-------
The emissions of H2S, SOg , and sulfur vapor from Glaus plants
are directly dependent ®m the -efficiency of sulfur recovery in the
Claus olant. For example, a TO LT/0 Claus plant operating at 95 percent
sulfur recovery efficiency would eimit 5 LT/B sulfur in the flora of H2S,
SO,,, -sulfur vapor ;, COS arad £$2 formed in the reaction furnace. Gas
iL.
stream compositions throughout a Claus plant are qiven ITU Table 3-1 for
a hypothetical 100 LT/D plant operating at :% percent recovery efficiency.7
Claus ®lant efficiencies in tuna are tjependent upon the-Jbllowing
:{T) dumber of catalytic cowersion stages,
i(2) Met feed stream composition.
(3') ©Derating temperatures and catalyst maintenance,
i(4'3 isiaimtalniira?! the proper stoidiioraetric ratio of fi
(5.) 'Bperatinq caparf-fey fact'or.
for 'Clau-s roll ante fisd Witt f90 -mol-e percent H2S, the sulfur reooverv
is approximately 85 parcent for om catalytic stage .and 95 percent for
two or three stages. The percentage of sulfur trecwery ails© increases
with increasing concentration of the acid gas fed to ttie Cilaus
plant. For plants having two -or flw^ee catalytic stages,, the
sulfur recoveries for warious «acM «oias c0iiicentrati©ns are appro'xii'mately
90 percent for a 15 mole ipercent iHgS feed stream, 93 penoerat for a
50 mole percent H2S stream, and 95 percent for 90 percent H2S concani-
trati on .
Contaminants in the feed -gas reduce Claus sulfw recovery efficiency.
Hydrocarbons in Claus feedstocks require extra aw for combustion- Tfee
3.5
-------
t£
»— * *•••••»
> •»->
< c
uj co
_, £
iiElL
«c«c
_JLO
CD a. en
H2T
t-< CO
35 CO v-
UJ C
>- 3
oo o:
2" LU to
_:f* H^ Xi
,
— J S-
U_ CO
CO >
C3 ^3
rr r— o
o co
e«! *-
i— tc_> s-
CO t-« 3
Ol— M-
0-UJ r—
S*~ — 3
1— to
O O
a_
_r>- ..
Sz: to
•r-
t-< to
O- n3
>- CO
to
ro
C£3
f-
£
to
3
ro
O
r
(
*
C
O
•r~
(O
S-
co
•r~
O
C
1— 1
£_
CO
M-
E
O
-p
ro
s_
CO
E
•r-
O
f^
H-I
s_
o
CO
CO
CO
3 n3
(O 4J
— C
!S. !i!-!^!!SSSS
0 CM 0 10 CD C
CM r-- CD
CnCTlLOLO-^fCMlO C5
LOCMOOOCMCO l§gg 10
o o o o o C3 o ' t— CD LO "CD
CM r--. o
i —
o
01 IO LO O CD
• III II ••! «li
r^t .- ••! •! 1 •
JJJ «d- O 1 LO 1 ICD
00 0
^™
LO
S E * S
CM 4-> CD
r— cC
cn
LO i—
^j* co co I^N
CO r— CO
CM CM
**
iO
_
ro
•
si- i — i — CO
O CM CM
i — O
CM
(O cu s:
•r- r— U_
U- CO O O
0 Q. S= OO
CO
OJ
3
05
to
o
a.
o
Ci
S-
o
Q.
ro
OO CM 00
CM O CO O
n; oo oo cj
CM
OO CD
O 0
o
ro
CM
O O
O ~
o
O
CM OJ
• -
CM CM
CO
s-
+J
rO
S-
co
a.
E
O)
(—
OJ
i-
3
co
co
a>
S-
Q.
-P
sz
3
O
^ "^
•i- CO
C -r-
3 ro
CTi—
01
co Q;
ro
o
-a
c
fO
S-
o
co
c
o
CO
to
CO
O
X
co
co
o
S-
cu
Q.
co
3
T3
CO
CO
fO
CQ
3.6
-------
added water and inert gas associated with burning hydrocarbons increases
the size of the sulfur plant equipment and lowers sulfur recovery since
the sulfur gas concentrations are decreased.U Higher-molecular-weight
hydrocarbons in the feed also reduce Glaus efficiencies because of soot
formation on the catalyst.13 High feedstock concentrations of C02
14
adversely affect catalyst Iffe.
Since the reactions in a Claus Plant are exothermic, sulfur
recovery is enhanced by removing heat, hence operation at as low a
temperature as practicable in the reactors without condensing sulfur
vapor on the Claus catalyst is necessary. Sulfur recovery is also
dependent upon catalyst performance. One vendor has reported a one
to two percent loss in recovery efficiency over the period of catalyst
life.15 Catalyst life generally varies from 2 to 5 years depending
on plant operation and contaminants in the feedstock discussed
previously.
Deviation above or below the 2:1 stoichiometric ratio of
H2S and S02 for the Claus reaction results in a loss of Claus
efficiency. Figure 3-2 illustrates the variation of recovery
efficiency with H2S/S02 concentration ratio in the Claus converters.
Operation of a Claus plant below capacity may impair Claus efficiency
somewhat. One vendor has reported a two or three percent loss in recovery
when the Claus plant was turned down to 20 percent of design capacity.
Another vendor reports no loss in recovery efficiency when operating
at two-thirds of capacity.
3.7
-------
95
tu
OS
34
0.5
1.0
2.0 3 4
H2S/S02 MOL RATIO
(TAIL GAS)
Figure 3-2. Theoretical Glaus sulfur recovery efficiency vs. Mole Ratio.
5 6 7 8 9 10
3.8
-------
Although the Impact of process variables as discussed above
represents ih most cases a loss of only one to three nercent efficiency,
at the relatively high efficiencies typical of Claus operations (95%),
a one to three percent efficiency loss represents a 20 to 60 percent
increase in uncontrolled sulfur emissions.
The Stretford Process
18
The Stretford process shown in Figure 3-3 is a one-step
process to convert H?S directly to elemental sulfur. At present
the Stretford process is the only commercial sulfur recovery
process which has supplanted conventional aroine treating and
CTaus sulfur recovery in a refinery. ATthough the Stretford
is installed- in only one U.S. refinery,, it has found application
in several refineries as part of the tail gas cleanup process for
Claus plants.
Refinery fuel gas or process gas is passed into a countercurrent
column where the hydrogen suTfide in the gas is absorbed in the
Stretford solution.
A summary of the Stretford reactions is as follows:
(a) Absorption of H2S
H2S + Na2C03 -»• NaHS + NaHC03
(b) Precipitation of sulfur
2NaV03 + NaHS + NaHC03 -> S* + Na2V205 + Na2C03 + H20
(c) Regeneration of sodium vanadate
Na2V205 + ADA* (oxidized) -»- 2NaV03 + ADA (reduced)
3.9
-------
in
in
0)
o
o
Q,
"S
O
CO
oo
I
CO
(U
3
o-.
3.10
-------
(d) Regeneration of ADA
ADA (reduced) + 1/2 02 (air) -*• ADA (oxidized)
(e) Overall reaction
H2S + 1/2 02 -»- S4- + H20
After recirculation through the H2S absorber,-the Stretford solution
is retained in a holding tank to allow completion of sulfur precipi-
tation. The Stretford solution is then regenerated by air blowing
and reduced vanadium is restored by oxygen transfer from the ADA*
Sulfur formed in the solution is floated to the top of the oxidizer
by air where the froths overflow to a settling tank. Sulfur settles as
a sludge in the settling tank and is separated and recovered.
Some adverse side reactions occur due to increased liguor
temperature and trace oxidizing gases contained in the fuel gas
(notably oxyqen, Sn2 and HCN) and result in the buildup of sodium
thiosulfate and related comoounds in the circulating liquor which
must be purged from the system. The rate of thiosulfate formation
depends on the partial pressure of the contaminant gas in the inlet
gas stream and the oH and temperature of the liguor. Formation
of thiosulfate is quite low,:he!ow- about 100°F.
Purge stream rates range from 1.5 to 15 gallons oer 100 moles
of feed gas to the Stretford absorber.
*Anthraguinone Disulfomc Acid
3.11
19
-------
The major advantage of the Stretford process is the overall
reduction in emissions, both in the desulfurized fuel gas or
process gas and the emissions from the sulfur recovery plant. Outlet
sulfur loadings have been designed for less than in ppmv H2S in the
fuel gas or process gas.20 COS and CS2 are not recovered by the
Stretford process; however, COS and CS2 in fuel gases or process
gases are much lower than the levels emitted from Claus plants.
Essentially no sulfur is emitted from the oxidizer tank; hence,
sulfur emissions from the Stretford sulfur recovery portion are nil.
Emissions are relatively unaffected by process variables, although
water wastes and chemical consumption are highly dependent upon
operation.
The primary disadvantage' of the Stretford process is the lack
of operating experience in the U.S. In the one U.S. refinery anplication,
the Stretford has yet to demonstrate good operability, experiencing
numerous design and operating problems.21 With time, however, the
design and operational problems will undoubtedly be resolved so that
the Stretford process will be competitive with the Claus process
in refinery applications.
Other Alternative Processes
Other sulfur removal and recovery processes which could conceivably
replace conventional amine treating and Claus sulfur recovery include
the Giammarco-Vetrocoke H2S Process and the Sulfox Process.22'23 The
Giammarco-Vetrocoke H2S removal process is similar to the Stretford
process, except that arsenate replaces vanadate in the scrubbing
solution. The Sulfox process uses an aqueous ammonia solution to
3.12
-------
absorb H2S. Elemental sulfur is produced by oxidizing the solution
over a catalyst.
The Giammarco-Vetrocoke process has been applied to natural gas
processing but has not yet been used in refinery sulfur recovery.
The Sulfox process is designed specifically for refinery sulfur
recovery but was only recently announced. Since neither process
has any operating experience in refinery applications, a detailed
discussion of operation and emissions is not included.
3.2 EXISTING EMISSION CONTROL REGULATIONS FOR SULFUR RECOVERY PLANTS
Table 3-2 summarizes current state emission control regulations
for both existing and new sulfur recovery plants. The most stringent
regulations for sulfur plant emissions are found in Florida, Los Angeles
County, and Philadelphia. Florida regulations specify maximum
allowable sulfur emissions as 0.004 Ib S02 per Ib sulfur input, or
a 99.8 percent sulfur recovery (equivalent to ^500 ppmv total sulfur
?4
on an undiluted basis). Los Angeles County restricts emissions
to 500 ppmv sulfur calculated as S02 equivalent,25 while Philadelphia
OC
restricts refinery sulfur plants to 500 ppmv S02-
State codes applicable to sulfur recovery plants are generally
expressed in pounds of sulfur input. A majority of states which
have sulfur emission regulations for sulfur recovery plants have
the same regulation, 0.01 Ibs S/lb S input, or 99 percent sulfur
recovery.
The format of standards in Table 3-2 shows some states regulating
S02 only, while several have standards for "sulfur emissions." The
3.13
-------
State
Alabama
Arkansas
Colorado
Connecticut
Del aware
Florida
Louisiana
New Hampshire
New Jersey
Ohio
Oklahoma
Pennsylvania
Texas
Utah
Vi rgi ni a
West Virginia
Tabl£ 3-2
State Regulations for Sulfur Recovery Plants
Regulation^
Existing Plants New Plants
0.16 Ib S/lb S input 0.08 Ib S/lb S input
ground level concentrations only
ground level concentrations only
0.01 Ib S/lb S input
2000 ppmv S02
0.004 Ib S02/Jb S input
0.01 Ibs S/lb S input
0.01 Ibs S/lb S input
15,000 ppmv S02
0.01 Ib S/lb S input
(a)
(b)
0.05 Ib S/lb S input
0.06 Ib S/lb S input
--0.5
0.01 Ib S/lb S input
2000 ppmv S02
0.004 Ib S0£/lb S input
0.01 Ibs S/lb S input
0.01 Ibs S/,lb S input
15,000 ppmv S02
0.01 Ib S/lb S input
0.01 Ib S/lb S input
(a)
2200 ppmv S02d(Ref. 28)
(c)
0.05 Ib S/lb S input
0.06 Ib S/lb S input
(a) According to A=0.32 E~u*° where E = plant rating in LT/D and A =
allowable emissions in Ib S02/lb S input.
(bj 214 Ibs S0?/hr/1000 scfm effluent flow rate.
(c) 80% control for new sources with uncontrolled sulfur emissions
;!250 ton/year.
(d) at incinerator outlet, calculated for 50% excess air.
3.14
-------
intent of sulfur emission codes is not specified, but assumed to
be S02 regulation only, since emissions of H2S or CS2 at these
levels would incur severe odor problems.
3.3 REFERENCES
1 Genco, Joseph M. and Tarn, Sanuel S., "Characterization of Sulfur
from Refinery Fuel Gas," EPA Contract 68-02-0611, Task 4, June 28,
1974, p. 28.
2. World Wide HPI Construction Boxscore, Hydrocarbon Processing,
Section 2, February 1974, pp. 3-10.
3. Reference 1.
4. Reference 2.
5. Reference 1, page C-l.
6. Beavon, David K., and Vaell, Raoul P., "The Beavon Sulfur Removal
Process for Purifying Claus Plant Tail Gas, presented at the 37th
Midyear Meeting, APl'Division of Refining, May 8-11, 1972, p. 272.
7, Reference 1, p. 30.
8. Beers, W.D., "Characterization of Claus Plant Emissions,"
EPA Contract 68-02-0242, Task No. 2, April 1973, p. 9.
9. Letter, O.C. Roddy, Vice President, Ralph M. Parsons Co.
to W.D. Beers, Processes Research Inc., March 23, 1972.
10. Letter, D.F, Cole, Contract Engineer, J.F. Pritchard & Co.
to W.D. Beers, Processes Research Inc., March 23, 1972.
11. Reference 8, p. 9.
12. Grekel, H. et al., "Why Recover Sulfur from H2S?" Oil arid Gas
Journal, Oct. 28, 1968, p. 95.
13. Reference 12.
14. Reference 10.
15. Reference 9.
16. Reference 9.
17. Reference 10.
18. Reference 1, op. 32-35.
3.15
-------
19. Reference 1, p. 36. •
20. Brochure, "Sulfur Recovery Qualifications and Experience," J.F.
Pritchard and Company, Kansas City, Mo., Sept. 1973, p. A-4.
21. Letter, L.W. Larsen, Sun Oil Company to James F. Durham, ESED,
OAQPS, EPA, dated January 2, 1975.
22. Reference 1, pp. 38-40.
23. Conser, R.C., "Here's a New Way to Clean Process Gases,"
01:1 and Gas Journal, April 1, 1974, pp. 67-70.
24. Duncan, L.J., "Analysis of Final State Implementation Plans -
Rules and Regulations," EPA Contract 68-02-0248, July 1972,
(APTD-1334) pp. 55-57.
25. "Rules and Regulations," County of Los Angeles Air Pollution
Control District, Rule 53.2.
26. Section II (B), Air Management Regulation III of the Air
Pollution Control Board, City of Philadelphia, adopted April 10,
27. Reference 24, p. 12.
28. Private communication, Frank J. Spuhler, Texas Air Control Board
to Charles B. Sedman", Industrial Studies Branch, ESED, OAP,
January 17, 1974.
3.16
-------
4. EMISSION CONTROL TECHNOLOGY
4.1 ALTERNATIVE EMISSION CONTROL TECHNIQUES
Removal of sulfur compounds from Claus plant tail gas is possible
by three general schemes:
(1) Extension of the Claus reaction to increase overall
sulfur recovery;
(2) Conversion of sulfur gases to sulfur dioxide (S02)
followed by S02 removal technology;
(3) Conversion of sulfur gases to hydrogen sulfide (H2$)
followed by H£S removal technology.
Option (1) is conceptually simple and requires only adjustments
in the operating temperature of the three-stage conventional Claus.
Ootion (2) is attractive because the S02 removed can be recycled
to the Claus plant. An incinerator which is standard equipment for
sulfur plants oxidizes all gaseous sulfur comoounds to S02 prior to
scrubbing.
Option (3) is attractive because H2S removal technology is advanced
and H2S can be recycled to the Claus plant or directly oxidized to
sulfur. Conversion of all sulfur species to H2S requires heat, a
reducing gas, and a reducing catalyst.
Figure 4-1 shows the various commercially available control schemes
and how each utilizes ootions (T), (2), and (3) to remove sulfur from
Claus tail gas. Although the literature cites many tail gas control
systems, those described below were chosen as renresenting available
control technology for Claus sulfur plants and are presently in
commercial operation.
-------
CO
CD o;
t *~~*
LU e: u. § §
f— o c =c >•
JO -J LU eC
CO g LU §"^
-J s: c\i i — a: a; .. ..
O _J 1 1
c£ _i a. Q-
1— UJ LU U_
2= 3C H-l l-H
O
f ^ * • • • • •
U- O LU U_
= s t fe
^ ^ T rv S
=> CO
«C CQ CJ Q Q LL: CO ..
»k i
CO
o; CD
LU CO
CQ ^^ r )
0 1 5!
co u. 1—
CQ U_
•a: O *
t
cc; — J «c _
*-*— — — > TJ; ^r5
O rv ~-^ CD ^^
1- U_ S h- ^ I ^
1 — Lu SZ ^ LJ_
LU — 1 eC Ll_
rv* *"~> i f^
1— co n_
co cc:
^-^ LU
^J. ^ ^ S .
-^ r^ to
*~ CO t-
L co 5
1 d; *
o
zz S = S
o j— z GZ o (— 52
I-H^ ooo; i— i=c *"
CO Qi S2=
0 UJ
CO CD N
CQ UI ^
1-1 co i— ^rs -rv^ co H- o;
O ro«=ccc _j^ — ^ cc. -z. *
™ — 1 _J LU =D O LU
, o cj o; co co CD
CO CQ LU
•=c oi
6 >
0
Qi CM * — *
LUO C "S"
2:1— co ci
1 — 1 ^ ^
1 — 1 "*
cC CQ
k ^ , i
^5 y , e
J C
— CO
>- cc: co
_1 LU O CM
a: > (— n:
a: o
IZ" "CD Jl
^
CD
1 1 1
i
CJ
CJ
LU
I UC.
CD CO
«c co oc D; .^
1 — -» ^-* 1 — -j ^T
t^ — J —J r— — j -*-
CO eC Li_ SZ >Lu
r _i _] •< _i
CD cj ra —i rs
3: CO Q. CO
>
-
CO
^
O
S-
O)
c^
qZ
£
0
q
E
QJ
1 ^
to
to
O
S-
c
o
CJ
0
co
£
LU
•
1
*%}"
£
CT
C£
£-
-------
Extension of the Claus Reaction
The conventional Claus plant, having two or three stages, was
discussed in Chapter 3. Such a unit has a recovery
efficiency of about 95%. To gain further improvement in sulfur
recovery efficiency via the Claus reaction, the temperature
can be lowered to shift the equilibrium of the Claus reaction
toward formation of additional sulfur. Two processes have been
developed for reducing emissions through extension of the Claus
reaction by operation at low temperatures. These are the IFP-1 and
Sulfreen processes.
IFP-1 PROCESS1'2
The basic reaction involved in the IFP-1 process (Figure 4-2)
is the same one which takes place in the reactors of a Claus unit:
Catalyst
2H2S + S02 * 3S* + 2H20
Tail gas which exits a Claus unit at 265-285°F can be fed directly
into the IFP reactor without cooling the gas. The reactor is a
packed column with a specially designed "boot" for collecting sulfur.
Metal salts catalyze the reaction which takes place in a hiqh boiling
polyglycol (PEG) solvent above the melting point of sulfur-generally
in the ranae of 250-260°F. The metal salts form a complex with H2S
and S02 in the feed gas, which in turn reacts with additional H2S and S02
to form elemental sulfur and regenerate the catalyst complex. Sulfur
coalesces and settles into the boot of the reactor, from which it
is drawn as a molten product.
4.3
-------
00
(— I-H
«=C C£
O in
2:
_1 i—i
t-l CO
-------
The gas leaving the reactor contains about 850 ppm H2S and 420 ppm;
S0?, assuming a 90 percent conversion. Conversion efficiency is
maximized at the 2:1 stoichiometric H2S/S02 ratio required by the
Claus reaction. Above or below this ratio the conversion efficiency
diminishes; hence, process controls must effectively keep the
H2S/S02 ratio in the !FP reactor feed as near 2:1 as possible. Also,
below 2:1 H2S/S02 ratio in the gas stream, the solvent tends to
evolve absorbed H2S, which also increases ultimate sulfur emissions.
Besides H2S and S02, about 300 ppm sulfur vapor is evolved with
the exit gas. This is roughly the equilibrium concentration of sulfur
in the vapor phase at 260°F. COS and CS2 formed in the Claus plant
are not affected by the IFP-1; hence, they exit at aporoximately
the same concentrations as in the feed gas. The reactor exhaust
containing 1500-2500 pom sulfur and some entrained solvent is
incinerated before discharge to the atmosphere.
SULFREEN PROCESS3'4
The Sulfreen process (Figure 4-3) reduces the sulfur content
in Claus olant tail gas by further promoting the Claus reaction
on a catalytic surface in a gas/solid batch reactor. Claus tail
gas is first scrubbed with liquid to wash out entrained sulfur
liquid and sulfur vapor. The tail gas is then introduced to a
battery of reactors where the Claus reactions are carried out at lower
temperatures (260-300°F) than those utilized in the sulfur plant.
Lower temoeratures oush the Claus reaction toward comoletion
due to favorable equilibrium conditions. The catalyst is usually
activated carbon, though alumina is also used.
4.5
-------
ce
LU
o
o
CD
O
£
Q..
t-
00
O)
S-
o
rO
i-
cr
n3
S-
3
cr.
4.6
-------
A regeneration gas, essentially nitrogen, periodically desorbs
the sulfur-laden catalyst beds. Nitrogen is heated and cycles
through the catalyst bed at approximately 570°F until all water and
C02 are driven off.
For the desorption of sulfur, the temperature is raised to 750°F
where sulfur vaporizes, is swept away with the nitrogen, and precipitates
in a condenser. The carrier gas is further scrubbed in a sulfur wash
before returning to the regeneration cycle.
The process reduces entrained sulfur, since the catalyst acts
as an absorbent for liquid sulfur. However, tail .gas passing through
the catalyst bed retains several hundred ppm sulfur vapor in equilibrium
with liquid sulfur. H2S and S02 are reduced by 80-85 percent to
levels of about 1800 ppmv H2S and 900 ppra S02. As with the IFP-1 process,
the levels of ^ and S02 are highly dependent upon maintaining the
2:1 ratio of H2S/S02 in the Glaus tail gas. COS and CS2 are not affected
by the Sulfreen process.
A Sulfreen unit may consist of as little as three reactors, two in
absorption and one in desorption service. The gases from the reactors
in desorption service are incinerated before discharge to the atmosphere.
Conversion to S02 Followed by S02 Recovery
Two processes have been developed to reduce emissions from
Claus sulfur plants through conversion of the sulfur compounds
present in the sulfur plant tail gas to S02, followed by recovery
of the S02. These are the Wellman-Lord and IFP-2 processes. Incineration
of the sulfur plant tail gas effectively converts all the sulfur
4.7
-------
compounds present to Sf)2 and this is recovered in a conventional
S02 scrubbing system. In both processes the S02 recovered is
recycled to the sulfur olant for eventual conversion to elemental
sulfur by the Glaus reaction.
WELLMAN-LORD PROCESS5'6
The Well man-Lord process (Figure 4-4) uses a wet regenerative
system to reduce the stack gas sulfur concentration to less than 250 ppmv.
Sulfur constituents in Claus nlant tail qases are oxid'ized to S02 in
the standard sulfur plant incinerator, then cooled and quenched to
reduce the gas temperature and remove excess water. The S02 rich
gas is then contacted countercurrently with a sodium sulfite (Na2S03)
and sodium bisulfite (NaHS03) solution which absorbs S02 to form
additional bisulfite. The principal reaction between S02 and the
absorbent solution is:
S02 + Na2S03 + H20 t 2NaHS03
The absorber off-gas is reheated and vented to the atmosphere
at less than 250 pomv S02 and negligible amounts of other sulfur comoounds.
Any S03 produced by the incineration is preferentially absorbed.
S02-rich solution is boiled in an evaoorator-crystallizer, wherein
the bisulfite solution decomposes to S02 and H20 vaoor and sodium
sulfite is precipitated according to the reaction:
heat
2NaHSOs * Na2S03* H20 + S02+
A. 8
-------
C
00
CO
CD
O-
D-
o 5
a. «a:
•=C UJ
> I—
UJ CO
in
OJ
o
CO
o
_J
I
C.
(O
E
OJ
S-
o
CO
s-
o;
-------
Strtfite crystals are separated and redissolved for reuse as lean
solution to the absorber.
The wet S02 gas flows to a partial condenser where most of the
water is condensed and reused to dissolve the sulfite crystals.
The enriched S02 stream is then recycled back to the Claus plant
for conversion to elemental sulfur.
IFP-2 PROCESS7'8
With the IFP-2 process (Figure 4-5), Claus plant tail gas is
incinerated to convert all sulfur species to S02. The incinerated
gas is cooled and then fed to an ammonia scrubber, where S02 is
absorbed and converted to ammonium sulfite and ammonium bisulfite
by the following reactions:
2NH4OH + S02 $ H20 + (NH4)2S03
NH4HH + S02 $ NH4HS03
Sulfates and thiosulfates are also formed in the ammonia scrubber by
the foil owing side reactions:
2NH4OH + S02 + 1/202 -»• H20 + (NH4)2S04
2NH4OH + S03 -*- (NH4)2S04 + H20
2NH4OH + S02 + S •* H20 + (NH4)2S20,
3
Gas leaving the absorber is reheated and vented to the atmosphere
at less than 250 ppm S02 concentration.
The S02 rich solution is fed to an S02 regenerator where the
sulfite and bisulfite are thermally decomposed to S02, NH3, and H 0.
A saturated solution containing ammonium sulfate and thiosulfate
is drawn from the bottom of the S02 regenerator and fed to a sulfate
reducer.
4.10
-------
O-
=>
UJ
UJ
o
co
CD
U. •*£
U_ <_3
°P
a; co
^ —
CQ o
g-
CO
QJ
O
O
O.
I
CD
(O
cr.
i
to
Z5
to
CM
O-
en
to
•
eC CO I— i
_j «=C o: i—
Q. CD UJ -=C
h- C£
CO — J U- UJ
=3 1-1
-------
In the sulfate reducer, the sulfates and thiosulfates formed in
the side reactions are thermally decomposed oer the following reactions
(NH4)2S04 -»- NH4HS04 + NH3+
(NH4)2S203 -j- 2NH3f + S02+ + S + H2(H
2fJH4HS04 + S -*• 3S02f + 2NH3f + 2H2(H
Gases from the S02 regenerator and sulfate reducer are combined
with an H2S rich stream (normally Claus plant feed gas) and fed to
a catalytic reactor where they are contacted with a polyethylene
glycol solvent. The'H2S and S02 react in the solution to form
elemental sulfur. Sulfur is withdrawn in the molten state and
sent to the Claus plant sulfur pit.
Gases from the reactor are cooled to condense out water and NHs
as NH4OH. The NH4OH solution is returned to the ammonia scrubber.
Any H2S or S02 which leaves the reactor -and is not absorbed by the
NH4OH solution is recycled to the incinerator and from there to
the ammonia scrubber.
Conversion to H2S Followed by H2S Recovery
Three processes have been developed to reduce emissions from
Claus sulfur plants through conversion of the sulfur compounds
present in the sulfur plant tail gas to H2S, followed by H2S recovery.
These are the Beavon, Cleanair and SCOT processes. In each of
these processes, the sulfur plant tail gas is mixed with a reducing
gas such as hydrogen and passed over a reducing catalyst. Most
of the sulfur compounds present are converted to H2S. Both the
Beavon and Clean air processes then use the Stretford orocess
4.12
-------
(discussed in Chapter 3} to absorb the H2S and convert it directly to
elemental sulfur. In the SCOT process the H2S is absorbed in a
conventional H2S scrubbing system and then recycled to the Glaus
sulfur plant for eventual conversion to elemental sulfur.
9,10
BEAVON PROCESS'
The Beavon process (Figure 4-6) begins by converting sulfur
present in the tail gas (S02, COS, CS2, and elemental sulfur) back
to HoS. This is done by hydrogenati on and hydrolysis under moderate
conditions of temperature and pressure similar to those in the
Claus plant. Before the tail gas enters the packed bed hydrogenati on
reactor, fuel gas is combusted substoichiometrically in an inline
burner to produce the reducing conditions necessary to convert
sulfur gases to H2S. The combustion products are mixed with the
tail gas to provide a reducing atmosphere. Extra hydrogen may be
made in the fuel-rich combustion chamber as required to supplement
hydrogen already in the tail gas by the reaction:
CO + H20 * H2 + C02
A cobalt-molybdenum catalyst promotes the hydrogenation and
hydrolysis reactions which reduce S02 to very low values and
(COS + CS2) to less than 100 ppmv. Elemental sulfur is completely
reduced to H2S. The reactions are:
8H
8H2S
S02 + 3H2 •> H2S + 2H20
COS + H20 + C02 + H2S
CS2 + H20 * H2S + COS
4.13
-------
LU
Z
a:
QQ
UJ
u.
U)
o
o
Q.
10
O
S-
q-
c
o
to
o>
03
O)
I
cr.
O)
CD
1
L
ta/
t '
LlJ
o:
4.14
-------
CS2 + 2H2S
CH4 + 4S
4H2 + CS2. ^ 2H2S + CH4
COS + 4H2
H2Q + H2S + CH
After hydrogenation, the tail gas is cooled and water is removed.
The H2S-rich tail gas is fed to the Stratford circuit for removal of
H2S to less than 10 opm. final incineration before discharge to the
atmosphere is optional .
CLEANAIR PROCESS11'12
The CLeanair process includes the Stretford process and two
confidential processes. Figure 4-7 is a process diagram. An optional
part of the Cleanair package includes a modification of the Claus
plant first stage to' include a reducing and hydrolysis catalyst.
This causes the conversion of COS and CS2 to H2S according to the
following reaction:
COS + H20 £ H2S + C02
CS2 + 2H20 £ 2H2S + C02
Carbon dioxide also is decomposed to CO to prevent the recurrence
of COS.
Claus tail gas, with essentially all gaseous sulfur as sulfur
vapor, H2S and S02, is quenched to reduce temperature and remove
water and entrained sulfur. The cooled gas is fed into a reactor
where H2S and S02 (in a 2:1 ratio) react, lowering S02 to less than
250 ppmv. Both water and sulfur (products of the Claus reaction)
are removed.
Next, the tail gas is sent to a Stretford unit where remaining
H2S is removed and oxidized to elemental sulfur. Residual S02,
4.15
-------
-------
although absorbed by the Stretford solution,, decomposes the solution
and therefore increases chemical consumption and liquid purge rate.
Residual COS and CS-2 will pass through unaffected. Purified gas
may then be sent to an incinerator to- oxidize residual sulfur to
S02 (guaranteed less than 250 ppmv) and CO to- C02 before discharge
to the atmosphere.
TO: "14
SHELL CLAUS OFF-GAS TREATING (SCOT) PROCESS '
Similar to the Beavon process, the SCOT process (Figure 4-8)
first converts all sulfur compounds, and free sulfur in the tail
gas to HoS. Hydrogen, or hydrogen and carbon monoxide mixtures
are used as the reducing gases while a cobal t/mO'lybdenura-on -alumina
catalyst- promotes the reactions:
S0
8
H2S + 2H20
8H2S
COS and CS2 are reduced in reactions identical with those in
the Beavon catalytic reactor. Where carbon monoxide is also present
as a reducing agent, the following additional reactions may occur:
S02 + SCO -»• COS + 2C02
S8 + SCO -»• 8COS
COS + H20 t C02 •*• H2S . .
CO + H20 * C02 + H2
CO + H2S S- COS 4- H2
From the SCOT reactor, the tail gas containing less than ion ppm
(COS + CS2) and 10 ppm S02 is cooled and excess water removed. The
H2S (at 20,000-40,000 oprnv) and some C02 are then removed by treating
4.17
-------
4->
•r—
E
to
ro
o
o
ro
cr
ro
s-
(I)
c:
HI
en
o
4-
OJ
E
E
ro
_j
s-
OJ
.Q
S-
o
to
JQ
•=c
S-
O)
re'
s-
s-
o
ro
CD
I
O)
o
O)
E
•i—
ro
ro
a-.0
C T3
•r- O)
i— ^4
O 0
o re
co a.
i s-
- s- oj
4-^ 3 o
C C..
S- OJ to -i-
OJ 4-3 4.
+-> ro cr.+J
re to E to
S E *r-
I QJ 4-> S-
S- T3 to 0}
3 E -i- 4->
o o x re
co o as 3
s-
o
U
re
•a
cu CT+-J
J3 E to
I t- >,
•o o •—
aj 3 ro
X T3 +->
•r- cu ro
u. s~ o
re
O)
--C
"V
\
r
CO
CO
LU
CJ
o
s
o:
-p-
u
+J c
E -r-
re
i— O
C.4->
w s-
3 C
re -i-
i — S-
cj c.
10 S-
re -t-
4.18
-------
16
with di-isoprooanolamine solution in an .absorption column. The H2S-
rich solution is regenerated by stripping H2S in a conventional
steam stripping column. Regenerator off-gas, mainly H2S and some-
C02 is recycled as feed to the first stage of the Glaus unit.
The absorber off-gas, containing less than 300 ppmv H2S is
incinerated in a standard Claus incinerator.
4.2 COMMERCIAL STATUS OF TAIL GAS TECHNOLOGY
The Wellman-Lord process has demonstrated the longest trouble-free
15
operation—over two years on a Claus plant, Wellman-Lord units on
Claus plants are in operation at three refineries with five
additional units scheduled for start-up in 1974 and early 1975.
The IFP-2 process is being installed in three Japanese refineries.'7
Of units designed specifically for Claus tail gas treating, the
Beavon, SCOT, and IFP-1 units have been successfully operated in refineries.
Four Beavon units have been running continuously at two refineries since
July 1973.18 Three other Beavon units started up during 1973, with three
more scheduled for completion in 1974.19 Two small SCOT units were
installed in June 1973, and are currently in operation followina some
modifications.20 At least seven more SCOT'S were scheduled for start-up
in 1974.21
The IFP-1 and Sulfreen processes have been applied to natural
77
gas plants in France and Canada for about five years. Also, the IFP-1
is currently in operation on six refinery sulfur plants in Japan
23
and two in the U.S. with several more announced.
The Cleanair process was installed at three U.S. refineries
24
in 1973; but all are still awaiting start-up after process modifications."
A detailed list of present and planned Claus tail gas control
systems is presented in Table 4-1.
4.1-9
-------
Table 4-1. LISTING OF ANNOUNCED TAIL GAS TREATING
UNITS FCR CLAtfS SULFUR PLANTS'AS OF 3/1/74
Type Unit
Beavon
Beavon
Beavon
Beavon
Beavorr
Beavon
Beavon
Beavon
Cleanafr
Cleanalr
Cleanair
Cleanair
WeTlman-Lord
Well man-Lord
Company/Locatton
Union Oil Company/
Wilmington, California
Mobil Oil Company/
Torrance, California
Atlantic-Rich field/
Philadelphia, Pennsylvania
Getty Oil Company/
Delaware City^ Delaware
Kobe Steel Co ./Japan
Texaco, Inc./
Long Beach, California
Unknown/
Carribbean
Union Oil Co./
Rodeo, California
Atlantic Richfield Corp. (ARCO)/
Wilmington, California
Gulf Oil Co./
Philadelphia, Pennsylvania
Santa Fe Springs Refinery
(Gulf Oil Co.)/Santa Fe
Springs, California
Techmashimport, U.S.S.R.
Standard Oil of California/
Richmond, California
Kashima Oil Co./
l^achnmar .lanaw
Onstream
Date-
July 1973
July 19.73:
September 1973
November 1973
October T973
March 1974
April 1974
November 1974
--
—
—
--
January 1975
February 1974
No. of Number/Capacity
Units Claus Plant* LT/
2' 2/1 QQ.
2 3/100
T 1/140
1 1/300
1 1/220
•j
2
3
1 3/90
1 1/46
1 1/27
2
1 1/290
1 1/180
4.20
-------
Table 4-1. (eorjt.) LISTING OF MN'OUICED TAIL GAS TREATING
PUTS FOR CUIUS SULFUR PLANTS AS «F 3/1/74
Type Unit
MeTnimami-Lord
WeTI mam-lord
Meirilman-1'O.rd
Mel taam-L'O rd
Bellman-Lord
WeTlman-Lord
SCOT
Company /Loca ti on
To a ,Ne nryo/.-AH ta 9 .
Japan
T>oa learyo/Mat-syS'hima:,
Toa ilfenryo !ief«e^/fe,wasa>kii ,
Japan
Standard Oil ®f 'Calif orrona/O
Segundo, California
Standard Oil
-------
Table 4-1. (Cont.) LISTING OF ANNOUNCED TAIL GAS TREATING
UNITS FOR CLAUS SULFUR PLANTS AS OF 3/1/74
Type Unit Company/Location
IFF
IFF
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
IFP
Nippon Petroleum Refining
Company (Caltex)/Negishi ,
Japan
Idemitsu Oil/Himeji, Japan
Showa Oil (Shell )/Kawasaki ,
Japan
Kyokuto Petroleum Industries
(Mobil )/Chiba, Japan
Chevron Standard Ltd.
(Chevron Research,)/Nevis,
Alta., Canada
Mitsubishi Oil Company (Getty)/
Mizushima, Japan
Mitsubishi Oil Company (Getty)/
Mizushima, Japan
Phillips Petroleum/ Borger
Texas
Ministry of Gas/Orembourg I,
U.S.S.R.
Ministry of Gas/Orembourg II,
U.S.S.R.
Ministry of Gas/Orembourg III,
U.S.S.R.
Stauffer Chemical Company/
Delaware City, Del., U.S.A.
Commonwealth Oil Refining/
Ponce, Puerto Rico, U.S.A.
Koa Oil No. 1/Marifu, Japan
Phillips/Sweeny, Texas
Koa Oil No. 2/Marifu, Japan
Unannounced/
Onstream
Date
1971
1972
1972
1972
1972
1972
1972
1973
1974
1974
1974
1973
1973
1973
1973
1974
1975
No. of Number/ Capacity o
Units Claus Plant, LT/D
1
1
T
1
1
1
1
1
1
1
1
Confidential
1
-
1
-
-
1/300
1/250
1/80
1/200
1/260
1/180
1/350
1/45
1/800
1/800
1/800
Confidential
1/60
--
1/45
--
1/400
. 22
-------
4.3 PERFORMANCE OF EMISSION CONTROL SYSTEMS
The Well man-Lord, IFP-2, Beavon, SCOT and Cleanair processes
all guarantee sulfur emissions less than 200-300 ppmv, for an overall
sulfur recovery including the Glaus sulfur plant of better than
99.9 percent. The Sulfreen and IFP-1 are capable of reducing sulfur
emissions- to 1500-3000 ppm, for an overall 99.0 percent sulfur
recovery including the Claus sulfur plant.
Table 4-2 summarizes the expected concentration and mass
emissions as discussed above both for emission control systems on
a 100-long-ton-per-day Claus sulfur plant. Uncontrolled emissions
are also included for comparison.
Emission Source Test Results
Four processes—the IFP-1, Wellman-Lord, SCOT and Beavon—were
tested to determine emissions of S02, H2S and reduced sulfur compounds
(H2S, COS and CS2). The data gathered by these tests are presented
in Figures 4-9 through 4-12. The data represented by black circles
is data gathered by the Agency, while the data represented by white
circles is data gathered by refinery personnel or the LA APCD.
Figure 4-9 summarizes the results obtained from an emission
test on the IFP-1 process. This process was not operating on a
petroleum refinery sulfur recovery plant, but was operating on a large
Claus sulfur plant installed in a carbon disulfide plant. The data,
however, is applicable to refinery sulfur plants and the emission
levels achieved representative of what would be achieved by the IFP-1
process on a refinery sulfur plant. As shown in Figure 4-9, emissions
of S02 ranged from 2300 to 2600 ppm.
4.23
-------
CO
>— 1
Of.
LU
U.
UJ
U
LU
_J
CC
LU T3
O.U
t-H CO
CM
LU
co c:
LU cr
o
o »
o. c
LU (O
—J "O
co to
«-§
§+•
LL. cr,
c
CO 0
ic
co cr
CO r—
X"
LU ••
t jQ
eC 03
^ **— ^
O
EC
CJ
cv*
1
cu
.£5
CO
-£
*^
-C
•13.
E
c
c
CO
c
•r-
V
•r—
E
LU
r^ S- "ci
CO 3 5
CUi— 1
> = a
cr co cc
O>
o
CO
^
"ro
2
cv.
CO
to
CO
CD
CJ
i.
O
C.
CO
CC
CO
c§*
CO
CO
CM
a:
..^
CJ
- c
3 CU
> -i—
2 u
" LU
CO
cu
o
o
S-
CL.
o
0
en
' 1
I
cu
CJ
£
4-*
1
1
_
cr
C7.
cr
cr
i cr
cr cr
cr i—
cc
i
i
ai
CJ
(O
s.
^
LO
CTl
CO
O
CO
CU
•r—
CJ
1 — 1
CO
CO
o
^ ^
1
tl
1
•o
CO
43
{
1
1
^
'5-'
cr
CM
1
CU
CJ
CO
S-
+-1
c
cr.
cr.
CO
cr
CJ
CO
CO
CO
CJ
^
^
ITS
, CM
CM
cr
•la
.
tcr
CO
Ifc
.1
11
1
1
1
1
1
cr
V
en
ct-
cn
c:
o
CO
cu
CQ
CO
CO
o
^_^
; i —
~- '
i tn
CM
LO
C
CO
iO
?r^-
i
i
1
i
i
i
0
V
cr-.
en
CT!
S-
CO
CO
cu
CJ
"*"
CO
CO
CJ
'.
J
il
(CU
1O
CO
•5-
|
.'1
cr
CM
CT
LO
CM
1
CU
o
CO
S-
cr.
cri
o
i
c:
CO
i—
cu
'
CO
CO
CJ
\ cr
'-- —
• O
LO
CM
: 1
CU
1C3
! 110
*J
II
ai
o
CO
S-
1
1
!l
1
O
CM
C
LO
CM
1
1
cu
o
CO
i.
cr
CTl
en
CM
a_
LL.
"r
CO
CO
CJ
c
cu
c
o
*r—
4->
CU
C
CJ
CO
t
S-
cu
4^
tj
tc
o
w
(/I
t/) d)
OJ >
-*J CD
O _J
s: re
-------
E:
0*
•i-LO
w t/r
•r- VI
E «B
tu O
o
01; s-
X'tr—
O> I-
S~
CO
ov
cu
S-
cr.
u.
O
(0
O
O
CD
«=!-
O
O
O
co
o
CD
O
CM
O
o
o
(uidd) SUOLSSLW3
4.25
-------
Figure 4-10 summarizes the results obtained from emission
tests on a Wellman-Lord process and a SCOT process. Test Bl
was conducted by the Agency, test B2 was conducted by refinery
personnel and tests B3 and C3 were conducted by the LA APCD.
The Wellman-Lord process is an oxidation-scrubbing process, while
the SCOT process is a reduction-scrubbing process followed by
incineration. Emissions of S02 from the Wellman-Lord process ranged
from 10 to 50 ppm and emissions from the SCOT process averaged
210 ppm.
Figures 4-11 and 4-12 summarize the results obtained from
emission tests on the Beavon process. Test Bl was conducted
by the Agency, test B2 was conducted by refinery personnel,
and Tests B3, E3 and F3 were conducted by the LA APCD. Emissions
of H2S were undetectable in four of these tests and ranged from
0 to 7 ppm in the fifth. Emissions of reduced sulfur compounds
(calculated as S02) generally ranged from 10 to 20 ppm in each
of these tests.
Vendor Guarantees
Vendors of the higher efficiency tail, gas treatment processes
normally guarantee emissions of total sulfur, expressed as S02
equivalent, not to exceed 200 to 500 ppmv. [SQ2 equivalent is equal
to (S02 + H2S + COS + 2CS2 + sulfur vapor) expressed in gas
concentrations only.] The J.F. Pritchard Company designed the
Cleanair tail gas process not to exceed 200 ppmv, 250 npmv, and 300 pnmv
total sulfur emissions for their first three installations.26>27,28
Union Oil Research designed the initial Beavon units not to exceed
4.26
-------
ro
rv
o
CO
(/) f
OCM
to to
to OXM
£ O to
uu s- to
D_ CD
CU 0
X O Du
O _l
••- I I—
Q C O
fO O
s- e oo
CU
C7>
ea
0}
T3
o
O)
to
+J CO
to 1— T3
•< ^:
-------
-a
o
in
CO
ro
CO
OT
CO
VI
a>
en
s_
01
DL.
UJ
t/)
O)
o
O
CD O
«=
cc
o>
O
•a
CO
CO
O
(O
Lu
3
D1
CM
03
CQ
o
CO
O
CM
(uidd) SUOLSSLIU3
4.28
-------
CO
PO
CO
<£>
OO..
in
to
LU
TD
II
3 a)
00(2
O)
o
CM
cu
CD
tZ
s
CU
cr,
(O
O)
-o
o
.c
-------
210 DDHIV total sulfur emissions.39 shell Development gave emission
Guarantees-of 500 and 400 ppmv hydrogen sulfide for their initial
installation of the SCOT process, though they have cited guarantees
as low as 300 DPIW H2S (before incineration) in their literature.40^! ,42
Davy Power Gas guaranteed 250 ppmv total sulfur as S02 on their
first U.S. Claus nlant application of the Wellman-Lord systen.43
Other tail gas processes which do not remove COS or CS2 usually
guarantee efficiencies of H2S and S02 removal. These guarantees
cannot be directly correlated to a total sulfur emission guarantee
because the concentrations of COS and C$2 vary according to Claus
feedstock comoosition and operating conditions in the first stage
catalytic converter. An example of vendor emission guarantees is
the IFF guarantee at one installation of 90 percent removal of
H2S + S02 in Claus tail gas.44
Neither of the vendor guarantees soecified whether the concen-
trations quoted were on a dry or wet basis or whether they were adjusted
to zero percent oxyaen. Therefore, the final levels of H2S and S02
could differ from the guaranteed levels when these conditions are
taken into account.
4.30
-------
4.4 REFERENCES
1. Genco, Joseph M. and Tarn, Samuel S., "Characterization of Sulfur
from Refinery Fuel Gas," EPA Contract 68-02-0611, Task 4, June 28,
1974, DP. 66-68.
2. Barthel et. al., "IFF Processes for Recovering h^S and S02 from
Claus Unit Tail Gas and for Cleaning S02 from Stack Gas," APCA
Paper 73-304.
3. Reference 1, pp. 84-90. •
4. Krill, H. and Storp, K., "HoS Adsorbed from Tail Gas," Chemical
Engineering 80<17), pp. 84-85, (July 23, 1973).
5. Reference 1, pp. 52-55.
6. Potter, Brian H., and Earl, Christopher B., "The Wellman-Lord S02
Recovery Process," presented to the 1973 Gas Conditioning Conference.
7. Reference 1, pp. 68-71.
8. Reference 2.
9. Reference 1, po. 43-49.
10. Beavon, David D. and Vaell, Raoul P., "The Beavon Sulfur Removal
Process for Purifying Claus Plant Tail Gas," Proceedings, API
Division of Refining', 1972, Volume 52, pp. 267-276. .
11. Reference 1, pp. 49-50.
12. Landrum, L.H., Corn, L.H. and Fernald, W.E., "The Cleanair Sulphur
Process," presented at the 74th AIChE Meeting, New Orleans, March 11-15,
1973.
13. Reference 1 , op. 60-66.
14. Naber, et. al., "The Shell Claus Off-Gas Treating Process,"
presented at the 74th AIChE Meeting, New Orleans, March 11-15, 1973.
15. Trip Report - "Visit to Wellman-Lord S02 Recovery Unit," C. Sedman,
Industrial Studies Branch, ESED, OAP, Nov. 2, 1973.
16. Reference 1, p. 56.
17. Reference 1, p. 71. .
4.31
-------
18. Trip Report - "Visits to Beavon Sulfur Removal Units," C. Sedman,
Industrial Studies Branch, ESED, OAP, Nov. 5, 1973.
19. Reference 1, D. 45.
20. Trip Report - "Visits to SCOT Units," C. Sedman, ESED, OAOBS, EPA,
Nov. 2, 1973.
21. Reference 1, p. 63.
22. Reference 1, pn. 72-87.
23. Reference 1, pp. 72-73.
24. Conversation, C. Sedman, ESED, OAQPS, EPA, with Jim Nance,
Los Angeles County APCD, Dec. 11, 1974.
25. Source Test Report No. 74-SRY-4, EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Engineering,
Gainesville, Florida, June 1974.
26. Letter, H.F. Schaffer, GuTf Oil Company of Pennsylvania to
Charles B. Sedman, ESED, OAOPS, EPA', October 16, 1973.
27. Letter, A.A. Muse, Jr., Atlantic Richfield Company to
Charles B. Sedman, ESED, OAQPS, EPA, November 20, 1973.
28. Letter, F.A. Becker, Gulf Oil Company-U.S. to Charles B. Sedman,
ESED, OAQPS, EPA, November 27, 1973.
29. Source Test Report No. 74-SRY-l , EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Engineering,
Gainesville, Fla., March 1974.
30. Letter, Thron Riggs, Standard Oil Co. of California to C. Sedman,
ESED, OAQPS, EPA,'dated August 9, 1974.
31. Source Testing Section Report No. C-1895, Los Angeles County
APCD, February 28, 1973.
32. Source Test Section Report No. C-2104, Los Anaeles County APCD,
April 25, 1974.
33. Source Test Report No. 74-SRY-2, EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Engineering,
Gainesville, Fla., March 1974.
34. Letter, George L. Tilley, Union Oil Company of California, to
C. Sedman, ESED, OAQPS, EPA, dated August 26, 1974.
4.32
-------
35. Source Test Section Report No. C-226, Los Angeles County APCD,
November 6, 1974, December 12, 1974.
36. Reference 30.
37. Reference 31. .
38. Reference 32.
39. Letter, D.L. Hanley, Union Oil Company of California to Charles
B. Sedman, ESED, OAQPS, EPA, December 4, 1973.
40. Private communication, David Hamilton, Douglas Oil Company,
to Charles B. Sedman, ESED, OAQPS, EPA, October 16, 1973.
41. Letter, R.S. Little, Champlin Petroleum Company to Charles B.
Sedman, ESED, OAQPS, EPA, October 24, 1973.
42. Brochure, "SCOT Process," Ford, Bacon and Davis, Dallas, Tecas,
1973.
43. Letter, Thron Riggs, Standard Oil Company of California to
Charles B. Sedman, ESED, OAQPS, EPA, December 12, 1973.
44. Letter, W.W. Turner, Stauffer Chemical Company to Don R. Goodwin,
ESED, OAQPS, EPA, dated June 14, 1974.
4.33
-------
-------
5. MODIFICATION AND RECONSTRUCTION
Existing sources which are modified may become subject to
standards of performance under section 111 of the Clean Air Act.
Any physical or operational change to an existing facility which
results in an increase in the emission rate of any pollutant to
which a standard apolies may cause that facility to be considered
a modified source, subject to compliance with standards of
performance. The full discussion of modification is included
under Subpart A - General Provisions of Part 60, 40 CFR.
5.1 MODIFICATION OF REFINERY SULFUR BLANTS
Under the General Provisions of Part 60 40 CFR, physical changes
to a facility resulting from routine maintenance, repair or
reolacement will not be considered modifications. Also, neither
an increase in production capacity if this does not require a
capital expenditure^; defined by §60.2(bb); nor the use of
alternative raw materials, if the facility was originally designed
to accommodate materials, will be considered modifications.
The H2S gases processed in refinery sulfur plants are
produced throughout a petroleum refinery in a number of
different process units. Consequently, almost any change within
a petroleum refinery,such as a change in the type of crude oil
processed, construction of new refinery process units, expansion
of existing process units, or even a change in the operation
of various process units, could potentially lead to a modification
of an existing refinery sulfur plant.
5.1
-------
Changes such as these could increase the volume of F^S gases
generated within the refinery, leading to a need for increased
sulfur recovery capacity. Most refinery sulfur plants-, however,
are overdesigned to some extent in anticipation of fluctuations
in the volume of ^S gases generated within the refinery due to
seasonal changes in refinery operation (i.e., maximum fuel
oil production in the winter months comnared to maximum gasoline
production in the summer months), or to changes in the types of
crude oil processed. As a result, most refinery sulfur plants
have excess sulfur recovery capacity and can readily accommodate
changes within the refinery which require only moderate increases
in sulfur recovery capacity.
Where the increased volume of H2S gases is more than
that which can be readily accommodated by existing sulfur plants,
additional plants have usually been constructed rather than expanding
existing plants. While some expansion of existing sulfur recovery
Plants would be possible through operation at higher pressure,
this would also probably require addition of another Claus catalyst
stage to compensate for the unfavorable shift in the equilibrium
of the Claus reaction which would accompany an increase in pressure.
In addition, alterations to the existing furnace and sulfur condensers
would probably also be necessary to provide increased combustion
capacity and increased sulfur condensation capacity. As a result,
the expense and problems associated with expansion of existing sulfur
recovery plants, plus the desire on the part of many petroleum
5.2
-------
refineries to have excess sulfur recovery capacity and more than one
sulfur plant to compensate to some extent for the shut-down of a
sulfur recovery plant due to a malfunction, has led refineries to
construct additional sulfur recovery plants rather than expand
existing ones.
Although as discussed in Chapter 3 the efficiency of sulfur
recovery in Claus plants and hence emissions from sulfur plants
are dependent on the composition of the H2S gases processed (lower
concentrations decrease efficiency), changes within petroleum
refinereies such as those mentioned earlier do not lead to significant
changes in the composition of the H2S gases processed by refinery
sulfur plants. The amine treating units which remove H2S
from the gases generated by various refinery process units essentially
recover a gas stream containing 80-90 percent H2S regardless of the
H2S content of the original process gas.
Occasionally, however, in response to local or state air
pollution control regulations, petroleum re fineries,:, have been
required to control sour water stripper gases in refinery sulfur
plants. These gases usually contain only 30-40 percent H2S and
similar concentrations of ammonia. In some refineries the total
volume of these gases can be about the same as the total volume
of H2S gases produced by the amine treating units. If any of these
gases are mixed with those from the amine treating units, the
concentration of H2$ in the gases processed by the sulfur plant
is lowered. In addition, since the ammonia present in the sour
5.3
-------
water stripper gases results in the formation of ammonia sulfates
in a Glaus plant which can lead to fouling of the Glaus catalyst,
the ammonia has to be removed by incineration of the sour water
stripper gases. This contributes to even greater dilution of the
gases processed by the sulfur plant because of the fuel and air
required for incineration.
Consequently, if sour water stripper gases are mixed with the
H£S gases prpduced by the amine treating units in a refinery to any
appreciable extent and fed to an existing sulfur recovery plant,
emissions from the sulfur plant would increase as a result of both
the higher sulfur throughput and the resulting decrease in sulfur
recovery efficiency. Whether or not this would be termed a
modification would depend on the circumstances involved. If prior
to introduction of the sour water stripper gases into the sulfur
plant they were incinerated and vented to the atmosphere, S02
emissions from the refinery as a whole would decrease. Under the
general provisions of 40 CFR Part 60, in this situation the sulfur
Plant would not be considered as a modified source even though its
emissions would have increased.
If, however, these sour water stripper gases had not been
incinerated and vented to the atmosphere prior to their introduction
into the sulfur recovery plant, this might be considered a modification,
If the sulfur recovery plant required alterations to accommodate these
gases, such as changes to the Glaus furnace which required a capita^
5.4
-------
expenditure to carry out, then this would be considered a modification
and the sulfur recovery plant would be subject to compliance with
standards of performance. If,oon the other hand, the sulfur recovery
plant required no alteration to accommodate these gases, the
decision as to whether this was a modification would hinge on the
question..."Is this a change in the method of operation?" (i.e.,
was the sulfur plant originally designed to process these gases?).
If this was concluded to be a change in the method of operation,
the sulfur recovery plant would be considered a modified source
subject to compliance with standards of performance.
The emission control techniques discussed in Chapter 4 are
as applicable to modified refinery sulfur plants as. they are to new
refinery sulfur plants, since these control techniques are essentially
"add-on" control techniques which operate independently of the
refinery sulfur plant. In addition, the potential modifications
of refinery sulfur plants discussed above would not .significantly
change the characteristics of the tail gas from the sulfur plant and
thus would not prevent the use of any of the control techniques
discussed in Chapter 4. While the Cleanair nrocess might require
a catalyst change in the first stage catalyst reactor of a modified
refinery sulfur plant to reduce formation of COS and CS25 and
the IFP-1 or the Sulfreen process might require the installation
of additional instrumentation and process controls to ensure main-
tenance of the proner H2S/S02 ratio in the tail gas from the sulfur
plant, these requirements are minor and could be readily accommodated
5.5
-------
in a modified refinery sulfur plant. The Wellman-Lord, Beavon or
SCOT processes, however, could be installed as readily on a modified
refinery sulfur plant as on a new sulfur plant. Each of these
emission control techniques, therefore, can be expected to function
as well on modified existing refinery sulfur plants as on new
refinery sulfur plants and emissions could be reduced to the same
levels in both situations.
5.2 RECONSTRUCTION OF REFINERY SULFUR PLANTS
Under the General Provisions of Part 60 40 CFR, existing
sources which are reconstructed may be considered as new sources
subject to compliance with standards of performance, regardless
of whether emissions increase or not. An existing source will be
considered reconstructed, however, only if most of the separate
components of the source are replaced.
An incremental expansion of a refinery sulfur plant,
therefore, is more likely to be considered a potential modification than
a reconstruction. In those cases where the desired increase in
sulfur plant capacity would be sufficient to necessitate replacement
of most of the separate components of an existing sulfur plant,
it would orobably be more economical to maintain the capacity of the
existing plant and construct a new plant to provide the additional
capacity as discussed above.
Construction of an additional catalyst/sulfur condenser
stage to an existing refinery sulfur plant to increase efficiency
would not be considered reconstruction. If sulfur capacity were
increased at the same time with a resulting increase in emissions,
however, this would probably be considered a modification.
5.6
-------
As with modified refinery sulfur plants, each of the emission control
techniques discussed in Chapter 4 are as applicable to reconstructed
sulfur plants as to new sulfur plants. Consequently, emissions could
be reduced to the same levels through the use of these control
techniques whether the refinery sulfur plants were new or reconstructed.
5.7
-------
-------
6. EMISSION CONTROL SYSTEMS
Based on the discussion presented tn the Dreceding chanters, two
alternative emission control systems emerge as possible candidates
to serve as the basis for standards of nerformance (i.e., the best
system of emission'reduction). These systems are the low temperature
Glaus reactor svstem and the tail gas scrubbing system.
6.1 LOW TEMPERATURE CLAUS REACTOR SYSTEM
As,indicated previously, most sulfur recovery nlants are two- or
three-stage Claus units which generally achieve a sulfur recovery
efficiency of about 95%. A low temnerature Claus reactor system
may be added to achieve an improvement in overall sulfur recovery.
The low temnerature Claus catalyst reaction system is much more
comnlex than a conventional Claus reaction stage and it increases
overall sulfur recovery efficiency to about 99%. Emissions are
reduced bv about 80 to 85% over those from conventional Claus nlants.
Like the conventional Claus, the emissions from the low-temperature
svstem consist essentially of S02. .Incineration of the tail oases
prior to release to the atmosnhere effectively converts any H2S
remaining, and any COS and CS2 formed in the Claus reactors, to S02.
As with the conventional two- or three-stage Claus unit, the
low-temnerature Claus reactor system denends on maintenance of the orooer
H2S/S02 ratio in the gases in the sulfur nlant. Consequently, minor
fluctuations in the gases processed by a Claus sulfur plant that would
not lead to a significant increase in emissions from a conventional
6.1
-------
two- or three-stage unit are likely to result in an appreciable increase
in emissions from a low-temperature Claus reactor system. The problem
of maintaining a 99% overall sulfur removal efficiency, therefore, is
more difficult than maintaining a 95% overall sulfur removal
efficiency. Hence, emissions from a low-temperature Claus reactor system
are likely to exhibit considerably more fluctuation than those from a
two- or three-staae conventional Claus unit, although the absolute
magnitude of emissions would be much lower.
6.2 TAIL GAS SCRUBBING SYSTEM
As discussed in Chapter 4, two different types of tail gas
scrubbing systems may be used to reduce emissions from Claus sulfur
plants-reduction/scrubbing systems or oxidation/scrubbing systems.
Either type of system increases overall sulfur recovery to about
99.9 percent, thereby reducing emissions from a conventional two- or three-
stage Claus sulfur plant by about 98-99 percent.
The Beavon, Cleanair and SCOT processes discussed in Chapter 4
are representative of the reduction/scrubbing systems. In these
systems the tail gases from the final sulfur condenser of a conven-
tional Claus sulfur plant are reduced through the use of a reduction
catalyst. Essentially all the S02 and about 80-85 percent of the
COS and CS2 is converted to H£S. In the Beavon and Cleanair processes
the H2S is then absorbed in.a scrubbing solution and converted
directly to sulfur via the Stretford process (see Chapters 3 and 4).
6.2
-------
In the SCOT process the H2S is absorbed in a scrubbing solution,
then desorbed and recycled back to the Glaus sulfur plant.
The Wellman-Lord and IFP-2 processes are representative of
the oxidation/scrubbing systems. In these systems the tail gases
from an incinerator following the final sulfur condenser of a con-
ventional Claus sulfur plant are absorbed in an S02 scrubbing system.
The S02 is then regenerated and in the Wellman-Lord process recycled
to the Claus sulfur plant. In the IFP-2 process the S02 is mixed
.with a small bypass gas stream from the H2S gases processed by
the sulfur plant and then sent to a low temperature Claus catalyst
reactor. The off-gases from the low temperature Claus reactor
are then recycled to the Claus plant incinerator.
Emissions from the oxidation/scrubbing systems are essentially
S02. Incineration of the tail gases from the Claus sulfur plant
prior to entering the S02 scrubbing system effectively converts
any H2S, COS and CS2 to S02. Emissions from the reduction/scrubbing
systems, however, can be either S02 or a mixture or H2S, COS and
CS2, depending on whether the tail gases from the scrubbing system
are incinerated before discharge to the atmosphere. Reduction of
the tail gases from the final sulfur condenser of a Claus sulfur
plant effectively converts all the S02 to H2S, but only about 80-85
percent of the COS arid CS2 present is converted to H2S. The COS
and CS2 remaining is not absorbed in the H2S scrubbing system and
is, therefore, released to the atmosphere in the tail gases as COS
and CS2 along with some H2S, unless these gases are first incinerated,
6.3
7.1
-------
Incineration effectively converts the COS, CS2 and HoS present
water pollution, solid waste generation and energy consumption also
need to be assessed in the same manner.
As discussed in Chanter 3, most SIP's require new refinery
sulfur plants to achieve an overall sulfur recovery of 99 percent.
This requires a level of emission control equivalent to emission
control alternative I. A few SIP's, however, require an overall
sulfur recovery of only 96 percent. This level of sulfur recovery
can be achieved by conventional three-stage Glaus plants which,
from the point of view of this document, are considered to represent
uncontrolled plants. Similarly, a few local air pollution control
agencies require an overall sulfur recovery of 99.9 percent. This
is equivalent to emission control alternative II. Finally, a number
of states which contain no refineries or oil and gas production facilities
located within their borders have no air pollution regulations limiting
emissions from refinery sulfur plants.
7.1 Ambient Air Quality Impact
The health and welfare effects of S02 have been well documented.
The health and welfare effects of H2S, COS and CS2, however, are
not as well documented as those of S02. Table 7.1 summarizes
the major health and welfare effects known to result from exposure
to var-ious levels of these pollutants. Of the three (H2S, COS,
and CS2), H2S appears to be the most harmful. It can lead to
death at concentrations exceeding 1,500,000 yg/m3 while continuous
exposure to concentrations as low as 15,000 yg/m3 leads to
1
7.2
-------
1
6
ia
£
o
S
5
-S
0)
>~.
o
3
T3
t
*r •
0
c
o.
c:
1
c
V>*
^
CO
0}
-CL
(a
g
-*J
4-
fO
E
0?
3
S-
ai
<£i
-.
i
c
C Q
O to-
U I-
1
•*-»
I
5! SS
UJ £
-
roco
>(-£
5
ca
S
I
g
g
. n>
OJ
ss
g
cT
s
7.3
-------
symptoms such as conjunctivitis, sleeplessness, and pain in the eyes.
For occuoational exposure the American Conference of Governmental
Industrial Hygienists (ACGIH) has set a 4-hour-per-week exposure
limit of 15,000 ig/m3. The effects of H2$ on domestic animals parallel
those on humans, while appreciable damage to plants may occur at lower
concentrations. HgS will also cause significant damage to materials,
especially metallic and painted surfaces. Metals, such as silver or copper,
will be tarnished when exposed to levels as low as 4 ifl/m3.2
Experience with the use of CS2 in the textile and rubber manu-
facturing industries, and to a lesser degree from toxicological studies,
has shown it to be highly toxic at concentrations approaching 1,000,000
Workers exposed to lesser concentrations over a number of years have
experienced a variety of symptoms, including mental illness, coronary
heart disease, high blood pressure, and changes in normal reproductive
processes in women (i.e. disruption of the menstrual cycle and
pathological changes in the cellular composition of vaginal smears).
Some of these effects have been observed with industrial exposures
as low as 14,00n Pq/m3. The ACGIH limit for occupational exposure
to CS2, however, has been established at 67,857 vg/m3. It should
be noted that several countries have established lower occupational
exposure limits, including the USSR where it is set at 4553 yq/m3.3'4
Comparatively little information is available relative to the
effects of COS. It appears, however, that this compound is less
toxic than H£S or C$2- One study usinq laboratory rabbits indicates
the lethal dosage is about twice that of CS2.4
To assess the environmental impact associated with standards of
Performance for refinery sulfur plants, the maximum impact on ambient
7.4
-------
air quality of emissions from sulfur recovery plants was analyzed.
The dispersion model used for estimating maximum ambient air concen-
trations for averaging times ranging from one hour to one year was the
Single Source (CRSTER) Model developed by EPA's Meteorology and
Assessment Division, Concentration estimates were extrapolated from
the Single Source Model estimates for averaging times less than one hour.
The meteorological input to the Single Source Model consists of one
year of hourly stability-wind-temperature data from Houston and
Chicago, two major petroleum refining cities. The emission source
model is a 100 LT/D Claus plant with a stack height of 50 meters,
which would likely accompany a refinery capacity of about 100,000 BBL/day.
In addition to assuming that emissions are from a single stack,
it was also .assumed that fugitive emissions do not exist, and adverse
downwash phenomena (which sometimes occur in the lee of stacks or
nearby structures) do not occur. If fugitive emissions or downwash
problems existed, a special dispersion analysis would be required to
estimate the resultant air quality impact. It should be noted in passing
that such problems can result in ambient concentrations several times
greater than those predicted in this analysis.
Maximum ambient air concentrations (see Table 7.2) were estimated
to occur from about 0.6 to 1.6 kilometers from the'sulfur recovery
plant. The 24-hour and 8-hour maxima will most likely occur on days
when there are several hours during which the wind is from a single
direction at about 3-7 meters per second and during which a neutral
or near-neutral atmospheric stability condition exists. The one-hour
7.5
-------
s_
£T
co
(O
o
03
-P
C
OJ
o
c
o
o
OJ
o
D_
•»->
C
co
CO
.E
5
CJ
S-
CU
i
X
ro
T3
CU
LO
10
r--
LO
•*
CO
O
oc
tc. oo
co i—
LO,
CM
co
CM CD tn
CO CO)
i
CM
LO o cr
CM oo, •*
CO LO
o s-
H-> £=
00 01
O
c
a
e.1
to
CM
cn
c
CM
O O LO
O CTi
CO r—
CU
O
S-
dJ
c..
LO,
as
CD
O
O
CU
(C
S-
cu
>•
CD'
to
o.
s-
.
CU
CO.
r-—
-t^
E
to
•!->
3
^~
^—
O
D,
cu
to
CO
CJ
CM
cu o
tor CO
res
CG
CU
5~
•i—
H-^
rci
E
S- CM
CU O
•»-> CO
5
CU- E E
> O O
•1 •! •!—
03 03 CMU
E T3 O 3
S- •!- CO T3
-------
maxima will occur during unstable or slightly unstable atmospheric
conditions, accompanied by light wind speeds on the order of 1 to 2
meters.per second. The downwind distance to the one-hour maxima will
range from about 0.4 to 1*6 kilometers.
For averaging times less than one Jhour3 the peak concentrations
will most likely occur under light winds and unstable atmospheric
conditions. The peak concentral!ows listed iia Table 7,2 for averaging
times of less than one incur will occur approximately 0,5 kilometers from
the source.
The predicted maximum ground-level 24-hour and 1-year ambient air
concentrations of '502 from a typical 'uncontrolled IQfD LT/day sulfur
plant are 175 ifl/m3 and 15 is/m3, respectively. These levels are
well below the correspondiing national ambient air quality standards
-------
It appears from Table 7.2 that emission control alternative I
(low temperature extended Glaus reaction) reduces the maximum 24-hour
ambient air S02 concentration by a factor of 7 and the maximum
one_year ambient air S02 concentration by a factor of 3. Alternative II
(tail gas scrubbing), on the other hand, reduces these maximum ambient
air concentrations to essentially zero for all practical purposes,
assuming an oxidation tail gas scrubbing system is employed. If a reduction
tail gas scrubbing system is employed, emissions of S02 are eliminated.
Use of a reduction system, however, leads to emissions of H2S, COS and
C$2 which results in low ambient air concentrations of these pollutants.
Specifically, alternative I reduces the maximum 24-hour ambient air
S02 concentration to 25 yg/m3 and the maximum one-year ambient air
SOo concentration to 5 ig/m3. If an oxidation system is employed,
alternative II reduces the maximum 24-hour and one-year ambient air
S02 concentrations to 4 yg/m3 and <1 yg/m3, respectively.
If a reduction system is employed, alternative II leads to maximum
2-10 second ambient air concentrations of F^S, COS and CS2 of
25, 380 and 540 ;g/m3 and maximum one-hour ambient air concentrations
of 1, 7 and 10 yg/m3, respectively. Although the maximum 2-10 second
ambient air concentrations of.H2S, COS and CS2 could lead to odor problems
or material corrosion problems if they persisted for longer periods
(one to two hours), the ambient air quality modeling indicates these
concentrations are only of short duration. As shown in Table 7.2,
beyond 2-10 seconds the maximum ambient air concentrations of H2S,
COS and CS2 decrease rapidly and fall well below the threshold limits
for odor or material corrosion problems.
7.8
-------
Under normal operation, emissions from the reduction emission
control systems are no higher than 10 ppm H2S and 95-100 ppm for both
COS and CS2 and-, as discussed above, the resulting ambient air
concentrations of these pollutants will be low enough to ensure
that no adverse health or welfare effects arise. Emissions of COS
and C$2, however, may increase considerably, approaching levels as
high as 1000 ppm, if the catalysts used in the reduction systems are
permitted to deteriorate.3 However, even at this level these
emissions are not likely to result in ambient air concentrations
sufficiently high to pose adverse health or welfare effects, based
on our present knowledge.
As noted earlier, most SIP's require an overall sulfur recovery
equivalent to emission control alternative I. Consequently, standards
based on alternative I would have little or no impact on ambient
air quality. Only if standards are based on emission control
alternative II will there be a beneficial impact on ambient air
quality. Based on the growth projections presented in Chapter 8,
by 1980 some 8100 LT/day of refinery sulfur plant capacity will be
subject to NSPS. Nationwide, therefore,.the beneficial impact on
ambient air quality of standards in 1980, based on emission control
alternative II, would be to reduce emissions by some 55,000 tons
per year of S02-
7.2 Water Pollution Impact
Petroleum refineries normally discharge large volumes of
waste water and can be major point sources of water pollution.
The amount of waste water discharged varies greatly from refinery
7.9
-------
to refinery, however, and depends on both the types of orocess
units Within the refinery and the degree of water reuse. The
notential water pollution impact of a typical 100,000 bbl/day
refinery is summarized in Table 7.3. As this table shows, even
with the use of best practical control technology to limit water
pollutant discharges, this impact can be significant.
Petroleum refinery Claus sulfur plants by comparison aenerate
an extremely small waste water stream. This stream results from
condensation of water vaoor contained in the H2S gases as they
flow from the amine scrubbing units to the Claus sulfur plant.
Although this sour water condensate normally contains 1500-2000 ppm
H£S and UD to 1000 ppm ammonia, the volume of water involved is
usually less than 0.25 liter per minute, and this is easily
handled in the refinery's waste water treatment facilities. In
terms of the examole presented above, this is about 0.25 m3/day,
or less than 0.002 percent of the total waste water effluent routinely
discharged by a moderate size refinery.
The notential water pollution impact of the two
alternative emission control systems outlined in Chapter 6 is
negligible. Table 7.4 summarizes the characteristics and flow rates
of the various waste water streams discharged by each of these
systems. As this table shows, althouah the volume of waste water
discharged by some of these emission control processes is larger
than that discharged by the Claus sulfur nlant, even in the worst
case this is less than 50 liters per minute, which is less than
7,10
-------
Table 7.3
Water Pollution Impact of a Typical
100,000 bbl/da.y Petroleum Refinery'
Effluent Flowrate, m3/day
Pollutant Load, kg/day
COD
BQ%
TQC
TSS
Oil
Phenolics
Ammoni a
Sulfides
13,500
Raw Water
BPCT2
1800
4500
1400
700
700
150
450
20
200
1100
350
150
70
1
100
1
Notes:
1. Development Document for Proposed Effluent Limitations
Guidelines and New Source Performance Standards for the
Petroleum Refining Point Source Category, Effluent Guidelines
Division, Office of Air and Water Programs, Environmental
Protection Agency, December 1973, EPA/440/1-73/014.
2. Best practical control technology.
7.11
-------
0.2:
ol
O Q
So
§1
in
C CO
01 co
'o'io s-
CO -t-> OJ
CU +J
OE: to
tOt—fcft
2tO«3-
-*sr-
55
tn
CM
to i-
co cu
,
o in
CO l-~
CO
.—
CO CO >
O O
.
s- oco
C- I— COOl
in
cu a)
E E
(C (O
CO CO
O
cu
E
1^
CO
o-
13 CO
-3.0.
*CO
u
in
5
4J
c
3.
c:
•r-
S-
ra
£
0)
o
c
o
m
co
CM
C
•r-
S- 5-
a> a>
CD in
CM i—
w- CVl
•
s- v
re
•<->
re
o
>
a> i—
+j to
CO 4J
o
CO
ID
•4->
CO
S-
0)
O
CO
5!
cu
ai o-.
4-> S-
CO 3
tn CL.
c
cu c:
C •!-
o 4->
O 3
s-'o
a> co
to cr.
3 c:
03
CO S-
o
CO
cu
4J
re
O
O
01
to _i
CJ
c
~
.£!
J3
"
O
—i
+J
to to i
S-
0)
CM
I
Q_
u
CD
t-
a>
eul01
+->!
of -
7,12
-------
0.5 percent of the total waste water effluent routinely discharged
by a tynical 100,000 bbl/day refinery.
The ootential water pollution impact associated with alternative I
emission control systems (low temperature extended Glaus.reaction)
is negligible. Although this alternative does not generate additional
waste water streams continuously, intermittent waste water streams
are generated. After about two years of operation, washing of the
catalyst is required to restore catalyst activity. As shown in
Table 7.4, some 28,000 liters of water are required and, when discharged,
this water contains about 1 percent by weight of the organic solvent
for the Claus catalyst (polyalkyene glycol), and about 20-25 percent
by weight of the Claus catalyst itself (alkali metal salts). This
waste water stream, however, can be easily treated in the refinery
waste water treatment facilities to prevent any adverse water
pollution impact from arising.
The potential water pollution impact associated with alternative II
emission control svstems (tail-gas scrubbing), although slightly
greater than that associated with alternative I systems, is also
negligible for all practical purposes. Generally, the waste water •
streams generated by the various tail-gas scrubbing processes consist
of a sour water condensate, similar to that generated by the Claus
sulfur plant, and a purge stream containing either organic or inorganic
salts. The amount and composition of these waste water streams
varies depending on the particular tail-gas scrubbing process used.
The sour water condensate is produced by cooling of the gases prior
7.13
-------
to the scrubbing tower, and the purge stream is necessary in most
cases to prevent a build-up of impurities in the scrubbing solutions.
As shown in Table 7.4, the amount of sour water condensate
generated by these tail gas scrubbing processes is in the range of
20 to 45 liters per minute for a 100 LT/day Claus sulfur plant.
This stream is usually acidic in nature with a pH of about 2 if an
oxidation/scrubbing emission control system is employed. If a
reduction/scrubbing system is used, this stream is usually only
slightly acidic, containing about 50 ppm H2S and a trace of ammonia.
These streams, however, can be added to the sour water condensate
stream discharged by the Claus sulfur olant for treatment in the
refinery's waste water treatment facilities.
For a 100 LT/day Claus sulfur plant, the purge stream from
an oxidation/scrubbing emission control system» such as the Wellman-
Lord process, is in the range of 15 liters oer minute. This
stream consists essentially of a solution of sodium salts
in water and generally has the following composition:5
Na2S04 5 percent by weight
NaHS03 5 percent by weight
Na2S03 15 percent by weight
H20 75 percent by weight
This stream could be treated in the refinery waste water
treatment system without difficulty, or it could be treated by
7.14
-------
the NICE process developed by Nittetu Chemical Engineerinq Ltd.
The NICE process recovers the sodium as sodium carbonate (Na2C03)
or soda ash, which can be used to provide the necessary make-up
sodium sulfite/sodium bisulfite solution to the Wellman-Lord process.
Davy Power Gas, the developer of the Wellman-Lord process, is also
developing a treatment process for this purge stream. Consequently,
disposal of this waste water stream will not lead to any adverse
water pollution impact.
Unlike the Wellman-Lord process, the other oxidation/scrubbing
emission control system, the IFP-2 process, has no purge stream. This
process Is essentially an upgraded alternative I process and consists
of a third-stage low temperature Claus reactor followed by an ammonia
scrubber. The only waste water streams discharged are the sour water
condensate stream which is common to all the tail-gas scrubbing processes,
and an intermittent waste water stream resulting from washing of the
low temperature Claus catalyst every two years as mentioned earlier.
Thus, there is essentially no potential water pollution impact
associated with use of the IFP-2 process.
The purge stream from a reduction/scrubbing emission control
system, such as the Beavon or Cleanair process, installed on a
100 LT/day Claus sulfur plant is in the range of 5 liters
per minute. As with the purge stream from the Wellman-Lord process,
this stream consists essentially of a solution of sodium salts
Q
in water and generally has the following composition:
7.15
-------
Na2C03
Na2S203
NaSCN
Misc. Sodium Salts
0.5 percent by weight
0.5 percent by weight
0.5 percent by weight
0.5 percent by weight
98.0 percent by weight
This stream could also be treated in the refinery
waste water treatment system without difficulty, or treated by the
NICE process to recover the sodium as sodium carbonate. The sodium
carbonate could be used to provide the necessary make-up solution
to the Stretford scrubbing unit of the Beavon or Cleanair processes.
The potential water pollution impact associated with the use of
the Beavon or Cleanair processes, therefore, is negligible.
The SCOT process, which is also a reduction/scrubbing emission
control system, does not generate a purge stream. As with the
IFP-2 process mentioned above, the only waste water stream generated
by this process is the sour water condensate stream, which can
be added to the sour water condensate stream generated by the
Claus sulfur plant for treatment in the refinery's waste water
treatment facilities. Thus, use of the SCOT process would result
in essentially no water pollution impact.
In conclusion, all the waste water streams generated by the
two alternative emission control systems outlined in Chapter 6
can be treated without difficulty in the refinery's waste water
treatment facilities. Considering the small amount of waste
water discharged by these emission control systems, the potential
water pollution impact of NSPS for refinery sulfur plants will be
negligible, whether they are based on alternative I or alternative II
emission control systems, because of the magnitude of dilution in the
refinery's waste water treatment facilities. Also, since these waste
7.16
-------
water streams are so small, they will have no impact on the ability
of petroleum refineries to meet water quality effluent regulations.
7.3 Solid Waste Impact
There is essentially no potential solid waste impact associated
with refinery sulfur plant standards based on either alternative I or
alternative II emission control systems.
The Claus process itself requires periodic replacement of the
reaction catalysts, the frequency of replacement depending upon the
impurities present in the acid gas feed. Usually the catalyst,
made of bauxite or alumina, is regenerated annually until a substantial
g
loss of activity occurs, normally in two to five years. For a 100 LT/D
Claus plant, replacement of the catalyst would require disposal of approxi-
mately 70 tons of spent catalyst, usually in landfills. Alumina (aluminum
oxide) and bauxite (a natural mixture of iron, aluminum and manganese
oxides/hydroxides) are both non-toxic materials. Neither emission
control system alternative will affect the rate or quantity of
catalyst replacement in the Claus plant.
As for the emission control systems themselves, neither the
alternative I systems ("low temperature extended Claus reaction) nor
the oxidation tail gas scrubbing systems (alternative II) generate
any .solid waste. The reduction tail gas scrubbing systems, however,
do require periodic replacement of the reduction catalysts about
every two years.10 These catalysts usually have significant salvage
value, being composed primarily of colbalt-molybdenum, and consequently,
are normally returned to a vendor for reprocessing. Hence, even the
reduction tail gas scrubbing systems generate essentially no solid
waste.
7.17
-------
7.4 Energy Impact
Although not generally recognized, petroleum refineries consume
a significant amount of energy in processing crude oil into various
petroleum products such as petrochemical feedstocks, gasolines,
and fuel oils, etc. The energy requirements of a typical moderate
or high conversion refinery, for example, usually represent about
10 percent of the crude oil throughput.11 Thus, the energy consumption
of a nominal 100,000 bbl/day refinery is equivalent to about
10,000 bbl/day of fuel oil, or some 250,000 kw-hr/tir.
The energy requirements of refinery sulfur plants are quite
small in comparison. A 100 LT/day Glaus sulfur plant, for example,
typically consumes less than 1000 kw-hr/hr of energy,12 or less
than 0.5 percent of the energy consumed within the petroleum refinery
itself. Consequently, the use of Claus sulfur plants to control
emissions of sulfur dioxide or hydrogen sulfide at petroleum
refineries does not significantly increase the energy requirements
associated with petroleum refining.
The energy impact associated with each of the alternative
emission control systems outlined in Chapter 6 is summarized in
Table 7.5. As this table shows, the impact is negligible in all
cases. Alternative I, for example, has a slight energy benefit
(i.e. overall energy consumption is reduced somewhat) due to reduced
tail gas incineration requirements. Alternative II, on the other
hand, has a slight energy nenalty or a moderate energy benefit,
denending on whether an oxidation or reduction tail gas scrubbing
emission control system is emoloyed and on whether tail gas reheat
7.18
-------
fs required to increase plume bouyancy. A few local air pollution
control agencies require that gases discharged into the atmosphere
must be above a certain minimum temperature to insure good plume
bouyancy and thus good plume dispersion. Under these conditions,
Alternative II has a slight energy penalty. Most local air pollution
control agencies, however, do not have requirements of this nature
and in most cases tail gas reheat is not necessary to obtain good
plume dispersion. Under these conditions, Alternative II has
either a slight energy penalty or a moderate energy benefit,
depending on whether an oxidation system or a reduction system is
employed. The moderate energy benefit associated with the reduction
tail gas scrubbing system arises because of reduced tail gas incineration
requirements.
As mentioned earlier in this chapter, most SIP's require new
refinery sulfur plants to achieve an overall sulfur recovery of 99 percent.
This requires a level of emission control equivalent to alternative I.
Consequently, if alternative I is selected as the basis for refinery
sulfur plant NSPS, there will be no energy impact associated with the
standards.
If NSPS are based on alternative II, there will be either
a slight energy penalty or'a moderate energy benefit associated
with the standards in all but those few localities which already
require this type of emission control system. This impact will
vary from refinery to refinery depending on whether an oxidation
or a reduction tail gas scrubbing system is employed to comply with
the NSPS. As shown in Table 7.5, use of an oxidation tail gas
7.19
-------
O cr
« c
.£) -r-
s-
ai u
>
CO "
CD" CM
LO i —
en •>
01
S-
cu
cu
cu
Q,
Q.
•r—
3
d'
CU
CU
cu
S-
0
cu
c:
o
c:
o
u.
c
o
is
UJ
0)
— S- in
t-i =j 3
•«-> m
co to i — c
> i- o o
•I- O) T-
•M O.T3 4->
S -4-1
i— O X
o o
i— s- c:
CU CL. I-H
i- CU
S-
3
4-5 S-
•r- S- CM
S- J=
4-5
i— S-
a.
to
t/l V)
a) in
OJ
E
(O
CU
:
-------
scrubbing system without tail gas reheat increases the overall
energy consumption of a Glaus sulfur plant by about 17 percent.
Use of a reduction tail gas scrubbing system without tail gas
reheat, however, reduces the overall energy consumption by about
50 percent.
Based on the growth projections presented in chapter 8, by 1980
some 8100 LT/day of refinery sulfur plant capacity will be subject
to NSPS. Standards based on alternative I will have essentially
no impact on national energy consumption, since most SIP's already
require the use of this type of emission control system. Standards
based on alternative II, however, will reduce national energy
consumption by some 54 million kw-hr/yr, or about 90,000 barrels
of fuel oil per year, assuming half of the refinery sulfur plant
capacity subject to compliance with NSPS install oxidation tail
gas scrubbing systems without tail aas reheat and half install
reduction tail gas scrubbing systems without tail gas reheat.
7.5 Other Environmental Impacts
No environmental impacts other than those discussed above
are likely to arise from standards of performance for refinery
sulfur plants, regardless of which alternative emission control
system is selected as the basis for standards. Furthermore, other
than those resources initially required to construct either alternative
emission control system (most of which could probably be salvaged
in one way or another), there do not appear to be any irreversible or
irretrievable commitment of resources associated with these standards.
As discussed above, there is even no overall increase in the energy
requirements associated with refinery sulfur plants, since both
7.21
-------
Table 7.6
Environmental Impact of No Standards or
Delayed Standards
Sulfur Plant Canacity
Nationwide S02 Emissions (M Tons/Yr)
Year Affected by Standards' No SIP*
1976
1977
1978
1979
1980
TOTAL
3150
1425
1815
850
850
8090
130
60
75
35
35
335
SIPJ Alternative IJ
25
10
15
5
5
60
25
10
15
5
5
60
Alternative II'
2
1
1
0.5
C.5
5
Note
1. LT/day.
2. 95% control.
3. 99% control.
4. 99.9% control.
7.22
-------
emission control systems result in a net reduction in energy
consumption.
Based on the growth projections presented in Chapter 8, the adverse
environmental impact of no standards or delayed standards on nationwide
S02 emissions is summarized in Table 7.6. If alternative I is
selected as the basis for standards, there is no adverse impact on air
quality since alternative I does not reduce emissions beyond levels
currently required by most SIP's. If alternative II is selected
as the basis for standards, on the other hand, the adverse environmental
impact of delaying standards or not setting standards is an increase
in nationwide S02 emissions, of some 5,000 to 25:,000 tons per year,
reaching a total of about 55,000 tons per year by 1980.
Since there are essentially no adverse water pollution, solid
waste disposal or energy consumption impacts associated with either
of the alternative emission control systems which could serve as
the basis for standards, there is no "trade-off" of potentially
adverse impacts in these areas against the resulting adverse
impact on air quality of delaying standards or not setting standards.
Furthermore, there does not appear to be any emerging emission
control technology on the horizon that could achieve greater emission
reductions or result in lower costs than that represented by
the emission control alternatives under consideration here. Con-
sequently, delaying standards to allow further technical developments
appears to present no "trade-off" of higher SOg emissions in the
near future against lower S02 emissions in the distant future.
7.23
-------
References
1. "Air Quality Criteria for Sulfur Oxides," U.S. Dept. of Health,
Education and Welfare, Public Health Service, January 1969.
2. "Preliminary Air Pollution Survey of Hydrogen Sulfide,"
U.S. Dept. of Health, Education and Welfare, Public Health
Service, October 1969.
3. Peyton, Thomas 0., Steele, Robert V. and Mabey, William R.,
"Carbon Disulfide, Carbonvl Sulfide, Literature Review and
Environmental Assessment,'" EPA Contract 68-01-2940, Task 23,
July 1975.
4. Patty, Frank A., Industrial Hygiene and Toxicology, Vol. 2,
2nd Ed., Interscience, New York, 1963.
5. Genco, J.M., and Tarn, S.S., "Characterization of Sulfur Recovery
from Refinery Fuel Gas*11 EPA Contract 68-02-0611, June 1974, p. 54.
6. Reference 5, pp. 35, 37-38.
7. Reference 5, p. 6n.
8. Reference 5, p. 36.
9. Letter, Cole, D.F., J.F. Pritchard and Co. to W.D. Beers,
Process Research, Inc. dated June 5, 1972.
10. Reference 5, pp. 52, 57 and 64.
11. "Impact of Motor Gasoline Lead Additive Regulations on
Petroleum Refineries and Energy Resources - 1974-1980 Phase 1,"
EPA Contract 68-02-1332, Task"4, May 1974, p. V-38.
12. Beers, W.D., "Characterization of Claus Plant Emissions,"
EPA Contract 68-02-0242, Task 2, Anril 1973.
°f C°a1 Gasification Emission Control
7.24
-------
8. ECONOMIC IMPACT
8.1 INDUSTRY PROFILE
As of April 1975, there were 259 petroleuifi refineries in the United
States with a total capacity of 14.8 million barrels per calendar day
(BCD). These refineries ranged in size from 200 BCD to 445,000 BCD,
with an average size of approximately 57,300 BCD.1 In general, new
refineries are expected to be considerably larger than the current
industry average. Information is available on 12 refineries that have
been projected to be built after January 1, 1975. These refineries vary
in size from 5,000 BCD to 200,000 BCD, with an average size of approximately
97,000 BCD.2
Not all petroleum refineries currently include sulfur recovery
plants. At the 259 domestic refineries referred to above, only 81 included
sulfur recovery plants. There are currently 122 sulfur recovery plants
within the industry, either currently installed or due to be installed
in 1975, ranging in size from 4 long tons of. sulfur per day (LTD) to 375
LTD, with an average size of 72 LTD.1'2'3 Sulfur recovery plants tend
to be found in the larger refineries. Table 8-1 illustrates this point.
For example, there are 175 refineries with individual capacities less
than 50,000 BCD, amounting to 67 percent of the total number of 259
refineries. These refineries, however, account for only 16 percent of.
t'he total number of 122 sulfur recovery plants. Stated in another way,
the average size of a refinery which includes a sulfur recovery plant is
approximately 140,000 BCD. This compares to an average size of approximately
27,000 BCD for those refineries without sulfur recovery plants.
8-1
-------
en
•»
CM
CO
LU
LU
LU
an
to
4J S_
o cu
o o
o
S- OJ 4->
cu ex c
•§*,,£
CO
oo
CO
co
c\j
o
CvJ
CO
CVJ
CU
LO
- t
O i.
O CX
o 2:
in
oo
CO
CM
o
o
CO
LU
o
o
rs
CO
O-r-
S-cB
cu E
H'C
=3 O)
LO
CD
CM
CO
CM
CM
cn
IO
CM
CO
H-4
O
CO
CD
.a
*«"^»
CO Q
NI O
•r— CO
CO^— •
o
OS E
t— I
CD
CD
CD
LO
1
CD
O
O
CD
A
CD
O
I
r—
CD
O
M
o
LO
o
o
CD
CD
LO
^_
CD
O
M
O
O
r111 •
o
o
o"
o
C\J
CD
CD
0
r~"
0
CD
O
0
LO
CM
1
O
CD
CD
CD
CM
CD
CD
CD
O
O
CO
I
O
CD
O
LO
CM
CD
O
CD
««
O
LO
CO
1
O
CD
A
0
CD
-------
Table 8-2 summarizes the current status of the domestic refining
industry with regard to sulfur recovery plants. Table 8-3 provides
additional, detail with regard to the sulfur recovery plants that are
currently installed or due to be installed in 1975.
8.2 COST OF ALTERNATIVE EMISSION CONTROL SYSTEMS
As outlined in Chapter 6, there are two alternative emission control
systems that could serve as the basis for refinery sulfur plant NSPS.
Alternative I, exemplified by the IFP-1 and the Sulfreen processes,
achieves an overall sulfur recovery of 99.0 percent measured against the
total sulfur in the Claus plant feed gases. Alternative II, exemplified
by the Beavon, SCOT, Wellman-Lord, Cleanair, and IFP-2 processes, achieves
an overall sulfur recovery of 99.9 percent.
Since more data was available, both from vendors as well as owners,
the IFP-1 and the Wellrnan-Lord and the Beavon processes were taken as
representative of the two alternative emission control systems. Costs
would have to be comparable for the other systems or they would not be
competitive in the marketplace. Tables 8-4, 8-5, 8-6, and 8-7 present
the operating costs for the basic Claus sulfur recovery plant, a Claus
plant with an alternative I emission control system (IFP-1), a Claus
plant with an oxidation alternative II emission control system (Wellman-
Lord), and a Claus plant with a reduction alternative II emission control
system (Beavon). The plant with the capacity of 10 long tons of sulfur
per day is believed to be representative of a unit required by a typical
small refinery. The plant with the 100 long ton per day sulfur capacity
is believed to be typical of a unit required by a typical large refinery.
A plant with a capacity of 5 long tons of sulfur per day, while believed
to be toe, snail to be generally utilized in typical refineries, is also
shown for comparison. — o o
~
-------
•r- O
rtS E OQ
O. to
"Sl- CT> CF»
«3- Cn CM
I I I
CM in CJV
•— co CM;
Lf)
I
CM
CO
co
If)
CO
co
CO
t/J
uu
1—4
ce
LLl
O) 4->--^
CD-i- Q
«J O CO
S- «O CO
O) Q.
> fas:
•a; o—•
t^ CM CD
CM CD CM CM
CM
in
LU
M- ( o
O 4-i (O
O £3.
^«|— ftf
CJ
•— oo co
8
I r- Q
ra o o
-l-> (O CO
o ex
i— CO
CM i—
CO
VO
CM
CM
o
10 in
CO CO
*»lj' ' t^J*
i,
i— Q)
<<- <0 J3
O •»-> E
CM
CM
CO
CO
CTi
WD
O
O
CD
LO
CM CM i—
CM| CO
CO
to
CM
CM
CO
0)
§
o
o
a>
a:
OO
t/>
O)
o
O) a;
DC DC
O
CJ
en
in
CTl
en en
en en
to
+->
0
in
O >>
'il aj
cu >
c o
••- o
0)
TO
4->
O
8-4
-------
a >,
s- •*>
O
«=C to
Q.
03
CJ>
i— CVJ CO|
in to cnl
CM
01
LU
t—i
ce
en
o
oo
LU
O
CD
•ZL
I—I
I—
CO
I—I
X
r— O
(O (O
4-i O.
O tO
H- CJ
O =5
13
co
S- Q
fl3 O
•4~> n}
C3 a.
in
i — co
CD
o
CO i—
CM cr>
I— If)
ro
in
n
CO
•a; in
r«.
co cr>
_i r-»
n.
>- ^
of. s-
UJ O.
>
Si O
•^ o
S- 05
CO Oi
CO
o
o
CO
O
o
CO
C£
LT> CT)
CO
O
O
oc
•&s.
01
cr>
(O
4->
O
8-5
-------
CTi
•*
CO
r-^
to
*»
to
o
o
o
CD
1
co
CM
•f}-
o
CD
CM
, —
r—
CD
CD
00
LO
OO
CD
O
OO
,_
*^-
CD
CD
LO
OO
CO
CD
O
"=^
co
CD
C.
CD
CD
CD
O
OO
CO
CD
O
CD
CT>
OO
CD
CD
' —
'CO
•==*-
o
o
en
LO
to
CM
CD
CD
CD
OO
co
r--.
«t
O
O)
tsl
•r—
OO
c
ro
•a:
Q-
>-
UJ
§ ~
o in
uj r^
Q£ O>
CM
"
CD
O
CO
CVJ
•bO-
O
O
CM
-
o
CD
UO
00
CD
CD
•rf-
0
o
CM
1^
Cvl
CD
CD
CO
CD
CD
CD
CD
OO
CO
CD
CD
r^.
CTi
O
O
OO
OO
un
1 —
CD
o
10
ro
CO
•to-
o
o
CD
cvl
o
cr>
-fa*
V)
O)
X
OJ
Q.
O
Q.
OJ
(/J
o>
E
0)
LO
u_
=3 --3
to s—'
CM
CD
CD
CO
CM
CD
CD
CM
, —
1 —
CD
CD
CO
r— •
CD
CD
CM
O
CD
r--.
CM
CM
CD
0
LO
—
o
CD
LO
r—
^
CD
CD
O
cr>
o
CD
CO
CM
o
o
r->
•S*
CD
CD
p^_
LO
-faO-
3
"o.
O)
o
on
OO
o
o
to
in
to
E
O
d.
o
OO
i.
03
O.
i-
cu
CL
E.
ro
o
ro
o
O)
O
*S
CD
to
O)
U-
d-
oo
o
CO
o
i-
O)
o
Q_
s: i—
o
LO
O
to
o
O)
I-U
o
•faO-
C2;
0)
^
OO
<&>
o
•faq-
-o
O)
i-
CJ)
ro
01
00
LO
CM
T3
i.
ti-
to
O
ro
s_
O)
d.
o
0)
(HJ 4->
to CD
CD O
cn
ro
•— -a
ro —'
O E
•— i—
•i- ro
a. +->
ro O
E
OJ
4->
a>
c
01
to
s-
cu
O
CD
LO
00
OJ
o
o
a>
LO
CD
co
to
un
s-
o
4-> X
CJ OJ
ro • T3
cn
LU
s_
cu ~o
> E
O «3
o
OJ •>
s- a>
o
i— E
ro ro
4-> S-
CL 10
ro E
CJ T-
8-6
-------
O
O
CD O O O CDO CD O
O O O O CD O CD CD
CM tf> t— CO to CO to Cxi
to CM
tO CM
•0*
i— CO CM
CO OO to
o
o
"=1-
CO
CD
0
CD
CO
to
CO
CO
o"
a
to
ID
10
to
0
O
CO
O
to
to
o
o
co
to
•fa*
0
o
o
co
CO
1 —
CM
O
o
CD
to
•=*•
O
*
0
0
o
CO
CM
CO
CO
•fa*
CO
00
- 1-1 rs
ex: <_> -o
o
CJ O
a>
IM
CO
-4->
C
1C
ro
CJ
CD CD CD O
C3 CD O O
CM to •* CM
CD CD
O O
i— CO
CD
O
UD CM
tO CM
•fcO-
o
o
CO
o
r— •
o
CD
co
to
CO
CD
CD
o
*3-
CM
CD
CD
CO
*d-
1 —
^—
O
CD
CO
CO
cn
•fa*
0
CD
CD
CM
0
CM
O
O
CD
CO
CM
i —
CD
CD
O
CO
CM
O
r,
c.
Ol
to
(U
Ills
O CD
CD O
r--. to
CD
CD
cn
U3 CM
to CM
-to-
O
CD
•3-
to
o
CD
CO
CO
•=i-
o
CD
CT>
, .
10
CD
O
CD
*3-
>S-
o
o
CD
to
CD
CM
•fa*
O
CD
O
1 —
to
r~-
CD
o
o
CD
cn
CD
CD
CD
«%
t~^
^f
co
•fa*
in
0)
CO CO
OO
to
to
ro
00
o
o
ir>
I
CO
01
s
o
O-
o
o
CD
to-
-O
CO
o
+J
M-
s_ s-
(I) O
ro
S_ _J
D. CU
ro -I—
CD
01
Q.
CO
S.
CD
CO
O
ca
lu
. •
> >
c: c
t—l t-H
^/^ ^^
co to
CO
>^
1 —
+J
ra
CJ
oB
•^
cr
aj
>
i —
o
8-7
cC
CD
E:
CD
ro
•
CD
fa*
C2»
OJ
-)— >
03
to
C
cu
T3
c~
O
CJ
E:
o
to
•
CD
fa*
4J
•p—
-a
QJ
S_
CJ
^
ro
OJ
-4_>
to
s_
CJ
ro
i-
03
Ct.
CD
CHj 4-1
to
to O
aj cj
S1 r-
ro ro
O C
to
3
ro
CJ
O
O
C1J
s-
ro
-I-J
•r-
O-
ro
CJ
ro
O
to
(U
ro
4->
O
ro
s_
O)
O-
o
o
to
CO
cn
CD
O •
<*- "O
O C
«£ -0
>i OJ
1- CJ
aj c
> ro
O S-
tj rs
tu to
s- c:
ro •*
4-> CO
•r- CU
O- X
ro re
CJ -M
-a
c
cn
UJ
-------
CD
O
CM
CM r-
§§l§l§§§§
in
•=»•
CM CT>
CO CO
to
en
CO
0 0
0 0
o co
•=*- CO
in co
^t" i^
CO
o
O
CO
^T
r—
oo
•so-
f
oo
CD CD
00
Oo'cO
00 00
P*N [•»*»
ft *\
CM>—
CD
CD
CD
^
CM
in
*3-
to
to
0)
IM
(O
EC
CO
ZJ
03
CM
0
CD
*3-
CM
4^V
CD
CD
CD
in
o
CD
CO
•^
CD
CD
"^
0 0
CD CD
i — in
CO CM
CM CM
CD CD o" CD
CD CD CD CD
CD CM CO LO
r^ en r**~
~-^ CO
>^-"'
CD
CD
CO
CM
r~~
-tO-
O
CD
CM
CM
O
CD
CM
CM
to
00
•to-
0 0
0 0
O CD
i— en
co **•
CD
CD
O
0
CM
OO
c
cu
to
cu
CM
O
CM
._,
O
O
CD
in
0 0
0 0
CO i—
00 CM
0 0
0 0
en oo
•— CO
CM .—
o
CD
in
oo
0
CD
CD O
0 0
CO 1^-
*^t~ OO
**fc^
0 0
CD O
CO CM
CO CO
in co
r— r— •
•to-
CD
O
CO
•=3-
oo
0 0
CD O
CD CD
CD Ln
oo co
r^ oo
O
CD
CO
in-
CD
cn
3
n-
CU
S-
cu
to
ra
O)
cu
o
o.
CO
s_
03
cu
•£. CO
s_
03
cu
S-
d)
o.
re
-o
en
s=
i.
cu
o
CJ
cu
03
S_
CU
ex
o
CD
LO
00
i-
o
S- o3
O C3
, cu
s- u
CU E
O S-
O
CD fl) CO
O3 •*
-)-> CO
•r— CU
D. X
(^ fO
CJ +J
Seven Selected
Technology and Costs
o .
C_> CTl
CM
C tO
O
•i- CU
4-> a>
3 03
r— O.
I—'
o «
D. CD
S- O
•r- I
*s. oo
O CD
if- ^t-
CO- X
CU UJ C,'
E —
to o >—<
•"-j
•a +-> cr
03 s- E
o LU
i. Q.
o cu
4J o; E
fO
-------
o
CD
CD
O
CM CM
1^
^
CD S
o £
0 C\
LO g
) O
> o
4 UD
1 00
1
<~~> CD
CD CD
CD
LUi—
_ico
0:1—
0)
N
4-> O
IO
CO
O
8888
§ § co cr,
CM CM in <*>
CD O
0 0
CD
CD
10 •
CM '
o"
0
CD
r—
CD
CD
in
CO
CD
CD
CTl
^-
0
o
CD
O
CO
o
CD
0
CM
•tO-
0 0
0 0
o-o
r — cr>
00 00
o
CD
0
in
to
W
a.
o
s-
o.
+->
c
o>
in
0}
I—I—
H-l *>
•soo
CO '—
I o
>- o
Of. LU
LU a:
> —'
8 s:
UJ LU
OL \-
oc. >-
=) to
to or
to •z.
ZD CD
*£. O
DC
O
CD O
O O
^j- CD
O o O
2 CD O
"*• CTl CO
CD
CD
CO
CM CM
•faO-
CD
CD
i£.
o
CD
1 —
CO
O
CD
CM
1 —
CD
O
in
•=3-
CM
CD
CD
O1
O1
?F
00
0 0
0 0
o to
CO CM
0
CD
O
in
in
a>
S-
ta
r-~
3 I
r— CO
"o o
a. in.
r— = LU
o
o
CO
01
03
C
O
Q.
O
o
OO
(U
Q-
S_
CD
S-
O
CD.
SS
CD
ca
Q)
c'
-r—
S-
res
1
CO
o
ra
i —
o
o
>
S-
S= O
OJ Q.
i-
> u
O. 0)
Z3 I—
00 I-U
UD
o
O
c
ro
c
CD
C/l
to
o
£
a)
r-
C_3
c —
t/1
t
to
to
o
c3
s:
CD
co
o
•b^
ca
j_
OJ
ta
•^
, %
tj
o
to
o
in
O
ca
4_3
-^-j
tu
s-
e:
to
o;
oo
_J
in
CM
•bO-
ca
4->
•r—
-o
dJ
S-
CJ
S-
t[
, —
^3
oo
•!->
IO
O
c_>
cr.
c
•r-
4->
(0
0?
c.
o
.4->
O)
2:
K
ca
t/i
O)
en
S-
to
i:
.c_>
"to
4->
•i —
O.
res
CD
to
CD
en
S-
tO
J=.
CJ
r—
to
3
C
C
ec
1 —
to
4J
O
1—
IO
o
CO
3
(O
1 —
o
* — '
4->
c
0>
B
IO
O)
>
c
1 — I
_
c
o
>
to
C
O)
E
• 4->
t/1
QJ
>
s=
r^
ns
-l->
O
\—
. •
IO
O)
o
2:
to
-a
en
•i —
ta
s-
^
j_
O)
>
o
o
ai
s-
r»
fO
>t^
O-
to
^_^
o
a)
o-
c
ia
&.
^3
to
c
•fv
•^
w
OJ
X
to
3
•a
to.
SL.
o
tC'
t-
O)
c
o
cr
1 — 1
ff_^
•o
«
=
«/>
ta
s_
^^
-a
at
4->
o
CO
'aj
oo.
•
en
f~*
LU
•
E
CO
-sz
0
in
i —
^^
to
^
O)
8-9
-------
Tables 8-8, a-9, and 8-10 summarize the economics of installing an
alternative I or alternative II emission control system on a Claus
sulfur recovery plant. For a 100 LTD plant, use of an alternative I
system reduces the annual return from $265,900 to $15,800, a loss of
$250,100 per year. Use of an alternative II system increases the annuaT
costs to $314,600-462,800, depending on whether an oxidation or reduction
process is employed, or a loss of $580,500-728,700 per year. For a 10
LTD Claus plant, use of an alternative I emission control system increases
the annual costs to $198,800, a loss of'$65,200 per year. Use of an
alternative II system increases the annual costs to $352,200-442,000, or
a loss of $218,600-308,400 per year. Finally, for a 5 LTD plant, use of
an alternative I system increases the annual costs to $205,900, a loss
of $58,200 per year, and use of an alternative II system increases the
annual costs to $344,800-419,900, or a loss of $197,100-272,200 per
year.
8.3 ECONOMIC IMPACT
Impact by Company Size
Two financial profiles have beer, developed to evaluate the economic
impact of the two alternative emission control systems. The first
profile represents a large, integrated oil company and the second represents
a small oil company. The profile of the large, integrated company is
biased upon an analysis of the published financial statements of Exxon
Corporation, Mobil Oil Corporation, Shell Oil Company, Phillips Petroleum
Company, Cities Service Company, and Ashland Oil, Incorporated. These
companies considered together should adequately represent the major
integrated refiner sector of the domestic oil industry. The resulting
financial profile iis presented in Table 8-11.
8-10
-------
E
4J 4.J. O
.,- V) -4J
ra o-t/*
CO
O O
to r—
1—
- 3
CO 'O
1 O
O 4->
di
i— -a
^z cu
CD •*-*
C-> c/)
33
^J- -r-j
o -o
i— i *zr
^J 4J <^ >,
E (a a-^
•a: cy — -
Q-
CD
1
4->
4-^ co
tn * — *
cu
E
> — E
'
4->
O
tJ>
1X3
O
| —
o>
Ifr If ^
n3 3
o; t- *--.
E^ CM^
O CO O -^
•i— CO tn
to i — jet
to res tn _l
iVl t
CM
CO
CT>
LXJ ^~
03
en
i — E * — -
O3 •!— -J-3 S-
3 +j in >j
E fO O •~~~
E S- C-5-faq-
•=c cu • — •
0
CD
CD
cn
u->
CM
4^
cu
4-> V»
t/1 **— *
G^
E
- •
O
^^
»«
n
CO
CM
E
O!
4J
tr
cr
r"~
c
•4-
c
c
c_
1 cu
to
CJ
cu
) (0
CO
10
^f
f^^
0
o
(— .
A
0
LO
CM
C
cr
0
to
CD
"
^™*
P^
CC
1—
JD
S"
CD
CO
LO
'
^
l-t
cz:
cr^
cc
CM
CO
fl
CO
1 — 1
cu
•r—
+J
£
cu
^tl
cf.
coco
to- ^0
^" r~~ •
CD O
oo
-J **O
oco
00 <*
CD CD
0 CD
CO CO1
CD CO
«5 l-»
*"
r"— '
CT-. CTV
^— f-™
^-^^-^
CD CD
UD CO
•* CM
i— IO
CO •*
*^~-»^-^»
-Q J3
CD C
^^ cr*
*t M
cviS
LO LO
M »l
•=1- LO
• — 1
1—1
CU E E
> o o
'43 +j +j
03 ro o
E -0 3
S- ••- -0
CU X CU
c—
tn
tn o
• 4-2 O
E tn
•i— O i —
03 O OS
en 4->
tn E
CO CU QJ
CU "^3 ^—
4-> 3 CU
0 r— S-
E U O
CU E E
tot
Qi !
^ > L— ~^ ^""-W *^~*^
O[03 -O O
21 K— ' ^^^--^— *•
s-n
-------
4-> +J O
•r- in 4->
C O
00
CM
CM CO
•a:
cu
u.
cc
O
CJ O1
to
>-
CO
cr
E
-^
cu
o
cu
s-
IX
J-
0)
o
,
re
•»—
+->
cu
cu
tf_
H—
•T~*
o
cu
4J S-
re 3
rv *|
r~~ ^^>
E 3 CMS-
O CO O -E
•r- CO ^,
to r— to
to re to .a
E o <—
UJ J—
1 ^- 1^ «Nls^
r*^» , '
03
J — C
re -r- .— N
3 4-> +3 S-
c: re to >;
«=C CU CJ-t/?.
Q. •
o
c
CD
E '^"^
to»-^
CU
1— 1
0 00
P 00
CM *^J- CM
W CO CO
r— CM
C O C
0 00
O CD O
VC CVJ f^
CM 01 CO
i— CM r^.
2= CU
o +•>
O 10
3
O-5*
CO
CO +*
&"-! £^J
5E O
UJ O*
~ 43
I— to
«=C 0
1C CJ
UJ r—
1— re
«C O
u. «~
CD
»— <
co
>™<
^^
^c
•gr-
O
CJ
•
cu
•4J I.
o±«^
c "3 CM'S^
o oo o ^:
•i— CO "»v.
to r— to
to re to ^i
E O > —
CO
•»-> S-
c: re to >-
c S- o "<
*-^ CU C_3 CO
o
^
E
CU
p:
"^-t/^-
cu"
1— 1
CD
CD
to
ft
CO
CO
CD
CD
O
CM
O
cy>
CT>
re
CD
O
CO
CO
CT>
re
CD
CD
co"
CM
S.
CM CM
re re
"0*0
0 0
CM O
*». A
CM CM
LO -51-
co «a-
(O fO
^^ f**^
00
CDO
CD ID
CM to
A #*
a*
e&
cu
S
re
•feO-
CU
to
re
cu
to
re
CQ
cu
re
s-
cu
CU E E
> O O
re re o
E-O 3
S-T- -o
Q) X CU
•<-> o oi
(O (U
"cx cu
O
0,0,
o
O CM
(U O
i-CO
3 O
«*-
i— C
3 O
V) •!->
CO I —
3 tO
fO +->
O O)
CO fP
to Jr
re o
a)
to *•
re
to
IO O
-M O
CO
O i—
u re
4->
to c:
cu cu
II
o o
c c
col
cu
o|re j
8-12
-------
•=t
o
oo
oo
o
o
CO
oo
LU
UJ
o
oo
>—i
o:
•a:
CL.
o
o
o
CO
CO
-•— *
c
•4-» +3 O
-r- CO -J-3
C O "*"**.
ID 0*0-
O)
CO 3
coi SI-'- 3 CMl-
•coi o to o jc
C-3 "r— X/} Jv**»
; -CO t— tO
ICT; ;CO CO CO .£1
E -r- -)J «C _J
•r- £ 0 *-^
~O LJLJ J—
.a
CTl ^LO
.CO P^.
ro
g~^
j
t^
CO
CO
o;
tD!
S- cn
0- -»jg £ ^^
S_, 3 -£ -P V-
£U !E CO CO >-
•> E S- 0 -v.
CD *C 'fD C-3 ^!)-
o.. — -
r- O
0
co
'LO
[_Q
CO
E 4->
£ g_
(4— +j ^^
CJ_ (/) s^^ -•
•r- CO
Q >
.E
1— 1
CO
J _\ *^
CO 3
o: M-
E "3 CM'S-
O 00 O .E
•r- 00 -^
CO i — CO
CO CO CO -O
•i- 4-> «=C .-I
E O ^-^
LUI—
O
1 ' -
cr>
CT!
**>•
LO LO
ID CD
^j-
co
CO - — .
O CD
0 r— E
CO *r- ^—"v
r— 3 4-> 4-> S-
CO E CO CO >-,
•M E S- O '^
O «=C CO CJ -t^-
1— d •
O
JO
^M.^ «
o o
o o
LO cn
r~- LO
^d~ f™^
r— CM
CO
"^
CO
JsS?
CO - —
QJ
>
E
»— i
E
CO
CO
>_
<.r>
•
r—
2
.)_>
E
O
CJ
O C
C C
c cr
r**** r**1*
LO ^d-
r-~ oo
i— i
i — i i-^
CO CO
> >
CO .,- .,-
CO 4-> 4-^
CO CO CC
<-> E E
S- S-
O) CL) CO
to •*-> 4->
CO i — i —
CO cf CC
8-13
r-> (~y
v-__^_irf.
r-— ^"
'CTi 'O>
CO CT>
LO LO
a -*
COCO
0 0
00
CT> O
CO sf
,r^ CM
0 0
o cr
o c
co"cn
•d- 0
CM VO
^— I —
-"^^- -s
CO CO
•_^«^^
O CD
O O
co cr>
«*• CTv
co -a
C (U
CO S-
•— <1J
Q. >
s-8
a, a,
o
O CM
3 O
«*-
r- C
3 O
(/J +J
CO i —
3 CO
CO +J
i — C
o ai
0) O)
co s-
CO O
o c:
co s-
CO O
CO
O r-
O CO
4->
in c
QJ CO
1!
o o
CO
4->
O <0
-------
CO
en
S-
Of
^^
^^
f*^
crv
t
CM
o
VO r—
LU
_I CO
K-4 c:
U- 0
f% f.-gg
« 4J
O- to
S-
_i cu
«C Q.
»-« O
5 S
HH 5
U- 5-
O
_I U,
f^f. {|3f;
S £
•o
U. =3
Ij
O •— *
t-« 1
H- I
cC co
>• 0
UU CO
i "to
LU CO
•— * >1
U. C
LU (O
PS 0.
LU O
§ °
3 5
o
xr
c
f^
r~
CO
*sd" ^jr to
tn co> CM
CD CM O
I* ft •!
CM CM CM
: cn tni -3- cr»
; CD; in I u> oo
CO
a.
•r—
B—
L 1—
-C
a.
r
, tn «3~ CM
CO OX CO.
CO f^>» [ **O*' CD^
1 •*- -r-8! CM IT-
^
r-«
CU
JC
CO
i — CM en
CD i — CO
tn co r^%
M « <\
LO ID in
CO' ooT in, CFJ
I^s, (»D | CD t*^»
tn r— 1 «*• co
,
CO
r—
•r—
: .a
^*
I>^ CM *^t*
CM CO IO
CO r— l-v.
i-^ CO CD
r-* r~* r~*
co ml co CD
av r>. CM i—
co in oo co
« «|
CM t— 1
CO
o co tn
§
j5
LU
! cn •— o
CTi ^O ^O
*<•*<*
tn CD co
CM CO CM
CO O| CO CO O
i — -^H r^ in •!-
o CD! co cr> •—
r*T «rf-| CM r— 5^-
E
t— • £1
t •!-
CO (0
C CO
(!)
10 CD* *QJ CU **— i~-
H- i- to . i— 4-> ••-
r— CU CU tO C 4-
ret Q. x co*— » res o
i— o ns s_ rs s-
CJ 1— E - i— i—
c: s- i — ~^~ i — to
T- co o o co «a; -i—
U. 4J <4_ 4J t. , — o
CO- CU •!— 4-> CU C
co o D3 f- cu t- •• IB
CO O; O Q_ i- LU E
« CO 4-» 5- 03 I — T-
OQ 4J to r— -r— CO D- r— CO CD Li-
CU CU- 03 *f- CU (O»— • ^
CO r— 4-> O X 4J 4->
» co(dO&»(8COO •
«C cC CO 1— Q. 1— z: h- CO
p_
cu
t- CO r** r— UO O> CD O
recotntn ocOi — • —
Vs^1 *>J' VO OO OO r""* r™™ GO
4^- r— r- r—
, 1
(O CO
i- cu co
CU X 4->
a. res cu
Oh- CO
"S £ ^
o c
4-> C|_ 4-> O
CO CO •!-
O DO C|_ C
cj o S-
CO 4-> S- 13
4-* CO i — • CO *r- CO Q- 4-*
a> cu to c "+- a; cu
co i— 4J o o x+-»a;
co ca o -i— s- 03 (u
«=Ccoi — +J a. H- 2: &9
.
CD
CO
s_
cu
B
cu
Q.
CU
co
O)
c:
•5
c
co
i.
(tS
co
fO
8-14
-------
The financial profile of the smaller, less complex firm is based
upon an analysis of the published financial statements of Murphy Oil
Corporation, Quaker State Oil Refining Corporation, APCQ Oil Corporation,
United Refining Company, and Edgington Oil Company, These firms con-
sidered together should adequately represent the small refiner sector of-
the domestic oil industry. The resulting financial profile is presented
in Table 8-12.
The economic impact associated with standards of performance results
from the incremental cost imposed on a source to comply with these
standards above those imposed on a source to comply with existing state
or local air pollution regulations. As discussed in Chapter 3, most
State Implementation Plans to meet the national ambient air quality
standards for SC^ require new plants to achieve an overall sulfur recovery
of 99.0 percent. A few local air pollution control regulations require
an overall sulfur recovery of 99.9 percent. Consequently, most state or
local regulations already require the installation of an alternative I
emission control system and standards of performance based on this
alternative will have no economic impact.
To assess the economic impact associated with standards based on
emission control system alternative II, the effect of compliance with
standards on the financial profile of a typical refiner was evaluated.
Eleven cases covering various refinery and sulfur plant capacities were
examined, three representing a large refiner and eight representing a
small refiner. These cases are presented in Tables 8-13 through 8-23
and are summarized in Table 8-24.
8-15
-------
a
ra
ai
•5
CM
0
Ol
c.
CT
-o
UJ
oo co to o en i— LO
r— LO CM CO LO r>^ O
en
Q.
3!
•r—
CO
. oo r—
CM
UJ
a
tu
o
o
a:
o:
UJ
•a
o
oo
CM
CM LO in
r~» LO en
vo LO to
cy>
•
CM
co cy»
to
U
•r~
i/)
4-)
ro
OO
%
a
a.
S
P_
(0
o
S-
OJ 0)
_S^ Tj^
ro ro
3 4^>
croo
g—
a.
s-
3
i — r^
t— CO
r^*« c-~
i — CM
• — CO
i— CM
r^. CTJ
r-~ LO
CO r— CO LO r-.
CTI ^i* ^* en co
I^ CO r— r—
r—
I— CM CO CTI O
CO VO CO CM i —
o oo «3- «d- LO
LO
o
o
OO
O
,__
OJ
i.
i.
ra
•fa5
to oo r~, CM co CTI ^5
VO "<^ i,O OO VO r™*" f"^
•!*• CO O1 CO (£, [^ cvj
i — i — O1 i — CD CD CO
r— - ^™
c
O)
•r—
Ol
4J
O
c
c
s-
oo
OJ 0)
to i—
oo ro
CC 00
•r—
4->
rO
s-
a>
ex
o
it—
o
4^
to
O
/—"•
ro
1 ^
0
f—
to
(U
X
rO
H-
^
O
^.
O
«t-
Ct)
ca
4->
•r- OO
M- O)
0 X
i- 03
a. I—
4.9
• r-»
O
^.
Q-
4->
CO
00
OJ
ra
oo
§g^
ro
CD a>
• — >-
o -^.
i- 00
4-> i —
££
S_
r— 03
ro ca
1 ^ ^^^
O
1—
•r-
4->
•r™
rO
cr
i—
t—
eC
..
UJ
f—
O
z:
a)
»^_
t[ ^
2
CL.
r—
(O
•r—
O
c-
ro
c:
ul
.
CO
a>
to
a»
^—
ro
OO
•r™*
ro
i-
a>
ex
CD
tf«
O
4J
00
O
o
r—
ro
4->
O
t—
00
O)
X
03
H-
O)
i-
o
C{«
OJ
ca
4-5
•r-
O)«
0
S-
Q.
[ .^
•r—
M~
O
^
00 O-
O)
X 4->
ro CD
H- 2:
00
CD
00
00
^^
C
O
£Z
s-
3
1 ^
a>
ct:
^^
t-*
o
ra
a.
«a
o
t{^
o
^^
CD
cr>
c
o
•a
O)
to
ro
CO
ro
8-16
-------
CM •=!-
O r—
r-: .«
o
(J>
CQ
3
D.
0 0
0 C5
o
o
s
c
i.
01
CO
01
-O
=
-a
l
o
o
o
c>
r—
-------
_)
8
cb
at
£
o
i,
o.
u
ro
c
•i—
u.
c
o
u
TO
:G
£
in
=co
o!
(/i
s:
o
o
RJ
i
1—4
CO
•5
s-
40
C
3
O
4J
O
CJ
E
>
c
* t-
4-J
^
O
5
•o
GJ
C
O
•1—
ra
•o
'x
c
QJ
**~
TO
C »"•<
S-
-4-J
t—*
*~~*
QJ
43
(13
C
s-
QJ
^£
c
43
•a
aj
C£
c
o
4J
(t)
•a
o
at
(>
43
«3
OJ
5
t— •
OJ
*r^
*Q
£
5
c
u
1
o
•a
x
o
4J
(O
J-
OJ
CM
r— OJ
(O (U V)
c*£
ro O QJ
d L. in
•»- 0_ ro
U_ CC
CO *** ^^
in t^ ^r
^r u> ID
•^ to ro
^ ^^* i«^
D r^ n
«3- U> ro
CM in co
o i — cr>
<0 T- J^.
r^« T**.
ro in TO
o i — o ;
CO r-^ CO
l i- I
LO iO
01 CM
O I O
<= C5
;
CO «5
CO i —
0 I 0
0 0
1 1 1
CM CM CM
to o <*•
t— O O
O O O
in CM «3*
a\ o co
0 00
o o o
I^ CM VO
o o o
c o o
2 «* CT>
co in in
•* «5 CO
to «tf-
in ic o
O r- i—
Co" r-^ CO
0*0 aj
+J *~ w
to QJ «a:
s- s-
Q> O C
C <4- 4J O
CO) •—
CD M- C
>t- 0 S-
l/i O 4-» S- 3
4J U) -r- Qt 4-»
W i — to C 4-> DT
*/) to O S— QJ
=t trt CJ 0- Z 4«
in
•o CM
O O
an
O r—
O O
:in
4^ —1
QJ CD
in CQ
•a; *«>• %^
c
0
c
s-
4J
E=
fO
E:
s
s-'
in
QJ
0
QJ
QJ
in
OJ
o
c
t— 1
0)
o
0-
Itt
*o_
t-
s.
o
•S-'g
^
co di
»— -3
•5 I
•
8-18
-------
o
Oi
CQ
et
o
in
vo
a •—
a
o
O
O
o
O-
t-
(U
4-t
3
O
CO
!
£*
-------
QJ
(*_
£
CL,
(13
U
C
c;
u_
c:
o
u
ro
t—t
ts>
O-
in
2Z
£
c.-.
s_
tw(
o>
1=
s-
•a
c
-
"5.
^
•— CO
co r*-. r—
^- o vo
i t i
I-. CO
CD CM
•— 1 O
d d
co to
=> 1 O
d d'
1 II
2 g i
d c" d
00 CM 0
in c co
°, c. °
d cd
o CM •*•
0 CO
d c" d
.e • co f*.
•a- co en
r^- i-^ en
CJ Oi
co r— r^<.
n r*. cv:
i-^ O i£3
tn
(U in
CX 4->
o ro oj
*^~ t~
4J !>;
£ g! •*
OI 0 C
C 4- 4J Q
C OJ -T-
CO 4- C
*4- C fc-
«/) O 4-> S- 3
4_> t/) .,_ o_ 4j
QJ OJ 4-> **- QJ
U» i— Crt C -M Or
t/> ro O i- OJ
D. IS fct
CM
O CO
O C3
cn
0 r-
'C3 0
V)
Q} CO
t/) CO
C
o
£
-3
C£
C
c
•1—
(O
o
S-"
C/1
tn
u
OJ
03
u
c
s
T
0-
8-20
-------
QJ
M-
S-
o
rtJ
.^
u_
C
o
-M
o
ro
D
CO
CO
1/5
Q_
CO
H:
• M-
O
(J
ro
»-i
*o
s-
-M
o
O
(^
o
•*-»
in
•— i
QJ
C
s_
QJ
5
OJ
*•£
id
c:
S-
s
*-•
QJ
02
C
S-
QJ
'5
QJ
(O
5-
EE
_
QJ
-M
QJ
£
QJ
4-J
(O
c
s-
QJ
-M
S
(O QJ
O -r-
C **-
££
«r- O-
LJ~
O
O
•o
S
,_
O
rt)
•^
X
c:
-
c:
c
4J
O
Z3
QJ
C
0
-M
rd
."^
X
O
C
o
U
T3
QJ
t>;
C
o
(O
•^
X
QJ
O
in
(O
cc
^) CO «=1-
-o «t3- cn
-o ro c^
i — i — en
£ § §
, , eft
t i i
0 CM
O 1 C?
o c:
( rn
n CM
0 I 0
o o
\ i '
LO ^- CO
r- 0 CO
•— c: o
d o o
r*. r- CM
& g 8
0 <~ 0
i i i
o c* o
10 co r^-
ID ^f 1C
«3- co er
r— • r— CT:
C
O
s-
QJ
C
c:
t/» O
QJ QJ -M
in i — v]
in TO O
•=C CO O
U *
LD O CJ1
|ro r-^ o
r— O tO
° O <3-
co r*^- r-~
f— O tO
1 1 1
CM or»
CO i — I--
co r- CM
i-^ 0 1C
tn
cu in
X 4J
03 a
t/
a;
O E
Q) •!-
"S E
4-> S- 3
•r- Q- *-»
e-M or
QJ
CO
CO tri-
CS co
0 o
CO
CM **
O CM
O C3
1 1
/)
U CO
V) CO.
in -^^
5
C
=]
•*->
QJ
CC
C
1
5
rd
in
WJ
QJ
O
QJ
QJ
(/)
£
o
c
QJ
(J
"£
ex.
CM
3
QJ
(Q
t—
QJ
Co?
.1™
QJ r— C
.2 C ?
QJ *r- t/J
a.**- a>
t/) i — .£1
t- QJ (O
ro -ot—
£— M E
*° ai4^
.c .c -c
VI QJ -t-»
QJ S- tO
^^5
"G E E
-------
g
CD
**-
O
L.
Q_
U
CZ
ns
•T^
U.
e
o
u
m
t— t
i
to
to
2=
M-
O
4_>
L)
ro
io
e
c
u
o
4.
C
QJ
(U
s
c
o
o
•o
c
o
ro
T3
*X
C
QJ
^
g^
s-
-«
(U
c
i.
5
c
o
u
•o
CD
C
0
"D
X
c
4->
ro
C H-t
fc.
•*-»
(U
4J
i
QJ
**•
c
o
• •*->
c
•c
a
c
c
(O
^
>
0
OJ
*-M
m
C I— I
t.
5
CM
r— 03
tO O tf)
•r* ^— tl
L> "r- C-5
(0 O 01
C i- in
T- CU fO
U. CC
CO 0
^- 01
co en
r—
02 !^
«a- co
co en
r- r-^ en
r—
0 O CM
* CD *O
r^ ^»
cn r«* i —
,-! o iO
t 1
T^ CO
IO CM
[ 1 0
CD O
S , 0
C3 O
1 1 1
CT» cn
r^ cva
0 1 0
0 0
CO O
CO CSJ
0 1 0
0 0
CM ^
c o
tc cc r-^
*£. ^- 1C
*r co en
r-^ *— en
co r-^ c\
•~ c *°
(A
OJ vi
CX 4^
o ro
*43 *~ j
0) 0
C <1) ••—
CO «*-
4J (/> -r- D- •
QJ QJ 4J **-
tn i— w c •*-> o_
t/) ro o J- OJ
e£ t/*. O & Z ^
CO
co en
O CM
0 0
CO
O i—
CD CD
j |
QJ CD
j ca
5
~
3
J
1^
-
J
d
Z"
»-.
n
in
u
u
V
s
u
c
s
T
Q-
f
CM
r—
QJ
^i
£
OJ
.1
1"
l"i
!- C
OJ •*-
C "4-
W I —
S- 01
03 -O
^i
Sis
•~5I
in O) QJ
QJ C i—
rs xi
r— S- «
ro ajt-
, 're o
wS
-------
a
o
Ol
I
CO
8-23
-------
c
E
s
o.
n
C
ro
c
LL.
c
o
I
n
•§
V)
a.
tr
co
31
M-
O
jj
U
re
§
t— »
CO
r—
O
4J
C
2
M-
O
3
1
=
•M
c
o
4
c
i
c
c
n
X
c
QJ
g_
ai
5
0)
•r-
4J
n)
ai
5
c
c
4J
(J
3
c
O
•r—
fO
•D
'x
4J
re
c
i— i
£
53
u
(U
5
h— 1
O
o
•c
£ £ 5!
•r- O. CO
LJU cc
f- en o
ir> -^ CD
irj co o
i— r— O
IO Crt CO
o «a- co
w> co en
;— i— en
t I t
i — i — CO
O C CO
i— 00
C O 0
«9- C CM
0 00
0 C5 0
01 CM
^ o r-^
ro (^ o
r— O l£
S °°
rO f*s. r—
"™ C3 <£
1- 1 1
Fi s °
er o' o
n o co
o oo
0 CO
r~~ cr.
; C
c: c"
i CC ^^
1C *3- Id
«3- CO CT.
r— i-^ CO
CM i—
3 o " g |
*-* f> T- D. 4J
QJ (U 4-> M- oj
fl i — tn C q_> rt"
*/> ro o S- QJ
=C co o a. z: s^
in
o ^
o o
ID
CM ro
o r».
o d
in
QJ CO
en ca
c:
o
_
u
1>
"~
f>-
•a
5
5
r\
tn
-------
UJ
z
*-t
U-
s
CO
I— O
oo '•a
*f CO
ro 01
oo 01
10 o
co r-^
r-^ O
o
o
o
01
o
o
E
en
Ul
CM
cb
O)
I—
JD
i— 01
to CO c/1
•r- I— (O
c >£
ro o cu
c s- w
s-
OJ
oo
cu
e
o.
a; «j
S- O
£|
l^
S2l
ro o
^ito
cu T
tn CO CU
CO S-i—
3 XI
t— s. ro
ro cut—
> r—
:5iiit
8-25
-------
QJ
O
S-
o.
u
c
TO
c
E
C
O
(J
to
S
tn
0,
to
2=
to
to
o
4-*
U
(O
CO
2
•M
C
O
o
o
01
5
£
g
£
QJ
5
C
O
•M
0
•o
QJ
ce
c
o
•M
(0
X
0)
4-*
CI-H
QJ
S
»— •
*"^
c
c
U
•o
QJ
C
o
4J
(O
•a
"x
o
QJ
rtJ
C H-l
<5
4Jl
»-«
QJ
T>
ro
S_
QJ
£
c
o
-M
U
•a
QJ
c
o
(O
TO
I
£
o>
•t- i — (O
O *r— 0
C QJ «r-
CQ •*- C
«f- 01-
01 O 4-> i- 3
4J V) •.- Q. 4->
QJ QJ 4-> M- OJ
Ol r— ol O 4-> Di
Ol (tJ O i- Q>
•a: tn tJ cu ET a-y
CO
in t--
o ^»-
CD CD
CO CO
O CM
0 CD
01
OJ CD
tA CQ
O
g
3
ce
C
cO
c
rO
Z"
c
1
Ol'
1
z:
CJ
Ol
S
o
c
QJ
n.
CM
CO
OJ
(O
H-
0)
2
• O-
S- 'tO
(0*0
g'C
(O "O
'o •
-0 J- 0
QJ *—
fl~ (^
V) QJ QJ
QJ i- t—
2 S-"S
id r—
$1
O . - •
£|r- CM CO
8-26
-------
o
c
C +- ^
rO O QJ
C S- l/l
-^ D. reJ
y_ cc
co *—
,— i— en
-------
a
•o-
xco
oca
•sf r^ oo
i— CM r—'
co ir> en
CM *3- CMS
o o o
CM
o
CM LO «3-
r- CM O
O O O
o
to
CO
o
CO
o
en
CM
LO
O
o
o
in
co
co
en
CM CO CO
•3- CO CD
O O o
CO
in
o
O CO «O
§ S 2
«3- ID
CM r-
o> • co co
>— CM r—
O O O
CM
o
o
CO
CM
10
CM
o
CO
CM
CM
O
CO
CM
r— CO
CO tO
O O
•U
U
01
cn
CO
o
CM
O
•* O
i— CM
cn to cr.
CO r— i—
CO
CM
co
CM
o
in
LD
«*
^
CM
1C
U
IO
a.
r— ta
3 a
§- o
to
0) U
CC
a
-------
As shown by the large refiner cases of Table 8-24, standards based
on emission control system alternative II decrease the profitability (as
measured by return on assets) of a large integrated refiner by 0.37 to
1.48 percent, if unaccompanied by price increases. To maintain an 8.10
percent return on assets, a large refiner would"have to increase prices
for petroleum products by 0.02-0.27 percent, or about 0.4-4.5 cents per
barrel. It is highly unlikely that this impact would retard industry growth
among the large integrated refiner sector of the domestic petroleum refining
industry. The reduced profitability, or the increase in product prices
necessary to maintain profitability, are for all practical purposes negligible.
As might be expected, the impact of standards on a small refiner is
greater than that on a large refiner due to economies of scale. As shown
by the small refiner cases of Table 8-24, standards based on emission
control system alternative II decrease the profitability of a small
refiner by 1.28 to 7.50 percent, if unaccompanied by price increases. To
maintain a 6.27 percent return on assets, the small refiner would have to
increase prices for petroleum products by 0.16 to 0.93 percent, or about
1.8 to 10.6 cents per barrel. Although it appears that the impact of
standards on the smallest refiner may be from three to five times as severe
as that on the typical large refiner, the magnitude of this impact is still
quite small and not likely to retard industry growth among the small refiner
sector of the domestic petroleum refining industry. As with the large
refiner, the price increases necessary to maintain profitability are negligible,|
especially in light of price increases over the past three to five years.
8-29
-------
The impact analysis by refinery size is clouded by the relatively
small size of the control investment and annual cost compared with the
overall refinery operation. There is, however, a more pronounced effect
when the incremental differences between control units is considered.
Table 8-25, which is derived from Tables 8-8 - 8-10, presents'the incremental
costs of achieving alternative II utilizing five sizes of control units.
Due to economies of scale, the cost of controlling an incremental ton of
S02 at the level of alternative II goes from $468-678 for a TOO LTD
sulfur plant to $3,891-5,994 for a 5 LTD sulfur plant.
Nation-Wide
Table 8-26 provides the number and size distribution of Claus plant
affected facilities in the period 1976-1980. It should be noted that
the growth is greater than the projected annual increase of refinery throughput
for three reasons. First, approximately 30 percent of the current refining
capacity is not controlled by Claus units so the base is narrow. Second,
it is assumed that all future refinery capacity, will be controlled and
third there will be an annual replacement of 5 percent of the existing
Claus plants.
Table 8-27 develops the 1976-1980 forecast to show the national
impact of required investment dollars, the annual costs and the potential
emission reductions. Standards based on alternative I will have no impact.
Standards based on alternative II, on the other hand will require an
Incremental investment by the domestic petroleum refining industry of some
$110 MM over the five-year period of 1975 to 1980. In 1980, these standards
will increase the annual operating costs of the domestic industry by some
$16 MM per year. In return, the standards will reduce national S02 emissions
by some 57,000 tons per year.
8-30
-------
o
O
co
co
co
CO
LJJ
O
UJ
u_
u.
UJ
CO
CD
O
in
CM
co
co
4-* O
10 H-
O
co 10 r-»
r^, co LO
to r^- CM
CM
CO
en
•»
•*
CO ^_,
to T3 $_
J^
to O £o
to s-
,E.
LU 0
0 00
CM
CO
o
CO CM
LO i —
oo CM
CD
LO
oo
OJ
OJ >
c_> res
IO CU
tO •)->
CQ r—
CM
O
CM CO
CO O
0 f—
•*>=>-
CD
CM
CD
O
CO
LO
CM
LO
cvj
00
LO
CO T3 S-
CO O OO
CO S_
.^-4-3 «
£ C CM
LlJ O O
C_3 c/5
LO
00
•k
CO
CO
LO
CO
CM
n
CO
CM
O
IO
O
CO
co
*v
oo
•a
t-t OJ
cu s-
T- 01
tj ^*
CO CO
I^E-a
4-> CU CD O
^^~ ^> «l^ [ —
CO O CO I—
o >> o
r— CU CO S-
4-> S- O O
C 3 S- O
O i— C CM
300
C CO O CO
O
•r- •&« £= <4-
w cn o o
tf) « «r-
•l- CT> tO C
TD CO CU r™*
C CO
(O T3 C 4->
0) O £=
+J r— -r- CD
(O O
•«-> Q
C H-
CO —1
CU
<+- O
•— AS
3 n.
co to
0
CD
0
LO
o
OO
S- i—
CU 3 C
> to o
O
CJ
o
3'
-o
£= CU
O S-
o
to
a> o
s-
3
£=
to O
tO •!-
•I- •«->
*-> E 10
to cu "a
o
S-
cu
CX
to
4->
CO
o
o
«f- O r—
tO
to
3 to
CO
O i— «r- 4J
. O TS to
CTi C "O O
en M
-------
8.4 POTENTIAL SOCIO-ECONOMIC.AND INFLATIONARY IMPACT
Since the emission control systems required to comply with standards
based on either emission control system alternatives represent such a
small proportion of the overall equipment or investment required by a
petroleum refinery, there should be no more socio-economic impact associated
with standards than associated with the addition of any new processing
unit to a refinery.
The inflationary impact associated with standards of performance
for refinery sulfur plants is negligible. If standards are based on
emission control system alternative I, there is no inflationary impact
at all. If standards are based on emission control system alternative
II, the fifth-year annualized costs are about $16 MM/year, and the price
Increases necessary to maintain the current industry average return on
investment varies from $0.004 to $0.106 per barrel, depending on the
size of the refinery affected. These are well below the Environmental
Protection Agency's guidelines for preparation of inflationary impact
statements.
8-34
-------
REFERENCES - CHAPTER 8
1. Oil and GSs Journal, April 7, 1975, pgs. 100-118.
2. Hydrocarbon Processing, February, 1975, pgs. 3-15.
3. Characterization of Sulfur from Refinery Fuel- Gas. (Contract No.
68-02-0611, Task 4), Battelle Columbus Laboratories, June 28, 1974,
pgs. 22-27.
4. Department of Labor, Monthly Labor Review, October 1975, p. 96.
5. Ford, Bacon and Davis, Sulfur Recovery Plants, 1971, p. 80.
6. Federal Power Commission, Typical Electric Bills, 1974 FPCR-83.
7. Beers, W. D., Characterization of Claus Plant Emission, Final
Report from Process Research, Inc., to the United States E.P.A.
Contract No. 68-02-0242, Task No. 2, Report No. EPA-R2-73-188
(April 1973). p. 75.
8. Letter, Hanley, D. L. Union Oil Company to Sedman, C. B, E.P.A,,
dated 12-4-73.
g. Comparative Assessment of Coal Gasification Emission Control Systems,
(Contract No. 68-01-2942, Task 007), Booz-Allen & Hamilton, Inc.,
October 1975, page A-3.
10. Letter, Andrews, J. W., J. F. P. to Genco, J. M., Battelle dated
12-21-75.
11. Letter, Turner, W. W., Stauffer Chemical Co. to Goodwin, D. R.
E.P.A., dated 6-14-74.
12. Letter, Ballard, B. F., Phillips Petroleum Co. to Sedman, C. B.
E. P. A., dated 12-18-74.
13. Trip Report: Sedman, C. B., Standard Oil Company of California, El
Segundo Refinery, 10-15-73.
14. Hydrocarbon Processing, April 1973, p. 116.
8-35
-------
15. Op. Cit, Booz-Allen & Hamilton, page A-ll, 12.
16. Moody's Industrial Manual, 1974 and 1975,
.17. Beers, W. D., op. cit., page 94.
18. Beers, W. D., op. cit., page 93.
19 Op. Cit., Booz-Allen & Hamilton, page A-5.
8-36
-------
9. RATIONALE FOR THE STANDARDS
9.1 SELECTION OF SOURCE FOR CONTROL
Sulfur dioxide emissions from petroleum, refineries are a
function of the sulfur content of the crude oil processed and the
complexity of the refinery itself. A major portion of the sulfur
which enters the refinery in the crude oil leaves the refinery in
the various petroleum products produced. Most of the sulfur
not accounted for in these petroleum products is recovered
as elemental sulfur, or emitted to the atmosphere as S02-
The major S02 emission sources in petroleum refineries are
gas and liquid fuel combustion, fluid catalytic cracking unit
catalyst regeneration and elemental sulfur recovery. Standards
of performance limiting S02 emissions from gaseous fuel combustion
were promulgated on March 4, 1974 (39 FR 9308). These standards
essentially require the removal of H2S from fuel gas before
it is burned, thus forcing increased elemental sulfur recovery
within petroleum refineries.
Petroleum refinery sulfur recovery plants, however, as mentioned
above, are responsible for a sizeable portion of the total S02 emissions
emitted from petroleum refineries. In 1975, nationwide refinery
sulfur recovery plant S02 emissions were estimated to be 0.272 x 10° MT/
yr, or about 10% of total domestic refinerv S02 emissions. Although
refinery sulfur recovery plants are responsible for only a small
portion (about 1%) of the total unabated U.S. S02 emissions, the
expected rapid growth of these facilities emphasizes the need for their
control. As of April 1975, there were 122 sulfur recovery plants
located in 81 domestic refineries. Between 1976 and 1980, 107 new
-------
olants are expected to be built. Well over half of these new
plants (about 62%) will have an average size of 100 LTD. The total
combined capacities of these new plants will be about 6,000 LTD
so that by 1980 the number of refinery sulfur recovery plants will
double.
A very important consideration of developing standards for
refinery sulfur recovery plants is that most refineries are located
in or near urban areas. As the size and number of these refineries
increase and the average sulfur content of the crude oil processed
increases, control of S02 emissions becomes much more critical.
9.2 SELECTION OF THE BEST SYSTEM OF EMISSION REDUCTION
As discussed in chanter 6, two alternative emission control
systems are considered candidates to serve as the basis for standards
of performance (i.e. best system of emission reduction, considering
costs). These systems are the third-stage low temperature Claus
reactor system (alternative I) and various tail gas scrubbing
systems (alternative II).
Considering only the nerformance of these systems, the
alternative II systems are clearly superior to the alternative I
systems. (See chapters 4 and 6,0 Use of an alternative II emission
control system increases the overall sulfur recovery of a refinery
sulfur nlant from about 95 percent to 99.9 percent, comnared to
99 percent with use of an alternative I emission control system.
In terms of emission reduction, the alternative II systems reduce
emissions by 98-99 percent, compared to an emission reduction of
only 80-85 percent achieved bv the alternative I systems. Also,
9.2
-------
the alternative II emission control systems are essentially insensitive
to fluctuations which might occur in the composition of the tail
gases from the sulfur plant. The alternative I systems, however,
require strict maintenance of a 2:1 H2S/S02 ratio to function properly.
Thus* the alternative II systems are much less prone to upsets and
are able to limit emissions to lower levels over a wider variety
of operating conditions than the alternative I systems.
Considering the various environmental impacts associated with
both alternatives, the alternative II emission control systems again
emerge as clearly superior to those of alternative I. (See chapter 7.)
In terms of ambient air quality, although the alternative I emission
control systems result in a significant reduction in the maximum
ambient air concentrations of S02 arising from the operation of a
refinery sulfur nlant, the alternative II systems result in a
substantially greater reduction of these concentrations.
More importantly, however, most State implementation plans (SIP)
already require a level of control essentially equivalent to the
alternative I emission control systems. Consequently, standards
based on this alternative would have little or no impact on emissions
of S02 from new or modified refinery sulfur plants. Standards based
on the alternative II systems, however, will reduce these emissions
by about 90 percent and will lead to a reduction in the growth of
national S02 emissions by 1980 of some 55,000 tons per year.
Considerinq possible environmental impacts in other areas,
there are essentially no potential adverse water pollution or solid
waste impacts associated with either alternative emission control
9.3
-------
system1. With regard to energy consumption, both alternatives
reduce the overall energy consumntion associated with a refinery
sulfur plant. Since most SIP's already require the installation
of alternative I emission control systems, no reduction in energy
consumption can be associated with standards based on this alternative.
If standards are based on the alternative II systems, however, the
growth in-national energy consumption will be reduced by some 54
million kw-hr/yr (90,000 barrels of fuel oil per year) by 1980.
With regard to other areas of potential environmental impacts,
there appear to be no noise or radiation impacts, or any irreversible
or irretrievable commitment of resources associated with either
of these alternative emission control systems. Neither does there appear
to be any incentive for not developing or delaying standards.
In terms of the economic impacts associated with the alternative
emission control systems, the alternative II systems generally cost
about twice as much to install and about 2 1/2 times as much to
operate as the alternative I systems. Again, however, since most
SIP's already require the installation of alternative I emission
control systems, there will be no economic impact if standards are
based on this alternative. If standards are based on alternative II,
the impact on a typical large integrated refinery will reduce its
profitability from about 8.10 percent return on assets to about
7.98-8.09 percent. To maintain an 8.10 percent return on assets,
the refiner would have to increase prices for petroleum products
by only 0.03-0.12 cents per gallon, or less than 0.5 percent. The
magnitude of this impact, therefore, is negligible.
9.4
-------
The impact on a typical small refinery is larger than that on a
large refinery due to the "economies-of-scale". In this case the
impact of standards based on the alternative II emission control
systems will reduce the profitability of a small refinery from about
6.27 percent return on assets to about 5.80-6.18 percent. To
maintain a 6.27 percent return on assets,, the small refiner would
have to increase prices on petroleum products by about 0..06-0.25
cents per gallon, or 0.16-0.94 percent. While this impact is about
twice as severe on the small refiner as on the large refiner,
its magnitude is still quite small and not likely to retard
growth among the small refiner sector of the domestic refining
industry. As with the large refiner, the price increases necessary
to maintain profitability are negligible, certainly in light of price
increases over the past three to five years.
In terms of the national impact on the domestic petroleum
refining industry, standards based on the alternative I emission
control systems will have no impact. Standards based on the
alternative II systems,, however, will increase the national investment
required by the domestic industry by some $115 MM over the five-year
period from 1975 to 1980; and the annual operating costs of the
industry will be increased by some $16 MM per year in 1980.
The potential inflationary impact of these standards is
essentially negligible. If standards are based on the alternative I
emission control systems, there is no impact, and if standards
are based on the alternative II systems, the increased fifth-year
annualized costs and increased,product prices are well below the
9.5
-------
Agency's guidelines of $100 MM per'year and 5 percent for signalling
potential inflationary impact.
It is clear, therefore, that the alternative II emission control
systems must be selected as the "best system of emission reduction,
considering costs" and that standards of performance for refinery
sulfur plants must be based on the use of these systems.
9,3 SELECTION OF POLLUTANTS FOR CONTROL
As discussed above, sulfur recovery olants in petroleum refineries
are major point sources of S02 emissions. .The objective of standards
of nerformance, therefore, is to reduce these emissions from new
and modified refinery sulfur plants. Selecting emission control
system alternative II as the basis for standards, however, complicates
the selection of pollutants for control.
Emission control system alternative II refers to two different
processes: oxidation-scrubbing and reduction-scrubbing. Residual
emissions released to the atmosphere from oxidation-scrubbing
processes consist of S02- Residual emissions released to the
atmosphere from reduction-scrubbing processes, however, consist
of S02 if the tail gases are incinerated before release to the
atmosphere, or a mixture of reduced sulfur compounds such as
hydrogen sulfide (HaS), carbonyl sulfide (COS) and carbon disulfide
(C$2), if the tail gases are not incinerated (see chapters 4 and 6).
A limit on S02 emissions alone, therefore, while appropriate for
the oxidation-scrubbing processes, and those reduction-scrubbing
processes with tail gas incineration, is inappropriate for those
reduction-scrubbing processes without tail gas incineration.
9.6
-------
While emissions of H2S from the reduction-scrubbing processes
without tail gas incineration can vary widely depending on the
design and operation of the process, emissions of COS and CS2
will not exceed 90-100 ppm unless the reduction catalyst in the
process is permitted to deteriorate. Consequently, the major
potential air pollution problem posed by these processes is emissions
of H2S.
Minimizing emissions of H2S from reduction-scrubbing processes
without tail gas incineration requires design and operation: of the
scrubbing portion of the process to reduce emissions of H2S to
very low levels initially. If standards of performance do not limit
emissions of H2S, owners or operators of these processes could
operate only the reduction portion of these systems and by-pass
the scrubber portion. The S02 originally present would merely be
converted to H2$ and released directly to the atmosphere. Although
this is possible, it is unlikely because at the ambient air concen-
tration of H2S which would result (500-2000 yg/m3, with short-
o
term peaks, 2-10 second, approaching 25,000 yg/m} an extremely
severe odor problem would result (see chapter 7). Sources would
tend to control these HLS emissions to very low levels due to
the offensive odors which would otherwise result if they were
not controlled.
Even in the more likely situation, however, where emissions
of H2S were reduced to the range of 200-300 ppm before release
to the atmosphere, ambient air concentrations of H2S ranging from
15 to 60 yg/m3 could arise with short-term peaks as high as 500
to 800 yg/m3. Since the odor threshold for H2S is 45 yg/m3, an
odor air pollution problem could still arise.
9,7
-------
Although emissions of COS and CS2 are normally very low,
as mentioned above, if the reduction catalyst is permitted to
deteriorate, emissions of these compounds from reduction-scrubbing
processes without tail gas incineration can approach 1000 ppm.
Under these conditions, ambient air concentrations of COS and C.^2
ranging from 100 to 400 yg/m3 could arise, with short-term peaks
approaching 3500 to 5500 yg/m3, respectively. While ambient air
C
concentrations of COS and CS2 at these levels probably do not
pose health problems, so little health effects data is available
on COS and CS2 that this is questionable. (The little data
available indicate adverse health effects occurring only at
levels greater than 15,000 yg/m3 for CS2, with no data available
for COS—see chapter 7.) The data do indicate, however, that
these short-term peak ambient air concentrations of CS2 are at
the odor threshold level for CS2, so that transitory odo,rs could
also arise if the reduction catalyst is permitted to deteriorate.
Developing standards of performance to limit emissions of S02
from refinery sulfur recovery plants, therefore, gives rise to
a rather unusual situation. In some cases, compliance with these
standards, while eliminating SOz emissions, would lead to emissions
of reduced sulfur compounds (i.e. H2S, COS and CS2), which could
lead to an odor air pollution problem. Consequently, two alternative
courses of action emerge with regard to selecting the pollutants
for control by standards of performance for refinery sulfur recovery
plants:
9.8
-------
1. Limit emissions of S02 only
2. Limit emissions of both S02 and reduced sulfur compounds
Limiting emissions of reduced sulfur compounds (i.e. ^S,
COS and CS2) from reduction-scrubbing emission control systems
without tail gas incineration, however, would make these pollutants
"designated pollutants," and all existing reduction-scrubbing emission
control systems without tail gas incineration installed on refinery
sulfur recovery plants "designated facilities," under section lll(d)
of the Clean Air Act. ERA's regulations implementing section lll(d)
(40 CFR §60) would require the Agency to issue a draft guideline
document containing the necessary information for states to develop
plans for controlling these pollutants from these existing facilities
and to solicit public comments on this document. Following considera-
tion of these comments, the Agency would make appropriate changes
to the guideline document and issue it in final form. Those states
containing reduction-scrubbing emission control systems without
tail gas incineration installed on refinery sulfur plants would
then be required to develop plans for controlling emissions of these
pollutants from these facilities with public hearings to solicit
the views of interested parties. These plans would then be submitted
to the Agency for approval or disapproval. If a State's plan were
disapproved, the Agency would have to develop a plan for that state.
Currently, there are about 25 reduction-scrubbing emission
control systems without tail gas incineration now operating on
petroleum refinery sulfur recovery plants in some seven states.
Developing State plans to limit emissions of reduced sulfur compounds
9.9
-------
from these facilities, therefore, could be a significant undertaking
requiring the expenditure of considerable resources at the Federal,
State and 1 oca! 1 eve!.
Emissions of reduced sulfur compounds, from these existing
reduction-scrubbing emission control: systems without tail gas
incineration, however, are quite low. These systems have been
installed to comply with State or local air pollution regulations
limiting emissions of SO^ from existing refinery sulfur recovery
plants. To ensure that the installation of these emission control
systems do not lead to local odor problems,., these.regulations
also limit emissions, of H^S, COS and CSz» either directly or
indirectly. Where emissions of reduced sulfur compounds are
limited directly, local regulations specify the maximum concentrations
of H£S> COS and CS2 that can be present in the tail gases discharged
to the atmosphere. In each case where this approach has been
followed, emissions of H2S are limited to 10 ppm and emissions
of total sulfur (H2S, COS and CS2) are limited to either 300 or
500" ppm.
Where emissions of reduced sulfur compounds are limited indirectly
by local regulations, these regulations require that the best available
emission control technology be installed. In the process of
specifying what the best emission control technology is, local air
pollution control agencies generally contact EPA, vendors of various
emission control systems and, other local air pollution control
agencies where various emission control systems have been installed.
9.10
-------
In terms of -emissions .of reduced sulfur compounds from existing
reduction-scrubbing emission control systems without tail gas
incineration, this approach lias achieved the same end result
as that above, and all 25 of these systems which are now operating
have been designed and guaranteed toy the -vendors of these systems
to limit emissions to less than 10 ppm ^S and less than 300 otf
500 ppm total sulfur (M?S, COS and CS^J
Consequently, existing reduction-scrubbing emission control
systems without tail gas incineration are not considered significant
sources of reduced sulfur compound emissions. Developing State
regulations to control emissions of these pollutants from the'se
facilities, therefore, would accomplish no additional reduction i;n
reduced sulfur compound emissions.
Probably the major reason why existing reduction-scrubbing
emission control systems without tail gas Incineration are not
sources of reduced sulfur compound emissions is that to date-,
these systems have only been installed in heavily industrialized
metropolitan areas. In these areas, the :need for air pollution
control is generally well recognized and the local air pollution
control agencjes have been in a position to develop and enforce
strong air pollution regulations.
Standards of performance for refinery sulfur plants, however,,
will require the installation of oxidation-scrubbing or reduction-
scrubbing emission control systems on new or modified refinery sulfur
recovery plants throughout the United States. In some areas
9.11
-------
where these systems might be installed, the need for stringent air
pollution control might not be as great as it is in these areas
where these systems have already been installed. In these areas,
owners and operators who installed a reduction-scrubbing emission
control system without tail gas incineration would not face stringent
regulation of reduced sulfur compounds and they might permit
emissions of H2S to increase to the range of 200 to 300 ppm. At
this point the resulting ambient air concentrations of F^S would
lead to noticeable but intermittent and transitory odors as discussed
earlier. It is quite possible, therefore, that in complying with
standards of performance, some new refinery sulfur plants could
become sources of emissions of reduced sulfur compounds unless
the standards specifically limit emissions of these pollutants.
Preventing new air pollution problems from arising, however,
is one of the primary goals of standards of performance. Also,
standards of performance are to reflect the best systems of emission
reduction, taking into account the costs of installing and operating
these systems. Considered from this perspective, since the technology
for reducing emissions of reduced sulfur compounds from reduction-
scrubbing systems without tail gas incineration is well demonstrated,
the costs of controlling these emissions are reasonable, and not
controlling these emissions could lead to an adverse environmental
impact in some cases; reduction-scrubbing emission control systems
without tail gas incineration can only be included among the best
systems of emission reduction if emissions of reduced sulfur compounds
are controlled.
9.12
-------
It is also through standards of performance that the Agency
identifies the best systems of emission reduction for various
industrial sources of air pollution. If the Agency were not to
limit emissions of reduced sulfur compounds, therefore, it would
imply that the Agency does not consider controlling these emissions
from reduction-scrubbing processes necessary. This view could
serve to undercut or weaken those local air pollution regulations
which are now effectively controlling these emissions from existing
sulfur recovery plants which have installed reduction-scrubbing
emission control systems without tail gas incineration.
A number of good reasons exist, therefore, for extending the
standards to cover emissions of reduced sulfur compounds. However,
a problem may arise if the burden of the state plan submission required
by 40 CFR §60.23 outweighs its possible benefits. To resolve this
problem, it is appropriate to consider the intent of both standards
of performance and section lll(d).
Briefly, the intent of standards of performance is to require
the installation of the best systems of emission control at the
time a source is being constructed or modified. The intent of
section lll(d) is to reduce emissions of pollutants emitted from
existing facilities which pose a danger to the public health or
welfare, but which on the basis of the information available
cannot be controlled under sections 108, 109 and 110 of the Act
as criteria pollutants, and which cannot be controlled under
section 112 of the Act as hazardous pollutants.
9.13
-------
Where an emission control system installed to comply with a
standard of performance might lead to the emission of a pollutant
not originally emitted by a source, the logical course of action
is to limit emissions of this new pollutant to ensure that it
does not lead to a new air pollution problem. In cases where the
new pollutant is a non-criteria pollutant, it must be decided
whether or not to initiate the chain of events leading to the
development of state regulations for limiting emissions of this
pollutant from existing sources already well controlled for this
pollutant.
The pollutants selected for control by these standards,
therefore, are S02 and reduced sulfur compounds. A determination
of the effort involved in developing state plans will enable EPA
to determine whether or not to develop a guideline document or
initiate the chain of events leading to the development of state
plans for controlling emissions of reduced sulfur compounds from
existing reduction-scrubbing emission control systems without
tail gas incineration which have been installed on refinery sulfur
plants.
9.4 SELECTION OF FORMAT FOR THE STANDARDS
A number of different,formats could be selected to limit
emissions from refinery sulfur plants. Mass standards limiting
emissions in terms of overall sulfur recovery (i.e. emissions
oer unit of sulfur produced or contained in the feed to the
9.14
-------
plant), or concentration standards limiting the concentration
of emissions in the stack gases discharged into-the atmosphere,
could be developed.
While mass standards may appear more meaningful in the sense
that they relate directly to the quantity of emissions discharged
into the atmosphere, enforcement of mass standards is more costly
and the results more subject to error than enforcement of concen-
tration standards. Determining mass emissions, for example,
invariably requires developing a material balance of some form
and this requires process data concerning the operation of
the plant, whether it be input material flow rates or production
flow rates. Gathering this data increases the. testing or monitoring
necessary and consequently increases the costs. Manipulation
of this data increases the number of calculations necessary,
compounding the error inherent within the data and increasing
the chance for human error.
Enforcement of concentration standards, however, requires
a minimum of data and information, decreasing the costs and
minimizing the chances for error in determining compliance.
Concentration standards are also somewhat more consistent with
the concept of basing standards of the "best systems of emission
reduction," since vendors of emission control equipment normally
guarantee the performance of their equipment in terms of the
concentration of emissions discharged.
The primary disadvantage normally associated with concen-
tration standards is that of possible circumvention by dilution
9.15
-------
of the gases discharged to the atmosphere lowering 'the concen-
tration of emissions, but not reducing the total mass emitted.
To ensure that compliance with concentration standards is not
achieved by dilution, however, section 60.12 of 40 CFR Part 60
specifically prohibits the use of dilution as a means of complying
with concentration standards.
Consequently, considered primarily from the perspective of
enforcement, concentration standards are selected as the format
for standards of performance for refinery sulfur plants. The
lower resource requirements of concentration standards over
mass standards far outweigh their drawbacks.
9.5 SELECTION OF EMISSION LIMITS IN THE STANDARDS
Specific emission limits need to be selected to limit
emissions of S02 from refinery sulfur plants which employ either
oxidation-scrubbing emission control systems or reduction-scrubbing
emission control systems with tail gas incineration. Emission
limits also need to be selected to limit emissions of H2S and
reduced sulfur compounds (H2S, COS and CS2) from refinery sulfur
plants which employ reduction-scrubbing emission control systems
without tail gas incineration.
The data and information to support selection of these emission
limits is summarized in chapter 4 and consists primarily of emission
source tests by the Agency or local air pollution control agencies.
Since the amount of emission data available is quite limited, a
number of factors need to be considered in selecting the specific
emission limits.
9.16
-------
Considering first the limit for S02 emissions, the emission
source test data from oxidation-scrubbing emission control systems
shows emissions in the range of 10-50 ppm (tests A-j, A2, A3,
Figure 4-10). These data, however, were collected from a unit
operating at less than half its design capacity immediately
following a major equipment turnaround. Consequently, the data
do not reflect emission levels that could be maintained by a
unit operating at full capacity over a period of time. According
to both vendors and owners and operators, unavoidable equipment
deterioration and chemical aging will lead to lower efficiency.
Generally, these systems operate for about a year between major
equipment turnarounds and during this time, emissions increase
by as much as 100 to 200 ppm. Basing the emission limit solely on the
basis of this emission data, therefore, would significantly
shorten the normal run length between major turnarounds and
increase maintenance and chemical replacement costs considerably.
The emission source test data available from reduction-scrubbinq
emission control systems with tail gas incineration shows emissions
of about 200 ppm S02 (test C, Figure 4-10). Although this data
is not EPA data, but is data from a test bv a local control agency
(EPA's source testing at this facility was invalidated due to
operating problems), the source testing method emploved by this
local aqency is considered comparable to EPA's.
9.17
-------
This emission level, however, is considered higher than normal
for a typical reduction-scrubbing emission control system with tail
gas incineration. According to the operators at this source, this
particular facility is significantly under-designed and, in fact,
shortly after this emission test data was gathered, it was "de-
bottle-necked" and expanded to improve its performance.
In EPA's discussions with the vendors of this facility, they agreed
with the operators that it was "under-designed." They also indicated,
however, that emissions from typical reduction-scrubbing emission
control systems with tail gas incineration are normally comparable
to those from oxidation-scrubbing emission control systems, and that
emissions from both systems increase to about the same extent between
turnarounds due to unavoidable equipment deterioration and chemical
aging.
As pointed out in chapters 4 and 7, the reduction-scrubbing
emission control systems with tail gas incineration (as opposed
to those without tail gas incineration) have a number of advantages
over oxidation-scrubbing emission control systems. Operation
of these systems involves techniques with which most refiners
have had a great deal of experience and thus refiners understand
these systems better, tend to experience fewer problems with them
and generally tend to favor these systems. More importantly, however,
these reduction-scrubbing emission control systems with tail gas
incineration produce no wastewater streams that require disposal.
Both the oxidation-scrubbing systems and the reduction-scrubbing
systems without tail gas incineration generate a wastewater stream.
9.18
-------
Considering all these factors, the emission limit for S02
emissions is set at 250 ppm. This limit applies to both the
oxidation-scrubbing emission control systems and to the reduction-
scrubbing emission control systems with tail gas incineration.
In view of the limited emission data available and the comments
of. both the owners and operators and the control system vendors,
this appears to be a reasonable emission limit consistent with the
performance of these emission control systems. This limit will
also ensure that these alternative II emission control systems are
installed and well operated.
Considering the emission limit for emissions of H2S and reduced
sulfur compounds, the available emission source testing data from
reduction-scrubbing emission control systems without tail gas
incineration shows that emissions of these pollutants from these
systems are quite low. Emissions of H2S, for example, were
frequently not detectable (tests B], B2, E1 and F-J , Figure 4-11).
In the one test in which emissions of H2S were detected (test B2),
an analytical method different from that employed by the Agency
was used and simultaneous testing by the Agency detected no emissions
(test 63). Review of both the Agency's test method and this other
test method indicated that this method which detected H2S emissions
was not as selective as the Agency's in identifyina I'2S, and was
including some of the COS and C$2 present as H?S.
The emission source test data available on emissions of reduced
sulfur compounds (H2S, COS and CS2) from these emission control
9.19
-------
systems shows that emissions of these pollutants are in the range
of 10-20 ppm (tests B-,, 82-,, B3, E-, and F-,, Figure 4-12). Here
again, however, this data was collected from these systems shortly
after major equipment turnarounds, or when they were operating
well below their design capacity (as low as 1/3 of design capacity
in one case).
Discussions with the vendors of these control systems also
indicated that unavoidable equipment deterioration and chemical
aging leads to a gradual increase in emissions with time. Pilot
plant data, for example, indicates that emissions increase by
about 200 ppm over a vear's operation. Thus, as discussed above
for the oxidation-scrubbing systems and the reduction-scrubbing
systems with tail gas incineration, basing the emission limits
for reduction-scrubbing emission control systems without tail gas
incineration solely on the emission data available would significantly
shorten the normal run length between major turnarounds and increase
the costs of maintenance and chemical replacement considerably.
Considering these factors, the limit on emissions of H2S
and reduced sulfur compounds from reduction scrubbing emission control
systems without tail gas incineration is set at 10 ppm and 300 ppm
respectively. The 10 ppm limit on H2S emissions will ensure that
installation of these emission control systems will not lead to a
local odor air pollution problem. The 300 ppm limit is equivalent
to the 250 ppm limit on S02 emissions in the sense that both limits
reduce sulfur emissions—be they S02 or H2S, COS and CS2—to the
9.20
-------
same level. (The 250 ppm S02 limit reflects the larger volume
of gases discharged to the atmosphere.)
9.6 SELECTION OF MONITORING REQUIREMENTS AND PERFORMANCE TEST METHODS
The objective of monitoring reouirements is to provide a quick
and easy means for enforcement oersonnel to ensure that an emission
control system; installed to comply with standards of performance
is DTOp-erly operated and maintained. For refinery sulfur recovery
olants» the most straightforward means of ensuring proper operation
ancf maintenance is to monitor emissions released to the atmosphere.
Consequently, where oxidation-scrubbing processes or reduction-
scrubbing processes with tail gas incineration are installed to
comply- with the standards, monitoring of S02 emissions is required.
Where reduction-scrubbing processes without tail gas incineration
are installed, monitoring of H£S and reduced sulfur compound emissions
is required.
Although monitoring requirements are included for l^S and
reduced sulfur compound emissions, the Agency has not yet developed
performance specifications for these monitors. Consequently, owners
and operators of reduction-scrubbing emission control systems without
tail gas incineration, who are subject to these requirements, will
not have to install these monitors until these specifications have
been promulgated in the Federal Register. The requirement for
monitoring is included in the regulations to ensure that when
these monitors become available, sources which became subject
to standards of performance before the monitors were available
will then be required to install monitors to aid enforcement personnel
in determining if the emission control system is being properly
operated and maintained.
9.21
-------
For determining compliance with the standards, Method 6 -
Determination of Sulfur Dioxide Emissions from Stationary Sources
will be used where oxidation-scrubbing processes or reduction-
scrubbing processes with tail gas incineration are installed.
Where reduction-scrubbing processes without tail gas incineration
are installed, Method 18 - Determination of Hydrogen Sulfide,
Carbonyl Sulfide and Carbon Disulfide Emissions from Stationary
Sources will be used. These methods were the methods used to gather
the emission data contained in chapter 4 and Appendix C, which support
the standards. Details as to why these methods were selected over
other methods for gathering this data mav be found in Appendix D.
9.7 REFERENCES
1. Telephone conversation, F.L. Porter (EPA) with G.L. Tilley
(Union Oil), April 28, 1976.
9.22
-------
APPENDIX A
EVOLUTION OF STANDARDS
A.I
-------
EVOLUTION OF STANDARDS
8/73
9/73
10/5/73
10/10-10/19/73
10/17/73
11/73
10/73-2/74
Surveyed and reviewed process operations and
emission control systems for all domestic
Glaus sulfur recovery plants.
Sent letters to control enuinment vendors
requestina desion data for the Wellman-Lord,
Beavon, Cleanair, SCOT, and Aquae!aus orocesses,
.and location of well-controlled Claus nlants.
Selected eioht refineries for initial nlant
inspections.
Contracted for detailed ennineerina studv of
Claus tail gas control systems v/ith Battelle-
Columbus.
Inspected eiqht refineries with well-controlled
Claus plants, pre-surveyed for emission testina,
and sent 114 letters to refineries.
Met with Los Anqeles APCD for discussion of
their regulations for sulfur recovery plants.
Contractor sent additional letters to vendors
for design data on the IFP-1, IFP-2, Sulfreen,
Cataban/and Chivoda Thoroughbred processes
for tail qas sulfur removal.
Test methods for determining gaseous sulfur.,
compounds in Claus tail gas ("i.e., COS, ,C.S2i
H2$, S0£, Sx) investigated and developed.
Presurveys made of three likely test sites.
1/11/74
2/25-3/13/74
Inspection made of IFP-1 process on a Claus
sulfur plant.
Emission tests completed for Mellman-Lord,
Beavon, and SCOT control svstems.
A.2
-------
4/1/74
5/7/74
6/10-6/12/74
6/74
8/74- 2/75
3/75
4/29/75
5/5-5/6/75
5-7/75
8-9/75
10/75-4/76
Inspection made of second IFP-1 orocess on a
Glaus sulfur plant. 114 letters on IFP-1
process sent to operators.
Meeting with API's Committee on Environmental
Affairs to discuss emission test.results and
Dotential standards.
Emission test of IBP-1 control system completed.
Study entitled "Characterization of Sulfur
Recovery From Refinery Fuel Has" completed
by BatteH e-Columbus Labs. Report circulated
to API for review.
Emission control costs developed, monitoring
and emission test methods finalized, and
dispersion analyses completed. Developed first
draft of EPA Standards Support and Environmental
Impact Document (SSEID).
Revised SSEID and sent to NAPCTAC and Workina
Group members.
Met with EPA Uorkinn Group to discuss findinas
of SSEID and the recommended standard.
Met with NAPCTAC to review the recommended
standard for refinerv sulfur plants and the
basis 'for the standard outlined in the SSEID.
Delayed action to resolve potential 111(d)
aspects of standards for refinery sulfur
plants.
Responsibility for SSEID informally trans-
ferred to Standards Development Branch.
SSEID reviewed, edited and rewritten by
Standards Develonment Branch to conform
with the general outline for a SSEID.
A.3
-------
-------
APPENDIX B
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS
B.I
-------
aa
X
0
2J
UJ
a.
CL.
:MPACT CONSIDERATIONS
t— I-
^
t—
z:
o
B£
5
z:
LU
O
H-
X
LU
a
z:
^•»4
s~
O +J
CL. C.
a. a
=r E
co a
•^
CO rt
s- c/~
OS
T34J
c a
OS ft
to E
i — i
a>
+j "ro
+J
4J E
T- O
3 S-
E >
O E
•.- LU
SI
O OS
D1^__
E to cn
•f- 4Ji—
S- E«d-
03 CU t--
cx S co
0) E
s- o o;
Q- S- U_
O E CO
Guidelines f
:ory Action E
Statements (
i-*
E 3 to
cu en a.
en cu E
E
O
£
O
«3T
•o
CU
CO
o
CL
£
Q-
<4-
O
E
O
•r—
4->
CL
•r—
S-
0
CO
S
-CJ
E
to
1 . Background
r
^~
£Z
O
•r-
•4-*
O
CU
CO
s-
01
tL
OS
o
•a
cu
si
as
CO
cu
s-
03
CO
03
TO
03
CO
cu
{5
CO
•a
s-
ta
•a
c
Proposed Sta
M-
0
ZJ
CO
c
•r~
E
cu
>
en
CO
•r—
-a
S-
03
-a
OS
c/r
cu
.0
M—
CO
CO
03
o
3CSX
03 S-
4-> CU
CO •*->
O.
CU 03
T3
i.
03
•a
03
CO
cu
CO
CO
OS
Statutory B
CO
•r—
•o
CU
O'
<»-
03
CU
a
•4-1
•4^
^
i
cu.
-l-> CO
I- £_
o cu
E CL
O 03
•43-5
n.,
si ••-
o
CO E
cu cu
-a >
=C co
-a
cu
CJ
cu
Facility Af
E
CU
>
• (—
co-
co-
••"
-o
€)
o,
cu
03
CU
0
CO
CO
s-
cx 01
si c.
U 03
CO JC
cu o
T3
cr
-d
cu
o
cu
CO
CO
cu
o
s
Q_
en.
a
"o
E '
-E
O
01-
Q
E
cr
o
^
0
•17-
r—
•i —
03
i — •
•r-
tO •
Information on the av
is given in chapter 4
^
01
"o
E
CJ
cu
£
o
CJ
<*-
o
Availabilit,
CO
S— -
CD
S- Ct
•4-> OS
3 O
TJ
E E
cuj "X31
C— ' Q^J
+J--T3
E i —
O CJ
CO t—
f-~
o co-
os CO
en to
CU T3
to
E- co
•4J CU
CO J=
o -o
cu
o o
•r- CU •
one f- co
co M—
rs to E
u o
CO CU •!-
•5 u
o cu
-a: +-> co
^
cu
cu
— i
03
o
1
ir
cu
+J
to
+J
CO
03
CO
O
•r-
03
en
cu
C£
en
E
4J
CO
X
LU
E
O
i <
-a
cu
CO
0
o
o
D-
CU
O
to
I. Alternative:
E
O
•£>
o
O
0
CO'
•o
OS
"O
c-
4->
CO
CU
jT
en
E
. y
E
£r'
cu: •
r— - t^**
Q.
03!
.IJ
O CO
CO
<*- S-
o cu
Environmental effects
are discussed in chap
CO
o
Cu
HH
"to
^
Environmei
cu
£-
O)
CO-
O'
t\
cu
si
03
S-
03
Jli
OJ
-4_»>
cu.
.E
4-=
en-.
•^
•4-* •
E i—
d) .
E 00
cu
"a. o
"~ u
•f-3 cu
O CO
E
n
<4- OO
CO CU
CO Q.
O 03
U .E
O
CU
(^ -S
CO
+J
CO
o
CJ
B.2
-------
INDEX TO ENVIRONMENTAL IMPACT CONSIDERATIONS (continued)
elines f or _ Preparing Location Within the Standards Support
Action Environmental and 'Environmental Impact Statement
T3 -1
5s--!
tn o o
>, to -
E 13
0) 05
en CU
lje - - The health/welfare impact of not implementing the
,d Welfare Impact standards is considered in chapter 7.
J E
o to
O c"
_J t —
J to
0 CU
3- :E
\
ve I - 99% Control
The environmental impacts associated with alternative -I
mental Impacts emission control systems are considered in chapter 7.
•r- w
4-> O
to i-
£ >
CU E
40 1 1 1
5
The cost impact of alternative I emission, control system
is considered in chapter 8, section 8.2.
CO
4J
tn
O
CJ
Tmnart The impact of alternative I emission control systems
and Welfare Impact on health and welfare is considered in chapter 7.
.E
r—
tO
CU
:n
-a
cu
S-
ar
-a
CO
c
o
CO
to
to
CO
cu
It-
CD •
o r-.
tO
Q. E
E 0
•I— •!—
E O
O CU
'r— CO
13 "
'o i-
O- CU
i- CL
•r- tO
tO .E
O
cu
.E E
1— ••-
E
O
CU
CO
o
O-
o
s-
0-
<*-
o
+J
0
to
Q-
1 — 1
r— E
to O
4-3 *r—
E 4J
E
0
•r-
^
UJ
CO
r-
"o
O-
i.
5
i
E
a
o
CO
•r—
E.
O
"o
O-
cu
tO CM.
E:
O E
O
in -i—
~S. (J
tO CU
"O CO
E
"t.
CU CU
^= 4->'
4J O-
tO
0 0
4-> E
O -i-
tO
Q--O
E CU
•r- $-
CU
cu ~o
\— tn
E
O
4J
3
i —
"o
Q_
S-
cu
4J
to
, The impact of the standard on solid waste disposal is
ste DlsP°sal considered in chapter 7, section 7.3.
The impact of the standards on energy use is considered
in chapter 7, section 7.4.
IO
3
-a
*o
00
?
cu
E
UJ
B.3
-------
-------
APPENDIX C
EMISSION! SfMECE TEST DATA*
C.I
-------
EMISSION SOURCE TEST DATA
C.I INTRODUCTION
This appendix summarizes the emfssion test data gathered during
the development of standards for refinery sulfur plants. Detailed
information on each facility tested is presented herein. Each
facility is identified by the -same coding used in Chapter 4. Any
reference in this appendix to commercial products or processes
by name does not constitute an endorsement by the Environmental
Protection Agency.
C.2 SUMMARY OF TEST DATA
Four different processes for removing sulfur from Claus
sulfur plant exhaust gases were tested by EPA to determine the
best available control technology as required by section 111
of the Clean Air Act. Pollutants measured included total sulfur
by EPA Method 18 (gas chromatograph/flame photometric detection),,
SOg by EPA Method 6, H2S by EPA Method 11, carbon monoxide .by
EPA Method 10, NOX by EPA Method 7, hydrocarbons by a flame
ionization detector, Orsat gases by EPA Method 3, and moisture
by EPA Method 4.
C.3 DESCRIPTION OF FACILITIES
Plant A - Plant A consists of three fdentfcal 150 long ton/day
(LT/D) Claus trains, two of which operate with the third on stand-by.
Emissions from each Claus train is controlled by a Wellman-Lord
scrubber. Design basis of these Wellman-Lord scrubbers is 250 ppmv
SOg. Tests Al and A2 were performed by EPA and refinery personnel,
respectively, during the period March 11-13, 1974. In Test Al
sulfur compounds were determined by EPA Method 18 (gas chromatograph/
flame photometric detection) for total sulfur and EPA Method 6 for S02,
C.2
-------
Carbon dioxide, carbon monoxide, and oxygen were determined by continuous
methods (non-dispersive infrared for C02 and CO and paramagnetic for 02)
and by the Orsat method. Nitrogen oxides were determined by EPA Method 7,
visible emissions by EPA Method 9, moisture and flow rates by EPA
Methods 1, 2, and 4, and hydrocarbon concentrations by a flame
ionization detector.
In Test A2 SO? and NO were determined by a fuel cell electro-
£ X
chemical method, COS and CS2 by a gas chromatograph flame photometric
detector, CO and C02 by Orsat, 02 by Orsat and a paramagnetic oxygen
analyzer, and hydrocarbons By a hydrogen flame gas chromatograph.
During Tests Al and A2 only one sulfur plant was operating
due to low refinery throughputs, caused by the OPEC oil embargo.
Sulfur feed rates during the tests averaged 113 LT/D for three runs.
Test A3 was conducted by the Los Angeles APCD, January 10-11, 1973,
on all three Claus plants. Hydrogen sulfide was determined by a ZnC03
impinger train, and S02 by impingers containing a 5% NaOH solution.
Nitrogen oxides, hydrocarbons, carbon monoxide, carbon dioxide,
moisture, and flow rates were determined according to methods described
In the Source Testing Manual of the Los Angeles APCD.
Sulfur feed to the three plants during Test A3 was 116.6, 76.9,
and 68.8 LT/D, respectively, averaging well below design rates.
plant B - Plant B consists of two parallel 100 LT/D Claus trains,
each of which exhausts into a Beavon tat! gas treating unit. Design
basis of each Beavon tail gas treating unit is 200 ppmv total sulfur,
with less than 10 ppmv H2S.
C.3
-------
Tests Bl and B2 were performed by EPA and refinery personnel,
respectively, during the period March 5-7, 1974. In Test Bl sulfur
compounds were determined by EPA Method 18 (gas chromatograph/flame photo-
metric detection) for total sulfur and EPA Method 6 for S02. Carbon
dioxide, carbon monoxide, and oxygen were determined by continuous
methods (a non-dispersive infrared instrument for C02 and CO, and
paramagnetic analyzer for 02) and by the Orsat method. Nitrogen
oxides were determined by EPA Method 7, visible emissions by EPA
Method 9, moisture and flow rates by EPA Methods 1, 2, and 4,
and hydrocarbon concentrations by a flame ionization detector.
In Test B2 mass spectrometry was used to determine CH4, COS,
S02> H£S, C$2, H2j CO/N2» 02, Ar, and C02. Gas chromatography
was then used to obtain the CO/N2 split.
During Tests Bl and B2 the sulfur plants were operating well
below design levels due to the OPEC oil embargo. Sulfur feed rate
to each or both Claus tra?n(s) averaged 34.2 LT/D for three runs.
Test B3 was conducted by the Los Angeles APCD on three separate
occasions: July 10, August 8, and September 24, 1974. The August 8
and September 24 tests were performed before and after overhaul to
determine the effect of overhaul on emissions.
Sampling techniques included a 5% HC1 solution in an impinger
to collect ammonia, a 3% ^2^2 solution in an impinger to collect
SOgj and a ZnCOs slurry in an impinger for H2S collection. Grab
samples were made for subsequent determinations of H2S, COS, and
CS2 by gas chromatograph with a flame photometric detector and
carbon monoxide by non-dispersive infrared absorption.
C.4
-------
Plant C - Plant C consists of one small 16 LT/D, two-stage
Claus unit followed by a SCOT tail gas treating unit. Design
basis of the SCOT unit is 400 ppmv total sulfur emissions calculated
as H2S. An incinerator oxidizes the 400 ppm H2S to S02 before
discharge to the atmosphere.
Test C3 was conducted by the Los Angeles APCD on February 14, 1974.
Hydrogen sulfide was determined by a ZnCOs train at the outlet of the
SCOT system. At the incinerator stack S02 was determined by an NaOH
train; combustion gases (CO, C02, and CH4) were analyzed by total
combustion analysis using a non-dispersive infrared instrument
and NOX was determined by hydrogen peroxide/sulfuric acid impinger
trains. Water vapor, gas flow and Orsat gases were also analyzed
using Los Angeles APCD methods.
During Test C3 sulfur feed averaged 11:1 LT/D.
plant D - Test Dl was conducted June 10-12, 1974,
by EPA on a large, 395 LT/D, three-stage Claus plant followed by
an IFP-1 tail gas process. The Claus nlant recovers sulfur from
acid gases produced in a carbon disulfide plant. Emission design
for the IFP-1 unit is 90 percent conversion of (H2S + S02) from the
Claus plant. This corresponds to 1640 ppmv total sulfur (wet) or
2540 ppmv total sulfur (dry).
In the first test total sulfur compounds were determined by EPA Method 18
(gas chromatograph/flame photometric detection). S02 was determined by
EPA Method 6, H2S by EPA 'Method 11, NOX by EPA Method 7, moisture
by EPA Method 4, gas flow by EPA Methods 1 and 2, and visible
emissions by EPA Method 9. C02, CO and 02 were determined by
C.5
-------
continuous methods (non-dispersive infrared for C02 and CO, para-
magnetic analyzer for 02). C0£ and 02 were also determined by a gas
chromatograph with thermal conductivity detection.
The second test was conducted using a Meloy Analyzer, which
oxidizes all sulfur compounds to S02 and then measures S02 by a
flame photometric detector. The third test was conducted using
a DuPont analyzer, which also oxidizes all sulfur compounds to
S02> but which then measures S02 by UV absorption.
During these tests, sulfur feed to the Claus unit averaged
336 LT/D.
Plant E - Plant E consists of two 96 LT/D Claus plants each with
a Beavon tail gas treating unit. Test El was conducted December 12,
1974, by the Los Angeles APCD. Analyses for COS, CS_, H-S, SO ,
, total sulfur, CO, and NOX were conducted. Details on test
methodology were not specified, though assumed the same as in previous
tests conducted by Los Angeles County (Tests A3 and B3). No process
data were available. The Beavon units were designed at 200 ppmv
total sulfur, 10 ppmv I^S emission levels.
Plant F - Plant F consists of a Beavon tail gas treating unit
which removes sulfur from four combined Claus plant tail gas streams
in a petroleum refinery. No design or operating data are available.
Los Angeles County conducted tests November 6, 1974, for total
sulfur, H2S, and carbon monoxide (test Fl). Again, test methods
are assumed to be Los Angeles APCD methods as described for
Tests A3 and B3.
C.6
-------
Table 1
FACILITY A
Summary of Results
Test Number
Run Number
Date
•Stack Effluent:
Flow rate - DNMJ/min
Water vapor - Vol. %
C02 - Vol. % drya
02 - Vol. % dry9
CO - Vol. % dry3
C02 - Vol
02 - Vol.
CO - ppmv dry
dryc
dryb
b
S02 - ppmv dryc
S02 - ppmv dry
COS - ppmv dry
CS? - ppmv dry
.
I^S - ppmv dry
TS - ppmv dry
NOX - ppmv dry6
THC - ppmv dry
Visible emissionsS
Al
1 2
3/11/74 3/12/74
197.1
13.0
4.3
0.9
95
5.9
38
3.2
2.5
<0.1
46.2
17.2
7.5
0
135.4
10.6
7.2
0.8
0.0
5.6
0.2
100
21.8
16
1.9
3.4
<0.1
24.7
9.0
6.2
n
3
3/13/74
209.7
11.2
5.35
2.95
0.0
3.8
1.5
39
7.4
10
0.9
1.1
13.1
21.0
4.6
0
Averagi
180.4
11.6
6.3
1.9
0.0
4.6
0.9
78
11.7
21
2.0
2.3
•f~ r\ i
^ w • 1
28
15.7
6.1
0
Orsat analysis
NDIR/Paramagnetic
CEPA-6
dGC/FPD (EPA-18)
SEPA-7
Total hydrocarbons as methane by flame ionization
9EPA-9
Source: Reference 1 ^"'
-------
Test Number
Run Number
Date
Stack Effluent:
Flow rate DNM3/MIN
Water vapor - vol.
C02 - vol. % drya
02 - vol. % dr.va
CO - vol. % drya
S02 - ppmv dryb
COS - ppmv dryc
CS2 - ppmv dryc
NOX - ppmv dryb
HC - ppmv dryd
TABLE 2
FACILITY A
Summary of Results
A2
1 2
3/12/74 3/13/74
6.6
1.3
0.3
10
0.3
1.5
21.7
3.0
15
25
Averane
6.6
1.3
0.3
12.5
0.3
1.5
23.4
3.0
Orsat analysis
fuel cell electrochemical
CGC/FPD (EPA-18)
Hydrogen flame chroma!oaraphy
Source: Reference 2
C.8
-------
TABLE 3
FACILITY A
Summary of Results
Test Number
•Run Number
Date
Stack Effluent:
Flow rate, DNM3/M
Water vapor - vol.
C02 - vol. % wet
CO - vol. % dry
SOg - ppmv dry
H2S - ppmv
COS - ppmv
C$2 - ppmv
NOX - Ib/hr
HC - Ib/hr
1
1/10/73
229.37
14.0
16.0
0.36
31
<.10
15
13
0.57
0.90
A3
2
1/11/73
150.08
10.0
16.0
0.20
38
<.10
1
2
0.59
0.44
3
1/11/73
127.43
12.0
19.0
0.067
47
1.7
<1
1
0.46
0.21
Averaqe
169.05
12.0
17.0
0.41
40
0.6
6.0
5.7
0.54
0.52
Source: Reference 3
C.9
-------
TABLE 4
FACILITY B
Summary of Results
Test Number
Run ftaber
Date
Stack Effluent:
Flow rate, DNM3/M
Water vapor - vol
C02 - vol. % drya
02 - vol. % drya
CO - vol. % drya
C02 - vol. % drvb
02 - vol. % dryb
CO - vol. % dryb
S02 - pwnv dryc
S02 - ppmv drvd
COS - ppmv dryd
CS2 - pomv dryd
H2S - ppmv dryd
TS - ppmv dryd
NOV - ppmv dry6
* f
THC - ppmv dry7
Visible emissions
1
3/5/74 .
65.5
4.2
5.4
0.6
0
5.8
0.02
566
3.6
1.5
17
0.15
19
1.1
Bl
2
3/6/74
71.6
5.0
5.5
0.5
0
5.7
0.09
565
3.8
0.7
17
-
17
0
3
3/7/74
68.8
3.3
6.0
0.3
0
5.9
0.02
604
4.5
0.76
15
-
16
0
Average
68.6
4.2
5.6
0.5
0
5.8
0.04
578
4.0
1.0
16
-
17
0.4
0
Orsat analysis
NDIR/Paramagnetic
cEPA-6
dGC/FPD (EPA-18)
eEPA-7
Total hydrocarbons as methane by flame tonization
9EPA-9
Source: Reference A
C.10
-------
TABLE 5
FACILITY B
Summary of Results
Test Number
Run Number
Date
Stack Effluent:
Flow rate, DNM3/M
Water vapor - vol,
H2 - vol. %
CO - ppm
CH4 - ppm
N2 - vol. %
02 - vol. %
H2S - ppm dry
Ar - vol. %
COo -vol. %
COS - ppm dry
SOg - ppm dry
C$2 - ppm dry
1
3/5/74
5.0
479
125
87.7
0
7
1.0
63
9
0
0
B2
2
3/6/74
6.0
620
206
87.0
0
1
1.0
6.0
9
0
0
3
3/7/74
5.8
595
332
86.9
0
0
1.0
6.3
9
5
0
Averaae
5.6
565
221
87.2
0
2.7
1.0
6.2
9.0
1.7
0
Source: Reference 5
c.n
-------
TABLE 6
FACILITY B
Summary of Results
Test Number
Run Number
Date
Stack Effluent:
Flow rate, DNM3/M
Water vapor - vol,
C02 - vol. %
CO - ppm
S02 - ppmv dry1
COS - ppmv dry1
CS2 - ppmv dry1
- ppmv dry1
B3
1
7/10/74
2
8/8/74
3
9/24/74
4
9/24/74
Averane
-
0
23.0
9.7
0
-
0
16.0
0
0
346
0
6.8
0
0
335
0
7.0
0
0
341
0
13.0
2.4
0
Source: Reference 6
Note:
T "Assume moisture level of 5.6% based on test 62-
C.12
-------
Table 7
FACILITY C
Summary of Results
Test Number
Run Number
Date
Stack Effluent:
Flow rate - DNM3/M
I-LS - ppmva
u c b
H9S - ppmv
S02 - ppmva
S02 - Ib/hr1
SO
NO
x
- Whr
b
- ppmv
CO - ppmv
02 - Vol.
C02 - Vol .
HC - Vol .
2/14/74
11.33
197
<10
0
5.5
0
14
1500
12.5
Incinerator inlet
Incinerator outlet
Source: Reference 7
C.13
-------
Test Number
Run Number
Date
Stack Effluent:
\
Table 8
FACILITY D
Summary of Results
(See below)
1 2 3
6/10/74 6/11/74 6/12/74
Average
Flow rate - DNM°/min
Water vapor - Vol . %
C02 - Vol . % drya
02 - Vol . % drya
CO - Vol. % drya
C02 r Vol . % dry5
02 - Vol. % dryc
CO - ppmv dry
S02 - ppmv dry
S02 - ppmv dry6
S02 - ppmv dry
COS - ppmv dryf
H?S - ppmv dry
f
CSp - ppmv dry
TS (l)""""- ppmv dryf
TS (2) - ppmv dry6
TS (3) - ppmv dry9
NO - ppmv dry
421
30.9
1.3
0.28
_
1.5
0.1
3240
59
420
82
132
1190
460
2310
2390
2380
7.8
431
37.7
1.1
0.35
_
1.6
_
2450
42
430
72
77
1410
180
1920
2590
2540
1.0
414
39.0
1.1
0.31
—
1.8
0.1
3140
42
360
80
133
1950
300
2760
_
2070
4.0
422
35.9
1 .2
0.31
..
1.6
0.1
2940
48
400
78
114
1520
310
2330
2490
2330
4.3
Visible emissions1
aOrsat
bNDIR
paramagnetic
dEPA-6
Q
DuPont Analyzer(includes elemental sulfur)
fGC/FPD (EPA-18)
9Meloy Analyzer
hEPA-7
EPA-9
Source: Reference 8
C.14
-------
Table 9
FACILITY E
Summary of Results
Test Number
Run Number
Date
Stack Effluent:
Flow rate, DNM /tnin
Water vapor - Vol. %
COS - ppm dry1
CS2 - ppm dry1
H2S - ppm dry1
S02 - ppm dry1
S04 - ppm wet
H2S04 - ppm wet
TSa - ppm
TSb - ppm
CO - ppm wet
N0v(as N0?) - ppm wet
X £•
El
1
12/12/74
5
0.5
<1
<0.4
2
<0.3
8+
14
250
I
aTotal of separately measured constituents
bTotal as measured by sulfur detector (includes mercaptans)
Source: Reference 9
Note:
1. Assume moisture level of 5.6% based on test 82-
C.15
-------
Table 10
FACILITY F
Summary of Results
Test Number
Run Number
Date
Stack Effluent:
Flow rate - DNM3/M
COS - ppm dry
C$2 - ppm dry
H2S - ppm dry
S02 - ppm dry
Total sulfur - ppm dry
CO - ppm dry
Source: Reference 10
Fl
11/6/74
311.49
15.2
0.1
0
0
15.4
670
C.16
-------
References
1. Source Test Report No. 74-SRY-l, EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Engineering, Gainesville,
Fla., March 1974.
2. Letter, Thron Rigqs, Standard Oil Co. of California to C. Sedman,
ESED, OAQPS, EPA, dated August 9, 1974.
3. Source Testing Section Report No. C-1895, Los Anqeles County APCD,
February 28, 1973.
4. Source Test Report No. 74-SRY-2, EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Enqineerina, Rainesville,
Fla., March 1974.
5. Letter, George L. Tilley, Union Oil Company of California, to
C. Sedman, ESED, OAQPS, EPA, dated Auaust 26, 1974.
.6. Source Testing Section Report No. C-2082, Los Anaeles County,
APCD, Dec. 27, 1974.
7. Source Test Section Report Mo. C-2104, Los Angeles County, APCD,
April 25, 1974.
8. Source Test Report No. 74-SRY-4, EPA Contract No. 68-02-0232,
Task Order No. 34, Environmental Science and Engineering Gainesville,
Florida, June 1974.
9. Source Testing Section Report No. C-2234, Los Anaeles County APCD,
Feb. 20, 1975.
10. Source Testing Section Report No. C-2226, Los Angeles Countv APCD,
Nov. 6, 1974.
C.17
-------
-------
APPENDIX D
EMISSION MEASUREMENT AND CONTINUOUS MONITORING
D.I
-------
D.I EMISSION MEASUREMENT METHODS
A review of the literature revealed that four different
analytical methods could be used for analysis of sulfur compounds:
colorimetry, coulometry, direct spectrophotometry, and gas
chromotography. Although these methods were developed in most
cases for measurement of ambient air concentrations, this did not
preclude their application to measurement of stack gas emissions.
Colorimetr.y. In this method a sample is bubbled through a
solution which selectively absorbs the component or components
desired. The solution is then reacted with specific reagents to
form a characteristic color which is measured spectrophotometrically.
An example of a colorimetric method is the methylene blue
method which involves the absorption of reduced sulfur compounds
in an alkaline suspension of cadmium hydroxide to form a cadmium
sulfide precipitate. The precipitate is then reacted with a
strongly acidic solution of N, N dimethyl-P-phenylene-diamine and
ferric chloride to give methylene blue, which is measured spectro-
photometrically. Automated sampling and analytical trains using
sequential techniques are available for this procedure. Inherent
deficiencies for stack sampling applications however include
variable collection efficiency, range limitations, and interference
from oxidants.
Another colorimetric method is the use of paper tape samplers
impregnated with either lead acetate or cadmium hydroxide. These
compoundsreact specifically with H2S to form a colored compound
which can be measured directly with a densitometer. Tape samplers
D.2
-------
would not be appropriate for all reduced sulfur compounds unless
they were first reduced quantitatively to H^S. In addition, the
range is limited, the method requires precise humidity control and
suffers from light sensitivity, fading, and variability in tape
response.
Coulometry. In this method a gas sample is bubbled through
a solution containing an oxidizing or reducing agent (titrant).
The concentration of the titrant in solution is buffered by the
presence of a titrant precusor. Passage of an electric current
through the solution causes the titrant precusor to break down,
releasing additional titrant into solution. Consequently, as the
titrant is consumed by reaction with specific compounds contained
in the gas sample, an electric current is passed through the
solution to maintain the titrant concentration. The electric
current required is a measure of the reactive compounds in the
gas sample. Normally, the titrant is a free halogen such as
bromine or iodine in solution as an oxidizing agent, or a metal ion
such as silver in solution as a reducing.agent.
For determining emissions coulometric titration has the
advantage of responding to a wide variety of sulfur compounds. The
response to each compound is quite different, however, and this
makes standardization and reporting of data difficult in many cases.
In addition, the method suffers from high maintenance and requires
frequent calibration to reduce drift to acceptable levels.
Spectrophotometry. Although infrared and mass spectrophotometric
methods are well established analytical techniques, most of these
D.3
-------
seven-fold that to S02> S02 cannot be analyzed by this method
if appreciable elemental sulfur is present. The Model 464
has not yet been used as a continuous monitor in sulfur recovery
pi ants.
Gas cJiromatography/flame photometric detection (GC/FPD) is
another method for monitoring SC^ emissions. Systems using
this principle include the Bendix Model 8700, Tracer Model 250H.
These systems cost about the same as the DuPont 464 system discussed
above and are also semi continuous in operation. Recently, however,
one vendor announced a complete sampling, analysis, and recording
system for $14,000 (Tracer Model 270H). Again, automated data
reduction can be added at additional cost. Integrators compatible
with GC analyzers can be programmed to print the concentration of each
sulfur compound. Cost of these integrators is in the range of
$3,000-$49000.
The GC/FPD system has several advantages. It can separate and
detect each individual sulfur comoound. These systems are extremely
sensitive, however, and require sample dilutions of 100:1 or more.
This presents a potential source of error and frequent calibration
is necessary to minimize such errors.
Other continuous monitoring instruments are commercially
available. Many of these are summarized according to their
capabilities by Nader et.al. (EPA-650/2-74-013). To date, however,
they have not been evaluated for use on sulfur plants.
D.6
-------
For montiroing emissions of total sulfur and H^S, the two
systems described previously can be used. The Da-Font Model 454
ultraviolet analyzer is capable of oxidizing the gas sample and
measuring all sulfur compounds including elemental sulfur as SCh?.
This system, however, is not able to monitor H^S emissions.
The GC/FPD system: is capable of monitoring individual sulfur
emissions (except sulfur vapor). Total sulfur can be determined
by adding the individually measured components tp the estimated
sulfur vapor. Sulfur vapor may be calculated from the partial
pressure of sulfur at the gas stream temperature. Since H^S
vapor is one of the compounds determined by GC/FPD„ it can be
reported separately.
Although continuous monitors are available to monitor emissions
of reduced sulfur compounds, compliance with the monitoring
requirements included in the standards will he delayed until
EPA promulgates performance specifications for these monitors.
Since the standards specify that emissions must be determined
at zero percent oxygen, continuous monitorinq of the oxygen concen-
tration in the tail gases discharged to the atmosphere is required.
A number of systems are available to monitor oxygen concentration
and performance specifications for these systems were promulgated
by EPA in 40 FR 46268 on October 6, 1975.
D.3 PERFORMANCE TEST METHODS
EPA Method 18, "Semicontinuous Determination of Sulfur
Emissions from Stationary Sources," has been prepared for use
in determining compliance with new source performance standards at
refinery sulfur plants. This method requires use of the GC/FPD
D.7
UU—UNC—[J UfcM-l-Vj—I-I=CC—U-l \^l-Itl-l-^C I-I-UIII . =
Public Information Center (PM-21:5), EPA,
2O. SECURITY CLASS (TKispage)
Unclassified
22rPRl'CE-
EPA Form 2220-1 (9-73)
-------
system discussed above and utilized during the emission testing
-------