P-351
PROCESSING IN-SITU OIL SHALE RETORT
OFFGAS WITH A STRETFORD PLANT
AT GEOKINETICS
HAL TABACK, P.E.
CONSULTING ENGINEER
i
AND
GREG QUARTUCY, P.E.
SENIOR ENGINEER
KVB. INC.
IRVINE, CALIFORNIA ;
EDWARD BATES
OIL SHALE MANAGER \
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO
PRESENTED TO THE
13th ANNUAL OIL SHALE SYMPOSIUM
AIRPORT HILTON HOTEL
GRAND JUNCTION, COLORADO
APRIL 22-24, 1985
-------
NOTICE ;
i
This work was sponsored by the U.S. Environmental Protection Agency and
was performed under subcontract to Metcalf & Eddy, Inc., Boston, |
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. i
-------
PROCESSING IN-SITU OIL SHALE RETORT OFFGAS WITH A
STRETFORD PLANT AT GEOKINETICS
By
Hal Taback, P.E. and Greg Quartucy, P.E.
Consulting Engineer Senior Engineer
KVB, Inc.
Irvine, California
Edward Bates
Oil Shale Manager
D.S. Environmental Protection Agency
Cincinnati, Ohio
ABSTRACT
The EPA transportable Stretford process pilot plant
was used on a 700 ACFM slipstream of in-situ shale oil
retort offgas to investigate HjS removal efficiency and
process compatibility. This was the fourth application
of the pilot plant which had demonstrated a continu-
ously improving performance. During this test the
pilot plant was operated first with a venturi contactor
•lone and then with the venturi followed in series by a
packed tower contactor. With the venturi alone, the
plant achieved 95 percent removal efficiency, but its
average performance was lower. With the addition of
the packed tower, the removal efficiency reached 99+
percent. Excessive foaming of the process solution was
experienced which was attributed to the presence of
organic aerosol in the retort offgas.
INTRODUCTION
Oil shale facilities proposed for Colorado and Utah
will produce substantial quantities of R2S and other
sulfur ganes which could impact Class I airsheds such
a» the Fliittops wilderness area, the Clean Air Act
requires ntringent control of such emissions through
the use oi! best available control technology under PSD
permits. This report provides data characterizing
in-situ oil shale offgases free the Ceokinetics (Seep
Ridge) plitnt in eastern Utah and assessing the
effectiveness of the Stretford process in controlling
the emission of HjS and other sulfur compounds. The
results 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
Ceokinetics, in eastern Dtah, contains approximately
0.15 percent (1500 ppmv) HjS, 22 percent CO2, and 0.10
percent NH3 in addition to N2 (60 percent), H2 (9
percent), CO (5 percent), CH, (1.5 percent), and other
*• i
(2.25 percent). While these percentages are presented
on a dry basis, the offgas is actually saturated with
•oisture. Also present, at levels of 0 to 10 ppmv
each, are organic carbon species such as carbonyl
sulfide, sercaptans, thiophenes, and carbon
disulfide. Lovell, et al. (1982) and Desai, et al.
(1983) evaluated various processes and concluded that
the Stretford process was a viable candidate for this
application. Therefore, a series of feasibility tests
were sponsored by the EPA in which' a Stretford pilot
plant was used to process a 700 ACFM slipstream of
retort offgas. This paper presents the results of the
fourth and latest test of that pilot plant. A
i
comprehensive test report has been published as
•Alkaline and Stretford Scrubbing 'Tests for HjS Removal
from In-Situ Oil Shale Retort Offgas,* EPA Contract 68-
03-3166 (Taback, «t al., 1985) !
Previously the Stretford pilot plant had been used
at Occidental Oil Shale, Znc.'s (OXY) Logan Hash oil
•hale development mine near De Beque, CO (July 1982),
at the Geokinetics* in-situ retort in eastern Dtah
(October 1982), and on a U.S. Bureau of Mines coal
gasifier in Minneapolis, UN. The HjS removal
-------
efficiency was higher on each of these tests than on
the previous one. On the coal gasifier test 99+
percent H.jS removal was achieved.
The objectives of this fourth test were to (1)
duplicate on retort offgas the 99+ percent HjS removal
efficiency achieved on the gasifier tests and (on
achieving that), (2) attempt to explain the low removal
•fficiency during the 1982 test by deliberately
introducing upsetting changes to the plant chemistry
and then returning it to the 99+ performance.
The Stretford process is regenerative and converts
H,S in this retort offgas to elemental sulfur. It uses
air oxidation to regenerate the chemicals reduced
during th« offgas treatment. It is a highly selective
process in that it removes H2S with negligible side
reactions with CO2. CO2, like H2S , is an acid
anhydride and is present in the offgas at much higher
concentrations than the HjS . Therefore, most
scrubbing processes must be concerned with selectively
scrubbing H^ in preference to CO2« Th« Stretford
process is highly selective for HjS. It has been in
use for more than 25 years with nearly 100 plants in
operation worldwide on such processes as:
. Coal gasification
. Coke oven gas
.. Refinery fuel gas
„ SNG (petroleum) plant gas
. Natural and associated gases
„ Claus tail gas
. Geothermal power generation
„ Carbon disulfide manufacture
» Ore roasting
„ Sewage sludge digester gas
Process Description
The process chemistry of the Stretford technology
is based on the absorption of H2S in an alkaline
scrubbing solution and subsequent liquid-phase oxida-
tion of 1i>e captured BjS to elemental sulfur. The
Stretford liquor is a dilute solution of sodium
carbonate (Na2CO3), sodium metavanadate (NaVOj), and
sodium salts of the 2:6 and 2:7 isomers of anthra-
quinone (lisulfonic acid (ADA). The solution is
maintained at a pH of 6.5 to 9.5 and a temperature of
approximately 43*C.
Removing the HjS from the gas stream and converting
it to elamental sulfur'is basically a six-step process,
•s follows (with simplified chemical reactions):
/
1. 'the HjS is absorbed in the alkaline Stretford
solution in a suitable gas/liquid contactor.
2. The H^ reacts with the sodium carbonate to
form sodium hydrosulfide and sodium
bicarbonate:
H2S + Na2CO3 + NaHS + NaHCO3 i
3. The hydrosulfide then reacts-with sodium
metavanadate to form elenental sulfur, a
quadravalent vanadium salt, and sodium
hydroxide:
2ttaHS + 4NaV03 + H2O + tJo2v4°9 + 2S + 4NaOH (2)
4. The quadravalent vanadium salt then reacts with
ADA to regenerate the sodium! metavanadate:
Na2V4O9 + 2NaOH + H20 + 2ADA +
4NaV03 + 2ADA-2H
(3)
5. The sodium hydroxide and sodium bicarbonate
i
reaction products further react to form sodium
carbonate: :
HaOH + NaHC0
(4)
6. The reduced ADA reacts with oxygen to
regenerate the ADA: ;
2ADA-2H + O2 + 2ADA + 2H2O
(5)
The overall process reaction can be written as the
oxidation of HjS to elemental sulfur:
2H2S
2S
(6)
Several side reactions that form nonregenerable
compounds, primarily oxidized sulfur compounds such as
sodium sulfate and sodium thiosulfate, are possible in
the Stretford process. These nonregenerable compounds
can build up in the system and eventually impede the
performance of the process by interferring with the
principal chemical reactions. These compounds must be
removed from the process either by purging them from
the system or by recovering them in.a regeneration
system.
The oxidized sulfur compounds form when the
dissolved oxygen in the process liquor is too high,
which occurs when the pH is too low. R2s absorption is
also reduced at low pH. The high Cp2 found in retort
offgas can reduce the pH unless tJ» process pH level is
maintained by sodium bicarbonate addition.
Figure 1 presents a simplified process flow diagram
of the pilot plant. This diagram depicts the basic
design configuration of the plant, 'including a variable
throat Tenturi scrubber gas/liquor ^contactor, reaction
tank, oxidizer, pump tank, and slurry tank. At
Geokinetics, before the retort offgas stream entered
!
the Stretford plant, it was conditioned in vacuum
blower and mist elimination equipment to remove
residual product oil and water. Thus, an attempt was
made to preelean the gas stream prior to its entering
-------
c
ID
ft,
A.
T3
O
4J
0)
U-,
O
10
(0
•H
n
•O
0)
•H
a
E
2
3
CT
-------
the Stiretford. B»e Stretford also had a knockout tank
to collect oil and water which further condensed on the
wall* of the inlot pipeline.
Tine retort offgas stream first enter* a venturi
scrubber, where the gas cones in contact with the
Stretford solution, the solution i» delivered to the
top of the venturi through a single feed line with a
•pray nozzle.
This plant was modified during the t«»t to add a 30-
esi-dianeter packed tower scrubber containing 2.5-cm-
diametor Raschig rings. The packed tower w«« located
downstream of the venturi. It provided a much longer
gas/liquid contact -time. The venturi solution was
•prayed into the top of the tower and drained into the
reaction tank. Figure 2 shows the installation of the
venturi and packed tower.
The reduced process liquor flows from the reactor
to the oxidizer (Figure 1). The function of the
oxidizer is to reoxidize the Stretford liquor
(replenish the reduced ADA), separate the sulfur
product froa the liquor by air flotation, strip
bicarbonate formed in the process fro* the liquor (as
carbon dioxide), and strip any ammonia absorbed fro*
the gai stream. The stripped carbon dioxide and
anmonia are removed from the process via an atmospheric
Timt stack in the oxidicer. Oxidation air is
introduced into the base of the oxidation tank through
a dispersion ring. The air is further dispersed into
the liquid by a mixer. The sulfur product is generated
aii a froth at the top of the oxidizer. This froth
contains approximately seven percent (by weight)
sulfur. The froth overflows a slurry weir into the
•lurry tank.
RESULTS
The Stretford pilot plant was operated for 200
hours. Three sampling and analysis system* were used
to characterize the retort offgas both upstream and
downstream of the plant. The first system measured
rttduced sulfur compounds according to SPA Nsthcids 15
and 16 (40 CFR 60, Appendix A, July 1, 1962) which
•wploys a gas chrcaatograph (GC) with • flame
photonetric detector (FPD). This system analyzed the
gas for hydrogen sulfide (H^B), carbonyl sulfide (COS),
carbon disulfide (CS2), Methyl mercaptan (MeSB), and
thiophenes. The OC/FFO was a Perkin-Zlmer Model 990,
•quipped with a ten-port valve for automatic Injection
of tho sample from the-sample loop. For backflushing,
this unit has a precolumn that traps high-molecular-
weighi sulfur and hydrocarbon compounds. A continuous
offgan sample is taken, bubbled through a cold
scrubbing solution' (citrate buffer), to remove SO2 and
dilutitd with damn, dry air. An aliquot of the diluted
gas was then analyzed by the OC/FPD.
Tine second system was a continuous real-time
sieasure of total reduced sulfur (TRS) in the retort
eiffgai. The sample stream was passed through the cold
scrubbing solution to remove CO2, diluted with air,
oxidized in a tube furnace, and then passed through a
TBCO continuous SO2 monitor. This; method was derived
from EPA Method 15A (40 CFR 60, Appendix A, July 1,
1982). The readout was calibrated in units of HyS
since there was little organic sulfur present in the
offgas.
The third system measured non-sulfur gases uiiing a
Baseline Industries, Inc. Model 1030-A GC with a
thermal conductivity detector.
Table 1 summarizes the operating conditions imd the
plant's H2S removal efficiency recorded throughout the
test. The high water and oil vapor content of the
retort of fgas caused the lines to both sampling iiystems
to clog frequently, limiting the number of removal
efficiency measurements that could be made. The
removal efficiency in Table 1 was 'computed from data
obtained from the TBS measurement*. For the finst
•even days the Ef removal efficiency averaged 80
percent and on May 6 and B reached a high, approaching
95 percent. When the packed tower was added on May 12
the removal efficiency stepped upito 99+ percent which
was sustained for ten hours. The 'flow to the tower was
discontinued to verify that it was the use of thie
packed tower which produced the 99+ efficiency. This
was confirmed when the efficiency dropped to 86
percent. Two days later the scrubbing solution was
•gain supplied to the tower, returning the unit to a
higher efficiency which had reached 99 percent when the
test was terminated due to a Geokinetics plant
shutdown. j
The daily reduced sulfur specie* data are
summarized in Table 2. These data were measured using
the OC/FPD technique. This table!shows good agreement
between the GC/FPD values for HjS| concentration and the
TOS values for BjS in Table 1. The B2S removal
efficiencies computed by the two methods also agree
well. Table 2 also shows the small concentration of
organic sulfur species in the offgas. The scatter in
the organic species data illustrates the difficulty in
Measuring these species in such small concentrations.
Zt is believed that there was no significant change in
concentration between inlet and outlet for these
species. !
Discussion of Results |
The original objective of achieving and maintaining
• 99+ percent HjS removal efficiency on retort offgas
was not achieved with the original plant configuration
and operating procedures. Attempts to adjust operating
parameters (e.g., decreasing the venturi throat area
and increasing the residence time of the solution in
the reaction vessel) did not produce any significant
change in H^ removal. Nevertheless, when the «»«-to-
liquid contact time was increased by adding the packed
tower, the 99+ percent efficiency was obtained imd
Maintained for ten hours. '
-------
Packed
Tower
Pump
Tank
Venturi
Contactor
Reaction
Vessel
Figure 2. Overall View of Stretford Plant Installed at GKI.
-------
en
B
CS
| I]
£4 O
i g
q c:
6
oooenenen o> en en CD r- p> oen P-P-P-P-P-CO ooencoooen oteioiencnencncncn
i
CMCMCMCMCMCM 'cMCMCMCMenen inin to to to vo to to vo P» P» r» r» p^cocoaocbaoaDODCD
O O O O O O O O O O O O O O O O O O O O O O O O O OOOOOOOOO
CD ^ oo en 01 in «- p» en o* in o
vo vo in in » T? « en en CM CM CM
o en
vo m
ocM««mto p- •- ^ en ID in»-o\o9ineno»p-»
oooooo oooenoi ^-oenoioioicoooco
en CM CM in »»
CM CM CM CM CM CM
»» en en en CM
CM CM CM CM CM CM
en CM CM CM CM
en en en en en
1 » « » en en en en
en en en en en en en
in in in p- r* r*
P. P. P. eo 09 co
p- vo vo vo tn m
o m
o en
CM o p- vo in «
oo oo r- r- p- p>
en vo 10 vo VD
p- « >» ^ «
inGnenGOP»p*tomtn
incnoicncnencncnoi
in en »» en en
p- P- r* r* p- p-
oooooo
oooooo
O O O 0 O 0
in «
p- P-
enenen*1* »rp-p-r»p-p-p-oop-
P-P-P-P-P- p-p-r-p-pfp-p-r-p-
O\ CO
ss
o «r » vcnoocoeo Omwvvoencnen
i .enenen enenencncn encMCMCMCMCM*-^*-
! • ^ o ^ o « e\ r* n OOOOOOOOO
§en «r i
•w «r '
* o o o o o o o
t» ^ vo vo p. r»
o «r »
en en en
o o o
o o o
o o o o o
o o o o o
OOOOOOOOO
vO»» inenenenen enencnoieMeMCMen
enenen en«««« enenenen»»««
<*i " 2 2 * — r-p-vovop-co IOP- enl lenenoi enenenenen en^ — o»-»-CMen'»
**™^^^^ ^^*^^f^w w^1 ^*l lenenen n *f f ^ ^ ^^^f^^^f^v^
en«-oir>to« ocoeo<-coo p-» o *r ^£cMvOp-vo Svenino Ototn«
o^enocnv ^oavmtop*
CM I IP-P-P* lOCMCMCMCM encDODP-OOO»-CM
VI lenenen m TT -r ^ «r encncnen«v«««
otnentocnoi CM CM p. » m en vo vo o enenen en co co in in en
cntnentntoto p- en CM in co 10 p- o co . . CD cc « <*o»-^o .00
r-CMCMCMCMCVI CM«*enCMv-^ CMCM CMJ j^^^ «VCMCMCMCM |«»
_«OIVOOICOCMCD
•-CM^CMCM^CM^-
OOOOOO OOOOOO OO
CM e- p- «- p. i
ooo ooooo
pooooooo
Ttoeo — p- o»p-09en«^»-eneno
in^encMv^- CMcnoenven eno v^vo*-OO ^-incMvocM CMOOO^«-O^^
r> p>'p> vo in vo p>p>onoienen vo r- eooip-ooooco oo co oo oo co 09090000000303000
Svn o m m o o o o o in o ^ CM m o m r- CM in moocMo Ovoooe^Oincotn —
tnen»enen encnenencMen CNO CMOO^-*--* « — en o CM en»-»CMen^cM»-eM
cocoovCMcn Q^CMcnivin «»CM eoocMCMent*) ovovocoo OP-P^coeno^cMen
.- — CM CM CM W OOOOOO CMCM — CM CM CM CM CM O ^ S- ^ CM St-V-iF-^CMCMCMCM
m
2?
-------
2
*
M
1-
4
|
•O
X
o
7"
&
fl
en
•g
*>
e
*
I
0 «
*4 M
JJ JJ j,
£
.0".
§55
"Sg^-
365
d§>
3ft
" S
o b •>
gdS
S^
iJg u
£ o y
X K
O *> ^v
tu B eo
SS 5
>!i
* o a
iJ S S1
S a 1
M a g
sf"
* • "
O *jri
S .M-
IB*
£
«s«
3§|
U a
»J
5 09
32
^ *- r* m o * o
T T 7 7 i *
co o «• en O O «-*-«•
CO r* CO 00 CD CD CO
oooomom
o ^ o« «n t-
.»>,
m« in « r^
CM
•V.
in
CM ^- o en en on
o e o o o o
en en en en CD CD
CM CM CM CM CM CM
r- i- r» vo in to
CM m en m m m
o o o o o o
O 0 0 O O 0
10 vo r- oo en o
* •. CM
iSSSSB
CD c r- en
en o o o o o
— CO f» 10 0 O
CM CM CM CM m en
10 r- r- r- o
....I.
O 0 O 0 O 0
o o o o o o
CD 00 f* CD CD CO
S8BR88
05.305"
o — «-
* * z o* o o
CO CO CO CO CD CO
in* « t^ eo cK o
vo
tn in
b o
CM CM
O O
en en
i
So
0
SS
i : :
i I !
1 i i !
o en , |
m CM ' '
°.°. , ,
• vo r» r*
m m m fn
O 0 O O
O O O O
Son en en
m m in
O CM CM CM
CM CM CM CM
<- CM O <*
vo ic m ^
in vo *- «-
r?2§
CD o in in
CM CM t- t*
m in m I
O CM CM CM
CM ^ ^ ^
O 0 0 O
T « r- a>
So o in
o m «-
SSSo
s
4J
•H
5
e?
B
S
1
1
o
D
C
g
s I
*O U
» 5
j f
3 3
is s,
US |
-------
TABLE 2. REDUCED SULFUR SPECIES EMITTED (ppm)1
Hos Concentration ''
Inlet
Date
1984
5/5
5/6
5/8
5/9
5/10
5/11
5/12
5/13
5/14
Avg
1584
1719
1377
1638
1314
1144
1141
981
1121
Range
1322-1730
776-2165
1367-1898
1398-1935
1245-1761
1015-1253
953-1249
718-1125
1091-1137
Avg
447
261
244
278
248
228
10
131
92
Outlet
Range
385-693
16-559
75-395
188-343
235-301
190-240
6-15
7-140
14-138
Remvl
Eff.
Avg
72
85
82
83
81
80
99
87
92
COS
Concen.
Inlet
45
190
N.D.
35
N.D.
26
99
88
94
Outlet
59
36
54
82
53 I
52
72
76
79 i
i
i
MeSH
Concen.
Inlet
N.D.*
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
N.D.
Outlet
N.D. •
7
5
4
N.D.
N.D.
14
18
18
* N.D. = none detected, minimum detectable level.
** GC/FPD measurements
-------
whether or not a packed tower la the most desirable
contactor for the stretford process was not resolved on
this test. Other types of contactors (e.g., a tray
tower, multiple venturi stages) nay also be
•infective, The packed tower could eventually become a
maintenance problem if sulfur or particulate matter in
the retort offgas were to collect. This test was
insufficient to investigate this aspect of the packed
tower performance.
It is, however, reasonable to conclude that if a
Tenturi contactor fails to provide the desired H2S
removal efficiency-on a Stretford plant, then changing
the contactor to increase contact time should increase
the removal efficiency.
During the test, one episode (on May 10) of
excessive foaming was experienced in the slurry tank
(Figure 1). An antifoaming agent was added at frequent
intervals, the oxidized air flow was reduced, and the
solution flow from the slurry tank to the oxidizer tank
was increased to prevent the slurry tank from
overflowing. Eventually the foaming was brought under
control when the apparent cause was discovered. This
upset in the solution had no noticeable effect on the
H2S renoval as can be seen in Tables 1 and 2.
Thn foam was attributed to excessive amounts of
•hale-oil-laden eondensate water carried into the
process by the retort offgas. A liquid knockout tank
van provided upstream of the system as mentioned
above. When the foaming began, it was discovered that
the knockout tank was filled and the overflow was being
carried! into the process. When the tank was drained
and the above antifoaming measures were taken, the
process was restored to normal.
REFERENCES '
Desai, B. O., Day, D. R., and Peters, J. A.,
"Air Pollution Investigations of Oil Shale
Retorting: In-Situ and Surface, Task Is :
Evaluation of Sulfur Removal Technologies,"
(Draft Report) EPA Contract No. 68-03-2784,
February 1983.
Lovell, R. J., Dylewaki, S. H., and Peterson,
C. A., "Control of Sulfur Emissions from oil
Shale Retorts," EPA Report 600/7-82-016, NTIS
PB82-231945, April 1982.
r
Taback, H. J., Ouartucy, G. C., and Goldstick,
R. J., "Alkaline and Stretford Scrubbing Tests
for H2S Removal from In-Situ Oil Shal« Retort
Offgas," KVB Report No.1 807430-1982, EPA
Contract 68-03-3166, EPA Air and Biergy :
Engineering Research Laboratory, Research
Triangle Park, NC, February 1985.
CONCLUSIONS
Based on this test the following conclusions were
reached:
1. With-an adequate contactor, the Stretford
process can obtain removal efficiencies of 99
percent or higher on retort offgas. If
adequate H2S removal cannot be achieved with a
venturi, then a packed tower or other method
for increasing contact time should be
,considered for improving performance.
2. To ensure satisfactory performance of a
Stretford plant in processing retort offgas, it
is important to provide effective removal of
hydrocarbon/ mist and other particulate matter
from the gas before it enters the plant.
------- |