SULFATE SAMPLING  DILUTION TUNNEL
        DESIGN AND  VALIDATION

                       by
                  Melvin N. Ingalls
                 FINAL REPORT
                       of
                    Task No. 6
                Contract 68-03-2196
                    Prepared for
           Environmental Protection Agency
      Office of Mobile Source. Air Pollution Control
          Emission Control Technology Division
              Ann Arbor, Michigan 48105
                    June 1977
              SOUTHWEST  RESEARCH  INSTITUTE
              SAN ANTONIO     CORPUS CHRISTI    HOUSTON

-------
        SOUTHWEST RESEARCH INSTITUTE
        Post Office Drawer 28510, 6220 Culebra Road
                San Antonio, Texas 78284
SULFATE SAMPLING DILUTION TUNNEL
        DESIGN  AND VALIDATION
                         by
                   Melvin N. Ingalls
                   FINAL REPORT
                         of
                      Task  INo. 6
                  Contract 68-03-2196

       EPA  Project Officer:  Richard D. Lawrence

                      Prepared for
            Environmental Protection Agency
          Office of Air and Waste Management
       Office of Mobile  Source Air Pollution Control
          Emission Control Technology Division
              Ann Arbor, Michigan 48105

                      June  1977

                          Approved:
                          Karl J. Springer, Director
                          Department of Emissions Research
                          Automotive Research  Division

-------
                                  FOREWORD

      This project was conducted for the U. S. Environmental Protection Agency
by the Department of Emissions Research of Southwest Research Institute (SwRI).
Work was begun on April 26, 1976, and completed on June 25, 1977.  The project
was conducted under Task 6 of Task Order Contract 68-03-2196.  It was identified
within Southwest Research Institute as Project 11-4291-008.

      The EPA Task Officer for this task was Richard D. Lawrence of the Emission
Control Technology Division, EPA, Ann Arbor, Michigan.  Karl J. Springer,
Director, Department of Emissions Research at SwRI served as Project Manager.
The task was under the supervision of Melvin N. Ingalls, Senior Research
Engineer.  The initial tunnel design was done by Terry Ullman, Research Engineer,
and the final drawings were made by Bob Nye.  Although a nurober of Department of
Emissions Research personnel were involved in the project, key personnel in-
cluded Nathan Reeh, technician, who constructed the tunnel;  Tom Jack, senior
technician and Dolores Bynum, laboratory assistant, who assisted in the test
phase; and James Herrington, lead technician, and Shelly Stevens, laboratory
assistant, who performed the various chemical analyses required.

      Several project reviews were held both at SwRI and at the EPA laboratory
in Ann Arbor, Michigan.  The reviews at SwRI were on April 15, 1976, and Novem-
ber 22, 1976.  The reviews at Ann Arbor were on October 1, 1976, and December
10, 1976.  In addition to these reviews, frequent telephone conversations pro-
vided close liaison between EPA and SwRI during the project.
                                      11

-------
                                   ABSTRACT

       This report describes the design features and development testing of
a dilution tunnel for sampling automotive sulfate emissions.  The design of
the dilution tunnel and sampling system was the result of a concensus of
ideas from representatives of the EPA and industry.  The development testing
included a propane traverse to verify exhaust mixing;  testing the critical
flow sampling system;  development and use of a sulfate mist generator to
determine system sulfate recovery;  and evaluation of system losses using
automotive exhaust.

       The propane traverse demonstrated acceptable exhaust gas mixing.  The
original design sample flow rate could not be maintained because the pressure
drop across the sulfate filter caused the critical flow sample system to un-
choke.

       While system losses were acceptable using the acid mist generator, the
system was found to have unacceptable sulfate losses at tunnel temperatures
above 85°C (185°F) when tested with automotive exhaust.  These losses appeared
to be a function tunnel temperature, with losses increasing as tunnel tempera-
ture increased.  The losses were minimized by increasing the dilution air flow
which lowered the tunnel temperature.  Sulfate losses in the connecting tubing
between the car exhaust pipe and sulfate tunnel were found to increase con-
siderably as the tubing length increased.  At the completion of the develop-
ment tests, the sampling system, including the tunnel and sampling and control
console, was delivered to EPA, Ann Arbor.  A complete set of drawings of the
tunnel was furnished with the tunnel.
                                       111

-------
                              TABLE OF CONTENTS
                                                                  Page

FOREWORD                                                           ii

ABSTRACT                                                          iii

LIST OF TABLES                                                     vi

LIST OF FIGURES                                                   vii

SUMMARY                                                          viii

CONCLUSIONS                                                         x

I.    INTRODUCTION                                                   1

      A.  Objective                                                  1
      B.  Approach                                                   1

II.   DESIGN FEATURES                                                3

      A.  Preliminary Design Considerations                          3
      B.  Tunnel Design                                              3
      C.  Sample Probe Design                                        4
      D.  Sampling and Control System                                7
      E.  Design  Modifications  Resulting from Test Program         11

III.  SULFATE MIST GENERATOR EXPERIMENTS                           12

      A.  Development of Mist Generator                            12
      B.  Aerosol Sizing Tests                                     14
      C.  Sulfate Recovery Tests                                   21
      D.  Tunnel Propane Traverse                                  22

IV.   TEST WITH AUTOMOTIVE EXHAUST                                 25

      A.  Equipment and Procedures                                 25
      B.  Initial Probe Loss Tests                                 25
      C.  System Sulfate Losses as a Function of Temperature       28
      D.  Effect of Connecting Tubing Length                       32

V.    SYSTEM MODIFICATIONS                                         34

      A.  Critical Flow Sample Rate Evaluation                     34
      B.  Methods to Reduce Tunnel Temperature                     34

          1.  Cooling Dilution Air                                 38
          2.  Increase Amount of Dilution Air                      38

      C.  Final System Modifications Chosen                        39

LIST OF REFERENCES                                                  41
                                      IV

-------
APPENDICES
      A.  Statement of Work
      B.  Velocity Traverse of 21 cm Diameter Tunnel
      C.  Mist Generator Calibration Data
                                       v

-------
                               LIST OF TABLES


Table                                                              Page

 1    Sulfate Recovery From Mist Injection Tests                     21

 2    Sulfate Probe Loss Tests                                       25

 3    Results of Probe Loss Tests                                    26

 4    Tunnel Temperature Effects Test Schedule                       29

 5    Summary of System Sulfate Losses                               29

 6    Effects on Connecting Tubing Length                            33

 7    Maximum Exhaust Temperature During Dynamometer Operation       38
       for Three Cars

 8    Test  Schedule with  Large  CVS                                   40

 9    Results  from Tests  with Large  CVS                              40
                                       VI

-------
                               LIST OF FIGURES


Figures                                                             Page

   1       Schematic of Sulfate Sample Dilution Tunnel                5

   2       Photographs of Sulfate Dilution Tunnel                     6

   3       Sulfate Sampling Probe Tips                ,                8

   4       Modified 47 mm Filter Holder                               9

   5       Sampling and Control Console                              10

   6       Baird-Atomic Mist Generator                               13

   7       Sulfate Mist Injection Equipment                          15

   8       Andersen Impactor                                         16

   9       Sizing  Tests Equipment Schematic                          18

  10       Modified Mist Generator                                   19

  11       Results of H2SO4 Mist Sizing Tests                        20

  12       Tunnel Traverse During Propane Injection                  23

  13       Percent Probe Loss Versus Sulfate Through Probe           27

  14       Percent 804= on Tunnel Versus Tunnel Temperature           31

  15       Percent SO^~ on Probe Versus Tunnel Temperature            31
  16       Percent 804" Lost on Sulfate Sampling System Versus
            Tunnel Temperature                                       32

  17       Calibration of SwRI Built Critical Flow Nozzle            35

  18       Pressure Drop Across Clean 47 mm Fluoropore Filter        35

  19       Capacity Curve for Cast 2565 Pump                         37
                                      vii

-------
                                  SUMMARY

      A dilution tunnel and sulfate sampling system to be used with both the
positive displacement pump (PDF) and critical flow venturi (CFV) constant
volume sampler  (CVS) emission sampling systems was built from EPA design cri-
teria.  The system uses a 2.7 m  (8.8 ft) long by 21 cm (8.4 in.) inside dia-
meter stainless steel pipe as the dilution tunnel.  Exhaust mixing length in
the tunnel is approximately 1.74 metres  (68.5 in.).

      The sulfate probe is of 2.54 cm (1 in.) stainless steel tubing approxi-
mately 61 cm  (24 in.) long with a removable tip which reduces the probe dia-
meter to that necessary to provide isokinetic sampling.  For the CFV-CVS system,
this removable  tip contains a convergent-divergent section designed to operate
in the choked flow regime.  The probe inlet is centered in the tunnel, with the
holder for the sulfate filter located outside the tunnel,

       An alarm system is furnished to provide an audible and visible alarm when-
 ever the tunnel temperature is outside a specified range.  Another alarm is
 activated whenever the pressure downstream of the probe critical flow nozzle
 increases to the point that the nozzle flow becomes- unchoked.  Controls are also
 provided for the air heaters installed in the dilution air plenum.

      The development testing performed included a propane traverse to deter-
mine the extent of exhaust mixing;  proof testing of the CFV-CVS sampling
system;  development and use of a sulfate mist generator to determine system
sulfate recovery;  and evaluation of system losses using actual automotive
exhaust.  The propane traverse showed that the maximum and minimum propane con-
centrations were within approximately ±5 percent of the average.  This spread
was considered small enough not to affect the accuracy of the sulfate sample.

      A SET-7*test was run with the critical flow nozzle sample probe used with
CFV-CVS units.  The test revealed that the originally planned sample rate of
0.028 m^/min  (1 cfm) resulted in the nozzle unchoking as the pressure drop
across the filter increased due to sulfate loading.  As a result, the sample
flow was changed to 0.014 m3/min  (0.5 cfm) for both the POP and CFV-CVS systems.

      A sulfuric acid mist generator was developed which injected sulfuric acid
into the tunnel at a rate approximating automotive exhaust.  Approximately 80
percent of the mist generated was less than one micron in size.  The sample
filter recovered an average of 105 percent of the acid injected.  Accuracy
limits of the analytical procedures are probably responsible for recoveries
greater than 100 percent.  Improved analytical methods were developed for
future work.  Probe losses were approximately 3 percent at a tunnel temperature
of 55°C (131°F).  While a great deal of progress was made on both the mist
generation and sulfate recovery procedures, more work is needed to perfect both.

      Tests with car exhaust  revealed that probe  and  tunnel  losses  were a func-
tion of tunnel temperature, increasing as temperature increased.  The total
system losses  (excluding the connecting piping) varied from 3.9 percent at
68°C (155°F) to 21.6 percent at 96°C  (205°F).  The type of probe material ap-
peared to make some difference in the amount of sulfate lost on the probe.
However, at 85°C  (185°F) average temperature probe losses were above  25 per-
cent with stainless steel, Teflon and glass probes.  There was also some  indi-

* SET-7  is  the congested freeway driving schedule.


                                    viii

-------
cation that the system would become "seasoned" and that sulfate losses would
decrease with test time.  A test of the critical flow sample probe nozzle
showed higher sulfate losses than the noncritical flow nozzle  (15% vs. <3%)
at about 70°C tunnel temperature.

      For this project,a 1588 kg  (3500 Ib) inertia  weight car with a 4.98
litre  (304 CID) engine was used.  The tunnel temperature during SET-7 tests
run with this car were higher than permissable for acceptable  sulfate losses.
A brief check of exhaust temperatures of several cars revealed that for
many cars, a 9.20 mVmin.  (325 cfm) CVS may not produce low enough tunnel
temperatures for acceptable sulfate losses.  To minimize system sulfate losses
the tunnel temperature should be as low as possible, yet still above the dew-
point of the dilute exhaust.  To do this, either the dilution air tempera-
ture must be lowered or its volume increased.  The choice for this study was
to increase the volume.

      A 14.73 m3/min (520 cfm) CVS was used for a series of tests with the
same test car used in the previous tests.  For a SET-7 test, the average
temperature was reduced to 59°C  (138°F) from about 85°C  (185°F) for a SET-7
test with a 9.20 m3/min (325 cfm) CVS.

       During these tests the sample probe with the critical flow nozzle was
 used.   Probe losses with the critical flow probe averaged 19 percent compared
 with the less than 3 percent for the noncritical flow probe at a CVS flow of
 9.2 m3/min (325 cfm)  and approximately the same average tunnel temperature.
 While time did not permit further investigation of the sulfate losses with the
 critical flow probe,  it is recommended that a thorough investigation of the losses
 with this probe be conducted prior to using this system for routine sulfate test-
 ing.

      Two different lengths of connecting tubing between the car exhaust and
the tunnel were tested for sulfate losses.  The car was run for 25 minutes at
64 km  (40 mph) to generate the exhaust sulfates.  The loss in the 0.61 metre
(2 foot) pipe was approximately 0.6 percent; the loss in the 4.57 metre (15
foot)  pipe was approximately 20 percent.

      The sampling system, including pump, flow meters and controls was fit-
ted into a console together with the alarm system and heater controls.  This
console and the dilution tunnel itself were shipped to EPA, Ann Arbor at the
completion of the test phase of this project.  A complete set of drawings of
the tunnel was also furnished with the system.
                                       IX

-------
                                   CONCLUSIONS

     The following conclusions can be drawn from the testing performed with
the 21 cm diameter dilution tunnel.

     1.  Tunnel and sample probe sulfate losses are a function of tunnel
         temperature, with losses increasing as temperature increases.

     2.  The sulfuric acid mist generator developed during this project can
         be used to determine system sulfate recovery, however the proce-
         dure needs further refinement.

     3.  The tunnel design provides adequate mixing of car exhaust gas
         and dilution air.

     4.  There is considerable sulfate lost when connecting tubing on the
         order of five metres long is used between the car exhaust and the
         dilution tunnel.

     5.  The sample flow rate for the critical flow sample probe nozzle
         designed for use with a CFV-CVS should be at or below 0.014 m3/min
         (0.5 cfm) because of high filter pressure losses.

     6.  Sulfate losses using the critical flow sample probe nozzle are
         higher than losses on the noncritical flow nozzle at the same
         tunnel temperature.

     7.  For many cars, a 9.20 m^/min  (325 cfm) capacity CVS may not pro-
         duce low enough tunnel temperatures for acceptable sulfate losses
         without precooling the dilution air.
                                       x

-------
                              I.  INTRODUCTION
     Past EPA studies showed that cars equipped with catalytic converters emit-
ted more sulfates than noncatalyst cars.^"^'    With the possibility of the
routine testing of prototype cars for sulfate emissions, the need for a stan-
dardized sulfate sampling system became evident.

     In an effort to develop an acceptable standard sulfate sampling system
for use by the EPA Mobile Source Pollution Laboratory at Ann Arbor, Michigan,
and hopefully by the auto industry as well, a meeting was held on December 4,
1975 at the EPA, Ann Arbor, facility with representatives of the EPA, the auto
manufacturers, and Southwest Research Institute in attendence.  From this meet-
ing a general design consensus and criteria for a sulfate sampling system
emerged.

A.  Objective

     The objective of this project was to build and test a suitable sulfate
dilution tunnel and filter collection system based on SwRI's mini-tunnel de-
sign which was used extensively in previous sulfate testing projects for EPA
(2,4)_   The new tunnel was to use design criteria established at the Decem-
ber 4,  1975 meeting at Ann Arbor between EPA and industry.  Once built, the
tunnel was to be used to investigate ways to achieve a calibration with sul-
furic acid mist; determine the effect of exhaust connecting tubing length;
determine the effect of dilution air humidity; and compare this  21 cm (8 inch)
diameter tunnel to  a  46 cm (18  inch)  diameter tunnel.

B.  Approach

     To meet the objectives of the project, a 12 item Statement of Work was
agreed upon by EPA and SwRI.  The complete Statement of Work is included as
Appendix A.  The Statement of Work was later modified to include furnishing
the necessary hardware for a complete sampling system and to add additional
testing to better define the system sulfate losses discovered during the
test program.

     Briefly the plan was to:  (1) build the tunnel to the agreed upon speci-
fications;  (2) develop a sulfuric acid mist injection system;  (3) determine
the sulfate sampling system recovery using the sulfuric acid mist injection
system;    (4) compare the sulfate levels obtained from steady-state and sul-
fate cycle tests on an actual car with sulfate levels obtained using an 18
inch dimater tunnel on the same car and test cycles and;  (5) evaluate the
effects of connecting tubing length between car and tunnel using a 2 foot
long piece of flexible tubing.  At the completion of the test program, a
complete set of drawings would be made incorporating any changes found nec-
essary during the test program.

     During the project the test plan was modified to delete the comparison
of the small tunnel and the 18  inch diameter tunnel and to delete  the tests

*Superscript numbers in parentheses refer to the List of References
 at the end of this report.

-------
to determine the effect of humidity.  In place of these tests, a series of
tests were performed to define the system losses as a function of tunnel
temperature.

-------
                           II.  DESIGN FEATURES

     The design considerations and physical layout of the sulfate tunnel,
probe and sampling and control system are explained in this section of
the report.

A.   Preliminary Design Considerations

     As mentioned in the Introduction, a planning meeting was held in Ann
Arbor between EPA and industry for the purpose of agreeing on a set of
design parameters for a sulfate dilution tunnel sampling system.  The fol-
lowing parameters were agreed upon:

               The dilution tunnel diameter will be approximately 8 inches

               The distance between the exhaust inlet and the probe tip will
               be 8 pipe diameters

               The distance between the probe tip and the downstream end
               of the tunnel will be 2.5 pipe diameters

               The tunnel material will be stainless steel

               A mixing orifice will be located at the plane of the
               exhaust inlet

               The probe will be made of stainless steel tubing

               The filter holder will be outside the tunnel

     The EPA added the following requirements to meet their particular
needs:

               Provide mixing of the exhaust and dilution air

               Be compatible with a CVS flow rate of 325 SCFM at
               typical dilution tunnel operating conditions of
               735 mm HgA and 66°C

               Be compatible with EPA certification test cells

               Sample system to be able to be used with a critical
               flow venturi  (CFV)-CVS and  a positive displacement
               pump  (PDF) -  CVS

B.   Tunnel Design

     From the criteria listed above the basic tunnel system was chosen.
The tunnel was constructed of 8 inch schedule 5 welded stainless steel pipe.
This pipe had an inside diameter of 21.4 cm  (8.41 inches) and a wall thick-
ness of 2.77 mm  (0.109 inches).  This fixed the distance between the exhaust
inlet plane and the probe tip at 170.7 cm  (67.2 inches), and the distance
between the probe tip and the tunnel exit  at 53.3 cm  (21.0 inches).  The

-------
overall length of the tunnel is 2.67 metres  (8.75 feet).  Provision was made
for the exhaust to enter from either side of the tunnel using a 7.6 cm (3
inch) diameter elbow through a 15.2 cm  (6 inch) Tri-Clover  (Ladish Co.)
coupling.  The exhaust discharges into the tunnel in the downstream direction.
A 15.2 cm (6 inch) mixing orifice is located in the tunnel at the plane of
the exhaust discharge.  The diluted exhaust exits the tunnel through a nominal
4 inch stainless steel pipe approximately 11 cm (4.5 inches) long.  A 15.2 cm
(6 inch) Tri-Clover coupling installed in the top rear of the tunnel is used
for the sample probe port.  To better fit the EPA certification test cells and
utilize minimum floor space, the dilution air filter box was placed above the
tunnel.  In addition to the usual filters, the dilution air inlet plenum was
equipped with duct heaters to allow the sulfate tunnel to be pre-heated.   It
also allows the tunnel to maintain a minimum temperature above the condensation
temperature of the mixed stream during a test.  Figure 1 is a schematic of the
tunnel showing some of the more pertinent dimensions.  Figure 2 contains photo-
graphs of the tunnel from various angles.

C.   Sample Probe Design

     The sample probe was made from 2.54 cm  (1 inch) stainless steel tubing
with 1.24 mm  (0.049 inch) wall thickness and is approximately 61 cm  (24 inches)
long.  The probe is fitted with an interchangeable nozzle tip(s) to provide iso-
kinetic sampling.  When installed, the probe tip points directly upstream.  There
is a single 45 degree bend in the probe.  The probe comes out of the tunnel at a
45 degree angle through a 15.2 cm  (6 inch) Tri-Clover coupling.  The filter
holder is located outside the tunnel at the  end of the probe.  The initial sample
flow rate specified by EPA was 0.028 m3/min  (1 cfm).

     Part of the design criteria for the tunnel was that the sulfate sampling
system be able to be used with a CFV-CVS.  The CFV-CVS is in reality a propor-
tional sampler.  Since the flow is controlled by a critical flow venturi, the
volumetric flow is a function of the gas stream pressure and temperature.  The
gaseous emissions sample lines of a CFV-CVS  also have choked flow sections so that
the sample taken is proportional at all times to the total volume flowing through
the CFV-CVS.  Therefore, the sulfate sampling system must also be proportional
at all times to the total volume flowing through the CVS.  To do this, the sulfate
sampling system must also have a critical flow section.  After much discussion of
the best way to accomplish this, it was decided to use a choked convergent-
divergent nozzle at the sample probe inlet.  This arrangement will give sulfate
sample flow proportional to the total CFV-CVS volume flow, providing the tem-
perature difference between the sulfate probe entrance and the CVS venturi is
negligible.

     Using a total pressure traverse previously performed on a similar tunnel,
the velocity at the center of the tunnel was found to be 125 percent of the
bulk velocity.  See Appendix B.  Thus,  for a 325 cfm CVS at 735 mm of mercury
and 66°C,the velocity seen by the probe is 5.89 m/sec  (19.3 ft/sec).  This
fixed the nozzle inlet diameter for 0.028 m3/min  (1 cfm) at 10.1 mm.  Using
standard compressible flow calculations(5),  the nozzle throat was calculated
to be 1.59 mm.

     For smooth transition, the angles  for the converging and diverging section

-------
        to sample
          pump
dilute exhaust
  to CVS
filter holder
                                                                                dilution
                                                                                   air
                                                                            filter
                                                                             pack
                                                                           Heaters
                                                                                                 2.49 metres
                      Figure 1.   Schematic of Sulfate Sample Dilution Tunnel

-------
Figure 2.  Photographs of Sulfate Dilution Tunnel

-------
of the nozzle were set at 15 degrees for each.  The nozzle length was set
so that the entrance was tapered  for approximately 2 nozzle diameters.

     For a positive displacement pump (PDF) CVS the critical flow section
is not used.  Since the testing to be performed at SwRI would be with a PDP-
CVS, a second nozzle was designed without the converging-diverging section.
Sketches of both nozzles are shown in Figure 3.

     At the request of the EPA, the standard 47 millimeter Fluoropore (Milli-
pore Corporation) filterholder was modified with quick connects on each end.
The inlet end was replaced with a new, longer, diverging section, and a Viton "0"
ring to seal the filter.  The male part of a 1 inch quick connect was machined
into the end of the filter holder entrance piece.  The discharge side was
drilled and tapped with 1/2 inch pipe threads and the stem of a 1/2 inch full
flow quick connect installed.  A sketch of the filter holder is shown in Fig-
ure 4.

D.   Sampling and Control System

     The original intent of this project was that SwRI would provide to EPA,
Ann Arbor, only the sulfate tunnel and probe, with the EPA to furnish the
remainder of the sampling system.  During the course of work, the EPA modified
the scope of the project to include a complete sampling and control system.
A picture of the sampling and control console is shown in Figure 5.  The con-
sole contains:   (1)  sample system including pump, flow meters and controls;
(2) alarm system to indicate loss of critical flow and failure to remain with-
in temperature limits;  and  (3) dilution air heater controls.

     The sample system consists of a sample drier tube filled with a molecular
sieve, a Cast Model 2565 vacuum pump, a rotameter for visual indication of
instantaneous flow rate and an American Meter Division Gas Sampling meter for
total sample volume measurement.  An interconnect is provided to allow the
sample pump to be turned on and off from the CVS control panel.

     The alarm system was designed at the request of the EPA to provide an
audible and visible alarm whenever the tunnel temperature is outside a speci-
fied temperature range  (high or low).  Another alarm is activated whenever
the pressure behind the sulfate sample probe critical flow nozzle increases
to the point that the nozzle becomes unchoked.

     The controls for the heaters in the dilution air plenum allow the amount
of heating to be controlled manually or automatically.  Three different levels
of heating are provided under manual operation.  A temperature controller pro-
vides for automatic heater control for a given tunnel temperature.  It should
be recognized that the temperature controller provides control on the minimum
tunnel temperature only.  During a test, if the dilute exhaust temperature
exceeds the tunnel set point temperature, the heaters would shutoff and the
tunnel temperature would be dependent on the car exhaust and ambient air tem-
peratures until the tunnel temperature again fell below the set point.  An
interconnect is also provided to permit the heaters to be operated only when
the CVS blower is on.  This keeps the heater elements from burning-out from
overheating.

-------
2.54 cm
                       Sulfate  Probe  Tip  for CFV-CVS  Systems
                      Sulfate Probe Tip  for  PDP-CVS  Systems
                      Figure 3.  Sulfate Sampling Probe Tips

-------
                                  To Sample Pump
 Modified Discharge side  	
 of Millipore 47 mm filter holder  f
                                                      Swagelok 1/2"  Full  Flow
                                                      Quick Connect  Coupling
Redesigned inlet portion of
filter holder with intergral
quick connect coupling stem
f
al
m


C,.

















•^
r
v
r












\



mm






i






's
j


^N^

	 __ ^ TT-J -l-f-»r» flnlf "D-1 nr*



^^^^^^^< rlosiirp r-ipg ^r
3%£&Z2^?'2&Z^' ~~ Mi -, -i • _ _ , -,
wi±±iport; 4/ nun
Filter Holder

^^^
\
}
\
h )

/

^^^^^ t>wageloK ±" lull flow auick
^jjjiiijjiii~~ connect coupling body



1
"^^-•s,

                                         t
                                 From Sample  Probe
                     Figure 4.  Modified 47 mm Filter Holder
                                       9

-------
gas meter
temperature
heater
switches
pressure
alarms
      i-o ff
switch
valve
n^n
      n
                                 n
•D
                                           Q^~
                   Q-"
                                                           sample flow gas

                                                           meter readout
                                   sample flow
                                   rotameter
                                                           sample flow
                                                           control valve
heater temperature
controller
                                   temperature
                                   alarms
                                                           sample flow drier
                                                           sample pump
                          ampling and Control  Console
                                  10

-------
E.   Design Modifications Resulting From Test Program

     As the result of the tests described in Sections III, IV and V, some
modifications to the sampling system were required.  The CVS volume flow
was increased to 15 m3/min  (530 cfm) with 19.8 m3/min  (700 cfm) as an al-
ternative for the CFV-CVS and the sulfate sample flow decreased to 0.014 m3/min
(0.5 ft3/min).  This required the resizing of the sulfate sample probe tips.
While the internal dimensions changed somewhat, the configuration of each tip
remained the same as that shown in Figure 3.
                                       11

-------
                 III.  SULFATE MIST GENERATOR EXPERIMENTS

      One of the goals of this project was to do preliminary development
and evauation of a sulfate injection method of calibrating the complete
tunnel sampling system similar to the propane test used for the CVS gaseous
emission sampling system.  This section of the report covers the development
of the mist generator and its use in determining system recovery.

A.    Development of Mist Generator

      The  first step in  developing a  sulfate  calibration  system was finding
a mist generator that would produce the required mist size  (98 percent less
than  1 micron) in an amount equivalent to  an  automobile exhaust  (about 7 mg/
min of H2SO4).  The design of the system should also be such that it is rela-
tively easy to determine the amount of sulfuric acid injected into the tunnel.
Several mist  generators  were investigated  including Collison type generators
from  several  manufacturers, Sierra Instruments Spinning Disc Aerosol Generator,
the "Misto-Gen"ultrasonic generator,  and others.  For most of these generators,
it appeared to be difficult to determine accurately how much sulfuric acid
mist  would be generated  for a given test.

      After discussing the problem with Dr. Paul Lamoth   of the EPA (Research
Triangle Park, North Carolina), who was working in the field of atmospheric
monitoring of aerosols and had used several of the mist generators under con-
sideration, it was decided to use the calibration mist generator from a Baird
Atomic atmospheric sulfuric acid analyzer.

      This mist generator was all glass with  short flow passages, so that it
would be easy to recover the residual sulfuric acid from  the generator.  A
drawing of this generator is shown in Figure  6.

      The  generator produces a mist by breaking up droplets of dilute sulfuric
acid  with  compressed air.  The acid is admitted into the  generator through a
capillary  tube attached  to a small funnel.  As the acid drips into the mist
chamber, it hits the air jet from the compressed air nozzle and is broken up
into  a mist.  A piece of glass tubing connects the generator to the tunnel.

      Not  all of the acid entering the chamber left as an aerosol mist.  The
larger particles from the droplet impinged on the side of the chamber and
ran down to the bottom of the chamber.  There was also considerable conden-
sation on  the walls of the mist generator.

      The  original generator was modified  by  placing a stopcock  in the cham-
ber drain  line.  If the  chamber became too full, some of  the residual acid
was drained from it.  However, some liquid was always left in the chamber
so that no mist would escape through  the drain.

      The  proposed test  procedure began with  feeding a given amount of dilute
sulfuric acid of known normality into the  mist generator. Knowing the amount
and normality of the dilute solution, the  amount of  sulfate put  into the gene-
rator could be calculated.  After the test, the residual  acid would be drained
from  the generator and the generator  and discharge tube  rinsed with a  small
amount of  distilled water.  The rinsing would be added to the residual acid
and the total volume measured.  The normality of the mixture would be  deter-
mined by titrating an aliquot of the  mixture  with sodium hydroxide.  From  the
                                      12

-------
                   Dilute sulfuric acid in
        Atomizer
       B.A.-02561
Generator chamber
   B.A. 02573
                              I  I
                              I  I
              Residual dilute
             sulfuric acid out
                                               Compressed air in
                                                    Stopper
                                                         Mist outlet  tube
                                                       Sulfuric  acid  mist  out
                                                     Drain valve
                   Figure 6.  Baird-Atomic Mist Generator
                                      13

-------
volume and normality of the mixture, the amount of sulfate remaining in the
generator could be determined.  Subtracting this amount from the total sul-
fate put into the generator would give the amount of sulfate injected into
the tunnel.

      Preliminary bench tests were run with the generator using water to de-
termine the amount of liquid converted to mist, and to determine the air pres-
sure that gave the highest mist generation rate.  The data from these tests
is contained in Appendix C.  It was found that approximately 12 percent of
the dilute acid feed into the generator was atomized into mist.  The optimum
air pressure was found to be about 138 kilopascals (20 psi).  From these tests,
it was decided to use 100 ml of 0.2 N sulfuric acid as the mist generator feed-
stock.  Figure 7 shows the mist generator set up to inject mist into the tunnel.

B.    Aerosol Sizing Tests

      An Andersen impactor was chosen as the easiest to use and most readily
available device for determining the size distribution of the sulfate aerosol
in the tunnel.  Figure 8 shows the details of the impactor.  The Andersen
Stack Sampling Head consists of a stainless steel case and 1/2" female pipe
fittings at the inlet and discharge.  It is designed to be inserted directly
into the stack.

      The Sampler contains nine jet plates each having a pattern of precision-
drilled orifices.  The nine plates, separated by Teflon "O" rings,  divide the
sample into eight fractions or particle size ranges.   The jets on each plate
are arranged in concentric circles which are offset on each succeeding plate.
The size of the orifices is the same on a given plate, but is smaller for each
succeeding downstream plate.  Therefore, as the sample is drawn through the
sampler at a constant flow rate, the jets of air flowing through any parti-
cular plate direct the particulates toward the collection area on the down-
stream plate directly below the circles of jets on the plate above.  Since
the jet diameters decrease from plate to plate, the velocities increase such
that whenever the velocity imparted to a particle is sufficiently great, its
inertia will overcome the aerodynamic drag of the turning airstream and the
particle will be impacted on the collection surface.   Otherwise, the particle
remains in the airstream and proceeds to the next plate.  Since the particle
deposit areas are directly below the jets, seven of the plates act as both a
jet stage and a collection plate.  Thus, No. 0 plate is only a jet stage and
No. 8 plate is only a collection plate.

      For this project, a backup 0.5 micron Fluoropore filter was placed
on a plate 8 to trap all the remaining particulate.  It should be pointed
out that the impactor is calibrated in terms of unit density (1 g/cc). It
can be readily seen that two objects of the same size, but different den-
sities, will have different inertial characteristics and would behave dif-
ferently in different environments.  By referencing the sizes to unit den-
sity, the particles are measured in what is termed aerodynamic size.  Par-
ticles of the same aerodynamic size behave aerodynamically as though they
were the same size, even though they have different physical sizes and
densities.

      The size distribution was determined by washing each of the nine im-
pactor plates with 20 ml of 60 percent isopropyl alcohol solution and proces-
sing the solution using the Barium Chloranilate  (BCA) procedure employed with
sulfate filters.     The backup filter was processed using the usual BCA pro-
cedure for sulfate filters.
                                      14

-------
Ring stand
                                   100 ml burette
                                         Sulfuric acid mist
                                             generator
                                                13 mm dia. glass tube
                                                              Sulfate
                                                               tunnel
                                   Vehicle exhaust inlet
                                       marmon flange
                                                    200 ml beaker
            Figure 7.  Sulfate Mist Injection Equipment
                                 15

-------
 SAMPLE FLOW
     OUT
                          HOUSING
PLATE 8-
PLATE 0-
                         PLATE HOLDER
                     CONE
   SAMPLE PLOW
       IN
                    ISOKINETIC NOZZLE
  Figure 8.  Andersen Impactor
                16

-------
      The test set-up for the sizing tests using the Andersen impactor is
shown in Figure 9.  The impactor is suspended in the tunnel at the same lo-
cation that is normally occupied by the sulfate probe.  Preliminary tests
showed that while the Baird Atomic mist generator produced a sufficiently
fine mist, it did not produce a sufficient quantity for the desired accu-
racy of size determination from the impactor plates.  Since the size distri-
bution was determined from the amount of sulfate on the plates, the more sul-
fate that could be put through the system, the more accurately the percent
of each size could be determined.  Also, it was desired to have the amount
of sulfate be close to that generated by car exhaust.  The target level was
about 130 mg per test.

      In an effort to increase the quantity of mist, the generator chamber
was lightly heated by wrapping it with electrical heating tape.  While this
increased the amount of aerosol to between 30 and 40 mg per test, this was
not deemed sufficient to represent automotive exhaust.  The mist generator
was then modified by increasing the size of the chamber as shown in Figure
10.  The objective of this modification was to allow more room in the gen-
erator chamber, reducing losses due to impingement of the droplets on the
walls.  The modified generator did produce a greater amount of aerosol.
Although the amount of aerosol was still not quite to the level desired,
it was deemed adequate for development tests and the configuration was used
for the remainder of the project.

      A total of nine sizing experiments were run, six of which were usable—
two tests with the original generator and four with the modified generator.
Each test consisted of placing the impactor in the tunnel and drawing a 22.7 £/
min  (0.8 cfm) sample through the impactor while sulfuric acid mist was being
injected into the tunnel.  The mist generator was fed with 100 ml of 0.2 N
sulfuric acid.  This normally required about 25 to 30 minutes.  The total
sulfuric acid injected into the tunnel varied from approximately 35 mg for
the original generator to approximately 60 to 110 mg for the modified generator.

      The results of the sizing tests are shown in Figure 11.  The amount of
sulfate mist below 0.4 microns  (except for one test) varied from approximately
50 to 72 percent.  The highest value was obtained with the sulfate mist gene-
rator in its original configuration.  The lowest percentage was obtained with
the revised mist generator configuration.  One test  (10/22/76) showed only
25 percent of the mist below 0.4 micron.  The reasons for this are not imme-
diately apparent.  It may have to do with the amount of heating of the gene-
rator, or it may be the result of high residual sulfates on some of the sampler
plates.  It was extremely hard to remove all residual sulfate from the plates.
Sometimes the residual sulfate indication would increase between two successive
cleanings.

      Several things were learned from this series of tests.  The first is
that the Andersen Impactor, though easy to use, has several deficiencies for
this type of aerosol sizing.  It cannot separate particles smaller than 0.36
micron at the flow rate used.  At the outset of this project, available in-
formation indicated that an aerosol mist below 1 micron would be sufficient
for simulating automotive exhaust.  However, during the project, results from
the General Motors sulfate dispersion experiment^  ' became available which
indicated that exhaust sulfate particle size in the atmosphere  was  less
                                      17

-------
 00
Diluted
exhaust
to CVS
                               Rotameter
                             Pump
                                                  Dry
                                                  Gas
                                                 Meter
                          I molecular
                             seive
                                        ndersen
                                        Sampler
                                        inside
                                        tunnel
Vehicle exhaust
  inlet  flange
                                                                                       Dilution air plenum
                                     Figure 9.  Sizing Tests Equipment Schematic
                                                                                             Sulfuric acid
                                                                                             mist generator

-------
Compressed
  air in
                                                  Modified 2000 ml
                                                  Erlenmeyer Flask
                          Figure 10.  Modified Mist Generator
                                            19

-------
TO
8
o
0)
N
U
•H
-P
                             Test Date

                             9/20/76

                             9/22/76

                             10/8/76

                             10/22/76

                             10/25/76

                             10/29/76
                         Open symbols - tests with
                                        original
                                        generator
                         Solid symbols- tests with
                                        modified
                                        generator
                   5    10     20  30   40   50 60  70   80    90
                         Percent of H2SC>4 below size indicated
95
98  99
                                     Figure 11.  Results of H2SO4 Mist Sizing Tests

-------
than 0.36 micron.   In light of this information, the Andersen Impactor has
insufficient sensitivity to be used to size automotive sulfates.  If it is
desired to generate and measure sulfuric acid mist below 0.36 micron range,
then additional development is needed on the sulfate mist generator and a
different method for measuring the mist size is needed.

     This test series also showed that it is extremely hard to rid the im-
pactor plates of all residual sulfates.  The time and care used to reduce
the residual level of the plates for this study would be inappropriate for
routine testing.  For routine testing, to have sufficient sulfates deposited
on the plates to make the usual residual level insignificant, either the amount
of sulfuric acid put into the tunnel must be larger than can be generated by
the mist generator used or the injection time must be extended to several hours.

C.     Sulfate Recovery Tests

       The determination of sulfate recovery was done using the mist generator
and the PDP-CVS sulfate sample probe  and filter hardware.  For these tests,
the tunnel was heated to 53 to 57°C  (128 to 135°F).  The recovery tests are
summarized in Table 1.  The average recovery was 105 percent, with a maximum
recovery of 120.8 percent and a minimum of 90.5 percent.
            TABLE  1.   SULFATE  RECOVERY FROM MIST  INJECTION TESTS
 Test Date

 10/18/76
 10/18/76

 10/19/76

 10/26/76

 10/27/76

 11/3/76
Average
 Tunnel
Temp. °C

   53
   54

   53

   57

   57

   56
Test
No.

 1
 2

 1

 1

 1

 1
    Amount of
H2SO4 Injected, mg

      49.0
      71.1

      76.0

      64.9

      33.1

      60.0*
   Percent H2SO4
Recovered on Filter

       119.0
        94.5

       120.8

       101.6

       103.6

        90.5
                                            Average Recovery   105.0
                                            Std. Dev.           12.5
                                            Coeff. of Var.      11.9
 * sulfate BCA procedure used to determine amount of mist injected


      The average amount of sulfate lost on the probe during the first three
recovery tests was determined by washing the probe after the 10/19/76 test,
(thethird test),with 40 ml of 60 percent IPA and processing the solution using
the BCA sulfate procedure.  The amount of sulfate on the probe(in terms of
yg/£ of sample  flow)was divided by the total sulfate(in yg/& of sample flow)
from the first  three tests to obtain the average percent probe loss.  The
probe loss for  these tests averaged 3.2 percent.
                                      21

-------
       Some  of the spread in the recovery test data is felt to be due to the
 accuracy of the  titration procedure for determining the amount of sulfuric
 acid  injected into the  tunnel.   Due to the small amount of sulfuric acid
 injected into the tunnel, a difference of 0.005 in the determination of the
 normality of either the fresh solution injected or the remaining solution
 (plus washing) could make a difference on the order of 15 percentage points
 in recovery.   The difference of 0.005 in normality translates to approximately
 0.25  ml  of  titrant in the titration procedure.   While the amount of titrant
 can be read to 0.1 ml,  the repeatability is normally around ± 0.2 ml.  To
 overcome this problem,  a procedure was developed to use the BCA method to
 determine the amount of sulfuric acid injected.

     The  procedure is to collect the residual acid and washing, then dilute
to 1000 ml total  solution with distilled water.  Next, a 40 ml aliquot of the
solution  is  diluted to 100 ml with 100 percent isopropyl alcohol.  This solu-
tion is then processed by the sulfate filter BCA procedure using sulfuric acid
standards.

      Sulfuric acid mist recovery tests using the CFV-CVS sampling system
were not  run, since it was felt that the pressing need was for tests with
actual automotive exhaust to investigate the high total sampling system los-
ses reported by other laboratories.^)   Thus work on the mist injection sys-
tem was concluded with the results considered satisfactory for the experiments
conducted.  However, more development of these procedures is necessary before
they could be used as a system  verification test.

D.    Tunnel Propane Traverse

      A traverse of the tunnel was also performed while injecting propane
into the  tunnel.   The results of the traverse are shown in Figure 12.  There
is a slight stratification from top to bottom, but it is  felt that it is not
large enough to affect the accuracy of the sulfate sample.  The minimum and
maximum concentrations were within approximately  ± 5 percent of the average.

-------
                      O  12/2/76  Traverse
to
0)
rti
                800
                700
                600
                500
                400
                300
. 1
' i '
1

;i !

: : C
i-J
n ~j: .
( |
i
i ;
n
!
" ~
t

1
I
i
w


•


>
H -•
!


'
1- -

-
.„.

_-

...








-^-
"'


'

-

^T-
J'J1

!
1

- ,
'





)






-±
.'.I
r r

"7








'"






j.i
!!:


i--








'






1-

— _

_




1


<^
— -
1


.



^t~






i '
:
i
i



• H -' -
i

i .
i


:: ~


•


1 "
:
r~ • . T

1

•)
._. I ..C
'
'

' ' ; j
,

1

•


1
1
,.r

• 1

















-r-


i

i: ;










--



••.H

--r»-






)










,:



:




c;










., . .

V--J






)








; '

1

'

,.
i;



C



•

1 1
;



M-1-

'-
'1
' • 1 •


i ,


) : C
i •

"4- '

i
i :
-'-
| ;
1




-,;H' 7
•: i
'
,

' L'

)



•:


•:^



•*-!--'

i"^






c



;:.


- \
- -
. ,




-
1 1 '

•' •

i'
?


1
.1



.,
-_. r
-
-1
1
— r
1 '

--,,






...C
J:-:
;.]
'


';•


•;-j
-.-.



~i-\ —
J U







' •
i:
ij.-
| ' ! •
iV
tr-—
' 1 j
~r< •


j-
.^
.1
1 ;
'.

t"
',
rjr-
I •



1
ji'f
n •'•*
"HTl"
lif
Ff
*ji-|
Vl"
j i
, - .
i. -
i - .
"jTj j


7: (T
j;
•• i



~^r
^;
r.'
,!,'•


j! !
~ '1
-'-U-
•;-t
i
( •
1 ' • , 1
!• i.
i^1
.
..




C
-1 — -
, 1 1
i
i4!
,t;

T~
Li-f

-^tj-r



i:{-:
:
'+1





j
)
T""
1"




1

-ir
-LJ

i
1
•
•p-
'
'
L .












	

•71


'
'
1
j



4-'


.j. i


:_.!
--
.L,
-4-

'

"+.•- .1
1
..J ' V •'

T j

! i
) Q
, ' 1 : i
[ _ j
:; l.i ; .



- 1- r i ••
= 4 .-

:
• i }
\
1 — i- T -.-,,_- —
i ; 1 , :: •'
' I i

                      0
                   Bottom
                              23456

                                       Inches from bottom of tunnel

                      Figure 12.   Tunnel Traverse During Propane Injection
                                                                                                    Top

-------
                       IV.  TESTS WITH AUTOMOTIVE EXHAUST

     This section  describes the test vehicle, procedure, and  test  results
from sulfate tunnel tests using actual  car  exhaust.

A.   Test Equipment and Procedure

     The tests described in this section were performed using the  21 cm dia-
meter tunnel described in Section  II.   All  tests described  in this  section
used the sample probe  designed for use  with PDP-CVS  systems.  That  is, the
sample probe tip did not have a choked  flow section.  A CVS flow rate of
approximately 9.2  m3/min  (325 cfm)  and  a sample flow rate of  0.023  m3/min
 (0.8 cfm) was used during these tests.

     The test car  used to generate the  exhaust sulfates was a 1975  AMC Hornet
Sportabout with a  4.98 litre  (304  CID)  V8 engine.  This car was tested ex-
tensively for sulfate  emissions under Contract 68-03-2118 and Task  2 of this
project's Task Order Contract.  The car is  equipped  with a  pelleted catalyst
and catalyst air injection system.  It  is known to be a higher than average
sulfate producer.  The fuel used was a  commercially  available unleaded fuel
with the sulfur level  adjusted to  0.04  weight percent using thiophene.

     Two driving schedules were used.   The  first was a 64 kph (40 mph) constant
speed cycle.  This cycle was chosen as  a condition which would produce ade-
quate sulfates, yet would not cause undue strain on  the engine, tires or dyna-
mometer.  The second cycle was the SET-7 sulfate cycle, or  as it is currently
called, the Congested  Freeway Driving Schedule  (CFDS).  This  cycle  is a
variable speed cycle with two stops and an  average speed of 56.0 kph  (34.8 mph).
The test is 23.3 minutes long.  The distance driven  during  the test is approx-
imately 21.7 km  (13.5  miles).

B.   Initial Probe Loss Tests

     While preparations were being made to  begin comparison testing between the
new tunnel and the EPA standard tunnel  using actual  car exhaust, the EPA Pro-
ject Officer requested that some probe  loss tests be run with actual car ex-
haust on the new tunnel.  These tests were  felt necessary since a  General
Motors study had indicated that there were  considerable probe losses using a
nominal 350 CFM CVS and a small  (approximately  21  cm ) tunnel. (8)

     The tests conducted are listed in  Table 2.   The results of the first
series of tests are shown in Table  3 under  Test Series A.   This test series
shows that probe losses apparently  can  be large and  seem to be at least par-
tially dependent on the amount of  sulfate drawn through the probe.   The per-
cent of total sulfate on the probe  from Test Series A as a  function of the
amount of sulfate  recovered (filter plus probe)  is shown in Figure  13.  For
the three SET-7 tests,  it appears  that  as the amount of sulfate put through
the probe increased,  the percent of sulfate on the probe decreased.  Note that
the tunnel temperature range for these  tests was from 60°C to 100°C (140 to
212°F)  with average temperature approximately 85°C (185°F).

     In an effort  to identify if the sulfate on the probe might be  the
                                      24

-------
                         TABLE 2.   SULFATE PROBE LOSS TESTS
       Date
                       Probe
                                   Type of Test
1.   Test Series A

      12/7-8/76    1" s.s.    SET-7

      12/16/76     1" s.s.    SET-7

      12/30/76     1" s.s.    SO2 injection

      1/7-10/77    1" s.s.    SET-7
2.   Test Series B

      1/17/77

      1/19/77

      1/19/77


3.   Test Series C

      1/26/77

      1/27/77

      1/28/77
                   3/8" s.s.  SET-7

                   1/4" glass SET-7

                   1" teflon  SET-7
                   1" teflon  40 mph  (155°F tunnel)

                   1" teflon  40 mph  (185°F tunnel)

                   1" teflon  40 mph  (197°F tunnel)
No. of
Tests
                                                       10

                                                        2

                                                        1

                                                       12
  1

  1
   Emissions Measured



H2SO4 and gaseous (4 tests)

H2SO4 and gaseous (1 test)

H2SO4 and S02

H2SO4 and gaseous (4 tests)
  3     H2S04

  3     H2S04

  3     H2SO4
H2S04
H2S04

H2S04
      Note:   A 9.2 m3/min (325 cfm) PDP-CVS was used for all tests.
                                          25

-------
                   TABLE 3.  RESULTS OF PROBE LOSS TESTS
Avg.
Tunnel
Date Temp°C
Test Series A
12/7-8/76
12/16/76
12/30/76 (S02 inj.)
1/7-10/77
Test Series B
1/17/77 (3/8" s.s.)
1/19/77 (1/4" glass)
1/19/77 (1" Teflon)
Test Series C
1/26/77
1/27/77
1/28/77
85
85
50
85
85
85
85
68
85
92
No. of
Filters
10
2
1
12
3
3
3
1
1
1
Sum of Percent of
SO4= on SO4= on Total Total 804=
Filters, yg Probe, yg* SO^=, yg on Probe
5178 1575 6753 23.3
524 577 1101 52.4
>1 23 	 	
4257 1226 5483 22.3
700 426 1125 37.8
813 341 1154 29.6
1521 542 2063 26.3
1963 69 2032 3.4
2214 221 2435 9.1
1595 679 2274 29.9
Note that the results for each line represent the total aggregate sulfates
from a number of repetative tests.

*From single analysis at end of each set of repetative tests.
                                       26

-------
70
           1000       2000      3000      4000      5000      6000



                      Total Sulfate Passed Through Probe  ~ yg
7000
8000
      Figure 13.   Percent Probe Loss Versus Sulfate Through Probe
                                   27

-------
result of a reaction of the probe wall with sulfur dioxide, (SO2), a test
was run injecting pure SC>2 into the tunnel.  The SC>2 was injected at a
rate approximately equal to exhaust SO2 during SET-7 test.  This test was
run at a tunnel temperature of 50°C  (122°F) which is close to the maximum
that can be obtained with the tunnel heaters without automotive exhaust
being put into the tunnel.  The sulfate found on the filter was equivalent
to the usual background level.  Sulfate above residual levels was found on
the probe.  However, the sulfate level was less than a tenth of the sulfate
levels seen from the two previous probe loss tests run with car exhaust.

     At the time the test was run, the results of the SC>2 injection test
were interpreted to mean that the probe sulfate did not come from a reaction
of the probe walls with SO2-  Later tests with car exhaust showed the probe
losses to be a function of temperature as will be explained below.  While no
tests with car exhaust were run at tunnel temperatures as low as 50°C, the
trend from Figure 15 indicates that probe losses with actual car exhaust
could be the same at 50°C as that seen from the S02 injection test.  Thus it
is not certain how the results of the SC>2 injection test should be interpreted.
Therefore, SO2 injection tests over a range of tunnel temperatures are recom-
mended as a possibility for future investigations.  One way to do this is to
use sulfur   sterile fuel during tests with an actual car and inject SO2 into
the exhaust at the tunnel entrance.

     A second series of tests was run to determine if probe material affected
the probe losses.  This series of tests is listed in Tables 2 and 3 as Test
Series B.  In this series, three consecutive SET-7 tests using the 1975 Hornet
were run with each of three types of probes.  These tests indicate that while
probe material did have an effect on probe losses, substantial losses can
occur using stainless steel, glass and Teflon probes at tunnel temperatures
of 85°C  (185°F).  The 9.5 mm  (3/8 inch) stainless steel probe had the highest
losses, 37.8 percent, and the 2.54 cm  (1 inch) Teflon lined probe the smallest
losses, 26.3 percent.

     Three tests were run at different tunnel temperatures to investigate the
effect of temperature on probe losses.  The test series is listed in Tables 2
and 3 as Test Series C.  Each test consisted of a  64 ]qph steady state run
for one-half hour using the 1975 Hornet.  These tests show that tunnel tem-
perature does affect probe losses, with the probe losses increasing from 3.4
percent at 68°C  (155°F) to 29.9 percent at 92°C  (197°F).

C.   System Sulfate Losses as a Function of Temperature

     The initial series of tests showed that substantial probe losses can
occur with stainless steel, glass and Teflon probes and that the probe losses
were a function of the amount of sulfate passed through the probe and the tun-
nel temperature.  Since there was such a large difference in the probe loss
between 85°C  (185°F) and 92°C  (197°F), it was  felt by both EPA and SwRI that
further testing was necessary to verify the shape of the probe loss curve with
temperature.  This test series is listed in Table  4.

     A series of six tests were run  using  the  1975 Hornet as a source of exhaust
                                     28

-------
              TABLE  4.  TUNNEL TEMPERATURE EFFECTS TEST SCHEDULE

              Tunnel
Test
1
2
3
4
5
6
Temp. , °C
68
77
85
88
91
96
Test Condition
1/2
1/2
1/2
1/2
1/2
1/2
hr
hr
hr
hr
hr
hr
at
at
at
at
at
at
40
40
40
40
40
40
mph
mph
mph
mph
mph
mph
Probe
1" s.s.
1" s.s.
1" s.s.
1" s.s.
1" s.s.
1" s.s.
Emissions
S04,
S04,
S04,
S04,
S04,
S04,
S02,
S02,
S02,
S02,
S02,
S02,
Measured
gaseous
gaseous
gaseous
gaseous
gaseous
gaseous
 Note:   Both the probe and the tunnel were washed
        with 60 percent IPA following each test.
                 TABLE 5.  SUMMARY OF SYSTEM SULFATE LOSSES
Sulfate Losses,

Test
1
2
3
4
5
6
Filter
S04=
jig/m?
3052.61
3002.47
2496.82
2673.02
2843.57
2457.98

Probe
59.32
72.03
175.85
264.12
271.89
548.38

Tunnel
52.36*
57.66
67.97
101.69
116.14
129.84
ug/m3 @

Elbow
12.08
9.04
5.12
5.76
8.16
6.71
9.20 m/min. CVS Flow Rate
0.46 m.
Flex
pipe
28.60
29.59
39.09
29.03
31.78
30.90
Tunnel
+ elbow
+ flex
92.94
96.29
112.18
136.48
156.08
167.45

Total
losses
152.26
168.32
288.03
400.60
427.97
715.83
Total
Sulfates
jug/m
3204.87
3170.79
2784.96
3073.80
3271.54
3173.73
Tunnel
Temp
°C
68
77
85
88
91
96
                          Percent of Total Sulfate
1
2
3
4
5
6
95.39
94.67
89.65
86.97
86.92
77.45
1.85
2.27
6.31
8.59
8.31
17.28
1.63
1.82
2.44
3.31
3.55
4.09
0.38
0.29
0.18
0.19
0.25
0.21
0.79
0.93
1.40
0.94
0.97
0.97
2.80
3.04
4.02
4.44
4.77
5.27
4.65
5.31
10.33
13.03
13.08
22.55
                                                                                68
                                                                                77
                                                                                85
                                                                                88
                                                                                91
                                                                                96
*estimate from curve of tunnel temperature vs sulfate loss
                                     29

-------
sulfate to determine the sulfate losses in the sampling system with changes
in tunnel temperature.  A Teflon lined probe was used during these tests.
Each test consisted of a 25 minute run at 40 mph.  An initial acceleration
from idle to 40 mph was part of each test.  At the conclusion of the test,
the sulfate level in the sample probe and in all system components upstream
of the probe was determined.

     Table 5 lists the sulfate on the filters together with the sulfate found
on each of the elements of the sample system.  In the top half of the table,
the sulfate levels are expressed in micrograms per actual cubic meter of
dilute exhaust flow.  The bottom part of the table shows the sulfate levels
as a percent of the total sulfate.  As can be seen from the table, the probe
and tunnel losses increase as the tunnel temperature increases.  The flexible
tubing connector losses measured for each test  are constant, as would be expected,
at about 1 percent of the total sulfate.   Figures 14 and 15 show the percent probe
and tunnel losses as a function of tunnel temperature.  Figure 16 shows the total
sulfate losses (probe, tunnel, elbow, and flex) as a function of tunnel tempera-
ture.

     It should be noted that these results are from one set of tests on a
particular car.  It may be that the sampling system losses are also a function
of the sulfate concentration.  Since the car used on this project has much
higher than average sulfate emissions, the sulfate losses shown here may not
be typical.

     Remember also that the probe loss tests covered in the previous
section seem to indicate that a stainless steel probe would give somewhat
higher losses than a Teflon probe.  Those tests also indicated that there is
a "seasoning" effect on the sampling system and that the losses could go down
considerably after a certain number of tests if the system was not washed
after every test.  Thus while a considerable amount has been learned about
the sulfate sampling system losses, there are still many areas where more
information is needed.

D.   Effect of Connecting Tubing Length

     The various laboratories engaged in automotive sulfate sampling use dif-
ferent lengths of flexible exhaust tubing to connect the car exhaust system
to the sulfate tunnel.  One of the items of investigation in this study was to
evaluate the effects of this connecting tubing length on sulfate losses. Two
different lengths of 10.2 cm  (4 inch) diameter tubing were used, a 0.61 m
(2 foot)  length and a 4.6 m (15 foot) length.

     Several tests,  each 25 minutes in length, at 64 kph (40 mph) steady state
were run with each length of connecting tubing.  The test results are sum-
marized in Table 6.   For these tests, probe and tunnel losses taken from
Figures 14 and 15 were used together with the sample filter sulfate value from
each test to obtain the total sulfate level.

     As can be seen from the table, the sulfate losses in the connecting tubing
increased considerably as the tubing length increased.  For the 0.61 m  (2 foot)
length, the loss was 0.6 percent and for the 4.6 m  (15 foot) length, the loss
was 20.4 percent.
                                     30

-------
 0)
 c
 c
 3
 EH


 0
-P
c
0)
u
    •I  —	
                60
            70       80        90



            Tunnel Temp. Deg.  C
                                                       100
110
 Figure 14.  Percent SO4~ on Tunnel Versus Tunnel Temperature



    20
0)
ft
o
 c
 o

-------
6
0)
-p
CO
H1
-H
 O

 -P
 10
O
CO
 id
 4J
 O
 O

 -P
 £
 0)
 O
                   60
70       80        90


Tunnel Temp. Deg.  C
                                                         100
                                      110
        Figure 16.  Percent  SO4~  Lost on Sulfate Sampling System

                        Versus  Tunnel Temperature
                                    32

-------
                    TABLE 6.  EFFECTS OF CONNECTING TUBING LENGTH
  Test
  Date

3/31/77
4/13/77
3/31/77
4/13/77
4/15/77
               Vehicle
  Flexible     Exhaust
 Pipe Length   Temp °C

0.61 m (2 ft)    263
0.61 m (2 ft)
4.6 m (15 ft)    254
4.6 m (15 ft)
4.6 m (15 ft)
 Exhaust
temp into
Tunnel °C

   224
163
Tunnel
Temp °C

  66.7
  66.7
               49.4
               53.9
               66.7*
 Percent of
total 804 on
  flex pipe

    0.63
    0.51
                                                           Avg 0.57
              20.0
              23.4
              17.8
          Avg 20.4
                                    Total
                                     S04
                                    mg/km

                                    37.59
                                    40.15
                42.54
                38.20
                41.87
*Tunnel dilution air preheated to maintain tunnel temperature
                                          33

-------
                            V.  SYSTEM MODIFICATIONS
     As a result of the testing explained in the previous sections of this
report as well as some additional tests reported in this section, some system
modifications were required.  This section explains the rationale for those
changes.

A.   Critical Flow Sample Rate Evaluation

     A series of tests were conducted to check out the components of the sulfate
sampling system designed for use with a CFV-CVS.  Recall that a critical flow
section in the sampling system is required for the CFV-CVS and that the probe
entrance nozzle was chosen as the location for this critical flow (choked) section.
For choked flow the static pressure at the nozzle throat must be 0.528 of the
nozzle inlet static pressure.     Although some pressure is recovered in the di-
verging section of the nozzle, to insure choked flow it was planned to keep the
pressure downstream of the nozzle at approximately 0.5 of the nozzle inlet pressure.
Since the pressure drop across the filter is large, it was recognized from the
start that a sample pump with high vacuum capability would be needed.

     Prior to obtaining a pump, the nozzle was calibrated using ambient air.
The calibration curve is shown in Figure 17.  The pressure loss curve for a
clean 0.5 micron Fluoropore filter was also determined and is presented in
Figure 18.  If a tunnel pressure of approximately 735 mm of mercury is assumed,
then from Figure 17 and 18 it can be determined that a sample pump capable of
0.028 standard m /min (1 scfm) flow at 636 mm  (25.0 inches) of mercury vacuum
is required.  After a considerable search, a pump was located with this capability.
The pump is a Cask Model 2565.  The manufacturers published flow ratings are shown
in curve form in Figure 19.

     A SET-7 test was run with the CFV-CFS sample probe using a car known to
be a high sulfate producer  (on the order of 50 mg/km).  As the filter loaded
during the test, its pressure losses increased.  At approximately 1100 sec-
onds into the test, the pressure loss became sufficient that the nozzle un-
choked.  The flow continued to drop for the remainder of the test.  From this
test it was concluded that 0.028 m /min  (1.0 cfm) sample rate was too large
to maintain choked flow with a 47 mm, 0.5 Urn Fluoropore filter.  After discussing
the problem with the Project Officer, it was decided to set the sample flow at
0.014 m /min (0.5 cfm).
        .01
B.   Methods to Reduce Tunnel Temperature

     The tests discussed in the previous section demonstrated that sulfate
losses on the probe and tunnel increase as the tunnel temperature increased.
This section briefly examines two methods of lowering the  sulfate tunnel
temperature and presents the modification made to arrive at the final tun-
nel configuration.  However, since the elevated tunnel temperature is caused
by the hot vehicle exhaust gas, it was first necessary to  determine what  tem-
perature should be used as exhaust gas temperature  in a system design.
                                        34

-------
 n
 •O
 C
 (B

 U
 CN

 +J
 id
 3
 C
 •H
n
 tc
 cu
 i
 (0
 3
 +J
FM


-------
                                        least squares straight line

                                        Y = 0.00110 + (7.8307 x  10~5)(X)

                                        r2 = 0.997
•d
G
a

u
-p
id

c

I
n
6
8
     0.030
     0.025
     0.020
     0.015
     0.010
     0.005
     0.000
                i—r
                      ' ' L." '-


                     3^
                                       &"
                                                          ^
                                                      *r.-".

                                                                     Snp7
                                                                       fe
                                    ••(.-
                             t	i	L
                                        - -I—f-
                                                                          ••I'--'
                                                                                r
            50
                      100
                                   150
                                                200
                                                            250
                                                                        300
                                                                                   350
                                  Differential pressure,  mm Hg

                                     (p upstream  = 743 mm HgA)
          Figure  18.   Pressure Drop Across Clean 47 mm FlUoropore Filter


                                           36

-------
   0.6
    0.5
    0.4
•d
    0.2
    0.1

                                  :1^

                                            ;-,l,.--l
                 100
200
300
400
                                                        500
                                      600
                                       700
800
                                   Inlet Vacuum, mm Hg  gage
                    Figure  19.   Capacity Curve for Cast 2565 Pump
                                           37

-------
     The exhaust gas temperature of three different  catalyst  cars were meas-
ured during dynamometer operation.  Table 7  shows  the maximum exhaust tem-
perature obtained on these cars.

     While this is certainly not a large sample, it  does  show that exhaust
temperatures on the order of 380°C  (716°F) are possible during the SET-7.
It is probable that other vehicles could have higher exhaust  temperatures.
Nevertheless, for this analysis, 380°C was used for  a design  exhaust gas
temperature.
                 TABLE 7.   MAXIMUM EXHAUST TEMPERATURE DURING
                     DYNAMOMETER OPERATION FOR THREE CARS
Car
1
1
2
3
Mfgr.
Ford, LTD
Ford, LTD
Chevrolet
Hornet
Engine
Size, liters
5.75
5.75
5.74
4.98
Inertia, kg
2268
2268
2041
1587
Cycle
Type
FTP
FET
SET-7
SET-7
                                                             Max.  Exhaust
                                                              Temp.  °C

                                                                285

                                                                350

                                                                378

                                                                356
      1.   Cooling Dilution Air

          One  method for lowering the tunnel temperature  is  to  cool  the dilu-
 tion  air.   For a simple air conditioning system,  the  cooling limit  for the
 dilution  air  should be about 4°C (39°F)  to prevent ice formation  on the coils.
 Assuming  24°C (75°F)  room air,  this is about a 20°C (36°F)  drop in  dilution
 air temperature.   Cooling the dilution air to 4°C results  in a tunnel tem-
 perature  drop of about 14°C (26°F)  using a nominal 9.20  m3/min (325 CFM) CVS
 with  a blower inlet cooled to 43°C (110°F)  and a  5.74 liter (350  CID) engine
 with  exhaust  temperature of 380°C (716°F).   If it is  desired to keep the tun-
 nel below 77°C (170°F), this amount of cooling would  not be sufficient for a
 5.74  liter engine in a 2041 kg car -

          With a more complex cooling  system,  the  air  could  be cooled to lower
temperatures.  Using the  example  5.74  liter car from above,  a dilution air
temperature of about -35°C  (-32°F) is  required to  lower  the  peak  tunnel tem-
perature to 77°C  (170°F).  This  low temperature could  also  cause problems with
 condensation.


           The above examples assume a CVS with a  heat exchanger to  maintain the
 blower inlet  temperature at about 43°C (110°F).   If a CFV-CVS  without a heat
 exchanger is  used,  more cooling would be required since  as  the tunnel tempera-
 ture  increases,  mass flow through the CFV-CVS decreases, causing  the tunnel
 temperature to rise.   Also, sophisticated temperature control  is  required  due
 to the rapidly changing exhaust temperature and flow  rate.
                                       38.

-------
     2.  Increase Amount of Dilution Air

         Another method for lowering the tunnel temperature is to increase the
CVS capacity, thus increasing the dilution air flow.  Using the 5.74 liter car
from the example above, a CVS with approximately 17.0 m^/min  (600 CFM) capacity
would be required to lower the peak tunnel temperature to 77°C  (170°F) .  A CFV-
CVS without a heat exchanger would have to be somewhat larger, on the  order of
18.5 m-^/min  (650 CFM) for reasons explained above.
 C.    Final System Modification Chosen

      After considering all the factors involved, it was decided by the EPA
 Task Officer that the small (21 cm)  tunnel should be used together with a
 larger CVS.   It was requested that the sampling system nozzle be sized for
 525  cfm CVS flow and 0.5 cfm sample flow.   The new nozzles were made and a
 series of three tests run on the completed sulfate sampling system with the
 critical flow nozzle and a 520 CFM CVS.   The tests run are described in Table
 8.   The results are shown in Table 9.

      From the Table 9, it is evident that the increased flow rate reduced
 the  sulfate losses on the tunnel walls.   The sulfate losses in the flexible
 connecting tubing were slightly higher than seen in previous tests with lower
 CVS  flow, while the losses in the exhaust entrance elbow were about the same.
 The  sulfate losses in critical flow sample probe were considerably higher than
 seen in tests with the noncritical flow probe and lower CVS flows .

      The probe losses of 15 and 22 percent from the tests 1 and 2, with average
 temperatures of 69°C  (156°F) and 74°C (165°F) respectively, are higher than
 desired for routine sulfate testing.  For the noncritical flow probe at tem-
 peratures below 77°C, the probe losses were less than 3 percent.  While time
 did  not permit further investigation of the sulfate losses with the critical
 flow probe, it is recommended that a thorough investigation of the losses
 with this probe be conducted prior to using this system for routine sulfate
 testing.
                                         39

-------
                       TABLE 8.  TEST SCHEDULE WITH LARGE CVS
Test No.
Test Type
Duration (min)
   1              64 kph steady                30
   2              SET-7                        23,
   3              SET-7                        23,
       For  All  Tests
Notes: 1.  CVS flow 14.7 m^/min  (520 cfm)
       2.  critical flow nozzle used on probe
       3.  sample flow = 0.014 m3/min  (0.5 cfm)
   heater condition

       Max heat
       Max heat
Min. temp controlled
 to 43°C  (110°F)
                    TABLE 9.  RESULTS FROM TESTS WITH LARGE CVS
                                               Sulfate Losses  yg/m"

Test
No.
1
2
3

Test
Type
64 kph
SET-7
SET-7
Average
Tunnel
Temp. °C
69
74
59
Filter
Sulfate
yg/m3
1624.82
2077.29
1902.15

CFV
Probe
307.80
615.45
125.67

Tunnel
10.76
9.60
3.64
Percent of
Test
No.
1
2
3
Test
Type
64 kph
SET-7
SET-7
Tunnel
Temp. °C
69
74
59

Filter
81.16
74.85
91,14
CFV
Probe
15,38
22.18
6.02

Tunnel
0.54
0.35
0.17

Elbow
7.20
12.86
12.43
Total Sulfate

Elbow
0.36
0.46
0.60

Flex
Pipe
51.23
60.13
43.15

Flex
Pipe
2.56
2.17
2,07

Total
Losses
376.99
698.04
184.89

Total
Losses
18.84
25.16
8.86
Total
Sul fates
yg/m3
2001.81
2775.33
2087.04





-------
                             LIST OF REFERENCES
1.    Bradow, R. L. and Moran, J. B., "Sulfate Emissions from Catalyst
     Cars, a Review." SAE Paper No. 750090, February 1975.

2.    Ingalls, M. N. and Springer, K. J., "Measurement of Sulfate and
     Sulfur Dioxide in Automotive Exhaust."  EPA Report No.  EPA-460/3-
     76-015, August 1976.

3.    Somers,J. H., et al, "Automotive Sulfate Emissions - A Baseline
     Study."  SAE Paper No. 770166, February 1977.

4.    Dietzmann, H. E., "Protocol to Characterize Gaseous Emissions as
     a Function of Fuel and Additives Composition." EPA Report No.  EPA-
     600/2-75-048, September, 1975.

5.    Shapiro, A. H., "The Dynamics and Thermodynamics of Compressible
     Fluid Flow."  Vol. 1, Chapter 4, The Ronald Press Co., N.Y., N.Y.,
     1953.

6.    Bradow, R. L. , Carpenter, D. A., Klosterman, D., Black, F. M., and
     Tejada, S., "Sulfate Emissions from Catalyst and Non-Catalyst Cars."
     SAE Paper No. 740528, October, 1974.

7.    Wilson, W. E., et al, "General Motors Sulfate Dispersion Experiment:
     Summary of EPA Measurements." Journal of the Air Pollution Control
     Association, Vol. 27, No. 1, January, 1977.

8.    Letter from C.J. Elder, Environmental Activites Staff, General Motors
     Corporation,  to Richard Lawrence, Emission Control Technology Division,
     EPA, Ann Arbor, Michigan, dated October 26, 1976.
                                       41

-------
   APPENDIX A





STATEMENT OF WORK

-------
                                                       EXHIBIT A
                                                       REVISED 2-13-76
                               STATEMENT OF WORK
     The purpose of this task is to provide EPA with a dilution tunnel
and sampling system compatible with the critical flow venturi CVS units
used in certification testing.

     1.  Design and build a dilution tunnel and sampling system which
          will meet the following requirements:

          (a)  Provide mixing of the exhaust and dilution air.
          (b)  Be compatible with a CVS flow rate of 325 SCFM.
               A typical dilution tunnel operating conditions will be
               735 mm Hg. A and 150F.
          (c)  Be compatible with EPA certification test cells, as
               determined by discussions with, project officer or other
               persons designated by project officer.

     2.  Design parameters such as dilution tunnel length, diameter,
material and configuration; mixing orifice diameter; probe location,
material and size; and sampling train components and configuration will
be determined by discussion with the task officer and others, as appropriate.
These parameters should be chosen within the first 10 days of the task
period.

     3.  Two planning meetings have been held in Ann Arbor and the
following parameters have been tentatively agreed upon:
                          £
          (a)  The dilution tunnel diameter will be approximately^
               8 inches,
          (b)  The distance between the exhaust inlet and the probe
               tip will be 8 pipe diameters.
          (c)  The distance between the probe tip and the downstream end
               of the tunnel will be two pipe diameters.
          (d)  The tunnel material will be stainless steel.
          (e)  A mixing orifice will be located at the plane of the
               exhaust inlet«
          (f)  The probe will be made of stainless steel tubing.
          (g)  The filter holder will be outside the tunnel

     4.  Design and build a sampling system compatible with critical
flow venturi CVS units.  Flow velocity into the probe shall be within +
10% of  the velocity of the dilute exhaust mixture adjacent to the probe
at all  times.  Sampling through a critical flow venturi or other means
such as electronic control can be used to accomplish the desired flow
control.

     5.  Design and build an apparatus which will inject a known amount
of sulfuric  acid aerosol into the-..dilution tunnel.  Dilution tunnel and
analytical system verification can be performed using this apparatus in
much the same way a propane injection is currently used for CVS

                                    A-2

-------
                                        2
                                                                 of
  .ification.  The size of  the aerosol  should be measured and 98%.should
 * less than one micron             aerodynamic diameter.  It is recog-
nized that the development  work specified  in this  item will not necessarily
result in a good sulfuric acid mass injection system.  A "best effort"
Ls expected.   Problem areas should be identified so  that subsequent
investigations can address  them.

     6.  Validate the tunnel and sampling  system by  running the following
tests:

          (a)  Sulfuric acid injection  using the apparatus developed
               in 5, above, if possible.
          (b)  Continuously inject propane while performing radial
               traverses with a probe in the vicinity of the sulfate
               probe location*  Record  HC  concentration continuously.
               Traverses in 2 radial directions should be performed.
               Plot: HC concentration versus radial distance.  Determine
               the difference between the  average  observed HC concen-
               tration and  the expected value.
          ,         ^
     7.* Compare the new tunnel at approximately 300 SCFM with the 18"
tunnel (on hand at SwRI) at 475 to 600  SCFM by:

          (a)  Alternating  vehicle exhaust from a  vehicle at .s_teady_
               jrrijise_conditions between the two systems.  Approximately
               20 samples  (H2 S04 + S02 +  gaseous  emissions) will be
               taken in each configuration.  Since a CFV-CVS is not
               available at SwRIf a standard PDP-CVS can be used with
               the new tunnel, however  the new CFV compatible sample
               system should be used for these steady state tests.
               During the  300 CFM tests both the CFV sampling system and
               a standard  probe should  be  operated in parallel.   Sulfuric
               acid level  and probe losses determined with each system
               can then be  compared,,

          (b)  Running consecutive suLfate_cy.c\e?f  alternating the
               vehicle exhaust from one system to  the other. Approximately
               six cycle*with each tunnel  will be  required.  A standard
               probe and sampling system may be required i-f a PDP-CVS is
               used instead of a CFVCVS.

Determine probe and funnel  losses with  each system.  A single determination
for all tests run in each  tunnel infeach of parts (a) and (b) above is
sufficient.  Continuously  record the temperature of  the sample entering
the filter holder.  Note that the CFV sampler is designed for 325 SCFM
but that operation at the  low end of the normal CVS  range, i.e. 300
SCFM,  is requested here.   This should not  cause a  problem since these
two flows are within 10% of each other, which is the range allowed in
item 4, above.
                                     A-3

-------
     8.  Determine the effect of the length of connecting tubing from
vehicle to tunnel (tubing specifications to be supplied by EPA).
Record temperature of exhaust gas at vehicle tailpipe and dilution
tunnel inlet using both short (2s) and long ( 15') lengths of tubing.
Measure sulfuric acid trapped in the connecting tubing.  The following
method, or equivalents should be used.

          (a)  Rin-se  the tubing with a 60/40 IPA solution until
               an analysis of the IPA shows little or no sulfate response
               on the barium chloranilate (BACL) analytical system.
          (b)  Run one or more emission tests (S7).
          (c)  Rinse  the tubing with IPA and analyze the rinse solution
               for sulfate.

     9. * Determine the effect of dilution air humidity by running S-7
tests at three dilution air humidity levels of 100 tollO, 70 to 80, and
20 to 40 grains/pound^  Measure probe losses and size distribution of
the collected particulate.

     10.  Deliver the dilution tunnel and sampling system to EPA in Ann
Arbor, Michigan.  Provide  technical assistance in setting up and checking
out the ccnplete system with a CFV-CVS.  For planning purposes qn estimate
of 1 man for 3 days at the Ann Arbor facility is reasonable.

     11.  Provide 3 complete sets of shop drawings and specifications,
•including one reproducible set.

     12.  Other.

          (a)  The sulfate analytical system should be dedicated
               to this test work so that preliminary results are obtained
               as quickly  as possible i.e. within a few hours. This is
               the only way  the number of tests to be run can be determined
               while  testing is in progress.
          (b)  The number  of tests requested above, are only esti-
               mates.  Close liaison with the Task Officer is required
               to insure that adequate testing is performed and that
               unnecessary testing is avoided.
          (c)  In the event  probe or tunnel losses are found, system
               modification  and additional testing will be done as
               agreed upon at that time between the contractor and the
               task officer.
                                    A-4

-------
13.  Provide additional equipment to complete sampling system
     including rotometer, dry gas meter, filter, sampling pump,
     and alarm system to indicate loss of critical flow.

14.  Provide mobile cabinet to house above components and install
     components to provide complete working system.

15.  Calibrate dry gas meter and verify that complete system is
     operational

16.  Provide dilution air filter and heater assembly similar to
     assembly in use for tunnel validation tests performed during
     Nov - Dec, 1976.
* Items 7 and 9 were deleted in.favor of increased testing on item 12(c)
                                 A-5

-------
                APPENDIX B





VELOCITY TRAVERSE OF 21 cm DIAMETER TUNNEL

-------
SOUTHWEST RESEARCH INSTITUTE  SHEET NO—!_OF_!
           DATA SHEET              PROJECT	
SUR.IFP.T  21 CM.  D/A.  TUNHEL,	   QATg    1 \ / 2G 174
             SORg T£ftN/g£SE	   BY	M.Kl. JL,
                                                 •SHEETS
   ®
CO- BOTTOM)
i
1 	 4-
L 2"
1
; r 3*
i A'1
\ — 1 	
;."H*
"Ij"
_L
i
._)_
_._.ToP
: i
STJvnc
PRESS.
1,05
4.0^
l.lfe.
4.2D
4.JO.
4. iO
4. jo
:
TOTAL
-L.
4, Lo
" 4", 1 5
^^5
..4.50
4.I5
4- i?
4, IS

0\f?i PRESS.
O.D5
i
0 , 06
O, 01
.. 0./0
^p , o5
0, 05
0,05

VELOC IT/"
,«
1
180
/?oo
1 1 •? /-^
(I 4. U
wo
no


i , ! 1 i i i

-





                              IK).
                  ."_._7_l°f:
VELOCITY
                      AIR \ti-ociTt FLPU) CHARTS   IM
,

i
!
- -











—




—


i ;
1 — j- — f 	
! H2
!


_,






i ! L i ' 1 ' j i i
i ' ! -: i 1 i
1 | • till
: ! I I i i ' ! f
                                                           _l_

-------
   Top
    CD


    3
    §
    H
    M-l

    05




    I
    M
Bottom
            800         900          1000       1100         1200


                                Air Velocity,  feet/second




               FIGURE B-l.  TUNNEL VELOCITY PROFILE
1300
                                  B-3

-------
                Comparison of Measured Average Velocity
           and Average Velocity from Volume Flow Calibration

                      (Area of Tunnel = 0.385 ft2)
1.  CVS blower flow at 28.0 in. Hg.A and 70°F = 381.7 ACFM from blower
    calibration curve
2.  Volume flow at tunnel condition  (29.1 in. Hg.A and 71°F) during
    traverse.



          VT = 381-7  x  Mir  x  §5-  =  368 ACFM

3.  Average velocity =   368    =    956 ft/min
                        0.385


4.  From velocity profile curve intergrated average velocity is 1009
    ft/min.  This is only 5.4 percent higher than average velocity from
    volume flow, which is considered good agreement.  Part of the dif-
    ference is undoubtedly because 890 ft/second was taken as the velocity
    all the way to the wall  (were the velocity is in reality zero) since
    the velocity profile was not obtained close to the wall.
5.  Maximum velocity =          X  Average velocity = 1.25 X average velocity
6.  If volume flow average is used, and maximum velocity is taken to be
    1.25 times average velocity, then the design velocity for the sulfate
    probe is 1195 ft/m  (956 x 1.25), since the probe is to be in the area
    of maximum velocity.
                                   B-4

-------
          APPENDIX C





MIST GENERATOR CALIBRATION DATA

-------
           APPENDIX C - MIST GENERATOR CALIBRATION DATA

Test I - Determination of Mist as a Percent of Total H2SO4 Input

Run
1
2
3
4
5
Test
Time
7'0"
6'43"
10 '39"
7' 57"
7 '05"
Total ml
in Generator
25
25
25
25
25
Mist
Ml
2
2
3
3
3
Percent
Mist
8.0
8.0
12.0
12.0
12.0
Air
Press (psi)
24
20
40
30
25
       Test II - Optimization of Compressed Air Pressure

Run
1
2
3
4
5
6
7
8
9
10
11
12
13
14
Test
Time (sec)
140
19.8
17.7
19.0
21.0
24.0
31.0
20.0
20.0
17.0
16.5
16.5
17.0
17.5
Total
ml
into Generator
1 ml
It
11
II
ri
it
ii
ii
ii
ii
ii
ii
IT
II














                                                              Air
                                                           Press, (psi)

                                                                 0
                                                                30
                                                                20
                                                                30
                                                                35
                                                                40
                                                                10
                                                                15
                                                                16
                                                                18
                                                                20
                                                                22
                                                                24
                                                                26
                                C-2

-------