United States Industrial Environmental Research EPA-600/2-79-048
Environmental Protection Laboratory February 1979
Agency Research Triangle Park NC 27711
v>EPA
Research and Development
Acrylonitrile Plant Air
Pollution Control
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental
Protection Agency, have been grouped into nine series These nine broad cate-
gories were established to facilitate further development and application of en-
vironmental technology. Elimination of traditional grouping was consciously
planned to foster technology transfer and a maximum interface in related fields.
The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
bS aSSl(?ned t0 th6 ENVIRONMENTAL PROTECTION TECH-
o ^?-SerieS describes research performed to develop and dem-
^IOni.equlpment' and m*hodology to repair or prevent en-
pos h new P°int 3nd n°n-point sour'es of P°llution This work
of PoSion ^ rPr°Ved technol°9y required for the control and treatment
or pollution sources to meet environmental quality standards.
EPA REVIEW NOTICE
iorU'S> Environmen'al Protection Agency, and
reflect the views and niP, does not si9nifVthat the contents necessarily
commercial nrodurtc EJL"?.0! the A9encV- nor does mention of trade names or
products constitute endorsement or recommendation for use.
*rough the National Technical Informa-
-------
EPA-600/2-79-048
February 1979
Acrylonitrile Plant
Air Pollution Control
by
M.T. Anguin and S. Anderson
Acurex Corporation
685 Clyde Avenue
Mountain View, California 94042
Contract No. 68-03-2567
Program Element No. 1AB604
EPA Project Officer: In/in A. Jefcoat
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, NC 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, DC 20460
-------
TABLE OF CONTENTS
Section Page
1 SUMMARY 1-1
2 INTRODUCTION 2-1
3 SOURCE DESCRIPTION AND CURRENT CONTROL PRACTICES 3-1
3.1 INDUSTRY OVERVIEW -. . 3-1
3.2 SOHIO PROCESS 3-2
3.2.1 Process Description 3-2
3.2.2 Waste Stream Description 3-8
3.2.3 Control Technologies 3-19
3.3 MONTEDISON PROCESS 3-22
4 ALTERNATE CONTROL METHODS 4-1
4.1 INTRODUCTION 4-1
4.2 FLARES 4-2
4.2.1 Technical Description 4-2
4.2.2 Suitable Waste Streams 4-3
4.2.3 Technical Feasibility 4-4
4.2.4 Efficiency, Cost, Reliability, and Energy
Requirements 4-4
4.2.5 Conclusions 4-6
4.3 THERMAL INCINERATION 4-8
4.3.1 Technical Description 4-8
4.3.2 Suitable Waste Streams 4-13
iii
-------
TABLE OF CONTENTS (Continued)
Section
4
(cont)
Page
4.3.3 Technical Feasibility 4-13
4.3.4 Efficiency, Cost, Reliability, and Energy
Requirements 4-18
4.3.5 Conclusion 4-19
4.4 CATALYTIC INCINERATION 4-20
4.4.1 Technical Description 4-20
4.4.2 Suitable Waste Streams 4-24
4.4.3 Technical Feasibility 4-24
4.4.4 Efficiency, Cost, Reliability, and Energy
Requirements 4-27
4.4.5 Conclusion 4-27
4.5 CARBON ADSORPTION 4-28
4.5.1 Technical Description 4-28
4.5.2 Suitable Waste Streams 4-31
4.5.3 Technical Feasibility 4-31
4.5.4 Efficiency, Cost, Reliability, and Energy
Requirements 4-34
4.5.5 Conclusion 4-34
4.6 SOLVENT EXTRACTION AND ABSORPTION 4-35
4.6.1 Technical Description 4-35
4.6.2 Suitable Waste Streams 4-35
4.6.3 Technical Feasibility 4-36
4.6.4 Efficiency, Cost, Reliability, and Energy
Requirements 4-38
4.6.5 Conclusion 4-38
IV
-------
TABLE OF CONTENTS (Concluded)
Section Page
5 GENERAL CONSIDERATIONS 5-1
5.1 CONTROL METHODS FOR SPECIFIC STREAMS 5-1
5.1.1 Absorber Vent Gas Stream 5-1
5.1.2 Liquid Stream on Way to Holding Pond 5-1
5.1.3 HCN and Acentonitrile Incinerators 5-1
5.1.4 Startup Emission Stream 5-2
5.2 RESEARCH AND DEMONSTRATION RECOMMENDATIONS 5-2
5.2.1 Absorber Vent Gas Stream 5-2
5.2.2 Liquid Stream on Way to Holding Pond 5-?
5.2.3 HCN and Acetonitrile Incinerators 5-3
5.2.4 Startup Emission Stream 5-4
REFERENCES R-l
APPENDIX A A-l
APPENDIX B B-l
LIST OF ILLUSTRATIONS
F i gure Page
3-1 Flow Diagram for a Representative Acrylonitrile Plant . . 3-4
4-1 Coupled Effects of Temperature and Time on Rate of
Pollutant Oxidation 4-9
4-2 Direct-Fired Afterburner with Tangential Burner
Arrangement 4-11
4-3 Low NOX Emission Incineration 4-12
-------
LIST OF ILLUSTRATIONS (Concluded)
Figure Page
4-4 Species Flux vs. Time, Flame 3 4-15
4-5 Comparison of the Relationship of Equivalence Ratio
and NO Yields 4-17
4-6 Schematic Diagram of Catalytic Afterburner 4-21
4-7 Schematic of a Catalytic Abater in Use at
Acrylonitrile Plant 4-22
4-8 Catalytic Incinerator 4-23
4-9 Effect of the Adiabatic Flame Temperature on the
Conversion of NH3 to NO 4-26
4-10 Acrylonitrile Adsorption on PPL Activated Carbon .... 4-30
LIST OF TABLES
Table Page
3-1 Stream Codes for Figure 3-5
3-2 Acrylonitrile REactor System Heat Balance 3-7
3-3 Material Balance for a Representative Arcylonitrile
Plant 3-10
3-4 Streams to and from Catalytic Oxidizer 3-13
3-5 Acrylonitrile Plant Waste Water 3-15
3-6 Untreated Process Water Load Discharged to Deep Wells . . 3-16
3-7 Streams to and from Incinerator 3-17
3-8 Startup Emission Stream 3-18
3-9 Control Technologies for Emission Streams 3-20
4-1 Comparison of Startup and Continuous Flow Gas Streams . . 4-5
VI
-------
SECTION 1
SUWARY
The Industrial Environmental Research Laboratory is responsible for
performing the research and development required to assess the impacts of
pollution from a variety of industries and to evaluate the applicability
of various control technologies for these industries. Pollution control
options must be evaluated for efficiency, reliability, economics, and
energy consumption. If secondary pollutants are generated by the cleanup
of the original pollutant, their impact must also be assessed. This
report on acrylonitrile plants addresses these aspects of control
technology evaluation.
The purpose of the report is to provide data for making decisions
about control technology. Control technologies are identified and ranked
in terms of efficiency, cost, and energy requirements. Control technology
demonstration opportunities in the acrylonitrile industry are also
identified.
There are six operating acrylonitrile plants in the U.S. Each has
several air pollutant emission sources. The effluent streams addressed in
this report are:
• The absorber vent gas stream
• The liquid waste streams that go to the holding ponds and
deep-well ponds
-------
• The HCN and acetonitrile incinerators and their off-gas streams
• The reactor startup emission streams
The absorber vent gas stream, when unabated, emits large quantities
of hydrocarbons. Thermal incineration is used for abatement of this
stream at one acrylonitrile plant, and catalytic incineration is used at
another plant. Data for these streams and their abatement by the
incineration processes were available from EPA contractors. Using these
data, the effectiveness of catalytic and thermal incineration was
evaluated. A quick review of the literature showed other methods to be
unsuitable: carbon adsorption because the pollutants are too low in
molecular weight, and hydrocarbon absorption because the stream is too
dilute. It was concluded that thermal incineration with waste heat
recovery is the best method for abatement of this stream; catalytic
incineration has a high unburned-hydrocarbon passthrough rate.
High levels of hydrocarbon emission occur from the holding ponds.
There are no reasonable pollution control technologies for open ponds, but
there are control technologies for hydrocarbon removal from waste water on
its way to the ponds. A review of studies and demonstration projects on
solvent extraction of organic nitrogen containing waste waters was made.
In addition, a patent for changing the acrylonitrile processing to
eliminate water scrubbing of the product was reviewed. This would also
eliminate most of the waste water production. These methods are still in
the research and development stage, and conclusions about their efficacy
cannot be drawn.
All acrylonitrile plants have HCN and acetonitrile thermal
incinerators; the emissions data available (from other EPA contractors)
for the exit streams from these incinerators showed 0.6 percent conversion
1-2
-------
of fuel nitrogen to NO . A review of the combustion literature revealed
A
that 20 to 80 percent of conversion of fuel nitrogen to NO could be
A
expected. (This discrepancy should be resolved by further study.)
Catalytic incinerators were evaluated as replacements for the existing
incinerators; a literature review shows that similar levels of NO
A
production could be expected.
When reactors at acrylonitrile plants are started up, the emissions
from these reactors are vented directly to the atmosphere. To control
this intermittent pollution stream, which contains up to 10,000 Ibs of
acrylonitrile per reactor per emission, flares and carbon adsorption were
evaluated. Flares (and other combustion methods) form unacceptable
amounts of NO . Carbon adsorption, and wet scrubbing followed by carbon
adsorption, appear to be more effective.
This report presents the following conclusions:
• Absorber vent stream: Thermal incineration is an acceptable
and efficient control method. Thermal incinerators are
currently in use, and no further development is required.
• Holding pond: Extraction of hydrocarbons from the waste water
before it is sent to a holding pond is the most desirable
control method. Bench and pilot-plant scale research on carbon
adsorption and hydrocarbon absorption (solvent extraction) is
recommended. A literature review of the waste water control
methods in use in Europe (e.g., the Montecatini plant) is also
recommended.
• Hydrogen cyanide/acetonitrile incinerators: Investigation of
the NO production of the existing incinerators is
X
recommended. Threre is a potential for high levels of NO
A
1-3
-------
emissions from these incinerators. A feasibility study of
advanced incineration techniques — two-stage (low NO )
A
thermal and catalytic incinerators — is also recommended.
Startup emissions: A study of the feasibility of routing
startup emissions to the absorber tower for scrubbing and a
demonstration of a combined wet-scrubber and carbon adsorption
abatement technique are recommended.
1-4
-------
SECTION 2
INTRODUCTION
The Industrial Environmental Research Laboratory is responsible for
performing the research and development required to assess the impacts of
pollution from a variety of industries and to evaluate the applicability
of various control technologies for these industries. Pollution control
options must be evaluated for efficiency, reliability, economics, and
energy consumption. If secondary pollutants are generated by the cleanup
of the original pollutant, their impact must also be assessed. This
report on acrylonitrile plants addresses these aspects of control
technology evaluation.
The exposure of the general public to emissions from acrylonitrile
plants has recently been evaluated by Monsanto Research Corporation under
contract to EPA. The result of that study was a Source Assessment
Document (SAD) (Reference 1), which describes the effluent streams from a
typical acrylonitrile plant and assesses the pollution problems of the
industry.
This report evaluates the control technologies available for the
effluent streams identified in the SAD. Where adequate control
technologies do not exist, this report evaluates the need to develop new
methods of control. To avoid redundancy, this document is organized
according to control technologies instead of waste streams. Section 3
2-1
-------
describes the waste stream sources and current industrial control
practices. Section 4 discusses alternative control methods. Section 5
includes a comparative evaluation of control technologies for each waste
stream and presents recommendations for further study.
2-2
-------
SECTION 3
SOURCE DESCRIPTION AND CURRENT CONTROL PRACTICES
3.1 INDUSTRY OVERVIEW
Acrylonitrile is a colorless liquid with a mild odor. It boils at
77.3°C. The molecular structure is:
The most common process for making acrylonitrile uses agricultural grade
ammonia and propylene as raw materials.
Acrylonitrile is a feedstock used to produce polymers for the
manufacture of acrylic fabrics and some synthetic rubber. The compound is
also used to make the plastic resins acrylonitrile-butadiene-styrene (ABS)
and styrene-acrylonitrile (SAN). Six plants produced 1.5 billion pounds
of acrylonitrile in the United States in 1977. Annual production is
projected to reach 2.5 billion pounds by 1981 (Reference 2). All six
American plants use the SOHIO process described in Section 3.2.
In 1974 a report by Air Products Corporation predicted that by 1982
acrylonitrile plants would rank second in total annual emissions from all
sources and first among the major petrochemical processes when those
emissions are weighted according to the amount and toxicity of the
effluent constituents. (In making this determination, hydrocarbons were
3-1
-------
given a weighting factor of 80, all particulates 60, and carbon
monoxide 1.) (Reference 3).
Evidence indicates that acrylonitrile is a potential public health
hazard (Reference 2). Acute toxicity has been observed in workers exposed
to 16 to 100 ppm acrylonitrile levels for twenty to forty-five minutes.
Acute intoxication resulting in death has been reported in a case where
children slept in a room that had been disinfected with acrylonitrile.
Acrylonitrile has been shown to be carcinogenic in laboratory
studies with animals (References 4 and 5). In addition, preliminary
epidemiological data indicate that it may cause cancer in humans. The
health records of 470 male employees working in the acrylonitrile
polymerization area of a plant were studied. These men breathed
acrylonitrile vapors. Sixteen cancer cases occurred in this group, while
only seven could be expected based on national cancer rates.
The Occupational Safety and Health Administration recently
estimated that a capital expenditure of $4,740,000 would be required in
order to lower the ambient concentration of acrylonitrile within a typical
plant to 1 ppm. The typical plant would have 167 employees and produce
255 million pounds of acrylonitrile a year (Reference 2).
3-2
-------
3.2 SOHIO PROCESS
3.2.1 Process Description
As previously noted, all six acrylonitrile plants in this country
use the SOHIO process. (Figure 3-1). The process streams are listed in
Table 3-1. The basic chemical equation for the formation of acrylonitrile
is:
2 CH2 = CH-CH3 + 2 NH3 + 3 02 — 2 CH2 = CH-CN + 6 H20
A bismuth-molybdenum catalyst (catalyst 41) currently is used; this
type of catalyst replaced a uranium catalyst, (catalyst 21) which replaced
an older version of the bismuth-molybdenum catalyst (catalyst A).
(Reference 6).
Approximately stoichiometric proportions of air, ammonia, and
propylene are introduced into a fluidized bed reactor. Once-through flow
is used since conversion of propylene is virtually complete. The reaction
is exothermic so heat removal must be provided. The heat of reaction is
normally used to generate steam by heat exchange and the effluent is then
sent to a water quench tower where acid is added to neutralize unconverted
ammonia. The stream then flows through a water absorber-stripper to
reject inert gases and recover reaction products. The operation yields a
mixture of acetonitrile, acrylonitrile, and HCN. The mixture is first
distilled to remove acetonitrile and water. Next, acrylonitrile is
separated from the HCN by fractionation. The final two steps involve
drying of the acrylonitrile stream and a final distillation to remove
heavy ends. The acrylonitrile product thus obtained is 99+ percent pure.
Primary raw materials for acrylonitrile production are agricultural
grade ammonia and propylene (more than 90 percent Co). No alternative
raw materials are available for the SOHIO process. Impurities in the
3-3
-------
CO
I
WASTE HEAT
BOILER
COOLING COILS
DENOTES MAIN PRODUCT ROW
DENOTES ALL OTHER STREAM FLOW
1
|te
|
O
)
ILE COLUMN \-i I
J
h-
(s
S COLUMN 3"~ *j
®
®
n
1
-1
®
TO DEEP WELL
T_J
i
CRUDE
A'CRYIONITRIU
"-[STORACEj-1
ABSORBER VENT CAS
-». RARE
-»» FUGITIVE EMISSIONS
ROOUCT
ACRYlONITRILt
STORAGE h
PRODUCT TRANSPORT
LOADING
FACILITY
INCINERATOR STACK CAS
DEEP WELL POND EMISSIONS
STORAGE TANK EMISSIONS
PRODUCT TRANSPORT LOADING
FACILITY VENT
TANK TRUCK
RAILROAD CAR
Figure 3-1. Flow Diagram for a Representative Acrylonitrile Plant
(Reference 1)
-------
TABLE 3-1. STREAM CODES FOR FIGURE 1 (Reference 1)
Stream Number Description
1 Propylene feed
2 Ammonia feed
3 Process air
4 Reactor feed
5 Reactor product
6 Cooled reactor product
7 Quenched reactor product
8 Sulfuric acid
9 Stripping stream
10 Waste water column volatiles
11 Waste water column bottoms
12 Absorber vent gas
13 Acrylonitrile plant waste water
14 Absorber bottoms
15 Water recycle
16 Crude acetonitrile
17 Crude acrylonitrile
18 Recovery column purge vent
19 Acetonitrile column bottoms
20 Acetonitrile
21 Hydrogen cyanide
22 Light ends column purge vent
23 Light ends column bottoms
24 Product acrylonitrile
25 Heavy ends
26 Product column purge vent
27 Flare
28 Fugitive emissions
29 Incinerator stack gas
30 Deep well pond emissions
31 Storage tank emissions
32 Product transport loading facility vent
3-5
-------
propylene feed with less than four carbon atoms are unaffected by the
reaction, but those with four or more participate and, therefore, are
undesirable.
Acetonitrile and hydrogen cyanide are produced as by-products in
the amount of about 0.1 pound of each per pound of acrylonitrile. These
products can be produced at saleable purity (99 percent purity); however,
they cannot always be marketed. Parts of each of these by-products are
usually incinerated (Reference 6).
Streams 1, 2 and 3 (propylene feed, ammonia feed and process air)
combine to form stream 4, the reactor feed. The reactor operates at 135
to 310 kPa and 400°C to 510°C as a fluidized bed; the catalyst is the
solid phase and the reactants and products are the vapor phase. Cooling
coils extract heat from the reactor.
Further heat is extracted from the product gases (stream 5) in a
waste heat boiler. Table 3-2 shows a heat balance for an acrylonitrile
reactor system.
Although the stoichiometric ratio of propylene/ammonia/air would be
1/1/7.14, the actual ratio fed in to the reactor is 1/1.06/8.40. The
excess ammonia forces the reaction closer to completion, and the excess
air continually regenerates the catalyst (Reference 1).
The reactor effluent stream is cooled in the heat exchanger of a
waste heat boiler (stream 6). The effluent next goes to the quencher
where sulfuric acid is added. The sulfuric acid reacts with the excess
ammonia to form ammonium sulfate. Catalyst fines are also scrubbed into
the liquid acid. Bottoms from the quencher are sent to the waste water
column, where the volatiles are stripped with steam and returned to the
quencher (stream 10). The residual (stream 11) goes to the deep well pond.
3-6
-------
TABLE 3-2. ACRYLONITRILE REACTOR SYSTEM HEAT BALANCE9 (Reference 1)
Heat out
kJ/g acrylonitrile
Steam generation
Cooling coils inside reactor
Waste heat boiler
Reactor heat losses
Quench (232°C to 43°C)
Incremental effluent heat content
15.66
5.02
0.10
9.52
-2.33
TOTAL
27.97
Heat in
kJ/g acrylonitrile
Exothermic heat of reaction
Acrylonitrile formation
Effluent neutralization
Feed vaporization and preheat
21.93
3.20
2.84
TOTAL
27.97
aBasis: Data shown later in Table 12; feed preheated to 149°C;
reactor outlet temperature 510°C (max.).
Difference in heat content of reactor product at 43°C and feed at
27°C (liquid) plus air at 38°C.
3-7
-------
Volatiles from the quencher go into the absorber, a tray-type
column where the acrylonitrile product and by-products (hydrogen cyanide
and acetonitrile) are absorbed by water. Absorber columns are operated
with and without auxiliary cooling. Most plants vent gas from the
absorber directly to the atmosphere (stream 12), but one plant uses a
thermal incinerator, and another plant uses a catalytic incinerator. The
absorber bottoms, water with absorbed product and by-products (stream 14),
are sent to the recovery column.
In the recovery column, the absorber bottoms are separated into two
streams: crude acrylonitrile (stream 17) and crude acetonitrile
(stream 16). The crude acetonitrile is sent to the acetonitrile column
where it is separated into acetonitrile (stream 20), water to be recyled
(stream 15), and acetonitrile column bottoms (stream 19). In most cases,
this acetonitrile is incinerated; however, in one plant it is purified and
sold. The recyled water is sent back to the absorber, and the
acetonitrile column bottoms are sent to the deep well pond.
The crude acrylonitrile (stream 17) is stored in tanks after
leaving the recovery column. From these tanks it is sent to the light
ends column, where materials with low boiling points are removed from the
stream. Process stream 21 from the light ends column contains hydrogen
cyanide, which is either incinerated or sold.
Bottoms from the light ends column are sent to the product column
(stream 23) where the acrylonitrile is finally distilled. Waste products
from the product column bottoms, made up of polymers, water, hydrogen
cyanide, and miscellaneous organics, are incinerated. The product is sent
to storage tanks.
3-8
-------
Table 3-3 shows a material balance for a representative
acrylonitrile plant.
3.2.2 Waste Stream Description
The recent Source Assessment Document for acrylonitrile plants
evaluated the severity of the pollution caused by various effluent
streams. Hydrocarbon emissions from the absorber vent gac and the deep
well pond were found to be the most serious pollution problems. In the
course of the present study, two other potentially serious pollutant
sources have been identified: NO emissions may be produced by the
A
hydrogen cyanide and acetonitrile incinerators, and hydrocarbon and
acrylonitrile emissions occur during reactor startup.
Absorber Vent Gas
Emissions data for the absorber vent gas are shown in Table 3-4.
These data taken from the Hydroscience Trip Report of September 1977,
(Reference 7) are more detailed than the emissions data in the Air
Products report (Reference 6). Data used in the Source Assessment
Document were obtained from T. Hughes at Monsanto (Reference 8). However,
the data from Mr. Hughes is for absorber vent gas that is mixed with other
streams before incineration, so these emission control data are not
directly comparable to the Hydroscience Trip Report data.
Deep Well Pond
Existing data for the deep well pond were not evaluated because
there are so few practical control methods for air emissions from open
ponds. The current control method (heavy oil on the surface) is probably
as effective as any. Another effective control method is to lower or
change the pollutant load of the pond. Therefore, data was assembled on
the waste water on its way to the pond and pollution control methods for
3-9
-------
TABLE 3-3. MATERIAL BALANCE FOR A REPRESENTATIVE ACRYLONITRILE PLANT3 (Reference 1)
Stream number
1 2
3 4
5
6
7
8 11
12
Description
Propylene Ammonia
Component feed feed
Nitrogen 2.5C
Oxygen
Carbon dioxide
Hater
Propylene 1,289
Propane 22
An»onia 553.1
Carbon monoxide
Acrylonitrile
Acetonitrile
Hydrogen cyanide
Sulfuric acid
Ammonium aulfate
Oxygenated hydrocarbons
Organic polymers
TOTALS 1,311 556
Process Reactor
air feed
5,865d 5,867
1,561 1,561
83.5 83.5
1,289
22
553.1
7,510 9,375
Reactor
product
5,867
103
185
1,653.
32e
22*
24
178
1,000;
118*
118e
95
9,395
Cooled
reactor
product
5,867
103
185
1,653
32
22
24
178
1,000
118
118
95
9,395
Quenched
reactor
product
5.867
103
185
1,653
32
22
178
1,000
118
118
9,276
Waste water Absorber
Sulfuric column
acid bottoms
3.470
0.9
70.6
93.2
79.3
70.6 3,643
vent
gas5
s.ae?*1
103
185
922
32
22
178
<0.1
0.6
0.3
7,310
(continued)
Note: Blanks indicate no component present in stream.
aAll values are g/kg of product acrylonitrile.
Emissions determined through field sampling.
cAmmonia used is agricultural grade as described in Section III.A.I.
d.
1 C3H6/1-06 NH3/8.4 air.
Composition of the combined reactor feed is based upon the following mole rates:
'conversion of propylene - 98%; conversion of propane = 0%; yield of acrylonitrile is 0.78 g AN/g C3H6; yield of
HCN and ACN -0.09 g/g C3H6.
-------
TABLE 3-3. Continued
co
i
Component
Nitrogen
Oxygen
Carbon dioxide
Mater
Propylene
Propane
Ammonia
Carbon monoxide
Acrylonitrile
Acetonitrile
Hydrogen cyanide
Sulfuric acid
AMftonium aulfate
Hydrocarbon*
Organic polymer*
13 14
Acrylo-
nitrile
plant Absorber
waste water bottoms
4.470 23,714
1,000
118
1.0 118
93
79 16
Stream number
15 16 17 19 20 21
Description
Aceto-
Crude Crude nitrile
Water aceto- acrylo- column Aceto- Hydrogen
recycle nitrile nitrile bottoms nitrile cyanide
23,000 23,000 10,000 1,000
1,000
118 118
118 118
17.5 1.2 7.5
7
23 24
Light ends Product
column acrylo-
bottoms nitrile
1,000 1,000
5.1 3.1
1.0 1
7.1 1
TOTALS
4,643
24,966 23,000 23,136 11,126 l.OOfl
118 118 1,013 1,005
(continued)
-------
TABLE 3-3. Concluded
CO
I—1
l\3
Stream number
25
27
28
29
30 31 32
Description
Heavy
Component ends
Nitrogen
Oxygen
Carbon dioxide
Water
Propylene
Propane
Ammonia
Carbon monoxide
Acrylonitrile
Acetonitrile 2
Hydrogen cyanide
Sulfuric acid
Ammonium sulfate
Hydrocarbons
Organic polymers 6.1
TOTALS 8 . 1
Flare
stack
gas
72.5
2
16
8
0.02
0.04
0.35
98.9
Fugitive
losses
0.006
0.0006
0.006
0.4
0.02
0.43
Incinerator
stack
gas
1,771
174
300
435
0.034
2,680
Deep well Storage Product transport
pond tank loading facility
emissions losses vent
0.802
0.81 0.0065
7
7 0.81 0.0065
-------
TABLE 3-4. STREAMS TO AND FROM CATALYTIC OXIDIZER (Reference 7
Organic reduction 42.5%
Combustion modules
Fuel 36.3 SCFM 97 Ib/hr
Air 55342 SCFM
Source3
Emission^
Stream No.
Flowrate, CFM, 7QOF
Flow determined by
Temperature
Pressure
Composition determined by
Component or formula
Acrylonitrile
HCN
Acetonitrile
Acetaldehyde
CO
Propylene
Propane
Ethane
C02
Argon
N2
02
H20
Totals
Total organics
81994
Flowmeter
40 oc
1.5 psig
Analysis
Wt. Ib/hr
0.001 5
0.002 7
0.03? Ill
0.002 /
1.643 5636
0.386 1334
0.790 2733
0.038 130
4.413 15270
1.429 4946
83.830 290065
1.114 3853
6.320 21868
100 346015
1.25 4327
138926
Calculation
400°
Atm
Analysis
Wt Ib/hr
1
1
0.005 28
0.001 7
0.227 1386
0.051 312
0.341 2080
0.009 54
5.322 32495
1.356 8282
79.524 485603
8.532 52076
4.632 28282
100 610627
0.41 7.489
3-13
-------
this waste water were evaluated. Waste water content is shown in Tables
3-5 and 3-6. Pollution control methods are evaluated in Section 4.
Incinerator Emissions
Incinerators at acrylonitrile plants burn hydrogen cyanide and
acetonitrile, as well as product column bottoms (precursor streams 20, 21,
and 25 and emissions stream 29). The data used in the Source Assessment
Document were for incinerators in which waste stream 12, the absorber vent
gas, is also incinerated. The major effluent problem that might arise
from these incinerators is the formation of NO from the fuel nitrogen
/\
of the acetonitrile (stream 20) and hydrogen cyanide (stream 21). The
streams to and from an acrylonitrile incinerator are shown in Table 3-7.
Reactor Startup Emissions
During startup, the reactor is heated to operating temperature
before the reactants (propylene and ammonia) are introduced. During
regular operation, effluent from the reactor is fuel rich and above the
upper flammability limit (see stream 3, material balance, Table 3-3).
However, during startup, the effluent stream from the reactor starts out
oxygen-rich, and passes through the flammable composition zone before
reaching the fuel rich zone.
In order to prevent explosions in the lines to the absorber, the
oxygen-rich reactor effluent is vented to the atmosphere during startup.
This effluent is high in reactant and product concentrations. It has been
estimated that the acrylonitrile emission rate during startup exceeds
10,000 lbs/hr., shown in Table 3-8.
Stream composition varies continuously during the startup.
Estimates in the Air Products Report are generally considered to be high.
3-14
-------
TABLE 3-5. ACRYLONITRILE PLANT WASTE WATER9 (Reference 1)
Material discharged
Concentration,
mg/1
Effluent factor,
g/kg
Raw waste water
Biological oxygen demand
Chemical oxygen demand
Total organic carbon
Total solids
Total suspended solids
Total dissolved solids
Oil and grease
Total nitrogen (as N2)
Ammonia nitrogen (as N2)
Nitrile nitrogen (as N2)
Phosphate
Phenol
Sulfate
Zinc
Chloride
Iron
Copper
Chromium
Cadmium
8,620
32,800
14,400
36,700 to 57,800
184 to 630
36,500 to 57,200
135 to 168
4,040 to 22,000
2,600 to 13,600
197 to 270
0.152 to 6.15
0.165 to 2.28
2,700 to 5,309
0.052 to 2.1
125 to 858
3.13 to 4.24
10.5
<0.05
£0.05
4,470
38.7
133
57.5
163 to 182
0.915 to 1.78
161 to 181
0.475 to 0.657
16.9 to 62.1
10.3 to 38.3
0.755 to 0.97
0.0004 to 0.0298
0.0007 to 0.0064
64.1 to 74.3
0.00002 to 0.0092
0.616 to 2.42
0.0088 to 0.0182
<0.00024
10.00014
10.00024
Other compounds which have been qualitatively identified include:
Acetaldehyde
Acrolein
Hydrogen cyanide
Acetic acid
Fumaronitrile
Acrylic acid
Acrylanide
Acrylonitrile
Acetonitrile
Maleonitrile
Organic polymers
Propionitrile
Ammonium formate
Methacrylonitrile
trans-Crotonitrile
cis-Crotonitrile
Allyl cyanide
Benzonitrile
Nicotinonitrile
Malononitrile
Furonitrile
Ticoline
Lutidine compounds
Benzene
Toluene
Ammonium acetate
Ammonium methacrylate
Ammonium acrylate
Succinonitrile
Acetone
Acetaldehyde cyanohydrin
Acetone cyanohydrin
Acrolein cyanohydrin
Pyrazole
Methyl pyrazine
Cyanopyrazine
Pyrazine
Personal communication to T. Hughes, Monsanto, A. W. Busch,
Regional Administrator, Region IV, U.S. Environmental
Protection Agency, February 1974.
3-15
-------
TABLE 3-6.
UNTREATED PROCESS WATER LOAD DISCHARGED TO DEEP WELLS
JANUARY 1, 1975 TO SEPTEMBER 30, 1976
TWO B PERMIT: WDW-100 & WDW-101 (2 WELLS) (Reference 7)
DAILY GRAB
SAMPLES
PARAMETER
Flow (MGD)
pH (pH units)
BOD5 (Ibs/day)
COD (Ibs/day)
TOC (Ibs/day)
TSS (Ibs/day)
TDS (Ibs/day)
NH3 as N (Ibs/day)
TKN as N (Ibs/day)
Phenol (Ibs/day)
Iron (Ibs/day)
Molybdenum (Ibs/day)
Nickel (Ibs/day)
Sodium (Ibs/day)
Zinc (Ibs/day)
Arsenic (Ibs/day)
Chromium (Ibs/day)
Copper (Ibs/day)
Sulfates (Ibs/day)
Cyanides (Ibs/day)
Acetonitrile (Ibs/day)
Acrylonitrile (Ibs/day)
Phosphates (Ibs/day)
MINIMUM
0.26
5.4
18000
50000
20000
23
41000
3900
9400
1
5
43
3
3000
<1
0.03
0.03
0.04
8900
290
42
70
<1
AVERAGE
0.49
N/A
58000
148000
61000
250
169000
17000
35000
17
22
200
12
6800
<1
0.06
0.44
0.15
50000
1500
970
900
23
MAXIMUM
0.68
9.2
156000
249000
139000
4000
300000
31000
61000
46
91
700
64
16000
6
0.25
20
0.71
140000
4900
3400
5400
480
MONTHLY
AVERAGES
MINIMUM
0.41
N/A
48000
1 20000
49000
65
1 30000
14000
27000
12
12
88
6
4900
<1
0.04
0.07
0.07
38000
1000
490
300
4
MAXIMUM
0.49
N/A
74000
171000
73000
500
200000
20000
41000
22
36
230
27
8600
2
0.09
1.35
0.22
67000
2100
1500
1700
43
3-16
-------
TABLE 3-7. STREAMS TO AND FROM INCINERATOR (Reference 8)
(CATALYST 21)
Emission Flow Rate to Incinerator Flow Rate from Incinerator
336,000 Ibs/hr
18,000
36,500
40
From Absorber Vent:
N2
°2
co2
CO
Methane
Ethyl ene
Ethane
Propane
Propylene
Butene
Acrylonitrile
Acetonitrile
HCN
Ally! Alcohol
Furan
Benzene
Toluene
Water
From Other Sources:
HCN
Acetonitrile
Combustion Air
Natural Gas
NOX (as N02)
Flow Rate
Uncertainties: inch"
167,000 Ibs/hr
5,760
6,790
2,360
8.3
57.5
23.2
965.5
614.5
22.3
2
34
14
1.3
22.3
7
4
38,200
625
2,800
218,900
750
To Incinerator
vidual HC ± 25%
CO ± 6%
HC total ± 8%
30
From Incinerator
3-17
-------
TABLE 3-8. STARTUP EMISSION STREAM (CATALYST 21)
(Reference 6)
Emission Rate (Ibs/hr)
Particulate 25
C02 1.593
CO 882
NH3 392
Propylene 417
Propane 711
HCN 466
Acrylonitrile 10,221
Acetonitrile 319
Nitrogen and Argon 84,094
Oxygen 5,686
Water 3,726
3-18
-------
However, these estimates were used to assess control methods since they
are based on worst case assumptions.
The two to six acrylonitrile reactors at a plant are each shut down
and restarted about four times a year; each startup lasts about one hour.
Therefore, there are between eight and twenty-four startup emission
incidents per year (References 1 and 6). Since these incidents are so
frequent and the emissions are so large, control technologies were
evaluated for these startup emission streams. The results are presented
in Section 4.
3.2.3 Control Technologies
Table 3-9 summarizes the emission streams reviewed in this report
and the control technologies currently in use. (Since most incinerators
can be fitted with waste heat recovery devices, and several are currently
in use on thermal incinerators at acrylonitrile plants, waste heat
recovery devices for incinerators are not evaluated separately in this
report.)
The absorber vent stream contains low molecular weight pollutants
in a dilute stream of high N,, and HUO content (see Tables 3-4 and
3-7). Thermal and catalytic incinerators are in use for this stream and
are further evaluated in this report. Since the hydrocarbons in this
stream are primarily of lower molecular weight, carbon adsorption and
hydrocarbon absorption are not economically feasible.
The startup emission stream has a high concentration of
acrylonitrile (Table 3-8). The stream is of short duration and has a
variable concentration. Flaring is a traditional method of dealing with
intermittent hydrocarbon emissions, and it was evaluated for this stream.
However, combustion of this stream could lead to 6,900 ppm NO emissions
3-19
-------
TABLE 3-9. CONTROL TECHNOLOGIES FOR EMISSION STREAMS
Emission Stream
Absorber Vent
Absorber Vent
Absorber Vent
Absorber Vent
Startup Stream
Startup Stream
Startup Stream
Starup Stream
Startup Stream
Holding Pond
Holding Pond
HCN and
Acetonitrile
Incinerators
Control
Technology In Use
Thermal Incineration Yes
Catalytic Incineration Yes
Carbon Adsorption No
Hydrocarbon Absorption No
Thermal Incineration No
Catalytic Incineration No
Flare No
Carbon Adsorption No
Solvent Absorption No
Carbon Adsorption of No
Effluent on Way to Pond
Solvent Extraction No
2-Stage Combustion No
3-20
-------
if 100 percent fuel nitrogen conversion occurred. Even 20 percent fuel
nitrogen conversion would lead to 1,300 ppm NO . Consequently,
A
combustion methods are considered unsuitable for this stream.
Carbon adsorption was also evaluated for the startup stream. The
stream is quite concentrated, consequently carbon adsorption might be used
after water scrubbing. Further study of a combination of water scrubbing
followed by carbon adsorption is recommended.
If the startup emission stream could be sent to the absorber (which
is basically a wet scrubber) the emissions from the stream could be
reduced, and further cleanup would not be required. However, this is not
possible because of the flammability problem caused by the mixture of fuel
rich and oxygen-rich streams in the lines to the absorber.
Hydrocarbon absorption of the startup emission stream is also
possible. However, disposal of the spent hydrocarbon could be a problem.
This hydrocarbon would have a high nitrogen content, making it potentially
unsuitable for burning in a waste-heat boiler. Therefore, further study
of hydrocarbon absorption was not undertaken for this stream.
Since hydrocarbon emissions from the holding pond are high, carbon
adsorption of the effluent on the way to the pond might be used to reduce
the hydrocarbon content of the pond.
Two patents for hydrocarbon absorption treatment of nitrile waste
waters were also reviewed. One describes a process change of
acrylonitrile plants that would greatly reduce the acrylonitrile waste
water emissions. The other describes solvent extraction of the waste
water. These patents are reviewed in Section 4.6.
3-21
-------
3.3 MONTEDISON PROCESS
Montecatini Edison makes acrylonitrile in a process similar to the
SOHIO process, except that a tellurium, cerium and molybdenum catalyst is
used (Reference 9). The process and emissions are not significantly
different. Chemical and biological treatment of the waste water, rather
than deep well injection are used. A further study of Montecatini's waste
water treatment techniques should be undertaken, with the intent of
eliminating the holding ponds as a source of air emissions.
3-22
-------
SECTION 4
ALTERNATE CONTROL METHODS
4.1 INTRODUCTION
This section discusses control methods currently in use and those
proposed for use in the acrylonitrile industry. The operation of each
control method and the waste streams suitable for each are described.
Then the technical feasibility of each control method is evaluated.
(Secondary pollutant formation, large energy requirements, or other
problems might eliminate a control method from further consideration.)
After identifying the best control method for a particular waste
stream, this section describes the efficiency, cost, reliability, and
energy requirements for that method. (Section 5 discusses the available
demonstration opportunities for single and combined control technologies.)
Combustion processes for waste gases and liquids are discussed
first (Sections 4.2, 4.3 and 4.4) because they are currently in use at
acrylonitrile plants. Waste heat recovery is also reviewed in these
sections. Problems with existing incinerator processes (e.g., the
possibility of high NO production and the lack of control devices for
A
the startup stream) are also discussed. Carbon adsorption and hydrocarbon
absorption processes, which are not currently used at any acrylonitrile
plant, are presented in Sections 4.5 and 4.6.
4-1
-------
4.2 FLARES
4.2.1 Technical Description
A flare is a diffusion flame, open to air, used as an inexpensive
method to dispose of temporary and suitable gaseous effluent without heat
recovery. Because a flare is capable of handling moderately large changes
in flowrate, it is typically used during system upsets and startups. (In
contrast, incinerators are designed for a specific, regulated volumetric
flowrate, commonly with heat recovery.) Flares can be classified into two
groups: elevated and low-level. Elevated flares are used for the
periodic combustion of high flowrate gases (1,000,000-2,000,000 Ib/hr);
whereas low-level flares, often located at ground level, are used for the
disposal of low, continuous flow gases (80,000 to 100,000 Ib/hr)
(Reference 10). Knockout drums, installed in the flare feed stream,
prevent liquid carryover into the flare, and natural gas pilot burners act
as a continuous ignition source. Adequate combustion of low Btu effluent
•5
(less than 200 Btu/ft°) requires the use of supplemental natural gas.
Depending on the duration of firing, this may represent a substantial
operating cost. Flame luminosity and particulate formation can be reduced
by injecting steam into the flare. Steam is usually controlled manually
to minimize its consumption.
Carbon monoxide, unburned hydrocarbons, particulates, NO , and
)\
SO are all potential flare pollutant emissions. The air/fuel ratio,
A
the flame temperature, the amount of steam injection, the effluent
composition, and the design of each specific flare determine the quantity
of pollutants. Unfortunately, due to sampling difficulties, there is
little data available on flare performance. (Sampling requires insertion
of a probe into a very unsteady, tall plume. This is difficult to
4-2
-------
accomplish with precision and accuracy.) Nonetheless, manufacturers
routinely claim that flares are capable of combusting 90 percent of the
hydrocarbons (Reference 6).
Flares are also a potential noise problem. Noise intensity depends
on the square of the exit velocity of the gas stream and on the properties
of the particular gas. With proper design, noise can be adequately
reduced to meet OSHA regulations (i.e., 90 dBA for 8 hours exposure).
Despite these drawbacks, flares are an effective low-cost method for
reducing transient gaseous emissions.
4.2.2 Suitable Waste Streams
In an acrylonitrile plant, two flares are designed into the system
for use during emergencies as back-up emission control devices. The HCN
flare burns any HCN vapors not removed by water scrubbers from the HCN
storage bin and railroad car loading procedure. The acrylonitrile flare
receives vapors from the top of the stripping and HCN columns if emergency
release of pressure relief valves occurs. (The Hydroscience Trip Report
states that the emergency valves venting to the acrylonitrile flare have
not been tripped in seven years (Reference 7).) Since these flares seem
to be performing satisfactorily, no further evaluation was conducted.
During reactor startup, gaseous effluent from the reactor is vented
directly to the atmosphere. Although this condition is temporary and
occurs only four times a year (per reactor), this one hour procedure
releases a substantial quantity of pollutants. Because this startup
stream is characterized by rapid concentration and flowrate changes, it is
quite suitable for control by flaring. The potential difficulty is that
large quantities of NO might be formed. This startup stream is
4-3
-------
described in Table 3-8 and compared with steady state flow from the
absorber vent in Table 4-1.
4.2.3 Technical Feasibility
During startup operation, the reactor gaseous effluent bypasses the
absorber and is vented into the atmosphere. The vented gaseous stream is
10 percent higher in mass than the normal absorber vent stream. This
additional flow is primarily an increase in acrylonitrile output to 10,221
Ib/hr, and to a lesser extent, increases in NH^, HCN, and acetonitrile.
(Table 3-8 describes this stream.) Table 4-1 compares the vented gaseous
stream to the normal absorber vent gas flow. Flares can be designed to
handle this volume of gas; however, the high acrylonitrile content may
lead to high NO emissions. This problem will exist for the startup
/\
streams with any combustion method.
4.2.4 Efficiency, Cost, Reliability, and Energy Requirements
Flares are designed to oxidize over 90 percent of the hydrocarbon
pollutant. Due to the low Btu content of the startup effluent stream
(calculated at 131 Btu/ft ), it would be necessary to use supplementary
natural gas -- around 0.016 cubic feet of natural gas per cubic foot of
effluent (Reference 11).
Combustion of the reactor startup effluent, which is high in fuel
nitrogen, could produce a considerable amount of NO . NO formation
is a strong function of flame temperature and the extent of fuel/air
premixing. Forty to ninety percent of the fuel nitrogen can potentially
form NO (Reference 12). Complete conversion of fuel nitrogen into
NO would yield a concentration of 29,220 ppm NO effluent, or
X *
4-4
-------
TABLE 4-1
COMPARISON OF START-UP AND CONTINUOUS FLOW GAS STREAMS
(CATALYST 21) (Reference 6)
Combustibles:
Continuous
Start-up
Component
CO
Aircnonia
Propylene
Propane
Hydrocyanide
Acrylonitrile
Acetom'trile
TOTAL
Noncombustibles:
Participate
co2
N2
°2
H?0
TOTAL
Ib/hr
1,498
0
470
740
6
3
80
2,803
TR
4,171
81,339
966
8,598
95,074
Weight Percent
1.5
0
.48
.76
--
--
.08
--
4.26
83.10
.99
8.78
Ib/hr
882
392
417
711
466
10,221
319
13,408
25
1,593
84,094
5,686
3,726
95,124
Weight Percent
.81
.36
.38
.65
.43
9.42
.29
.02
1.47
77.48
5.24
3.43
Total Flow:
97,877
108,532
4-5
-------
3,370 Ib NO per hour (per reactor) — considerably above allowed
X
emission standards (see Appendix B for calculations). However, this high
NO emission might be preferable to the present 10,221 Ib/hr
A
acrylonitrile output during startup. The health hazards of such
trade-offs should be assessed.
The capital cost of a flare system capable of handling the startup
effluent stream is approximately $20,000 including control equipment, hut
not including the site specific installation costs, which may be equally
high.
Approximately 25,000 scf of natural gas (900 Btu/ft3) are
required to fire the flare for one startup (Appendix B). At about
$0.25/lb acrylonitrile, the cost of the flared acrylonitrile would be
$2,500.
Because the increased reactor flowrate during startup is only 10
percent higher than during normal operation, it may be possible to burn
the reactor startup effluent in the existing incineration equipment used
for the gaseous absorber effluent. This would require supplementary
monitoring of the incineration during startup with probable modification
of fuel/air mixture and temperature to accommodate the different gaseous
compositions and flowrate.
4.2.5 Conclusions
Although startup occurs infrequently, it is necessary to implement
a method capable of reducing the exceedingly high acrylonitrile emissions
during this procedure. Flaring is a reliable approach capable of handling
large concentration changes typical of the reactor startup streams.
Although it has a low capital cost, a flare would require supplementary
fuel and could produce a large amount of NO
A
4-6
-------
High NO emissions due to acrylonitrile combustion are a
potential problem for any combustion method used to control emissions of
the startup stream. Research into the health trade-offs of high
acrylonitrile emissions vs. high NO emissions could be performed to
J\
provide information for evaluating flares. Such research, however, is not
recommended. Instead, low NO emission control technologies for the
/\
startup stream are recommended in Section 4.6.
4-7
-------
4.3 THERMAL INCINERATION
4.3.1 Technical Description
A thermal incinerator is an enclosed system in which waste gases
are pyrolyzed and then oxidized. Hydrocarbons and carbon monoxide in the
gases oxidize to carbon dioxide and water. The incinerator is often fired
with supplemental fuel and air since the waste stream will usually not
support a flame by itself. Indeed, more Btus are usually supplied by the
added fuel than by the waste gas stream. Heat exchangers are frequently
added to incinerator systems to recapture energy.
Incinerators usually operate continuously, rather than
intermittently. Constant thermal cycling of the incinerator would shorten
its life expectancy, due to increased cracking of the refractory lining.
Incinerators are designed to provide for adequate temperature,
retention time and turbulence for complete oxidation of the waste gas
fume. Temperature and retention time are coupled variables. Low
temperatures and long retention times will achieve the same degree of
control as high temperatures and short retention times (Figure 4-1).
Temperature profiles through the incinerator are very important, since
many reactions take place downstream of the flame front and these
reactions occur rapidly only at high temperatures.
Turbulence and the extent of premixing of fuel and air are also
important in promoting rapid oxidation reactions. By increasing
turbulence and premixing, the chance for a pollutant molecule to come into
close contact with an oxidizing species is increased. Modern incinerator
designers have discovered that the amount of turbulence can be rate
determining and that shorter residence times and less supplementary fuel
are needed for incinerators with high turbulence (Reference 13).
4-8
-------
10
o
INCREASING RESIDENCE TIME
1NO
12M
14M
TEMPERATURE. "F
1800
1000
2000
Figure 4-1. Coupled Effects of Temperature and Time
on Rate of Pollutant Oxidation
(Reference 14)
-------
Turbulence is controlled by burner, nozzle, baffle, and air flow
arrangements, and tangential fume entry.
A simple fume incinerator is illustrated in Figure 4-2. Air and
gas enter on one side, with fumes admitted tangentially, below the flame
front. Tangential admission of the fume and a refractory baffle ring
promote turbulence. The fire box is lined with refractory and insulated
to ensure a sufficiently high temperature. The size of the fire box
guarantees a sufficiently long residence time for complete incineration.
Newer designs promote even better mixing and carefully control the excess
air requirements, which can reduce operating costs by 20 to 30 percent
(Reference 13).
Special purpose incineration systems are designed to meet specific
needs. Two types of special purpose incineration systems may be useful
for acrylonitrile plants; these are the rich fume incinerator and the low
NO incinerator.
A
Rich fumes are fumes which do not contain enough oxygen for
combustion. Careful incinerator design is needed to handle these fumes;
the additional air dilutes the mixture, and lowers the overall
temperature. Rich fume incinerators differ from ordinary incinerators
chiefly in the careful attention given to addition of air for firing
(References 15 and 16). Since the absorber vent gas is oxygen-depleted,
rich fume incinerators were briefly evaluated for use on this pollutant
stream. However, rich fume incinerators are designed for use on very
concentrated hdyrocarbon streams; so they are not appropriate.
Low NO incineration is a two-stage combustion process: fuel
rich combustion followed by fuel lean combustion. Figure 4-3 shows a
diagram of this process. The first stage is a reduction furnace in which
4-10
-------
REFRACTORY
LINED STEEL SHELL
GAS
AIR
EXHAUST
REFRACTORY
BAFFLE RING
o:.:
FIRED BURNERS
fd-
\
TANGENTIAL.
FUME INLET
Figure 4-2. Direct-Fired Afterburner with
Tangential Burner Arrangement
(Reference 13)
4-11
-------
WASTE CATEGORY
"
GAS
LIQUID
EXAMPLE
NH5
NITROSAMINE
PRODUCTS OF OXIDATION
FG, NOX
FG, NOX
WASTE
AIR
I
1-*
ro
CONDITIONING
TOWER
C02
H2
CO
H20
FLUE GAS
RECYCLE GAS
Figure 4-3. Low NO Emission Incineration
/\
(Reference 17)
-------
a high temperature reducing environment (less than stoichiometric air)
converts fuel nitrogen to N,, and the supplemental fuel to water gas.
The quench section cools the water gas to approximately 1400°F, by
directly contacting it with recyle gas. A Thermal Oxidizer (TO) next
converts the H^ to H^O and CO to CO,,. A heat recovery boiler then
produces steam in cooling the flue gas to 350°F and is followed by an
unlined vent stack. Recycle gas cooling, not the use of air, steam or
water, is an integral part of this process to minimize NO formation and
maximize heat recovery.
Rich fume incinerators are not useful for acrylonitrile plants;
further evaluation of low-NO incinerators is recommended in Section 5.
X
The rest of this section will evaluate traditional incineration only.
4.3.2 Suitable Waste Streams
Thermal incinerator use has been evaluated for three waste streams:
1. HCN, acetonitrile and absorber vent off-gas streams currently
combined and sent to a thermal incinerator at an acrylonitrile
plant (Table 3-7).
2. Absorber vent off-gas streams described in Hydroscience Trip
Report (Table 3-4).
3. Startup emission stream described in Volume 2; Air Products
Report (Table 3-8).
Incinerators are not considered reasonable choices for the waste
water effluent stream on the way to the deep well pond.
4.3.3 Technical Feasibility
The first step in assessing feasibility for a control method is to
determine that the method:
4-13
-------
• Is efficient in eliminating the pollutants in question
• Does not form unacceptable amounts of a secondary pollutant
This evaluation is especially important for assessments of
combustion methods for nitrogen-containing waste streams. For these
streams, the potential for NO formation must be evaluated.
A
The degree of conversion of fuel nitrogen to NO depends upon:
A
• The form of the nitrogen in the fuel
t The percent of fuel nitrogen in the fuel
• The equivalence ratio (excess air) in the combustion chamber*
• The amount of mixing of air and fuel
• The temperature and residence time of the combustion system
Many experiments on fuel nitrogen conversion to NO use ammonia as a
source of fuel nitrogen. The sources of fuel nitrogen in the incinerators
in use at acrylonitrile plants are hydrogen cyanide and acetonitrile.
Acetonitrile and hydrogen cyanide both contain the C - N bond, and it is
assumed that they react similarly during combustion. However, these
cyanide-containing molecules would not be expected to follow the same
reaction paths as ammonia. The ammonia reacted in the flame front; some
of the hydrogen cyanide reacts downstream. (Figure 4-4, Reference 18).
Since ammonia and hydrogen cyanide show similar overall conversion to NO
during combustion, the results of experiments performed with ammonia are
assumed to be applicable to direct flame incineration of hydrogen cyanide
and acetonitrile, providing there is a sufficient residence time at post
flame-zone temperatures.
*The equivalence ratio, , is defined as /ly§T\ //lyel)
Vair / \ air/
'actual/ v stoichiometric
4-14
-------
-\2S
i
(—•
en
TIME, MSEC
Figure 4-4. Species Flux vs. Time, Flame 3, HCN Addition with an Equivalence Ratio
of 1.5 Flux Mole/am2-Sec, Nitrogen Species at Left, Carbon at Right
Flux Units of Graph. (Reference 18).
-------
.These NO conversion experiments (Reference 18) were performed with
premixed flames near their adiabatic flame temperature (1700°C). The
incinerators at acrylonitrile plants are thermal oxidizers; the pollutants
are oxidized at temperatures below 900°C. The amount of NO which would
be formed from hydrogen cyanide and acetonitrile in these temperature
ranges is uncertain (Reference 19), but is expected to be substantially
lower.
Figure 4-5 shows the effect of equivalence ratios on NO yield. The
dashed line shows the change in NO yields from 10 percent to 65 percent
with changing equivalence ratios. Homogeneous conversion processes are
also sensitive to changes in the fuel nitrogen content. (References 18
and 20).
The absorber vent gas (stream 2) currently sent to a catalytic
incinerator (Table 3-4), contains 42.8 Ibs/hr of fuel nitrogen, which
would yield 130 Ibs/hr of NO (240 ppm — weight) with 100 percent
A
conversion. At conversion rates of 10 to 60 percent, this would be 24 to
144 ppm NO . Based on this analysis, showing acceptable NO
A X
production rates, thermal incineration seems appropriate for this stream.
The hydrogen cyanide, acetonitrile, and absorber vent gas streams
Stream 1 (Table 3-7) are currently incinerated. Assuming all fuel
nitrogen is converted to NO 4361 Ibs NO /hr could be expected or
A A
9000 ppm (weight) of NO However, the exit stream was measured at 30
A
Ibs/hr of NO , or 67 ppm (weight) NO . This is equivalent to a 0.68
A X
percent conversion to NO . For fuel lean combustion, the examples from
X
the literature imply 5 to 60 percent conversion to NO , a ten-fold to
J\
one hundred-fold difference. This discrepancy cannot be resolved without
detailed information on the internal temperature-time profile within the
4-16
-------
FN] = 5.0 WT. % OF FUEL
0.8 0.9 1.0
= 1.75 WT. % OF FUEL
\—[FN] =1.5 WT % OF FUEL
\ (HOMOGENEOUS;DATA
\ OF MALTE et ol, 1976 )
\
O
\
\
\
\
\
"7
1.2 1.3 1.4 1.5
EQUIVALENCE RATIO
(Note: <|) is greater than one for fuel-rich combustion)
Figure 4-5. Comparison of the Relationship of Equivalence
Ratio and NO yields Between Catalytic Combustion
(solid curves) and Homogenous Reaction (dashed
curves). Ammonia Addition.
(Reference 20)
4-17
-------
incinerator. Without such information, this low conversion rate is
assumed to be correct. Further research on the topic is recommended in
Section 5.
Stream 3, the startup stream, described in Table 3-8, could yield
19,619 Ibs NO /hr, with 100 percent fuel nitrogen conversion. If this
A
stream were sent to one of the HCN incinerators used on stream 7 and only
0.6 percent conversion occurred, 268 ppm NO (weight) would be
A
expected. This would be 100 times as great for 60 percent conversion.
The NO emissions from combustion of this stream are potentially high,
A
and incineration is not a suitable method for control of this stream.
(Section 4.2 also discusses combustion of this stream.)
4.3.4 Efficiency, Cost, Reliability, and Energy Requirements
Theoretical efficiency of thermal incineration for degradation of
hydrocarbons and CO is very high. Incinerators that do not go through
frequent thermal cycling are quite reliable. Incinerator reliability has
been aided by heat sensitive paint. When the outside temperature of an
acrylonitrile plant incinerator becomes high, above 500°F, the paint at
that spot becomes lighter in color, until at 750°F it is white. This
assists in locating cracks in the refractory lining (Reference 21).
For an incinerator burning stream 1 (hydrogen cyanide, acetonitrile,
and absorber off-gas), capital cost was $3.8 million dollars in 1976,
without heat recovery (Reference 22). Heat recovery could be expected to
add 20 to 40 percent to the cost (Reference 23). This incinerator requires
148 MBtu/hr of supplementary fuel. If there is 50 percent heat recovery,
the net energy requirement would be 74.4 MBtu/hr. (This incinerator handles
the very dilute absorber off-gas stream. If the incinerator handled only
the liquid waste stream, the energy requirement would be lower.)
4-18
-------
For an incinerator burning stream 2 (absorber off-gas only),
capital cost estimates are $800,000 to $1,000,000 without heat recovery
and up to $1,400,000 with heat recovery. Because the design would have to
be unique (standard incinerators are much smaller), this estimate is
probably low (Reference 23). This incinerator requires 128 MBtu/hr of
supplementary fuel. If there is 50 percent heat recovery, the net energy
requirement would be 64 MBtu/hr (Reference 23).
4.3.5 Conclusion
Thermal incineration is inappropriate for stream 1, the hydrogen
cyanide - acetonitrile waste streams. Lower-than-expected NO emissions
A
are observed from these incinerators. Further investigation of this
phenomena and of the suitability of low - NO incineration for this
A
stream is recommended. Thermal incineration is appropriate for stream 2
(absorber vent gas stream) for which NO emissions are not a problem.
A
No combustion methods are suitable for stream 3, the reactor startup
stream, because of the high NO emission potential. No further
investigation of combustion methods for this stream is recommended, except
investigation of routing it to a low NO incinerator, if one were
A
installed for the HCN stream.
4-19
-------
4.4 CATALYTIC INCINERATION
4.4.1 Technical Description
In catalytic incineration, pyrolysis and subsequent oxidation of
waste gases occurs on a catalytically active surface in an enclosed
system. Catalytic incinerators usually operate at a lower temperature and
with much less supplementary fuel than thermal incinerators. The catalyst
itself degrades in time and needs replacement. These systems are designed
for a certain hydrocarbon flowrate, and unburned hydrocarbon may pass
through when this flowrate is exceeded. A diagram of a catalytic
afterburner suitable for hydrocarbon abatement is shown in Figure 4-6. A
schematic of a catalytic abater in use at an acrylonitrile plant is shown
in Figure 4-7. Catalytic incinerators can be designed with or without
heat exchangers, as shown in Figure 4-8.
Important parameters in the design of catalytic incinerators are
the temperature, residence time, air/fuel ratio, fuel and waste stream
passthrough rates, choice of catalyst metal (usually precious metals), and
expected catalyst degradation rates.
Catalytic incinerators are often chosen in preference to thermal
incinerators in order to lower thermal NO output, as well as save
A
fuel. In the temperature range around 1400°C, nitrogen and oxygen in
the air react to form "thermal NO ". Thermal incinerators operate in
this temperature range and potentially have thermal NO problems.
A
Catalytic incinerators usually operate below 1000°C at temperatures
where thermal NO does not form. However, incinerators at acrylonitrile
plants may have fuel NO problems. The relationship between thermal
A
incineration and fuel NOV is discussed in Sections 4.3.2 and 4.3.3. The
A
4-20
-------
CLEAN.HOT
GASES
CATALYST
ELEMENTS
OVEN
FUMES
PREHEATER
Figure 4-6. Schematic Diagram of Catalytic
Afterburner Using Torch-type
Preheat Burner with Flow of
Preheater Waste Stream Through
Fan to Promote Mixing
(Reference 14)
4-21
-------
CATALYTIC ABATERS
NATURAL
£
SILENCER
fl-
CATALYST
COMBUSTION
BLOWER
I
ro
r\>
SILENCER
DILUTION AIR
BLOV;ER
ABSORBER
OFF GAS -
s
T
A
C
K
FROM
OTHER
UNITS
TO
OTHER
UNITS
Figure 4-7. Schematic of a Catalytic Abater in Use at an
Acrylonitrile Plant. (Reference 7)
-------
Basic Catalytic System
tmr
«T AMMIMT TO I
Contaminated •xhaust enters the unit and Is preheated to the selected
temperature for optimum catalytic reaction. A filter screen traps large
non-combustible particles and provides a burn-off surface for cap-
tured lint and char. The hot, filtered effluent enters the reaction cham-
ber and passes through the catalyst-coated modules where rapid
oxidation of hydrocarbons occurs. The clean, hot exhaust then can
be vented to atmosphere or recycled to process.
Heat exchanger added to supply fresh process air
WAITI CFFLUCNT
AT AMMCNT TO MO'F.
MAT nCMAMQO
Adding the optional air-to-air heat exchanger yields important heat
recovery benefits. Heat from the clean, hot exhaust is transferred to
incoming fresh air for apace heating or process purposes.
Naat exchanger added for maximum fuel economy
HOT* AU TMMMATUNt* MOWN AM TTMCAL
VALUM AND NOT POM •HOWCATIO* mH
The heat exchanger transfers heat from the reactor exhaust to the
Incoming effluent, raising the waste gas temperature as much as
400 *F. The pre-heatlng requirement la thus reduced — or eliminated
completely. Reeultant fuel savings can reach 10%, compared to con-
ventional Incinerators.
Figure 4-8. Catalytic Incinerator (Reference 24)
4-23
-------
relationship between catalytic incineration and fuel NO will be
/\
discussed in detail in Section 4.4.3.
The discussion of catalytic incineration will focus on information
from a catalytic incinerator currently in use at an acrylonitrile plant.
It is assumed that optimization of all parameters for catalytic
incineration are incorporated with the existing incinerator (Table 3-4).
Development of low fuel NO catalytic incineration is discussed in
/\
Section 5.
4.4.2 Suitable Waste Streams
The waste streams for which thermal incineration were evaluated are
also logical candidates for catalytic incineration. These streams were:
1. HCN, acetonitrile, and absorber vent off-gas streams currently
combined and sent to a thermal incinerator (Table 3-7)
2. Absorber vent off-gas stream sent to a catalytic incinerator
which was described in the Hydroscience Trip Report (Table 3-4)
3. Startup emission stream (Table 3-8)
Since the data for stream 2 were detailed and specific, they were
evaluated first. Conclusions about the suitability of catalytic
combustion for streams 1 and 3 were derived from that detailed
evaluation.
4.4.3 Technical Feasibility
Data for stream 2, the absorber vent gas, were studied first
(Table 3-4). This is a dilute pollutant stream, characterized by low
molecular weight compounds and carbon monoxide. Combustion methods are
appropriate for this stream because only 240 ppm (weight) NO emissions
A
would occur if there were 100 percent fuel nitrogen conversion
(Section 4.3). However, the reduction of organic compounds after passing
4-24
-------
through the catalytic abater is only 42.5 percent (Reference 7), which
implies that the abater is not well suited for this use from a hydrocarbon
abatement efficiency standpoint. Consequently, consideration of NO
J\
emissions is secondary.
Stream 1, the currently incinerated hydrocarbon cyanide and
acetonitrile stream (Table 3-7) and stream 3, the startup emission stream
(Table 3-8), have the potential for significant NO pollution problems
A
with combustors described in Sections 4.2 and 4.3. The use of a catalytic
combustor is not expected to reduce the fuel nitrogen conversion problem.
As illustrated in Figure 4-5, fuel nitrogen conversion to NO in a
/\
catalytic abater ranges from 20 to 80 percent as does fuel nitrogen
conversion in a homogenous system, but the conversion in a catalytic
abater is much more sensitive to the fuel-oxygen (equivalence) ratio.
Figure 4-9 illustrates that the conversion rates are independent of
adiabatic flame temperature and are dependent on fuel nitrogen content.
Figures 4-5 and 4-9 show that fuel rich catalytic combustion gives
low NO yields from fuel nitrogen, while lean combustion can give very
A
high (up to 80 percent) NO yields. Unfortunately, fuel rich catalytic
A
combustion would accentuate the already-existing unburned hydrocarbon
passthrough problem. Consequently, catalytic incineration is not
appropriate for these streams because it would either produce high NOX
content (for lean combustion) or high unburned hydrocarbon (for rich or
near-stoichiometric combustion).
Two-stage (fuel rich followed by fuel-lean) catalytic combustion
might be appropriate for this stream. Such methods are currently under
development, and review of their potential is recommended in Section 5.
4-25
-------
2-
O
£ 80
LU
>
•z.
O
O
0
_J
LJ
Ul
O
O
a:
40
20
t-
u
n_
[FN] = 5.0 WT. % OF FUEL
D
n_
= 1.75 WT. % OF FUEL
I
I
I
1700 1800 1900 2000
ADIABATIC FLAME TEMPERATURE (°K)
Figure 4-9. Effect of the Adiabatic Flame Temperature on the
Conversion of NH3 to NO for Two Fuel-N Concentrations
at = 1.03, (Slightly Fuel-rich) Catalytic Combustion.
(Reference 20)
4-26
-------
4.4.4 Efficiency, Cost, Reliability, and Energy Requirements
Catalytic inefficiency for hydrocarbon reduction, and the potential
for high fuel-nitrogen conversion to NOV have eliminated catalytic
X
combustion from further detailed consideration. The initial cost of the
abaters for stream 2 ($6,800,000 in 1976) was quite high (Reference 7),
and catalyst replacement is also expensive.
4.4.5 Conclusion
Traditional catalytic fume abatement is not suitable for use at
acrylonitrile plants. Two-stage catalytic abaters currently under
development might be suitable for the high fuel NO streams. Further
A
evaluation of the potential of these systems is recommended in Section 5.
4-27
-------
4.5 CARBON ADSORPTION
4.5.1 Technical Description
Carbon adsorption is a technique used to remove organic compounds
from either liquid or gaseous streams. The organic impurities physically
(and reversibly) adsorb in multilayers on the carbon surface without
chemical reaction. Effective adsorption requires a highly porous material
with extremely high surface area per unit mass.
P
Carbon adsorbent beds with high surface areas (over 1000m /gm)
are used to adsorb large quantities of organic compounds from effluent gas
streams (up to 70 Ibs organics/100 Ibs carbon). Carbon adsorption is a
suitable cleanup method for effluent streams with organic pollutant
concentrations ranging from a few ppm to one percent (Reference 24).
Water streams containing high molecular weight hydrocarbons can be
effectively scrubbed by carbon adsorption. However, the technology
description and evaluation in this report center on the use of carbon
adsorption for gaseous pollutant streams.
Carbon adsorption is a suitable method for gaseous pollutant
control when the effluent:
1. Flowrate and concentration fluctuate (or when they are stable)
2. Contains dilute concentrations of pollutants up to one percent
3. Contains valuable organics which can be recovered during carbon
bed regeneration
4. Has a low Btu content, thus making it difficult to burn
5. Contains non-polar organics
The adsorption capacity of a specific form of activated carbon
depends primarily on the specific characteristics of the adsorbate and on
the system conditions. High molecular weight and high boiling point
4-28
-------
organics preferentially adsorb on a carbon surface. Low system
temperature and high organic concentration (partial pressure) will also
increase carbon capacity. Carbon capacity is typically expressed via an
adsorption isotherm in which bed capacity is plotted as a function of the
partial pressure of the organic vapor with system temperature as a
parameter (Reference 25). Acrylonitrile adsorption isotherms are given in
Figure 4-10 (Reference 26).
The adsorbate can be desorbed by raising the bed temperature 50°C
greater than the boiling point of the adsorbate. This is commonly
accomplished by passing low pressure stream or hot air through the bed.
The highly concentrated (up to 25 percent by weight) by-product gas can
then be incinerated or processed further to recover the organic
materials. For example, if steam were used for regeneration, distillation
or decanting could be used to recover the organics.
Despite the versatility of carbon adsorption, it does have certain
inherent limitations. Depending on the particular adsorbate, activated
carbon may catalyze cracking or polymerization of the adsorbed organic
material. To prevent plugging, particulates must be removed prior to
passage of the effluent through the bed. High water vapor concentrations
(high humidity) may reduce the capacity of the bed. In general, 40
percent is considered to be the maximum allowable water vapor
concentration in the effluent (Reference 14).
Physical adsorption releases 200-300 Btu per Ib-mole of adsorbed
material (Reference 14). Therefore, bed overheating may occur if rapid
adsorption of a concentrated effluent occurs.
4-29
-------
TOO
CO
o
O
O
o
CQ
O
CQ
ct:
I—
»—I
z
o
_J
>-
GC.
o
-
h-
1—4
o
<:
D-
<:
o
_i
-------
4.5.2 Suitable Waste Streams
Three waste streams were evaluated for the applicability of carbon
adsorption: (1) the absorber column vent gas, (2) the acrylonitrile
waste-water stream, and (3) the reactor startup stream.
The absorber vent off-gas (Table 3-4) contains mainly low molecular
weight organic compounds (less then C3). Because carbon adsorption is
not effective for removing these compounds, it is not considered useful
for control of this stream.
The acrylonitrile waste water stream (Tables 3-5 and 3-6) contains
large amounts of high molecular weight compounds. Carbon adsorption is
suitable for control. This waste stream is currently sent to a settling
pond where evaporation of the organics becomes a pollution problem.
Unfortunately, this waste stream is not well characterized. Further work
on waste water characterization and evaluation of carbon adsorption of the
waste water is recommended in Section 5.
During reactor startup (stream 3) high concentrations of
acrylonitrile (10 percent by weight) are vented directly to the
atmosphere. This effluent stream has fluctuations in concentration and
flowrate. It appears to be suitable for emission control using carbon
adsorption. Acrylonitrile is readily adsorbed by activated carbon.
Therefore, the rest of this section evaluates the use of carbon adsorption
to control emissions from the reactor startup stream.
4.5.3 Technical Feasibility
Due to the high concentration of acrylonitrile in the reactor
startup stream, use of carbon adsorption alone would require large amounts
of carbon (50,000 Ibs), and would require careful design to prevent carbon
bed overheat or pollutant breakthrough.
4-31
-------
These problems can be solved by having the effluent stream pass
through a wet scrubber before the carbon adsorption bed. The wet scrubber
reduces the acrylonitrile content of the stream from 10 percent to 1
percent, and the carbon adsorbs the remaining 1 percent. Process steam
(150-200°C) can regenerate the carbon bed. The water from the wet
scrubber and the carbon bed regeneration could be sent directly (or with
temporary storage) to the absorber column for recovery of the
acrylonitrile (Figure 4-11).
This system is designed to reclaim virtually all acrylonitrile lost
during reactor startup. (Effluent acrylonitrile emissions are reduced to
1 ppm). The system also releases none of the NO which would result
A
from flaring or incinerating of the same stream (see Section 4.2). At an
acrylonitrile plant, reactor startup occurs once or more a month (four
times a year for each reactor). The same scrubber and carbon adsorption
system could be used for each reactor. Appendix A describes sizing and
costing of this system.
The startup effluent stream also contains low concentrations of
hydrogen cyanide and low molecular weight hydrocarbons. Although a
significant portion of the HCN should be removed by the scrubber, any
remaining low molecular weight hydrocarbons and HCN would pass through the
adsorption beds into the atmosphere as a pollutant due to their poor
adsorption on carbon. It should be possible to direct this effluent
stream to the existing incinerators. Further study should be done to
determine whether this is feasible. Since adsorption is exothermic, the
rapid rate of acrylonitrile adsorption (1022 Ib/hr) could require external
cooling. Detailed design of the carbon adsorber must take into account
4-32
-------
REACTOR
START-UP
EFFLUENT
WET
SCRUBBER
90% EFF.
CARBON
ADSORPTION
BED
99% EFF.
co
co
» 1 PPM ACRYLONITRILE
r~
I STORAGE |
i r
i i
RECEIVER
ADSORPTION
COLUMN
Figure 4-11. Proposed Wet Scrubber Carbon Adsorption Module
-------
the need to dissipate the heat of adsorption, and the role that the wet
scrubber itself will play in heat transfer.
4.5.4 Efficiency, Cost, Reliability, and Energy Requirements
Water scrubbing followed by carbon adsorption is an efficient
method for removing acrylonitrile from an effluent stream and it allows
almost complete recovery of the acrylonitrile. Carbon capacity is high,
33 Ib acrylonitrile per 100 Ib carbon, at these operating pressures. Exit
pollutant concentrations as low as 1 ppm are possible. No NO is produced.
X
This system should be quite reliable. Carbon bed cooling may be
required and should be investigated. The same scrubber and adsorption
system can be used for each reactor.
Capital costs for the scrubber-adsorber combination total $212,000.
Sizing and costing assumptions for this system appear in Appendix A.
There may be an energy cost associated with chilling the carbon
beds; more detailed analysis is needed to ascertain if they require
chilling. The main energy cost is the steam for regeneration. Fifteen
pounds of steam are required for each pound of adsorbed organic, and
22,000 SCF of natural gas would be required to generate steam for each
startup (Reference 26).
At 1978 prices, energy for each startup would cost less than $100
(Reference 27). $2500 worth of acrylonitrile would be reclaimed.
4.5.5 Conclusion
A wet scrubber-carbon adsorption system would effectively recover
all acrylonitrile from the reactor startup stream. Lower molecular weight
hydrocarbons may not be removed completely, requiring additional pollution
control techniques. A demonstration program for a scrubber-adsorber on
the startup stream is recommended in Section 5.
4-34
-------
4.6 SOLVENT EXTRACTION AND ABSORPTION
4.6.1 Technical Description
The design of hydrocarbon pollution control systems is based on the
fact that organic pollutants often dissolve readily in other
hydrocarbons. There are two principal types of hydrocarbon absorption
pollution control:
• Solvent extraction of hydrocarbon pollutants. In this process,
a water solution containing hydrocarbon pollutants is brought
into contact with a different liquid hydrocarbon. The
pollutant hydrocarbon selectively dissolves in the liquid
hydrocarbon. Then the water and the pollutant containing
hydrocarbon liquid are separated.
• Absorption of hydrocarbon gas or mist into an organic solvent.
This is a wet scrubbing process using a hydrocarbon as the
scrubber rather than water.
4.6.2 Suitable Waste Streams
Scrubbing of hydrocarbon gases into hydrocarbon liquids was
evaluated for use on the absorber vent gas stream and the reactor startup
stream. Hydrocarbon scrubbing is practical for concentrated hydrocarbon
streams; it is most commonly used for gas streams containing 40 percent to
85 percent hydrocarbon by weight (References 28 and 29). Since the vent
gas and startup streams are 1.5 percent and 11 percent hydrocarbon by
weight respectively, the further evaluation of hydrocarbon absorption was
not performed.
Organic pollutants in waste water can also be cleaned by
hydrocarbon scrubbing (solvent extraction). Another way to reduce organic
pollution of acrylonitrile waste water is to change the acrylonitrile
4-35
-------
manufacturing process to eliminate water scrubbing and substitute
hydrocarbon scrubbing. Neither method for reducing pollutants in organic
waste water has been put into practice, but both seem feasible and will be
discussed in Section 4.6.3.
4.6.3 Technical Feasibility
Solvent extraction for organic-nitrogen containing waste water was
examined by Union Carbide Corporation under contract to the Robert S. Kerr
Environmental Research Laboratory of the Environmental Protection Agency,
in Ada, Oklahoma (Reference 30). This work was based on previous research
on solvent extraction for caprolactam plants. Union Carbide holds patent
#3,433,788 (granted March 18, 1969) on this process. Although solvent
extraction could be used on acrylonitrile waste water, this process has
only been tried on caprolactam waste water. It is considered
theoretically possible to reduce acrylonitrile water emissions down to ppm
levels; however, any polymers in the waste water stream are expected to
reduce the effectiveness of this process by forming emulsions that are
difficult to separate.
Tests of the solvent extraction process on solutions containing
caprolactam were performed in 1977 (Reference 30). Efficiency was found
to be low in the primary extractor. Unwanted separation of pollutant and
solvent occurred in the stripping column, and air leakage into the column
caused solvent decomposition and oxidation. Further research is
recommended to overcome these problems.
From these tests, it is impossible to draw conclusions about the
effectiveness of solvent extraction on acrylonitrile plant waste water.
The process was tested on amine-containing water while acrylonitrile plant
waste waters contain nitriles. Extraction efficiencies will be different
4-36
-------
for these different species. Polymers from the heavy ends (stream 25} in
the waste water may form emulsions. To summarize, the feasibility of
using solvent extraction for acrylonitrile waste water cannot be
sufficiently evaluated with these data.
Badger Company holds a patent (#3,895,050, Disposal of Waste
Materials from Unsaturated Nitrile) on prevention of waste water
contamination by nitriles. The abstract of the patent is:
A method of recovering and disposal of waste materials from a
plant for manufacturing unsaturated aliphatic nitriles or
aromatic nitriles whereby waste water, unreacted ammonia and
by-products such as HCN and acetonitrile are not condensed but
remain with the absorber off-gas for ultimate disposal by
incineration. The method employs a hydrocarbon solvent to
adiabatically quench the reactor effluent and, after removal
of polymer by-products, the partially quenched effluent is
passed to a hot absorber column where the nitrile product but
no ammonia and only some of the HCN are absorbed by the
hydrocarbon solvent. The nitrile-solvent mixture is distilled
to separately recover the solvent and nitrile product. The
solvent is recycled. The hot absorber off-gases are cooled to
recover water and then incinerated with ammonia, HCN,
acetonitrile and some vaporized solvent furnishing the
necessary fuel values. In a preferred embodiment, solvent in
the absorbed overheat vapors is recovered by scrubbing with a
high boiling oil.
This is basically a control technology process change: absorption
and quenching of the products in a hydrocarbon solvent rather than
absorption in water. In this process, waste streams which are sent to
incinerators are expected to be high in fuel nitrogen. Consequently, the
incinerators are expected to be high NO emitters.
A
4-37
-------
4.6.4 Efficiency, Cost, Reliability, and Energy Requirements
Solvent Extraction
The efficiency, cost, reliability, and energy requirements for
solvent extraction cannot be estimated from the data available. Data
would be needed on:
• Extraction efficiency of the solvent on waste water containing
acrylonitrile
• Effects of polymer and other contaminants on the process.
This process is at an early stage of research and development;
further study is needed before recommendations can be made. A theoretical
and laboratory study of solvent extraction is recommended in Section 5.
Hydrocarbon Absorption Process Change
This method has not yet been put into practice. It would solve
many waste water problems because water would never contact the product
mixture. However, the small amount of water produced in the reaction
would still need cleaning and disposal. Incineration of the
solvent-by-product mixture might lead to high NO emissions, and the
A
economics of this process are unknown.
4.6.5 Conclusion
Hydrocarbon absorption does not seem suitable for any of the
gaseous streams at acrylonitrile plants, although such a process change
might be suitable for eliminating waste water discharge.
More research needs to be done on solvent extraction for waste
water at acrylonitrile plants. Research should include study of:
0 Optimum solvent for nitriles (not amines)
• Solvent extraction efficiency for nitriles
• Effect of polymer on the solvent extraction.
4-38
-------
In-pi ant process changes to eliminate water as a sorbent are
untried and possibly expensive. Further work on this subject is not
recommended.
4-39
-------
SECTION 5
GENERAL CONSIDERATIONS
5.1 CONTROL METHODS FOR SPECIFIC STREAMS
5.1.1 Absorber Vent Gas Stream
Two control methods were evaluated for this stream: thermal ano
catalytic incineration. Since the hydrocarbon abatement efficiency of the
catalytic incinerator was less than 50 percent (42.5 percent), no furtner
evaluation of catalytic incineration was performed. For the thermal
incinerator at Monsanto's plant in Chocolate Bayou, Texas, capital costs
were $3.6 million dollars (1976). Supplemental fuel required was 143
MBtu/hr or, with 50 percent heat recovery, 74.4 MBtu/hr net. For a
catalytic incinerator, capital costs would be approximately a factor Wv
two higher and fuel requirements a factor of three lower (Reference 23).
5.1.2 Liquid Stream on Way to Holding Pond
Information on hydrocarbon absorption (solvent extraction) for this
stream was incomplete. Carbon adsorption appears promising, but it was
impossible to determine the efficiency, cost, reliability, or energy
requirements of this technique.
5.1.3 HCN and Acetonitrile Incinerators
The literature on fuel nitrogen conversion implies that these
incinerators may have serious NO emissions problems. Their efficiency,
costs, and energy requirements were not evaluated further.
5-1
-------
5.1.4 Startup Emission Stream
Combustion methods for abatement of this stream would produce large
quantities of N0x. For example, flaring this stream would yield 2^,000
pom NO . Ninety percent hydrocarbon abatement would be achieved, with
25,000 SCF of natural gas used per startup and a capital cost of around
$30,000.
A preferred method of abatement for this stream is a portable wet
scrubber-carbon adsorption module. All the acrylonitrile absorbed in the
wet scrubber and carbon adsorber could go back to the process for
recovery; there would be no secondary pollutant stream. Abatement for
acrylonitrile emission would be over 99 percent, capital cost for the
system would be $212,000 with natural gas costing less than $100 (2?,000
SCF) would be required to generate steam for regeneration of the carbon
bed. While this method requires several times the capital cost of
flaring, $2500 worth of acrylonitrile would be recovered at each startup.
5.2 RESEARCH AND DEMONSTRATION RECOMMENDATIONS
5.2.1 Absorber Vent Gas Stream
Thermal incinerators for this stream do not need further
development. An advanced catalytic combustion method, graded cells, might
increase the effectiveness of catalytic incineration in the future by
reducing hvdrocarbon passthrough (Reference 31). No specific research or
development program is recommended; however, when catalytic abaters are
developed in other contexts, they may be applied here.
5.2.2 Liquid Stream On Way to Holding Pond
Two programs are recommended:
• A program to evaluate carbon adsorption and alternate waste
water treatment methods used in Europe,
5-2
-------
A bench-scale program to assess the efficiency of using the
solvent extraction method under development for caprolactam
plant waste waters for acrylonitrile plant waste waters. This
program would assess:
The efficiency of these solvent extraction methods at
extracting the waste water hydrocarbons
The extent of interference of the polymer fraction of waste
water
The stability against oxidation of the solvents (solvent
oxidation has been a problem in the caprolactam
applications)
The costs and energy requirements of scaling up this
solvent extraction system.
HCj^LAceJ^^
A study is recommended of the existing incineration methods to
determine the factors affecting their N0x emissions. If a NOX
emission problems exists, an advanced incineration abatement feasibility
study should be undertaken.
Two-stage low NOX incineration should be evaluated for combustion
of these compounds. Both thermal and catalytic systems should be studied,
as well as processes where incineration is followed by alkali scrubbing
and ammonia injection.
After this feasibility study, a demonstration program of the best
incineration process would be appropriate.
5.2.3
5-3
-------
5.2.4 Startup Emission Stream
Two programs are recommended:
• A feasibility study of routing the startup emission stream to
the existing absorber. A demonstration program is recommended
if this proves feasible.
• A demonstration program for the wet-scrubber, carbon-adsorption
module described in Section 4.
5-4
-------
REFERENCES
1. Hughes, T. W. and Horn, D. A., "Source Assessment: Acrylonitrile
Manufacture (Air Emissions)," EPA-600/2-77-107J, September 1977.
2. "Economic Impact Assessment for Acrylonitrile," Final Report,
submitted to David R. Bell — COTR, Occupational Safety and Health
Administration, February 21, 1978.
3. Pervier, J. W., et al., "Survey Reports on Atmospheric Emissions from
the Petrochemical Industry," Volume 1, EPA-450/3-73-005A, March 1974.
4. Clements Associates report by Dr. Robert Squire, dated September 2,
1977, Pursuant to DOL Contract No. 2-9-f-7-0099 for the Critical
Scientific Evaluation of Data on the Carcinogenic Potential of
Acrylonitrile (quoted in Reference 2).
5. NIOSH Memo of December 7, 1977, including transmittal of pathologists1
reports on reviews of MCA and duPont studies, (quoted in Reference 2).
6. Schwartz, W. A., et al., "Engineering and Cost Study of Air Pollution
Control for the Petrochemical Industry," Volume 2: Acrylonitrile
Manufacturing, EPA-450/3-73-006b, February 1975.
7. "Emission Control Options for the Synthetic Organic Chemicals
Manufacturing Industry," Trip Report, Acrylonitrile Production Plant,
September 7, 1977.
8. Personal communication with T. W. Hughes, Monsanto, February 2, 1978.
9. Caporali, G., "How Montedison makes Acrylo," Hydrocarbon Processing,
pp. 144-146, November 1972.
10. Klett, M. G., et al., "Flare Systems Study," Lockheed Missiles and
Space Company, Huntsville, Alabama, U.S. Department of Commerce,
EPA-600/2-76-079, March 1976.
11. Personal communication with Bill McNew, National Air-Oil Burner Co.,
April 1978.
12. Sternling, C. N., and Wendt, J. 0. L., "On the Oxidation of Fuel
Nitrogen in a Diffusion Flame," AIChE Journal, Volume 20, No. 1, pp.
81-85, January 1974.
13. Hemsath, K. H., Thekdi, A. C., and Lewis, F. M., "Application of
Reaction Kinetics and Mixing Studies in Design of a Fume
Incinerator," No. 73-299, 66th Annual Air Pollution Control
Association Meeting, June 24-28, 1973
R-l
-------
14. "Control of Volatile Organic Emissions from Existing Stationary
Sources, Volume 1: Control Methods for Surface-Coating Operations,"
EPA-450/2-76-028, November 1976.
15. Hemsath, K. H., and Thekdi, A. C., "Rich Fume Incineration,"
Pollution Engineering, pp. 38-39, July 1973.
16. Davis, A., "Incineration of Low Oxygen Fumes," Pollution Engineering,
p.36, December 1972.
17. Cegielski, J. M., Jr., "The Technology of Modern Incineration
Processes," AIChE National Meeting, Houston, Texas, March 20-24, 1977
18. Axworthy, A. E., et al., "Chemistry of Fuel Nitrogen Conversion to
Nitrogen Oxide in Combustion," EPA-600/2-76-039, February 1976.
19. Personal communication with W. Macon Sheppard, Environmental
Consultants, Clemson, S. Carolina, July 1978.
20. Matthews, R. D., "The Nature and Formation of Nitrogenous Air
Pollutant Emissions from Combustion Systems," Lawrence Berkeley
Laboratory, LBL-6850, October 1977.
21. Eckel, J. A., "Hydrocarbon Emissions Abatement by Incineration,"
AIChE 83rd National Meeting, March 20-24, 1977.
22. Personal communication with James Eckel, Monsanto Co., St. Louis,
Missouri, May 3, 1978.
23. Personal communication with W. Fielding, Midland-Ross, Chicago,
Illinois, May 4, 1978. y
24. "The Practical Solution to Hydrocarbon Emissions Control, the DuPont
Catalytic Abatement System," brochure, Applied Engineering Company,
Inc., Orangeburg, South Carolina.
25. Cheremisinoff, P. N., "Carbon Adsorption of Air and Water
Pollutants," Pollution Engineering, pp 24-32, July 1976.
26. Personal communication with R. I. Cooley, Calgon Corporation, May 8,
i.y / o •
27. "Adsorption Bed Costing," pamphlets, Nos. 23-2001a and 23-1040
Calgon Corporation, March 1, 1977.
28. Pruessner R. D and Broz, L. D., "Hydrocarbon Emissions Reduction
Systems UtT^zed by Petro-Tex," AIChE annual meeting, March 1977.
29. Personal communication with L. D. Broz, February 1978.
3°' March^l C^nication with w- DePrater, EPA, ADA, Oklahoma,
R-2
-------
3!. Persona, co»n,cat1on ;1th **- Kes-lrln,. Acurex Corporation, Mt.
View, California, May 1978.
R-3
-------
APPENDIX A
ASSUMPTIONS AND RESULTS FOR SCRUBBER-ADSORBER SIZING AND COSTING
Wet Scrubber
Assumptions:
Table 3-8 describes the reactor startup effluent stream which is to
be controlled by the scrubber-adsorber unit.
The scrubber was designed to adsorb 90 percent of the acrylonitrile
in this stream: the effluent from the scrubber will be 1 percent
acrylonitrile. This concentration of acrylonitrile is suitable for
cleanup by carbon adsorption.
The cost of a wet scrubber is S5.40/ACFM (Reference 32).
Results:
Input to scrubber: 108,532 Ibs effluent/hr
10,221 Ibs acrylonitrile/hr
Output: 99,332 Ibs effluent/hr
1,022 Ibs acrylonitrile/hr
Partial pressure of
acrylonitrile: 0.16 psig
Temperature of acrylonitrile: 40°C
Cost of wet scrubber: $135,000
A-l
-------
Carbon Adsorber
Assumptions:
Input to adsorber is output from scrubber.
Theoretical capacity of the carbon adsorber is 33 Ibs
acrylonitrile/100 Ib carbon.
Regenerating with 250°F steam would leave 10 Ib acrylonitrile/100
Ib carbon unregenerated.
Operating capacity would be 23 Ibs acrylonitrile/100 Ib carbon if
unit were operated to 100 percent breakthrough.
Since the adsorption cycle will be terminated shortly after the
initial breakthrough, before all the carbon is saturated, the overall
carbon capacity is approximately 15 Ib acrylonitrile/100 Ib carbon
(Reference 19).
The carbon bed must adsorb 1022 Ib acrylonitrile in one hour during
startup, and then be regenerated.
Output is less than 1 ppm acrylonitrile.
Results:
For 1022 Ibs acrylonitrile, 6810 Ib carbon would be required.
Carbon beds are frequently purchased in 6000 Ib units (References
25 and 26). Since the carbon bed requirement is more than 10
percent over the standard size, the engineering assumption was made
that two carbon beds would be needed.
Cost of two 6000 Ib carbon beds = $66,000
Cost of carbon at $0.90/lb = $11,000
Carbon adsoprtion system cost = $77,000
A-2
-------
Energy requirement for regeneration of carbon: 22,000 SCF of
natural gas used to form steam, gas costing approximately $50 at 1978
prices.
Combined System
Assumptions:
All acrylonitrile from wet scrubber and carbon adsorber
regeneration is reclaimed.
Acrylonitrile is worth $0.25/lb (January 30, 1978, Chemical
Marketing Reporter).
Results:
System capital costs:
Cost of energy:
Credit for acrylonitrile:
$135,000 scrubber
$ 77,000 adsorber
$212,000 total
$50/startup
$2500/startup
A-3
-------
APPENDIX B
POTENTIAL NOV FORMATION FROM REACTOR STARTUP
X
Assuming that all fuel nitrogen is converted to NO , and that no
A
N (in air) is converted to NO , the maximum amount of fuel nitrogen
X • X
available for reaction is shown in Table A-l. (Data from Reference 6).
Table A-l
Combustible
Compound
Ammon i a
Hydrocyanic Acid
Acrylonitrile
Acetonitrile
Molecular
Weight
17
27
53
41
Weight %
in Effluent
0.36
0.43
9.42
0.29
3.106
Weight %
Nitrogen
in Effluent
0.296
0.223
2.488
0.099
Weighted Change
in Moles upon
Reaction
-0.00148
-0.00215
-0.04710
-0.00145
-0.05218
Therefore, the total number of moles of N0£ in the effluent is:
0.03106 Ib N
1 Ib Effluent
X A Ibmole N0?\//1 Ibmole EffluentN
14 Ib N V\ 30 Ib Effluent I
0.067 Ibmole N02_
Ibmole Effluent
B-l
-------
Addition of 0.016 Ibmole natural gas per 1 Ibmole effluent is
required to raise the Btu content to combust the flare (Reference 11).
The amount of air required to combust the effluent is 1.05 Ibmoles
air/lbmole effluent. The amount of air required to combust the natural
gas is 7.52 Ibmoles air/lbmoles natural gas. Therefore, the total volume
exiting the flare (corrected for mole change upon reaction) based upon 1
Ibmole effluent is:
(0.016 Ibmole natural gas + 0.120 Ibmoles air) x 1 Ibmoles Output
Ibmoles Input
+ (1 Ibmole effluent + 1.05 Ibmole air) x 1.052108 Ibmoles Output
Ibmoles Input
2.29 Ibmoles Output
Thus the maximum concentration of N0 in the effluent is:
0.067 Ibmoles NO? _
Ibmoles Effluent
- = 29,220 ppm
2.29 Ibmoles Output
Ibmoles Effluent
B-2
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the reverse before completing)
1. REPORT NO.
EPA-600/2-79-048
2.
3. RECIPIENT'S ACCESSION- NO.
4. TITLE AND SUBTITLE
Acrylonitrile Plant Air Pollution Control
6. REPORT DATE
February 1979
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
M.T. Anguin and S. Anderson
B. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Acurex Corporation
685 Clyde Avenue
Mountain View, California 94042
10. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO.
68-02-2611, W.A. 15
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final: 1/78 - 12/78
14. SPONSORING AGENCY CODE
EPA/600/13
IB. SUPPLEMENTARY NOTES project officer j A jefcoat is no longer with IERL-RTP. For
details contact Bruce Tichenor, MD-62, 919/541-2547.
i6. ABSTRACT
on avaiiable literature , the report identifies and ranks (in terms of
efficiency, cost, and energy requirements) air pollution control technologies for
each of four major air pollutant emission sources in acrylonitrile plants. The sour-
ces are: (1) absorber vent gas streams, (2) liquid waste holding ponds, (3) hydrogen
cyanide/acetonitrile incinerators, and (4) reactor startup streams. It also identifies
control technology research and development needs. Conclusions concerning emis-
sions from each source include: (1) absorber vent gas streams --large amounts of
hydrocarbons (HCs) are emitted; thermal incineration is an acceptable and efficient
control; (2) liquid waste holding ponds--high levels of HC emissions occur; no
controls are available for these emissions at the ponds; reduction of the HC levels
prior to discharge to the ponds is feasible; research should be conducted on carbon
absorption and solvent extraction; (3) hydrogen cyanide/acetonitrile incinerators --
high levels of NOx may occur; more data are required; advanced incineration tech-
niques should be investigated; and (4) reactor startup streams --large amounts of
acrylonitrile are vented to the atmosphere during reactor startup; incinerating this
stream produces high levels of NQx; carbon adsorption and wet scrubbing appear
feasible; and demonstrations are required.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COSATi Field/Group
Air Pollution
Acrylonitriles
Industrial Processes
Ponds
Hydrogen Cyanide
Acetonitrile
Incinerators
Hydrocarbons
Nitrogen Oxides
Scrubbers
Activated Carbon
Adsorption
Air Pollution Control
Stationary Sources
13B
07C
13H
08H
07B
07A/13I
11G
18. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report)
Unclassified
21. NO. OF PAGES
86
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (8-73)
B-3
------- |