ALKALINE SCRUBBING  OF IN-SITU OIL
      RETORT OFFGAS AT GEOKINETICS
                 HAL TABACK, P.E.
                CONSULTING ENGINEER
                      KVB, INC.
                 IRVINE, CALIFORNIA


                ROBERT GOLDSTICK
                     PRESIDENT
               ENERGY DESIGN SERVICE
                  OJAI, CALIFORNIA


                  EDWARD BATES
                 OIL SHALE MANAGER
        U.S. ENVIRONMENTAL PROTECTION  AGENCY
                  CINCINNATI, OHIO
                                             P-352
SHALE
                  PRESENTED TO THE

        18th ANNUAL OIL SHALE SYMPOSIUM
         /       AIRPORT HILTON HOTEL
              GRAND JUNCTION, COLORADO
                  APRIL 22-24, 1985

-------
                                    NOTICE

       This work was sponsored by the U.S. Environmental Protection Agency and
was performed under subcontract to MeteaIf fi Eddy, Inc., Boston,  i
Massachusetts, under EPA Contract No. 68-03-3166.  It has been subject to the
Agency's peer and administrative review, and it has been approved for
publication as an EPA document.  Mention of trade names or commercial products
does not constitute endorsement or recommendation for use.

-------
                                       ALKALINE SCRUBBING OF IN-SXTU OIL SHALE
                                            RETORT OFFGAS AT GEOKINETICS
                                                         By
           Hal Taback, P.E.
         Consulting Engineer
             KVB, Inc.
      Irvine, California  92714
                                      Robert Goldstick
                                     |    President
                                   Energy Design Service
                                     ;ojai,  California
                                                    Edward Bates
                                                  Oil  Shale Manager
                                        U.S. Environmental Protection Agency
                                                   Cincinnati, OH
ABSTRACT
    The EPA's «obile wet scrubber was need on a 200
ACFM slipstream of the Geokinetics retort offgas to
investigate the R^s removal efficiency and selectivity
(percent H^ removal/percent CO2 removal) as a function
of liquid/gas contact time, alkaline solution OH"
concentration, and the specific scrubbing chemical.  A
venturi and a tray tower were used to produce contact
times of approximately 0.003 and 0.2 second,
respectively.  Three alkaline solutions, HaOH, KOH, and
NH4OH were employed on each contactor at various
concentrations for a total of 22 runs.  To analyze
these results and provide design criteria for future
alkaline scrubbers a sophisticated computer model
employing the penetration theory for liquid-phase mass
transfer was developed.

INTRODUCTION
    Oil shale facilities proposed for Colorado and Utah
will produce substantial quantities of BjS and other
sulfur gases which could impact Class X airsheds such
as the Flattopn wilderness area.  The Clean Air Act
requires stringent control of such •missions through
the use of best available control technology under PSD
permits.  This report provides data characterizing in-
situ oil shale offgases from the Geokinetics (Seep
Ridge) plant in eautem Utah and assessing the
effectiveness at alkaline scrubbing processes in
                           -.
controlling the emission of BjB and other sulfur
compounds.  This renults should assist developers and
permit writers in selecting appropriate  controls  for
the treatment of oil shale offgases. '
    The offgas from the horizontal in-situ retort at
Geokinetics, in eastern Utah, contains approximately
0.15 percent (1500 ppnv) of H2s,  22 percent CO2,  ».nd
0.10 percent NH3 in addition to N2 (60 percent),  H2  (9
percent), CO (5 percent), CH4 (1.5 percent), and  other
(2.25 percent).  While these percentages are presented
on a dry basis, the offgas is actually saturated  with
moisture.  Also present, at levels of 0 to 10 ppmv
each, are organic sulfur species  such as carbonyl
sulfide, mercaptans, thiophenes,  and carbon
disulfide.  Lovell, et al. (1962) and Desai, et al.
(1983) evaluated various sulfur-control processes and
concluded that caustic (NaOH) scrubbing could be  a
candidate process if the selectivity of the scrubbing
process were sufficiently high.   Selectivity here is
defined as the percent removal of H,S (and the organic
                                   * \
sulfur) divided by the percent removal of the CO2.
Both HjS and CO2 are acid gases and CO2 i« present in a
concentration 150 times greater than that of H2s.
Therefore, any caustic or other alkaline scrubbing must
take advantage of relative solubilities and reaction
rates to achieve a high selectivity of HjS relative  to
co2.
    The full sulfur removal scheme for employing  an
alkaline scrubber is shown in Figure 1.  The retort
                                     i
offgas enters the alkaline scrubber and gives up  H2s
and CO2 to the scrubber liquid.  The scrubber liquid is
cycled through a regenerator or stripper where the H2s

-------
             flC

             u_
             CO
 CO

 o
  §N
   I f*. ^
           CM
           o
           o

           CO
           CM
           I
o
z
CO
CO
LU
O
O
DC
Q.
                    cc
                    o
                    CE
                    LJ
                    Z
                    111
                    O
                    LU
                    CC
                                                O
                                                •p
                                                0)
                                                cs

                                                3
                                                4J
                                                •H
                                                to
                                        O
                                        •H

                                        4J
           U
           U)

           10
           in
           0)
           o
           o
          o
                                        3
                                        14
CO
<
O
I
            as
            |S
            Si
                f
HI
                           2o
                           _i in
                             cc
0)

•l-l
rt
10
A;
              tSS
              Sgo

              DO"8
              5WCO
               UJU.

-------
•nd CO2 are distilled off and sent to a sulfur recovery
plant (such as Claus) where solid sulfur is produced
and the CO-, is released to the atmosphere.  In this
system an important parameter is the H2S concentration
in the feed gas to the Claus or other sulfur recovery
plants.  Dcsai, et al. (1983) indicated that the Claus
process night work with a feed of B percent H2s but at
least 15 percent was needed for confidence and 25
percent or higher was desirable.  »e sulfur recovery
plant inlet: gas H2S concentration is equal to the
product of the H2s/CO2 ratio in the retort offgas times
the selectivity factor "for the alkaline scrubber.  For
an oil shale offgas such as that from Geokinetics, the
•electivity required in the alkaline scrubbing process
can be calculated from the ratio of H2s to CO2 in the
retort offgas and the desired percentage of H2S in the
feed to the sulfur recovery plant.  So»e values are:
Sulfur Plant Inlet Gas
H2s Percentage Desired
          8>
          15
          25
 Scrubber
Selectivity
 Required
     12
     23
     38
Thus, a scrubber selectivity  of  over  40 is  desirable,
25  could  be  acceptable,  and  10 is  marginal  for retort
offgas similar  to  Geokinetics.
     Selectivity, while  important,  is  only one
performance  parameter determining  H^ removal.
Unfortunately,  many of  the factors influencing scrubber
performance  have conflicting effects.  Typical examples
•re that  increasing the hydroxyl ion  concentration
 (IOH~J) in the  scrubbing solution and increasing the
gas-to-liquid contact time may increase removal
•fficiency of H^  while decreasing selectivity of H2S
relative  to CO2.
     To investigate these effects the  EPA sponsored a
field testing program which  is reported in detail by
Tabaclc,  «t; al.  (1985).   This paper presents a summary
of those  results and conclusions.

 EXPERIMENTAL APPROACH
     The EPA's nobile wet scrubber was installed at the
 Geokinetics site to process a 220 ACFM slipstream of
 retort offgas.  The objectives of these tests were to
 measure H<>S removal .efficiency and selectivity as a
 function of  (1) liquid/gas contact time, (2)  scrubbing
 solution  [OB~], and (3) specific scrubbing chemical.
 The mobile scrubber was equipped with  both a  venturi
 and a tray tower  contactor which produced liquid/gas
 contact  times  of  approximately  0.003 sec and  0.2 sec
 respectively.  Alkaline solutions of NaOB, KOH, and
 KHfOH were  employed alternately on each of the
 contactors  at  various  concentrations as summarized in
 Table 1.  A run consisted of discharging the contents
of the 1-n3 mix tank once through the\contactor for a
period of approximately 40 minutes.  This simulates the
operation of the scrubber module in the system shown in
Figure 1.
    The retort offgas was sampled upsjtreaa and
downstream of the scrubber and analyzed for specific
sulfur compounds and total reduced sulfur.  The
sampling and the analytical procedure? that were used
for the specific reduced sulfur compounds are
essentially those specified in EPA Methods 15 and  16
(40 CFR 60, Appendix A, July  1, 1982);.  This method
employs a gas chromatograph  (GC) with a flame
photometric detector  (FPD).   In this procedure, a
continuous gas sample is extracted from the emission
source at a known rate, scrubbed to remove SO2 and
diluted with clean  dry air.   An aliquot of the diluted
sample is then analyzed for  the following sulfur
compounds: hydrogen sulfide  (H2S), carbonyl sulfide
 (COS), carbon disulfide  (CS2), methyl  mercaptan  (MeSH),
and thiophene.   In  addition,  continuous real-time
analyses of  total reduced  sulfur  (TRS)  in the  retort
offgas were  made by first  removing SO2« oxidizing  the
•ample gas stream  in a tube  furnace,  and  then reading
 the total sulfur as SO2 using a  TECO!continuous  SO2
monitor.  This  technique  was derived from EPA Method
 ISA (40  CFR  60,  Appendix  A,  July,  19$2)  and was  used to
 provide  a real-time display of removal efficiency
 during the  test runs.
                                     t
                                     i
 EXPERIMENTAL RESULTS
     The test results are summarized in Table 2 and
 Figure 2.  Three different solution concentrations were
 used for each alkali except for the last four runs (No.
 19-22) where only the tower was used to make two high
 concentration runs for both HaOH and KOH.
     As expected, it was found that the highest
 •electivity (percent removal of HjS divided by percent
 removal of CO2> »as obtained*at the lowest solution
 concentrations and at the shorter liquid/gas contact
 times (i.e., with the venturi contactor).  Conversely,
 the highest HjS removal efficiencies were obtained at
 the higher solution concentrations and the longer
 contact times (i.e., with the tray tower contactor).
 Figures 3 and 4 Bhow the variation of Bf removal
 efficiency with lOH").  A limit of 94 percent removal
 efficiency was  reached at an loir) of approximately  0.9
 gram moles/liter where the  selectivity is estimated  at
 approximately ten  (analysis  of spent  scrubber solution
 was not performed  on that test as indicated  in  Table
  2).  At the low IOH-)  of  O.012 grami«ole/liter  the
  nelectivity with  the venturi reached as  high as 79 with
  a removal efficiency of  just over 50 percent.
     The effect of  IOH"]  on selectivity for the  venturi
  and  the tower  is plotted  in Figure 5.  Mote that the
  venturi selectivity is more sensitive to tOH~]  than
  that of the tower.

-------
                      TABLE 1.   SCRUBBING TESTS CONDUCTED
Run No.
Alkali
   Cone.
gmoles/liter
                                                               Contactor
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Ammonia
it
H
n
M
•
NaOH
ft
M
*
•
H
KOH
M
tl
M
M
M
H
•
NaOH
•
2.0
If
0.05
H
0.3
• H
0.05
H
0.012
«
0.023
M
0.05
•
0.012
•
0.023
•
0.9
1.8
1.25
2.5
tower
venturi
tower
Iventuri
i tower
venturi
tower
Iventuri
tower
iventuri
tower
j venturi
tower
venturi
tower
'venturi
tower
[Venturi
tower
tower
tower
tower
i
Gas flow approximately 0.1 sm3/s  (220 ACFM)
Liquid flow approximately 0.38   £ /s  (6  gpm)                    i
Pressure differential across contactor:  Venturi:     34  nmHg  (18" H20)
                                         Tray tower:  23  mmHg  (12" H2O)

-------
                 TABLE 2.  SUMMARY OF ALKALI SCRUBBING RESULTS
Contactor
Venturi
Venturi
Venturi
Venturi
Venturi
Venturi
Venturi
Venturi
Venturi
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Tower
Alkali
NaOH
KOH
NaOH
KOH
NaOH
KOH
NH4OH
NH4OH
NH4OH
NaOH
KOH
NaOH
KOH
NaOH
KOH
NH4OH
NH4OH
NH4OH
KOH
NaOH
KOH
NaOH
OH~ Cone.
gmole/liter
0.012
0.012
0.023
0.023
0.045
0.046
0.049
0.29
2.0
0.012
0.012
0.023
0.023
0.045
0.046
0.049
0.29
2.0
0.89
1.25
1.79
2.5
Removal
Efficiency %
52
53
48
48
70
71
60
62
67
52
54
54
59
83
88
64
91
93
94
93
92
94
Measured
Selectivity*
79
71 !
60
51 :
[84]t
21 !
71
56 '
11 i
i
52 ;
43 i
41
49
36
41 ;
29
29
9
N/A'
N/A|
N/A
N/A
Run
No.
10
16
12
18
8
14
4
6
2
9
15
11
17
7
13
*
5
I
1 »
21
20
22
* Selectivity - A measure of the preferential removal of H2S over CO2 talcing
into account the relative difference in concentration between the two gases.
In this paper, selectivity is the ratio of percent removal of H_s to percent
removal of CO^*                                               i

t Data in brackets are suspected to be erroneous.             ;

N/A - Not available.  Selectivity values for these runs were not available
because the spent scrubbing solution was not analyzed.        ;

-------
100
                                  TRAY TOWER
                            VENTURI
                              N=  NaOH
                              K=  KOH
                                = NH4OH
                         x
                         g OPEN SYMBOLS: VENTURI
                         Q. SOLID SYMBOLS: TRAYi TOWER
                         Q.
                 SELECTIVITY
40            60
  % REMOVAL H2S
  % REMOVAL CO2
   Figure 2.  Removal Efficiency vs. Selectivity for Alkaline Scrubber

-------
                                        oc
                                        UJ

                                        o
                           O co
                           X Z

                           DO
                                              oc
                                        s
                                             O CO
                                             * Z
                                                    in
                                                    q
                                                    d
                                                    q
                                                    d
                                                    eo o
                                                    9 Ss
                                                    o -^
                                                       CO
                                                       1®
                                                       o
                                                    CM
                                                     q
                                                     d
                                                              g
                                                              cr>
                                                       o

                                                       o

                                                       v
o
o
o
o>
o
00
o
N.
o
(D
O
in
                                                       •H
                                                        u
                                                        3
                                                       4J
                                                        c
                                                        Q)



                                                       •0
                                                              o
                                                 u
                                                 c
                                                 o
                                                 •H

                                                 O
                                                 •H
                                                 14J
                                                 U-J
                                                        (0

                                                        O

                                                        1
0)
1-1


IT>
•H
&-,
       !N30d3d 'AON3IOIdd3 IVAOINSd '

-------
                  i	1—i—r
                                                  o
                                                   t

                                                  CO
-6
   D
                                           O)
                                            I
                                           X
                                    x  _  o

                                    S  S  5
                                    2  ^  Z


                                    OD<
O
O
      O
      O
o
oo
o
f-.
o
tt>
o
in
                                     o
O
TO
                                                  in
                                                   •

                                                  CM
                                                  O
                                                   •

                                                  CVJ
                                                     0)
                                                     CO

                                                     £
                                                   in O

                                                   ^ e
                                                     O)

                                                     i—K
                                                     I

                                                     ±

                                                     C>
                                                   in
                                                    •
                                                   o
                                                            w
                                                            (U
                                                            in
                                                            O
                                                             I


                                                             O
                                                            I


                                                            •D


                                                            10
                                                             o


                                                             (U


                                                             U
                                                             •H
                                                             UJ
                                                (0
                                                s
                                                s.
                                                            T

                                                             
-------
Discussion of Results
    All three of the alkaline solutions performed
•inilarly.  The plot of removal efficiency vs.
selectivity in Figure 2 indicates the specific chemical
at each data point.  All three solutions can produce
removal efficiencies above 90 percent at a selectivity
to be considered a candidate for use with Claus or
other sulfur recovery processes.  All three show high
selectivity at recovery efficiencies high enough that,
with the use of multiple venturi stages, a satisfactory
H2S removal efficiency (>95 percent) could be obtained
for the system.  Since the system envisioned for using
these chemicals (Figure 1) involves recycling the
alkali, the relative cost of the individual chemicals
is insignificant.  Hhat may be significant are factors
of corrosion, safety and availability as well as the
cost of stripping the gases from the alkaline solution.
    The test results indicate that the selectivity  for
the venturi is highly sensitive to the  [OH~] with a
rate of change, ds/d[OH~], of -1700 liter/gmole in  the
[OH~J range of 0.01 to 0.04 gmole/liter.  The tower
results! show a rate of change of only -300  liter/gmole
in the sane  [OH~] range.  This effect is believed due
to the presence of NH3 in the retort gas.   The short
residence tine in the venturi  (0.003 seconds) results
in a high dependence of selectivity on  [OH~] due to the
direct dependence of CO2 enhancement.   In other words,
the short residence time Beans that the CO2 has a
limited time to react.  However, as the H-,S absorption
is controlled by the gas film, its absorption rate  is
independent of  [OH~1 at values less than 0.03
gooles/liter.
    Figure 5 shows that, at concentrations greater than
0.03 gmoles/liter, the tower provides higher
selectivity than the venturi.  This is due to the      :
combined effect of the gas film coefficient and the    ;
presence of NH^.  The higher gas film diffusion
coefficient in the venturi essentially increases the
availability of CO2 at the scrubbing liquid
interface.  Consequently, the liquid-phase chemical
enhancement factor, which is a direct function of      ;
(OB~), has a substantial effect!'on the CO2 absorption  •
rate.  In the tower, the gas film coefficient is lower
which decreases the relative importance of the liquid
film, and, therefore, decreases, the dependence of the
CO2 absorption on  (OH~],  Since, the HjS removal is
determined solely by the gas film coefficient  (due to
the presence of NH3 in the gas), the sensitivity of H2?
absorption to IOH~] in both tower and venturi is       .
decreased.                     ,
                               i                    '   ;
    These results  indicate a clfcar choice of alterna-
tives in deciding between a tower or venturi scrubber
based on process requirements. j If selectivities
greater than 50 are needed, the venturi is required to;
take advantage of  the high selectivity at the short
contact time.  However, the venturi scrubber will only,
provide 50 to 60 percent remova'l efficiency per
stage.  If a selectivity less than 50 is acceptable,
the  tower is more  effective in that combined removal
efficiency and selectivity is greater than with  the    '
venturi.                       i
     A theoretical computer model was developed by Aiken
(1985) using the  penetration theory to correlate the
venturi test results  obtained in this program  at
                         TABLE  3.  COMPARISON OF THEORETICAL  AND EXPERIMENTAL SELECTIVITIES;
Selectivity (Unitless)* '
IOH-)
gaoles/liter
• 0.045
0.023

0.012


24 cat
25
43

66

Theoretical
20.5 cut
31
53

82


Average
28
48

74

Experi-
mental
21
55

75
Avg.
Difference,
percent
25
14.6

1«4
13.7
                    •Selectivity is the percent H^ removal/percent CO2 removal
                    tDistance along venturi where liquid is injected

-------
                                      1-1
                                      
                                      U
                                      0)
                                      0!
                                      W
                                      in

                                      OJ


                                      
                                      •H
U/SSHU= A1IAI10313S

-------
        %.i.ca witli tdie -tes-t. conditions And provide a
 basis for detailed design of future alkaline
 scrubbers.  One of the sensitive parameters in the
 analysis was the point of injection of the scrubber
 fluid into the gas stream.  The actual injection point
 is 20.5 cm from the beginning .of the venturi
 contactor.  However, the injected liquid requires a few
• centimeters to atomize.  Therefore, computer runs were
 •ade for distances of 20.5 and 24 cm.  The computed
 •electivities are shown in Table 3 compared to the test
 results,  test agreement is between the test results
 and the numerical average of the computed results at
 20.5 and 24 cm.
     The theoretical selectivities are in good agreement
 with the test results with respect to both trend and
 absolute values.  There is excellent agreement at the
 low concentration range (1.4 percent) while at the
 higher concentration (0.045 gmoles/liter) there is a 25
 percent deviation.  This information is also shown in
 Figure 5 as a range of predicted values for each
 concentration.  The agreement of the theoretical model
 with the test results; particularly at the lower
 concentrations (which are of primary interest when
 evaluating a venturi scrubber) indicates .that the model
 can be used for predictive studies of multi-stage
 performance.
     Some other interesting results derived from the
 modeling study, which included the chemical reaction of
 C02, H2S, and NH3 in the retort offgas with the
 alkaline solution in a venturi contactor are:

         1. The KH3 in the retort offgas reacts with the
            H2S at the gas/liquid interface.  Since the
            HHj present in the offgas is at
            approximately the same concentration as the
            H2s , the removal efficiency is only
            marginally dependent upon -*he IOH~].

         2. The removal efficiencies for HH3 and H2S are
            similar.

         3. The HjS selectivity is significantly
            affected by contact time with a maximum
            •electivity of 110 occurring at
            approximately 0.0015 sec contact time.

         4. Variations in temperature and liquid droplet
            size can have a significant effect on
            •electivity.

 Scrubber Concept Designs
     These experimental and theoretical results suggest
 two alternative alkaline scrubber design concepts for
 futuro  consideration.  One system combines the high
 •electivity  of''the  venturi with the high removal
 efficiency of the tower.  The other uses a tray tower
 for miiximum  H2S removal and isolated liquid scrubbing
•Cages to maximize the selectivity for use with a
•ulfur recovery plant.         |
    For the first concept, the 'first stage is a venturi
designed for peak selectivity based on contact time and
[OR~].  The theoretical model indicates that a maximum
•electivity of 110 will result in a 50 percent removal;
efficiency.  The CO2 removal efficiency is 0.40
percent.
    The second stage is a tray tower designed! for 90
percent H2S removal efficiency Iwith a stage aielectivity
of 40 using an [OH~] of 0.045 gmoles/liter.  These
                               \
values are based on the experimental results obtained
in this program as shown in Figure 2.  The CC>2 removal
efficiency is 2.2 percent.                            ;
    The overall performance for this design is a 95
percent H^ removal efficiency ;with a selectivity of
37.  These values are calculated as follows:
"2s
CO2 cone.


ppm
*
Inlet
(Assumed)
1500
22
; Stage 1
Venturi Exit
! 750
! 21.9
Stage 2
flower Exit
75
21.4
 H2S removal eff. -  L^

                     22-21.4
          CO2 removal  eff

             Selectivity
                       22
                                         95*
                                         2.7%
                        #-»
    The second concept  is essentially two towers in
series; but,  in  fact, the two tower stages could be
combined  into one partitioned vessel.  Each stage  has
90 percent B2s removal  efficiency and a selectivity of
40.  The  overall performance  of  the scrubber is 99   :
percent removal  efficiency  with  a selectivty of 22.  ;
These  values  are calculated as,follows:
                               !

                    Inlet      i
                  (Assumed)  Stage 1 Exit  Scrubber Exit
 H2S  cone.
  °2
            ppm
                    1500 ,
H2s removal eff.    -

CO2 removal eff.   » ,

  Selectivity      - i
150
 21.5


1500-15
  1500

 ~22
                                      15
                                      21.0
                                           99%
                                          4.5%
 CONCLUSIONS
     The findings reported above lead to the following;
 conclusions:                   |                       ;

-------
t. For shale ell retort offgas similar In
   composition to that fro* the Geokinetics
   process, tho alkaline scrubber, in
   combination with • stripper and • Cl»u»
   plant, could be a technically viable Beans
   of Bf removal.  A. design/cost study for  •
   such a system would determine its economic
   viability.

2. Per eeokinetics-type process offgas and
   based on these tests, the performance of an
   alkaline scrubber with a tray tower
   contactor similar to that in the EPA pilot
   plant can achieve an H2S removal efficiency
   of at least 90 percent with a selectivity of
   approximately 30.  Under the sane conditions
   a single venturi contactor in place of the
   tray tower would remove only 50 to 60
   percent HjS but with a selectivity of 70 to
   80.

3. Bused on the computer model developed to
   analyze these test results, the removal
   •ffficiencies and selectivity above are
   applicable to offgas with  lower H2S
   concentrations than  found  at Geokinetics.
   This  suggests a concept of multiple
   scrubbing stages to  increase the  H2s
   nimoval.  ••cause this increased  removal
   efficiency is  accompanied  by a reduced
   selectivity  (which could present  a  problem
   by lowering  the H2S  concentration in the
   f*ed  gas to  the sulfur  recovery plant),  the
   cost  effectiveness  of this concept requires
   • design study.

 4. Based on a theoretical analysis of the
    three-gas  component system (B2S,  HH3,  and
   CO2), the principal reactant for the H2S in
    the retort offgas  is the KH3 in that same
    offgas.  In that HH3 is present in the
    Geokinetiea offgas in similar mole
    qinantities to that of the HjS , the scrubber
    performance observed on these tests may not
    tie applicable to. retort offgas with little
    <>r no HH3«  this suggests that the water and
    1die HH3 in the'off gas would be an effective
    Dcmbbing ageat without any alkali additive
    to the watery  Scrubbing in this manner
    irould certainly improve the selectivity, but
    1th* effect oa removal efficiency obtainable
    :ls uncertain.
          offgas itself  that is reacting the H2s  .
          Since the HH3  and H^ concentration* are
          variable, it is  likely  that  some of the H2s
          is  reacted  by  the alkali;  Bierefore, !lt is
          likely some alkali will always be needed.
          However, the choice  of  scrubbing alkali may
          be  made  on  such  factors as co»t,maintenance,
          safety,  availability, and  crew comfort.

        6. The absorption of HjS   and CO2 in  the
          alkaline solution occurs;in reactions which
          should be  fully  reversible by
          distillation.  The sulfur in the scrubber
          solution is primarily in the fora  of
          sulfide.  The  sulfate  or sulfite  level
          determined in the scrubbing solution was
          equal to that in the water supply.   The
           •ulfide will distill off as H2s (along  with
           CO2) while the sulfate and sulfite will not.
RETEREHCgS
           Aiken, R. C., "Selective Scrubbing of H2s
           from C02 in Shale Oil Retort Offgas Bused on
           the Penetration Theory,"! Appendix B to EPA
           Report cited below, Taback, et al., 1985.
                                   !

           Oesai, B. O., Day, D. R., and Peters, 3. A.,
           "Air Pollution Investigations of Oil lihale
           Retorting: In-Situ and Surface, Task  li
           Evaluation of Sulfur Removal Technologies,"
           (Draft Report) EPA Contract Ho. 68-03-2784,
           February 1983.

           fcovell, R. 3., Dylewaki, S. ¥., and
           fetarson, C. &.,  •Control of Sulfur
           Emissions from Oil Shale Retorts," EPA
           Report 600/7-82-016, HTIS PB82-231945,  April
           19S2.

           Taback, B. J.,  puartucy, C.  C., and
           Coldstick, R. 3., "Alkaline  and Stratford
           Scrubbing Tests for  HjS Removal from In-Situ
           Oil Shale Retort Offgas,-  KVB Report Ho.
           •07430-1982,  EPA Contract 68-03-3166, EPA
           Air mad  Energy Engineering Research
           laboratory,  Research Triangle Bark,  SC,
           February 1985.          ,
  5.  'Die alkaline •ember removal «fficiency and
     aeleetivity aeemed to have  little  dependence
     en the alkali used.  This is consistent with
     the above concept that it is the HH3 in the

-------