Stationary Source Enforcement Series
         EPA 340/1-78-OOla
         APRIL 1978
        JET ENGINE TEST CELLS -
        EMISSIONS AND CONTROL
        MEASURES:   PHASE 1
SMI
U.S. ENVIRONMENTAL PROTECTION AGENCY
      Office of Enforcement
    Office of General Enforcement
      Washington, D.C. 20460

-------
                                         EPA-340/l-78-001a

                                         April  1978
  JET ENGINE TEST CELLS - EMISSIONS AND
        CONTROL MEASURES:  PHASE 1
                    by

               D. E. Blake
     Contract No. 68-01-3158, Task 4
   EPA Project Officer:  James Herlihy
          Acurex Report TR-78-102
               Prepared for

   U.S. ENVIRONMENTAL PROTECTION AGENCY
Division of Stationary Source Enforcement
         Technical Support Branch

-------
This report was furnished to the U.S. Environmental Protection Agency by
the Aerotherm Group of Acurex Corporation, Mountain View, California, in
fulfillment of Contract No. 68-01-3158.  The contents of this report are
reproduced herein as received from the contractor.  The opinions, find-
ings, and conclusions expressed are those of the author and not necessar-
ily those of the U.S. Environmental Protection Agency.
The Enforcement Technical Guideline series of reports is issued by the
Office of Enforcement, Environmental Protection Agency, to assist the
Regional Offices in activities related to enforcement of implementation
plans, new source emission standards, and hazardous emission standards to
be developed under the Clean Air Act.  Copies of Enforcement Technical
Guideline reports are available -- as supplies permit — from the Air
Pollution Technical Information Center, Environmental Protection Agency,
Research Triangle Park, North Carolina, 27711, or may be obtained, for a
nominal cost, from the National Technical Information Service, 5285  Port
Royal Road, Springfield, Virginia, 22161.

-------
                              ACKNOWLEDGEMENT

       The author wishes to acknowledge the help and guidance of
Dr. James Herlihy, the Project Officer, and Mr. Charles Seeley of  EPA
Region IX.  In addition, many persons provided useful  information  dur-
ing the course of this program.  Appendix B presents a list of their
names.

-------
                                  ABSTRACT

       This report discusses the current  state of the art of pollutant
emission measurement and cleanup technology related to military jet  engine
test cells.  Considerable emissions  data  from jet engines is available,
but data from test cell stacks are sparse.  Gaseous emission data  for the
major pollutant species (CO, NO, N02,  S02, S03)  are reliable; partic-
ulate, opacity, and unburned hydrocarbon  data are less so.
       The five types of test cell cleanup methods that have been  experi-
mentally evaluated — electrostatic  precipitator, nucleation scrubber,
fuel additives, thermal converter, and fuel atomization improvement  — are
described in detail.  Other, less promising methods are briefly discussed.
Several methods are quite effective  in reducing  test cell emissions.  Fuel
additives are effective in reducing  test  cell plume opacity.  Capital and
operating cost data on these methods are  presented.  For the nucleation
scrubber, the best-developed cleanup technology, three cost estimates from
different sources are given.
       Phase II of this study will discuss operating parameters, state
regulations, additional control device cost data, environmental impact,
clean combustor development status,  and additional emissions data.

-------
                             TABLE OF CONTENTS
Section                                                                Page
   1       INTRODUCTION	    1
   2       SUMMARY	    3
   3       JET ENGINE TEST CELLS — PROCESS DESCRIPTION	    9
   4       LOCATION OF MILITARY TEST CELLS	   19
   5       EMISSIONS DATA	   25
           5.1  Opacity	   25
           5.2  Gaseous Emissions	   27
           5.3  Participate Emissions	   37
   6       TEST CELL EMISSION CONTROL TECHNOLOGY	   47
           6.1  Electrostatic Precipitator	   50
           6.1.1  ESP Experimental Apparatus	   52
           6.1.2  Experimental Results	   56
           6.1.3  Costs	   59
           6.2  Nucleation Scrubber	   60
           6.2.1  Scale Model Scrubber Tests	   75
           6.2.2  Proposed Scrubber Facilities	   79
           6.2.3   Costs	   80
           6.3  Fuel Atomization	   83
           6.3.1  Methane Absorption	   83
           6.3.2  Emulsions	   84
   6.4.    Thermal Converter	   86
           6.4.1  Thermal Converter Configuration for Test Cells	   86
           6.4.2  Pollutant Conversion Performance	   89
           6.4.3  Costs	   90
           6.5  Fuel Additives	   90
           6.5.1  Civilian Airline Additive Use	   97
           6.6  Other Abatement Methods	   99
           6.6.1  Cyclone Separators	  103
           6.6.2  Fabric Filters	  103
           6.6.3  Venturi Scrubbers	  104
                                  vn

-------
                       TABLE OF CONTENTS (Concluded)


Section                                                                Page

           6.6.4  Coanda Noise Suppressor	  105
           6.6.5  Hush  House	  107

           6.7  Summary of Cleanup Technology	  109

           REFERENCES	  Ill

           APPENDIX  A — SCRUBBER  RETROFIT  COST ESTIMATE	  115

           APPENDIX  B — INFORMATION  SOURCES	  133
                                 vi

-------
                           LIST OF  ILLUSTRATIONS
Figure                                                                  Page
   1       Stationary Test Cell (Oceana Type  Shown)	    10
   2       Travis AFB Test Cell Building...	    11
   3       Tinker AFB J79 Engine Test Cell	    12
   4       Augmenter Tube for Tinker AFB Test Cell	    14
   5       Test Cell with Open-End  Augmenter	    16
   6       SAE Smoke Number Versus  Opacity  for Jet  Engines	    28
   7       Relationship Between SAE Smoke Number  and  Soot
           Density for Jet Engines	    29
   8       NOX Emissions from Jet Engines	    32
   9       Comparison of Typical Turbine Engine Exhaust Gas
           S02 Concentrations with  Various  Air Pollution
           Control District Limits	    34
  10       Unburned Hydrocarbon (UHC) Tailpipe Emission Levels
           of Jet Engines at Idle	    35
  11       Particulate Mass Loading for Jet Engines,  Measured
           at the Tailpipe	    44
  12       Electrostatic Precipitator Experimental  Setup --
           Black Point No. 1 Test Cell	    53
  13       Electrostatic Precipitator Test  Unit	    55
  14       Nucleation Scrubber Schematic	    61
  15       Teller Environmental Systems, Inc. Augmenter	    63
  16       Nucleation Scrubber at Jacksonville NAS	    71
  17       Aerial View of Jacksonville Scrubber	    73
  18       Pilot Scrubber System on Black Point No. 1 Test Cell
           at the Jacksonville Naval Air Station	    76
  19       Schematic Diagram of Basic Thermal Converter
           Configuration	    87
  20       Effect of Ferrocene on Particulates	    98

-------
                     LIST OF ILLUSTRATIONS (Concluded)
Figure                                                                  Page

  21       Principle of Coanda Device	   106
  22       Hush House at Miramar Naval  Air Station	   108

-------
                                LIST OF  TABLES
Table                                                                   Page
   1       Operational Navy Test Cells  and  Test  Stands	     20
   2       Operational Air Force Test Cells and  Test  Stands	     21
   3       Operational Army Test Cells  and  Test  Stands	     23
   4       Total Test Cells —  All Services	     23
   5       Participate Size Distribution  from  057  Jet Engine	     39
   6       Comparison of Engine and Stack Participate Measurement...     43
   7       Jet Engine Test Cell Control Measures	     47
   8       Emission Data from the  Electrostatic  Precipitator
           at Jacksonville Naval Air Station,  Black Point
           No. 1 Test Cell — J79  Engine	     58
   9       Nucleation Scrubber  at  Jacksonville Naval  Air
           Station, Emission Levels from  Black Point
           No. 1 Test Cell	     65
                                                  (
  10       Observations During  Test Cell  Cleanup System Test	     68
  11       Test Results on Pilot Scrubber at the Black  Point
           No. 1 Test Cell at Jacksonville  Naval Air  Station,
           Using a J79 Engine	     78
  12       Scrubber Retrofit Cost  Estimate  --  Naval Air Systems
           Command	     81
  13       Scrubber Retrofit Cost  Estimate  —  Jacksonville  NARF	     81
  14       Comparison of Scrubber  Cost  Estimates	     82
  15       Effect of Water/Fuel Emulsion  on Combustion	     85
  16       Thermal Converter Cost  Estimate	     91
  17       Metallic Fuel Additives Evaluated by  the Naval Air
           Propulsion Test Center	     93
  18       Engines Suitable for Use with  Ferrocene Additives	     94
  19       Effect of Fuel Additive on Stack Opacity	     95
  20       Fuel Additive Use by Commercial  Airlines	    100

-------
                         LIST OF TABLES (Concluded)





Table                                                                  Page



  21       Summary of Particulate Abatement  Systems	    101



  22       Comparative Data  -- Alternate Gas Cleaning Systems	    102
                               xn

-------
                                  SECTION  1
                                 INTRODUCTION

       Jet engine test cells  are  used  for testing  jet  engines  after  main-
tenance or overhaul.  They  house  the instruments,  fuel  delivery  systems,
and noise suppression devices  required for  the  testing.   Test  cells  are
maintained by the U.S. Army,  Navy,  and Air  Force,  civilian  airline main-
tenance bases, and gas turbine manufacturers.   When  jet  engines  are  tested,
a dark sooty particulate  is produced by the gas turbines.   And because
test cells usually have vertical  exhaust  stacks, these  emissions  are often
visible from a considerable distance.   Several  jet engine test cell  facil-
ities have recently been  cited for  violation of local Air Pollution  Con-
trol District opacity regulations.  Some  test cells  may  be  in  violation of
particulate mass loading  and  NO   stationary source emission  standards as
                               ^
well.
       The military services  are  aware of these problems, and  have put
into effect a number of programs  to characterize and control test cell
emissions.  The U.S. Air  Force is primarily responsible  for  determining
the nature of test cell emissions and  how they  affect overall military
base and regional pollutant levels.  The  U.S. Navy has  the  responsibility
for evaluating ways of reducing emissions and has  investigated many  dif-
ferent methods.  Civilian airlines have not reported any comparable
studies.

-------
        This  report  is  limited  to military test  cells.   It considers both
 the  emissions  data  available and the current  state  of  test cell cleanup
 technology.  Because of the great amount  of current activity in these
 areas,  a great deal of recent  information  has not been  published.   Accord-
 ingly,  much  of the  information in this report was obtained from interviews
 with people  actively working on test cell  emissions  programs.   Mr.  W.  C.
 Morhard, Environmental Coordinator, Naval  Air Systems Command,  helped  a
 great deal in identifying people working on various  Navy  programs and  in
 obtaining permission for the author to visit and discuss  their  work.   Ap-
 pendix  B gives a list of people who provided significant  useful  informa-
 tion.   The guidance of the Project Officer, Dr.  James Herlihy and Mr.
 Charles Seeley of EPA Region IX,  is also gratefully  acknowledged.
       Additional aspects of test cells will be  considered  in Phase  II  of
 this study.   These will include (a)  discussion of process variables  that
 affect emission,  (b) summary of state and  local  regulations, (c) addi-
 tional  capital  and operating costs,  (d) environmental impact of test
cells,  (e)  discussion  of clean  combustors, and (f)  a summary of available
emission data.

-------
                                  SECTION 2
                                   SUMMARY

        In  this  report,  Aerotherm Division of Acurex Corporation analyzes
what  is currently  k.vjwn about  pollutant emissions from jet engine test
cells and  about cleanup technology for these cells.  This analysis is
limited to military  test cells in the United States and was undertaken for
the Stationary  Source  Enforcement Division of the Environmental Protection
Agency.
       To  provide  a  background for discussion of test cell emissions and
control technology,  the report begins with a description of typical  jet
engine test cells  and  a discussion of how they operate.  This discussion
describes  how augmenter tube design and exhaust water spray cooling  prac-
tice  can affect emissions and  cleanup technology.  The design of the aug-
menter affects  the dilution  ratio of the exhaust stream, and therefore the
opacity, while  the presence  or absence of water spray cooling of the en-
gine exhaust affects particulate and unburned hydrocarbon emissions  in
several ways.   In  addition,  the report also points out that test cells are
not standardized,  and  that  similar engines can produce different opacity
levels in  different  test cell  geometries.
       Next, U.S.  military  test cells are
military service,  and  location.   There anex"a total of 130 permanent^-te'st

-------
cells, 88 demountable test cells, and 273 test  stands  for  the Army,  Navy
and Air Force.
       In Section 5 of the report, the validity of existing  data  and mea-
surement methods is discussed and analyzed.  Opacity,  gaseous emission,
and particulate emission data are treated separately.   Data  on the  opacity
of plumes from test cell stacks are found to be widely available.   Actual
emission data are given in Section 6 of the report.
       The correlation between stack opacity and engine exhaust smoke
number is also discussed.  This correlation is not particularly good, but
it still gives useful information about the relationship between  engine
particulate output and test cell opacity.
       Gaseous emission data from jet engines are also widely available,
but have usually been taken at or immediately downstream from the engine.
Samplings of test cell stacks are sparse.  Methods for measuring  the con-
centration of NO, NO^, CO, and C02 are reasonably accurate and reli-
able in the hands of knowledgeable analysts.  Furthermore, engine tailpipe
data should be readily transformable to stack emissions data  by accounting
for augmentation air, because these gaseous pollutant  species  are sub-
stantially unchanged between engine exhaust and stack  outlet.  After the
exhaust gases leave the tailpipe,  cooling is so rapid  that essentially no
more NO  is formed.   SO  is usually not measured but is calculated
       A               A
from fuel  sulfur specifications.  Measuring unburned hydrocarbons (UHC)
presents problems.   Flame ionization adequately measures overall hydro-
carbon vapor levels.   But analyzing individual hydrocarbon species,  which
requires collection  and subsequent analysis, is complicated  by the  fact
that UHC can exist  in both vapor and particulate form.  This  means  that
the sampling method  used to collect the UHC must be carefully chosen.

-------
       Particulate emission  data  are  the most  difficult  to obtain and are
in general the  least reliable.  There are  almost  no  good particulate data
obtained by sampling at the  top of  test cell stacks;  and,  the  present
state of knowledge does not  permit  accurate  prediction of particulate
stack emissions based on particulate  measurements at  the engine.  Special
problems in measuring particulate emissions  include  particle loss and
reentrainment from noise suppression  equipment in the stack and the
inability to account for the effects  of exhaust water spray cooling.
       In Section 6 of the report,  five test cell exhaust cleanup methods
which have been experimentally evaluated are discussed at some length.
Several other possible cleanup methods which have been judged  infeasible
                     -\
are also discussed in less detail.
       Fuel additives are generally organometallic compounds added in
small concentrations to the  fuel  burned in turbine engines.  The metallic
component of the additive acts catalytically to improve  or at  least modify
combustion.  The main effect of fuel  additives is on  opacity,  although  a
moderate but inconsistent reduction in particulate loading has also been
reported.  Both the Navy and the  Mr  Force are pursuing  experimental pro-
grams to evaluate different  kinds of  fuel  additives.  The Navy has settled
on an iron-based additive for its engines, while  the  Mr Force is con-
sidering a manganese-based additive.   Both appear able to reduce opacity
of test cells to below 20-percent opacity  without damage to the engines
being tested.   Problems include toxicity (for  the manganese compound)  and
thermal sensitization (for the iron compound).  The  cost of using fuel
additives at test cells would be  negligible.
       Improving fuel atomization is  another exhaust  cleanup method now
being tested.   With this method,  an emulsion of about 10-percent water  in

-------
 jet  fuel  is  used  for  testing  in  cells.   The  water globules present in the
 emulsion  flash  in the burners.   This  causes  finer atomization of the fuel,
 and  thus  better combustion.   Improvement in  particulate  and N0x emissions
 and  opacity  have  been reported.   This work  is  in  a relatively early stage.
        A  third experimental cleanup method  is  the thermal  converter.   A
 thermal converter consists of additional fuel  burners  downstream from the
 engine  which oxidize  CO,  carbonaceous particulate,  and UHC.   In experiments
 conducted to date,  the thermal converter has proven  very effective, reduc-
 ing  smoke number  by a factor  of  10.   NO   production  by the additional
                                       A
 fuel  burners is apparently not a  severe  problem.   The  main difficulty with
 the  thermal  converter is  its  high  fuel consumption  and operating cost.
 Capital cost for  a  Type A test cell retrofit has  been  estimated at
 $500,000,  with a  similar  amount  required annually for  operation.
        A  fourth cleanup device,  the electrostatic precipitator (ESP), has
 been  evaluated in 1/65  scale  model tests on  a  test  cell  at Jacksonville
 Naval Air  Station.  The ESP showed reasonable  collection efficiency,
 averaging  59 percent  when particulate was measured  by  EPA  Method 5, and
 higher  (80 to 95  percent) when measured  by a dry  filter  method.
       Problems with  arcing and  high voltage lead failure  were noted  by
 the Jacksonville  Naval personnel testing the ESP.   These problems were
 reported to  be related to water  droplet  carryover into the ESP from the
quench system at  the  augmentor.  Similar problems would  be anticipated in
the demonstration of  a full-sized test cell  ESP unless excess water
carryover  is eliminated.  On  the other hand, operating completely dry (no
quench water) means higher volumetric flowrates in  the ESP.
       According  to reports from Jacksonville  Naval  Air  Station, more
development work  is required,to determine if a satisfactory operating

-------
condition can be maintained.  The cost of  installing  an  ESP  on  an  existing
500,000 acfm test cell has been estimated  at  $850,000.   Annual  operating
expenses are estimated at $32,000.
       The fifth cleanup device considered is  the  wet packed cross-flow
scrubber.  It is the only test cell cleanup device that  has  been given
full-scale evaluation on a cell.  A prototype  installation on the  Black
Point No. 1 cell at Jacksonville NAS  is  a  nucleation  scrubber mounted on
the top of the cell stack.  Water is  injected  into the engine exhaust at
the augmenter tube, causing the nucleation and particle  growth.  The
larger particles are carried  up the stack  and  collected  in the  packed
scrubber.  Limited sampling of the cleaned exhaust showed that  at  least 90
percent of the particulate had been removed and the opacity  reduced  to
below 10 percent.  More extensive sampling of  a scale model  packed bed
scrubber showed that 50 to 70 percent  (average of  55  percent for all
tests) of the particulates and nearly  all  of  the condensible hydrocarbons
had been eliminated.  Two Navy estimates for  the cost of retrofitting
500,000 acfm test cell facilities with nucleation  scrubbers, including
closed-loop water treatment,  are $1,515,000 and $1,944,735.   Teller
Environmental Systems, Inc.,  the system  engineer/contractor  for the
prototype scrubber, estimates the cost of  retrofitting at $705,650.
       In summary, several methods are currently available for  consider-
ably reducing test cell pollutant emissions.   Particulates can  be  effec-
tively removed by the scrubber, the ESP, and  the thermal converter.
Hydrocarbons are substantially removed by  the  scrubber and the  thermal
converter.  The major technical issues relate  to true capital costs, and
the impact on air quality of  controlled  versus uncontrolled  test cells.

-------
                                  SECTION  3
                JET ENGINE TEST  CELLS  ~  PROCESS  DESCRIPTION
                                        /

       Jet engine test cells  house  the engine  and test  instrumentation
during engine testing.  Test  cells  are maintained by the  military services,
commercial airlines and turbine  engine manufacturers.   Most  test  locations
contain only a few cells, although  some large  engine manufacturers main-
tain up to 70 test cells  at their test facilities.   Existing test cells
were designed mainly  to suppress noise, and  they  are reasonably effective
at reducing the very  high sound  levels at the  engine tailpipe (in excess
of 180 dB for some engines) to tolerable  levels.
       Five kinds of  test facilities are  used  by  the military for out-of-
aircraft engine testing.  These  are depot permanent  test  cells, Type A  and
Type C permanent test cells,  demountable  test  cells, and  test stands.
Depot cells are permanent masonry structures,  fully  instrumented.  Type A
cells are similar, but less fully instrumented.   These  kinds of cells are
used to checkout engines that have  been overhauled (completely rebuilt).
       Figures 1, 2,  and 3 are diagrams of typical depot  or  Type  A mili-
tary test cells.  While there are superficial  differences between the
three cells shown, the basic  structures are  similar,. The test cell itself
is usually a massive  reinforced  concrete  structure,  designed to withstand
the intense vibrations generated by the engine.   Air is drawn into the
cell through various  kinds of acoustical  baffling by the  engine acting  as

-------
                                     EXHAUST
                                      OUTLET
                 SOUND
         TREATMENT BAFFLES
    AUGMENTING  AIR
PRIMARY
ENGINE AIR
              TEST ENGINE
MULTIPLE SPRAY  RINGS     EXHAUST
                         AUGMENTER TUBE

                 SECTION B-B
                                                                           CONTROL
                                                                           ROOM
                                                                              &
                                                                           PUMPING
                                                                           FACILITIES
                                                              SECTION A-A
                   Figure 1.  Stationary test cell (Oceana type shown)
                            (Reference 10).

-------
     COMBUSTION
AND AUGMENTER AIR

Kv»>XvXvIv>X
pvXv>»Xv»>»
tV.V.V.V.V.V.V.V

V.V.V.V.V.V.V.V.'.V
••*•*•••••••••*••••
r/ v.v.v %v v.v.v.v
fev>>X%vXvXvv:-
                                                                EXHAUST
                                                          SOUND TREATMENT7
\.
THRUST
r D A M C
r KM m t

-^






•«^
                     Figure 2.  Travis AFB test cell  building
                              (Reference 10).

-------
(\3

                HORIZONTAL ACOUSTICAL BAFFLE


                ACOUSTICAL DUCTS


                VERTICAL ACOUSTICAL BAFFLES
                             COMBUSTION AIR
                                                           AUGMENTER  AIR
EXHAUST

                                      Figure 3.  Tinker AFB J79  engine  test cell
                                                 (Reference 17).

-------
an ejector.   (As with  all  ambient-pressure  test  cells,  no booster fans are
used.)  Although not shown  on  the  figures,  adjustable dampers are often
provided near the  inlet-air acoustical  baffler to control the engine
airflows and test  cell  interior  pressure  levels.
       The test engine  is  fastened to a thrust frame or stand.   This is in
turn firmly anchored to  the cell.   The  thrust frame contains the instru-
mentation used to  monitor  engine thrust.   Immediately behind the engine is
mounted the augmenter  tube, so named because it  augments the flow of air
through the cell.  The  augmenter is simply  an ejector that serves several
purposes:  (1) powered  by  the  high energy engine  exhaust stream, the aug-
menter reduces the test  cell pressure to  a  level  equivalent to  the pres-
sure at the engine compressor  inlet; (2)  the air  drawn  over the engine
provides some of the cooling normally obtained by the motion of the air-
craft in flight; and (3) the air entrained  by the augmenter cools and di-
lutes the engine exhaust.   The cooling  produced  protects the integrity of
the test cell by keeping the concrete below the  spall ing temperature of
350°F to 400°F, and protects the noise  control equipment.  Since a
test cell augmenter must be used with many  different types of engines,
each with its own  air  pumping  characteristics, exhaust  diameter, and
exhaust temperature, the main  concern is  usually  to assure that thermal
damage to the test cell  is  minimized, rather than to optimize the aug-
menter 's airflow properties.
       A typical augmenter  tube  is shown  in Figure 4.  The bellmouth posi-
tion is adjustable so  that  the spacing  between the engine and the aug-
menter can be varied according to  the type  of engine being tested.  The
engine exhaust gas and entrained air are  carried  down the steel augmenter
tube to a position under the exhaust stack.  The  augmenter tube is usually
                                      13

-------
   •MOVABLE BELLMOUTH
                                                                              EXHAUST GAS
K \ /i
' t
5 FT 6 IN.
I
'W
1
1
1
1
X
COOLING SPRAY RINGS"
"T
1
1
!
1
1
._!
y

^t
i
8 FT 4IN.
ooooooooooooooooooooooooooo\
oooooooooooooooooooooooooooA
OOOOOOOOOOOOOOOOOOOOOOOQOOOO
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo*
oooooooooooooooooooooooooooo
OOOOOOOOOOOOOOOOOOOOOOOOOOOOl
0000000000000000000000000007
nnnnrmnnnnnnrinnnnnnnnnnnnnn/
/ PERFORATED BASKET
                          Figure 4.   Augmenter  tube  for Tinker AFB test  cell
                                     (Reference  10).

-------
perforated at this point to  allow the  gas  to  diffuse  out  more  or less  uni-
formly.  This prevents hot spots which could  cause  damage to the concrete
cell.
       Some augmenters, however, are open-ended.  Figure  5 shows a test
cell with an open-ended augmenter.  Exhaust gases are directed up the
stack by permanently mounted turning vanes.   The Black Point No. 1 cell  at
Jacksonville Naval Air Station, where  considerable  work on scrubber and
electrostatic precipitator cleanup  devices has  been done,  is of this type.
       The cooling spray rings shown in Figure  4 provide  important addi-
tional cooling at high engine power settings.   The  cooling spray rings
have nozzles mounted on the  inside  surface that spray water radially into
the hot gas stream.  Water spray cooling is universally used during after-
burner operation.  In most test facilities, water spray is used at all
engine power settings above  idle, but  some test cell  operators choose  to
use cooling water only at military  (100-percent) power and afterburner
settings.  In some test cells, cooling water  is automatically  turned on
when the exhaust gas reaches a predetermined  temperature  (usually 350°F).
       Water spray cooling of exhaust  gases affects pollutant  emission in
several ways.  First, the water spray  scrubs  some particulate  out of the
exhaust stream.  Excess runoff water flowing  out of the augmenter is often
quite black. Second, under some operating  conditions  (idle and after-
burner), considerable amounts of unburned  hydrocarbon vapors are found in
the exhaust.  Quenching the  exhaust stream with water causes the condensa-
tion of these vapors to form additional  particulate.   Third, if excess
amounts of water are sprayed, a saturated  or  supersaturated condition  is
attained.  Water vapor condenses on particulate nuclei, causing particle
growth.  In addition, water  droplets present  in the stream will not
                                      15

-------
                                                Exhaust gases
 Combustion and
augmentation air
                       suppresson
                         baffles
                 Engine
                              Augmenter
 Turning
'vanes
        Figure 5.  Test cell with open-end augmenter.
                             16

-------
evaporate.  The result, often seen when  too much  spray cooling water is
used, is a dark, heavy plume that settles  rapidly,  causing considerable
annoyance to those downwind of the test  cell.
       After the exhaust gases move  out  of the  perforated  basket  (or col-
lander) section of the augmenter tube, they  leave the  cell  through  a ver-
tical stack.  These stacks are often quite large:  200 to  400 square feet
in cross-sectional area and 40 to 60 feet  high  are  common  dimensions.   The
stack is fitted internally with sound baffling  and  absorbent  structures.
       Permanent test cells in the United  States  are not standardized.
While all cells share the same operational features (air inlet, engine
test room, augmenter, stack, sound deadening  structures),  their internal
arrangement can vary considerably.   Stack  diameters, engine-to-stack
spacing, turning vane design, and augmenter tube  design in general  differ
from cell to cell.
       Differences in augmenter design have  a particularly significant
effect on air pollution control.  As the design of  the augmenter  is more
art than science, each test cell designer  makes such changes  in the
augmenter as he thinks appropriate.   As  a  result  the air pumping  charac-
teristics and water spray cooling pattern  of  each augmenter seems to be
unique.  The augmentation ratio of test  cells (air  entrained  divided by
engine exhaust) is adjustable, usually by  varying the  engine-to-augmenter
distance.  It ranges from about 0.6  to 3,  depending on engine power level.
But the pattern of airflow in the augmenter  tube  and the water pattern  are
not adjustable.  The amount of particulate scrubbed out of the exhaust
stream, the amount of hydrocarbon vapor  condensed,  the particulate  wall
losses, and the overall dilution ratios  are therefore  generally differ-
ent from one cell to another, even when  the  same  engine is being  tested.
                                      17

-------
This  is  a complicating factor both  in  determining  emissions from test
cells  and in  designing control devices.
Other  Types of Test Facilities
       Like Type A cells, Type C cells are  usually constructed of rein-
forced concrete.  Some, however, are made of metal.   They are permanent
fixtures, similar in shape to Type  A cells  but commonly  only 70 to 80
percent  of their size.  Type C cells are not fully instrumented,  and often
lack thrust measurement instrumentation.  They are used  for testing en-
gines  that have undergone intermediate level repair  -- that is, repair
that does not involve the hot section of the engine.
       Demountable test cells are of metallic construction  and are usually
smaller  than Type C test cells.   These cells can be  taken  apart and moved
if necessary, although they are large enough and complex  enough that mov-
ing them is not a simple matter.   They are  instrumented  to  about  the same
extent as Type C cells.
       Test stands are simply unenclosed frames designed  to hold  the en-
gine while simple tests are performed.   In test stands,  the engine ex-
haust is  not  confined.
                                     18

-------
                                 SECTION 4
                      LOCATION OF MILITARY TEST CELLS

       Military jet engine test facilities are widely dispersed throughout
the United States.  Tables 1 through 4 show a census taken in July 1975.
"Cells" include both Type A and Type C permanent test cells.
                                      19

-------
TABLE 1.   OPERATIONAL NAVY TEST CELLS AND TEST
          STANDS (Reference 18)
Location
Alaska
Arizona
California
Florida
Georgia
Hawaii
Louisiana
Maine
Maryland
Massachusetts
Mississippi
New Jersey
North Carolina
Pennsylvania
South Carolina
Texas
Virginia
Washington
TOTAL
Cells

1
31
7

1


1


9
10

1

9
2
72
Stands
2
2
16
10
2
6
2
1
3
1
3
1
6
3
1
7
6
1
73
                        20

-------
TABLE 2.  OPERATIONAL AIR FORCE TEST CELLS AND TEST STANDS (Reference 18)
Location
Alabama
Alaska
Arizona
Arkansas
California
Colorado
Connecticut
Delaware
Florida
Georgia
Hawaii
Idaho
Illinois
Indiana
Iowa
Kansas
Kentucky
Louisiana
Maine
Maryland
Massachusetts
Michigan
Minnesota
Mississippi
Missouri
Nebraska
Nevada
New Hampshire
New Jersey
New Mexico
New York
roil,- Demountable
Cel1s Cells
2

4
2
8 9
1

1 2
4 3
3
1

2
1
1


2

1 1
1
3

4
2
2
1
1 1
1
5
1 1
Stands
3
3
11
5
18
3
1

10
5
4
3
5
3
2
4
6
6
2
4
4
7
2

3
2
4
1
3
6
5
                                  21

-------
TABLE 2.   CONCLUDED
Location Cells
North Carolina
North Dakota
Ohio
Oklahoma 15
Oregon
Pennsylvania
South Carolina 1
South Dakota
Tennessee
Texas 10
Utah 3
Vermont
Virginia
Washington 1
West Virginia
Wisconsin
Wyoming
TOTAL 47
Demountable
Cells
1
2
2
3


3
1

15
2
1

3

1

88
Stands
5
3
10
2
1
3
6
1
1
12
2

5

1

1
188
         22

-------
                       TABLE 3.  OPERATIONAL ARMY TEST CELLS AND TEST STANDS  (Reference  18)
                                Location
        Cells
         Stands
                             Pennsylvania
                             Texas
                             Virginia
                                 12
                                            TOTAL
         10
          1

         11
           12
JNS
CO
                               TABLE  4.  TOTAL TEST CELLS - ALL SERVICES (Reference 18)

Navy
Air Force
Army
Cells
72
47
11
Demountable c*a^«.
Cells Stands
73
88 188
12
                                TOTAL
130
88
273

-------
                                  SECTION 5
                                EMISSIONS DATA

       This  section  discusses  each  of the three main classes of jet engine
test cell pollutants --  visible,  gaseous, and particulate emissions.   Pol-
lutant measurement methods  are analyzed, problems  and drawbacks are
pointed out,  and  an  assessment of the general validity of published data
is made.  Actual  emission data are  given in Section  6, Test  Cell  Emission
Control Technology.   A compilation  of emission data  will  be  included  in
Phase II of  this  study.
5.1    OPACITY
       The typical plume from  a jet engine  test cell is quite dark, be-
cause of the  carbonaceous nature  of the  particles  that are emitted.  It
can be easily seen by anyone  in the area.  Plume opacity  is  the only  basis
to date upon  which test  cells  have  been  cited for  violation  of local  Air
Pollution Control District  requirements.  Plume opacity varies with the
engine tested, the type  of  cell,  the power  setting used,  and the  kind of
fuel being burned.
       A basic problem exists  in  attempting to predict the opacity of the
plume produced by the same  engine tested in a different cell.   In general,
different opacities  will result —  in some  cases,  markedly different.
There are several reasons for  this.   One is the physical  differences  that
exist between the test cells.   Another  is the different methods used  in
                                      25

-------
operating  test  cells.  And a third reason  is the modification  of particle
size  distribution  that occurs between engine tailpipe  and  cell  stack exit
plane.
        Test cells  are  individually designed and constructed, as are their
augmenter  tubes, exhaust quench spray devices, and noise abatement  equip-
ment.   Accordingly,  different cells, when  testing the  same engine,  will
show  different  augmenter air dilution ratios, exhaust  gas  quench rates,
stack  velocities,  and opacities.  Particle residence time  will  thus vary,
as will  unburned hydrocarbon condensate particle size  distribution, and
stack  wall  loss and  reentrainment rates.   The result will  be different
particle mass loadings and size distribution for different cells testing
the same engine.   In addition, the opacity of the plume can be  affected  by
the augmentation ratio (ratio of entrained air to engine exhaust).
        Because  of  these many variables, not only will  a given  engine give
different  opacity  readings in different test cells, but the same engine  in
the same cell can  from time to time give different readings because of
variations in operating procedures.
       Another  parameter used to measure opacity is called the  smoke
number.  The smoke number is usually used  to quantify  the  opacity of the
gas stream at or near the engine tailpipe.  There are  several  methods for
measuring  smoke number.  Although they are generally similar,  they  none-
theless result  in  different values for the smoke number of an  engine
(Reference 21).
       The Society of Automotive Engineers (SAE) has established Committee
E-31 which is responsible for development of recommended practices  for
measuring smoke emissions.   Aerospace Recommended Practice (ARP) 1179 has
resulted from their efforts (Reference 24).  The technique for  quantifying
                                     26

-------
smoke emission specified by ARP  1179  requires  sampling  and  transfer  of the
exhaust stream, filtering of  a known  quantity  of  the  sample gas  through
filter paper, and measurement of the  reflectance  of the soiled filter
paper.  A quantitative number (SAE  smoke  number between 0 and  100)  is
produced as a result of this  technique.
       The U.S. Navy has established  a similar procedure (Method AS  1833)
to give an opacity characteristic called  the Navy smoke number.   Although
the SAE smoke number is more  commonly reported, Navy  smoke  numbers can be
converted to SAE smoke numbers by means of  a good correlation  (Reference
22).
       Figure 6 shows  the experimentally  determined relationship between
SAE smoke number and opacity.  It is  important to remember  that  the  smoke
number is measured at  the engine, while the opacity is  measured  at the
stack.  As can be seen, the correlation is  only approximate.
       A correlation also exists between  SAE smoke number and  particulate
mass concentration, the Champagne correlation  (Reference 23).  Figure 7
shows this correlation compared  to  actual engine  data.   Using  Figures 6
and 7, particulate mass loadings at the engine can be compared to opacity
readings at the stack.
5.2    GASEOUS EMISSIONS
       There is a very large  body of  data for  jet engine gaseous emis-
sions.  Most of these  data have  been  collected at the engine tailpipe or
somewhat downstream.   Little  data have been obtained  by sampling at  test
cell exhaust stack exits.  In spite of this lack  of data, there  is evi-
dence that concentration measurements for gaseous emissions made at  the
stack exit are consistent with those  made at the  tailpipe when allowance
                                      27

-------
                                                          A-134878
  100
   80  -
   60  -
Q.
o
   40  -
   20
                                                                       5.0
TF 30-P-6
TF 30-P-8 (smokeless)
J 52-P-6A
J 52-P-8A
J 65 (clean)
T 56-A-10
J 57-P-8
                                          correlation
                               40          60
                        SAE  smoke number  (engine)
         Figure 6.  SAE smoke number versus  opacity for jet engines
                    (Reference 22).
                                    28

-------
    O  T56  (501)                                 £ T65-GE-413

    0  J57-P-8                                   Q T58-GE-10

    Q  JT8D  (conventional) J52-P-6A  (conventional combustor) smoke numbei
    O  JT8D  (smokeless) J52-P-8A  (clean combustor) smoke number

    A  T4-0  (max particulates) at max power
    100
5-
01
0)
O

CT.
E
in
c
O)
"O
O
O
   10.0
    1.0
    0.1
                                                                      0
                                                                      vo
                   20          40          60
                         SAE smoke number (ARP 1179)
80
100
            Figure 7.   Relationship between SAE smoke number and
                       soot density for jet engines  (Reference 22).
                                   29

-------
 is made for entrained air  (Reference 25).  This conclusion  seems  reason-
 able because the major gaseous pollutants — CO, C02,  NOX,  SOX  — are
 unlikely to react chemically or to be adsorbed or absorbed  between the
 engine tailpipe and the top of the cell stack.  Unburned  hydrocarbons are
 an exception to this rule  in those cases where water spray  causes
 condensation.
       There are many data on gaseous emissions from jet  engines  in Refer-
 ences 22, 26, 27, 28, 29,  and 30.  Although it is beyond  the  scope of this
 report to reproduce these  data, each major pollutant gas  is identified and
 briefly discussed in the following sections.  In some  cases,  emission
 factor data are summarized.
 Nitrogen Oxides
       The nitrogen oxides NO and N02, which appear with  limited  amounts
 of NpO and N^Oc, are emitted from jet engine test cells,  sometimes in
 significant concentrations.  Nitrogen oxides are formed primarily in high
 temperature, high pressure combustion reactions like those  that take place
 in jet engines at the higher power settings.  Data taken  in all studies
 confirm this prediction.  The larger, newer jet engines (which burn with a
 higher combustion zone temperature) tend to have noticeably higher concen-
 trations of NOX in the exhaust, principally NO and NOo.   N02  consti-
 tutes 45 to 100 percent of NOX at low engine power settings,  and  10 to
 30 percent at maximum power settings (Reference 30).
       NO is eventually converted to N02 in the atmosphere  by a slow
oxidation reaction.   Therefore, nitrogen oxide concentrations are usually
reported as "total  nitrogen oxides expressed as N02".  Since  concentra-
tions are reported  by volume,  this means adding concentrations of NO and
N02 plus twice the  concentrations of N20 and N20g.
                                     30

-------
       Large variations  in NOX concentration  levels  have  been found  for
different engines, and even between engines of the same type.   Figure  8
gives NOX tailpipe concentration measurements (ppm actual) for  several
common engines.  All measurements were made with the engine at  intermedi-
ate power level which typically gives the highest N0₯ concentration
                                                     A
levels.
       The Society of Automotive Engineers' standard test method ARP 1256
specifies techniques for detecting various gaseous pollutant species.  The
instrumentation method specified for measuring NO is nondispersive infra-
red (NDIR).  For measuring N02, nondispersive ultraviolet (NDUV) is
preferred.
       Recently, however, a considerable amount of data has been gathered
using chemiluminescent techniques to measure NO  concentrations.  In
                                               A
fact, most NO  measurements are currently made using chemiluminescent
             A
monitors.  NOX measurements made by NDIR, NDUV, and  chemiluminescent
methods are believed to  be reasonably accurate.  The problems involved in
measuring NO  for gas turbines are not significantly more difficult  than
            n
for other combustion processes.
Sulfur Oxides
       Sulfur oxide concentrations in jet engine exhaust  streams are in
general far lower than the most stringent current Air Pollution Control
District regulations.  Based on maximum fuel sulfur  content allowed  by the
fuel specifications, the greatest fuel/air ratio (in maximum afterburner
mode), and no dilution by augmenter air, the maximum possible S02 con-
centration in the engine exhaust is on the order of  100 ppm by  volume,
This theoretical value is a factor of three less than the most  stringent
current S02 standards (San Diego APCD).
                                     31

-------
              Range  of  scatter of experimental
              measurements
200 1
150
_ 125-
• —
o
E 100'
o.
Q-
x 7C
o 75 •
•z.
50'
25-


•









•V^M



I



—









•
•





m



—



I



	









I |









•^••P-









I
s
?
J
^




A
  J79
J52   J57-P-8    TF30
T56
Figure 8.   NOX emissions  from jet engines
           (Reference 22).
                        32

-------
       Measurements of SOX  at  the  top  of  test  cell  stacks  have  been  made.
Concentration  levels far below the theoretical  tailpipe  concentrations
have always been recorded.   Figure 9 compares  actual  stack  measurements
made at Albany Naval Air Station with  theoretical  tailpipe  concentrations
and several APCD limits.  The  stack measurements  are  low because  of  the
dilution by augmentation air,  and  also because  the  fuel  had a lower  sulfur
content than specifications required.
       Sulfur  oxide concentrations in  gas turbine  exhaust  streams are
usually determined by standard wet-chemical  methods.   A  recent  measure-
ment (Reference 31) indicates  that the ratio of S03 to S02  in aircraft
turbine engine exhausts was 0.138.
       The conclusions to be drawn regarding SO  emissions  by jet engine
test cells seem clear:  (1) because of the very clean fuel  used,  SO  con-
                                                                   A
centrations are well below  regulatory  requirements; and  (2) measurement
methods are well developed  and reasonably accurate.
Unburned Hydrocarbons
       Unburned hydrocarbons (UHC) are emitted  from jet  engines as a
result of incomplete combustion.   The  UHC include  gases  and condensible
vapors.  Jet engines usually emit  the  highest  concentration of  UHC at
idle, although the maximum  amounts (Ib/hr) are  emitted in  afterburner
mode.  Figure  10 shows UHC  levels  measured at  the  tailpipe  of several
engines (Reference 22} at idle.  The engines are  either  in  compliance or
marginally in  violation of  Bay Area Air Pollution  Control  District (BAAPCD)
standards, except for the J57. At all engine  power settings above idle,
UHC concentration falls to  insignificant  levels (Reference  19).
       Unburned hydrocarbon emissions  are typically measured by flame
ionization detectors.  Both condensible and  noncondensible  hydrocarbons
                                      33

-------
500
400 '
300
*:
100 •
80 '
60
40
20

tic
iy Area HI'LIJ i inn L
|

San Diecio County APCD limit
t l -
i






o





<




F
             0

    (^ Maximum SO;? in gas turbine engine exhaust  based on maximum
        specification sulfur content of fuel and maximum fuel/air
        ratio (afterburning)


    LJ SO;? measured at stacks by Source Emissions Team at Albany
        Naval Air Station  (J79 with afterburner)


Figure 9.  Comparison of typical turbine  engine  exhaust gas
            SOp concentrations with various air pollution
            control  district  limits (Reference 22).
                                   34

-------
                      I
                                = Range of  scatter of experimental
                                  measurements
    1200


 o  1000


j|   800  •
o

=   600


     400


     230
      BAAPCD limit
       IT
                                                          in
       J79      J52   J57-P-10  J57-P-8    TF30     T56
Figure 10.   Unburned hydrocarbon (UHC)  tailpipe emission levels
            of jet engines at idle (Reference  22).
                         35

-------
 are  included  in these emissions,  a fact  which  complicates the measurement
 of particulates and  (in  some cases)  of UHC.
        In the  case of particulate measurement,  differences between the EPA
 and  Los Angeles Air  Pollution Control District  (LAAPCD)  particulate sam-
 pling  trains  (discussed  in more detail below)  can make  significant differ-
 ences  in particulate concentrations, depending  on whether condensible UHC
 is considered  to be  particulate.  In the case of UHC  itself,  test cell
 operating procedures can significantly affect  the readings obtained.   For
 example, the  water sprayed to cool exhaust gas  can  cause considerable con-
 densation of  UHC —  so much so that  at idle power settings, more  condensed
 UHC  aerosol than solid particulate may be present.  Some of this  condensed
 material will  be scrubbed out by  the spray water, but most of it  will be
 carried out the test cell stack.  If sampling  is being  done at the cell
 stack,  the condensed UHC may be measured as particulate,  as total hydro-
 carbon, or not at all, depending  on  the way in  which the sampling train  is
 set  up.  Some  of the UHC measurement discrepancies  reported in the litera-
 ture have been ascribed  to this cause (Reference 25).
 Carbon Monoxide
       Jet engine CO emissions range from 2 to  15 lb/106 Btu  for  various
 engines at idle (Reference 19) and fall to insignificant  values at higher
 power settings.  CO concentrations are commonly measured  by nondispersive
 infrared techniques.   Unless scrubbers are used, concentrations measured
 at the tailpipe should be very much the same as stack concentrations
 (taking into account  augmentation air).  This is because  almost no CO is
converted  to C02 at the low stack temperatures  common in  test  cells.
                                     36

-------
5.3    PARTICULATE  EMISSIONS
       Mass emissions from  jet  engine  test  cells  are  much  more  difficult
to measure accurately than  visible  or  gaseous  emissions.   There are  three
main reasons why:   (1)  the  nature of the  particulate  emitted; 2)  the fre-
quent confusion between condensed unburned  hydrocarbons  and solid particu-
late; and  (3) the fact  that particles  emitted  by  jet  engine change in size
as they move away from  the  tailpipe.  In  addition to  these fundamental
problems,  there are several operational  difficulties  that  arise.   Jet
engine test cell stacks are often 20 feet in diameter or  larger.   Isokin-
etic sampling conditions are  difficult to assure, because  the presence  of
sound baffling grossly  distorts the velocity profiles,  setting  up large
velocity gradients  and  even backflows.  Because of these  difficulties par-
ticulate emission data  tend to  be incomplete.
Nature of  Jet Engine Particulate
       The particles emitted  by jet engines are about 95  percent  carbon by
weight, and 5 percent oxygen  and hydrogen (Reference  32).   These  carbon
particles, which rapidly grow into  soot  in  the tailpipe  region, oxidize
completely at the high  temperatures downstream of the tailpipe  unless
quenching  occurs.   In turbine systems, the  formation  of  very large soot
particles  is prevented  by the vigorous mixing  that occurs.  In  gas turbine
burners, back mixing to recirculate combustion products  with unburned
liquid droplets in  the  primary  zone is deliberately induced. Quenching
occurs because combustor walls, turbine  blades, and other  engine  internals
are cooled with excess  air.  Any soot  particles which come into contact
with these cooled engine parts  are  quenched immediately.
       Measuring the size of  particles emitted by jet engines  is  compli-
cated by several factors (References 33,  34,  12).  At the  tailpipe,  as has
                                      37

-------
  been stated, most of the participate  is in the form of  soot.   Further away
,  from the engine, however, particles composed of condensible  hydrocarbons
  begin to become significant, particularly at low power  settings.   The soot
  particles are substantially submicron (on both a number  and  a  mass basis)
  at the engine tailpipe, and grow slowly by agglomeration  as  they  move away
  from the engine.  By contrast, the condensed hdyrocarbon  particulate rap-
  idly agglomerates into particles on the order of 10 microns  in diameter
  (Reference 19).  At high power settings, however, the exhaust  stream
  remains hot enough so that condensible hydrocarbon aerosols  do not form.
         The size of the particles emitted by jet engines  has  not been fully
  established.  Collecting and characterizing these particles  under  the ex-
  treme conditions in the engine exhaust is a difficult task.  Tests carried
  out to date report average particle diameters of 0.02 to  0.06  micron,  with
  some instances of particles with diameters up to 0.12 micron (on  a number
  median basis).  On a mass basis, particles are reported to average between
  0.2 to 0.4 micron in diameter.
         Electron photomicrography shows that soot particles from jet  en-
  gines are more or less spherical.  Usually several particles are  grouped
  to form a chain.  This agglomeration is probably the main reason  for the
  growth in size observed as soot particles move away from  the engine.
  Table 5 shows, for a J57 engine, the average size of soot particles  at
  varying distances from the engine tailpipe.
         There is at present very litte information about size distributions
  of the  aerosol produced by condensation of hydrocarbon vapor.
                                       38

-------
TABLE 5.  PARTICULATE SIZE DISTRIBUTION FROM J57 JET ENGINE
          (Reference 35)
Downstream Location from
Nozzle in Nozzle Outlet
Diameters
0


2-1/2


10


Engine Power
Approach
75% Normal
Cruise
Approach
75% Normal
Cruise
Approach
75% Normal
Cruise
Geometric Mean
Diameter, ym
—
0.053
0.052
0.084
0.084
0.076
0.096
-
0.13
Standard Deviation
Mean in.
—
1 .63
1.46
1.33
1.40
1.51
1.38
-
1.40
                                39

-------
Sampling Methods

       The participate emitted by jet engines has been sampled mainly at


three locations in test cells.  Some data has been obtained  at the  engine


tailpipe or a few nozzle diameters downstream, where temperatures,  vibra-


tion levels, and thrust loads are very high.  Because gas  velocities  are


also extremely high here, isokinetic sampling is difficult.  Most  sam-


pling at the tailpipe has been done at a fixed location  and  a single  sam-


pling flowrate.  A Method 5 sampler capable of isokinetic  sampling  at


supersonic velocities with full traversing capability has  recently  been


developed by the Aerotherm Division of Acurex Corporation  with support


from the Naval Aircraft Environmental Support Office and the Air Force


Civil Engineering Center.  This sampling instrumentation should soon  pro-


vide much better particulate data than now exists.


       Sampling conditions are much better 15 to 20 feet away from  the


engine, near the end of the augmenter tube.  The augmentation air  is  by


this point well mixed with the engine exhaust stream; the  combined  gas


stream is considerably cooler and is moving more slowly.   Most of  the par-


ticulate emission data that have been reported were collected under these


conditions.


       Measurements made 15 to 20 feet from the engine tell  more about


engine emissions than test cell emissions, for two main  reasons.   First,


the cell  is almost always run dry (without exhaust cooling spray water)
                                                         V,

when sampling is done at the augmenter, so that the water  spray does  not


affect particulate concentration.  This means that no particulate  is


scrubbed  out of the exhaust stream and there is no nucleation leading to


particle  growth or condensation of hydrocarbon vapor.  Second, the  exhaust

gas is sampled before it passes through the stack with its acoustic
                                     40

-------
baffling.  As a result,  the  wall  losses  and particle reentrainment which
occur as gas passes  through  the  stack  are not taken into account.
       The third  location  where  sampling has been done is at the top of
the test cell stack.   Very few attempts  have been made to sample particu-
late at this location.   In fact,  to our  knowledge, no successful attempt
has ever been made to  obtain complete  particulate emission data by multi-
point isokinetic  sampling  at the exhaust plane of a Type A test cell.
There have been a very limited number  of samplings made at the exhaust
plane of smaller  cells.  Reference 25  discusses a very complete sampling
at the exhaust plane of a  test cell with a small  (6-foot diameter) stack,
testing a J57 engine.   Partial sampling  of a large (700 sq.  ft.) test  cell
stack has recently been reported (Reference 36).
       A joint effort  by the Environmental Protection Agency,  the  Coordi-
nating Research Council, the U.S.  Air  Force and the U.S. Navy (Reference
25) to completely characterize the emissions from a small test cell  illus-
trates some  of the problems  that can arise'in particulate sampling from
test cells.  Eight sampling  points were  selected at the top of the 6-foot
diameter stack.   Based on  the velocity profiles,  this was probably not
enough to assure  a completely representative sampling.
       Sampling for  particulate  was done using both the EPA sampling train
and the LAAPCD train.   The EPA train consists of a heated probe and filter,
four impingers, an ice bath, and an exhaust gas metering system.  The  probe
and filter are kept  heated to 250°F to prevent condensation of water.   As
a result, most hydrocarbon vapor does  not condense.  The particulate catch
is the total amount  of material  on the filter and in the probe.  The sys-
tem is designed so that only solid particulate and hydrocarbons condensing
                                      41

-------
above 250°F are captured on the filter and upstream  of  it.   Hydrocarbons
condensing between 70°F and 250°F are captured  in the impingers.
       The LAAPCD train consists of three impingers  in  a  70°F  bath,  fol-
lowed by a filter thimble and an exhaust gas metering system.   The  tested
gas  is thus cooled before filtration, and condensible UHC knocked
out.  Unlike the EPA train, however, the condensible gaseous material  col-
lected in the LAAPCD train is included as particulate.  Because of  this
difference, one would expect the particulate concentrations measured by
the  LAAPCD method to be consistently higher than those  measured by  the EPA
method.  In fact, just the opposite result was  obtained during  this
testing program.  One possible cause for this anomalous result  is that the
EPA  matted filter is highly efficient and may exceed the  capability  of the
LAAPCD thimble.  Generally, however, data reported in the literature for
test cell sampling shows that the LAAPCD train  does give  higher particu-
late mass loadings than the EPA train, as expected.
       This test result shows that condensible .hydrocarbons can have a
significant effect on particulate emissions measurement.  It also shows
that confusing and contradictory results can be obtained  even  by
experienced and knowledgeable sampling crews following  standard practices.
Particulate Emissions Data
       Three kinds of particulate emissions data are obtained  from  jet
engine test cells:  tailpipe, uncontrolled stack, and controlled stack
emission measurements.
       Tailpipe data are defined as measurements taken  at or immediately
downstream of the engine exhaust nozzle.  Because of the  extremely  high
gas velocities,  temperatures and pressures existing  in  this region,  mea-
surement of particulate loadings and size distributions are difficult  to
                                     42

-------
obtain.  Those participate  data  that  do  exist  are  somewhat  questionable,
because it is doubtful that isokinetic conditions  were maintained  during
sampling.  Instrumentation  is  now  being  developed  (see Sampling Methods,
Section 5.3) which will  soon allow accurate  tailpipe  particulate data  to
be obtained.  Figure  11  shows  representative tailpipe data  for several
engines tested at two different  power settings.  More complete tailpipe
emission data will be presented  in the report  of Phase II of  this  study.
       Uncontrolled  stack  data are very  sparse, due mainly  to the  diffi-
culty of accurately  sampling the very large  stacks common to  test  cells
(see discussion  in Section  5.3).   Table  6  gives data  obtained on a rela-
tively small  (6-ft diameter) metal stack in  one of the few  carefully per-
formed sampling  tests that  have  been  conducted.

      TABLE 6.   COMPARISON  OF  ENGINE  AND STACK PARTICULATE  MEASUREMENT
Measurement Method
EPA sampling train

LAAPCD sampling
train
J57 Engine
Power Setting
IDLE
CRUISE
IDLE
CRUISE
Particulate
Engine Tai
0.041
0.076
0.025
0.016
Concentration, Grains/DSCF
Ipipe Stack Exit Plane
0.009
0.013
0.009
0.012
       Tailpipe  data,  which  were  obtained simultaneously with  the stack
data, are  presented  in Table 6  for  comparison.   Note  that both measurement
methods used  (EPA  and  LAAPCD) show  a  considerable  reduction  in particulate
mass  loading  from  the  engine to the stack.   This is  due  to dilution  by aug-
mentation  air, by  expansion  of  the  gas  at the lower  downstream pressure,
                                      43

-------
U. £U
£ 0.15 -
oo
o
c
•i —
(O
en
en °'10
c
-o
(O
o
1/5
1/5
£ 0.05 -








D


O

J79








D


O








D










D
>-k p~i
0 DO u
O
01
J52-P-8A J57-P-10 J57-P-8 TF30 T56-A-7

O Maximum power setting
f~l TH1 e nnu/or cottinn
P-6
P-8
P-408
Figure 11.   Particulate  mass  loading  for  jet  engines, measured  at  the  tailpipe
            (Reference  22).

-------
and probably also to wall  losses.  A  summary of  available  stack  emission
data will appear in the report  of  Phase  II  of  this  study.
       Controlled stack data  are nearly  nonexistent.   Only one full  sized
controlled test cell stack exists  (the scrubber-equipped Black Point No.  1
cell at  Jacksonville Naval Air  Station), and only one attempt has  been
made to  sample  the  cleaned exhaust gases.   Various operational difficulties
during sampling make the  results obtained  of very limited  value; this test
is  discussed  in detail  in Section  6.2.1.
                                      45

-------
                                 SECTION 6



                   TEST CELL EMISSION CONTROL TECHNOLOGY



       The military services recognize emissions from jet engine test

cells as a serious problem requiring investigation and solution.  The U.S.

Navy, which has responsibility for  investigating control technology, has

funded many research programs in this area.  As a result of their research,

a variety of methods for controlling jet engine test cell emissions have

been suggested.  Table 7 shows the  different control technologies that

have been proposed.



              TABLE 7.  JET ENGINE  TEST CELL CONTROL MEASURES
                        Thermal Converters
                        Electrostatic Precipitators
                        Fuel Additives
                        Fuel Atomization  Improvement
                        Nucleation Scrubbers
                        Spray Towers
                        Packed Towers
                        Cyclone Separators
                        High Spray (Ejector) Towers
                        Venturi Scrubbers
                        Wet Cyclones
                        Filters
                        Impingement Separators
                        Acoustic Precipitators
                        Magnetic Filters
                                     47

-------
          The Navy has determined that the first five of these control tech-
   nologies warrant experimental  study.   Most of the remaining methods have
   been briefly explored and then discarded because of obvious deficiencies.
   In this section, each of the five control technologies that have been
   evaluated is described in some detail.  A description of the method, the
   current state of its development, and capital and operating cost estimates
   are presented.   In addition, some of  the less promising concepts are
   briefly described.
          Control  technologies  are considered here mainly as retrofit devices
   for existing test cells.  In view of  th'is, any control device must be able
   to function within the operative parameters of the test cell to which it
   is added.  The  pressure drop of the control devi ce^for example, is one of
   the most important considerations:   The operational characteristics of the
x^engine under test can be affected when back pressures exceed about 20-inch
   water (Reference 10).  Current acoustical  treatment of the exhaust stream
   causes a back pressure of about 10-inch water.  This means that a control
   device fitted to the exhaust stack  should not cause a pressure drop of
   greater than about 10 inches since  a mechanical means of exhausting the
   test cell to overcome high control  device pressure drops is very expensive
   due to the extremely high gas flowrates involved.
          Any control device considered for jet engine test cells must be
   able to withstand the very high sound levels existing in the gas exhaust
   stream.  Average sound levels at the augmenter  range from 170 dB for a J57
   up to 180 dB for a J75 (Reference 16).   The reinforced concrete walls of
   test cells crack fairly frequently.
          Water availability is an important consideration for some cleanup
   devices,  particularly wet scrubbers.  At many installations, the water
                                        48

-------
supply is limited.  Any  device which  uses  large  amounts  of water would not
be feasible at these  locations.   Nucleation  scrubbers, for example,  have
the highest water usage  of  the cleanup  devices considered  here.   Even
though the scrubber packing irrigation  water can be  continuously cleaned
and recycled, a good  deal of water  is lost in the quench spray system.
Nucleation scrubbers  must saturate  the  exhaust gas downstream  of the aug-
menter in order to produce  the nucleation  and particle growth  necessary
for particle collection.  The amount  of water evaporated to saturate the
exhaust gas varies with  engine power  setting, but in general will  always
be greater than that  required for an  uncontrolled cell.
       The physical dimensions of a cleanup  device and  its support equip-
ment are also important  for retrofit  installation.   Test cells are commonly
built side-by-side in pairs sharing control  and  ready areas.   In some
cases (for example at Tinker AFB) many  cells are positioned in a row with
minimal clearance betweeen  them.  A control  device which takes up  a  lot of
room on either side of the  cell  stack will probably  not  be suitable  for
pairs or groups of cells.   Similarly, the  space  required for support
systems could become  a factor.   In  some locations, the ground  space  neces-
sary for the cooling  tower, pumping station, and water cleanup plant asso-
ciated with nucleation scrubbers, for example, may be unavailable.
       These restrictions and limitations, combined  with the diversity in
test cell design noted earlier,  make  a  standard  cleanup  device for all
test cells impractical  (with the  obvious exception of those methods  such
as fuel additives that require no alteration of  the  test cell  structure).
Different cleanup technologies may  be required in different locations
because of water availability, fuel cost,  electricity cost, physical space
availability, and local  air pollution control regulations.
                                      49

-------
6.1    ELECTROSTATIC PRECIPITATOR
       The electrostatic precipitator  (ESP)  is  an  obvious choice for a
test cell cleaning device.  It  is efficient  in  collecting the micron- and
submicron-sized particles that  make up  the test  cell  exhaust plume.  It is
capable of operating at elevated temperatures and  is  a  well  developed,
proven method  in current use  in a number of  diverse  applications.  ESP's
are also  low pressure drop devices, an  important consideration for the
test cell application.
       Particle collection in an ESP  is accomplished  by allowing the
particulate-laden  gas to pass close to  a wire,  point, or sharp edge which
is raised to a high enough electrical potential  in relation  to other
nearby parts of the apparatus to continuously ionize  the gas near the
wire.  The  ions thus created  rapidly  sweep through the  gas to be cleaned.
Particles present  attract and then adsorb the ions,  acquiring an electro-
static charge.  In a single-stage precipitator,  the  charged  particles are
carried to nearby  low potential surfaces by  the  field which  exists between
the ion source and the collecting surfaces.  In  a  two-stage  ESP, the par-
ticles are charged as described and then collected downstream in a separ-
ate high-field region.
       The efficiency of the ESP is limited  by  the ion  flux  density in the
charging zone,  the particle size, the magnitude of the  field in the col-
lecting zone, the distance a particle must travel  to  a  collecting surface,
the particle resistivity, and the residence  time of  a particle in both the
charging and collecting zone.    The designer  has  little  control over charg-
ing and collecting fields.   These are always the maximum values which can
be attained without excessive sparkover.  The performance of an ESP for a
given  gas  stream is thus determined by the residence  time of a particle in
                                     50

-------
the ESP, or by the gas  velocity and  available collection area for any spe-
cific precipitator.
       In  industrial  applications,  gas  velocity commonly ranges between 3
and 15 ft/sec  (Reference  8)  with 8  ft/sec (500 ft/min)  being considered a
maximum for good  practice in most applications.  This  low operating veloc-
ity is a drawback of  test cell  use.   For a cell designed for 500,000 acfm
(typical of a  TF  30 engine with conventional  augmenter), the required pre-
cipitator  cross-sectional area  to maintain a  500 ft/min gas  velocity in
the ESP would  be  1000 ft2. Providing this large an  area is the main rea-
son for the high  cost of  installing  retrofit  ESP's  on  test cells.
       There are  several  peculiarities  of test cell  operation that con-
strain ESP design.  The carbonaceous nature of the  particulate makes fire
a  real danger  if  the  precipitator is operated dry.  This danger is  increased
by the inevitable sparking that occurs  between high  voltage  elements in
precipitators.   In  general,  however, because  the exhaust gases are usually
cooled by  water  sprays  at the augmenter, the  gas leaving a test cell often
contains  liquid  water in  the form of entrained droplets and  condensation.
       This water creates several severe problems for  the designer of an
ESP for a  test cell.  First, the damp or wet  particulates make a sticky
paste that clings tenaciously to collecting surfaces.   Some  effective
means of removing this  material must be provided.  Secondly, and more
seriously, the carbonaceous  particles,  when dampened with water, make a
                                                                         f
very conductive  paste.   If conductive material is allowed to build up on
insulating components inside the precipitator, arcing  and system failure
may occur. Some  means  of periodically  cleaning all  internal parts of an
ESP designed for  test cells  is  therefore necessary.  Wet ESP's, that
                                      51

-------
operate with a continuous flow of water over  the  collection plates, may
solve these problems.  However, there  is  no present  experience with wet
ESP's on test cells.
6.1.1  ESP Experimental Apparatus
       An experimental program to evaluate the performance  of an electro-
static precipitator has been conducted at Jacksonville  Naval  Air Station
by United Engineers and Constructors,  Inc.  A final  report  on the results
of the program is due to be  issued  in March 1976.
       The test cell which has been used for  the  ESP tests  was Black Point
No.  1 at Jacksonville Naval  Air Rework Facility.   This  is the same test
cell on which the prototype  nucleation scrubber is mounted.   As shown in
Figure 12, a bypass port is  provided in the back  wall of the  test cell.  A
12 inch by 12 inch duct extends into the cell exhaust area,  approximately
coaxial with the augmenter.  A venturi section immediately  outside the
cell is used to measure the  gas flow through  the  experimental  test setup.
The  gas stream is then expanded to  30  inches  by 48 inches,  the cross-
section of the ESP.  A 25-foot length of duct stabilizes the  gas flow, at
which point a sampling port  is provided.  After passing through the pre-
cipitator, the gas enters another section of  straight duct,  toward the end
of which is another sampling port.  The gas is then  exhausted to the
atmosphere.
       The gas flow through  the test duct is  powered by the  positive pres-
sure in the test cell exhaust space.  No auxiliary blowers  are required,
but note that the test duct  expands from 12 inches by 12 inches at the gas
takeoff to 30 inches by 40 inches at the ESP.  This  is  because the gas
must be slowed down for effective operation of the precipitator.  For most
of the tests that were made, the gas flowrate in  the test section was about
                                     52

-------
              AUGMENTS* TUBE
    TFST EM6INE -I
cn
CO
                                                                                                       66' 0"
                                                                        8'0       8'0"     6'0   37
                                                                    FALSE BOTTOM .
                                                                    IN DUCTWORK
CAS DISTRIBUTION.

  BAFFLE
.PROTOTYPE
 PRECIPITAT0R
OUTLET SAMPLING
    POUT
                              Figure  12.   Electrostatic precipitator experimental  setup  -Black
                                             Point  No.  1  test cell  (Reference 9).

-------
9000 scfm.  Since gas flow in the test cell  stack  is  about 600,000 scfm,
the test  apparatus was about 1/65 of full  scale.
       Figure 13 shows details of the ESP, which was  provided by American
Air Filter, Inc.  As shown in the plan view, the precipitator is a double
two-stage ESP.  In effect, it is two identical  units  in  series.   This
design increases the overall collection efficiency of the  device (compared
to a single ESP).
       The dirty gas enters the precipitator through  a distribution plate
and flows past wash nozzles to the ionizing  section.   The  ionizer section
consists  of a parallel array of tungsten wires, separated  by  grounded
metal plates.  The wires are raised to a potential (about  13,000 V) high
enough to cause ionization of the air near the  wires,  and  movement of air
ions in the space between the wires and the  grounded  metal  plates.  Ions
attach themselves to particles passing through  this region  of high-ion
density,  giving the particles an electrostatic  charge.
       The gas, carrying the charged particulate,  then enters the collect-
ing section.  This is a parallel array of metal plates which  are alter-
nately grounded and raised to about 6000 volts.  The  electrical  field
between the plates in the collecting section causes the  charged  particles
to migrate to one of the plates, where they  are deposited.
       After leaving the collecting region, the gas passes  through an
identical charge/collection section, where some of the particulate not
collected in the first stage is trapped.  The gas  then leaves the
precipitator.
       After the ESP has been in operation for  some period  of time, the
accumulated deposit on the collection plates and on the  insulating compo-
nents must be removed.  The method used in the  Jacksonville tests was
                                     54

-------
CJ1
en
                         1.
                         2.
                         3.
                         4.
                         5.
                         6.
                         7.
                         8.
                         9.
                        10.
                                            Elevation
                                Legend
                                                                                                           Plan
Gas distribution baffle
Ionizing section
Collecting section
Trans former-rectifier
Wash nozzles
Deterpent pump
Solenoid valve
Wash control unit
Flexible duct connection
Support legs w/vibration isolators
                                    Figure  13.   Electrostatic  precipitator test  unit  (Reference  9).

-------
to direct a high pressure spray of water  and  detergent  at  the interior
parts of the ESP-  Wash nozzles were premanently mounted  inside the
precipitator, and a fully automatic sequencer performed the  washdown.   The
manufacturer of the ESP recommended washdown  every 8 hours.   Somewhat  more
frequent cleaning was found to be desirable.
       6.1.2  Experimental Results
       A fairly extensive series of tests of  the scale model  ESP were  per-
formed at Jacksonville.  Sampling for particulate was done using Method 5
procedures at the ports marked "inlet sampling port" and "outlet sampling
port" on Figure 12.
       Collection efficiency varied considerably from one test  to another.
ESP efficiencies ranged from over 95 percent  in some cases to very low or
even negative values  in others.  To add to the confusion, two different
types of efficiency data were collected during evaluation tests of the
scale model ESP-  The manufacturers of the ESP (American Air  Filter Com-
pany) performed single point isokinetic sampling measurements immediately
upstream and downstream of the ESP-  Their method measured dry  particulate
only; typically, efficiencies between 80  and  95 percent were  recorded.
       Another series of efficiency measurements was made by  Jacksonville
NARF personnel, using EPA Method 5.  For  these tests, particulate was
determined by combining the filter catch  with the impinger catch.  Con-
densibles were thus defined as particulate.   The Method 5 tests used a
20-point sampling grid at the inlet and outlet of the unit.   An average
particulate removed by the ESP of 59 percent was measured for all of the
Method 5 determinations.
       Comparison of the efficiency results obtained with the AAF method
and EPA Method 5 is not really meaningful, since dry particulate was being
                                     56

-------
measured in the first  case,  and  dry participate plus condensibles in the
second.  Also, the AAF tests were done at a single point,  rather than over
a grid.
       Finally, Jacksonville NARF personnel believe that inability to
maintain constant engine  loads  during the long times required for Method 5
sampling may  have affected  those data.  For purposes of discussion,  the
EPA Method 5  results will  be used, since their data were obtained using
equipment and procedures  specified by most regulatory authorities.
       Table  8 shows three  measurements on a J79 engine.  The operating
personnel at  Jacksonville consider these "good" data, representative of
the overall performance of  the  system.  The overall particulate removal
efficiency of the ESP  varied between about 52 percent and  65 percent for
the data shown in Table 8.
       There  are several  possible reasons for the low collection efficien-
cies  in these tests.   As  indicated in Table 8, the gas velocities at the
duct  were 700 fpm.  This  means  that at the ESP, they would have been about
500 fpm.  This velocity is  at the top of the range commonly used and may
have  contributed to low efficiency, especially for small particles.
Smaller particles move toward the collection plates at a lower velocity
than  do large ones.  As a result, small particles (such as those emitted
by turbine engines) are more likely than large ones to pass uncollected
through an ESP if the  gas velocity is too high.
       There  were several  operational problems noted during the evaluation
of the ESP device.  One was a recurring problem with shorting in the first
ionizing section.  Operating personnel believe that this problem was
caused by excessive water droplets carried over from the exhaust cooling
sprays, and that a mist eliminator upstream of the ESP would be necessary
                                      57

-------
                      TABLE  8.   EMISSION  DATA  FROM  THE  ELECTROSTATIC  PRECIPITATOR AT JACKSONVILLE  NAVAL
                                AIR  STATION, BLACK  POINT  NO.  1  TEST CELL  -  J79  ENGINE
on
CO

Date
Scrubber water flow at #1 sump gal /mi n
Vol. of dry gas sampled, scf
Stack flowrate, scfm, dry
Stack gas velocity, at stack
conditions, fpm
Moisture, % by volume
Stack gas temp, degree F.
Isokinetic sampling rate, %
Parti cul ate Results
(a) Probe and filter catch
grains/scf dry, x TO"3
(b) Total catch
grains/scf, dry, x 10~3
(c) Parti cul ate from #1 sump water sample
grains/scf x 10~3
Particulate Removal Efficiency
Based on air sample (a), %
Based on air sample (b), %
Based on total (air and water, a+c), %
Based on total (air and water, b+c), %
Entrained Water Removal, %
Inlet Outlet
4-17 4-17
9.66
92.79 81.66
9276.0 8149.4
763.8 665.8
15.75 15.15
132 131
97.5 97.6
6.17 2.90
7.30 3.54
5.41
53.0
51.5
74.9
72.1
11.5
Inlet Outlet
4-18 AM 4-18 AM
9.89
87.99 83.31
8404.3 7621.5
703.8 620.5
17.13 14.89
132 131
102.0 106.0
3.48 1.05
3.92 1.20
5.64
69.8
64.9
88.5
86.8
44.8
Inlet Outlet
4-18 AM 4-18 AM
9.61
84.27 84.82
8607.6 7920.2
698.0 698.3
16.66 15.39
132 131
95.0 104.3
3.14 1.05
3.21 1.27
5.14
66.6
60.4
87.3
84.8
22.7

-------
for a full-scale unit.   There  was  also a problem with the automatic wash-
ing system:  an increase in  water  pressure over that originally provided
was required for satisfactory  operation.  After proper water pressure was
supplied, the cleaning  system  appeared to work satisfactorily.   The clean-
ing cycle took about  20 minutes  and appeared to remove deposited particu-
late from the ESP  as  it was  designed to do.   The precipitator could be run
immediately after  the cleaning cycle.   If the ESP was allowed to stand
unused for a long  enough time  to completely dry the  collection  plates,
however, residual  particles  not  removed by the automatic  cleaning system
would be reentrained  into the  gas  stream and substantially lower the
apparent efficiency in  subsequent  test runs.
6.1.3  Costs
       United Engineers and  Constructors, Inc., have estimated  the capital
and operating costs for retrofitting test cells with electrostatic precip-
itators capable of attaining performance levels similar to those reported
here.  The following  is quoted from Reference 9:
           "Capital costs for  a  pollution abatement  system incorporating
       the precipitator concept  would  be on the order of  $850,000 for a
       500,000 acfm system or  $1,690,000 for a 1,200,000  acfm system based
       on a 500 fpm design velocity.
           Annual  operating  costs  for  a pollution abatement system incor-
       porating the precipitator concept can be on the order of $32,000
       for a 500,000  acfm system testing 500 J79 engines  per year or
       $65,000 for a  1,200,000 acfm system testing 500 350-lb/sec engines
       per year."
       These costs are  based on  complete installation of  a double two-
stage ESP system,  including  an automatic washing system and a pressure leaf
                                      59

-------
filler for removing participate from the washwater.  The operating costs
include maintenance and electric power consumption.  Operating  personnel
at Jacksonville NARF have stated that they expect much higher maintenance
cost, based on their experience with the scale-model ESP, but they could
not provide expected maintenance cost figures.
6.2    NUCLEATION SCRUBBER
       At the present time, only one test cell cleanup method has  been
evaluated in full scale tests.  This is the Teller Environmental Systems,
Inc., (TESI), nucleation scrubber installed on the Black Point  No.  1  test
cell at the Jacksonville Naval Air Station.  This scrubber consists of
three parts:  (1) an exhaust gas pretreatment section, in which the size
of the particles is increased by nucleation of water vapor;  (2) a  proprie-
tary packed bed scrubber section where the particulate is transferred to
the scrubber water; and (3) a water cleanup and sludge removal  system (not
provided as part of the Jacksonville prototype scrubber facility).
       A schematic diagram of the prototype system installed at Jackson-
ville is shown in Figure 14 (Reference 10).  Exhaust gases leaving  the
engine are carried into the augmenter tube, where quench water  is  injected.
This water changes the dew point of the gas to the proper conditions  for
nucleaction. Nucleation occurs in the augmenter tube and stack.  The
exhaust gas then passes through a packed bed scrubber section where irri-
gating water removes the particulate from the gas.  Dirty scrubbing water
is piped to a nearby river.
       A special augmenter was designed and installed in the TESI  system
to keep the volume of exhaust gases to a minimum.  With conventional
augmenters, the ratio of entrained air to engine exhaust gas is about 2.5
                                     60

-------
     Scrubber
     water
     inlet
Augmenter
    tube
ttttt
Exhaust gases
                          Packed
                          bed
                                            Demist
                                            section
                                                   Wind
                                                   screens
                                                           Used
                                                           scrubber
                                                           water
           Figure  14.  Nucleation scrubber schematic.
                             61

-------
in afterburner mode.  With the special  augmenter  (Figure  15),  the flow is
choked down to give a ratio of 1.0.  This  significantly reduces the flow-
rate of gas to be cleaned, which means  that the superficial  scrubber area
is also smaller.  The TESI augmenter is also designed  for efficient quench-
ing of the exhaust gas stream.
       The augmenter is designed with a converging  throat and  a vena con-
tracta section with pentrating radial sprays and  width large enough to
permit recycle.   It also has a diverging zone with  circumferential  sprays
and an annulus between the diverging zone  and the outside shell  to  permit
recycle of gas and entrained water droplets to the  vena contracta zone.
The quantity of water injected and the  engine exhaust  and entrained air
flows are such that the submicron particles emitted by the engine act as
condensation nuclei for water vapor in  the saturated or supersaturated
regions downstream of the augmenter.  In this way,  particles grow to a
size of 3 to 5 microns (Reference 11).
       After leaving the augmenter, turning vanes direct  the exhaust gases
vertically up the cell stack.  At the top  of the  stack, other  turning
vanes direct the  gas horizontally into  two packed scrubbing  beds con-
structed on the periphery of the stack.  Each bed is 30 feet long by 18
feet high by 6 feet thick.  The inner 5-1/2 feet  of each  bed consist of
2-inch Tellerettes ( a proprietary packing material),  irrigated  with face
sprays and overhead distributors.  The outer 9 inches  is  a demisting sec-
tion.   It is packed with 1-inch Tellerettes and is  not irrigated.
       The maximum irrigation rate of the  two packed beds is 8300 gpm (used
during operation of the J79 engine in the  afterburner  mode).   This  flow  is
reduced  to 400 gpm in military mode, and only quench water is  required
below cruise levels.  Because of the high  cost of electricity  to provide
                                     62

-------
               Augmenting
                  air
           Engine exhaust
CO
              Augmenting
                     air
Outlet flow
                         Figure 15.   Teller  Environmental  Systems,  Inc. augmenter.

-------
scrubber irrigation water at maximum irrigation  levels,  it  is  important to
reduce water flowrates as much as possible.
       Unfortunately, accurate particulate emission data from  the  proto-
type scrubber is not available.  Sampling at the face of the scrubber  bed
is a formidable task, and has only been attempted once.  For this  one
test, multiple sampling nozzles were arranged in parallel at 12 test
points on the outlet side of the paced bed.  Isokinetic sampling condi-
tions were not maintained.  .Instead, a constant nozzle suction velocity of
8 ft/sec was held.  Under ideal conditions, exhaust gas exit velocities of
6 to 10 ft/sec were expected.  In reality, however, the wind caused severe
disturbances.  Air sampled by the nozzles were carried through manifold
pipes to a sampling train.
       A sampling run consisted of the following sequence:  (1) start
flushing water flow; (2) engine start to idle; (3) engine runup to desired
thrust; (4) take sample; (5) engine runup to blow out gases; (6) engine
shutdown; (7) stop flushing water flow.  During all this time the  sampling
nozzles were in place.  Although the gas samples were drawn through the
nozzles for only a part of this time, particulate material could
accumulate on and in the nozzles and their manifold during the entire
time.  Since the material accumulated in the nozzles and manifold  could
not be distinguished from the rest of the sample, accumulation was added
as an unwanted but unavoidable portion of the sample.  In addition, a
great deal  of water was sprayed into the gas stream upstream of the
scrubber and the intake nozzles.  During the sampling, some of this water
could be seen passing down the flexible tubing to the wet impinger, where
it accumulated,  contributing to the sample.
                                     64

-------
        With these sorts of problems, the  validity  of  the  participate  sam-
pling data obtained is questionable.  The  results are  shown  in  Table 9.
      TABLE 9.  NUCLEATION SCRUBBER AT  JACKSONVILLE NAVAL  AIR STATION,
                EMISSION LEVELS  FROM  BLACK POINT NO.  1  TEST  CELL
Engine
J79


TF30


J79



Mode
Idle
Normal
Military
Idle
Normal
Military
Idle
Normal
Military
Max. A/B
Part icu late
Emissions
Grains/scf
0.0024
0.0029
0.0024
0.0019
0.0014
0.0018
0.0052
0.0029
0.0062
0.0033
Ringelmann
Less than 1/2
H ii n
ii n n
n n n
n n n
n n n
n n n
n n n
n n n
n n n
       Attempts  were  also  made  to determine  the  concentration  of  CO,  S0£
and hydrocarbons with the  same  sampling system used for  particulate col-
lection.   Because of  various  equipment  failures  and analytical  difficul-
ties, reliable gaseous emission data were not obtained.
       In  spite  of the lack  of  reliable quantitative data,  independent
observers  were  impressed by  the performance  of the TESI  nucleation  scrub-
ber system.   The following is quoted from Reference 10.
            "During a  visit to Jacksonville Naval Air Station  by the Con-
       tractor,  a TF30 was tested in the cell.  The test cell  cleanup
       system was run under  two modes of operation, scrubber  water  on and
       scrubber  water off.  One immediate effect of the  presence  of
                                      65

-------
the scrubber system regardless of the operating mode was the
lowering of the noise level.  In an adjacent cell, a TF411 engine
(15 000 Ibs thrust) was being run at military rating.  The noise
was such that even at a distance of over 500 feet conversation was
almost impossible.  When this engine was shut off and the TF30 run
at military rating it was possible to converse within 50 feet of
the test cell without difficulty.
    The test cell was first operated with the scrubber water in
use.  During engine idle it was noticed that despite the 1100 gpm
being injected into the augmenter tube and the 8000 gpm being
showered down on the scrubber section, the exhaust (gas) was
noticeably dirty.  Yet the scrubber exhaust water was only slightly
blackened by the collected particulate matter.  Little, if any,
water vapor appeared at the scrubber exhaust ports.  Since one side
of the scrubber section contained only the demisting section, this
side tended to produce the water vapor mist.  Considerable dirtied
water was being exhausted from a tube placed in the rear of the
augmenter tube during idle.  This seemed to indicate that the
injected water had condensed prior to exiting the augmenter tube.
    As the engine power was increased to military rating, a
dramatic difference in the exhaust was noticed.  The scrubber
exhaust became black with particulate matter.  The water ceased
flowing from the augmenter tube drain and air began being sucked
into the rear of the augmenter tube through the drain.  This last
occurrence would indicate that the water vapor no longer was
condensing in the augmenter tube but was being carried further
downstream.  The exhaust from the test cell became quite clean and
heavily laden with water vapor.  By standing under the plume, a
slight mist could be felt.  Nevertheless, it was mentioned that
there had been no ill effects experienced from the mist at the
base.  Prior to installing the cleanup system, the vapor plume
would extend hundreds of feet and the resulting mist and noise were
objectionable.  Despite the lack of the 5-1/2 feet packing on one
side of the scrubber, the plume from both sides seemed to be
equally dense and equally as long.
    For the second phase of the test, the 800-gpm scrubber water
was turned off with only the 1100 gpm of augmenter water being left
on.  At idle power the same phenomena of dirty exhaust gas and the
slight wisps of water vapor were again observed.  The scrubber
exhaust water (the water vapor carried downstream from the aug-
menter tube and being condensed without benefit of scrubbing water)
was much dirtier than observed earlier at idle when the scrubbing
water was on, but the flow was less.  The augmenter tube exhaust
water exited through the drain was just as dirty to the naked eye
as observed previously.
    Lastly, the engine was  increased to military rating.  The
scrubber exhaust behaved as before at military rating, producing  a
100-foot plume of clean water mist.The augmenter tube drain
behaved as before, ceasing  flow and sucking air in.  Only the
scrubber exhaust water changed characteristics.   It became dirty  as
before, but the flow increased from a light flow experienced under
idle rating to a more moderate flow under military rating."
                              66

-------
       Table 10 presents  a  summary of these observations.   Note that the
performance of the  scrubber was  observed to be about the same with and
without scrubber  irrigation water.  Personnel  at Jacksonville NAS have
also suggested that the  amount of irrigation water and the thickness of
the packing can probably be considerably reduced without lowering system
performance.  This, if confirmed, would have some effect on capital  cost,
and a major effect  on operating  costs.
       Personnel  from Acurex and EPA viewed the scrubber system in opera-
tion on February  4, 1976.   A TF41 engine was being tested  in Black Point
No. 1, and a similar engine was  being tested in a nearby cell,  which was
identical except  for the scrubber.  The scrubber irrigation water and
quench water were turned on first.  The engine was then started,  held
briefly at idle,  and then moved  up to the 90-percent power setting where
it was held for about 5  minutes.  A very heavy white plume was  produced at
the faces and top of the scrubber.  The plume  lasted for about  100 feet
downwind  in the very light  winds prevailing.  After the steam plume  evap-
orated, no trace  whatever of a pollutant cloud could be detected.  A very
rough visual estimate was made that perhaps 20 to 30 percent of the  steam
produced  came from  the top  of the scrubber assembly (presumably through
leaks) rather than  through  the faces of the scrubber beds.   The color of
the plume coming  from the top was no darker than the plume  coming from the
faces.
       After a few  minutes  of operation at 90-percent power setting,  the
engine was reduced  to idle  setting for  about 5 minutes more.  At  idle, the
velocity  of the plume from  the scrubber was markedly lower,  but it looked
about the same as the plume from the 90 percent power setting.   A clean
white plume persisted about 50 feet downwind,  where the water evaporated
                                      67

-------
TABLE 10.   OBSERVATIONS DURING TEST CELL  CLEANUP SYSTEM TEST
           (Reference 10)
Scrubber Mode
Scrubbing water
on (8000 gpm)


Scrubber water
off


Engine Mode Test Cell Exhaust
Idle Dirty mist
Water vapor wisps
through unpacked
sides
Military Clean cloud
Water vapor plume
about 100 ft long
Idle Dirty mist
Water vapor wisps
through unpacked
sides
Military Clean cloud
Water vapor plume
about 100 ft long
Scrubber
Water Exhaust
Slightly dirty
Heavy f 1 ow
Dirty
Heavy flow
Slightly dirty
Light flow
Dirty
Medium flow
Augmenter Tube
Water Exhaust
Dirty
Heavy flow

No flow, air
being sucked in
Dirty
Heavy flow

No flow, air
being sucked in
                            68

-------
No trace of a residual pollutant  plume  could  be  observed.   Relative  humid-
ity was very high, however,  and this  made  steam  plumes  very dense  and  per-
sistent.  By the time the  stream  dissipated the  plumes  were greatly  spread
out.  This may have contributed to  our  inability to  detect  a residual  pol-
lutant plume.
       After the scrubber-equipped  cell  was shut down,  a series  of test
were begun in the adjacent uncontrolled  cell  on  a nominally identical  TF41
engine.  This provided an  interesting opportunity to compare plume appear-
ance directly.  At idle, the plume  from  the uncontrolled cell was  essen-
tially invisible, with only  a slight  darkening apparent.  At 100-percent
power setting, a very definite dark plume  was visible.  Several  Ringelmann
readers present agreed that, although proper  conditions for accurate opac-
ity determination were not present, the  plume opacity corresponded to
approximately 20 percent.  The excellent scrubber performance must there-
fore be weighed against the  fact  that a  relatively clean engine  was  being
tested.  Jacksonville NARF personnel  observed that the  TF41 opacity  of 20
percent at military power  setting was consistent  with past  experience  at
the relatively small, close-coupled,  metal Black  Point  cells.  They  said
that if the same engine were to be  tested  in  a longer concrete cell  a  few
hundred yards away, the plume at  military  power  would have  an opacity  of
at least 40 percent.
       Figures 16 and 17 show the Black  Point No.  1  test cell and  nuclea-
tion scrubber in operation.   These  pictures were  taken  in 1971,  but
represent the plume appearance we saw fairly  accurately.  Figure 16  also
shows the ductwork extension out  the  back  of  the  cell that  was used  to
test the scale model scrubber and electrostatic  precipitator.
                                     69

-------
Figure 16.   Nucleation scrubber at Jacksonville  NAS.

-------
CO
                              Figure 17.   Aerial view of Jacksonville scrubber.

-------
6.2.1  Scale Model Scrubber  Tests
       Even through good  sampling  data from the full-scale prototype
scrubber is lacking, considerable  information  has  been  obtained  on  a  1/50
                                                           /
scale model known as the  pilot  scrubber,  attached  to  the  Black Point  No.  1
test cell at Jacksonville NAS.   Essentially the same  ductwork was used for
the pilot scrubber tests  as  was later  used  for the electrostatic model
experiments.  The differences were that for the scrubber  tests the  inlet
duct size was 4 feet x  4  feet in cross section and gas  flow measurement
was achieved with an adjustable orifice.  The  ESP  tests were conducted
with a 30 inch x 48 inch  inlet  duct and a venturi  upstream of the ESP.
Figure 12 shows the ducts which led engine  exhaust from the cell to the
scrubber.  Figure 18 (Reference 13) shows the  pilot scrubber.
       The pilot scrubber was designed expressly to overcome the sampling
problems noted with the full-scale prototype scrubber.  Gas velocities and
scrubber bed thicknesses  on  the pilot  scrubber were made  similar to those
on the full-scale prototype  unit.   Pilot  model sampling procedures  were
evaluated by EPA representatives,  and  were  determined to  be satisfactory.
It is therefore believed  that the  data obtained with  the  pilot scrubber
are representative of actual nucleation scrubber performance.
       In a sample run, an engine  was  set at a constant power level.  Air
flow through the pilot  scrubber was set by  adjusting  the  orifice at the
entrance to the system  ductwork.  The  flow  of  water in  the augmented  and
pilot scrubber were set and  stabilized.   Air sampling was  done at the
locations marked "inlet sampling port" and  "outlet sampling port" on
Figure 12.  At each sampling port, air samples were taken  at 25  locations
specified by a 5 x 5 grid uniformly spaced  within  the 4 feet by  4 feet
duct.  Isokinetic sampling methods and train components were as  required
                                      75

-------
                           Flowmeter
    Inlet gas
sampling port
         Drain  for
         Sump No. 1
                Outlet gas
                sampl ing .port
                             Packing box.
                              4' x 4' x 6'
                         Drain control valve
Drain for
Sump No. 2
   Figure 18.   Pilot  scrubber  system  on Black Point  No. 1  test cell  at the Jacksonville
                 Naval  Air Station.

-------
for Method 5.  Air was  sampled  before and after the scrubber section.  In
addition, water  samples were  taken at the locations marked "Sump No.  1"
and "Sump No. 2" on  Figure  18.   The sample from the No.  1 sump gave the
concentration of particulate  which was removed by the action of the water
sprayed  into the exhaust stream at the engine tailpipe,  by particle im-
pingement on the augmenter  tube and all  stack internals, and by the water
condensation and fallout that occured between the tail  pipe and the
scrubber face.
       Part of the difficulty in assessing the performance of any scrubber
system requiring exhaust gas  quenching is that the augmenter with its
water spray system removes  particulate independently of  the packed
scrubber bed.  Therefore, while the overall  system may be highly efficient
in removing particulate, the  scrubber section (as measured by air sam-
pling) may be much less effective.  The  relative particle-removing effi-
ciency of a fully quenching system, compared to a conventional augmenter
with water spray cooling is unknown.   In the data reported in Reference
13, and  repeated here,  both scrubber  efficiencies (based on air sampling)
and overall efficiencies (including particulate collected in the No.  1
sump) are presented.
       Table 11  shows  a representative set of data obtained with the  pilot
scrubber system  on three different days.  Particulate removal efficiency
of 51 to 62 percent  for the packing section, and overall efficiency,  in-
cluding both packing and augmenter spray collection, of  77 to 81 percent
were representative  of the  much larger number of data points given in
Reference 13.
       Among the parameters investigated in  the pilot model scrubber  tests
were type of packing,  effect  of packing  thickness, and effect of water
                                      77

-------
                     TABLE  11.   TEST  RESULTS  ON  PILOT  SCRUBBER AT THE BLACK POINT NO. 1 TEST CELL
                                AT  JACKSONVILLE  NAVAL  AIR STATION, USING A J79 ENGINE (REFERENCE 13)
00
1 	
Date
No. ft. of packing
Engine operating condition
Flow at #1 Sump gal./min.
Flow at #2 Sump gal./min.
Vol. of dry gas sampled, SCF
Stack flowrate, SCFM, dry
Stack gas velocity, at stack
conditions, f.p.m.
Moisture, % by volume
Stack gas temp, degree F.
Isokinetic sampling rate, %
Particulate Results
(a) Probe and filter catch
Grains/SCF, dry, x 1CT3
(b) Total catch
grains/scf, dry, x 10 3
(c) Particulates from #1 Surnj^ water
grains, scf, x 10'3
Particulate Removal Efficiency
Based on air sample (a), %
Based on air sample (b), %
Based on total (air and water, a+c),
Based on total (air and water, b+c),
Entrained Water Removal, %
Inlet
1-30
3
Normal Rating
12.3
.2
73.04
6539.8
542.4
16.6
129.6
104.4
6.90
7.27
sample
10.69
55.2
53.1
% 82.4
% 81 . 0
42.5
Outlet Inlet
1-30 1-31
3
Normal Rating
16.2
.5
62.75 91.46
5448.0 8919.5
437.1 746.1
14.1 17.4
126.8 130.2
107.7 95.9
3.09 6.67
3.41 6.85
9.33
62.0
60.9
84.2
83.5
75.3
Outlet Inlet Outlet
1-31 2-1 AM 2-1 AM
3
Normal Rating
14.7
75.0
85.15 91.89 75.01
8075.4 9046.4 7510.6
638.8 760.4 566.4
12.9 17.7 10.5
127.6 131.0 115.4
98.60 95.0 93.4
2.53 6.55 3.20
2.68 7.06 3.74
9.30
51.2
47.0
79.8
77.2
67.0

-------
irrigation rates over the packing.   It was found that none of these vari-
ables, within the ranges tested,  significantly affected the scrubber bed
collection efficiency for particulate.  These findings suggest that the
full-scale prototype scrubber  and the  four new scrubber-equipped  cells
under construction at Jacksonville  and Norfolk may be overdesigned.   All
were designed with thick packing  beds  and high scrubber water irrigation
rates, which may not be necessary.
       Gaseous emission measurements of the pilot model scrubber  were con-
ducted by personnel from the Aircraft  Environmental  Support Office,  Naval
Air Rework Facility, North  Island,  California.   Measurements of carbon
monoxide, carbon dioxide, nitric  oxide and total nitrogen  oxides  were made
at both the inlet and outlet sides  of  the scrubber.   The effect of water
flowrate and packing depth  on  gas concentrations was investigated.   For
all tests, the average change  in  the concentration of each of the gases
from the inlet to the outlet sides  of  the scrubber varied  from 0  to 5 per-
cent.  It was concluded that the  scrubber had little or no effect on the
removal of CO, C02, NO, and NOX  (Reference 14).
6.2.2  Proposed Scrubber Facilities
       Based on the results obtained with the prototype scrubber  on the
Black Point No. 1 test cell, four new  scrubber-equipped test cells  are  now
under construction — two at Jacksonville NAS and two at Norfolk  NAS.
These are very large cells, designed for engines in  the 70,000-lb.  thrust
category, whereas typical Type A  cells are designed  for 20 to 25,000 Ib.
thrust engines.  The cells  themselves, therefore, and the  associated
scrubber systems are considerably larger than the Black Point No. 1 cell.
       The new cells will be equipped  with complete  scrubber installations,
including water cooling and water treatment systems  that
                                      79

-------
allow closed-loop water recyle.  Each cell scrubber bed will be  supplied
with 14,000 gpm of irrigation water, compared to 8,000 gpm maximum  for  the
prototype scrubber.  Initial engine testing at the new Jacksonville cells
is scheduled for fall 1976, with full operation scheduled for December
1976.  The Norfolk cells are approximately 1 year behind.
6.2.3  COSTS
       Since a full scale prototype retrofit scrubber has been in opera-
tion for several years, and four new scrubber-fitted test cells  are near-
ing completion, it might be thought that the cost of providing scrubbers
for retrofitting existing test cells could be readily established.   Unfor-
tunately, this is not the case.  We requested separate estimates from the
Navy and a private contractor and found these to differ considerably.
These differences existed despite the use of a uniform cost estimation
basis through the stipulation of the following conditions:
       •   The scrubber was to be retrofitted to a Type A permanent  test
           cell capable of testing engines of up to about 20,000-lb. thrust
       •   A complete facility was to be provided including scrubber,
           cooling tower, water cleanup plant, all site modifications,  and
           installation of all equipment
       §   Costs of bringing utilities (water, electric power) to the test
           cell site were not to be included
       Three cost estimates were received.  An estimate prepared by the
Naval Air Systems Command (Reference 15) in early 1975 is shown  in  Table
12.  A separate estimate prepared in January 1976 by personnel at Naval
Air Rework Facility, Jacksonville, is shown in Table 13.  A very detailed
cost proposal prepared by the original system contractor for the
                                     80

-------
 TABLE 12.  SCRUBBER RETROFIT COST ESTIMATE -- NAVAL AIR SYSTEMS COMMAND
                         Item
                 Cost
           Packed scrubber
           Cooling tower basin
           Cooling tower
           Water system for irrigation and quench
           Water treatment plant
           Exhaust stack modifications
           Electrical work
             $  515,000
                160,000
                250,000
                280,000
                160,000
                 50,000
                100.000
                                             Total   $1,515,000
      TABLE 13.  SCRUBBER RETROFIT COST ESTIMATE -- JACKSONVILLE NARF
          Item
           Cost
 1.   Cooling tower well

 2.   Water treatment system

 3.   Scrubber and quench water
      system including cooling tower

 4.   Cooling tower and water
      treatment stairs

 5.   Water treatment building
      (miscellaneous)

 6.   Columns, beams, and
      miscellaneous for cooling
      tower pipe support

 7.   Cooling tower and water
      treatment building --
      general construction work

 8.   Upper concrete work for
      scrubber enclosure including
      additional foundation

 9.   Additional piles for
      building structure


10.   Extra electrical work on
      outside and inside of cell


                           Total
$   31,138

   250,000

   731,000


    21,995


     1,560


    60,450



   251,515
   250,000 (due to pollution  abate-
           ment).   Prorated from a
           larger  figure.

    40,000 (due to pollution  abate
           ment).   Prorated from a
           larger  figure.

   307,077 (due to pollution  abate-
           ment).   Prorated from a
  	 larger  figure.
$1,944,735
                                    81

-------
Jacksonville prototype scrubber, Teller Environmental  Systems,  Inc.,  is
reproduced in Appendix A.
       Direct comparison of these three cost estimates  is  difficult,  be-
cause equipment costs have been grouped in different ways.   Examination of
the bottom line costs (Table 14) shows considerable variation,  however.
It is beyond the scope of this task to specify which,  if any, of  these
estimates is "correct."

              TABLE 14.  COMPARISON OF SCRUBBER COST ESTIMATES
                                                 Complete Cost
                 Estimate                      One Cell Retrofit
              Jacksonville NARF                    $1,944,735
              Naval Air Systems Command             1,515,000
              Teller Environmental Systems, Inc.      705,650
       Although some inconsistencies may be noted between the two Navy
estimates, the main difference between their analyses and the TESI
estimate reflects a difference in the quality of construction specified,
particularly for supporting and ancillary structures.  The Navy  estimates
are based on construction of reinforced concrete stack modifications  and
concrete structures to house the water pumping and water treatment  equip-
ment.  The TESI estimate is based on provision of minimum construction
consistent, in their view, with satisfactory equipment performance  and
reliability.  A more detailed cost breakdown and analysis will appear  in
the Final Report to Phase II of this study.
       An estimate for annual maintenance and operating costs has been
made by Naval Air Systems Command (Reference 15).  Based on two  engine
                                     82

-------
tests per day, 250 days  per  year,  annual  maintenance was estimated at
$50,000 and annual operating costs for the scrubber and cleanup system
were roughly estimated at  $80,000  (operating costs will vary with type of
engine and length of  tests).  The  electrical power cost, which is the
major nonlabor operating cost,  is  $18,400 per year, based on 300 hp pump
capacity, two tests per  day, 250 days per year,  and $0.035 per kw-hr (the
cost of power at Jacksonville NAS).   This assumes maximum water flow at
all times.  In fact,  maximum water flow is used  only at afterburner engine
settings, which constitute a very  small fraction of test hours.  Also,  as
noted previously, there  is reason  to believe that the water flowrates
originally specified  for the system can be reduced with no loss of perfor-
mance.  These considerations indicate that actual power costs will be less
than $18,000 per year.
6.3    FUEL ATOMIZATION
       The fuel atomization  improvement concept  for reducing jet engine
emissions involves attempts  to  introduce  fuel into the engine in a more
highly atomized state than is presently the case.  The premise is that
this finer atomization would bring about  more efficient combustion and,
hence, reduced emissions.  This  concept has been experimentally evaluated
in two distinct forms -- methane absorption and  water or alcohol  emulsion.
Both programs have been  sponsored  by the  Naval Air Engineering Center,
Trenton, New Jersey.
6.3.1  Methane Absorption
       A series of experiments  has been conducted in which methane gas  is
dissolved in the fuel at high pressure.  When the fuel stream is exposed
to the low pressure and  elevated temperature existing just downstream from
the fuel nozzles, the methane is expected to flash out of solution and,
                                      83

-------
hence, aid in the atomization process.  In fact,  it was observed  that the
methane gas tended to slowly bubble out of solution when  the  pressure was
reduced, rather than violently disrupting the  liquid as had been  antici-
pated.  In a series of tests using a single burner, methane concentration
was varied between 0 and 2 percent by weight in JP5 or Jet A  fuels.   Pol-
lutant emissions monitored were NO, NO   UHC,  CO, C0?, and smoke  num-
                                      /\             £•
ber.  No significant change in any of these emissions was observed when
methane gas was added to the fuel.
6.3.2  Emulsions
       In this concept, an emulsion of an immiscible liquid with  the  reg-
ular fuel is created.  The liquid in the discontinuous phase  (water or
ethanol in the experiments conducted to date)  is  in the form  of small
globules which flash after the fuel stream leaves the nozzles.  This
vaporization of the volatile globules disrupts the fuel stream, leading to
finer atomization and more complete combustion.
       A series of combustion tests using a single burner indicated that
the presence of water or alcohol droplets in the  fuel stream  can  signifi-
cantly affect the pollutant emission (Reference 3).  A standard milk
homogenizer was used to prepare emulsions of water or ethanol  of  0 to 10
percent by volume in JP5.  Table 15 shows a representative sample of  the
results obtained.  It should be noted that the energy value of the fuel is
reduced proportionately to the water concentration.
       These results were reasonably repeatable in similar tests.  The
results show that opacity and particulate concentration were  significantly
reduced, NOX was slightly reduced, CO was nearly  unaffected,  and  unburned
hydrocarbon significantly increased.  The data in Table 15 raise  several
                                     84

-------
            TABLE 15.  EFFECT OF WATER/FUEL  EMULSION ON  COMBUSTION
Fuel
JP5
JP5
Water
Cone.
0%
10%
NOX
Ppm
66.5
54.5
UHC
ppm
2.5
5.3
CO
%
0.038
0.039
Smoke
No.
33.6
23.8
Particulate.
Conc., gm/m
5.55
3.11
 questions.  The reduction in smoke number and particulate with the water
 emulsion are consistent with the proposed mechanism of better atomization
 and combustion, but the increase in unburned hydrocarbon is difficult to
 understand.  It is also not clear whether the reduction in NO  is due to
                                                              A
 the finer atomization, or simply due to the lower combustion temperature
 of a 10-percent water in JP5 fuel mixture compared to straight JP5.
        There are two main problems inherent in the implementation of a
 fuel emulsion system for test cell use, even if further testing confirms
 the lower smoke numbers and pa£tjj:uj!ie-jnassJ4>^^             the spe-
 cific gravity pf-a-iO^pefcent water in JP5 emulsion is sufficiently
 ferejj^from straight JP5 that modification to the engine fuel control
/•system may be required for some engines.  Second, the engine thrust and
 other performance parameters are different for an emulsion than for JP5.
 Since the emulsion would be used only for test cell operation, careful
 calibration would be required so that an engine's performance in a cell
\sing an emulsion could meaningfully be compared to that same engine^:
 operation in the field on JP5.
        Despite these problems, the emulsion concept  is considered suffi-
ciently promising that further investigation, probably involving full-
scale  engine tests, are now being planned.
                                      85

-------
6.4    THERMAL CONVERTER
       The thermal converter, sometimes called an afterburner  (not  to be
confused with the afterburner stage of a jet engine), raises the  tempera-
ture of the exhaust stream with a secondary burner stage.  The  exhaust
stream is maintained at an elevated temperature for the  length  of time
needed to oxidize combustible components of the exhaust  — CO,  hydro-
carbons, and carbonaceous particulate to C02 and H^O.
6.4.1  Thermal Converter Configuration for Test Cells
       A thermal converter used in an engine test cell and installed  in  an
in-line arrangement takes the general form illustrated in Figure  19.   The
high-velocity exhaust jet from the engine and the induced secondary air
from the test cell both enter the cylindrical section called the  mixer.
In this section the two streams mix together to achieve  as uniform  a
velocity profile as possible.  The flow leaving the mixer is then diffused
to the full cross-sectional area through a simple conical diffuser.   This
reduces the average velocity of the stream to an acceptable level.  In the
burner, the required amount of fuel is added and ignited, and the result-
ing flame is stabilized.  Basically, the burner consists of a row of  fuel
spray bars followed by a row of vee-gutter flameholders.
       The converter is located immediately downstream from the burner.
It consists of a cylindrical section long enough to complete combustion  of
the added fuel and oxidation of the pollutants.  The region identified as
the combustion zone in Figure 19 represents the length required to  com-
plete combustion and attain a uniform temperature distribution.   In the
reaction zone, the temperature is maintained at a selected level  for  the
length of time needed to achieve the required degree of  pollutant
conversion.
                                     86

-------
00
                                                                                           Movable Sleeve
                                                                                Outer Liner
                     Tailpipe          Induced Cooling^

                           Induced Secondary Air

                                                          Combustion
                                                             Zone
           Reaction
             Zone
                                                 Burner
Converter
                                                                                     Cooler
                                                                     Water Spray
                                                                        Bars
                                                                                                      Back-Pressure
                                                                                                      Control Device
                                                                                                         (Exducer)
                                                            Stationary
                                                            '   Plug
                           Figure  19.   Schematic diagram of basic  thermal  converter  configuration
                                         (Reference 6).

-------
       The flow from the converter then passes through  a water  spray sys-
tem which consists of a row of pipes that inject water  into  the stream.
In the cooler, the evaporated water reduces the gas temperature enough  to
protect the exhaust stack.  Before entering the exhaust stack,  the  flow
passes through a variable-area device located at the exit of the cooler.
This consists of a stationary plug and a movable sleeve which can be used
to adjust the exit flow area.  This device serves two functions.  Its
primary function is to control the backpressure on the thermal  converter.
This makes it possible to adjust the amount of secondary induced air enter-
ing the converter to the desired level for any given engine.  The back-
pressure control device also acts as an ejector.  It induces cool ambient
air to flow through the annulus between the outer liner and the thermal
converter shell.  This air provides the necessary convective cooling of
the shell during operation.
       As described in Reference 7, the operation of a thermal  converter
installed in a test cell would be virtually automatic.  In starting  a test
sequence, the engine is first mounted on the test stand.  The converter's
adjustable-length mixer section is then locked into place as a  predeter-
mined distance from the engine's tailpipe.  The backpressure control
device is also locked into a predetermined setting.  Both adjustments
would be determined empirically and would be standard for a given engine
model.  For an engine being tested in the afterburning mode, the tempera-
ture of the exhaust is high enough to achieve oxidation of the  pollutants
without supplying additional fuel in the thermal converter, provided the
amount of induced air is maintained at a low value.  The burner  system is
designed so that upon command of the operator it can be retracted for
protection just prior to ignition of the afterburner and replaced at the
                                     88

-------
conclusion of the afterburning  sequence.   In  all  other  cases, the thermal
converter would be operated  automatically.   Its control  system would  ini-
tiate and adjust the flow of fuel  to  the  burner and water to the spray
system to maintain the required performance.
6.4.2  Pollutant Conversion  Performance
       The emissions of gaseous pollutants  and particulate from the ther-
mal converter have not been  adequately determined.  However, sufficient
testing has been done to show that the thermal converter is quite effec-
tive in reducing smoke number (that is, solid particulate is largely
eliminated).  It was also found that  NOX  emission  is not unduly in-
creased.  Specifically, data presented in Reference 6 show that a reduc-
tion in smoke number by a factor of 10 is readily  achieved.  While no
measurement of particulate mass loading was made,  the reduction in smoke
number probably implies a substantial reduction in particulate emission.
NO  emission from the thermal converter was measured at  about 1 lb/1000
Ib. fuel burned.  This is small  compared  to the NO output of most jet
                                                   A
engines.  The emissions of CO and  UHC from  the thermal converter were not
adequately measured.  But it seems unlikely that the converter would
markedly increase the emission  of  these pollutants over  emissions of the
jet engine itself, and it may actually lower  the concentration of CO and
UHC.
       Based on the information presented in  References  6 and 7, the
thermal converter would be effective  in greatly reducing the opacity of
the exhaust plume, would probably  significantly lower particulate emis-
sions, and would not greatly increase emission of  gaseous pollutants.
                                      89

-------
6.4.3  Costs
       Table 16 (Reference 6) shows the capital and operating  costs  for a
thermal converter retrofit to a standard type A test cell.   The  capital
cost is relatively low compared to other emission control methods  con-
sidered.  However, the operating costs are very high, due primarily  to the
high fuel consumption of the thermal converter burners.  As  a  rough  rule of
thumb, it requires as much fuel to operate the thermal converter as  it does
to operate the engine being tested.  It should also be noted that  the  costs
in Table 16 are based on a fuel cost of $0.35/gal.  It seems likely  that
this figure considerably underestimates fuel cost during the next  5  to 10
years.  The availability of the additional fuel that would be  required if
thermal converters were in wide use is also a matter of concern.
6.5    FUEL ADDITIVES
       Organometallic fuel additives used to modify the character  of emis-
sions from combustion processes have been studied for many years.  Many
additives, containing such metals as barium, manganese, iron,  calcium,  and
various rare earths,  are sold in the United States by a number of  manufac-
turers.  The primary market for these additives is conventional  fixed  power
generation equipment, although there is a growing market in  turbine-powered
processes.
       The emission modification produced by additives is apparently due to
the catalytic activity of the metallic component, although the exact mechan-
isms have not been determined.  Reductions in emissions of SCL (40 to  50
                                             ^—"*"    ___^—  ~ - ——
percent reduction with No. 5 fuel oil ),"^T~~(2Q to 30 percent  reduction
                                          x                        ^
with pulverized coal), and parti culate^^percent reduction with ^pulver-
ized coal) have recently been reported for stationary power  plants (Refer-
ence 1).  An additional benefit from using fuel additives is reduction
                                     90

-------
             TABLE 16.  THERMAL  CONVERTER  COST  ESTIMATE  (Reference 6)
    Capital Costs
                  Total Equipment  a                  $305,000
                  Assembly  and  Installation           88,000
                  Management                          84,000
                  General Contractor  Profit           14,000
                       Total  Installed Cost          $491,000

    Daily Operating Costsb
                         J52                     J57              TF30
   Fuel           $2010 (5740 gal)     $2760  (7890 gal)     $4960 (14,100 gal)
   Water             23 (188,000 gal)     25  (214,000 gal)     35 (289,000 gal)
   Power             11 (500 kW-hr)       12  (520 kW-hr)       16 (610 kW-hr)
   Purge Nitrogen 	6 (270 scf)      	8  (320 scf)      	4 (160 scf)
     Total Cost   $2050                $2805                $5015
    Includes material, fabrication, labor, pumps, and controls.
    Based on two 120-minute test cycles per day for each of the specified
    engines.
in plume opacity.   This  reduction  is  probably due  mainly to  a  change  in  the
particle size distribution,  rather than  to  a  reduction  in particulate mass
emissions.
       The first time  fuel  additives  were used  to  modify turbine  aircraft
engine emissions occurred  in Southeast Asia during the  1960's.  Some of  the
dirtier jet engines were at  a disadvantage  in combat  situations because
their trajectory could be  followed easily.  A manganese-containg  additive,
                                      91

-------
CI-2 (Ethyl Corporation), was quite successful in reducing the opacity  of
the exhaust plume.  For awhile, this additive was used extensively  in com-
bat operations.  But several serious problems with CI-2 led to curtailment
of its use.  CI-2 is toxic in bulk form and is easily absorbed through  the
skin.  Also the prolonged use of CI-2 damaged engines by causing a  buildup
of deposits on turbine blades and other engine parts.  These deposits lead
to hot spots and engine burnouts.
       In a number of programs, the Naval Air Propulsion Test Center
(NAPTC) has investigated usage of organometallic fuel additives to modify
jet engine emissions.  In one program, different additives were compared to
determine their ability to reduce plume opacity.  One engine (a J57-P8) was
used for all of the tests.  The test procedure was to gradually change  the
concentration of the additive being evaluated, while a Ringelmann reader
observed the opacity of the test cell plume.  The engine was operated at a
power level known to give the maximum plume density.  Since the Navy used
proprietary compounds and information in conducting the tests, the complete
ranking of the effectiveness of the various additive compounds was not
released.  The most effective compound (lowest percent metal in fuel to
produce a Ringelmann No.  1 reading) was CI-2.   Next in effectiveness, and
only slightly belpw CI-2, was Fe55 (Araphahoe Chemical Company), usually
referred to a^fFrroceneT^ Table 17 shows the additives evaluated and their
manufacturers.
       Because ferrocene is effective in reducing plume opacity, and be-
cause it is low in toxicity, a program to evaluate ferrocene in various
engines has been under way for some time at the Naval Air Rework Facility
at Alameda Naval Air Station.  A final report on this work is now in the
final stages of preparation at NAPTC.  It is scheduled for completion in
                                     92

-------
TABLE  17.   METALLIC FUEL ADDITIVES EVALUATED  BY THE
              NAVAL AIR PROPULSION  TEST  CENTER3
    Trade Name                           Manufacturer

 Smokeless #9                   Ellis Chemical, Inc.
 DGT-2                          Apollo Chemical Co.
 E/C 200                        G&M International
 Technol D                      Oil Technology Corp.
 Technol T                      Oil Technology Corp.
 Technol TT                     Oil Technology Corp.
 Fe55                           Arapahoe Chemical Co.
 RS 211242                      Arapahoe Chemical Co.
 RS 20521                       Arapahoe Chemical Co.
 PD 1471                        Arapahoe Chemical Co.
 PD 1472                        Arapahoe Chemical Co.
 CI-2                           Ethyl Corp.
 PSP MC955                      Petroleum Specialty Products Co.
 PSP MC900                      Petroleum Specialty Products Co.
 Cerium                         United Technology Research Center
 Rare Earths Mixture            United Technology Research Center

 aNot necessarily in order of effectiveness.
                             93

-------
June 1976.  Some of the findings of the report are presented here, along
with results from a recently presented paper (Reference 2).
  , "   It has been es^atTf^heTthaTmost Navy~engTnTs are~not-a^.etsely
Effected by the use of ferrocene additives, as long as the ferrocene con>
ckntration is kept at reasonable levels (0.1 percent) and additive use is
conf>ne4^to testing opej:atmo5_njiL_e^medlnj_2.._to 4JIQUES...-Table 18 shows
a list of the engines that have been tested with ferrocene additive.  For
each of the listed engine and model types, it has been found that ferrocene
additive at an appropriate concentration can reduce the test cell plume
density to Ringelmann No. 1 or below.  It was also found that ferrocene use
during testing is not injurious to engine performance or life; further
testing is required to determine the effects of continued or long term
use.  Currently, approved test durations with ferrocene range from 0.75
hour for the TF41 engine to 2 hours for the 052.
        TABLE 18.  ENGINES SUITABLE FOR USE WITH FERROCENE ADDITIVES
Engine Model Type
J79
J52
J57
TF30
TF41
T56
T56
aSuitability for
consideration.
GE8a
P6
P10
P6
A2
A10
A16
use with
Smokeless Combustion
Model Type
GE10-C
P408, P8B

P8, P408, P412



ferrocene additive is under
                                      94

-------
       Table 19 shows,  for  various  engines,  the observed stack opacity
without additive  at  the "dirtiest"  engine power setting.  The table also
shows the ferrocene  concentration  required to reduce the plume opacity to
20 percent  (Ringelmann  No.  1)  at  the same power setting.  These tests  were
conducted at the  Alameda Naval  Air  Station test cells.   Repeating the  same
tests at another  cell or with  a different engine of the same type might
give somewhat different results.   Also,  the  differences betweeen the find-
ings of Reference 2  and Reference  38 indicate that these test results  are
not very reproducible.   The tests  do indicate,  however, that adding small
concentrations of ferrocene to  the  fuel  can  significantly reduce plume
opacity for the engines tested.

            TABLE 19.   EFFECT  OF  FUEL ADDITIVE  ON STACK OPACITY
Engine
J57-P10
J52-6
TF41-A2
T56-A16
TF30-P6
J79-GE8
T56-A10
Maximum Smoke Opacity
(without additive)
40%
55%
40%
30%
35%
50%
35%
% Ferrocene for
Ref. 2
0.05%
0.10%
0.02%
0.02%
0.04%
0.04%
0.03%
20% Opacity
Ref. 38
0.05%
0.06%
0.02%
0.02%
0.06%
0.10%
0.03%
       The toxicity  of  the  additive  CI-2  has  been  a major  reason  why the
Navy has stopped using  it.   Concern  has also  been  expressed  about dis-
persion of manganese (the metallic component  of CI-2)  by the test cell  ex-
haust gases.  The toxicity  of  ferrocene,  the  additive  the  Navy feels is
                                      95

-------
most suited to replace CI-2, has been evaluated recently.  The NAPTC  re-
ceived a report of tests conducted at the U.S. Navy Toxicology Unit of  the
National Naval Medical Center.  Tests completed include:
       •   Ocular and dermal irritation with
           ~  Ferrocene (dicyclopentadienyl iron)
           --  Five percent ferrocene in JP-5
           —  Twenty percent ferrocene in toluene
       •   Inhalation studies with
           --  Ferrocene dust
           --  JP-5 with ferrocene
           --  Toluene with ferrocene
           --  Xylene with ferrocene
       •   Acute oral toxicity of ferrocene (l-D™)
       •   Acute peritoneal toxicity of ferrocene
The results of the study are summarized by the following quotation
(Reference 39):*
       "Ferrocene alone and in solution causes very little if any ocular
       irritation.  Irritation studies on both abraded and unabraded  skin
       were negative.  Rat ID™ for ferrocene is 1890 mg/kg  (oral)  and
       1520 mg/kg (intraperitoneal ).  Studies with ferrocene dust showed
       no significant toxic symptoms during and following exposures of  up
       to 150 mg/m .  Acute exposures to ferrocene and the solvents
       showed no increase in toxicity over the solvents alone."
*The ferrocene tested in this study was Arapahoe Fe 55R which  is  a
 specially purified form of ferrocene.  The intraperitoneal rat ID™
 data presented here are somewhat higher than reported by other
 laboratories with other sources of ferrocene.
                                     96

-------
  The health effects of emissions from  test  cells  in which CI-2 or ferrocene
  is  being used have not been determined.           _
         The effect of ferrocene additive on  gaseous and particulate emi
  si on has been studied in the Alameda  NAS tests.   For gaseous emissions
/ (CO, S0x> C02, NOX> UHC), the additive produced  no changes in pgj
                t i on .  Howev«*v-£^-i-wj-3^d4^^^             reduction in
  particulate emissions when ferrocene was used.  Figure 20 shows particle
  mass loading as a function of engine power setting, with and without the
  additive.  In Figure 20, NR refers to Normal Rating power level, and MR
  means Military Rating power level.  These data were collected at one sam-
  pling location at the exit plane of the test cell stack.  EPA Method 5
  sampling was used.
         One problem related to continuous (inflight) use of ferrocene is
  the reduced thermal stability with ferrocene-treated fuel.  A high temper-
  ature (400-500°F) thermal stress test indicated that ferrocene — an
  oxidation catalyst ~ causes significant increases in carbon and sludge
  deposition.  A similar, though less severe, problem occurred with the
  manganese additive.  As a result of an extensive experimental program at
  Wright Patterson AFB, the Air Force now believes that manganese based
  additives with minor amounts of a barium compound to promote stability
  offers the best compromise between effectiveness and reducing opacity,
  freedom from operational problems, and safety.  Air Force personnel
  believe that this additive is also capable of reducing test cell plumes
  from most engines in their inventory to 20 percent opacity or lower.
  6.5.1  Civilian Airline Additive Use
         Although this study was primarily concerned with military test
  cells and test cell pollution abatement programs, a brief investigation of
                                       97

-------
IO
CO
             
-------
commercial airline  practices  was conducted.   Letters were also sent to the
executives responsible  for  the  maintenance operations of the major air-
lines.  These  letters requested information  about any methods or devices
used to control test cell emissions.   Requests were also made for any
information specifically related to their use of fuel additives.   Replies
were received  in most cases.   The responses  are shown in Table 20.   In
general, additives  are  used by those  airlines with a significant inventory
of older dirtier, OT3,  JT4, and JT8 engines.   Several airlines stated that
the additives  would be  phased out in  the near future as  they switched over
completely to  smokless  combustor engines.
6.6    OTHER ABATEMENT  METHODS
       During  the approximately 8 years  that  test cell  emission abatement
has been seriously  pursued, numerous  techniques and devices  have been pro-
posed.  The five methods previously discussed -- nucleation  scrubbing,
fuel additives, thermal  converter, fuel  atomization, and electrostatic
precipitation  -- are the ones the Navy has considered promising enough to
warrant experimental investigation.  Many other methods  have been con-
sidered and ruled out because of some clearcut technical or  economic
problem.  The  characteristics,  costs, and application to test cells of
many of these  methods are summarized  in  References 10,  17,  19, and 20.
       Table 21 (Reference  10)  shows  the relative characteristics of a
number of abatement methods.   Table 22 (Reference 17) is more quantita-
tive.  It includes  estimates  of capital  and operating costs  as of mid-
1973.  The control  equipment  in Table 22 is  sized to baseline gas flow of
1,200,000 acfm.  This is a  much higher airflow than is  commonly encoun-
tered in cells.  For example, the workhorse  J79 engine  produces stack gas
flow of about  500,000 acfm  at military power  setting and 700,000 acfm in
                                      99

-------
                                   TABLE 20.  FUEL ADDITIVE USE BY COMMERCIAL AIRLINES
o
o
Airline
United
National
Bran iff
Western
American
TWA
North Central
Pan American
Allegheny
Delta
Continental
TWA does overhauls
The concentration
for a given engine
Additive used
Apollo DGT-2
See below3
CI-2
CI-2
NONE
CI-2
NONE
CI-2
NONE
NONE
CI-2
— data included in
of additive is that
type.
Engines with which Amount of Approximate % Engine
additives are used Additive used tests using additive
JT4, JT8, JT30, JT3, JT4 Technician's
discretion
All JT3D and some JT8D - -
JT3D 15 drops/ 50%
1000 gal fuel
- - -
JT4, JT3D 0.05-0.1% 15%
by volume
- - -
JT3D, JT4, JT8D Technician's 44;i
discretion^
- - -
- - -
JT3D, JT8D 1 gal/ -
1000 gal fuel
TWA figure.
which past experience has shown to give adequate plume opacity control,

-------
                             TABLE  21.    SUMMARY  OF  PARTICULATE  ABATEMENT  SYSTEMS  (Reference  10)
                          ^    E
                          85    '<
PARTICULATE
ABATEMENT METHOD
                                                                                                  H . High
                                                                                                  M • Modsrat*
                                                                                                  L " LOT
                                                                                                  •A- Bat Applicable
Dry Cyclone          L

Fabric Filter         H

Acoustical Filters    L

Magnetic Filters     H

Vet Scrubbers
    Spray            L
    Venturi          H
    Plate            H
    Baffle           L
    Impingement      H
    Moving Bed       H
    Filter Bed       N
    Packed Bed       H
    Centrifugal      H

 Electrostatic
    Dry              H
    Vet              H
L     H     L

H     L     H

ILL

MIL
 L
 H
 L
 L
 L
 M
 L
 H
 H
NA

L

H

H
H    NA

L    NA

H    NA

L    MA
                            HA
                            M
H     L

M     L

M     H

M     H
NODE        H    H   L

UHC( liquid)  H    M   H

NONE        H    H   H

NONE        H    H   H
                            CO, UIIC
                            CO, UHC
                            CO, UHC
                            co. imc
                            CO, UHC
                            CO, UUC
                            CO. UHC
                            CO, UHC
                            CO, UHC
                            HONE
                            CO, UHC
                                       H   H
                                       L   N
Would not neet visibility or case enisslon goals

Potential high AP if clogged,  Potential fire hazard with
UHC - only noderate visibility emission abatement
Vould not meet visibility or mass emission goals

Very early development  stage - requires massive auxiliaries
used only as secondary system
All comments predicated on exhanst preconditioning
Would not meet visibility goal
Will not handle  test cell flow variations - high AP could r»atilr» auxlllmrr fan
Vould not meet visibility goal
Vould not meet visibility goal
AP could require auxiliary fan
Does not easily  handle test cell  flov variations
Vould not meet visibility goal
Mow In use on  prototype Installation
Potential for  fcest cell use - minlnixed vatar losses
                                     Vould not meet visibility goals - high lost
                                     Preconditioning enhances performance - potential for test eell. nac
 Incineration
     Conventional
     Innovative
     Infrared
M      H    M     H     L    NA    M     L    CO. UHC     L    H
H      H    H     H     H    NA    M     M    CO, UHC     L    H
H      HMHKNAM     H    UHC (liquid)  M    H
                                                     Livlted particulate removal - high cost
                                                     Development required - moderate potential for test cell use - ctmld salt
                                                     Requires  large development effort - costs could be high - effectiveness
                                                     on visible emissions questionable

-------
                               TABLE  22.   COMPARATIVE DATA - ALTERNATE GAS  CLEANING SYSTEMS  (Reference 17)

DIRECT FLAME AFTERBURNER

FLOATING BED SCRUBBER

VARIABLE THROAT VENTURI

fMPINGEMENT PLATE SCRUBBER

REBOUND ENTRAPMENT SCRUBBER

JET SCRUBBER

NUCLEATION/PACKEO BED
SCRUBBER
IRRIGATED PRECIPITATOR

DRY SINGLE STAGE PRECIPITATOR

INTERMITTENTLY WASHED TWO -
STAGE PRECIPITATOB
BAGHOUSE

VISCOUS FILTER

POLLUTANTS
ABATED
SMOKE, SOL. & Lia PART.
CO, GASEOUS C,Hy
SMOKE. SOLID PART.
SOME NO2& LIO- PART.
SMOKE, SOLID PART.
SOME N02& LIQ. PART.
SMOKE. SOLID PART.
SOME NO2 & LIO. PART.
SMOKE. SOLID PART.
SOME N02 & LIO. PART.
SMOKE. SOLID PART.
SOME NOj & Lia PART.
SMOKE, SOL ID PART.
SOME N02& Lia PART.
SMOKE. SOL. & Lia PART.
SOMEN02
SMOKE. SOL. PART.
SOME Lia PART.
SMOKE, SOL. & Lia
PART.
SMOKE, SOL. PART
SOME Lia PART.
NONE

DESIGN PARAMETERS
MAX. GAS
FLOWHI
(ACFMI
3.320.000

1,200.000

1,200.000

1.200,000

1.200.000

1,200,000

1,200.000

1.200.000

1.460.000

1,200.000

1,460,000

1.200,000

MAX. GAS
TEMPERATURE
°F
1500

167

167

167

167

167

167

167

400

167

400

167

MAX. GAS
VELOCITY
IFPMI
1500

BOO/
800
400/ 121
500
400/
500
600/
650
400/ «l
500
600

1200/
1400
300/
350
300/
350
3-6

500

GAS FLOW
AREA
IFT2)
900/
1000
1500/
2400
2400/ 121
3000
2400/
3000
1B50/
2000
2400/ «)
3000
2000

B50/
1000
4200/
4850
3400/
4000
244.000/
485,000
2400

MAX. GAS
PRESS, DROP
(INHjOl
0

1012

4070

6-10

2-3

PROVIDES
HEAD
2-3

5-S

0.5-1

05-1

3-5

0.3-0.5

WATER
REQUIREMENTS
PRETREATMENT
SPRAY
(GPMI
2000131

200013)

2000(31

2000(3)

2000131

2000(31

2800(4)

2800(4)

2000(3)

2800«l

2000(3)

2800(41

GAS
CLEANING
IGPM)
_

60,000

8.400

4,800

11.000

7.200

14,000

4,200

—

60

_

_

POWER
REQUIREMENTS
WATER
PUMPING
IKWI
_

3450

700

510

1950

3710

1100

574

250

350

250

350

OTHER
(KW)


_

_

-

-

-

300

250

1900

120

100

100

TOTAL
(KW/1000CFM)


2.90

0585

0.425

1.63

3.10

1.16

0.685

1.79

0.35

a 29

a 375

COSTS
CAPITAL
COSTS
10* t
1.0 - 1.5

1.2- 1.5

1.2- 1.5

10- 1.2

1.2- 1.5

1.2- 1.5

0.9- 1.2

2.5 -3.0

1.5-20

0.9 • 1.2

2.5 - 3.0

0,2-0.3

ANNUAL
OPERATING
COSTS'*)
103 $
1300/
1500
80.'
90
im
25
IS.'
20
45/
55
90'
100
25'
35
20'
25
20'
30
to/
IS
91
13
IS/
20
o
no
                    11) LEAVING EQUIPMENT
                    (2) LIQUID/GAS SEPARATION COMPONENT
                    (31 BASED ON 400°F ENTERING EQUIPMENT
                    14) BASED ON 167°F (SAT.) ENTERING EQUIPMENT
                    (5) BASED ON 500 TESTS/YR OF 350 LB/SEC ENGINE:
                        EXCLUDES MAINTENANCE AND SOLID&LIOUID SEPARATION

-------
afterburner mode,  including augmentation airflow.   The J79 is one of  the
larger engines  tested  in  military test cells.   The cost figures  in Table
22 should be used  cautiously as  they are only  rough approximations based
on an "average"  test cell.
       Discussed below are  some  of the test cell cleanup methods
considered, but  not experimentally evaluated,  by the Navy.
6.6.1  Cyclone  Separators
       Cyclones  are  inertia! particle collection devices in  which carrier
gas containing  particles  is spun rapidly within a  closed cylindrical
shell.  Centrifugal force moves  the particles  to the wall  of the  cyclone.
From there they move to lower velocity regions of  the device and  are  col-
lected.  Cleaned gas exits  at the center of the cyclone cylinder.   Cyclone
dust collectors  are commonly used because of their low capital cost,  abil-
ity to run hot,  and good  collection efficiency for particles with dia-
meters greater  than about 5 microns.   It is easy to design cyclones that
can efficiently collect particles in the submicron range (such as those
found in test cells);  however, this cannot be  done without incurring  an
increased pressure drop.  For effective collection of 0.5  microns,  for
example, simple  calculations show that for reasonably sized  cyclones,
pressure drops  on  the  order of 100 to 300 inches of water  would be re-
quired.  This is far in excess of the 10 to 20 inches of water allowed for
a test cell cleanup device.
6.6.2  Fabric Filters
       Fabric filters  are commonly used to separate solid  particles from
liquid or gaseous  media.  These  filters are capable of collecting par-
ticles much smaller than  the space between the filter fibers.  This is due
to the fact that most  of  the filtration is accomplished by the coating of
                                     103

-------
 material  already collected,  rather  than  by the fibers themselves.  If
 glass  cloth  fabrics  are  used,  fabric  filters  can operate at temperatures
 approaching  600°F.   Fabric filters  are  simple, effective,  and well
 developed for  numerous applications.  However, there are several charac-
 teristics of test cells  that make these  filters of doubtful value as
 cleanup  devices.
       Fabric  filters  are  low  velocity  devices.  When fiberglass cloth is
 used,  superficial  velocities of  about 2.0  acfm/ft2 should  not be
 exceeded  (Reference  37).  This means  that  for  a J79 engine with sufficient
 augmentation air  to  drop the gas stream  temperature to 550°F, about
 400,000 feet2of  filter area  would be required.   With  water spray cool-
                                      j
 ing  (which introduces other  problems), a filter area  of  about 250,000
  9
 ft  would  still be required.  Considering  the  various  constraints  on  bag
 arrangement, bag cleaning devices, and dust hopper  arrangement,  a  baghouse
 to handle  this flow would have dimensions  of about  110 ft.  x  110 ft.  x
 40 ft. high  (Reference 10).  This is comparable  to  the size of  the  test
 cell.  Finding space for such an installation  would be impossible  in  most
 cases.  The  installed cost of such a baghouse  would be about  $2.50/acfm,
 or $1 to $1.5 million.   Finally, maintenance costs  due to  bag tearing  and
 blinding have been estimated at twice that of wet scrubbers or  ESP'.s.
 Fabric filters  pose other problems,  also.  They  create a fire hazard
because of the  large area coated with carbonaceous  particles.   Also,
filters do not  work well  in conjunction  with exhaust water  spray cooling
because the water can blind the fabric.   Considering all of these factors,
fabric filtration does  not  seem to be a  promising approach  to test  cell
cleanup.
                                    104

-------
6.6.3  Venturi  Scrubbers
       Venturi  scrubbers  are particulate collection devices in which the
gas to be cleaned  is  accelerated to a high velocity, then sprayed with
water droplets.  The  high relative velocity between water droplets and
particles in the gas  causes impaction between the particles and the water
drops.  The dirty  water drops are collected in a demisting section down-
stream from the venturi scrubber.  Collection of the particulate is en-
hanced if condensation  is induced by saturating the gas  in the reduced
pressure region at the  venturi  throat.   Water vapor condensation on the
particles causes particle growth and increased probability of  collection.
       Venturi  scrubbers  are very effective for collecting particles  with
diameters greater  than  about 1  to 2 microns.   But to collect smaller  parti-
cles, large gas pressure  drops  are required.   To effectively collect  par-
ticles of 0.5 micron  requires a water pressure drop from 40 to 80 inches.
These pressure  drops  increase operating  cost  and tend to make  venturi
scrubbers less  competitive with packed scrubbers.
6.6.4  Coanda Noise Suppressor
       The Coanda  system  (References 4 and 5) is designed to control
noise, but if affects pollution control  systems in several  ways.   The
Coanda can best be understood by studying  Figure 21.   Figure 21(a)  shows a
freestream jet  exhaust.   The hot gases are confined to a cone  with its
apex at the exhaust nozzle.   Because of  the high temperature in the gas
stream, the noise  is  refracted  away from the  exhaust.  This is the reason
a jet engine sounds much  louder 45° off-axis  than  directly behind the
engine.
       Figure 21(b) shows the principle  of the Coanda system.   When a
curved plate is placed directly above the  engine exhaust outlet,  the  hot
                                     105

-------
                                                 Noise
     a.   Unconfined jet exhaust
                              Hot gases
                                      Sound-
                                      absorptive
                                      material
       I/ ///////////// /Tl.A
    b.   Coanda gas-noise separator
Figure  21.  Principle of Coanda  device.
                 106

-------
gases of the exhaust  stream cling to the plate (the Coanda effect).   The
noise is refracted  away from the gas stream and is absorbed in sound-
absorptive material.   The  sound-absorptive material is not exposed to
thehot gas stream,  as is the case in conventional  test cells.   This  is  the
primary advantage of  the Coanda system.   The hot  gas stream and the  noise
are separated,  and  each can be handled separately and, hopefully,  more
effectively than when both must be controlled simultaneously.
       One obvious  advantage to separating the gas flow and noise  is  that
the test cell does  not require water spray cooling of the  exhaust.   In
conventional cells, this water cooling is needed  to protect the sound
suppression equipment.  Elimination of water cooling of the exhaust may
significantly reduce  the opacity of the  test cell  plume.   As a further
advantage, dry  operation allows a much wider choice of sound absorptive
material.
       A full-scale Coanda cell  is now being tested.   A final  report  on
these tests is  scheduled to be issued sometime in  March 1976.   In  addition,
there is a current  project at  the U.S.  Naval Academy to consider the  ef-
fects on test cell  operation of pollution control  devices  that could  be
installed on a  Coanda-equipped test cell  stack.
6.6.5  Hush House
       The hush house is simply a room large enough to contain an  entire
aircraft.  There is a long,  wide-mouthed  exhaust  tunnel with a deflector
baffle at the end which directs the exhaust vertically. Figure 22 shows
the hush house  at Miramar  Naval  Air Station.  The  hush house was designed
to suppress noise when testing engines in aircraft.  There is, however, an
unexpected additional  result.   Opacity readings from the hush  house  stack
are lower than  readings from the same engine tested in a conventional test
                                     107

-------
o
CO
                 //
                                      \
                                                  \
              r
                                                                                     -91
                                                                                 J

                                                                                 rfi
                                                                                  i
                                                                                 4.
                                                                                                Deflector baffle
                                                                                       19'—«-)
                  Mpch room

     Control rooni


Figure 22.   Hush house at  Miramar Naval  Air Station.
                                                       Cross-section

-------
cell.   In fact,  the  opacity readings from all  engines tested in the hush
house are less  than  20 percent,  except for the J79 engine in the F4,  which
has an  opacity  of  about 20 percent.
        The  lower plume opacity from  the hush house is apparently due  to
exhaust  stream  dilution brought  about by the large quantity of  air  en-
trained.  The wide-open exhaust  tunnel entrance (no augmenter is used),
combined with the  great length of the tunnel,  produces  a  diluted, well
mixed plume with a relatively low opacity.
6.7     SUMMARY  OF  CLEANUP  TECHNOLOGY
        Several  exploratory studies of possible test cell  cleanup methods
have been made.  Based on  these  studies,  the Navy,  which  has  been given
responsibility  for military test cell  cleanup  device development, has
sponsored experimental  work in five  areas.   These  areas are:
        •    Electrostatic precipitators
        •    Nucleation  scrubbers
        •    Thermal converters
        •    Fuel  atomization improvement
        •    Fuel  additives
A sixth method,  clean  combustion engine modifications, will  be  discussed
in Phase II of  this  study.
       Of the above  five methods,  only two  --  nucleation  scrubbers  and
fuel additives  —  are  reasonably well  developed.   Fuel  additives reduce
test cell plume  opacity, and may reduce particulate loadings  somewhat.
Use of fuel additives  is easily  and  inexpensively  implemented.   They  do
not seem to cause  serious  engine performance degradation  when employed
within recommended time and concentration limits.   The  extended use of
some fuel additives  (e.g.,  CI-2) is  known to be detrimental  to  engine
                                     109

-------
life.  The point at which the extended use of ferrocene will  cause  detri-
mental engine changes is not known.  Both the Navy and the Air  Force  are
now studying this question.  A nucleation scrubber is the only  test cell
cleanup device that has been experimentally evaluated on a full-scale test
cell.  Incomplete emissions sampling indicates that particulate and
condensible hydrocarbons are substantially removed.
       The electrostatic precipitator, fuel atomization improvement,  and
thermal converter concepts have been tested experimentally only in  small
scale models.  All seem to be capable of reducing test cell emissions.
Their drawbacks are high operating cost (thermal converter),  operating
problems (ESP), and limited pollutant reduction (fuel atomization).   Each
of these three methods would require further development before  it  could
be considered ready for full scale demonstration.
       Several concepts not experimentally evaluated by the Navy —
cyclones, fabric filters, venturi scrubbers — are briefly considered.   It
is concluded that none of these methods seems worthy of experimental
study.  Two recent noise control concepts are discussed because their
implementation would impact the other cleanup technologies.   These  are the
hush house and Coanda nose control systems.
       Cost data for each of the five experimentally-evaluated  test cell
cleanup methods is presented.   For the nucleation scrubber, the most
developed method,  three different cost estimates were obtained.  These
estimates differ by a factor of nearly 3 to 1.
       Foreign test cell cleanup technology was not investigated directly.
No such cleanup technology was discovered in the literature or  in the
discussions with the people listed in Appendix B.
                                    110

-------
                                  REFERENCES

  1.    Kikin,  I.  and R.  Bennett, "Chemical Reduction of S03, Particulates,
       and  NOx Emissions," presented at The International Energy
       Engineering Congress,  Chicago, November 4-5, 1975.

  2.    Klarman, A. F.  and J.  E.  Horling, "Participate Sampling from Gas
       Turbine Engines," Paper 27-5, presented at the Third International
       Conference on Environmental Sensing and Assessment, Las Vegas,
       Nevada, September 14-19,  1975.

  3.    Moses,  C.  A., "Reduction  of Exhaust Smoke from Gas Turbine Engines
       by Using Fuel Emulsions," Interim Report AFLRL No. 68,  July 1975.

  4.    Naval Air  Engineering  Center Report No. NAEC-GSED-80, Final
       Technical  Report, "Initial  Model Studies/Coanda Refraction Noise
       Suppression Concept — Exploratory Development," May 1974.

  5.    Naval Air  Engineering  Center Report No. NAEC-GSED-81, "Final
       Technical  Report, "Scale  Model Studies/Coanda Refration Noise
       Suppression Concept -- Exploratory Development," May 1974.

  6.    Demetri, E. P.,  "Model Testing of a Thermal  Converter for  the
       Control  of Exhaust Emissions from Navy  Jet Engine Test  Cells,"
       Northern Research and  Engineering Corporation Report  No. 1219-1  for
       Naval Air  Engineering  Center,  July 11,  1975.

  7-    Northern Research and  Engineering Corporation,  "Preliminary Design
       of a Thermal  Converter for  the Control  of Exhaust Emissions from
       Navy Jet Engine Test Cells," Report No.  1188-1,  1973.

 8.    White,  H.  J.,  "Industrial Electrostatic Precipitation,"
       Addison-Wesley, REading,  Pennsylvania,  1963,  p.  359.

 9.    United  Engineers  and Constructors,  Inc.,  "Test  and Evaluation of a
       Pilot Two-Stage Precipitator for Jet Engine  Test Cell  Exhaust Gas
       Cleaning,"  Final  Draft Report,  September  1975.

       Robson,  F.  L., et al.,  "Analysis of Jet Engine  Test Cell Pollution
       Abatement  Methods," AFWL-TR-73-18,  May  1973.   Available from NTIS
       as AD-763119.

11.    Teller,  A.,  "Turbine Emission  Control —  A Systems Approach,"
       Teller  Environmental Systems,  Inc.,  Worchester,  Massachusetts.

12.    Teller  Environmental Systems,  Inc.,  "Jet  Engine  Test  Cell  TESI
       Augmenter-Scrubber System,"  Contract No.  N62467-70-C-0240, December
       1971.

13.    Kemen,  R.  J., et  al.,  "Jet  Engine Test  Cell  Pollution Abatement
       Efficiency Tests," Naval  Air Rework Facility, Jacksonville,
       Florida, March 1973 -  May 1974
                                     111

-------
  14.     Krimmel,  J.  A.  and  B.  A.  Longley-Cook,  "The Effect of a Model
         Scrubber  on  Gaseous EMissions  from a Gas Turbine Engine Test Cell,
         Report  No. AESO 113-74-1,  August  1973.

  15.     Morhard,  W.D.,  "Joint  Navy-Air Force Jet Engine Test Cell  Study,
         Phase  II  —  Draft Report."   Final  report scheduled for issuance in
         July 1976.

  16.     Bailey, D. L.  and P. W. Tower,  "Production  Test Facilities for
         Turbojet  and Turbofan  Engines  —  1975 to 1995," Naval  Postgraduate
         School  Report  No. NPS-57Ba-FO-72061A.

  17.     Ferner, J. A.,  et al.,  "A  Study of Means for  Abatement of  Air
         Pollution Caused by Operation  of  Jet Engine Test Facilities," Naval
         Facilities Engineering  Command No.  5685-000,  August 1973.

r^
(  18J    Morhard,  W.  C.,  Naval  Air  Systems  Command.  Private Communication.

  19.     Kelly,  C. M.,  "Air  Pollution Abatement  for  Jet  Engine  Test
         Systems," Naval  Air Engineering Center  Report No.  NAEC-GSED-64.

  20.     C.  F. Braun  and  Co., "Turbojet Aircraft Engine  Test Cell Pollution
         Abatement Study," Naval Facilities Engineering  Command Report No.
         CR74.001.

  21.     Lindenhofen,  H.  E.,  "State-of-the-Art Review  of Gas Turbine Engine
         Exhaust Smoke  Monitors," Naval  Air Propulsion Test Center  Report
         No.  NAPTC-AED-1956,  August  1971.
 '\22.  j   Lindenhofen,  H.  E.,  "A  Survey of the Air  Pollution  Potential  of Jet
  ^_^'   Engine  Test Facilities,"  Naval  Air  Propulsion  Test  Center  Report
         No.  NAPTC-PE-3,  October 1972.

  23.     Champagne, D.  L.,  "Standard Measurement of  Aircraft Gas  Turbine
         Engine  Exhaust Smoke,"  ASME Paper No.  71-GT-88.

  24.     SAE  Aerospace  Recommended  Practice  1179,  "Aircraft  Gas Turbine
         Engine  Exhaust Smoke Measurement,"  Society  of  Automotive Engineers,
         New  York,  N.Y.,  May 4,  1970.

  25.     Klarman, A.,  "Gas  Turbine  Engine Particulate Measurement Technique
         —Summary  of  Coordinating  Research  Council  Program," Interim
         Report, Naval  Air  Propulsion Test Center, November  6,  1974.

  26.     Bogdan, L., et al.,  "Analysis of Aircraft Exhaust  Emission
         Measurements," Cornell  Aeronautical Laboratory,  October  15,
         1971.   Available from NTIS as PB204.879.

  27.     "Aircraft  Engine Emissions Catalog," Report No.  AESO 101 --
         Revision 3, June 1974.
                                      112

-------
28.    Environmental Protection  Agency,  "Compilation  of Air  Pollutant
       Emission Factors," 2nd  Edition,  USEPA,  Research  Triangle  Park,
       N.C., April 1973.

29.    Poth, E. W. and E. R. Lozano,  "Air  Pollution Emissions  from  Jet
       Engines," USAF Regional Environmental Health Laboratory,  Kelly AFB,
       Texas, February 1967.

       U.S. Bureau of Mines, "A  Field Survey of  Emissions  from Aircraft
       Turbine Engines," USBM  Report  No. R17634, May  1972.

       Slusher, G. R., "Sulfur Oxide  Measurement in Aircraft Turbine
       Engine Exhaust," National  Aviation  Facilities  Experimental Center
       Report No. FAA-NA-75-10,  September  1975.

32.    Faitani, J. J., "Smoke  Reduction  in Jet Engines  through Burner
       Design," SAE Paper 680348,  1968.

33.    DeCroso, S. M., et al., "Smokeless  Combustion  in  Oil Burning Gas
       Turbines," ASME Paper 67-PWR-5,  1967.

34.    Lieberman, A., "Composition  of the  Exhaust from  a Regenerative
       Turbine System," JAPCA  18,  pp.  149-153, 1968.

35.    Stockman and Betz, "Study of Visible Exhaust Smoke  from Aircraft
       Engines," SAE Paper 710429,  May 1971.

36.    Grems, B. C., "Plume Opacity and  Particulate Emissions  from  a Jet
       Engine Test Cell," M.S. Thesis,  University of  California, Davis,
       1975.

37.    Spaite, P. W., et al.,  "High Temperature  Fabric  Filtration of
       Industrial Gases," JAPCA,  May  1961.

38.    Klarman, A., Naval Air  Propulsion Test  Center.   Private
       communication, January  17,  1976.
                                     113

-------
                                 APPENDIX A
                      SCRUBBER RETROFIT COST ESTIMATE
       The following detailed cost estimate is reproduced as received from
Teller Environmental Systems, Inc.
                                     115

-------
ESTIMATED COST FOR A TELLER ENVIRONMENTAL SYSTEMS, INC.



                 EMISSION CONTROL SYSTEM



                           FOR



               A STANDARD JET ENGINE TEST CELL



                      (J 79 ENGINE)
                                             14 January 1976



                                      by  Charles B.  Wyman



                                          Denis R.  J.  Roy
                               116

-------
I.   INTRODUCTION
    The size of the Nucleation scrubber system is predicated on
    testing the J-79 jet engine in its maximum afterburner mode
    (17,500 Ibs. thrust).  Also, the use of a TESI designed aug-
    menter is assumed in order to reduce the amount of dilution
    air required.

    Basis for the estimate is the prototype installation now
    operational at the Black Point test cell number one (1) NARF-
    JACKSONVILLE, FLORIDA.  The estimate reflects refinements in
    material selection as well as improvements in design developed
    from operation of the prototype unit.

    The cost per cell was established using the following
    assumptions:
    1.   A common cooling tower is to be used for two test cells.
    2.   The Cooling Tower is located between the two test cells
        (Est. 200 ft. between cells).
    3.   One solids treatment system would be required for two
        test cells.
    4.   Structural design - no snow loads incorporated.

    The flow sheet attached is representative of typical flow
    conditions for gas and liquid streams.
                              117

-------
     fTI
0
     0
     o
     I
 o
 0
2
o
M
(/>   fff   o
           r

           I
2
E
n
z
<

§

Z
PI
z

>
r
(A

(fl

fl

01

Z
0

-------
I.I.   SYSTEM DESCRIPTION




      The Nucleation system to be employed for emission control




      would consist of  five major components:




           1.  TESI Augmenter




           2.  Quench System (Existing)




           3.  Nucleation Scrubber System




           4.  Recycle Cooling System




           5.  Solids Treatment System






      The TESI Augmenter (proprietary) has approximately the




      same physical size as the conventional augmenters (6 feet




      diameter x 12 feet overall lenght).  It has a similarity




      in the converging inlet section and quench spray ring.




      The main differences are quench design, the diverging




      section, and the recycle gas system.  Material of construc-




      tion would be carbon steel with the quench ring of stainless




      steel.






      The quench systems employed with the present augmenters




      could be adapted to the new TESI augmenter at minimal cost.






      The Nucleation Scrubber (proprietary)  design is essentially




      the same as the prototype design, that is the scrubber




      consists of two identical sections mounted with independent




      supports parallel to sides of the exhaust stack.  Typical




      overall scrubber dimensions can be approximately 22 feet



                             119

-------
II.   SYSTEM DESCRIPTION (CONT'D)
                               ,>

      long x 27 ft high x 28'-6" wide including discharge

      louvers.  Although dimensions can be modified to adopt

      to orientations.  Typical packed sections have a face

      dimension of 20 ft wide x 24 ft high and each would con-

      tain a 4 ft depth of 2 inch nominal polypropylene

      Tellerettes followed by a divider support plate and a

      one ft depth of 1 inch nominal polypropylene Tellerettes,

      Each side would be made of three 6'-8" wide modules

      constructed of FRP material.  The 2 inch packing is to

      be irrigated from above by two spray headers with poly-

      propylene nozzles and from the inlet side with face sprays.


      Above the center of the exhaust stack outlet would be lo-

      cated turning vanes to distribute the gas flow to the

      packed beds.  These vanes would be carbon steel with a

      high grade epoxy coating.


      The gas are discharged to the atmosphere at an angle of

      45° upward via gull wings (stationary louvers)  constructed

      of FRP and mounted on the scrubber section.  The sides

      and top of the unit would be enclosed with a heavy duty

      corrugated FRP sheathing.


      The liquid effluent is collected in clopped bottom troughs

      containing baffles to prevent gas bypass.  The collected

                                  120

-------
II.   SYSTEM DESCRIPTION  (CONT'D)

      liquid is discharged by gravity to the cooling tower
      sump.  These troughs would be constructed of carbon
      steel with epoxy coating inside and outside.

      All structural steel would be of ASTM A-36 specifica-
      tions and would be  prepared for and coated with an
      epoxy paint system.

      All piping is carbon steel schedule 40 and includes
      recycle water lines between the scrubber system and
      the cooling tower.  Two pumps with drives capable of
      delivering each 5500 GPM at 110' TDK would be installed.
      Carbon steel wetted parts would be employed.

      Instrumentation would provide for local indication of
      liquid flow, pressures, temperature and level as well
      as alarm circuits.

      The recycle cooling system is comprised of a tower
      capable of handling the peak thermal load of two test
      cells, with the J-79 engines in afterburner mode.

      A standard cooling tower approximately 90 ft x 60 ft
      constructed of wood or plastic with a concrete sump
      would be required.  Thermal load on the unit would be
      of the order of 840 M BTU/hr for both cells.  The re-
     mainder of  the cooling  system  would be comprised of
                             121

-------
II.   SYSTEM DESCRIPTION (CONT'D)






      level, temperature and pressure indicators, miscellaneous




      piping and structural steel for access ladders and plat-




      forms.






      The solids treatment system would consist of a 500 gallon




      mix tank (FRP) with agitator, a lime feeder (40 #/hr




      normal), a 5000 gallon (FRP) settling tank and a rotary




      vacuum filter (400 ft filter surface, wetted parts 304 s.s)




      Connecting piping and pumps would be mild steel construc-




      tion.  Instruments would include a flow control loop and




      a level control loop.  Structural supports for feeder




      would be to ASTM A-36 specifications and would be prepared




      for and coated with an epoxy paint system.
                                 122

-------
         ENGINEERING SERVICES PROVIDED BY TEST
Process Flow Diagrams
Piping and Instrument Diagram
General Arrangement Drawings
Equipment List and Specifications for Purchased Equipment
Fabrication Drawings for Manufactured Process Equipment
Isometric Piping Drawing in
Plan and Elevation Piping Drawings Including Supports § Hangers
All Foundation and Concrete Design
All Structural Steel Design Including Pipe Supports and Hangers,
    Platforms etc.
All Painting
All Electrical Work  (including one line and elementary diagrams,
    specifications, etc.) Required For Installation, Including
    Local and National Codes etc.
All Instrumentation and Electrical Location Plans.
Lighting
Noise Silencing Provisions that may be Required in the system with
      such equipment as a vacuum pump
All Erection Specifications
All Erection Supervision and Inspection
Purchasing of All Equipment and Materials
Operating and Maintenance Manuals
Recommended Spare Parts Lists
Preparation and up-dating monthly of  a schedule for  the  job
     showing key milestones, activities  and  action dates.
Total Responsibility  For All Erection and Construction
                              123

-------
   ENGINEERING SERVICES PROVIDED BY TESI (CONT'D)






Total Responsibility For Receiving, Unloading, Inspecting, and



Storing All Equipment and Materials Delivered to Job Site.








Turnkey - Engineering



     Cost For Two Test Cells, One Cooling Tower, One Solids



     Treatment System, and TESI Augmenter       $225,000








     All Engineering Listed Above Including (but does not



     include direct construction labor cost or equipment cost)



     Supervision of Erection $ Purchasing
Performance Guarantee





     <10I Opacity



     <0.004  gr/scf



      All services and equipment to bring system into per-



      formance guarantee.





Equipment Guarantee



Equipment guaranteed for one (1) year from date of start-up



Life expectancy of the unit 15-25 years.
                                124

-------
                 SUMMARY OF COSTS



     SECTION                         TOTAL COST



TESI Augmenters                      $  36,600

Nucleation Systems                     526 200

Cooling System                         336,000

Solids Treatment System                179,500
                          Sub-Total $1,078,300


Contingency 10%                        108,000

Engineering & Royalty                  225,000
TOTAL COST FOR INSTALLED SYSTEMS ON
TWO JET ENGINE TEST CELLS           $1,411,300
EQUIVALENT COST PER TEST CELL       $  705,650
                          125

-------
     BREAKDOWN OF COSTS FOR ONE TEST CELL
Nucleation Scrubber

TESI Augmenter

Cooling Tower

Solids Treatment System

Piping

Pumps

Instruments

Electrical

Miscellaneous



101 Contingency




Engineer ing
                                    (Prop-Rated from Cost
                                      of two cells)
$ 167,600

   17,800

  125,000

   50,600

   70,500

   10,000

   19,900

   52,500

   24,750

  538,650

   54,500
  593 ,150

  112,500

$ 705,650
                            126

-------
         ESTIMATED  COST  OF  TEST  EMISSION CONTROL SYSTEM FOR A STANDARD JET ENGINE TEST CELL
  ITEM
MATERIAL
  COST
LABOR
 COST
FREIGHT
  COST
  (J-79 Engine)

TOTAL
  Nucleator  System For  One  Test  Cell

  Nucleation Scrubber - Modular  construction
    6 -  6'-8" modules x 27'  high

  Packing -  2" Tellerettes-Polypropylene
                                   4032 ft
            1" Tellerettes-Polypropylene
                                   1056 ft

  Spray Nozzles - Polypropylene
                         Est.  100 nozzles

_, Scrubber Side Walls $ Top Covers
™           2000 ft2  FRP

  Stationary Louvers - FRP
    21'  long x 6' wide  Approx.  700 ft

  Structural Supports - incl.
    Framework for module
    Turning Vanes
    Sump Steel
    Pipe Supports
    Support Legs
    Side Plate Reinforcing  M0,000#

       All  steel primed § epoxy coated

  Piping §  Valves  -  incl.
    Manifolds for  nozzles
    Piping  to and  from cooling tower
    Strainers.       Carbon Steel Sch. 40
  25,000
  56,000
  27,000
 2,500
20,000
16,000
    800
  2,000
  2,000
  28,300
22,000
10,000
2,000
17,000
2,200
1,000
400
50
3,500
600
1,200
400
50
500
400
24,200
10,800
2,100
21,000
3,200
  78,000
  45,000

-------
        ESTIMATED COST OF TESI EMISSION CONTROL SYSTEM FOR A STANDARD JET ENGINE TEST CELL
ITEM
MATERIAL
  COST
 LABOR
  COST
FREIGHT
  COST
  (J-79 Engine)

TOTAL
Instrumentation -  incl.
  Pressure Gauges
  Orifice Meter
  Temperature Indicators
  Level Indicator

Pumps - includes
     5500 GPM at 110'  TDH pumps w/drive
  Carbon Steel Wetted Parts
  200 HP

Electrical - incl.
  Starter for one  - 200 H.P.  motor
  Wiring $ conduit

Painting - incl.
Instruments -  incl.
  Pressure Gauges
  Temperature  Indicators
  Level Indicators
  Alarms
  3,600
  6,000
 13,500
 4,000
   500
10,000
  5,000
 5,500
   400
   500
    500
    500
   8,000
  7,000
  24,000
Touch-up and finish
TOTAL NUCLEATOR SYSTEM
Cooling System For Two Test Cells
Cooling Tower to handle 840M BTU/hr
approx. 90' x 60'
Misc. Structural Steel for ladder,
platforms and supports
1,400
185,700
190,000
17,000
10,000
68,550
20,000
12,000
100
8,850
10,000
1,000
11,500
263,100
220,000
30,000
  11,000

-------
           ESTIMATED COST OF TESI EMISSION CONTROL SYSTEM FOR A STANDARD JET ENGINE TEST CELL
   ITEM
MATERIAL
  COST
LABOR
 COST
FREIGHT
  COST
                                                                                             "(J-79 Engine
                                                                                        TOTAL
ro
Piping and Valves - incl.
  Strainers

Electrical - incl.
  Starters for tower fans
  Wiring and conduit

Painting - incl.
  Touch-up and finish

Site-Prep for Tower

          TOTAL COOLING SYSTEM

Solids Treatment System For Two Test Cells

Lime Feeder - 304 SS
     40#/hr normal capacity

Mix Tank - 500 gallon  (FRP)

Settling Tank - 5000 gallon  (FRP)

Agitator for mix tank  -  304  SS
  impeller and shaft

Rotary Vacuum Filter -  3,04 SS
  Wetted Parts    400 ft   surface  incl.
  vacuum receiver and vacuum pump

Pumps  -
  1  -  160 GPM @  50' TDH
  2  -  80 GPM @  50' TDH
       Mild Steel  wetted  parts
                                                   11,500
                                                   24,000
                                                      900
                                                   75,000
                                                    1,000
                                                    2,500
                 8,000
                14,000
                   100
                 3,000
                   500
                 1,000
              500
            2,000
              100
            2,000
              500
              500
              20,000
              40,000
450
1,500
249,450
2,800
500
4,800
5,000
8,000
72,500
500
50
500
50
-
14,050
200
50
200
5,500
9,500
336,000
3,500
600
5,500
               1,100
              80,000
               2,000
               4,000

-------
           ESTIMATED COST OF TEST EMISSION CONTROL SYSTEM FOR A STANDARD JET ENGINE TEST CELL
   ITEM
MATERIAL
  COST
LABOR
 COST
FREIGHT
  COST
                                                                                             (J-79 Engine)

                                                                                          TOTAL
   Piping  §  Valves  -  incl.
    All inter  connecting piping  between
    filter  tower,  and  tanks.
    Carbon  Steel

   Instruments  -  incl.
    Pressure gauges
    2  - Flow control loop
    Level control  loop
    Alarms
   Structural Steel - incl.
    Access  platform  for tanks
    Support for  feeder

   Electrical -  incl.
    Starters for  agitator  -  pumps
o   Wiring  § conduit

   Painting  - incl.  touch-up  and  finish

   Site Prep -  incl.  foundation  for  tanks
    and filter

            TOTAL  SOLIDS TREATMENT  SYSTEM
00
                                                   15,000
                 15,000
             1,000
               31,000
6,000
6,000
9,000
450
1,000
124,950
6,000
4,000
7,000
5,000
5,000
47,650
800
500
1,000
50
-
6,900
12,800
10,500
17,000
5,500
6,000
179,500
  TESI Augmenter  For  One  Test  Cell

  Augmenter tube  consisting  of a  converging
  and diverging section with integral  Quench
  Ring.
       6' diameter x  12'  long
       Shell Carbon Steel
       Quench Ring Stainless Steel
 16,000
 1,000
    800
                                                                                            17,800
  Note:  Electrical  costs  do  not  include  sub-station costs if required.

-------
                                  APPENDIX B
                             INFORMATION SOURCES

The following  is  a list  of persons contacted during the course of this
study who provided significant useful information.   Their assistance is
gratefully  acknowledged.
Mr. Dominic  Croce
Naval Air  Engineering  Center
Lakehurst, New  Jersey   08733
(Coanda device,  hush-house)

Major Peter  S.  Daly
Air Force  Civil  Engineering  Center
Tyndall Air  Force  Base
Florida  32401
(Emissions)

Mr. Richard  J.  Keman
Code 610
Naval Air  Rework Facility
Jacksonville Naval  Air Station
Jacksonville, Florida   32212
(Nucleation  scrubber,  electrostatic
precipitator)

Mr. Anthony  Klarman
Code PE-71:   AFK
Naval Air  Propulsion Test Center
Trenton, New Jersey 08628
(Fuel additives)

Dr. Barbara  Longley-Cook
Naval Air  Rework Facility
North Island Naval  Air Station
San Diego, California   92135
(Emissions)

Mr. Larry  Mihalic
Naval Air  Rework Facility
North Island Naval  Air Station
San Diego, California   92135
(Emissions)

Mr. Don Munson
United Engineers and Constructors
100 Summer Street
Boston, Massachusetts   02110
(Electrostatic precipitator)
Mrs. Sarah H. O'Brien
Naval Air Rework Facility
Jacksonville Naval Air Station
Jacksonville, Florida  32212
(Nucleation scrubber, electrostatic
precipitator)

Lt. Alfred Roth
Wright-Patterson Air Force Base
Dayton, Ohio  45433
(Fuel additives)

Mr. R. J. Salazar
Naval Air Rework Facility
Alameda Naval Air Station
Alameda, California  94501
(Emissions, test cell operation)

Dr. Wayne Sule
Naval Air Engineering Center
Lakehurst, New Jersey  08733
(Thermal converter, fuel
atomizat ion)

Dr. Aaron J. Teller
Teller Environmental Systems, Inc.
10 Faraday Street
Worchester, Massachusetts  01605
(Nucleation scrubber)

Mr. Ed Thomas
Air Transport Association
1709 New York Avenue, N.W.
Washington, D.C.  20006
(Test cell operations)

Mr. James A. Tomich
Bay Area Air Pollution Control
District
939 Ellis Street
San Francisco, California  94109
(Emissions)
                                    131

-------
Capt. Dennis F. Naugle                 Mr. W. W. Wilcox
Air Force Civil Engineering Center     United Airlines Maintenance Base
Tyndall Air Force Base                 South San Francisco
Florida  32401                         California  94104
(Emissions)                            (Fuel additives)
                                 132

-------
                                   TECHNICAL REPORT DATA
                            (Please read Instructions on the reverse before completing)
1. REPORT NO.
  EPA  340/1-78-001 a
                              2.
                                                           3. RECIPIENT'S ACCESSIOWNO.
4. TITLE AND SUBTITLE

   JET ENGINE TEST CELLS
   MEASURES:  PHASE  1
                                  5. REPORT DATE
- EMISSIONS  AND CONTROL
                                               January 1978
                                  6. PERFORMING ORGANIZATION CODE
                                   EPAOE;  Project 7203
 7. AUTHOR(S)

   D.  E.  Blake
                                  8. PERFORMING ORGANIZATION REPORT NO.
                                   ACUREX  FINAL REPORT 76-218
                                   1978  Revision
 9. PERFORMING ORGANIZATION NAME AND ADDRESS

   Acurex Corporation/Aerotherm Division
   485 Clyde Avenue
   Mountain View, California  94042
                                  10. PROGRAM ELEMENT NO.
                                  11. CONTRACT/GRANT NO.

                                    68-01-3158,  Task 4
 12. SPONSORING AGENCY NAME AND ADDRESS
   U.S.  Environmental  Protection Agency
   Division of Stationary Source Enforcement
   Washington, D.C.   20460
                                                           13. TYPE OF REPORT AND PERIOD COVERED
                                                             Final - 8-13-76 - 9-30-76
                                  14. SPONSORING AGENCY CODE
 15. SUPPLEMENTARY NOTES
   Stationary Source  Enforcement Series
 16. ABSTRACT
      This report discusses  the current state of  the  art of pollutant emission measure-
   ment and cleanup technology related to military  jet  engine test cells.  Considerable
   emissions data from  jet engines is available,  but  data from test cell stacks is
   sparse.  Gaseous emission data for the major pollutant species (CO, NO, N02, S02*
   SOa) are reliable; particulate, opacity, and unburned hydrocarbon data are less so.
      The five types of test cell cleanup methods that  have been experimentally evalu-
   ated -electrostatic precipitator, nucleation  scrubber, fuel  additives, thermal con-
   verter, and fuel atomization improvement -are described in detail.  Other, less
   promising methods are briefly discussed.  Several  methods are quite effective in re-
   ducing test cell emissions.   Fuel additives are  effective in  reducing test cell plum
   opacity.  Capital and operating cost data on these methods are presented.  For the
   nucleation scrubber, the  best-developed cleanup  technology, three cost estimates
   from different sources are given.
                                                              ume
17.
                                KEY WORDS AND DOCUMENT ANALYSIS
                  DESCRIPTORS
                                              b.lDENTIFIERS/OPEN ENDED TERMS
                                                                           COSATl Field/Group
   t  Jet engines
   •  Emission exhaust gases
   •  Jet engine test cells
                        Enforcement
                        Jet engine test cells
                        Air facilities
13B
14D
01E
 3. DISTRIBUTION STATEMENT


   Release unlimited
                     19. SECURITY CLASS (ThisReport)
                        Unclassified
                     20. SECURITY CLASS (Thispage)
                        Unclassified
                                               22. PRICE
EPA Form 2220-1 (9-73)
                                             133

-------