions  increased  slightly when  the test  engine was
operated on  fuel  supplied  from  the  dissociator.   This increase
in emissions,  (.56  g/min),  while noted,  is not definitive;  the
engine was  emissions-tested only  once using  the  dissociator as
the  fuel  supply  system.   Formaldehyde  levels appear to  vary
considerably, from an average 1 mg/min using fuel  injector-
supplied methanol  versus no  detectable  concentration when  the
dissociator  supplied   fuel.    Again,   however,   only  limited
testing was conducted.   Additional testing must be conducted to
properly define the difference in emission levels.

                             Table  5

        Emission  Levels Over Cold  Start  and  Idle  (73°F)

                                                         Forma I-
                                                   UBF   dehyde
Engine Fuel System Configuration   No. of Tests   g/min  mg/min

Methanol injectors functioning        2           2.76    1.0
Oissociator fuel system              1           3.32

VII. Conclusions

     1.    The  test engine started  and  idled   at  43°F  using
dissociator  product  gas  as  the  fuel,  under   the  conditions
described  in  Table  2.   It  is  not  possible  to  state to  what
extent   the   limited   amount  of   H2   produced    under   these
conditions would assist a cold start at temperatures below 43°F.

     2.    Coking of the fuel and  subsequent  catalyst poisoning
was  a  problem with the  system at  gas temperatures  greater than
600°.   Some   coking   was   noticed,   however,   at   lower   gas
temperatures into the dissociator.

     3.    At   the  conditions  of   529  watts   applied,   gas
temperature  into  the dissociator  of  570°F  and a methanol  flow
rate  through  the  system of  1900  g/hr,  a  H2 concentration  of

-------
                               -13-
4.    UBF  emission   rates  from   the   test  engine
                                                 '
                                                             were
                                                             "
      a.     Methanol  from the engine fuel injectors; and
      b.     Dissociator  product   gas   with   the   engine   fuel
            injectors disabled.
 n/minro?-°n,  levels  measur^  "»re  2.76  g/m i n and  3.32
 g/mn   respectively.   Formaldehyde  emission  levels  over  this
 rnn^n/r°arti«neeSe  fuel  systems «wre 1.0 mg/m i n  and  no detectable
 «r!m}««?™   fme«.aSUred'  resPect j ve > Y •   Only  a  limited  number
 of   emission   tests  were  conducted,  however;   two  tests  were
 conducted  using  the engine  fuel  injection  system and one  test

 to I Ing "system.     9  ***  dlssocjator  Produc*  9as  only  as  the
 VI 1 1. Future Effort
                  *.that   wi"   more   favorably   promote   the
          on  reaction at temperatures  below 500°F  is  currently
 being sought.  A  lower temperature  reaction would have  the  dual
 advantages   of   requiring   less  activation   energy   and   not
 promoting undesirable reactions such  as the tendency   to  coke.
 A  lower  operating temperature  would also  make  possible a  more
 compact heat exchanger/dissociator design.

 irim.*?f7!ri.  ca!;dldate    cata'yst    configurations   have    been
 identified,   permission   is   being   sought  to   try  these   new
 catalysts on the  resistively heated ceramic  elements described
 here.   The  evaluation  of  these  new  catalysts  will   be   the
 subject of a separate technical report.

 IX.  Acknowledgments

     The methanol dissociator  used   in  this experimentation was
          l ColoroM'. P'c"  a  u"ited  Kingdom  corporation.  The
                                               a"d  Development,
            c*^ afPn:ciates  the   efforts   of   James  Martin,
            Standards Development  and Support Branch,  Emission
      u-     °'°?y  Division.  who greatly  assisted  the  author
with this project.   Jim was  also largely  responsible for  the
5®!? ,°t.   u   engine used  for  this testing.   The  efforts  of
Michael  Murphy,  electrical  engineer,  also of SDSB,  with  several

                                                     t"tln« are
     In   addition,   the  author   appreciates   the  efforts   of
Jennifer  Cnss and Marilyn  Alff of  the  Control  Technology  and
ftfirH yn™,,ranC,h>  E5T5;  Wh°  typed  thls nianuscript,  Ind  J.
Oil lard Murrell, also of CTAB, who drew Figure 1.

-------
                              -14-

X.   References

     1.    "Development  of  Methanol  Lean Burn  System,"  Katoh,
K. Y.  Imamura, and T.  Inoue, SAE Paper 860247, February 1986.

     2.    "Interim  Report  On  Durability Testing  of  Low  Cost
Catalysts  for  Methanol-Fueled  Vehicles,"  Wagner,   R.  and  L.
Landman, EPA/AA/CTAB/TA/84-4, August 1984.

     3.    "Using  Methanol   Fuels  In   Light-Duty   Vehicles,"
Brown,  D.,  F.  Golden, E.  Gons,  R.  Potter,  SAE  Paper 872071,
November 1987.

     4.    "Unassisted  Cold Starts  to  -29°C and  Steady-State
Tests  of  a   Direct-Injection   Stratified-Charge   (DISC)  Engine
Operated On Neat Alcohols," Siewert, R.  and  E. Groff, SAE Paper
872066, November 1987.

     5.    "Evaluation  of  ColorolI  Methanol  Dissociator  For
Cold Start Assist Application," Piotrowski, G.,  EPA/AA/CTAB/
87-08, December 1987.

     6.    FID  Methanol   Response,   Memo,   Edward   A.  Barth,
OAR/OMS/ECTD/TEB, Ann Arbor, Ml, August 1987.

     7.    Formaldehyde Measurement  In Vehicle Exhaust At MVEL,
Memo, Gilkey, R. L., OAR/OMS/EOD, Ann Arbor, Ml,  1981.

     8.    "Engine   Cold  Start  With   Dissociated   Methanol,"
Greiner,   L.    and   E.   Likos,   Proceedings  of    the   Third
International Symposium  on  Alcohol  Fuels Technology,  May 29-31,
1979.

     9.    Dissociated  Methanol Fuel  Requirements  to Start  A
Four-Cylinder   Engine,   Memo   from   Gregory  K.   Piotrowski,
OAR/OMS/ECTD/CTAB, Ann Arbor, Ml, 1986.

     10.   "Study of the  Methanol  Reformed Gas Engine," Hirota,
T., Japan Society of Automotive Engineers Review, March 1981.

     11.   "Dissociated   Methanol   Citation:    Final   Report,"
Finegold,  J.,  G.  Glinsky,  and  G.  Voecks,  SERI/TR-235-2083,
DE85000505, August 1984.

     12.   Private Communication,  Karpuk, M.  E., to  U.S.  EPA,
1987.

     13.   Briefing  to C. Gray, Jr.,  by Nissan   Motor  Company,
March 1987.

     14.   Fundamentals  of  Gas Analysis  by  Gas  Chromatoqraphy,
Thoopson, B., Varian Associates, Inc.,  Palo Alto, CA,  1977.

     15.   Basic Gas Chromatography, McNair, H., Bonelli,  E.,
Consolidated Printers, Berkeley, CA,  1968.

-------
                           APPENDIX A

               DISSOCIATOR ELEMENT SPECIFICATIONS


     The  dissociator  substrate  consists  of  a  highly  porous
ceramic to which electric  current is applied.   The fluid  to  be
heated is passed through  the  void spaces in the material.   Heat
transfer  is  encouraged  by the  very  large surface  area of  the
ceramic  (the material  contains  greater than  80 percent  void
space).

     Specification    ranges   for   certain   properties   of   the
material   used  in  the  dissociator are  given below.   The  exact
specifications for the  dissociator material  are  proprietary  to
the manufacturers  of the  elements,  Coloroll,  pic.,  Havenside,
Boston,  Lincolnshire, U.K.

     	Property	   	Range of Values	

     Power density:                      10-1600 W/cm3

     Normal  range:                        10-300 W/cm3

     Power dissipation:                   0.01-0.75 W/cm2

     Heatup/response time:               Milliseconds

     Heat transfer surface/volume:       400-750 cm2/cm3

     Operating temperature:              Up to 1000°C

     Material density:                   0.1-0.5 gm/cm3

-------
                           APPENDIX B

                 HYDROGEN CONTENT DETERMINATION


Background

     The  basis   for   gas   chromatographic  separation  is  the
distribution  of   a  sample  between  two  phases.   One of  these
phases  is  a  stationary  bed,  and  the  other  is  a   gas  which
percolates  through  the stationary bed.    An  inert carrier  gas
carries   the   components   to  be  separated  through  a  column
containing  the stationary  phase.  The active component of  the
stationary  phase  selectively  retards  the  sample  components
according to  their  distribution  coefficients,  until   they  form
separate  bands  in  the  carrier gas.  These component bands leave
the column  in  the gas  stream and  are recorded  as  a function of
time by a detector.

     If  the stationary phase is  a  solid, this  particular  gas
chromatographic   technique   is    referred   to   as   gas-solid
chromatography.  Common packings  used  are  silica gel, molecular
sieve and charcoal.  Gas-solid  chromatography was  used  in this
experimentation,  and   the  details  of  the procedure  are  given
below.    More  complete  explanations  of   gas  chromatographic
technique are provided by Thompson and McNair.[14,15]

-------
                       APPENDIX  B  (cont'd)
Spec!fI cat ions:

Chromatograph model

Detector:
  Operating principle
  Temperature control
  Carrier gas
  Detector elements
  Noise
  Drift

Inject ion:
  Number of ports
  Control

  Operating temperature

Column Oven:
  Temperature range
  Control

Co Iumn:

Gas flow system;
Thermal conductivity
  bridge control:
Electrical :
  Power requirements
  Circuit breaker

Physical :
Compressed Gas Auxiliaries:
  Zero gas
  Span gas

Output:
  Stripchart recorder
                              GOW-MAC Model  69-550
                              Thermal  conductivity type
                              Ambient  to 300°C
                              N2
                              Four (4) rhenium tungsten elements
                              10-micro volts maximum
                              40-micro volts/hour maximum
                              Two
                              Solid-state,  variable-voltage
                                phase control
                              Ambient to 300°C
                              Ambient to 300°C
                              Solid state time apportioning

                              5' x 1/4" molecular sieve

                              Qua I-co Iumn with dual-inject ion
                                ports and exits

                              Continuous current adjust 50-300
                                mA.  Bridge zero adjust.
                                Attenuator for bridge output, 10
                                posi t ions to 512.
                              105-125 volts, 50/60 H2
                              7 amps

                              Two-section construction.  Upper
                                section houses column oven,
                                detector and vaporizers.  Lower
                                section contains power supply,
                                bridge control circuit and
                                temperature controllers.
                              N2
                              40 percent H2/60 percent N2
                              So I tech model 3318

-------
                           APPENDIX C
                   TEST
Manufacturer
Basic engine designator
Displacement
Cylinder arrangement
Valvetrain
Combusti on chamber
Bore x stroke
Compression ratio
Compression pressure
Fuel control system
EGR
Valve clearance
Idle speed
Eng i ne oil

Fuel
Engine cranking speed
ENGINE SPECIFICATIONS
     Nissan Motor CO.,  LTD.
     CA18E
     1809 cc
     4-cyIinder,  in-Iine
     Single-overhead camshaft
     Semi-spherical,  2-spark plugs
       per cyIinder
     83 mm x 83.6 mm
     11.0
     17.0 kg/square  cm  (350  rpm,  80°C)
     Electronically  controlled  fuel
       inject ion
     EGR not used
     0.30 mm HOT,  intake and exhaust
     700 rpm
     Special formulation supplied by
       Nissan for methanol engine
       operat ion.
     M100 neat  methanol
     240 rpm

-------