NOTICE
The research described in this article has been funded wholly or in part by the United States,Environmental Protec-
tion Agency through Contract 68-03-3166 to Metcalf and Eddy, Inc.  It has been subjected to Agency review and is
approved for publication.                                                                     ;
                            THE EFFECT OF OIL SHALE RECOVERY PROCESSES ON AIR EMISSIONS
                                                         By
           Hal Taback, P.E.
          Consulting Engineer
             KVB, Inc.
      Irvine, California  92714
                                Robert Goldstick        '
                                    President
                              Energy Design Service
                          Santa Cruz, California  95062
                                                    Edward Bates
                                                  Oil Shale Manager
                                        U.S. Environmental Protection Agency
                                   Air and Energy Engineering Research Laboratory
                                    Research Triangle Park, North Carolina  27711
     Development of  oil shale production processes has
 led to a variety of retort designs.  Figure 1 shows the
 total facility emissions as reported in the PSD permit
 applications for seven potential shale oil recovery
 plants.  The actual facility emissions based on the
 reported processes, in some cases, would be signifi-
 cantly higher if assumptions made by the developers
 regarding low levels of organic sulfur and nitrogen
 species in retort gases prove incorrect.  The degree of
 variation in total facility emissions is considerable
 with complicating tradeoffs.  For example, the Clear
 Creek facility with a Chevron solids recycle retort has
 very high NOX and CO emissions.  Conversely, the Union
 facility with a gas recycle retort has much lower CO
 and NO  emissions but higher SO... emissions.
       X                        A
     The purpose of this analysis was to evaluate these
 various processing schemes and determine the effect of
 improved air pollution  controls.   The information pre-
 sented below will show  that, with the proper selection
 of air pollution control  techniques, the air emissions
 for each of these processes  can be held to essentially
 equivalent  values.
      In developing  the analysis  for this process  compar-
 ison,  the EPA's Pollution Control Technical Manuals
 (PCTMs)  for various  shale oil processes (References  1,
 2, 3)  and various  PSD Permit Applications  (References
 4, 5,  6,  7, 8, 9,  10) were used.   The PCTMs present  a
comprehensive analysis of the heat and material flows
in a complete oil shale recovery process.  The PSD
applications provide controlled air emissions for the
specific process considered by the developer.  Thus,
the comparison presented in this paper is based on the
actual design conditions expected in a full-scale
operation.                                              .
SHALE OIL RECOVERY PLANT
    A shale oil recovery plant is quite  complex involv-
ing many varied operations.  The unit operations
required to recover the oil from' the shale include:
     — mining
          - below-ground
          — above—ground
     — retorting
     - product  recovery
     — removal  of nitrogen (ammonia and  organic
         nitrogen gases) from  the  retort gas
     - removal  of sulfur  (hydrogen sulfide  and
         organic sulfur gases) from the  retort gas
     - gas  utilization (retort gas combustion)
     — end  of pipe  controls
     - upgrading                 :
     - spent  shale  disposal                            ,
     Within each of  these  unit  operations there can be a
 number of  process alternatives.  ; The  shale  can be mined
 in an open pit  mine or a room arid pillar mine,, or the

-------
oil can be recovered without mining with an in-situ
process*  The retort heat can be provided by combustion
of the spent char within the retort or with a recycled
stream which can be either gas or solid.  Each of these
variations can affect the emission rates.  Conse-
quently, consideration of all the potential processing
schemes can be quite complicated.
    Figure 2 presents some possible processing combina-
tions considered as viable alternatives for full scale
processing.  The specific process combinations used for
this analysis are presented in Table 1.
         Table 1.  FIVE PROCESSES FOR ANALYSIS
Case No.
1
2
3
4
Mining
Open pit
Room & pillar
Room & pillar
Room & pillar
Retort
direct combustion heat^8'
(e.g., Paraho)
direct combustion heat -
circular grate
(e.g., Superior, Dravo,
Allis Chalmers)
indirect combustion' ' gas
recycle (e.g. , Union)
indirect combustion solids
                             recycle
                             (e.g., Lurgi, Chevron)
 5         Modified in-situ  in-situ and indirect heat
                             gas recycle above ground

'a' Direct combustion heat - the heat for retorting is
    provided by combustion of the spent char within the
(b)
retort.
Indirect combustion heat - the heat for retorting
is provided by combustion of retort gas or spent
shale outside of the retort.
    In this paper the methodology used to evaluate the
various processes will be presented first.  Then an
evaluation of the sulfur and nitrogen gases produced by
the retort for the various processes will be presented
along with a discussion of the effect on the acid gas
removal processes and net facility emissions.  Emis-
sions from other facets of the shale oil recovery oper-
ation (e.g., mining, solids handling, spent shale dis-
posal) will then be presented.  This is followed by an
analysis of the emissions for five typical processes
being considered for full scale development.
    For each analysis, a base case scenario representa-
tive of the process configuration proposed by the
developers is presented, and the criteria pollutant
emissions are determined.  Then two alternative pro-
cessing schemes to reduce these emissions to their
lowest levels are considered.
    METHODOLOGY
    For this analysis, the oil recovery facility was
divided into three basic categories: mining, retort,
and upgrade.  The emissions associated with mining and
solids handling (primarily participates, carbon monox-
ide, and nitrogen oxides) are similar to other mining
operations.  The data provided in the PSD permit appli-
cations were used to develop emission rates for each
type of unit operations (e.g., blasting, drilling,    ;
vehicles, conveying, crushing).:  The Bwissioiss aBsoci —
ated with the upgrading process (hydrocarbons from
storage and fugitive sources), other than those from
the combustion of the retort gas, are similar to other
oil refining operations and, again, the PSD permit
applications were used to develop estimated emission  '•
levels.
    The combustion of the retort gas can be the princi-
pal source of emissions from the facility and the
source most affected by the particular retort process
and gas cleanup scheme used.
    The primary concern of this analysis is emissions
of nitrogen and sulfur oxides and particulates.  Emis-
sions of carbon monoxide and hydrocarbons are generally
consistent for all processes, with a few exceptions.
The combustion of the spent shale can produce very high
carbon monoxide emissions, and this will be discussed ,
below for that particular process.  High hydrocarbon  ,
emission rates can result from certain types of retort
processes (e.g., Tosco II) that involve direct contact,
heating of raw shale or a heat carrier with flue gas.
However, this process was not included in this
analysis.
                               i
    To evaluate the many process variations and develop
the data necessary to estimate emission levels it was
first necessary to determine what process combinations
are feasible.  The alternatives considered are shown in
Figure 1.
    The basic design parameters that affect the pollu-
tant emissions for each process were defined.  These
design parameters are shown in Table 2.
      Table 2.   DESIGN PARAMETERS  USED IN ANALYSIS
                                                        Unit Operation
                                                        Mining
                                                                                 Design Parameter
                         Type of Mining               ;
                             open pit
                          room ;and pillar             ;
                             irt-situ
Retort           . Retort Gas Produced, m /m  of oil
                 . Heating Value of Retort Gas, kj/mj
                 . Partitioning of sulfur and nitrogen
Product recv'ry    None- The product recovery process
                   has no significant effect on emiss.
NHo removal      . NHj exit concentration, ppm
                 . Organic nitrogen content of retort
                   gas, %
R^S removal      . HoS exit concentration, ppm
                 . Organic sulfur content of retort
                   gas, %
                 . Organic sulfur gas removal effic., %
Gas utilization  . Boiler-dilution ratio (dry gas/fuel)
                 . Spent shale combustion exit concen-
                   trations for NOX, SOX & CO, ppm
End—of—pipe controls           '
           Particulate - Baghouse - exit loading, g/m
                 . Sulfur - FGD - exit SOX, ppm
                 . Nitrogen
                   Ammonia inje'ction - exit NOX, ppm
                   Staged combus. - exit NOX & CO, ppm
                                                            For each of the unit operations, the design parame-
                                                        ters were applied as indicated by either the  retort

-------
process conditions, performance of the pollution con-
trol equipment, or the reported emissions from the PSD
applications.  The following discussion presents the
rationale for choosing the specific design parameters
used in the analysis.

MINING, SOLIDS HANDLING. AND UPGRADING EMISSIONS
    The data provided in the PSD permit applications
were evaluated to determine typical emission rates for
the various unit operations.  Figures 3 through 7 show
the emission rates from individual sources for the cri-
teria pollutants.  For this analysis, all emissions
associated with combustion of the retort gas (i.e.,
upgrade heater, retort heater) are considered as part
of the emissions from the retort operation.
    The carbon monoxide emission sources (Figure 3) are
blasting, below ground vehicles, above ground vehicles,
and the combustion of the retort gas in the retort and
upgrading process.  The hydrocarbon emission sources
(Figure 4) are primarily mining vehicles, storage, and
fugitive emissions in the upgrading and retort gas com-
bustion.  Nitrogen oxide emission sources (Figure 5)
are primarily from retort gas combustion and mining
vehicles.  The only significant sulfur oxide emissions
source (Figure 6) is the combustion of the retort gas,
with mining and upgrading adding a relatively small
amount.
    The piarticulate emission sources  (Figure 7) are
those  associated with below ground mining (drilling,
blasting, conveying, crushing, engines), above ground
mining (surface soils removal, second and third degree
crushing, conveying, storage, and spent shale dispos-
al), retort gas combustion  (steam generator, retort
heater, and upgrade heaters), above ground vehicles,
and fugitive emissions from truck traffic.
    The values from the PSD permit applications pre-
sented for mining and upgrading emissions were used in
the oversill facility emission estimates presented
below.  The emissions from  retort gas combustion were
calculated as  described below.
    The choice of mining technique determines  the  emis-
sion  rates.  The values used were developed  from the
PSD analysis for room and pillar mining and  from  the
literature for open pit mining  (Reference  1).   The
emission  rates used are shown in Table  3.
           Table  3.  EMISSION  RATES  FOR MINING
                (Above-  and  Below—Ground)
Type of Mining
Open pit
Room & pillar
Emissions , kg/1000 m3 of Oil
CO HC NOX SOX PM
370 50 470 40 410
150 20 350 20 180
                               RETORT GAS                       i
                               ~"""~™"~~~""~~~"'~                      !
                                   The emissions from  the  combustion  of  the  retort  gas
                               are determined by:                                      :
                                   .  volume and heating value  of  the retort gas
                                   .  presence of sulfur in  the ;gas
                                   .  presence of nitrogen as ammonia or organic
                                      nitrogen compounds
                                   The retort gas flow rates for  three types of
                               retorts are shown in Table  4.                          '
                                        Table 4.  RETORT GAS PRODUCTION RATES
                                Type
                               Retort
                Gas Produced
                m3/m3 of Oil
Heating Value
  g-cal/L
                               In-situ               7000
                               Direct combustion    1800
                               Indirect  combustion    180
                                            9000
                                            9000
                                          90,000
 Room & pillar with
   catalytic converters
     on engines
15   2  350
               20
                    180
    The in-situ process produces the highest retort gas
flow rate due to the combined effect of the higher
retort temperatures converting more of the kerogen to
gas and the higher dilution gas flow required t:o pro-
vide adequate oxygen to burn the 'shale.  The direct
combustion retort has similar conditions (i.e.,, high
temperatures and requirement for adequate oxygen for
combustion) but to a lesser degree than the in-situ
retort and, consequently, has lower retort gas flow
rates.  The indirect combustion process has the lowest
retort gas flow rate due to the lower retort tempera-
tures and low gas flow rate with no dilution required
to provide oxygen.               ;
    The heating value of the retort gas is determined
by the amount of dilution gas.  In-situ and direct com-
bustion retorts produce low heating value gas at
9000 g-cal/L (100 Btu/scf) and the indirect cotabustion
retort produces high heating value gas at 90000 g-cal/L
(1000 Btu/scf).                  ;
    The retort gas flow and heating value determine the
net exhaust gas flow after combustion.  As the perform-
ance of the air pollution controls is often determined
                                 '   o
by an exit concentration (ppm or:g/m ), high gas flow
rates result in higher pollutant'emissions.
                      Sulfur Gases
    During the retorting, the sulfur in the raw shale
is partitioned to the spent shale  (60 percent), oil
(10 percent), and retort gas (30 percent).  The signif-
icant variations in raw shale sulfur content, percent-
age of sulfur partitioned to the gas phase, and chemi-
cal structure of these sulfur gases result in the
sulfur gas cleanup strategy being  quite difficult.
    The sulfur emission problem  can be solved either by
removing  the sulfur prior to combustion or by adding a
flue gas  desulfurization process'after combustion.  As
the combustion process dilutes the pollutant concentra-

-------
tions and increases the gas flow rate, the economically
prererred technique usually is sulfur removal prior to
combustion.
    The form of the sulfur in the gas is extremely
important when considering sulfur removal processes.
Sulfur recovery processes that have been considered for
cleaning the, retort gas prior to combustion are not
effective in removing organic sulfur compounds which
can amount: to as much as 10-16 percent of the total
sulfur in the retort gas.  Consequently, the effective-
ness of these cleanup processes depends on the relative
amounts of organic sulfur to H2S.  Even high efficien-
cies of H^,S removal (99 percent) are not sufficient to
reduce the sulfur emissions below the 850 kg/1000 nr of
oil (0.3 Ib/bbl) regulatory level for Colorado if there
are significant amounts of organic sulfur gases.
    To avoid the costly alternative of adding an end-
of—pipe flue gas desulfurization, two alternatives can
be considered.  The first, the activated carbon-
hypochlorite HoS removal process, is an improvement on  '
the H2S scrubbing process which also removes organic
sulfur species (Reference 11).  Therefore, this process
is effective for removing sulfur gases prior to combus-
tion, eliminating the need for more expensive post com-
bustion control.  The activated carbon-hypochlorite
process reports removal of 99+ percent of the H2S and
90-98+ percent of the organic sulfur gases.  This
results in a net sulfur removal efficiency of 99 per—
                                                ^
cent and sulfur emissions (SOX) of 500 kg/1000 m  oil
(0.17 Ib/bbl) even when the organic sulfur gases are
15 percent of the total sulfur.
    The second alternative to the use of post-
combustion SOX control is the indirect combustion-
solids recycle retort process which limits sulfur gas
emissions by the chemistry of the retort and combustion
process.  The sulfur contained in the retort gases from
the recycle solids process can be as low as 1 percent
of the total sulfur content in the feed with proper
design of the retort.  [The remaining sulfur is parti-
tioned to the oil (10 percent) and the spent shale
(89 percent)].  Therefore, the amount of organic sulfur
is minimal, and the H2S removal processes alone are
sufficient to reduce the sulfur emissions below the
regulatory limit.
    The design conditions used for the H2S removal pro-
cess determine the residual H2S and organic sulfur in
the retort gas that eventually are emitted as sulfur
oxides.  '.Che processes considered are:
    1. Direct or indirect conversion of the sulfur
       (e..g., Stretford, Lo-Cat, Unisulf, alkaline, or
       amine scrubbing)
             H2S exit concentration - 50 ppm
             organic sulfur assumed at 5 percent
             of total sulfur in retort gas
               - no removal
    2. activated carbon-hypochlorite process
             HnS exit concentration ~ 1O ppm
             organic sulfur - 90 percent removal
    Two process operations result i in the direct emis-
sion of sulfur oxides from the retort:  the circular
grate direct heated retort and the fluidized bed com-
bustion of the spent shale.  The design conditions for
these two processes were taken from the literature
(References 8 and 9).
           .
    Circular Grate Retort - SOX - 175 ppm in retort gas
    Fluidized Bed Spent Shale Combustor in Retort Gas -
                            SOX =,20 ppm
                     Nitrogen Gases
    The removal of nitrogen gases,is also difficult to
predict due to the degree of variability of nitrogen
content in shale, partitioning between gas, oil, and
spent shale, and chemical form of the gaseous nitrogen
species.
    Using the data reported in the PSD permit applica-
tions and the Pollution Control Technical Manuals, the
partitioning of the nitrogen was estimated as shown in
Table 5.
    Table 5.  PARTITIONING OF NITROGEN IN RETORT
Process
                     % of Raw Shale Nitrogen in Product
                    Spent Shale   '  Oil      Retort Gas
In-situ (MIS)
Solids Recycle
Direct Combus'n
(Lurgi)
(Chevron)
(Paraho)
21
5
33
1 25
55
37
54
30
a)
^a' Remaining nitrogen content in the spent shale after
    retorting is burned in the lift pipe or fluidized
    combustor and exits with the flue gas.
    The nitrogen content of the retort off-gas consists
primarily of ammonia with smaller amounts of other
nitrogen compounds.  In a semi-quantitative investiga-
tion of nitrogen-containing species from an in-situ and
above-ground retort process, hydrogen cyanide, various
nitriles, pyrrole, pyridine, methyl and diethyl ani-
line, and other nitrogen gas species were identified
(Reference 12).  The organic nitrogen content of the
retort gas was found to be as much as 1—2 percent of
the ammonia content.
    The presence of organic nitrogen species presents
the same problem for limiting fuel—related NOX emis-
sions as that described above for the SOX emissions;
namely that the removal processes,generally considered
are not effective in reducing the organic nitrogen con-
tent of the retort gas.
    The primary nitrogen removal technique considered
is removal of ammonia from the retort gas by a water-
wash absorption tower followed by'an ammonia recovery
stripper.  The outlet ammonia concentration is deter-

-------
mined by the effectiveness of the ammonia absorber.  At
atmospheric pressure the equilibrium exit partial pres-
sure for ammonia at 50°C is 0.5 mm Hg.
    The nitrogen content of the treated retort gas, and
the subsequent NO  emissions from combustion of the
retort gas, is determined by:
    .  the exit gas ammonia concentration (660 ppm NHg)
    .  the amount of retort gas produced by the retort
       (Table 4)
    .  the amount of nitrogen partitioned to the retort
       gas (Table 5)
    .  the percentage of nitrogen present as organic
       nitrogen compounds (2 percent)
    The design conditions used to determine the NOX
emissions from burning the retort gas are:
    .  water wash - NH3 in exit gas based on
                    NHg partial pressure - 0.5 mmHg
                    organic nitrogen  based on 2% of
                    nitrogen in retort gas and no
                    removal with water wash
                  — thermal NOX from  retort gas rate,
                    heating value  and  0.2  lb/106  Btu
     .  acid wash  - same as water wash  except  KH3
                    assumed =  10 ppm
     The processes that utilize the  combustion  of high-
nitrogen-content spent shale for energy recovery can
produce high NOX emissions if  proper staging of the
combustion is  not used.  For PSD permit applications,
NOX  emissions  were based on a  high  estimate of 15  per-
cent for  the conversion of the nitrogen in the spent
shale to  NOX in the combustor. This level of  conver-
sion was  also  found by Lawrence Livermore Laboratory
investigators  who did not attempt  to stage the combus-
tion (Reference 12).
     The  principle of NO reduction  in a staged  combustor
can  be applied to reduce these ,high NOX emissions  to
approximately  3 percent nitrogen  conversion to NOX.
RETORT GAS COMBUSTION AND END-OF-PIPE
     The  end-of-pipe controls  are  those either  added
after combustion  of the  retort gas to remove particu-
 late, NOX, and SOX, or  incorporated as part of the com-
bustion  process as  in  staged combustion for NOX
 control,.
                       Particulate
     For particulate control, two alternatives were con-
 sidered.  The first is the base case using a standard  .
 baghouse.  The second control technique is the combined
 dry venturi-baghouse (Reference 13).
     The dry venturi—baghouse combination provides for
 particulate control that is somewhat independent of
 type of particulate.  The applicants for PSD permits
 all considered a minimum particulate exit loading of
 0.07 g/m^ (0.03 gr/scf) which was based on standard
technology within the limits of the unknowns associated
with oil shale particulate.  By capturing the small
particles on larger target particles of specified phys-
ical properties, the dry venturi eliminates the major
uncertainties in designing baghouses with respect to   .
particulate type and size.
    Two design conditions based ,on the face velocity in
the baghouse of 0.5 and 1.5 in/sec were used (Refer-
ence 14):
             g/n>3
       particulate loading  face velocity
                            m/sec
             0.02            0.5
             0.001           1.5
                     Sulfur Oxides
    If the H2S (and organic sulfur) removal is not suf-
ficient to reduce the sulfur emissions to an acceptable
level, a post-combustion flue gas desulfurization sys-
tem must be added.  This could be either a wet or dry
scrubber.                       i
    Another sulfur control technique is the use of a
spent shale combustor.  The combustion of the spent
shale has two important advantages:  1) recovery of the
energy value of the char, and 2) reduction of the
sulfur oxide emissions due to the scrubbing nature of
the spent shale.  However, spent shale combustion also
has two distinct disadvantages:  1) high emissions of
NOX from the nitrogen in the spent shale, and 2) high  .
emissions of CO due to incomplete combustion.  These
emissions (NOX and CO) are discussed in the following
section.                        :
    The design conditions for SOX emissions used in the
analysis are:                                          :
    .  Flue Gas Desulfurization  - 50 ppm SOK exit con-
       centration
    .  Combustion of spent shale with retort gas
                                 - 10 ppm SOX
                                ! - 300 ppm N0,{
                                ; - 1000 ppm CO
                    Nitrogen Oxides
    Two controls were considered for reducing post-
combustion NOX emissions.  The first is ammonia injec-'
tion;  this technique has been applied successfully in
utility boilers and could be used when  the retort gas
is burned in a conventional boiler for  steam and elec-
trical generation.
    The second NOX control  considered is  staged combus-
tion  which has particular advantages for  spent shale
combustion due to its ability  to adequately  control
fuel  related NOX.  The  staged  combustion  could be
applied  to either the conventional boiler or  the spent
shale combustor.
    Selective  catalytic reduction  (SCR) was  not consid-
ered  due  to  its  potential  for  poisoning the  catalyst.  ,

-------
The retort gas participate contains a wide variety of
heavy metals which have a deleterious effect on cata-
lyst life.  In addition, there are still a number of
unknown factors which can affect the long term catalyst
performance that have not been completely identified.
For example, it had been assumed that the mercury asso-
ciated with the retort gas was in the form of elemental
mercury and would be substantially removed prior to
combustion during the standard gas cleaning (ammonia
and hydrogen sulfide removal) processes.  However, it
has been shown that the mercury is present primarily as
methyl mercury which is volatile and is present in the
retort gas during combustion (Reference 15).  Conse-
quently, any post-combustion control processes must be
capable of handling these emissions of elemental and
oxides of. mercury.  This is only one instance where
unknown factors could have a negative effect on cata-
lyst performance.  Consequently, due to the inherently
variable nature of the retort gas from an oil shale
retort and the known presence of many catalyst poisons,
the use of SCR was not considered.
    There is a tradeoff between the NOX and CO emis-
sions in the spent shale combustor.  As indicated
above, the NOX and CO emissions are quite high (300 and
1000 ppm,, respectively).  Higher combustor temperatures
increase the NOX emissions but decrease the CO emis-
sions.  In the range of 600-800°C the NOX emissions can
range from 250 to over 600 ppm, while the CO emissions
can vary from 200 to over 1000 ppm at the lower temper-
atures with low excess oxygen.
    In addition, the staged combustion technique of
controlling fuel related NOX depends on low excess
oxygen (perhaps sub—stoichiometric combustion) which
would further increase the CO emissions.
    The fluidized bed combustor has limitations in pro-
cess control which result in its inability to provide
conditions that result in adequate staging for NOX con-
trol.  However, a cascading bed combustor may be
designed as a staged device (Reference 12).  Conse-
quently, combustion conditions can be controlled at
each stage of the process, alternating between fuel
rich and fuel lean zones to reduce the NO formed to ^
and complete the combustion of the CO formed to C02.
    There are as yet no specific test results of the
staged combustion with spent shale.  However, the reac-
tion kinetics of the reduction of NO to N, with spent
shale have been investigated (Reference 12), and the
engineering design of the cascading bed combustor is
ideal for a staged system with easy means for control-
ling the process conditions.
    The design conditions for post-combustion NOX con-
trol are:
    .  Nlij injection — 20 ppm NOX exit concentration
    .  Cascading bed combustor — 50 ppm NO
                               - 50 ppm CO
EMISSIONS FROM RETORT GAS COMBUSTION
    The results for the emissions from the retort gas
combustion from the analysis for all five cases are     !
shown for particulates, sulfur oxides, and nitrogen
oxides in Tables 6, 7, and 8, respectively.  The total
facility emissions for each of the five cases  are shown
in Tables 9 through 13.  The emissions from combustion
of the retort gas for particulates, nitrogen oxides,
and sulfur oxides, are also shown in Figures 8, 9, and
10, respectively.
    These figures indicate that there is wide  variation
in emission levels for the five processes based on the
present day technology (Base Case conditions).  For
particulates, the emission levels vary from 200 to
800 kg/1000 m3 of oil; for nitrogen oxides the emission
levels vary from 1000 to 8000 kg/1000 m3 of oil; for
sulfur oxides the emission levels vary from 350 to      ;
3000 kg/1000 m3 of oil.
    The first alternative considered was the use of the'
activated carbon enhanced I^S removal process, an acid
wash for improved ammonia removal, and the addition of
a dry venturi-baghouse for post-combustion particulate  .
control.  Referring to Figures 8, 9, and 10, the emis-
sion levels for alternate No. 1 show considerably less
variation, particularly for sulfur oxides (range from
100 to 250 kg/1000 m3 of oil) and particulates (range
from 50 to 200 kg/1000 m3 of oil).  The variation of
nitrogen oxide emissions is still considerable, ranging
from 1000 to 4000 kg/1000 m3 of oil.  Essentially, the
acid wash removes only the residual ammonia (without
affecting the organic nitrogen content) and has no
effect on the thermal NOX; therefore, there is rela-
tively little improvement in the NOX emission rate.
    The second alternative considered was the use of
ammonia injection for NOX control from boiler and/or
furnace combustion, the use of staged combustion for
control of NOX emissions from the spent shale combus-
tor, and the dry venturi—baghouse with an increased
space velocity which improves collection performance at
the expense of increased pressure drop.  Again, refer-
ring to Figures 8, 9, and 10, it:is apparent that the
addition of these controls essentially levels the per-
formance of all five processes.
    Figures 11, 12, and 13 show the particulate, nitro-1
gen oxide, and sulfur oxide emission levels for alter-
nate No. 2 conditions along with'the total facility
emissions.  The particulate emissions, Figure  11, still
                                     n
show variation from 4 to 12 kg/1000 m  of oil.  How-
ever, the absolute value is considerably less than par-
ticulate emissions from the mining and solids handling
operations, and the total facility is essentially
equivalent for all five cases, ranging from 180 to
200 kg/1000 m3 of oil.
    The nitrogen oxide emissions'(Figure 12) range from
75 to 500 kg/1000 m3 of oil. While this is still a sig-

-------
nificant variation, again the absolute magnitude of the
values is such that the net variation in the total NOX
emissions for the five facilities is less than a factor
of 2, ranging from 400 to 800 kg/1000 m3 of oil.
    The sulfur oxide emissions (Figure 13) range from
100 to 250 kg/1000 m3 of oil and are essentially the
same for the total facility as there are no other sig-
nificant sources of sulfur emissions.
    The basic conclusion derived from the above anal-
ysis is that:, although the air emission levels for the
different retort processes with controls considered to
be Best Available Control Technology (BACT) can vary
considerably, sometimes by as much as two orders of
magnitude, the application of control techniques that
are either improvements over existing technology or
more suitable for a specific application, results in
similar emission levels for all five processes consid-
ered.  This statement does need to be qualified by the
fact that some of the control techniques considered
have not been applied specifically to the oil shale
recovery process, and therefore cannot be considered as
BACT.  However, these techniques have been proven at
the full scale level in various other difficult control
applications.
    REFERENCES
                                  i
1.  Denver Research Institute, Pollution Control Tech-
    nical Manual for Lurgi Oil Shale Retorting with
    Open Pit Mining, EPA-600/8-83-005 (NTIS
    PB83-200204), April 1983.
2.  Denver Research Institute, Pollution Control Tech-
    nical Manual for Modified In-Situ Oil Shale Retort-
    ing Combined with Lurgi Surface Retorting,
    EPA-600/8-83-004 (NTIS PB83-200121), April 1983.
3.  Denver Research Institute, Pollution Control Tech-
    nical Manual for Tosco II Oil Shale Retorting,
    EPA-600/8-83-003 (NTIS PB83-200212), April 1983.
4.  Union Oil Company, PSD Permit Application for
    Union's Phase II Oil Shale Mining and Retort Facil-
    ity, December 1982.           '.
5.  Paraho Development Corporation, Permit Application,
    Paraho - Dte Facility, November 1981.
6.  Quintana Minerals Corporation,; PSD Permit Applica-
    tion, Syntana - Utah Commercial Shale Oil Facility,
    undated.
7.  Magic Circle Energy Corporation, PSD Permit Appli-
    cation, Utah Cottonwood Wash Project, Submitted to
    State of Utah, Department of Health, April 22,
    1982.                         ;
8.  Chevron Shale Oil Company, PSD Permit Application,
    Clear Creek Shale Oil Project, Submitted to State
    of  Colorado, August 1982.     '•
9.  White River Shale Project, PSD Permit Application,
    White River Shale Project, Submitted to State  of
    Utah, August 1981.
10. Cathedral Bluffs Shale Oil Company, PSD Applica-
    tion, Cathedral Bluffs Shale Oil - MIS-Union,  Sub-
    mitted  to State of Colorado, July  1981.
11. Bhatia, S. D., et al., Novel Wet Scrubbing Tech-
    niques  for the Removal of H2S and  Organic Sulfur
    Compounds, Presented  at  the 4th International  Clean
    Air Congress, Tokyo,  May 1977.
12. Lawrence Livermore National Laboratory, 5th Brief-
    ing on  Oil  Shale Research, Livermore, CA, October
    1985, Report MISC-4293.
13. Teller, A. J. and Roy, D. R., The  Dry Venturi,
    Private Publication by Teller;Environmental
    Systems,  Inc.,  Shrewsbury, Massachusetts,  1985.
14. Ondich, G. G.,  Control Technology  Appendices  for
    Pollution Control Technical Manuals,
    EPA-600/8-83-009  (NTIS PB83-214734), April  1983.
 15. Fox, J. Phyliss,  Private Communication, December
    1985.                         i

-------































03
Z
O
C/3
03
S
u
u
B
M
m
•a!
Cu


vO
0)
r~(
I



































u

51 o
CM •<-( 0)
O 03
4J tH 1j
3 * *"
m
0) •
CJ CO
CO
[K

n
^
r> o
— •HO
O CO
• 0 -^
4J tH 4J
»a1 !> "*"*
in
8 J
CJ "H
CO
[14







0)
CO
pa

















































CO
^^
bO O
M 0
o



bo
• e

CO *^
PL, O
CO

00 O
& O
0



be
• C
J-) t-l
CO *CC
EH 0
CO
e

bO o
^ o
o




. c?
• c
<-> -rf
PL, O
^
CO
§ g 4
O iH CO
1*. B


e
c
4J O
CQ t-(
J3 s
B -3
8 3
CO
B
CO
B
4J
O CO ^-,
Li CO CO
« m

>
33
















CO

CJ
o

Cu

111
CO
Ct
u
^
0

o




CO
^
bo
r-l


U-l
O





ea
t3
bO


•H
O


0




CO
bo

r-l
o





0
•H
CO



r-t
O

g

^
^
CO
CJ

























•a-






0
0
o


o







CN
o
0



-a
CN







O
O

CM
CO





CM



_^
VO
_H

O
o
o
o

o
8












e
o
•H
4-1
CO
S
e
o
o

u
o
cu
u

Q
^
— <
-a-






o
0
o


o







CM
o
0.



-a-
CN







o
o

CM
m
CO





CM



(— 1
VO
.— 1

O

o
o
o
a*








o
o

u






in ^
.—4


yj -(
r^ co
~H CO
CT\

O 0
0 0
0 0
~-t CO
m
o
o
o



^l •
o u->
oj S

\ CJ
CO

•H OJ
O
CO T3
1 
-------






























en
2
0
M
en
en
£
Ed

W
O
1-1
S

OS
g
£3






01
l-t
.0
CO
H


































































CO
CO
01
o
o
Pu
B

•£
. <3
•o
01
.u
CO

1-1
4J




































X
o
en
B
O

' X CO
O CO
en T-I
m
a
B
0
•H
X CO
O CO
en -H
e
w


CO •

§14-1
U
Oj




en
CM


X
o
en


B
O
•H
CO
X co
O i-l
en a
W


B
O
•H
X CO
O ca
en i-<
e
u



en ,en
CM
0 33
-rt
B tu
CO O
bO
o m






































4J
•H
X
sa

0
o
o
t-H

J?


§
o


bo
^



u
o





'"x
o
en


•H

W



o
o
o
•H
~bO
J(i



o
o
o
1— 1

bO





0
o
o


bo





e
Cu
ex

t-H rH
i-l CO
° S
ro H
S


,_,
•H
O en
CM
co 33
a
CO
O
B
CO 9-8
bO
IJ
O



a
ex
ex


g

ex



i-H i-l
•H CO
O U
o






•H
O en
CM
co ys
e





iH
•H
O


a





en s
CM O.
33
































o.




















CO
CO
01
0
o

Ol

01
eo
CO
0
m o
-a m



ON NO
m o







o





o
ON




CO
O ^0
-H m



\c m
CO O
r*> •-*




CO Q
m co
M






ON










r» O


— «


^^
co

o m
m i-.



0)

co
u
bO


co

1
•H
O
1
B B
O O
•H -H
4-1 4J
CO CO
3 3
"a "a
O O
O CJ

4J 4J
0 0
01 Ol

•H i-l


« «
-H CM
r*.
-a-




CM













O
ON





0
l—t



-a-

ON




-a-
-a-
CM






0
CM









I—
ON

~4





O
in




01
i-l
O


01


CO
CO
M
o
•H

CO
1

o

4J
O
01
u
•H

B
M

•
CO
CO -*
CM



oo -a-
CM







-a-





.0
ON





o
•— i



o -*•
^r CM





2 a







o
CM







J3

CO








0 O
in CM


0)
*H
o

CJ
01
h

•o
•H T3
i-l 0)
O N
E a
O 3
•H i-l

CO

a
o
o

4J
O
01
u
•H


I-l

•
-a-
CO
^H




ON
t—l







ON

^H



0
ON





o




CM
ON
t-H




o cn
en en
CO O
CM






O !•»
ON
m








oo








0
in
















3

i-t
m
1
•H

•0
0)

<4_|

•0

y

«
in
^
-a-



O —1
in -a-
CM







CM





O
ON





o




vO

ON




00 O
NO O
ON O
m






00










o
00
-a-






o
in


•

u
0)
u

CO
CO
bo

4J
CJ
0)
u
1-1
•o
B
1-1
1
1
o
tJ
bo

0)

Q

CO







i








































.






JJ
I-l
o

01
14

. B
1 1-1

en
f)
X
^U-l
0)
3
4J
ex
; co
•• u

CO
73


0
CO

: o


01
;-§
u
3
U-l

3
CO
X CJ
S 3
«
S W)
& ' 1-4
& O

m ! S





cd .^.Q
*-S ':*+*

-------
O i-4
	 1 ,-^ 	 I























CO
z
0
M
CO
to
i
s
o
i
2
3
o
9
H
z
^
CO

01
•§




















e eg o o
0 X U —
1-1 o o — en
u z H bo B
CM 0 Ji
0)
J 1? "0
H M 1-1 B x a
g * o o o.
en w o z a.
55




u u
•H c x e
x o o o.
w o z a.
iH
•H
o
X 0> «-"
2 i o °
z IH en
n s
J2 Z
co 6 O

» O.1-4 r-4 O
-< C »* 01 X -<
• -a • to CM 3O —
4J  o
0
i— 1
"hft
"i o* "*

e
fcj 01 z
•H BO
C 0 IH
CO M O
in bo u
• M -H »*
O O Z CM
11

CO O t—4
en to O i-<
33 01 4J O O
Z t< iH *-"
cu X ' — en
W M B
• •%
u
tt}



CO
to
oi
u
o
P
Cu
01
to
o
o en so ^ m cjs
inmenensoeM oomcMr-.
QOCOsOsDinsD *3"CM CM




O 0 O O 0 O O
CM CM CM CM in CM CM

•n **n •*-} •*"> -O
J 5 5 5 4i c1 c1
1 1 1 1 "O "71 T*
en en en 
OS OS .3- -* so «J\ -»
i~. r~- en en en so en

CMCM^H^^O\»-^ r^CMCTS
rtrfcMCMinco in-a-so
oo oo os c?^ f^ ^^ *^" "
c\j CM co *^ c^ r*^




oooooo ES
r>. t^ r^ r-. "•" £J
in in in m C<




CMCM-HrfOsO CM--;
•IH
: O
§
1-t
m
0)
§
u
,»<
i2
' IS
o
T3
01
to
.s
i ff^
n)
«k^

-------
            Table  9.   TOTAL  FACILITY  EMISSIONS

                Case  1 - Direct  Combustion
                    (Paraho type  retort)
Base Conditions; ^S removal (50 ppm) , NH-j water wash
o
Particulate loading = 0.07 g/nr
o l
Emissions, kg/1000 nr of oil
Base Case
Source
Open pit
Retort gas
Upgrade
Total
Alternate 1
Pollutant

CO
370
125
25
520
S0x -
NOX -
PM -
CO
Room & pillar 150
Retort gas 125
Upgrade 25
Total 300
Reduction 220
% Reduction
Alternate 2

Pollutant
42

HC NOX SOX
50 471 38
35 2812 2593
150
235 3283 2631
H^S removal - activated carbon
NHg removal - water & acid wash
dry venturi & baghouse
room & pillar mining
HC NOY SO..
A A
20 345 20
35 1156 159
150
205 1501 179
30 1782 2451
13 54 93

PM
r
410
246
656 :
r
PM
181
70
251 : '
405 ;
62
1
NOX - ammonia injection
PM - dry venturi/baghouse '
HC & CO - vehicles ;

CO
Room & pillar 15
Retort gas 125
Upgrade 2
Total 142
Reduction 158
catalytic converter
HC NOX SOX
2 345 20
35 85 159
60
97 430 179
108 1071 0

PM ;
181 '
4
185
67 :
% Reduction
52
53
71
27

-------
            Table 10.  TOTAL FACILITY EMISSIONS

                 Case 2 - Direct Combustion
                   Circular Grate Retort
Base Case
Pollutant
Room & pillar
Retort gas
Upgrade
Total
Alternate 1 SOX
NOX
PM
HC &
Emissions,
CO HC
150 20
125 35
25 150
300 205
kg /I 000 m3
NOX
345
2812
3157
of oil
S°x
20
370
390

PM
181
246
: 427
- Flue Gas Desulfurization
- NHo removal - water & acid wash
- dry venturi/baghouse
CO - vehicles
catalytic converter
Pollutant
Room & pillar
Retort gas
Upgrade
Total
Reduction
% Reduction
Alternate 2 NOX
PM

Room & pillar
Retort gas
Upgrade
Total
Reduction
CO HC
15 2
125 35
2 60
142 97
158 108
52 53
— ammonia injection
- dry venturi/baghouse
CO HC
15 2
125 35
2 60
142 97
0 0
NOX
345
1156
1501
1656
52

NOX
345
85
430
1226
S0x
20
106
126
264
68

s°x
20
106
126
0
PM
1 181
70
251
' 176
; 41

' PM
181
4
185
66
% Reduction
74
26

-------
            Table 11.  TOTAL FACILITY EMISSIONS

                Case 3 - Indirect Combustion
                 Gas Recycle Retort (Union)
Base Case
Process CO
Room & pillar 15
Retort 125
Upgrade Z_>
Total 165
Alternate 1 NOV -
sox-
PM -
Pollutant CO
Room & pillar 15
Retort 125
Upgrade 2
Total 142
Reduction 22
% Reduction 14
Alternate 2 NOV -
X
PM -
Pollutant CO
Room & pillar 15
Retort gas 125
Upgrade 2
Total 142
Reduction 0
Emissions, kg/ 10 00 m^ of oil
HC NOX SOX
20 345 38
35 921 2414
205 1266 2452
NHo removal - water & acid wash
HoS removal — activated carbon
dry venturi/baghouse
HC N0v SO,,
X X
2 345 20
35 756 124
60
97 1101 144
108 166 2308
53 13 94
ammonia in j ect ion
dry venturi/baghouse
HC NOX SOX
2 345 20
35 63 124
60
97 408 144
0 692 0
PM
181
185 :
366 :
!
PM ;
181
53
234 .
132 i
36
:
PM '
181 1
3
184 ,
50 •
% Reduction      0        0       63        0       21

-------
             Table 12.  TOTAL FACILITY EMISSIONS

                Case  4 - Indirect Combustion
           Solids Recycle/Fluidized Bed Combustor
Emissions, kg/1000 m3 of oil

Process
Room & pillar
Retort gas
Combustion
Upgrade
Total

CO
150
125
11197
25
11497

HC
20
35

1 cr\
J.D\J
205

N0v
A.
345
921
3359

IN O L/3X 3.
4626
t
sox
20 i
106 :
224


350 :

PM
181
185
653


1019
Alternate 1
    removal - water & acid wash
    removal-activated carbon
NOX - NH
S0x ~ H2
PM  - dry venturi/baghouse
HC & CO - vehicles
                       catalytic converter

Room & pillar
Retort gas
Combustion
Upgrade
Total
Reduction
% Reduction
CO
15
125
11197
2
11340
158
1
HC
2
35

60
97
108
53
NOX
345
756
3359

4460
166
4
S0x !
20
8
224 ,

252
98 ;
28 ,
PM
181
53
187

420
599
59
Alternate 2
             PM  -
ammonia injection for retort
staged combustion for combustor gas
cascading bed spent shale combustor
dry venturi
Room & pillar
Retort gas
Combustion
Upgrade
Total
Reduction
      CO

      15
     125
     423
       2
     565
   10775
                     HC

                      2
                     35

                     60
                     97
                      0
 NO..
   A.

 345
  63
 423

 831
3629
S0x

 20
  8
169

197
 55
 PM

181
  3
  9

193
227
% Reduction
      95
                      0
  81
 22
 54

-------
Table 13.  TOTAL FACILITY EMISSION

     Case 5 - Modified  In-situ
 Indirect Combustion Above-Ground
Emissions, kg/1000 m3

Room & pillar
Retort gas in-situ
Above ground
Upgrade
Total
. Alternate 1 NOX - NH3
SOX - H2S
PM - dry
HC & CO -
CO HC
150 20
72 35
53
25 150
300 205
removal - water &
removal-activated
venturi/baghouse
vehicles
NOX
345
7342
369

8056
of Oil
S0x
20
2033
968

3020

PM
i 181
739
; 74
l

994
acid wash '.
carbon








catalytic converter s

Room & pillar
Retort gas in-situ
Above ground
Upgrade
Total
Reduction
% Reduction
CO HC
15 2
72 35
53 0
2 60
142 97
158 108
52 53
NOX
345
2375
302

3022
5034
62
S°x
20
191
50

261
2760
91
PM
181
211
21

413
581
58
Alternate 2 NOX - ammonia injection
PM - dry

Room & pillar
Retort gas in-situ
Above ground
Upgrade
Total
Reduction
% Reduction
venturi/baghouse
CO HC
15 2
72 35
53 0
2 60
142 97
0 0
0 0

NOX
345
254
25

624
2398
79

S0x
20
191
50

261
0
0
r
' PM
! 181
11
1

! 193
221
53

-------
0)
I-l
    a:
c7>
co
2:
q
co
CO
2E
LU
    13
    O
    u.
    O
                             CO
                             c
                             o
                            "55
                             co

                             o
                             "co
                             CO
CO

o
"co
CO
CO   -IS
                                     Ej-   OT
                                     c   •=
o
'co
CO
                            LU
                                LU  LU
                                     111
                                 X  X
                            o  o  o  o
                            O  2  CO  I
co

o
'co
_to


LU
                                            D
CO           !

CO M O »     i-l
M U S I td   rt
•o   6 o e   ;
o> n o us a c eg
J- C8 4J CO 4J O 4->
4J V 4J M C -H -H
IS <—I O 0) P^ C »C
O O O (X, to C3 5
             H
             §
               I  I  I


              O O PL.
                                                             I
                                                            8
1 	
o
o
o
CM
O
o
o
o
0
o
o
CO
o
o
0
CD
0
0
o
^t
o
o
o
CM
o



                          jo  9iu  OOOL/*>I

                          SNOISGIIN3

-------
bO
•H
    ICO
    ;z

    o
    o
    o
QJ   H
ICO
ICO
ILU
O
O
IDC
iQ.

ILJJ
     ffl
     ico
     iCO
     o
     Q.
LLJ
Q.

Q.

LL

O

Q


111
t
CO
o
  -
o
HI
DC
                                         UJ
                                         CO

                                         o

                                         o
                                         <
                                         CD
      CO
     .=>
      OQ
      2
      O
      o
                       CO
                       «
                       9 2
                       O UJ
                       < oc
                                        OC
                                        UJ


                                        6
                                        CD
                                                      x
                                                      O

                                                      UJ
                                                      o
                                                        63
                                                        UJ <
                                                        CLZ
                                                        cow
                                                                       111
                                                                       o
                                                                       <
                                                                       tr
                                                                       o
                                                                       D.
                                                                       ID
                               DC
                           o
                           oc
                           2
                                            CO
                                               i3
                                  o
                                  UJ
                                  OC
                                  5
         o
         o
       QC Q.
       UJ >-
       CD X

< 3 S §> CO
^ OC — OC OC
_J O 2 O <
< CO < CO O
                                                  UJ UJ
                                                  z CD
                                                  j m
                                                           UJ
                             -o ±-~ m
                              •o
                               oc

                                           .
                                                     O

                                                     UJ
                                                     _i

                                                     I
                                                     CO
                    2
                 O
                 OCQ
u.
t
                                            UJ
                                    5
                           oc
                           o
                       o
                       X


                       1   K   g  S
                       (LOo-O

                       i    i    S   i
                                                        o
                                                        UJ
                                                        OC
                                                       o
                                                       CO
                                                 V /     x
                                                                  OC
                                                                  o

                                                                  UJ
                                          5

                                          OC
                                              UJ
                                              co
                      a.


                      c?
                      O
                                                                 co
                 UJ

                 < -i
              «'•  X ^
              o  WS
                                                       CO

-------
0)
f-l

a
CO

g
co
CO


111

111
Q
X
o
   o
   CD
   DC
   O
                   BO jo pui ooo i/EH

                    SNOISSIW3 OO

-------
CO
z
o
CO
CO
IU

z
o
QQ
CC
O
o
CC
X
             HO
O  O O  O  O O  o
|
              SNOISSIIAI3  OH

-------
                                            UJ CO
                                         CO
                                              x
                                              UJ
                                              >
in
a
•H
   CO

   o
   CO
   CO
UJ

CO
UJ
g
X
o

UJ
O
O
cc
                   SNOISSIIAI3 XON

-------
  CO
  z:
  Q
  CO
  CO
00
LLJ

CO
LLJ
Q
X
Q
ec
  CO

-------
  c:  -^3
  E:   co
  ^  J5
  -D  JD
           0>


           1

           O)
       O
                CO
                CD
O  CD

               D
                         CO
 o   °
 CO

 CD
 O
 CO
tl  -O

 CO  CsJ
               0
                                  0
                                  O)
                                  CO
                             .£   O
o:t

">i
0   w

?   "°
O   •
8   £
                           CD
                           D)

                           2
                           c5
                          •*—•
                           CO

                           CO
                           CD
                             CD
                             CO
                             C
                             CD
                             Q.
                             CO
                                 CD  3
CD
D>


CO
                                      CO
                                      CD
     •c
     O
&   "S
co   i-
 co
 CD


 CD
TD
 CO

 O>
 CL
r-»

(U

s

•H
CO
z:

LLI

111

a:
<
a.
       o
       CD
                                       ss:
                  %&S&^^
o
to
                       +
                                              II  I  Hill
                                          CO
                                       ^- co
                                          HI
                                          o
                                           1
                                       i o
                                          DC
                •₯
                                          +
                          o
                          CM
                                  O
                                  CO
                      HO *o e UJ 0001.

                     SNOISSIJN3 UUVd

-------
DO


0)
60
        CO
        111
        CO
        o
     ILL. U.

     co cc
     Z O
     O u.
     CO 2
     COQ
|

-------
CO
zco
OLU

60
•H
  j
OOC
xO
gu.


                   T-  |

                 SNOISSIW3  XON

-------
CM
^fe
                                         CO
                                   Ill  111  111  HI  LLJ
                                   CO  CO  CO  CO  CO

                                   O  O  O  O  O
CO
x "•
u.


5? CD
o
o
o
CO
o
o
to
CM
O
o
0
CM
O
o
to
1—
o
o
o
f—
o
o
10

o



                    HO jo gin O00t/B>|

                                  XOS

-------
CO
ill
CO
Q

LU
DC

O
LL.

CO


O

CO
CO



LU

LLJ
g

oc
<
a.
           CM co
                  in
        LU  LU LU  LU LU
        CO  CO CO  CO CO

        O  O O  O O


        • H M M D
                               LUOjZ

                               
-------
    CO
    III
    CO
 co<
 LUO
 gill
 x
(50
ZDC
82
^Q
  CD
  LLJ
           111 LU  LLJ LU 111
           CO CO  CO CO CO
           < <  < < <
           o o  o o o

O
o
0
1
o
o
00
—— 1 —
0
o
1^
— 1 —
o
o
CD
— 1—
o
0
IO
— l —
o
o
^f
— 1 —
0
o
CO
— 1 —
o
o
CM
•— 1 —
o
o


o


                   HO jo eui O00l/I5>|
                    SNOISSIW3 XON

-------
     CO
     HI
     CO
  LU HI
  9>
  XLJ.
00
ccO
:DU-
   co
  coO
     CO
     CO
     111

            HI  \\t  in

            CO  CO  CO CO  CO
                       HO jo  gtu~oOC>r/6>l

                       SNOISSIIAI3 XOS

-------