EPA-600/2-77-188
September 1977
Environmental Protection Technology Series
PHTHALIC ANHYDRIDE PLANT
AIR POLLUTION CONTROL
Industrial Environmental Research Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Research Triangle Park, North Carolina 27711
-------
RESEARCH REPORTING SERIES
Research reports of the Office of Research and Development, U.S. Environmental Protection
Agency, have been grouped into five series. These five broad categories were established to
facilitate further development and application of environmental technology. Elimination of
traditional grouping was consciously planned to foster technology transfer and a maximum
interface in related fields. The five series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumenta-
tion, equipment, and methodology to repair or prevent environmental degradation from point
and non-point sources of pollution. This work provides the new or improved technology
required for the control and treatment of pollution sources to meet environmental quality
standards.
EPA REVIEW NOTICE
This report has been reviewed by the U.S. Environmental Protection Agency and approved
for publication. Approval does not signify that the contents necessarily reflect the views and
policy of the Agency, nor does mention of trade names or commercial products constitute
endorsement or recommendation for use.
This document is available to the public through the National Technical Information Service
Springfield, Virginia 22161. vtw P,
-------
EPA-600/2-77-188
September 1977
PHTHALIC ANHYDRIDE PLANT
AIR POLLUTION CONTROL
by
C.T. Chi and T.W. Hughes
Monsanto Research Corporation
Station B, Box 8
Dayton, Ohio 45407
Contract No, 68-02-1320
Task No. 25
ROAP No. 21AXM-011
Program Element No. 1AB015
Project Task Officer: E.J. Wooldridge
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
CONTENTS
Figures v
Tables vii
1. Introduction 1
2. Summary 3
3. Conclusions and Recommendations 5
3.1 Thermal Incinerator/Steam Generation System 5
3.2 Wet Scrubber/MAN Recovery System 6
3.3 Carbon Adsorber/Waste Incineration System 6
3.4 Additional Considerations 6
4. Source Description and Present Control Practices 8
4.1 Phthalic Anhydride Processes 8
4.1.1 o-Xylene Based Process 8
4.1.2 Naphthalene Based Process 13
4.2 Comparison of o-Xylene and Naphthalene Based
Processes 19
4.3 Waste Gas Characterization 22
4.4 Present Control Practices 24
5. Alternative Control Methods and Their Technical
Evaluation 27
5.1 Thermal Incinerator/Heat Recovery 28
5.1.1 System Description 28
5.1.2 Technical Evaluation for the Incinerator 30
5.1.2.1 Thermodynamic Considerations 30
5.1.2.2 Kinetic Considerations 31
5.1.2.3 NOX Formation and CO Reduction 36
5.1.2.4 Conclusion 36
5.1.3 Heat Recovery 37
5.2 Catalytic Incinerator/Feed Gas Preheating 38
5.2.1 System Description 38
5.2.2 Technical Evaluation for the Incinerator 40
5.3 Wet Scrubber/Waste Disposal 41
5.3.1 System Description 41
5.3.2 Technical Evaluation for the Scrubber 43
5.3.2.1 Theoretical Considerations 43
5.3.2.2 Practical Considerations 46
5.3.2.3 Conclusion 47
5.3.3 Liquid Waste Incineration 47
5.3.4 Maleic Anhydride Recovery 49
5.3.4.1 Process Description 49
5.3.4.2 Technical Evaluation 51
5.3.5 Biological Wastewater Treatment 52
5.3.5.1 Process Description 52
5.3.5.2 Technical Evaluation 55
iii
-------
CONTENTS (continued)
5.4 Carbon Adsorber/Waste Incineration 56
5.4.1 System Description 56
5.4.2 Technical Evaluation 60
5.4.3 Conclusion 61
5.5 Condensation/Fiber Bed Filtration 61
5.6 CHAUNY Process (Process Modification) 62
5.6.1 Process Description 62
5.6.2 Technical Evaluation 64
5.7 Summary of the Technical Evaluation 65
6. Economic Evaluation and Energy Requirements 67
6.1 Economic Evaluation 69
6.2 Energy Requirements 70
7. Most Feasible Alternatives and Recommended
Demonstration Programs 72
7.1 Selection of Most Feasible Systems 72
7.2 Recommended Demonstration Programs 77
7.2.1 Thermal Incinerator/Steam Generation 77
7.2.1.1 Increase of Operating Temperature 77
7.2.1.2 Development of Improved Incinerator 79
7.2.1.2.1 Measurement of Temperature
Distribution 79
7.2.1.2.2 Selection of burner and Mixing
Devices 79
7.2.1.2.3 Full-Scale Demonstration 79
7.2.2 Wet Scrubber/MAN Recovery 80
7.2.2.1 Pilot-Scale Testing of Improved Scrubber 80
7.2.2.2 Full-Scale Demonstration of Scrubber 80
7.2.3 Carbon Adsorber/Waste Incineration 80
7.2.3.1 Laboratory-Scale Testing of Carbon Process 80
7.2.3.2 Full-Scale Operation of the System 81
References 82
Appendices
A. Material Balance for Existing, Improved, and New
Control Systems 85
B. Cost Model and Economic Assumptions 92
c! Detailed Estimates of Capital, Capitalized and
Operating Costs 95
IV
-------
FIGURES
Number Page
1 Breakdown of phthalic anhydride production in
the continental United States 9
2 BASF process for manufacture of phthalic
anhydride from o-xylene 10
3 Badger-Sherwin-Williams process for manufacture
of phthalic anhydride from naphthalene 15
4 Thermal incinerator with waste heat boiler 28
5 Thermal incinerator with feed gas preheater 29
6 Steps required for successful incineration of
dilute fumes 30
7 Coupled effects of temperature and time on rate
of pollutant oxidation 33
8 Hydrocarbon oxidation rates in absence of flame 35
9 Rhone-Poulenc catalytic afterburner for phthalic
anhydride off-gases 39
10 Wet scrubber/waste disposal system 41
11 Schematic diagram of scrubber for phthalic
anhydride plant 42
12 Vapor pressure vs. temperature 44
13 Schematic diagram of thermal incinerator for
scrubber purge liquor 48
14 Maleic anhydride recovery process 50
15 An activated-sludge-based biological wastewater
treatment system for 1977 BPCTCA requirements 53
16 Dual-media filters and granular carbon columns to be
added to the BPCTCA system to form a BATEA system 54
-------
Number
FIGURES (continued)
Page
17 Schematic diagram of the carbon adsorption/ ^
incineration system
18 Adsorption efficiency of a carbon bed for a ^
single organic compound
19 Adsorption efficiency of a carbon bed for a ^
three-component mixture
20 Schematic diagram for the condensation/ 61
filtration system
fi T
21 CHAUNY process for PAN manufacture
vi
-------
TABLES
Number Page
1 Steam Code for BASF Process Illustrated in
Figure 2 11
2 Steam Code for Badger-Sherwin-Williams Process
Illustrated in Figure 3 16
3 Overall Material Balance for a 5.9 x 1014 Metric
Tons/Yr o-Xylene Based Phthalic Anhydride Plant 20
4 Overall Material Balance for a 5.9 x 101* Metric
Tons/Yr Naphthalene Based Phthalic Anhydride Plant 21
5 Comparison of o-Xylene and Naphthalene Based Phthalic
Anhydride Processes (5.9 x 10* Metric Tons/Yr) 22
6 Switch Condenser Off-Gas from a Representative
o-Xylene Based Plant 23
7 Switch Condenser Off-Gas from a Representative
Naphthalene Based Plant 24
8 Control Systems Presently used by the PAN Industry
in the U.S. 26
9 Composition of the Purge Liquor from Scrubbing of
the Waste Gas from a Representative o-Xylene
Based Plant 51
10 Analysis of a Waste Stream from Condenser
Off-Gas Wet Scrubber 55
11 Typical Analysis of Condenser Off-Gas from the
CHAUNY Process 65
12 Summary of Technical Evaluation of Control
Alternatives 66
13 Summary of Results of Economic Evaluation and
Energy Requirement Study 68
14 Technical, Economic, and Energy Considerations
for Thermal Incinerator/Steam Generation 73
vii
-------
TABLES (continued)
Number Page
15 Technical, Economic, and Energy Considerations
for Wet Scrubber/Waste Incineration System 74
16 Technical, Economic, and Energy Considerations
for Wet Sctubber/MAN Recovery System 75
17 Technical, Economic, and Energy Considerations
for Carbon Adsorber/Waste Incineration System 76
18 Summary of Estimated Cost and Time Requirements for
the Recommended Demonstration Programs 78
A-l Overall Material Balance for a Typical Existing
Thermal Incinerator/Steam Generation System 85
A-2 Overall Material Balance for a Typical Existing
Thermal Incinerator/Feed Gas Preheating System 86
A-3 Overall Material Balance for a Typical Existing
Wet Scrubber/Waste Incineration System 87
A-4 Overall Material Balance for an Improved Thermal
Incinerator/Steam Generation System 88
A-5 Overall Material Balance for an Improved Wet
Scrubber/Waste Incineration System 89
A-6 Overall Material Balance for a New Wet Scrubber/MAN
Recovery System 9°
A-7 Overall Material Balance for a Carbon Adsorber/
Waste Incineration System 91
C-la Capital Cost of a Typical Existing Thermal
Incinerator/Steam Generation System 96
C-lb Operating Cost of a Typical Existing Thermal
Incineration/Steam Generation System 96
C-2a Capital Cost of a Typical Existing Thermal
Incinerator/Feed Gas Preheating System 97
C-2b Operating Cost of a Typical Existing Thermal
Incinerator/Feed Gas Preheating System 97
C-3a Capital Cost of a Typical Existing Wet Scrubber/
Waste Incineration System 98
viii
-------
TABLES (continued)
Number Page
C-3b Operating Cost of a Typical Existing Wet Scrubber/
Waste Incineration System 98
C-4a Capital Cost of an Improved Thermal Incinerator/
Steam Generation System 99
C-4b Operating Cost of an Improved Thermal Incinerator/
Steam Generation System 99
C-5a Capital Cost of an Improved Wet Scrubber/Waste
Incineration System • 100
C-5b Operating Cost of an Improved Wet Scrubber/Waste
Incineration System 100
C-6a Capital Cost of a New Wet Scrubber/MAN Recovery
System 101
C-6b Operating Cost of a New Wet Scrubber/MAN Recovery
System 101
C-7a Capital Cost of a New Carbon Adsorber/Waste
Incineration System 102
C-7b Operating Cost of a New Carbon Adsorber/Waste
Incineration System 102
C-8a Capital Cost of a Typical Existing o-Xylene Based
PAN Manufacturing Plant 103
C-8b Operating Cost of a Typical Existing o-Xylene Based
PAN Manufacturing Plant 103
C-9a Capital Cost of the CHAUNY PAN Manufacturing
Process 104
C-9b Operating Cost of the CHAUNY PAN Manufacturing
Process 104
IX
-------
SECTION 1
INTRODUCTION
The Industrial Environmental Research Laboratory (IERL) of EPA
has the responsibility for insuring that air pollution control
technology is available for stationary sources. If control
technology is unavailable, inadequate, uneconomical, or socially
unacceptable, then IERL financially supports the development of
the needed control techniques. Approaches considered include:
process modifications, feedstock modifications, add-on control
devices, and complete process substitution.
The decision to develop additional control technology is based on
information provided to EPA in the form of a source assessment.
The information necessary to determine the need to develop con-
trol technology includes: 1) identification and quantification
of the emissions of criteria pollutants, hazardous pollutants,
and potentially toxic materials, and 2) the hazard potential
of the materials emitted relative to permissible concentration
levels of the materials.
A source assessment of phthalic anhydride manufacture indicated
that this industry is a source of particulate, nitrogen oxides,
sulfur oxides, carbon monoxide, and hydrocarbons from its ortho-
xylene (herein designated as p_-xylene) and naphthalene based pro-
cesses. Through this source assessment IERL has determined that
hydrocarbon emissions, particularly maleic anhydride (MAN) and
phthalic anhydride (PAN) from the main process vent stream (from
both o-xylene and naphthalene based processes) represent a
potential air pollution problem.1 This is true despite the fact
that the emissions are being reduced by 96.5% using dual incin-
eration. Therefore, EPA decided that control technology must be
developed to reduce these emissions by 99% to eliminate the
potential problems.
R. W., and T. W. Hughes. Source Assessment: Phthalic
Anhydride. Contract 68-02-1874, U.S. Environmental Protection
Agency, Research Triangle Park, North Carolina, 1976. 144 pp.
-------
Monsanto Research Corporation was contracted by EPA to evaluate
the technical and economic feasibility of control system candi-
dates for abating air emissions from phthalic anhydride manufac-
ture. The objective of this study was to select the control
system that shows the most promise for reducing, by 99%, maleic
and phthalic anhydride emissions from the main process vent
streams of the manufacturing processes.
This study involves: 1) characterization of the waste gas from
manufacturing processes; 2) survey of the present control
practices in the industry; 3) theoretical determination of the
possibility of improving existing control technology to obtain
the control efficiency desired; 4) identification of existing,
improved, and emerging unit processes that can be used to con-
struct alternative control systems; 5) technical evaluation of
alternative control systems, including manufacturing process
modification, to select technically feasible candidate systems;
6) economic evaluation and energy requirement study for existing
and candidate add-on control systems, conventional PAN manufac-
turing process, and modified manufacturing process; 7) selection
of most promising control alternatives; and 8) recommendation of
demonstration programs for the most promising alternatives.
-------
SECTION 2
SUMMARY
There are currently ten phthalic anhydride (PAN) plants in the
U.S. One plant does not have control devices for air emissions
fr?m.tlle condenser off-gas (main process vent gas). Five plants
utilize direct thermal incineration, with either feed gas pre-
heating or steam generation for waste heat recovery. These con-
trol systems have total organic removal efficiencies in the range
of 80% to 96.5%. Two plants use wet scrubbers for the off-gas
and incinerators for the scrubber purge liquor, with a reported
overall organics removal of 96%. One of these two plants has
recently installed a maleic anhydride (MAN) recovery process for
treating part of the scrubber purge liquor. The ninth plant em-
ploys a wet scrubber and a biological waste treatment facility.
The tenth plant uses a wet scrubber only. The control efficien-
cies for the last two plants are not available.
The efficiencies of the existing control systems are mainly deter-
mined by the design and performance of their corresponding direct
thermal incinerators and wet scrubbers. Theoretical analysis was
therefore performed for these two devices to determine whether
they can be improved to achieve at least a 99% efficiency for
control of PAN and MAN in the condenser off-gas.
The analysis for thermal incinerators utilized thermodynamics,
reaction kinetics, and reported conversion at different operating
conditions. This analysis revealed that 99% conversion can be
obtained by increasing the mixed gas temperature of the best unit
incinerator in the PAN industry to about 860°C. It was also con-
cluded that the same improvement in control efficiency can be
accomplished by increasing the residence time in the combustion
chamber by 55%.
I
The analysis of wet scrubbers included considerations of equilib-
rium vapor pressure, solubility of PAN and MAN at different con-
ditions, and removal of solid deposits in the scrubber. This
analysis suggested that a three-stage scrubber, consisting of
two fluid-bed stages and a packed-bed stage, would be able to
remove both MAN and PAN at greater than 99% efficiency.
Eight add-on control systems and one modified manufacturing pro-
cess (with its control device included) were identified for
technical evaluation. These alternatives were derived so as to
ensure that they will generate no secondary pollution problems
3
-------
and will recover energy when applicable. The technical evaluation
of these alternatives included their: 1) ability to reduce PAN
and MAN emissions by 99%, 2) applicability to both the o-xylene
and naphthalene based PAN processes, 3) design and operating para-
meters for achieving 99% control efficiency, 4) unique operating
or maintenance requirements, and 5) other technical advantages
and disadvantages.
As a result of the technical evaluations, four new add-on candi-
date control systems and a modified manufacturing process were
identified as technically feasible alternatives; they were sub-
jected to further economic evaluation and energy usage study.
These alternatives are direct thermal incinerator/steam genera-
tion wet scrubber/waste incineration, wet scrubber/MAN recovery,
carbon adsorber/waste incineration, and the CHAUNY PAN manufac-
turing process. Economic evaluation included estimation of
capital, capitalized and operating costs. The energy require-
ment study considered the following items, where appropriate:
1) fuel consumed in incinerator, 2) fuel needed for steam genera-
tion, 3) energy needed for generating electricity, and 4) energy
contained in the steam recovered.
Among the add-on control systems, the thermal incinerator/steam
generation is characterized by the highest operating cost and
the highest energy consumption. The wet scrubber/MAN recovery
system has the highest capital and capitalized cost, but at the
same time it has a negative operating cost, meaning that there
is a net profit from the system if the product is marketed. The
carbon adsorber/waste incineration system has the lowest capital
cost and energy consumption. The difference in capital and
operating costs between the conventional manufacturing process
and the CHAUNY process is negligible. However, the CHAUNY pro-
cess is much less energy intensive, consuming only one-seventh
of the energy required for the conventional process.
-------
SECTION 3
CONCLUSIONS AND RECOMMENDATIONS
Ninety-nine percent reduction of maleic anhydride and phthalic
anhydride emissions from phthalic anhydride plants appears to be
both technically and economically feasible. Based on a compari-
son of the advantages and disadvantages found in the technical,
economic and energy consumption evaluations, the most feasible
add-on control systems for existing plants are the following:
thermal incinerator/steam generation, wet scrubber/MAN recovery,
and carbon absorber/waste incineration. The most feasible tech-
nology for new plants is the CHAUNY process although the current
technologies in the U.S. could be used if the above add-on
systems were also employed.
Because of the current energy crisis, it is important that,
insofar as practicable, accomplishment of emissions control be
accompanied by increased product yield and decreased energy con-
sumption. Hence, long term solutions must be considered along
with those providing immediate emissions reductions.
The recommended demonstration programs for the most feasible add-
on systems are based on the development stage of the individual
unit operations comprising the control systems and the technical
evaluation of each system. They are summarized in the following
subsections.
3.1 SHORT-TERM MEASURES
The short-term measures involve only the use of the thermal
incinerator/steam generation add-on control system. For the two
plants which currently use this system, 99% reduction of maleic
anhydride and phthalic anhydride emissions could be achieved by
increasing the operating temperature to about 860°C. The higher
incineration temperatures are acceptable only as a temporary
expedient since this mode of operation increases the plant's
consumption of natural gas by 11%.
Better short term solutions may be available for plants which
may use a thermal incinerator/steam generation system in the
future. The first alternative appears to be to increase the
-------
contact time of the flue gas in the incinerator. Theoretical
calculations indicate that a 55% increase in the combustion
volume may provide a 99% reduction in maleic anhydride and
phthalic anhydride emissions. Another alternative appears to
improve the burner design thereby increasing the turbulence in
the combustion zone.
Use of the thermal incinerator/steam generation system results
in two advantages which should not be overlooked. The system
can be used to replace a portion of a phthalic anhydride plant's
boilers. It will also control carbon monoxide emissions which
may be required by state or local EPA agencies.
3.2 PERMANENT MEASURES
The permanent measures involve the use of add-on control devices
which reduce emissions of maleic anhydride and phthalic anhydride
without imposing an energy penalty on the plants. There are two
alternatives which warrant further study. These are the wet
scrubber/MAN recovery and carbon absorber/waste incineration.
Development of a wet scrubber along with maleic anhydride and
phthalic anhydride recovery should be pursued. Wet scrubber data
should be sought from industry which will permit the design of
99% efficient scrubbers. Concurrent with this, a process should
be developed for recovering both maleic anhydride and phthalic
anhydride from the scrubber liquid purge (blowdown). The
demonstration program should include:
1. Pilot plant testing of the improved three-stage
scrubber design recommended in the design study
(Section 7.2.2.1) .
2. Full-scale demonstration of the improved wet
scrubber and waste product recovery processes
(Section 7.2.2.2) .
Development of a carbon absorber/waste incineration add-on
control system should be pursued. High priority should be
given to obtaining laboratory-scale and/or pilot-scale data
on the use of activated carbon for emissions control. Specifi-
cally, the absorptive capacity and the regeneration efficiency
of the activated carbon for maleic anhydride, phthalic anhydride
and by-product organic materials need to be demonstrated.
Applicability of a maleic anhydride and phthalic anhydride
recovery process for the adsorbed materials should be determined.
In the event that the recovery process is not feasible, then
incineration of the recovered organics is needed. Initial
development should be made at the laboratory scale. If the
laboratory results are sufficiently promising, development work
should follow at the pilot-scale level.
-------
Use of either a scrubber or an activated carbon recovery system
will not control carbon monoxide emissions. If this is required
by a state or local EPA agency, an additional control device,
probably an incinerator, will be needed.
-------
SECTION 4
SOURCE DESCRIPTION AND PRESENT CONTROL PRACTICES
4.1 PHTHALIC ANHYDRIDE PROCESSES
Phthalic anhydride (PAN) is manufactured by the catalytic vapor-
phase oxidation of either o-xylene of naphthalene. In 1975, 67%
of the PAN produced in the continental United States was obtained
from o-xylene and 33% from naphthalene as shown in Figure 1.
4.1.1 o-Xylene Based Process
Phthalic anhydride is presently produced in the United States
from o-xylene by the von Heyden process, which is licensed in
somewhat different forms by Chemiebau and Lurgi, and by the
Badische Anilin und Soda-Fabrik (BASF) process, developed in the
late 1960s. In 1975, 79% of the o-xylene based production in the
continental United States involved use of the BASF process and
21% utilized the von Heyden process.2'3 Figure 2 is a schematic
flow diagram for the BASF process. The numbered streams in
Figure 2 are identified in Table 1.
The vapor-phase oxidation of o-xylene to phthalic anhydride
takes place on a vanadium pentoxide catalyst. The reaction can
be represented by the following equation:
3+ 302 catalyst
'CH3
o-xylene oxygen phthalic water
anhydride
The theoretical yield is 140 kg PAN/100 kg o-xylene.
2Anderson, E. V. Phthalic Anhydride Makers Foresee Shortage.
Chemical and Engineering News, 53(26):10-11, 1975.
3Phthalic Anhydride. Chemical Marketing Reporter, 207(1):9, 1974
8
-------
PHTHALIC ANHYDRIDE PLANTS
10 PLANTS
o-XYLENE BASED
~ 7 PLANTS
67% OF CAPACITY
NAPTHALENE BASED
3 PLANTS
33% OF CAPACITY
BASF PROCESS
4 PLANTS
53% OF CAPACITY
VON HEYDEN PROCESS
3 PLANTS
14% OF CAPACITY
BADGER-SHERWIN-WILLIAMS
PROCESS 2 PLANTS
24% OF CAPACITY
OTHER PROCESS
1 PLANT
9% OF CAPACITY
Figure 1. Breakdown of phthalic anhydride production
in the continental United States.
-------
ATMOSPHERIC EMISSIONS
STEAM
•T~r~~r-^
JQ^ J© " *~ f <3
1 1 *
A IB 1 •* —
1
"—• ' V-J STEAM , Q
PROCESS ©X » 1 E]—
AIR L/
- OIL
TO SCRUBBER K -*-j
TO SECONDARY JM rSlCTEAM-.
SCRUBBER -<~H. ^ iiom-"
OR INCINERATOR j^1-^, ® WATER -
N OIL
^V
BASF PROCESS FOR THE MANUFACTURE OF
PHTHALIC ANHYDRIDE FROM 0-XYLENE
A o-XYLENE STORAGE L INCINERATOR
B SO 2 STORAGE M- HEATER
C. 0-XYLENE PREHEATER N. CRUDE PRETREA1
D COMPRESSOR 0- PRECOOLER
E. AIR PREHEATER P. STRIPPER
F. REACTOR (FIXED BED) Q. EVAPORATOR
G MOLTEN-SALT HEAT EXCHANGER R. EVAPORATOR
H. WASTE HEAT BOILER *. RECTIFIER
1. SWITCH CONDENSERS T. PHTHALIC ANHY
I. CRUDE PRODUCT STORAGE U. FLAKER
K. SCRUBBER V. BAGGING MACH
1
i85 f~{ ' TT°'
S 3 -fc^STEAM t
JO-I® H "^S)W/ | '
" ' ® f 1*011
- G f*" *^ >*-
LJ | |
1 'IT"
^JL^i
j
TO SECO
SCRUBBE
INCINER
iVACUUM
i rY# ^ ' 1 (
r^ rS,
)». P 0 OIL R
T V
L
®
i
, i i^fc. SCRUBBER VENT GAS
A®
•«-WATER
K .— ^-*- INCINERATOR STACK GAS
^~L-J i ®». CRUDE PHTHALIC ANHYDRIDE
1 • SIORAGt IANK VtNl tAi
MDARY ®1 M
R OR-< 1 _^ 1 1
ATOR ! ,^tY \ (S)
STEAM
WATER
IVACUUM , •«— AIR
/Ti-i k. L ,— L
(jy *^ ®
"*^8<
^ ,J ,— ^ rilGITIVE EMISSIONS
TO SECONDARY
INCINERATOR
A
^Y I
U ' ' ** J C»)^REFINED PHTHALIC ANHYDRIDE
|® ® ^ Fl ftKFR ANP BflfifiFR WT 4<;
,^J^_ i
1
I -r*
© E
WENT TANK
j ,
DRIDE STORAGE FWHALIC ANHYDRIDE
1 * TRANifUKI
LOADING FACILITY
INE
J-*- TO WAREHOUSE
1*- TRANSPORT LOAPIN^ F""" ITV
— »- rn T«UK CAR1;
Figure 2. BASF process for manufacture of phthalic anhydride from o-xylene.
-------
TABLE 1. STREAM CODE FOR BASF PROCESS ILLUSTRATED IN FIGURE 2
Stream
Identification
Stream
Id-entif ication
1 o-Xylene feed
2 Sulfur -dioxide
3 Filtered air
4 Reactor feed
5 Reactor product
6 Boiler effluent
7 Crude product
8 Condenser off-gas
9 Scrubber vent
10 Scrubber liquid purge
11 Crude PAN product
12 Pretreatment exhaust
13 Pretreated crude
14 Stripping column exhaust
15 Stripping column overhead
16 Rectifying column feed
17 Rectifying column vacuum exhaust
18 Rectifying column bottom product
19 Distillation light ends
20 Refined PAN
21 Water to steam generator
22 Generated steam
23 Water to waste-heat boiler
24 Steam from waste-heat boiler
25 Scrubber makeup
26 Incinerator stack gas
27 Water to cooling coil
28 Steam from cooling coil
29 Water to cooling coil
30 Steam from cooling coil
31 Incinerator fuel
32 Combustion air
33 Flaker and bagger vent
34 o-Xylene storage vent
35 Crude product storage vent
36 Refined PAN storage vent
37 Loading facility vent
38 Fugitive emissions
-------
The principal side reactions are the result of the further oxi-
dation of phthalic anhydride to maleic anhydride, carbon oxides,
and water. Based on the work of Bernardini and Ramacd,4 the
main sequence of reactions in the oxidation of o-xylene is as
follows:
o-methyl-
benzyl
alcohol
o-tolual-
dehyde
COOH
CH3
o-toluic
acid
phthalide phthalic
anhydride
O
maleic
anhydride
CO2, CO, H20
Filtered air (stream 3) is compressed to a gage pressure of 48.2
kPa to 55 2 kPa and preheated to 149°C using steam. The liquid
xyLne feel is also preheated and vaporized by injection -to the
hot air stream. A small amount (0.5% to 2.5% by weight) of sul-
fur dioxide (stream 2) is added to the feed stream in order to
maintain catalyst activity. In the BASF process, the weight
?atio of air to xylene in the feed stream (stream 4) to the re-
actor (unit F) is 25, which is equivalent to 1 mole percent
xylene in the feed to the reactor. (Weight ratios as low as 20
and as high as 34 have been reported in the literature.5 Excess
air is employed to ensure that the mixture is below the lower
explosive limit of 1.5 mole percent o-xylene After removal of
unvaporized feed, the reactants are fed to the fixed-bed reactor,
which operates at about 380°C within a range of 300°C to 390°C.
4Bernardini, F. and M. Ramacci.
to Phthalic Anhydride. Chimica
1966.
5Spitz, P. H. Phthalic Anhydride Revisited.
cessing, 47 (11):162-168, 1968.
12
Oxidation Mechanism of o-
e 1'Industria (Milan),
. 9 17,
Hydrocarbon Pro-
-------
A molten salt (sodium-potassium nitrate-nitrite eutectic6) is
circulated on the shell side of the reactor to remove the heat
produced in the exothermic reaction. The molten salt passes
through an external heat exchanger (unit G) where high pressure
steam is produced. The gases leave the reactor at 375°C (stream
5) and pass through a waste heat boiler (unit H) for additional
steam generation. The process vapors then pass from this waste
heat boiler to the switch condenser where the crude PAN product
is separated from the process air stream.
Due to the excess (567%) of air employed in the reactor, the par-
tial pressure of the phthalic anhydride in the reactor effluent
gas stream is such that the dew point is below the melting point
(130.8°C) of phthalic anhydride. Hence, the product condenses
as a solid. This operation is carried out in a parallel bank of
tubular condensers (unit I) which are alternately heated and
cooled by separate heat-transfer oil streams on an automatically
controlled cycle. During the cooling portion of the cycle,
phthalic anhydride crystallizes on the outer surfaces of the
finned tubes. During the heating portion of the cycle, the
solid phthalic anhydride is melted and then transferred to the
crude product storage tank (unit J). It is the process gas
leaving these condensers that offers the greatest potential for
an air pollution problem. The control of emissions from this
stream is the center of this study. For illustrative purposes,
a wet scrubber and a liquid waste incinerator (units K and L) are
shown in Figure 2 for control of this stream.
The crude product which contains 99% to 99.5% phthalic anhydride
as phthalic acid is stored at 149°C under atmospheric pressure.
The refining operation consists of two steps: a heat treatment
step followed by a vacuum distillation. The crude product
(stream 11) passes through a preheater (unit M) and then to the
pretreatment tank (unit N), where it is held at an elevated
temperature under vacuum for 8 to 12 hours. The purpose of the
heat treatment is to decompose color-forming compounds and to
convert them to higher molecular weight materials that can be
separated from the main product by distillation. At the same
time, dissolved phthalic acid is dehydrated to the anhydride,
and the associated water and other low-boiling materials, such
as maleic anhydride and benzoic acid, are partially evaporated
and removed through the vacuum jet ejector exhaust stream
(stream 12). This exhaust stream is sent to either the main
control device (unit K) or a separate control device (which may
be another scrubber or another incinerator). The pretreatment
tank is discharged to a continuous distillation system consist-
ing of a stripping column (unit P) and a rectifying column
(unit S), both of which operate in the range of 2.67 kPa to
6Schwab, R. F., and W. H. Doyle. Hazards in Phthalic Anhydride
Plants. Chemical Engineering Progress, 66(9):49-53, 1970.
13
-------
26 7 kPa absolute pressure. The discharge stream passes through
a precooler (unit O) and enters the stripping column at about
186°C Maleic anhydride and benzoic acid are separated in the
overhead stream (stream 15) which is sent to the main incinerator
(unit L) or to a secondary control device (via stream 19) for
disposal The bottom stream from the stripping column passes to
the evaporators (reboilers, units Q and R). Part of the stream
is recycled to the stripping column, and the remainder (stream 16)
enters the rectifying column (unit S).
The bottom stream (stream 18) from the rectifying column^(the
residue) is sent to the main incinerator or a secondary incinera-
tor for disposal. Phthalic anhydride (99.99%) is taken overhead,
condensed, and sent to the product storage tank (unit T) where it
is stored at atmospheric pressure and 149°C.
The phthalic anhydride in the storage tanks can be pumped to tank
cars for shipment as a liquid, or it can be sent to a flaking
machine (unit U) for solidification. The flaked product is
weighed and packaged in 36-kg bags by a bagging machine (unit V).
4.1.2 Naphthalene Based Process
All of the phthalic anhydride currently produced from naphthalene
in the United States is made via the fluid-bed Process. Two of
the three plants now in operation, which account for 73% of the
naphthalene based production, use the ^"r^erwin-Williams
process, first commercialized in the early 1960s. The Koppers
Company plant in Bridgeville, Pennsylvania, uses a fluid-bed pro-
cess developed by American Cyanamid, which built and operated the
plant in the 1950s. This process was never licensed, and the
plant was eventually taken over by Koppers when they acquired all
of Cyanamid-B facilities at the Bridgeville location 5 Figure 3
is a schematic flow diagram for the Badger-Sherwin-Williams pro-
cess, which is described below. Process streams numbered in
Figure 3 are identified in Table 2.
The vapor-phase oxidation of naphthalene to phthalic anhydride on
a vanadium oxide catalyst can be represented by the following
equation:
catalyst . rpV1 \Q + 2COz + 2H2O
4 1/2 02
naphthalene oxygen phthalic carbon water
anhydride dioxide
Theoretical yield is 116 kg PAN/100 kg naphthalene.
14
-------
ATMOSPHERIC EMISSIONS
®
PROCESS.
AIR""*"
(D
STEAM.
WATER-
t
I
-4 H K?IL
^
(5L
T
®
WATER
PROCESS
AIR
• NAPTHALENE STORAGE TANK VENT GAS
VACUUM TO
SCRUBBER OR
INCINERATOR
- SCRUBBER VENT CAS
INCINERATOR STACK GAS
-*- FUGITIVE EMISSIONS
• CATALYST STORAGE HOPPER VENTS
CRUDE PHTHALIC ANHYDRIDE
STORAGE TANK VENT GAS
REFINED PHTHALIC ANHYDRIDE
STORAGE TANK VENT GAS
*- TRANSPORT LOADING FACILITY VENT
*- FLAKER AND BAGGER VENT
A. NAPTHALENE STORAGE
B. COMPRESSOR
C. AIR PREHEATER
D. REACTOR (FLUID BED)
E. CATALYST FILTER
F. STEAM GENERATOR
G. PARTIAL CONDENSER
H. SWITCH CONDENSERS
I. CRUDE STORAGE
J. SCRUBBER
K. INCINERATOR
L. HEATER
M. CRUDE PRETREATMENT TANK
N. PRECOOLER
0. DISTILLATION COLUMN
P. PHTHALIC ANHYDRIDE STORAGE
Q. FLAKER
R. BAGGING MACHINE
TO WAREHOUSE
BADGER-SHERWIN-WILLIAMS PROCESS FOR MANUFACTURE OF
PHTHALIC ANHYDRIDE FROM NAPTHALENE
Figure 3. Badger-Sherwin-Williams process for manufacture
of phthalic anhydride from naphthalene.
-------
TABLE 2. STREAM CODE FOR BADGER-SHERWIN-WILLIAMS
PROCESS ILLUSTRATED IN FIGURE 3
Stream
Identification
Stream
Identification
1 Naphthalene
2 Air
3 Reactor product
4 Crude from partial
condenser
5 Partial condenser
off-gas
6 Crude from switch
condenser
7 Crude product
8 Switch-condenser
off-gas
9 Scrubber vent
10 Scrubber makeup
11 Scrubber liquid
purge
12 Crude PAN
13 Pretreatment vacuum
exhaust
14 Distillation column
feed
15 Distillation column
vacuum exhaust
16 Distillation column
light ends
17 Distillation column
bottom product
18 Refined PAN
19 Incinerator stack gas
20 Incinerator fuel
21 Combustion air
22 Water to steam
generator
23 Generated steam
24 Cooling water
25 Steam
26 Flaker and bagger
vent
27 Naphthalene storage
vent
28 Crude product
storage vent
29 PAN storage vent
30 Catalyst storage
hopper vents
31 Loading facility
vent
32 Fugitive emissions
16
-------
The side reactions are the consequence of the further oxidation
of phthalic anhydride to maleic anhydride, carbon oxides, and
water. Based on the work of loffe, et al.,7 the main sequence of
reactions in the oxidation of naphthalene is as follows:
H
H
naphthalene 1,4-dihydroxy-
naphthalene
1,4-naphtho-
quinone
phthalic
anhydride
C02, CO, H2O
maleic anhydride
Feed materials used in the Badger-Sherwin-Williams process con-
sist of naphthalene and air. Filtered air (stream 2 in Figure 3)
is compressed and heated to ]49°C. The hot air then enters the
bottom of the reactor (unit D) and passes through an air dis-
tributor plate before entering the fluidized catalyst bed.
Liquid naphthalene (stream 1), which is stored at 90°C to 102°C,
is pumped from the storage tank (unit H) and injected directly
into the fluidized bed above the air distributor plate. The
naphthalene is immediately vaporized and dispersed throughout
the bed upon contact with the hot catalyst and reaction air.
In the Badger-Sherwin-Williams process, the weight ratio of air
to naphthalene fed to the reactor is 10:1 to 12:1, which places
the mixture within the explosive limits.8 This low air-to-
naphthalene ratio is made possible by the inerting effect of the
7Ioffe, I. I., et al. Kinetics and Vapor Phase Oxidation
Mechanism of Aromatic Hydrocarbons, VI. Zhurnal Fizicheskoi
Khimii (Journal of Physical Chemistry), 29 (4):692-698, 1955.
8Graham, J. J., P. F. Way, and S. Chase. Phthalic Anhydride
by Fluid Bed Process. Chemical Engineering Progress,
58(1):96-100, 1962.
17
-------
fine catalyst particles in the fluidized bed reactor.8 Within
the fluid bed, a uniform temperature is maintained (to ± 5°C) in
the range of 340°C to 380°C.^ Reactor pressure is set by the
backpressure of equipment downstream.
Removal of the heat produced by the exothermic reaction is accom-
plished by means of cooling tubes located directly in the cata-
lyst bed. Water is circulated in these tubes and steam is thereby
generated directly without the need of a secondary heat-transfer
fluid.
Effluent gases leaving the reactor are cooled to about 260°C be-
fore entering the catalyst filter unit (unit E) in order to pre-
vent secondary reactions taking place on the filters. The en-
trained catalyst particles are separated from the gas stream by
specially designed porous ceramic filter elements. Process air
is used to periodically blow back the filters, and the catalyst
particles are returned directly to the reactor bed.
Due to the low air-to-naphthalene ratio employed in the fluid-
bed process, the crude phthalic anhydride is recovered both as
a liquid and a solid. Between 40% and 60% of the product is
obtained directly as a liquid in the partial condenser (unit G),
and the remainder is condensed as a solid in the switch con-
densers (unit H). The latter operate in the same manner as those
described for the BASF process. However, the partial condensation
of product, plus the reduced air rate, greatly reduces the load
on the switch condensers with a concomitant reduction in size
and cost. The tail gases (stream 8) leave the switch condensers
at 66°C and are sent either to a water scrubber (unit J) or
directly to the incinerator (unit K) or other control devices.
In Figure 3, a scrubber followed by an incinerator is shown for
illustrative purposes.
The crude product is stored as a liquid at 149°C and atmospheric
pressure (unit I). Crude phthalic anhydride (stream 12) from the
crude product storage tank passes through a preheater (unit L)
and then to the pretreatment tank (unit M) in much the same manner
as in the BASF process. Treating chemicals consisting of maleic
anhydride (1 g per kg PAN), and sodium hydroxide (0.05 g per kg
PAN) are added to promote the pretreatment process. In the
Badger-Sherwin-Williams process, the final product is obtained
by means of a batch distillation column which operates at an
absolute pressure of 2.66 kPa. The light ends (stream 16) are
taken off at a reflux ratio of 40 and sent to the incinerator for
disposal. The main.cut is then taken at a reflux ratio of 1,
9Graham, J. J. The Fluidized Bed Phthalic Anhydride Process,
Chemical Engineering Progress, 66(9):54-58, 1970.
18
-------
condensed, and sent to the phthalic anhydride storage tank
(unit P), where it is held at 149°C and atmospheric pressure.
The residue, or bottoms product, is sent to the incinerator.
The refined phthalic anhydride (99.7% minimum) can be pumped to
tank cars for shipment in liquid form, or it can be sent to flak-
ing and bagging machines (units Q and R) for shipment as a solid.
4-2 COMPARISON OF O-XYLENE AND NAPHTHALENE BASED PROCESSES
The flow diagram for the BASF process for the manufacture of
phthalic anhydride from o-xylene was given in Figure 2. Table 3
is an overall material balance for a typical 5.9 x 104 metric
tons/yr plant.1 The flow diagram for the Badger-Sherwin-Williams
process for phthalic anhydride manufacture from naphthalene was
given in Figure 3. Table 4 is an overall material balance for a
typical 5.9 x 104 metric tons/yr plant.1
Table 5 presents a comparison of the o-xylene and naphthalene
based phthalic anhydride processes derived from data presented in
Tables 3 and 4. Based on the data shown in Table 5, the following
conclusions about process yields and losses can be made:
1. The selection of o-xylene or naphthalene as a feedstock
in the manufacture of phthalic anhydride is based on the
cost of the feedstock since the yields on a kg PAN/kg
FEED are essentially equal.
2. The naphthalene based process is a more selective process
for PAN production since its yield on a moles PAN/moles
FEED is higher than the o-xylene based process.
3. The naphthalene based process emits 24% more carbon from
the switch-condenser off-gas than does the o-xylene
based process as expected from its process chemistry
(deliberate byproduct formation of carbon dioxide).
4. The o-xylene process emits 150% more organic material
than does the naphthalene based process.
5. The o-xylene process also emits 200% more carbon monoxide
than does the naphthalene based process.
6. The amount of total PAN and MAN losses are 483 kg/hr and
199 kg/hr for the o-xylene and naphthalene based processes,
respectively. Hence, recovery of PAN and MAN could be
very economically attractive. Recovery of both materials
would essentially solve the most troublesome emission
problems.
19
-------
TABLE 3. OVERALL MATERIAL BALANCE FOR A 5.9 x 104 METRIC TONS/YR
O-XYLENE BASED PHTHALIC ANHYDRIDE PLANT1fa
to
o
Stream number:"
Description:
Temperature, °C:
Gage pressure, kPa:c
Component
Sulfur dioxide
Carbon monoxide
Carbon dioxide
Nitrogen
Oxygen
Phthalic anhydride
Maleic anhydride
Benzoic acid
o-Xylene
m- and p-Xylene
Misc. organics
Water
TOTAL
18 18
o-Xylene Switch-condenser Rectifying
off-gas column bottom
product
27 66
0 26
Flow rates, kg/hrd
34.1
1,095.9
3,784.6
140,862.8
32,819.1
167.3 7.3
315.4
20.4
7,098.6
334.6
34.1
6,339.1
7,433.2 185,438.7 41.4
20
Product PAN
149
20
7,272.7
14.6
7,287.3
Original data developed by Houdry.
Stream numbers are the same as those shown in Figure 2
'Blanks indicate data not available.
Blanks indicate no mass flow of component.
-------
TABLE 4. OVERALL MATERIAL BALANCE FOR A 5.9 x 101* METRIC TONS/YR
NAPHTHALENE BASED PHTHALIC ANHYDRIDE PLANT1'3
ro
Stream number:'3
Description:
Naphthalene
8
Switch-condenser
off-gas
17
Distillation column
bottom product
18
Product PAN
Temperature, °C:
Pressure, kPa:c
Component
Naphthalene
Phthalic anhydride
Maleic anhydride
Naphthoquinone
Misc. organics
Oxygen
Nitrogen
Carbon dioxide
Carbon monoxide
Water
TOTAL
93 66
0
Flow rates, kg/hr1*
7,497.3
148.2 48.2
50.9
5.0 48.2
37.7 37.7
12,208.2
68,177.3
7,030.4
366.4
2,290.9
7,535.0 90,277.3 134.1
149
0
7,272.7
14.6
7,287.3
Original data developed by Houdry.
Stream numbers are the same as those shown in Figure 3.
•»
'Blanks indicate data not available.
Blanks indicates no mass flow of component.
-------
TABLE 5. COMPARISON OF o-XYLENE AND NAPHTHALENE BASED
PHTHALIC ANHYDRIDE PROCESSES ( 5.9 x 10^
METRIC TONS/YR PLANTS)a
Parameter
o-Xylene
process
b
Naphthalene
process
0.97C
0.84C
Yield
kg PAN/kg FEED 0.98
Moles PAN/moles FEED 0.70
Losses^
Total carbon, kg/hr 1,778
Phthalic anhydride, kg/hr 167.3
Maleic anhydride, kg/hr 315.4
Carbon monoxide, kg/hr 1,095.9
Carbon dioxide, kg/hr 3,784.6
Benzoic acid, kg/hr 20.4
Naphthoquinone, kg/hr _
2,198
148.2
50.9
366.4
7,030.4
5.0
aDerived from data in Tables 3 and 4.
^FEED =95% o-xylene; balance of FEED is not converted to
PAN. ~
CFEED is 99.0 to 99.7% naphthalene.
dLosses correspond to materials in switch-condenser off-gas.
Future PAN plants may be based on naphthalene feedstock due to
tight supplies of petroleum derived o-xylene. Naphthalene will
be readily available from future coal conversion plants. Tight
supplies of o-xylene will not prevent the PAN producers using the
BASF process from operating since they can operate on naphthalene.
The impact of this changeover in feedstock on emissions is not
known.
4.3 WASTE GAS CHARACTERIZATION
In order to evaluate the different control systems on the same
basis for the switch condenser off-gas (main process vent gas),
a representative plant was defined for each of the two types of
manufacturing processes described above. The waste gases from
the main process of these two representative plants were then char-
acterized for the subsequent technical and economic evaluations.
A representative o-xylene based phthalic anhydride plant is
defined to be one using the BASF process and having a production
capacity of 5.9 x 104 metric tons/yr.1 The BASF process is
22
-------
specified because it is the newest and most widely used tech-
nology, accounting for 79% of the o-xylene based production in
1974. *,3 The standard size BASF plant has two production trains
of two reactors each, and a capacity of 5.9 x lO1* metric tons/yr.
The flow rate and composition of the switch condenser off-gas are
shown in Table 6.
TABLE 6 . SWITCH CONDENSER OFF-GAS FROM A REPRESENTATIVE
O-XYLENE BASED PLANT
Component
Flow rate,
kg/hr
Volume,
%
SO2
CO
CO 2
N2
02
PAN
MAN
Benzoic acid
Water
TOTAL
34.1
1,095.9
3,784.6
140,862.8
32,819.1
167.3
315.4
20.4
6,339.1
0.01
0.60
1.32
76.94
15.69
0.0173
0.0491
0.003
5.39
185,438.7 100.00
Partial
pressure,
mm Hg
0.08
4.9
10.7
624.8
127.4
0.141
0.399
0.02
43.8
812.0
Volumetric flow rate = 2,665 m3/min (94,150 scfm)
A representative naphthalene based phthalic anhydride plant is
defined to be one using the Badger-Sherwin-Williams process and
having a production capacity of 5.9 x 104 metric tons/yr.1 The
Badger-Sherwin-Williams process is the dominant fluid-bed tech-
nology and accounted for 73% of the naphthalene based production
in 1974.2'3 There is no standard size for the Badger-Sherwin-
Williams fluid-bed reactors. Hence, the representative plant
capacity was chosen to be the same as that for the o-xylene based
Plant. The flow rate and composition of the switch condenser
°ff~gas are presented in Table 7. 1
of the maleic anhydride and most of the phthalic anhydride
contained in the waste gas are in vapor phase. There is a short-
term presence of PAN dust in the stream due to the opening and
closing of the valves used to isolate the switch condensers.
rhe waste gases from both of the representative plants can be
Characterized as large-volume, low-temperature, low-pressure,
low-heat-value gases containing air, CO, C02, and hydrocarbons.
23
-------
TABLE 7. SWITCH CONDENSER OFF-GAS FROM A REPRESENTATIVE
NAPHTHALENE BASED PLANT
Component
PAN
MAN
Naphthoquinone
02
N2
CO 2
CO
H2O
TOTAL
Volumetric flow
Flow rate,
kg/hr
148.2
50.9
5.0
12,208.2
68,177.3
7,030.4
366.4
2,290.9
90,277.3
rate = 1,270
Volume ,
%
0.0321
0.0167
0.001
12.24
78.09
5.12
0.42
4.08
100.00
m3/min (44
Partial
pressure,
mm Hg
0.261
0.136
0.01
99.4
634.1
41.6
3.4
33.1
812.0
,880 scfm)
4.4 PRESENT CONTROL PRACTICES
There are four types of control systems currently being used by
the phthalic anhydride industry for controlling the switch
condenser off-gases:
• Direct thermal incinerator
• Wet scrubber and incineration of the scrubber waste
liquor
• Wet scrubber and recovery of MAN from the scrubber
waste liquor
• Wet scrubber and biological waste treatment facility
24
-------
The type of control system and its control efficiency for each of
the ten plants in the United States are given in Table 8.10-16
Among these ten plants, one does not have devices for control of
the off-gas.10 Five plants utilize the method of direct thermal
incineration, with total organics removal of 80% to 96.5%.11~1't
Two plants use wet scrubbers for the off-gas scrubbing and in-
cinerators for the scrubber purge liquor, with a reported overall
organics removal of 96%.12 One of these two plants has recently
installed a maleic anhydride recovery process for treating part
of the scrubber purge liquor. The ninth plant employs a system
containing a wet scrubber and a biological waste treatment facil-
ity. The tenth plant uses a wet scrubber only. Available data
for the last two plants were not sufficient to develop an estimate
of the control efficiency.
10Personal communication with L. B. Evans, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina (data
originally supplied by Stepan Chemical Company), 12 June 1975.
^Personal communication with A. B. Netzley, Air Pollution Con-
trol District, County of Los Angeles, 28 March 1975.
l2Schwartz, W. A., et al. Engineering and Cost Study of Air
Pollution Control for the Petrochemical Industry, Volume 7:
Phthalic Anhydride Manufacture from Ortho-Xylene. EPA 450/
3-73-006g, U.S. Environmental Protection Agency, Research
Triangle Park, North Carolina, 1975. 108 pp.
l3Personal communication with H. M. Lacy, Monsanto Company, Texas
City, Texas, 7 April 1975.
llfPersonal communication with T. Brennan, Bay Area Air Pollution
Control District, San Francisco, California, 23 June 1976.
l5Personal communication with L. B. Evans, U.S. Environmental
Protection AGency, Research Triangle Park, North Carolina (data
originally supplied by Koppers Company), 12 June 1975.
16Personal communication with M. A. Pierle, Monsanto Company, St.
Louis, Missouri, April 1976.
25
-------
TABLE 8. CONTROL SYSTEMS PRESENTLY USED BY THE PAN INDUSTRY IN THE U.S.
to
Plant Nominal capacity,
no. Company Location 10 3 metric ton/yr
1 Allied Chemical El Segundo, Calif. 15.9
2 BASF Wyandotte S. Kearny, N.J. 59.0
3 Exxon Corp. Baton Rouge, La. 40.8
4 Koppers Co. Cicero, 111. 79.4
5 Monsanto Co. Texas City, Texas 59.0
6 Std. Oil Calif . Richmond, -Calif . 22.7
7 Stepan Chemical Millsdale, 111. 22.7
Co.
8 Koppers Co. Bridgeville, Pa. 40.8
9 Monsanto Co. Bridgeport, N.J. 40.8
10 U.S. Steel Neville Island, Pa. 68.0
Raw material
o-Xylene
o-Xylene
o-Xylene
o_-Xylene
o-Xylene
o-Xylene
o-Xylene
Desulf ur ized
naphthalene
Petroleum
naphthalene
Desulf ur ized
naphthalene
Process Type of control system
Chemiebau (von Heydon) Thermal incinerator
BASF , Wet scrubber for the
waste gas and in-
cinerator for the
scrubber purge
liquor.
BASF Thermal incinerator
BASf Wet scrubber for the
waste gas and either
incinerator or MA
recovery process for
the scrubber purge
liquor.
BASF Thermal incinerator
Lurgi (von Heydon) Thermal incinerator
Chemiebau (von Heydon) None1"
OWN(fluid bed) Wet scrubber15
Badger-Sherwin-Williams Wet scrubber and bio-
logical treatment
facility.16
Badger-Sherwin-Williams Thermal incinerator
Control efficiency
89% total organics11
96% total organics12
96.5% total organics13
96.5% total organics
for the system using
incinerator.12 Un-
known for the system
with MAN recovery.
96.5% total organics13
23% reactive
hydrocarbons * "*
NA3
-b
NA
80% total organics12
Not applicable.
Insufficient data for accurate calculation.
-------
SECTION 5
ALTERNATIVE CONTROL METHODS AND THEIR TECHNICAL EVALUATION
The alternative methods considered for controlling PAN and MAN
emissions from the switch condenser off-gas include the
following:
• Direct thermal incinceration with heat recovery
• Direct catalytic incineration with feed gas pre-
heating
• Wet scrubbing with waste disposal
• Carbon adsorption and incineration
• Condensation and filtration
• Process modification
When direct thermal incineration or direct catalytic incineration
is applied, the waste gas is directed to the stack after waste
heat recovery without further treatment. In the case of wet
scrubbing, the purge liquor presents a waste disposal problem.
The treatment alternatives for this liquid stream include incin-
eration, MAN recovery, and biodegradation in an activated sludge
Process. The carbon adsorption/incineration method consists of
carbon adsorption and subsequent incineration of the concentra-
ted organic stream. When condensation/filtration is applied,
PAN and MAN vapors are condensed into solid particles by cooling
and the water soluble particles are subsequently removed by a
fiber bed. The process modification includes the use of a modi-
fied PAN manufacturing process (the CHAUNY process) developed by
Rhone-Poulenc Industries in France.
The individual processes involved in the above-mentioned alter-
native control methods are further described in the following
subsections. Also included is the technical evaluation of each
control method in relation to (a) the ability to reduce MAN and
pAN emissions by 99%; (b) the important process and operating
Parameters that will affect emissions from the control process;
(c) the changes in process or operating parameters that will
affect 99% removal efficiency; (d) the applicability of the con-
trol method to both o-xylene and naphthalene based PAN processes;
(e) the unique operating or maintenance requirement of the pro-
cesses involved; and (f) technical advantages and disadvantages.
27
-------
5.1 THERMAL INCINERATOR/HEAT RECOVERY
5.1.1 System Description
Thermal incincerators are used to control the release of hydro-
carbon fumes from industrial processes. Inside the incinerator
the hydrocarbon pollutants are oxidized to form carbon dioxide
and water in the presence of oxygen at a temperature above their
ignition temperatures. The incinerator consists of a refractory-
lined combustion chamber with a raw gas burner fired at one end
to supply heat. Since a large amount of fuel is needed to heat
the total volume of waste gas, and because the price of fuel is
high, some form of heat recovery is practiced to make use of the
heat energy contained in the flue gases discharged from the in-
cinerator.
Schematic diagrams of two types of incinerator/heat recovery
systems which are applicable to phthalic anhydride plants are
shown in Figures 4 and 5. The heat recovery devices and the
problems associated with them are discussed in Section 513
WASTE GAS
400 °C 650 STM ,
^\ yj PURGE /STARTUP B
\Sy \J AlK BLOWER
• *NAIURAL / |
r GAS / !
670 °C TO 820 °C |
/V\/1 INCINERATOR 1
N\ '
~~\^ BURNER V-V
\
^
OILER
rS-
n
w
r >
X
"Trt
j STA
1
i,
D^ r
^^
V '
CK
Figure 4. Thermal incinerator with waste heat boiler.
28
-------
WASTE GAS INLET
510°CT0570°C>» 6\C
1 670
°C TO 820 °C
f-
AUXILIARY FUME
BURNER INCINERATOR
FEED GAS PREHEATER
Figure 5.
Thermal incinerator with
feed gas preheater.17
Figure 6 shows the steps which must occur to provide the three
Ts (temperature, turbulence, and time) for achieving satisfactory
destruction of pollutants in a dilute stream. First, the supple-
mental fuel is burned by utilizing part of the oxygen contained
in the dilute fume stream to produce high temperature combustion
products (>1200°C). At the same time, the rest of the cold fume
bypasses the flame and is mixed with the combustion products in
the second step. The mixing gives a nearly uniform temperature
to all fumes flowing through the incinerator. This is done as
rapidly as possible without causing flame quenching so that
sufficient residence time can be provided at the required temper-
ature in the third step for the oxidation of pollutants contained
in the bypassed fume. Residence time is governed by the size of
the combustion chamber. Mixing (turbulence) is caused by baffles
or by tangential entry of either the dilute fume or the combus-
tion products from the incinerator burner.
Ross, R. D. Pollution Abatement: Incineration of Solvent-
Air Mixtures. Chemical Engineering Progress, 68(8):59-64,
-L " / ^ •
29
-------
SUPPLEMENTAL
FUEL — *"
FUEL COMBUSTION
MIXING OF FUME
AND HOT
COMBUSTION GASES
RETENTION OF FUMES
AT HIGH TEMPERATURE
FOR SUFFICIENT TIME
(OXIDATION)
CLEAN
~"~ EFFLUENT
CONDENSER
OFF-GAS-
DILUTE FUME)
Figure 6. Steps required for successful
incineration of dilute fumes.
5.1.2 Technical Evaluation for the Incinerator
5.1.2.1 Thermodynamic Considerations—
As can be seen from Table 8/ the existing incinerators used by
the phthalic anhydride industry for control of condenser off-gas
have control efficiencies on organics (including PAN, MAN,
benzoic acid and naphthoquinone) in the range of 80% to 96.5%.
A 99% reduction of PAN and MAN was not reported. An attempt was
therefore made to determine whether there is a thermodynamic
limitation on the combustion efficiency of PAN and MAN for
reaching the 99% reduction.
The oxidation of these two compounds to form the final combustion
products of H20 and CO2 can be expressed as follows:
8CO2 + 2H20
PAN
C4H2O3
MAN
30
4CO
H,O
The first step in determining the equilibrium conversion effi-
ciency is to calculate the equilibrium constants for the above
reactions according to the following equation:
= e-AF/RT
(1)
where AF =
F.
vi
KP
:
. v.F. products - . v.F. reactants = change in free
1 J_ J- Ill
energy due to reaction
free energy of formation for component i
stoichiometric coefficient of component i
equilibrium constant at constant pressure
30
-------
To obtain the equilibrium constants as functions of temperature,
the free energies of formation for H2O, C02, PAN, and MAN in the
temperature range of interest are needed (O2 has a free energy of
formation of zero).
The free energy data for CO2 and H2O are available for a wide
range of temperatures. Unfortunately, similar data for PAN and
MAN at incineration temperatures are not available to MRC. How-
ever, the available data for some hydrocarbons and oxygenated
hydrocarbons indicate that the free energy of formation at tem-
peratures around 1,OOOK ranges from -20 to 40 kcal/mole.18 The
data for H20 and C02 at the same temperature are -46.0 and 94.6
kcal/mole, respectively.18 For a worst case estimate, assume
that the free energy of formation for PAN and MAN is -20 kcal/
mole. The resulting Kp's, calculated from Equation 1 for both
PAN and MAN, are infinity.
The above calculation shows that there is no theoretical limita-
tion on the conversion efficiencies for PAN and MAN in the in-
cinerator. One-hundred percent control efficiency for organic
removal can be achieved under equilibrium conditions. The re-
latively low control efficiencies in the existing incinerators
are therefore due to kinetic limitations.
5.1.2.2 Kinetic Considerations—
Oxidation involves bimolecular reactions between the combustible
compound and the oxidizer. The rate at which oxidation proceeds,
r, can be approximately represented by the overall expression:
r = k(c02)(Ccombustible) (2)
where k is the rate constant and C is the concentration. The
rate constant, k, is an exponential function of temperature in
the following form:
. , -E/RT
k = A e (3)
With the activation energy, E, typically in the range of 10 to
60 kcal/mole for this overall rate constant.19 For an activation
energy of 30 kcal/mole, the reaction rate will double for an in-
crease of temperature from 700°C to 750°C.
8JANAF Thermochemical Tables, Second Edition. NSRDS-NBS 37,
U.S. Department of Commerce, National Bureau of Standards,
Washington, D. C., June 1971. 1,141 pp.
l9Rolke, R. W., et al. Afterburner Systems Study. EPA R2-72-062,
U.S. Environmental Protection Agency, Research Triangle Park,
North Carolina, 1972. 336 pp.
31
-------
In the condenser off-gas, the concentrations of combustible
materials (including PAN, MAN, benzoic acid, naphthoquinone, and
CO) are very small as compared to that of O2. Therefore, the con
centration of 02 in the post-flame area in the incinerator is al-
most constant despite the consumption due to the oxidation of
pollutants. Equation 2 can thus be rewritten as:
dX
= k' CAO (1 - XA>
where CAQ = the inlet concentration of pollutant A
XA = the fractional conversion
Integration of Equation 4 over time, t, gives:
-In (1 - XA) = k't = A'e~E/RT t (5)
This equation shows the fractional conversion as a function of
residence' time, t, and temperature, T. It should be noted that
in deriving Equation 5, it is assumed that perfect mixing (uni-
form temperature) in the combustion chamber is attained. Figure
7 schematically indicates the general effects of temperature and
residence time on the fractional conversion of combustibles in
an incinerator.
In actual incineration operation, the mixing time for the gas
mixture is about the same order as, and sometimes even greater
than, the reaction time, depending on the design. To estimate
the temperature increase for obtaining the desired control effi-
ciency, the activation energies for complete oxidation of PAN
and MAN are needed. Also the effect of mixing on the overall per-
formance of a particular incinerator has to be quantified. Un-
fortunately, none of the above information is available. There-
fore the following approach was used to estimate the temperature
increase for achieving 99% removal of organics. In this approach,
the organic components in the off -gas were considered as a single
substance. The mixing effect and oxidation reaction were then
combined to obtain an "apparent" activation energy for later use
in estimating the temperature effect of the conversion efficiency.
This approach utilizes the information obtained from three sources.
An industry survey indicates that an incinerator (at plant 5 shown
in Table 8), operating at 700°C, attains a conversion efficiency
of 95%. 20 Two other information sources show that the same unit
was originally designed to operate at 760°C at a different gas
20Personal communication with L. B. Evans, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina
(data originally supplied by Monsanto Company), 12 June 1975.
32
-------
OJ
u>
INCREASING
RESIDENCE
TIME
1000 1200 1400
INCREASING TEMPERATURE
1600 1800 2000
Figure 7. Coupled effects of temperature and time on rate of pollutant oxidation.
-------
flow rate for a 96.5% conversion efficiency and that it is the
best unit presently available.
Based on the above information on different conversion efficiencies
at different temperatures, and considering the differences in re-
sidence time and in mixing efficiencies of the incinerator due to
a change in flow rate, it was estimated (by using Equation 5) that
the "apparent" activation energy for the oxidation of the mixture
of organic material is 6 kcal/mole. With this apparent activation
energy, the temperature at which 99% conversion can be obtained
was calculated to be 860°C.
An increase in operating temperature from 700°C to 860°C will re-
quire more fuel, thus passing a larger volume of gas through the
incinerator and therefore decreasing the residence time. However/
it is determined that this decrease is less than 1% and is hence
negligible.
The same efficiency improvement can also be obtained by increasing
the mixed gas residence time without changing the operating tem-
perature. This increase of residence time can be calculated by
using Equation 5 and the conversion efficiency at 700°C (95%) .
From this calculation, it was estimated that a 55% extension in
the length of the mixed gas combustion chamber will give enough
residence time for 99% destruction of organic substances.
From past experiences in the incineration of other organic pollu-
tants, it was reported that the destruction of most hydrocarbons
occurs very rapidly at temperatures in excess of 590°C to 650°C.ld
Possible exceptions are methane, cellosolve, and benzene deriva-
tives, like toluene, which are stable molecules and require a
higher temperature (^760°C) for complete oxidation to occur in a
few tenths of a second.19 Since both PAN and MAN are oxygenated
hydrocarbons and are intermediates in the oxidation of hydrocar-
bons, it is expected that the destruction of these two compounds
would not be more difficult than that of toluene.
Experimental data on the conversion efficiency of toluene and two
other hydrocarbons are shown in Figure 8.19 The curves show that
the rate of hydrocarbon disappearance was first order with re-
spect to hydrocarbon concentrations and that with mixed gas re-
sidence time of 0.21 sec a 99.9% destruction of toluene is possibl0
at 766°C.
It is indicated in a systems study that temperatures of 760°C to
816 °C are sufficient to obtain nearly complete conversion of most
substances in 0.1 sec to 0.3 sec.19 However, incinerators are
usually designed to have total residence times of 0.3 to 1 sec.
Even with so much excessive residence time, some of the units stil^:
have poor conversion efficiency because time is required for the
mixing step and many designs fail to complete the mixing in the
34
-------
10
2
o
CD
O
O
cc
10
UJ
10
-1
0
INCINERATION TEMPERATURE 766 °C
0.05 0.1 0.15 0.2 0.25
TIME, SECONDS
Figure 8. Hydrocarbon oxidation rates
in absence of flame.19
35
-------
distance (time) available. The greatest variation among differ-
ent incinerator (afterburner) designs is in how well they achieve
the goal of raising all of the fume to the required temperature
for the required time.19 Most cases of poor performance are due
to nonuniform temperatures and flows which allow some of the com-
bustible pollutants to escape without treatment.
5.1.2.3 NQX Formation and CO Reduction—
There are no nitrogen-containing compounds in the condenser off-
gas. The formation of NOX in the incinerator results from the
fixation of nitrogen and oxygen which are present in the waste
gas. Substantially all of the NOX is formed in the high temper-
ature region (>1150°C) of the burner flame. At the temperature
of the main combustion chamber (700°C to 900°C), the overall
rate of reaction of nitrogen with oxygen is too slow for signi-
ficant formation of NOV.19
J\.
When the combustion chamber temperature is raised to achieve a
better control efficiency, the increased supplemental fuel in-
jection into the burner will result in a larger flame size and
thus a longer residence time in the high temperature region to
form NOX. It has been estimated that an increase of the in-
cinerator operating temperature from 650°C to 815°C will result
in a gain of NOX emissions from 18 ppm to 22 ppm.19 Based on
this information, a 160°C temperature rise as recommended in
Section 5.1.2.2 for achieving 99% removal efficiency will in-
crease the NOX emissions by 20%. Since a relatively low source
severity (S = 0.14) has been reported for NOX emissions from the
incinerator, this increase is considered as unimportant.1
On the other hand, incineration at the temperatures considered
can reduce the CO contained in the waste gas by greater than
99.9%.llf This is an advantage, especially for cases where CO
emissions require control due to state or local regulations.
5.1.2.4 Conclusion—
From the technical evaluation of the incinerator, it is concluded
that 99% removal efficiency can be obtained by operating the in-
cinerator at a mixed gas temperature of 860°C (as compared to
95% conversion at 700°C). The desired efficiency can also be
achieved by increasing the length of the main combustion chamber
in either of the existing incinerators by 55%. If high conver-
sion cannot be obtained by raising the operating temperature
(which can be accomplished by increasing fuel consumption), the
problem may be caused by poor mixing efficiency of the incinera-
tor. In this case, an investigation is needed to define changes
in design of the burning-mixing section for obtaining near-
uniform temperature distribution of the mixture.
36
-------
5.1.3 Heat Recovery
When an incinerator is used to control PAN, MAN, and other com-
bustible materials from the condenser off-gas, large amounts of
fuel are needed to heat the large volume of waste gas to the re-
quired temperature. Both the high cost and inadequate supply of
fuel become problems in operating the incinerator. As a result,
some form of heat recovery becomes necessary.
Two types of heat recovery equipment are considered in this
study. One is the waste heat boiler and the other is the feed-
effluent heat exchanger shown in Figures 4 and 5, respectively.
The thermal incinerator/waste heat boiler system has been used
in one plant (plant 5 as shown in Table 8) for several years,
and a similar unit was recently installed in another plant (plant
3). Satisfactory operations have been reported. In this system,
the feed gas to the incinerator is not preheated. This permits
maximum steam generation but substantially increases the fuel
requirement. Since most of the required heat is supplied by
supplemental fuel, it is relatively easy to sustain combustion.
The major drawback of the system is that steam consumers must be
found.
!n the thermal incinerator/feed-effluent heat exchanger system,
the condenser off-gas is preheated by the effluent gas from the
incinerator; this reduces the fuel requirement for heating the
feed gas to the incineration temperature. This system had been
installed in one plant which, as originally designed, incorpo-
rated an extensive amount of feed-effluent heat exchange (plant
3). However, due to safety and other operational problems, this
sYstem was recently replaced by the incinerator/waste heat boiler
system described above. Similar problems have also been en-
countered in the incineration system in plant 6.
ne safety problem comes from the development of hot spots which
exceed the autoignition temperature of PAN (approximately 455°C).21
ue to the opening and closing of the valves used to isolate the
switch condensers, short-term high concentrations of PAN dust may
e present in the condenser off-gas. As PAN dust accumulates in
he heat exchanger, it melts, enters the tubes, and is vaporized.
the tube has a hot spot which is above the autoignition tem-
perature with sufficient PAN present, ignition will occur. There-
ore, ignitions are probably common in this type of system, but
ue to the small quantity of combustible material and relatively
igh velocity, ignition is a transient condition. However, when
n ignition and a sufficient puff of PAN dust occur simultaneous-
y/ an explosion occurs.21 If the operating temperature of the
2 i
Moores, c. W. Control PA Emissions. Hydrocarbon Processing,
54(10):100-103, 1975.
37
-------
incinerator is raised to achieve a better control efficiency
(99%) , the possibility of explosion is even greater.
Another problem associated with this system is that preheating of
the incinerator feed stream will reduce the amount of burner flame
present and thus reduce the radicals generated by the flame for
effective destruction of combustible pollutants. Because of the
safety problem and the lower conversion efficiency mentioned
above, the feed-effluent heat exchanger is not recommended for
use in connection with an incinerator. Therefore, further econo-
mic evaluation was not made for this system.
5.2 CATALYTIC INCINERATOR/FEED GAS PREHEATING
5.2.1 System Description
A catalytic incinerator is an alternative to a thermal inciner-
ator as a means of oxidizing gaseous hydrocarbons (including oxy-
genated hydrocarbons) to carbon dioxide and water. Contact of a
waste stream with a catalyst bed allows the oxidation reaction to
occur rapidly at a lower temperature than that required in a ther-
mal incinerator. This lower temperature will result in savings
in fuel consumption.
Five steps are involved in the solid-catalyzed vapor-phase re-
action:
• Diffusion of the reactants through the stagnant fluid
around the surface of the catalyst
• Adsorption of the reactants on the catalyst surface
• Reaction of the adsorbed reactants to form products
• Desorption of the products from the catalytic surface
• Diffusion of the products through the pores and sur-
face film to the bulk vapor phase outside the
catalyst.
The various noble metals such as platinum, palladium, rhodium,
etc., in varying concentrations cause different reaction rates
for each specific hydrocarbon. However, in air pollution con-
trol, it is not practical to undertake a research program to
develop a specific catalyst for each problem. Therefore, com-
mercial catalyst manufacturers have attempted to make available
a universal catalyst which is effective in oxidizing the entire
range of organic materials over an extended period of exposure
time with minimum maintenance and replacement.
A catalytic incineration system, developed by a European phthalic
anhydride manufacturer (Rhone-Poulenc Industries) for use in con-
nection with its new manufacturing process (CHAUNY process), is
shown in Figure 9. This system incorporates a feed gas preheat-
ing device and is said to be self-sufficient in energy balance.
38
-------
HOT GAS
(FOR STARTUP ONLY)
CATALYST LAYER
POLLUTED GAS FROM
SWITCH CONDENSERS
65°C
183°C
X
3m5^^&;^^-^^^
369°C
BYPASS
VALVE
Figure 9.
Rhone-Poulenc catalytic afterburner
for phthalic anhydride off-gases.
39
-------
The calories necessary to raise the gas temperature to the cor-
rect reaction level are provided by means of the gas-gas heat
exchanger and by the exothermic oxidation itself. The catalyst
was developed specifically to treat the off-gas from the new manu-
facturing process.
Although gas preheating is usually considered as a means for fuel
savings, it becomes a necessity when treating the phthalic anhy-
dride condenser off-gas. This is due to the occasional presence
of the PAN dust which, without preheating to vaporize, will pos-
sibly cause burning or explosion in the catalytic incinerator.
This becomes possible when hot spots are developed in the cata-
lyst bed which have a temperature in excess of the autoignition
temperature of PAN (455°C).
5.2.2 Technical Evaluation for the Incinerator
Catalytic incinerators were previously used in some naphthalene-
based phthalic anhydride plants which are no longer in operation.
Experience with these units was not satisfactory. Conversion
efficiencies as low as 40% to 60% were reported.22 One other
problem was associated with this type of unit: moderate catalyst
life with possible danger of catalyst fouling and poisoning.
Possible polymerization of PAN and MAN on the catalyst surface
can produce a polymer coating that covers the active sites of
the catalyst.
Past experience in the use of catalytic units for other sources
also indicates that while conversions up to 90% to 95% can be
attained with reasonable catalyst volumes, the catalyst volume
required for very high conversion (e.g., >98%) generally makes
these units uneconomical, or the unit must be operated at tem-
peratures close to that required for a thermal incinerator.19
As a result of this technical evaluation, catalytic incineration
utilizing commercial catalysts is not recommended for use in
treating the off-gas from conventional phthalic anhydride manu-
facturing processes. The use of the catalytic incineration unit
developed by Rhone-Poulenc Industries (shown in Figure 9) in con-
nection with the new and modified manufacturing process (CHAUNY
process) is discussed in Section 5.6.
22Fawcett, R. L. Air Pollution Potential of Phthalic Anhydride
Manufacture. Journal of the Air Pollution Control Association/
20(7):461-465, 1970.
40
-------
5.3 WET SCRUBBER/WASTE DISPOSAL
5.3.1 System Description
Scrubbers can be used to remove the vapor-phase pollutants con-
tained in the waste gas by absorption and to remove the PAN dust
which may also be present in the waste gas by collision. Because
the scrubber can operate at a temperature lower than that of the
waste gas, condensation of part of the PAN and MAN into particles
will occur in the scrubber. This will further enhance the re-
moval efficiency of PAN and MAN.
A schematic diagram for the wet scrubber system is presented in
Figure 10. Note that waste disposal is included in the system.
This is necessary because wet scrubbing produces a secondary
pollution problem caused by the discharge of the purge liquor.
If this liquid discharge is not properly handled, an air pollu-
tion problem is just traded for a water pollution problem.
CLEAN VENT GAS
TO STACK
SWITCH CONDENSER
OFF-GAS
AftAA A
A7Y7VAA
LlJ
CD
CO
o
to
FRESH WATER MAKEUP
POSSIBLE EMISSIONS
PURGE LIQUOR
DISPOSAL
Figure 10. Wet scrubber/waste disposal system.
Phthalic and maleic anhydrides are hydrolyzed to the correspond-
ing acids upon absorption into the scrubbing liquor. To concen-
trate the purge liquor for easier disposal (by minimizing the
volume of the effluent stream), a certain degree of scrubbing
liquor recycling is necessary. The means of purge liquor dis-
posal considered in this study included incineration, MAN re-
covery, and biodegradation. These are described in the sub-
sections following the technical evaluation of the wet scrubber,
41
-------
A new wet scrubber design for treating the switch condenser off-
gas is illustrated in Figure II.23 The unit consists of two
stages, each of which contains a conventional fluid-bed packing
on a supporting grid. The two stages are separated by a cone-
shaped deflector plate and collection tray. The unit is designed
to operate between 35°C and 40°C with an organic removal effi-
ciency of 98% to 99%. This high efficiency is possible because
most of the organics are solids at the scrubber operation con-
ditions.
VENT GAS
MIST ELIMINATOR
CONDENSER
TAIL GAS
RECYCLE _
2nd STAGE
PURGE ,
i i
k.
\^.
^
*-w*
STAGE
/ \
1st
STAGE
V
-« FRESH V
is* STAGE
^ RECYCLE
^
1st STAGE PURGE
TO INCINERATOR
Figure 11. Schematic diagram of scrubber for
phthalic anhydride plant.23
23Ferrari, D. C., and C. G. Bertram. Method and Apparatus for
Removal of Organic from Chemical Waste Gases. U.S. Patent
3,624,984 (to the Badger Company), December 7, 1971.
42
-------
The condenser tail gas enters the first stage at approximately
66°C. The scrubbing liquor for this stage consists of the purge
from the second stage and the first stage recycle stream. The
liquor is a slurry, containing 10% to 12% by weight of organic
solids. The concentration of organic compounds, both dissolved
and in the form of slurry solids, ranges from 10% to 50% by
weight, depending on the composition of the gas being treated.
A portion of the recycle stream is continuously purged and sent
to a waste disposal process.
In the second stage, the gas is scrubbed with a dilute solution
(0.5% to 3% by weight) of organic pollutants (chiefly maleic acid)
Fresh water is added to the second stage at a rate sufficient to
replace the water removed from the scrubber in the two exit
streams. The makeup water combines with the second stage re-
cycle stream to form the second stage scrubbing liquor. The
scrubbed gas stream passes through a mist eliminator and is then
vented to the atmosphere at 38°C.
5-3.2 Technical Evaluation for the Scrubber
5-3.2.1 Theoretical Considerations—
In order to obtain an overall removal efficiency of 99% for PAN
and MAN, the wet scrubber should be able to remove these pollu-
tants with an efficiency greater than 99% because of possible
emissions from the waste disposal area which should be accounted
for as emissions from the control system.
In this particular case, the removal of PAN and MAN in the scrub-
ber is achieved by two distinct mechanisms. The first is conden-
sation into solid particles and subsequent removal of the parti-
cles by scrubbing. The second is absorption of the vapors by the
scrubbing liquid. Because the dew points of PAN and MAN in the
waste gas are below the corresponding melting points, these two
Pollutants will be condensed into solid particles upon cooling
in the scrubber.
The waste gas (switch condenser off-gas) is hot (66°C) and not
saturated with water. Evaporation of water in the scrubbing
liquor provides a cooling effect. This, together with the addi-
tion of sufficient cooling (e.g., room temperature), fresh makeup
water, can maintain a temperature around 38°C in the scrubber.
TO evaluate the theoretical removal efficiency of the scrubber,
it was necessary to consider the equilibrium vapor pressure of
both PAN and MAN over their solid particles and over their aque-
ous solutions.
The equilibrium vapor pressures of pure PAN and MAN as functions
°f temperature are shown in Figure 12. At 38°C, the vapor pres-
sure of PAN is 0.0027 mm Hg. When this is compared with the
Partial pressure of PAN in the waste gases from the o_-xylene
based and from the naphthalene-based plants (0.141 mm Hg and
43
-------
en
u_T
CO
CO
LU
ex.
Q_
O
Q_
MAN
30 40 50 60 70 80 90 100
, °C
0.0001
Figure 12. Vapor pressure vs. temperature.
44
-------
0.261 mm Hg, respectively, as shown in Tables 6 and 7), one would
expect that, given enough time, more than 98% of PAN will condense
into solid particles at the scrubber temperature. From Figure 12
the equilibrium vapor pressure of MAN at the same temperature
(38°C) is 0.6 mm Hg, which is higher than the partial pressures
of MAN in both waste gas streams. Therefore, condensation of MAN
will not occur in the scrubber.
Prom the above analysis, it is obvious that the removal of PAN
will be achieved mainly by the first mechanism (particle collec-
tion) and that MAN removal will be achieved solely by the second
mechanism (absorption). In reality, condensation and absorption
of PAN will occur simultaneously. Therefore, the solubilities
of both PAN and MAN in water are the next things to consider.
At the scrubbing temperature, MAN is soluble up to 40 weight per-
cent of maleic acid.^3 PAN is only slightly soluble in water at
this temperature (1 wt percent in terms of phthalic acid).
However, with increasing MAN concentration in the aqueous solu-
tion, the solubility of PAN can increase to 3 wt percent.^
Therefore, in order to enhance the absorption of PAN, it would
be desirable to have a scrubber design that provides high concen-
tration of MAN in the first stage (when a multistage scrubber is
used).
An example was given in the Badger patent23 for the application
of the two-stage wet scrubber for treating the switch condenser
off-gas from a naphthalene based phthalic anhydride operation.
prom the material balance provided, the removal efficiency is
98.5% for MAN, and 98.4% for PAN.
Three PAN plants in the United States currently use wet scrubbers
for treating the waste gases from the switch condensers (see
Table 5). Plant 2 uses a proprietary cocurrent scrubber with a
l9-m3/min (5,000-gpm) recirculation of scrubbing liquid for clean-
ing about 3,400 mVmin (120,000 scfm) of off-gas. Makeup water
is sprayed into the scrubber system between the two demisters on
the scrubber vent gas. In plant 4, a similar quantity of waste
gas is processed in two parallel water scrubbers, each consisting
°f a venturi contactor (2.3-m3/min water rate) followed by a
Packed column, countercurrent mist eliminator with a 7.6-m /mm
water recirculation rate. The scrubber purge liquors in both
Plants 2 and 4 are incinerated in liquid waste incinerators.
At plant 4, a MAN recovery process is also used to treat part of
the purge liquor.23 Plant 8 utilizes two venturi scrubbers, a
tangential cyclonic separator, and a multinozzle spray tower.
Recirculation is practiced in the secondary scrubber (spray tower)
2ltPersonal communication with R. A. Mount, Monsanto Company,
St. Louis, Missouri, 3 June 1976.
45
-------
at a 0.3-m3/min water rate. The wastewaters from the scrubbers
are collected for treatment (means unknown).
From the data obtained from a Russian article,25 the Henry's law
constant for MAN at the scrubber temperature was estimated to be
1.55 x 10"5. This constant defines the vapor content of a parti-
cular component in a gas mixture which is in equilibrium with a
dilute solution of the said component. This small value of
Henry1s_law constant, together with the high solubility of MAN in
water, indicates that even if the scrubbing liquid is recycled to
provide concentrated purge liquor (and hence reduce the volume of
the liquid effluent), the resulting MAN vapor in the gas phase is
still very low.
Supposing^the purge liquor has a concentration of 20% maleic acid,
under equilibrium conditions, the concentration of MAN in the gas
phase will be 6.99 x 10~5 volume percent. This concentration,
when compared with the volume percent of MAN contained in the
waste gases from o-xylene and naphthalene based plants, repre-
sents 99.9% and 99.2% reduction of MAN in the two gas streams,
respectively. In the presence of phthalic acid and either benzoic
acid or naphthoquinone in the solution, the vapor pressure of MAN
will fall below the value mentioned, giving a higher theoretical
removal efficiency in a one-stage fluid-bed scrubber.
5.3.2.2 Practical Considerations—
A maleic anhydride removal of "about 100%" from PAN plant waste
gas in a one-tray foam scrubber with overflow baffle has been
reported under pilot plant conditions.25 In another test of water
scrubber handling 835 m3/min (29,500 scfm) switch condenser off-
gas, it was also reported that the unit was capable of removing
in excess of 99% of all organic acids.22 However, in the latter
case, the scrubber water discharge had a much lower acid concen-
tration (1.7% to 2.5% total acidity as maleic acid).
In the two plants (2 and 4) where scrubbing liquid is recirculated
for producing concentrated purge liquor for incineration or for
MAN recovery, data on the scrubber feed composition are not avail-
able for an accurate calculation of the removal efficiency. Usin
the waste gas composition presented in Table 6 and the data on the
emissions from scrubber vent gas, it was, however, possible to
make a rough estimate. This estimate shows that the removal of
PAN and MAN in these two units ranges from 97.5% to 99.5%.
25Yevzel'man/ I. B. , and G. D. Kharlampovish. Neutralization of
Exhaust Gas from Phthalic Anhydride Production. Khimicheskaya
Promyschlennost (Moscow), 49 (6):379-380, 1973.
46
-------
With the high actual removal efficiencies mentioned above, it is
believed that with design modifications, the scrubber is able to
remove both PAN and MAN by more than 99%.
5-3.2.3 Conclusion—
Prom the theoretical considerations of the scrubber operation, it
is obvious that the following conditions should be met in design-
ing a scrubber for high efficiency removal of PAN and MAN.
• Since a large quantity of PAN becomes solid particles
after entering the scrubber (not before) and is re-
moved as such, the scrubber bed should be able to pro-
vide good mixing and at the same time be able to re-
move the slurry without causing clogging in the bed.
• To enhance the absorption of the uncondensed PAN, it
is necessary to operate a portion of the scrubber at
high MAN concentration.
TO meet the first condition, a fluid bed of packing as described
•tor the design illustrated in Figure 11 is recommended. Venturi
scrubbers such as those used in plants 4 and 8 are not particu-
larly effective in achieving this. The packing can be made up of
relatively large, but light, smooth-surfaced pieces such as
spheres or rings.
TO meet the second condition, a multistage operation would be
necessary (refer to Figure 11). The first stage can remove
°st of the PAN by condensation and by absorption. Further ab-
sorption of PAN and MAN can be achieved in the second fluid-bed
stage where the MAN concentration is well below that of the first
(0.5% to 3%). To ensure a greater than 99% removal, a
stage of packed bed is recommended.
Liquid Waste Incineration
he liquid purge discharged from the multistage wet scrubber con-
ains about 20 wt percent of maleic acid, 10% of phthalic acid,
na 1% of benzoic acid (based on scrubbing of the waste gas from
,. £~xylene based plant). This concentrated liquid waste can be
isposed of by using a thermal incinerator which is shown schemat-
ically in Figure 13.
j,e liquid purge stream is atomized immediately before entering
. combustion area of the incinerator by either a steam or an
^ir stream. Fuel gas (e.g., methane) and combustion air (25% in
c^Cess of the stoichiometric amount) are introduced into the in-
r^!!era^or in a conventional manner to provide an operating tem-
ire of 760°C to 870°C. Heat recovery can be achieved by
of a coil which can be used either to preheat the combus-
air (as shown in Figure 13) or to generate steam.
47
5.3.3
-------
TO
ATMOSPHERE
LIQUID PURGE
FROM SCRUBBER
COMBUSTION
AIR
FUEL GAS
ATOMIZING STEAM
OR AIR
Figure 13.
Schematic diagram of thermal incinerator
for scrubber purge liquor.23
The incinerator reportedly attains an efficiency of 99.9% conver-
sion of organic materials.23 This high combustion efficiency be-
comes possible due to the smaller size of this incinerator (as
compared with the direct thermal incinerator mentioned in Section
5.1) which permits better mixing in the combustion chamber. The
liquid waste (which contains solid particles) is first evaporated
and then burned in the main combustion chamber of the incinerator.
From the emission data available and the waste gas concentration
shown in Table 6, the estimated combustion efficiency of the in-
cinerator at plant 2 becomes 99.1% to 99.5%. In this incinerator
(which is used to incinerate the scrubber purge liquor) , the
organic content in the feed stream is about one-third of that
which can be obtained from the multistage scrubber mentioned be-
fore. With a higher organic content in the feed stream, a better
conversion efficiency can be expected.
48
-------
Supposing 99.9% conversion can be achieved in the incinerator (as
reported in one literature source23), a removal efficiency of
99.1% in the wet scrubber will give an overall control efficiency
of 99% for the control system.
5.3.4 Maleic Anhydride Recovery
5.3.4.1 Process Description—
The maleic anhydride recovery process is another alternative for
disposal of the scrubber waste liquor resulting from wet scrub-
bing of the condenser off-gas. The technologies of recovering
the maleic anhydride contained in the scrubber waste liquor have
been developed by UCB (Brussels, Belgium) and BASF (Ludwigshafen,
Germany) separately. As is shown schematically in Figure 14,26
the recovery process basically consists of three sections: con-
centration, dehydration, and distillation.
In the first section, the organic compounds from the scrubber
purge stream are separated from water in two steps. The first
step evaporates water at a closely controlled temperature to avoid
any crystallization of organics. In the second step, the remain-
ing water is flashed off in an evaporator, leaving a bottom stream
of liquid organics. The water removed in both stages, which con-
tains some residual organic material, is condensed and recycled
to the scrubber.
In the second section, the liquid organic stream is sent to a de-
hydration unit where the organic acids are thermally dehydrated
to the anhydrides. The maleic anhydride and dehydration water
are distilled and separated by a selective condensation process.
The water is recycled to the scrubber, while the crude maleic
anhydride is sent to the vacuum distillation column for purifica-
tion in the third section.
According to one process developer, the maleic anhydride which is
recovered by this process meets the norms of purity of the stand-
ard MAN obtained by catalytic oxidation of hydrocarbons.26 Such
a product can be substituted for standard MAN in all its outlets.
Impurities present in the scrubber purge liquor are separated by
the recovery process as anhydrous melt of phthalic anhydride,
citraconic anhydride, benzoic acid, and a little maleic anhydride.
It was indicated that this melt can be pumped and burned in the
same way as a normal heavy fuel.27
26Weyens, E. Recover Maleic Anhydride. Hydrocarbon Processing,
53(11):132-134, 1974.
27Personal communication with A. David, UCB S.A., Brussels,
Belgium, 21 May 1976.
49
-------
FROM PAN SCRUBBER
Ul
o
CONCENTRATION *+• DEHYDRATION
RESIDUES TO BURNER •*•
TO PAN SCRUBBER
MALEIC ANHYDRIDE TO STORAGE
Figure 14. Maleic anhydride recovery process.26
-------
When this recovery process is used, demineralized scrubbing water
is needed because alkali metal ions act as polymerization cata-
lysts for maleic anhydride and such polymerization may cause
explosions.28
A commercial plant employing this MAN recovery process (annual
capacity: 4,000 metric tons of MAN) has been in operation by
BASF in Germany since 1971.29 Another unit has been recently
installed at plant 4 (see Table 8).30 .
5.3.4.2 Technical Evaluation—
When the MAN recovery process is utilized for disposing of the
scrubber purge liquor, all the organic constituents in the liquid
feed to the process should be dissolved in water. Prom the tech-
nical evaluation of the wet scrubber, it is obvious that deposits
or crystals of phthalic acid will be present in the purge liquor
if a ^20% concentration of maleic acid is to be achieved in the
scrubber. Therefore, some kind of solid-liquid separation device
is necessary to remove the solids from the scrubber purge liquor.
In addition to PAN particles, the purge liquor may contain a small
percentage of benzoic acid particles due to the low solubility of
benzoic acid. Table 9 shows an estimate of the flow rate and com-
position of the purge liquor from a multistage wet scrubber
(described in Sections 5.3.1 and 5.3.2), when treating the waste
gas from a representative o-xylene based plant.
TABLE 9. COMPOSITION OF THE PURGE LIQUOR FROM SCRUBBING OF THE
WASTE GAS FROM A REPRESENTATIVE O-XYLENE BASED PLANT
(kg/hr)
Component
Phthalic acid
Maleic acid
Benzoic acid
Water
TOTAL
In solution
55.4
369.7
1.2
1,422.2
1,848.5
In particulate form
130.3
19.0
149.3
Total
185.7
369.7
20.2
1,422.2
1,997.8
28Wirth, F. Recover MA from PA Scrubber Water. Hydrocarbon
Processing, 54 (8):107-108, 1975.
29Personal communication with I. V. Heese, BASF A.G.,
Ludwigshafen, Germany, 11 June 1976.
°Personal communication with R. T. Wedell, Koppers Company, Inc.
Chicago, Illinois, 5 October 1976.
51
-------
It was also estimated that a reciprocating conveyor centrifuge
filter with a 10-in.-diameter basket will be able to remove the
crystals and deposits from the purge liquor continuously. This
device was included in the economic evaluation of the wet
scrubber systems utilizing the recovery process.
From Tables 6 and 7, one can see that the MAN content in the waste
gas from a naphthalene based plant is only 16% of that from an
o-xylene based plant. It is therefore also recommended that the
MAN recovery process not be applied to naphthalene based plants.
A disadvantage in use of the wet scrubber/MAN recovery system
includes the generation of an anhydrous melt of organics from the
recovery process and the solids separated from the purge liquor
by filtration. The total waste product from these two sources
amounts to about 180 kg/hr. It is suggested by a process
developer that the anhydrous melt of organics, which contains
phthalic anhydride, citraconic anhydride, benzoic acid, and a
small amount of maleic anhydride, can be pumped and burned as a
heavy fuel.27 Presumably the solids separated from the purge
liquor can also be burned in a conventional burner.
5.3.5 Biological Wastewater Treatment
5.3.5.1 Process Description—
Biological treatment systems are living systems which rely on
biocultures to break down waste organics and remove them from
solution. In order to apply biological methods to wastewater
treatment, the organic wastes should be biodegradable. Figure
15 shows a system that contains an equalization tank, a neutrali-
zation and nutrient addition tank, an aeration tank, a settling
tank, and a sludge treatment process. This system is designed
to meet the 1977 "Best Practical Control Technology Currently
Available" (BPCTCA) wastewater treatment requirements. To meet
the 1983 "Best Available Technology Economically Achievable
(BATEA) treatment requirements, dual-media filters and granular
carbon columns (shown in Figure 1631) should be added to the
BPCTCA system.
31Development Document for Interim Final Effluent Limitations
Guidelines and New Source Performance Standards for the
Significant Organic Products Segment of the Organic Chemical
Manufacturing Point Source Category. EPA-440/1-72/045, U.S.
Environmental Protection Agency, Washington, D.C., 1975.
361 pp.
52
-------
NaOH
SCRUBBER
PURGE LIQUOR
EQUALIZATION
NUTRIENTS
NEUTRALIZATION
LTI
CO
SUPERNATANT
OR FILTRATE
AERATION
TANK
RETURN SLUDGE
WASTE
.SLUDGE
SLUDGE
PROCESSING
EFFLUENT
DRIED SLUDGE
OR ASH
FOR ULTIMATE
DISPOSAL
Figure 15,
An activated-sludge-based biological wastewater
treatment system for 1977 BPCTCA requirements.
-------
BIOLOGICAL TREATMENT
**"»
BACK WASH
HOLDING TANK
PLANT EFFLUENT
01
FILTER INLET
WELL
REGENERATED CARBON
STORAGE TANK
FILTER WATER
HOLD ING TANK
DUAL MED IA
GRAVITY FILTERS
CARBON
COLUMN
FEED PUMPS
BACK WASH
PUMPS
PLANT
EFFLUENT
PULSED BED
CARBON
COLUMN
TRANSFER
TANK
DRY ING TANK
AIR BLOWER
DRY STORAGE TANK
SCREW FEEDER
REGENERATION
FURNACE
VIRGIN
CARBON
STORAGE
Figure 16. Dual-media filters and granular carbon columns to be
added to the BPCTCA system to form a BATEA system.31
-------
The removal of organic substances contained in the wastewater
occurs in the activated sludge process which consists of an
aeration tank and a settling tank. In this process, organic sub-
stances are removed by microorganisms either through oxidation
to C02, H20 and other derivatives, or through conversion into a
settleable form that can be removed by sedimentation. The pro-
duction of C02 and H20 is called the respiration stage, while
conversion of organic contaminants to new bacterial cells which
can be settled out by gravity is referred to as synthesis.
Both respiration and synthesis occur in the aeration tank where
oxygen is provided to oxidize the organic material and satisfy
the biochemical oxygen demand (BOD) of the wastewater The
microbial cells formed in the aeration tank are then separated
trom the water in the settling tank. Part of the settled acti-
vated sludge is recirculated back to the aeration tank to provide
sufficient microbial cells for effective destruction of organic
wastes.
Although both phthalic and maleic acids contained in the scrubber
Purge liquor are readily biodegradable, neutralization of the
acids is necessary to provide the proper environment (PH) for
microbial growth. Furthermore, adequate amounts of nutrients
such as nitrogen and phosphorus must be added to the water to
maintain a ratio of BOD:N:P of 100:5:1 for the synthesis of
bacterial cells.
5.3.5.2 Technical Evaluation—
Biological wastewater treatment systems such as the one described
above can only be applied to waste streams containing dilute
organic contaminants. An analysis of such a stream from a con-
denser off -gas wet scrubber has been given in an EPA report 31
The 5-day Biochemical Oxygen Demand (BOD5), Chemical Oxygen De-
mand (COD), and Total Organic Carbon (TOC) of that stream are
snown in Table 10.
TABLE 10. ANALYSIS OF A WASTE STREAM FROM CONDENSER
OFF-GAS WET SCRUBBER
BOD* COD TOP
Concentration, mg/1 215 1,080 34
Raw waste loadings,
g/kg of PAN product 0.128 0.642 0.02
55
-------
The maximum strength of the wastewater that can be properly han-
dled by the biological treatment system is about 2,000 mg/liter
of COD. To obtain this waste strength, enough water should be
used in the wet scrubber to give a flow rate of 400,000 liters
per hour for the scrubber purge liquor. This flow rate is equi-
valent to 1.1 x 1014 m3/d (2.5 million gallons per day). To treat
this stream, a municipal size treatment plant is needed, and a
large piece of land is required.
In this case, since concentrated scrubber purge liquor is not
needed, the recirculation requirement of the wet scrubber is re-
duced. In addition, sodium hydroxide can be added to the scrub-
bing liquid for neutralizing phthalic and maleic acids, hence pre-
venting the formation of phthalic acid solid deposits and subse-
quent clogging of the scrubber bed. Therefore, the scrubber of
the special design described in Section 5.3.1 may not be needed.
The neutralization step is costly, however. Based on the mass
balance of the system and assuming a cost of $150/ton of NaOH,
the annual cost of chemicals for neurtalization alone is $422,000.
This cost is equivalent to $7.7 per ton of phthalic anhydride pro-
duced and is comparable with the operating costs of other alter-
native control systems given later in Table 13. Also, the neutra-
lization and subsequent biodegradion will add a large quantity of
dissolved solids (inorganic salts) to the water discharge.
Because of the large volume of wastewater to be treated, the
large land requirement for the facilities, the extremely high cost
of neutralization, and addition of a large amount of dissolved
solids in the water discharge, the biological treatment of scrub-
ber purge liquor was not given further consideration.
5.4 CARBON ADSORBER/WASTE INCINERATION
5.4.1 System Description
Due to its nonpolar surface, activated carbon has the ability to
adsorb organic and some inorganic materials in preference to
water vapor. The amount of materials adsorbed is dependent upon
the physical and chemical characteristics of the specific com-
pounds. The equipment used for adsorption processes consists of
vessels which hold a bed of the adsorbent and the auxiliary equip'
ment for regeneration of the adsorbent and recovery or disposal or
the adsorbate.
A schematic diagram for the carbon adsorption/incineration system
is shown in Figure 17. This system can be separated into two
areas: one for adsorption-regeneration and the other for disposaJ-
(i e , incineration) of organics (adsorbate) which are desorbed
upon regneration of the carbon bed. There are three vessels<
(carbon beds) in the adsorption-regeneration area, with two in
operation while the third is being regenerated.
56
-------
7 —
WER
HX-<
i
i
[-I*--
1
[*h-
I
:STE'AM
SUPPLY
1
PURE STRIPPED GAS
i
^±±C ADSORBER 11 )-tXj-*-
-M-
f"
^HX^T ADSORBER #2 VtX}-»
,_N/1 ..
HXJ-
*^
-------
3 50
3 40
LLJ
a:
a 30
20
10
o
g
4 6
TIME, HOURS
10
12
Figure 18. Adsorption efficiency of a carbon bed
for a single organic compound.
the breakpoint is reached, the exit vapor consists largely of the
more volatile material. This will continue until the second
breakpoint is reached, after which benzoic acid will start to
desorb from the carbon surface. This operating characteristic
is shown in Figure 19. To obtain effective control efficiency
in this case, the adsorption cycle should be stopped at the first
breakpoint as determined by the detection of vapors in the dis-
charge.
Regeneration is accomplished by passing low pressure steam through
the carbon bed. Since the boiling points of the organic contami-
nants in the condenser off-gas range from 198°C to 295°C, the tem-
perature of the steam should be greater than 300°C in order to
drive off the organics from the carbon surface in the regeneration
cycle.
The stream leaving the regenerating carbon bed has a high concen-
tration of organics. This stream is directed to a thermal incin-
erator for destruction of organic contaminants, including PAN and
MAN. The ability of the activated carbon to increase the concen-
tration of combustible materials, so as to decrease the fuel re-
quirement for operating the incinerator, permits significant
energy consumption savings.
58
-------
468
TIME, HOURS
10
12
Figure 19. Adsorption efficiency of a carbon bed
for a three-component mixture.
The incinerator vent gas is a heated gas stream with the potential
of being used for carbon bed regeneration. This usage was not
selected because the temperature of this gas stream is above the
ignition temperature of the organic compounds that are to be de-
sorbed. This high temperature, together with the presence of
oxygen in the gas, imposes a risk of explosion of the carbon bed.
Furthermore, the oxygen contained in the incinerator vent gas
might undergo reaction with PAN and MAN in the presence of acti-
vated carbon under high temperature to form a polymer-type mate-
rial that is difficult to remove from the carbon surface. There-
fore, steam was chosen as the regenerating gas for reactivation
of the carbon.
The service duration of a carbon bed between regenerations can be
determined by the following relationship:32
t =
6.43 (106) SW
eQMC
(6)
32Kohl, A. L., and F. C. Riesenfeld. Gas Purification. McGraw-
Hill Book Company, Inc., New York, New York, 1960. Page 428.
59
-------
where t = service life, hr
S = retentivity, fractional
W = weight of carbon, Ib
e = adsorption efficiency, fractional
Q = quantity of air processed by adsorption equipment, cfm
M = average molecular weight of contaminants
C = concentration of contaminants, ppm
The above formula can also be used to calculate the weight of
carbon needed (W) for a desired service life. The retentivity
of the carbon varies with the organic materials to be removed
and the type of carbon to be used. If the organic vapor or mix-
ture has not been previously handled by the equipment manufacturer
or the supplier of the activated carbon, laboratory tests of the
material must be run before design conditions can be determined.
5.4.2 Technical Evaluation
From the mass balance of this control system, which is shown in
Appendix A, the concentration of organics in the gas stream leav-
ing the regenerating carbon bed is 23 times the concentration in
the condenser off-gas. The volume of gas to be incinerated is
reduced by 15 times as compared to direct thermal incineration.
With the supplemental air to be added to the incinerator for
supplying oxygen to facilitate thermal oxidation, the total
volume to be heated to the incineration temperature becomes one-
tenth that of the condenser off-gas. This reduction in the
volume of gas to be incinerated and the high concentration of
organics in the gas provide good conditions for effective destruc-
tion of pollutants in the incinerator unit.
Supposing this incinerator can achieve a 99.9% destruction of PAN
and MAN, the removal efficiency of the carbon adsorption bed
should be equal to or greater than 99.1% in order to have an over-
all control efficiency of 99% for the system. From the theory of
carbon adsorption described in the process description, this pro-
cess can achieve close to 100% removal of organics. With the re-
tentivity of the carbon system unknown, it is difficult to deter-
mine whether a reasonable service life of the carbon bed can be
obtained using reasonable amounts of activated carbon (refer to
Equation 6 for their relationship).
Judging from the physical properties of PAN and MAN, it seems
reasonable to believe that their retentivity on a carbon bed
would be relatively high and would give favorable conditions for
the carbon system. This speculation is based on the relatively
high molecular weights and high boiling points of the organic
components.
Since steam is to be used to regenerate the carbon bed, hydrol-
ysis of PAN and MAN to corresponding acids and the high tempera-
ture needed will impose corrosion problems on the carbon handling
60
-------
equipment. Stainless steel vessels, stainless pipes, and corro-
sion resistant valves are necessary in constructing the carbon
system. Another problem is the existence of particulate PAN which
appears occasionally in the condenser off-gas. Fortunately, these
PAN solid particles, to be retained by the carbon bed, can be
vaporized during the regeneration cycle, eliminating the clogging
of the carbon bed that is usually encountered when treating gases
containing particulate materials.
5.4.3 Conclusion
From the above discussion, an activated carbon/incineration system
seems to be an attractive alternative for controlling the conden-
ser off-gas. Since this type of waste gas has not been handled by
any equipment manufacturer or carbon supplier, laboratory tests
are recommended to confirm the operating parameters such as reten-
tivity and steam rate that have been assumed and used in the tech-
nical evaluation (see Appendix A). The possibility of polymer-
like material forming on the carbon surface when high temperature
steam is used for carbon bed regeneration should also be investi-
gated.
5-5 CONDENSATION/FIBER BED FILTRATION
After examining the equilibrium vapor pressure of PAN and MAN at
different temperatures, it was found that a substantial amount of
PAN and MAN can be condensed to form solid particles if the con-
denser off-gas is cooled (by refrigeration) to a lower temperature
(e.g., the cooling of the off-gas from an o-xylene based process
from 66°C to 20°C will result in 99.82% condensation of PAN). A
control scheme was therefore considered which consists of cooling
the off-gas and subsequently removing the water-soluble particles
with a fiber filter bed. A schematic diagram of this control sys-
tem is shown in Figure 20.
VENT
MAKEUP WATER
SWITCH CONDENSED
OFR5AS *
MAN RECOVERY
OR INCINERATION
Figure 20.
Schematic diagram for the condensation/
filtration system.
61
-------
A vendor of a fiber filter bed system was contacted to learn
whether the system can remove PAN and MAN from off-gas. This
fiber filter system has the advantage that it is almost an abso-
lute filter for particles even in the submicron size range. The
water-soluble particles collected can be washed off the filter
bed by irrigating or spraying with water. The liquid stream can
then be directed to the MAN recovery process or to a liquid waste
incinerator.
The technical evaluation indicates that water condensates will be
formed in the cooling section, causing the formation of slurry,
in addition to the formation of condensed particles. Also, PAN
and MAN will condense and build up on the surface of the heat
exchanger. The surface condensation and slurry formation in the
cooling section will decrease the cooling efficiency and block
the flow path.
In addition, the cooling will probably not produce particles
small enough to require a fiber filter bed.33 Therefore, this
control scheme was not given further consideration in this task
effort.
Another alternative control scheme consists of cooling the waste
gas to its water dew point (to prevent the formation of slurry),
then scrubbing with a venturi scrubber to remove the particles
and the uncondensed vapors. In this case, the formation of
slurry can be prevented; however, the surface condensation of PAN
and MAN still remains.
5.6 CHAUNY PROCESS (PROCESS MODIFICATION)
5.6.1 Process Description
A new process for the production of phthalic anhydride from
o-xylene has been developed by Rhone-Poulenc Industries of France-
The flow diagram for this new process (shown in Figure 213I+) is
very similar to that for the BASF process (Figure 2), with the
following differences:34/35
• No sulfur or sulfur dioxide is required in the reactor
feed to maintain catalyst activity. Hence, this
source of pollution is eliminated.
33Personal communication with E. D. Kennedy, Monsanto Enviro-
Chem, Inc., St. Louis, Missouri, 8 June 1976.
3l*Phthalic Anhydride Process, CHAUNY. Rhone-Poulenc S.A. ,
Neuilly sur Seine, France, 1976. 17 pp.
35Personal communication with J. C. Zimmer, Rhone-Poulenc S.A.,
Neuilly sur Seine, France, 13 May 1976.
62
-------
OXIDATION
CONDENSATION | [POST. COMBUSTION
[DISTILLATION
COLD OIL
HOT OIL
U)
— # STEAM 6b
LIGHT
PRODUCTS
TO COMBUSTION
OR MAN RECOVERY
STEAM 6b
PHTHALIC
ANHYDRIDE
SOLID
RESIDUES
Figure 21. CHAUNY process for PAN manufacture.34
-------
• The switch condenser off-gas is controlled by means
of a catalytic incinerator which was developed for
this particular use. The control system is energet-
ically self-supported due to low temperature combus-
tion and a lower air-to-o-xylene feed ratio as com-
pared with the BASF process (22/1 versus 25/1).
• The 22/1 air-to-o-xylene ratio enables a high steam
recovery efficiency and its conversion to mechanical
energy for air compression by using steam turbine,
instead of an electric motor.
• The liquid waste stream from the distillation section
is further processed to produce solid pellets, which
are disposed of by burning.
Using this process, a 15 x 103 metric tons/yr plant has been in
operation since 1971 in Aisne, France. The process has also been
licensed to Resins, Incorporated (Philippines) and to Reposa
(Spain) . r
5.6.2 Technical Evaluation
It is claimed by the process developer that the catalytic incin-
erator incorporated in the process enables a complete oxidation
into CO2 and H2O of all the polluting substances contained in the
waste gas.^f36 The information supplied by the developer indi-
cates that in Europe, where there is no specific regulation on
phthalic anhydride plant emissions, the adopted working conditions
of the incinerator have given the following results in the CHAUNY
process : d 5
Polluting substance Percent conversion
CO 100
MAN 97
PAN 100
Miscellaneous 100
A typical analysis of the condenser off-gas from the CHAUNY pro-
cess is shown in Table 11.36 The mass flow rate Qf thig offl s
was not available to MRC for determing whether the 97% reduction
of MAN in the CHAUNY process can result in an emission of MAN
which is equal to or smaller than that from a conventional process
of the same capacity with 99% control efficiency. However, it was
indicated that depending upon the target and the local regulations/
it is possible to select the type and the quantity of catalyst
will provide practically a 100% conversion on all the polluting
components.
36Personal communication with J. C. Zimmer, Rhone-Poulenc S.A.
Neuilly sur Seine, France, 1 July 1976.
64
-------
TABLE 11. TYPICAL ANALYSIS OF CONDENSER OFF-GAS
FROM THE CHAUNY PROCESS36
(Percent by weight)
Phthalic anhydride 0.041
Maleic anhydride 0.243
Citraconic anhydride 0.015
Benzoic acid 0.031
Acetic acid 0.015
o-Toluic aldehyde 0.012
Miscellaneous organics 0.002
o-Xylene 0.09
H20 3.42
N2 76.3jQ
CO 0.40
C02 2.2
02 17.21
If this process is to be used for the purpose of controlling PAN
and MAN emissions, a new PAN plant will have to be built. Thus,
this control alternative is not applicable to existing PAN plants,
and it should be considered only as a process for future new
Plants.
5-7 SUMMARY OF THE TECHNICAL EVALUATION
The results of the technical evaluation for the nine alternative
control methods discussed above are summarized in Table 12. The
last line in the table indicates whether a specific alternative
was selected for economic and energy utilization evaluations.
65
-------
TABLE 12. SWMASY CF TECHNICAL EVALUATION OF COUTBOL ALTERNATIVES
Is th* control system
camntlv being1 ntml
by tlM PAD indwtxy?
Can 99% control effi-
ciency b« reasonably
**peetSd?
Important parameters
at f eetinqr control
efficiency
Applicability to
*Mf plants
Unique operating or
melnUiiiiiHf require-
ments
Are secondary wastes
generated?
Other technical
Advantage*
Other technical
Uaadvantaaea
.MHVBntoqea
Thermal incinerator/
steaai generation
Tea
Y«
Teaperafcure
Residence time
Mixing
Tas
Ho
Ho
Easy to sustain
combustion (good
reliability)
Can increase effi-
ciency by simply
increasing fuel
consumption
Low Maintenance
requirement*
Long operating life
CO is converted
to OCX;
Ho thsrmodynomic
ll^tation on con-
version efficiency
High fuel consumption
Difficult- to achieve
good Mixing when
treating large volume
of vest* gas
There should be a
market for steam
generated
Thermal incinerator
feed gas preheating
Ho
Tea
Temperature
Residence time
Mixing
Degree of pratreating
No. due to possi-
bilities of heat
exchanger explosion
Mo
Ho
Fuel savings due to
preheating
CO is oxidized to COj
Feed gas preheating
f las* and Urns the
combustion efficiency
Sustained fires in
the exchanger efflu-
ent gas side will
damage the exchanger
Safety and operations
problems in existing
plants
Catalytic incinerator/
HO
Ho
Temperature
Catalyst volume
Catalyst activity
Yes, but subject to
catalyst fouling and
possible polymer
coating
Frequent replacement
of catalyst
Ho
Fuel savings by oper-
ating at lower temp-
erature
CO ia converted to
c°2
Subject to catalyst
and possible polymer
coating
Large volume of cata-
lyst or high tempera-
ture needed for high
conversion efficiency
Met scrubber/liquid
Yes
t«
Solubilities of PAN
and MAN in water
Equilibrium; vapor
pressure
Yes
Corrosion-resistant
materials needed
So
Dilute waste can be
concentrated for ef-
fective Incineration
low equilibrium FAN
and MAN vapor pres-
sures enable efficient
removal
with a change in de-
sign, the clogging
problem in scrubber
can be avoided
Haste streams from
the purification
section can also be
treated in the same
system
CO emissions are not
The hardness of the
water contained in
the purge liquor is
emitted in the in-
cinerator effluent
gas
Excessive emissions
can occur from scrub-
ber when there is a
recirculating water
pump failure
Low removal effi-
ciency for aldehydes
which may be present
in the off-gas
Wet scrubber/
Yes
Yes
Solubilities of PAN
and HAN in water
Equilibrium vapor
pressure
Yes, for o-xylene
based plants only
Corrosion-resi stant
materials needed
Solid organic waste
and anhydrous melt of
organics (can be
burned in convention-
al burners as fuel)
Valuable chemical is
recovered from waste
Dilute waste can be
concentrated for easy
handling
lav equilibrium vapor
pressures of PAH and
MKH
With & change in de-
sign, the clogging
problem in scrubber
can be avoided
CO emissions are not
Need disposal of by-
roduct waste
Demin«rali£ed water
is needed for the
scrubber
Excessive emissions
can occur from scrub-
ber when there is a
recirculating water
pump failure
Filtration needed to
remove solids in
purge liquor
Low removal effi-
ciency for aldehydes
which may be present
in the off-gas
Not applicable to
naphthalene-based
plants
Het scrubber/bio logi-
Yes
Yes
Solubilities of FAN
and HAH in water
Equilibrium vapor
pressure
Ho, due to large BOD
loading
Corrosion-resistant
material needed
Have to modify the
process to meet
future water efflu-
ent standards
Ho
HaOH can be added to
the scrubbing water
and thus prevent
clogging of the
scrubber by solids
buildup
CO emissions are not
Large amount of KaQR
needed for neutrali-
zation
Large quantity of
waste water to be
treated
Large quantity of
dissolved solids in
the water discharge
Carbon adsorber/
HO
Yes
Carbon capacity
Boiling points
of organic pollutants
Hay be
Carres ion— res is tant
material needed
NO
Dilute waste can be
concentrated for ef-
fective incineration
High control effi-
ciency can be obtained
by using either •ore
carbon or more fre-
quent regeneration
•ever been applied to
Possible polymeriza-
tion of PAH and KAN
on carbon surface at
high temperatures
CO emissions are not
controlled
CHftUMY process
(with control )
Ye*
Yes
_
Yes, but for future
new plant* only
Ho
Ho
Ho saUfwr or sulfur
dioxide is required
ia the reactor feeti
The associated cata-
lytic incinerator ia
energetically self-
supported
The catalytic inciner-
ator can achieve high
conrersioo efficiency
The liquid waste
stream from distilla-
tion section ia pro-
ceased to form solid
pellets for burning
Can be used only when
build * new plant
bed filtration
Ho
Ho
equilibria vapor
pressure
Cooling temperature
Bo, due to blocking
of g«a passage in the
cooler
Refrigeration of gas
needed
Ho
Tbtt fiber filter is
almost an absolute
filter for particle*
Hater condeiuuttes will
ing section to form
slurry
PAH and MAN will build
up ob the. surface of
heat'*exchanger
Above phenomena will
decrease cooling
efficiency and will
clog the heat~exchang«r
Selected for
economic and energy
utilization walua-
tloo?
-------
SECTION 6
ECONOMIC EVALUATION AND ENERGY REQUIREMENTS
The economic and energy requirement evaluations were performed
for the existing control systems and for the candidate control
systems which can give a 99% removal of PAN and MAN. Evaluations
were made for the control of switch condenser off-gas from a rep-
resentative o-xylene based plant with flow rate and composition
as described~in Section 4.2.
Evaluations for the existing control systems were based on typ-
ical designs and typical operating conditions for control systems
in the existing plants. Details of these systems including mass
balances are presented in Appendix A. The existing control sys-
tems considered here include the following:
• Direct thermal incinerator with steam generation
• Direct thermal incinerator with feed gas preheating
• Wet scrubber with liquid waste incineration
The candidate control systems considered are those recommended
alternatives identified in the technical evaluation summary
(see Table 12). The candidate systems, derived either from
modifications of existing control technologies or from use of
emerging control technologies, are the following:
Direct thermal incinerator with steam generation
Wet scrubber with liquid waste incineration
Wet scrubber with MAN recovery
Carbon adsorber with waste incineration
CHAUNY PAN manufacturing process
The first four candidates are add-on control systems. The last
one is a modified PAN manufacturing process with its associated
control equipment; it was described in detail in Section 5.6.
The mass balance and operating conditions for each of the four
add-on systems are given in Appendix A.
Cost and energy usage estimates for existing and candidate add-
on control systems, for the conventional PAN manufacturing
Process, and for the CHAUNY process are summarized in Table 13
and discussed further below.
67
-------
TABLE 13. SUMMARY OF RESULTS OF ECONOMIC EVALUATION AND ENERGY REQUIREMENT STUDY
CO
Alternative
1. Existing control systems
Thermal incinerator/steam generation
Thermal incineration/gas preheating
Wet scrubber/waste incineration
2. New or improved add-on control systems
Thermal incinerator/steam generation
Wet scrubber/waste incineration
Wet scrubber/MAN recovery
Carbon adsorber/waste incineration
3. Conventional and CHAUNY manufacturing
processes
Conventional process (without control)
CHAUNY process (with control)
Capital,
cost,
106$
2.34
1.62
2.60
2.41
2.94
5.22
1.59
15.58
16.77
Capitalized
cost,
106S
4.16
2.88
4.61
4.28
5.22
9.27
2.82
27.68
48.93
Operating -
1 cost, a
$/metric 1
ton PAN
13.15
9.30
10.95
15-62
11.94
-8.61
6.66
218
' 202
Energy Usage
Recovery Net
latural gas,^ Electricity, (credit) TotalC
106 m3/yr 105 kWH/yr 1013 J 1013 J
34.9 -d 88 34
14.8 19.8 - 54
4.1 30.0 - 18
44.5 - 118 40
1.0 39.5 - 7
3.0 46.0 - 15
2.7 7.8 - 9
590 - 57
88.5 - 9
comparison purpose, the sale price of PAN is $510/ton as of November 22, 1976.37
bAlso includes the natural gas needed for producing steam which is to be consumed by the control systems.
cAssume that the lower heating value of natural gas is 33,600 J/m3 (950 Btu/ft3) and that the efficiency in electric power
generating is 37%.
applicable.
37Chemical Marketing Reporter. 210(21):50, 1976.
-------
6.1 ECONOMIC EVALUATION
Economic evaluation includes estimates of capital, capitalized,
and operating costs for both existing and candidate control
systems. Items included in the capital cost calculations are:
equipment cost, installation cost, taxes and insurance, startup
cost, interest on construction capital, and working capital.
Capitalized cost is a fund large enough to permit the interest
it earns to pay for the periodic replacement of the equipment for
an indefinitely long period of time. Items covered in the oper-
ating cost calculations include: labor, raw materials, utilities,
maintenance, overhead, taxes and insurance, depreciation, annual
interest charge, and byproduct credit.
The cost model and economic assumptions utilized in the estima-
tion of these costs are presented in Appendix B. The detailed
estimates of these costs, which are summarized in Table 13, are
contained in Appendix C. The cost figures for existing systems
were based on the data reported in a previous study which was
based on a plant survey and equipment manufacturer quotations.13
The cost figures for new or modified control systems were derived
from information supplied by process developers and from data in
the previous study mentioned above. The time frame for cost
estimates is July 1976.
Among the candidate control systems (which are either new or
improved systems), the wet scrubbing with MAN recovery system
has a negative operating cost, due to the MAN byproduct credit.
This means that there will be a net return when this system is
in operation and the byproduct is marketed. However, the system
is also characterized by a high capital investment requirement
and a high capitalized cost. The capital related costs of this
system are about double the corresponding costs for other alter-
natives. It should be mentioned here that the costs for dis-
posal of the anhydrous melt of organics from the recovery
Process and the solids separated from the purge liquor by
filtration are not included. According to a process developer,
disposal is not difficult because these organic substances can
be pumped and burned as a heavy fuel.
Direct thermal incineration with steam generation has the high-
est operating cost because of the large amount of fuel needed
to heat the condenser off-gas to a sufficiently high temperature
for efficient destruction of organic pollutants. Without the
waste heat boiler for generating byproduct steam, the operating
cost will be even higher.
!n the wet scrubber/waste incineration system, the organic
Pollutants contained in the off-gas are effectively concentra-
ted in the purge liquor. This reduces to a great extent the
69
-------
size of the waste incinerator and the cost of supplemental fuel
needed, resulting in a lower operating cost than that of direct
thermal incineration. In this case, a large portion (60%) of
the operating cost is derived from depreciation and the inter-
est charge on capital investment.
The carbon adsorption/incineration system is characterized by
both low capital investment and low operating cost. It should
be noted, however, that the cost figures were derived from a
rough estimate of equipment cost obtained from a carbon and
carbon-handling manufacturer. A representative of the manufac-
turer indicated that their preliminary laboratory evaluation was
inconclusive because insufficient data were obtained to determine
the design and operating conditions required to achieve a 99% re-
moval of PAN and MAN from the condenser off-gas. Further engi-
neering study is necessary to confirm that the activated carbon
process can remove organic pollutants to the desired level, and
that the capital and operating costs presented in Table 13 are
realistic. It is therefore recommended that these cost figures
be used only for order-of-magnitude comparison purposes.
Table 13 also shows the cost estimates for both conventional and
CHAUNY manufacturing processes. There is no significant differ-
ence in capital cost between these two processes. There is,
however, a large difference in capitalized cost, with that of
the CHAUNY process being about 50% more than that of the conven-
tional process with control systems included. This means that/
while producing the same amount of PAN and achieving the same
degree of air pollution control, the CHAUNY process will need
more capital at the start to earn enough interest to pay for
the periodic replacement of the process equipment. This is due
to the relatively short useful life period (5 years) of the
CHAUNY process equipment.38 Since the interest on capital cost
and depreciation based on a 5-year life are included in the oper-
ating cost, this high capitalized cost becomes less important.
6.2 ENERGY REQUIREMENTS
There are basically four items related to energy consumption in
the operation of control systems: (1) fuel burned in the incin-
erator to raise the temperature of the mixed gas; (2) fuel needed
to generate steam for use in evaporation (in MAN recovery) and
carbon regeneration; (3) energy needed to generate the electri-
city required to drive the draft fan and pumps; and (4) energy
contained in the steam recovered from the waste heat boiler.
38Zimmer, J. C. New Phthalic Anhydride Process. Hydrocarbon
Processing, 53 (2):111-112, 1974.
70
-------
The energy requirements for the existing and candidate add-on
control systems, the conventional manufacturing process, and
the CHAUNY process are summarized in Table 13. In this table,
items 1 and 2 mentioned above were combined and expressed as
the amount of natural gas required. Item 4 was subtracted from
the sum of the other three items to arrive at the net total
energy usage shown in the last column of Table 13.
Among the add-on candidate systems, direct incineration/waste
heat boiler has the highest energy requirement. The energy re-
quirement of this system is 2.6 times that of the second highest
energy-intensive control system (wet scrubber/MAN recovery) in
the same category. This is due to the large volume of low-
heating value dilute gas which is to be incinerated. The wet
scrubber/MAN recovery system requires a large amount of steam
for the MAN recovery process. This steam requirement consti-
tutes 70% of the system's total energy usage.
Wet scrubbing/waste incineration and carbon adsorption/waste in-
cineration both have low energy requirements because the dilute
organic contents in the condenser off-gas are effectively con-
centrated by scrubbing and adsorption processes to produce high-
heating-value waste streams for incineration. Without this con-
centration step, the direct incineration system requires fuel 25
times that of the incineration portion of these two systems.
With the new design of scrubber, the energy consumption of the
improved wet scrubber/waste incineration system is less than half
°f that of the corresponding existing system.
A difference in energy requirements between the conventional and
the CHAUNY processes is also shown in Table 13. The conventional
Process without control consumes about seven times the energy
required for the CHAUNY process with control. The CHAUNY process
enables a high steam recovery efficiency (due to its low air/
°_-xylene ratio) and conversion of the steam into mechanical
energy for air compression by using a steam turbine, instead of
an electric motor. There is another reason for its low energy
requirement: according to the process developer, the associa-
ted catalytic incinerator for air pollution control is energy
self sufficient and does not need supplemental fuel.
71
-------
SECTION 7
MOST FEASIBLE ALTERNATIVES AND RECOMMENDED DEMONSTRATION PROGRAMS
7.1 SELECTION OF MOST FEASIBLE SYSTEMS
The technical, economic, and energy considerations for the new and
improved candidate add-on control systems that can achieve 99% con-
trol of PAN and MAN are summarized in Tables 14 through 17. In
these tables, the advantages and disadvantages of each candidate
under the above three areas of consideration are listed for com-
parison purposes. Based on these comparisons, and the rationales
for selection of alternatives presented below, the following con-
trol systems were chosen as the most feasible:
1. Thermal incinerator/steam generation
2. Wet scrubber/MAN recovery
3. Carbon adsorber/waste incineration
Although thermal incinerator/steam generation has some drawbacks/
such as high fuel consumption, high operating cost, and the need
of a market for the steam generated, it has the highest operating
reliability among the alternatives. This is because external
fuel supply is used, and it is therefore easy to sustain combus-
tion regardless of transient changes in the main process opera-
tion. Also, carbon monoxide, a pollutant which is present in the
condenser off-gas, is oxidized to form C02. The control effi-
ciency of the incinerator can be increased simply by increasing
the fuel consumption, without having to replace the unit. In
addition, it has low maintenance requirements and long operating
life. Based on the above considerations, the direct incinerator/
steam generation was chosen as the most favorable system.
Wet scrubbing/MAN recovery is the second choice, mainly because
of its profitability resulting from the recovery of MAN. This
economic advantage is based on the assumption that the MAN re-
covered can find a market. This system is also characterized by
its high initial capital requirement, and its economic advantage
is strongly influenced by the capacity of the PAN plant. With
the future trend to larger PAN plants, the use of the MAN re-
covery process is even more attractive. On the other hand, owing
to its inability to control CO emissions, this system cannot be
applied to certain locations where state regulations limit the
72
-------
TABLE 14. TECHNICAL, ECONOMIC, AND ENERGY CONSIDERATIONS FOR THERMAL INCINERATOR/STEAM GENERATION
Technical
considerations
-j
u>
Advantages
Disadvantages
Economic
cons iderations
Easy to sustain combustion (good
reliability)
Can increase efficiency by simply
increasing fuel consumption
CO is converted to CO2
No thermodynamic limitations to
conversion efficiency
Low maintenance requirements
Long operating life
Difficult to achieve good mixing when
treating large volume of waste gas
Fuel availability is a potential
problem
• Highest operating cost among the
alternatives
• A market (or use) of steam generated
is needed
Energy
considerations
• Highest energy consumption among the
alternatives
-------
TABLE 15. TECHNICAL, ECONOMIC, AND ENERGY CONSIDERATIONS FOR WET SCRUBBER/WASTE INCINERATION SYSTEM
Advantages
Disadvantages^
Technical
considerations
Economic
considerations
Energy
considerations
Low equilibrium PAN and MAN vapor
pressures enable efficient removal
Dilute waste is concentrated for
effective incineration
With a change in scrubber design as
recommended, clogging problem can
be avoided
Waste streams from the purification
section can also be treated in the
same system
CO emissions are not controlled
Low removal efficiency for aldehydes
which may be present in the off-gas
The hardness of the water contained in
the purge liquor is emitted in the incin-
erator effluent gas
Relatively high operating cost (next to
incineration/heat recovery system)
• Low energy consumption
-------
TABLE 16. TECHNICAL, ECONOMIC, AND ENERGY CONSIDERATIONS FOR WET SCRUBBER/MAN RECOVERY SYSTEM
Advantages
Disadvantages
Technical
cons iderations
ui
Economic
considerations
Energy
considerations
Low equilibrium vapor pressures of
PAN and MAN
Dilute waste is concentrated for
recovery of valuable chemical
With a change in design as
recommended, clogging problem in
scrubber can be avoided
Extremely low operating cost (net
profit if product can be marketed)
CO emissions cannot be controlled
Demineralized water is needed for the
scrubber
Byproduct waste must be disposed of
Low removal efficiency of aldehydes which
may be present in the off-gas
Not applicable to naphthalene based plants
Highest capital cost among the alternatives
Relatively high energy consumption (next to
incineration/heat recovery)
-------
TABLE 17. TECHNICAL, ECONOMIC, AND ENERGY CONSIDERATIONS FOR CARBON ADSORBER/WASTE INCINERATION SYSTEM
Advantages
Disadvantages
Technical
considerations
Economic
considerations
Dilute waste is concentrated
for effective incineration
High control efficiency can
be obtained by using either
more carbon or more frequent
regeneration
Lowest capital and operating
costs among the alternatives
Has never been applied to PAN industry
CO emissions are not controlled
Possible polymerization of PAN and MAN
on carbon surface at high temperatures
Energy
considerations
Low energy consumption
-------
discharge of CO. Also, the scrubber has a poor absorption effi-
ciency for aldehydes which might be present, in some cases, in
the condenser off-gas.22
The wet scrubber/waste incineration system was not selected be-
cause the scrubber has the disadvantages mentioned above, and the
system has a much higher operating cost.
For a new o-xylene based PAN manufacturing plant, use of the
CHAUNY process, rather than the conventional process, appears to
be a better choice. This is due to its much lower total energy
consumption, efficient pollution control for both organics and
carbon monoxide, and compactness of the process equipment. In
addition, since no sulfur or sulfur dioxide is required in the
reactor feed, there are no emissions of sulfur oxides from the
CHAUNY process.
7.2 RECOMMENDED DEMONSTRATION PROGRAMS
Recommended demonstration programs for the control systems chosen
as most feasible were developed based on the development stage of
the individual unit processes comprising the control systems and
the technical evaluation of each control system as mentioned in
Section 5. The demonstration programs, including the time needed
and costs involved, are described below for add-on control systems,
They are also summarized in Table 18. (A demonstration program
for the CHAUNY PAN manufacturing process was not recommended be-
cause the process has been in operation in France since 1971. It
has also been licensed to manufacturers in the Philippines and
Spain, and therefore is well demonstrated.)
7.2.1 Thermal Incinerator/Steam Generation
Systems similar to this type are in operation at two existing
Plants. These two incinerators are being operated at a tempera-
ture of 700°C, giving control efficiency of 95% (design control
efficiency is 96.5% at 760°C at a different flow rate). d'i4'
The demonstration program for this system includes testing at one
of the two units to verify that the temperature increases or de-
sign changes described below can yield a 99% destruction of PAN
and MAN.
7-2.1.1 increase of Operating Temperature—
The operating temperature of the incinerator should be increased
to about 860°C, and measurements should be made to determine
whether this change will give 99% removal of combustible pollu-
tants, particularly PAN and MAN. If the desired removal effi-
ciency can be obtained at 860°C or any other reasonable tempera-
ture, this existing incinerator/waste heat boiler can be used for
further demonstration purposes. This project would take about 6
months and cost about $20,000.
77
-------
TABLE 18. SUMMARY OF ESTIMATED COST AND TIME REQUIREMENTS FOR THE RECOMMENDED DEMONSTRATION PROGRAMS
oo
1.
2.
3.
Demonstrated program
DIRECT INCINERATION/WASTE HEAT BOILER
A. Increase of operating temperature
B. Development of improved incinerator3
1. Measurement of temp distribution
2. Selection of burner and mixing
devices
3. Full-scale demonstration
WET SCRUBBING/MAN RECOVERY
A. Pilot-scale testing of improved scrubber
B. Full-scale demonstration of scrubber
CARBON ADSORPTION/INCINERATION
A. Laboratory testing of carbon process
B. Full-scale operation of the system
Time required,
months
6
1
6
12
18
30
6
30
Estimated cost,
$
20,000
6,000
20,000
100,000
120,000
2,500,000
25,000
1,700,000
alf the desired removal efficiency can be obtained by increasing the operating
temperature, this phase of the demonstration program is not needed.
-------
7.2.1.2 Development of Improved Incinerator —
If an increase in operating temperature does not result in the
desired removal efficiency, further steps should be taken to im-
prove the performance of the incinerator, particularly the mixing
efficiency in the burner-mixing section. The three steps involved
in this phase are described below.
7.2.1.2.1 Measurement of temperature distribution— The purpose of
this measurement is to determine whether a near-uniform cross-
sectional temperature is obtained before the gas mixture gets in-
to the retention chamber for effective destruction of combustible
organics. Thermoelectric thermometry, resistance thermometry,
radiation and optical pyrometry, or other techniques in flame re-
search can be used in the on-site temperature measurement. This
step would take 1 month and cost $6,000. If it is found that the
cross-sectional temperature variation is great in the retention
chamber, the next step should be taken.
7.2.1.2.2 selection of burner and mixing devices— The purpose of
this step is to select the combination of burner and mixing de-
vices which will provide good mixing between flame and inlet gas
to provide near-plug-flow conditions at the entrance to the reten-
tion section of the incinerator . Either a cold air or water model
can be used to test these devices. Such a model can be set up in
the laboratory for a small fraction of the cost of constructinga
complete incinerator for testing. It can be built of sheet metal
or wood although plexiglas is much to be preferred, since it
allows visual probing and observation with smoke tracers and dyes.
Mixing effects can bi studied while blending warm air or water in-
to room temperature air or water.
f Su S£ < ^
corrections should be made to account for difference
nneerence
the stream being considered.
It would take about 6 months to complete this project, at a cost
°f $20,000.
7 • 2 • 1 • 2 . 3 ^.n-««l. demonstration-After identifying the best
time involved9arer$ToO?000 and 12 months, respectively
79
-------
7.2.2 Wet Scrubber/MAN Recovery
There are essentially two processes included in this control
system. The wet scrubber process actually removes pollutants
from the condenser off-gas, whereas the MAN recovery process is
used as a means of waste disposal to eliminate secondary pollu-
tion problems resulting from operation of the wet scrubber. Be-
cause a MAN recovery process has been in operation since 1971 and
another new unit has been recently installed, a demonstration pro-
gram for this process is not needed. The demonstration program
for this system therefore involves pilot-scale testing and full-
scale operation of the three-stage scrubber described in Section
5.3.2.
7.2.2.1 Pilot-Scale Testing of Improved Scrubber—
This includes design, construction, and testing of the three-
stage, pilot-scale wet scrubber to determine the water recircula-
tion rate for each stage and other operating conditions of the
scrubber that would be required to obtain greater than 99% removal
of PAN and MAN. This pilot-scale scrubber should be designed to
operate at a gas flow rate of about 30 m3/min (1,000 scfm). After
the pilot unit has .been tested, data will be taken for further
scale up to full-scale units. The time needed for completing the
project is estimated to be 18 months. The cost of the project
would be about $120,000, including $85,000 of capital cost.
7•2•2.2 Full-Scale Demonstration of Scrubber—
This phase of the demonstration program includes the design of a
full-scale unit by using pilot-scale test results, construction/
and start up of the improved wet scrubber. Again, tests should
be performed to verify that over 99% removal of PAN and MAN is
achieved. The cost associated with this project would be about
$2,500,000, including $2,000,000 for equipment and installation
costs. It would need 30 months to complete.
7-2.3 Carbon Adsorber/Waste Incineration
Due to the lack of data on carbon adsorption of this type of
waste gas, laboratory tests would be necessary to determine the
conditions for achieving greater than 99.1% removal efficiency by
an activated carbon process alone. Laboratory tests are also
needed to determine the steam rate for regeneration and the gas
flow that is to be incinerated. There are two steps involved in
the demonstration program for this system.
7.2.3.1 Laboratory-Scale Testing of Carbon Process—
A laboratory-scale carbon bed should be used to handle a gas fl°w
of about 15 m3/min (500 scfm) . Operating parameters such as re-
tentivity and steam rate would be determined. Enough data should
be taken to determine the thickness of the carbon bed, linear gas
velocity, type and particle size of carbon, and other design
80
-------
factors for a full-scale carbon adsorption process. The possibi-
lity of forming polymer-like material on the carbon surface should
also be evaluated, especially when the carbon bed is subjected to
high temperature treatment during the regeneration cycle. The
time and cost of this step are estimated to be 6 months and
$25,000.
7.2.3.2 Full-Scale Operation of the System—
Based on the lab-scale test results, a full-scale carbon adsorp-
tion/incineration system should be designed and constructed at one
of the PAN manufacturing plants for full-scale testing and demon-
stration. Particular care should be given to the incinerator
unit. Although high efficiency of incineration is expected due to
the small gas flow to be treated, the burning-mixing combination
and incinerator configuration should be carefully selected.
Operating conditions should be adjusted to obtain 99% combined
removal efficiency for PAN and MAN. The project would take about
30 months, with a cost of $1,700,000.
81
-------
REFERENCES
1. Serth, R. W., and T. W. Hughes. Source Assessment: Phthalic
Anhydride. Contract 68-02-1874, U.S. Environmental Protec-
tion Agency, Research Triangle Park, North Carolina, 1976.
144 pp.
2. Anderson, E. V. Phthalic Anhydride Makers Foresee Shortage.
Chemical and Engineering News, 53(26):10-11, 1975.
3. Phthalic Anhydride. Chemical Marketing Reporter, 207(1):9,
1974.
4. Bernardini, F., and M. Ramacci. Oxidation Mechanism of
o-Xylene to Phthalic Anhydride. Ceimica e 1'Industria
(Milan), 48(1):9-17, 1966.
5. Spitz, P. H. Phthalic Anhydride Revisited. Hydrocarbon
Processing, 47(11):162-168, 1968.
6. Schwab, R. F., and W. H. Doyle. Hazards in Phthalic Anhy-
dride Plants. Chemical Engineering Progress, 66(9):49-53/
1970.
7. loffe, I. I., et al. Kinetics and Vapor Phase Oxidation
Mechanism of Aromatic Hydrocarbons, VI. Zhurnal Fizicheskoi
Khimii (Journal of Physical Chemistry), 29 (4):692-698, 1955.
8. Graham, J. J., P. F. Way, and S. Chase. Phthalic Anhydride
by Fluid Bed Process. Chemical Engineering Progress,
58(1):96-100, 1962.
9. Graham, J. J. The Fluidized Bed Phthalic Anhydride Process.
Chemical Engineering Progress, 66(9):54-58, 1970.
10. Personal communication with L. B. Evans, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina
(data originally supplied by Stepan Chemical Company),
12 June 1975.
11. Personal communication with A. B. Netzley, Air Pollution
Control District, County of Los Angeles, 28 March 1975.
82
-------
12. Schwartz, W. A., et al. Engineering and Cost Study of Air
Pollution Control for the Petrochemical Industry, Volume 7:
Phthalic Anhydride Manufacture from Ortho-Xylene. EPA-450/
3-73-006g, U.S. Environmantal Protection Agency, Research
Triangle Park, North Carolina, 1975. 108 pp.
13. Personal communication with H. M. Lacy, Monsanto Company,
Texas City, Texas, 7 April 1975.
14. Personal communication with T. Brennan, Bay Area Air Pollu-
tion Control District, San Francisco, California, 23 June
1976.
15. Personal communication with L. B. Evans, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina
(data originally supplied by Koppers Company), 12 June 1975.
16. Personal communication with M. A. Pierle, Monsanto Company,
St. Louis, Missouri, April 1976.
17. Ross, R. D. Pollution Abatement: Incineration of Solvent-
Air Mixtures. Chemical Engineering Progress, 68(8):59-64,
1972.
18. JANAF Thermochemical Tables, Second Edition. NSRDS-NBS 37,
U.S. Department of Commerce, National Bureau of Standards,
Washington, D.C., June 1971. 1141 pp.
19. Rolke, R. W., et al. Afterburner Systems Study. EPA-R2-72-
062, U.S. Environmental Protection Agency, Research Triangle
Park, North Carolina, 1972. 336 pp.
20. Personal communication with L. B. Evans, U.S. Environmental
Protection Agency, Research Triangle Park, North Carolina
(data originally supplied by Monsanto Company), 12 June 1975.
21. Moores, C. W. Control PA Emissions. Hydrocarbon Processing,
54(10):100-103, 1975.
22. Fawcett, R. L. Air Pollution Potential of Phthalic Anhydride
Manufacture. Journal of the Air Pollution Control Associa-
tion, 20(7) :46]-465, 1970.
23. Ferrari, D. C., and C. G. Bertram. Method and Apparatus for
Removal of Organic from Chemical Waste Gases. U.S. Patent
3,624,984 (to the Badger Company), December 7, iy/i.
24. Personal communication with R. A. Mount, Monsanto Company,
St. Louis, Missouri, 3 June 1976.
25. Yevzel'man, I. B., and G. D. Kharlampovish. Neutralization
of Exhaust Gas from Phthalic Anhydride PfJ^ion.
Khimicheskaya Promyshlennost (Moscow), 49 (6)-.379-380, 1973.
83
-------
26. Weyens, E. Recover Maleic Anhydride. Hydrocarbon Process-
ing, 53(11) :132-134, 1974.
27. Personal communication with A. David, UCB S.A., Brussells,
Belgium, 21 May 1976.
28. Wirth, F. Recover MA from PA Scrubber Water. Hydrocarbon
Processing, 54 (8):107-108, 1975.
29. Personal communication with I. V. Heese, BASF A.G.,
Ludwigshafen, Germany, 11 June 1976.
30. Personal communication with R. T. Wedell, Koppers Company/
Inc., Chicago, Illinois, 5 October 1976.
31. Development Document for Interim Final Effluent Limitations
Guidelines and New Source Performance Standards for the
Significant Organic Products Segment of the Organic Chemical
Manufacturing Point Source Category. EPA-440/1-72/045, U.S-
Environmental Protection Agency, Washington, D.C., 1975.
361 pp.
32. Kohl, A. L., and F. C. Riesenfeld. Gas Purification.
McGraw-Hill Book Company, Inc., New York, New York, 1960.
Page 428.
33. Personal communication with E. D. Kennedy, Monsanto Enviro-
Chem Systems, Inc., St. Louis, Missouri, 8 June 1976.
34. Phthalic Anhydride Process, CHAUNY. Rhone-Poulenc S.A.,
Neuilly sur Seine, France, 1976. 17 pp.
35. Personal communication with J. C. Zimmer, Rhone-Poulenc S.A-
Neuilly sur Seine, France, 13 May 1976.
36. Personal communication with J. C. Zimmer, Rhone-Poulenc S.A-
Neuilly sur Seine, France, 1 July 1976.
37. Chemical Marketing Reporter. 210(21):50, 1976.
38. Zimmer, J. C. New Phthalic Anhydride Process. Hydrocarbon
Processing, 53 (2):111-112, 1974.
84
-------
APPENDIX A
MATERIAL BALANCE FOR EXISTING, IMPROVED, AND NEW CONTROL SYSTEMS
TABLE A-l.
OVERALL MATERIAL BALANCE FOR A TYPICAL EXISTING
THERMAL INCINERATOR/STEAM GENERATION SYSTEM
(moles/hr)
Incineration temperature: 700°C
Stack gas temperature: 280°C
Component
Sulfur oxides
Carbon monoxide
Carbon dioxide
Phthalic anhydride
Maleic anhydride
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Nitrogen oxides
Total
Process vent gas
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
6,538,784
Natural gas Stack gas
533
1,962
265,084
57
161
8
694,032
571,607
5,030,814
146,510
16,280
197
162,790 6,564,455
85
-------
TABLE A-2.
OVERALL MATERIAL BALANCE FOR A TYPICAL EXISTING
THERMAL INCINERATOR/FEED GAS PREHEATING SYSTEM
(moles/hr)
Incineration temperature: 700°C
Preheated feed temperature: 480°C
Stack gas temperature: 370°C
Component
Sulfur oxides
Carbon monoxide
Carbon dioxide
Phthalic anhydride
Maleic anhydride
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Nitrogen oxides
Total
Process vent gas
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
6,538,784
Natural gas Stack gas.
533
1 957
227,130
113
322
17
499,675
885,842
5,030,814
n n a R rl
/ O^ D
197
74,305 6,646,600
86
-------
TABLE A-3. OVERALL MATERIAL BALANCE FOR A TYPICAL EXISTING
WET SCRUBBER/WASTE INCINERATION SYSTEM
(moles/hr)
Scrubber vent temperature: 38°C
Incinerator mixed gas temperature: 925°C
Component
Sulfur oxides
m Carbon monoxide
*-J Carbon dioxide
Phthalic anhydride (or acid)
Maleic anhydride (or acid)
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Process
vent gas
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
Scrubber Scrubber purge Natural Combustion Incinerator
vent liquor gas air vent
533
39,139
86,014
23
65
4
536,000
1,025,597
5,030,814
1,
62,630
1,107 22
3,153 64
164 3
214,400 276,850
91,000 12,340
343,700 343,700
18,670
2,194
Total
6,538,784 6,718,189
218,824
20,864
434,700
695,610
-------
TABLE A-4.
OVERALL MATERIAL BALANCE FOR AN IMPROVED
THERMAL INCINERATOR/STEAM GENERATION SYSTEM
(moles/hr)
Incineration temperature: 860°C
Stack gas temperature: 300°C
Component
Sulfur oxides
Carbon monoxide
Carbon dioxide
Phthalic anhydride
Maleic anhydride
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Nitrogen oxides
Total
Process vent gas
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
6,538,784
Natural gas Stack gas
533
391
314,252
11
32
2
787,900
579,496
5,030,814
186 .740
J- V w f 1 ^ \J
*"l rt *7 A f\
J\\ I 4 Q
*• v / / •* y
225
207,489 6,713,656
88
-------
TABLE A-5. OVERALL MATERIAL BALANCE FOR AN IMPROVED WET SCRUBBER/WASTE INCINERATION SYSTEM
(moles/hr)
Number of scrubber stages: 3
Scrubber vent temperature: 38°C
Incinerator mixed gas temperature:
925°C
Process vent
Component gas
Sulfur oxides
Carbon monoxide
g Carbon dioxide
Phthalic anhydride
(or acid)
Maleic anhydride
(or acid)
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
Scrubber purge
Scrubber vent liquor
533
39,139
86,014
10 1,120
28 3,190
1 166
536,000 79,011
1,025,597
5,030,814
Combustion Incinerator
Natural gas air vent
27,379
1
3
97,863
33,633 5,606
134,531 134,531
3,695
411
TOTAL
6,538,784
6,718,136
83,487
4,106
168,164
265,383
-------
vo
o
TABLE A-6. OVERALL MATERIAL BALANCE FOR A NEW WET SCRUBBER/MAN RECOVERY SYSTEM
(moles/hr)
Number of scrubber stages: 3
Scrubber vent temperature: 38°C
MAN concentration in MAN recovery process feed: 20 wt%
Component
Sulfur oxides
Carbon monoxide
Carbon dioxide
Phthalic anhydride (or acid)
Maleic anhydride (or acid)
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Process vent
gas
533
39,139
86,014
1,130
3,218
167
352,172
1,025,597
5,030,814
Scrubber vent
533
39,139
86,014
10
28
1
536,000
1,025,597
5,030,814
Scrubber purge
liquor
1,120
3,190
166
79,011
MAN product
3,028
TOTAL 6,538,784 6,718,136 83,487 3,028
-------
TABLE A-7. OVERALL MATERIAL BALANCE FOR A CARBON ADSORBER/WASTE INCINERATION SYSTEM
(moles/hr)
Steam rate: 15 Ib steam/lb organics
Retentivity: 15%
Incinerator feed gas temperature: 315°C
Incinerator temperature: 925°C
Component
Sulfur oxides
Carbon monoxide
Carbon dioxide
Phthalic anhydride
Maleic anhydride
Benzoic acid
Water
Oxygen
Nitrogen
Methane
Ethane
Process vent
gas
533
39,139
86,014
1,130
3,219
167
352,172
1,025,597
5,030,814
Adsorber vent
533
39,139
86,014
10
28
1
352,172
1,025,597
5,030,814
Incinerator
feed gas
1,120
3,190
166
415,083
Combustion Incinerator
Natural gas air vent
33,500
1
3
441,313
43,830 7,305
175,317 175,317
8,703
967
TOTAL
6,538,784
6,718,136
419,599
9,670
219,147
657,439
-------
APPENDIX B
COST MODEL AND ECONOMIC ASSUMPTIONS
Contained herein are the cost model and economic assumptions
utilized in generating the detailed cost estimates presented in
Apprendix C and summarized in Table 10.
1. COST MODEL
Items included in capital requirement calculations include:
equipment cost, installation cost, taxes and insurance, start up
cost, interest on construction capital, and working capital. Con-
tingencies and land costs were excluded. Taxes and insurance were
assumed to be 2% of installed equipment cost. Start up cost and
working capital were each estimated to be 20% of annual operating
cost. Interest on construction capital (11% total capital require-
ment) was calculated based on an annual 8% interest rate with the
capital requirement split between engineering and design (10%) an<^
plant investment (90%). Total construction time was assumed to be
3 years, with 1.5 years of engineering and design and 2.5 years of
construction.
Calculation of capitalized cost assumed a 10% plant salvage value
and an annual 8% interest rate. Plant lifetime was assumed to be
10 years for all options except the activated sludge process (20
years) and the Rhone-Poulenc manufacturing process (5 years).
Items covered in the operating cost calculation include: labor,
raw materials, utilities, maintenance, overhead, taxes and in-
surance, depreciation, annual interest charge, and byproduct cre-
dit. Labor cost was assumed to be $5/hr. Fuel cost was assumed
to be $0.60/ 106 Btu. Electricity cost was assumed to be 2£/kW
hr. Maintenance charges ranged from 3% to 5% of installed equip-
ment cost, with 4% assumed if not otherwise specified. Plant
overhead was assumed to be 90% of operating labor cost. As in the
capital cost estimate, taxes and insurance were assumed to be 2%
of installed equipment cost. Depreciation was assumed to be 10%
of installed equipment cost for all options except the activated
sludge process (5%) and the Rhone-Poulenc manufacturing process
(20%). The annual interest charge was calculated assuming 75%
equity financing requiring 15% return on investment and 25%
financing requiring 8% return on investment.
92
-------
2. SPECIFIC PROCESS ASSUMPTIONS
Following is a list of the additional assumptions required for
each specific process not listed above or in the text.
Specific Process Assumptions
- Incineration/Waste Heat Boiler
- Maintenance, 4% installed equipment cost
- Boiler feed water, 40£/103 gal
- Byproduct steam credit, 75C/103 lb
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
- Incineration/Heat Exchanger
- Maintenance, 5% installed equipment cost
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
- Wet Scrubber
- Maintenance, 5% installed equipment cost
- Process water, IOC/103 gal
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
- Maelic Anhydride Recovery Process
- Maintenance, 3% battery limits installed equipment
cost
- Battery limits, 71% total installed equipment cost
- Cooling water, 2C/103 gal
- Deionized water, $1.89/103 gal
- Steam, 75C/103 lb
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
- Maleic anhydride byproduct credit $0.30/lb
- Carbon Adsorption Process
- Installation cost is 40% of equipment cost
- Maintenance, 4% of installed equipment cost
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
• CHAUNY Manufacturing Process
- Outside battery limits installed equipment cost
assumed to be 1/3 battery limits installed equipment
cost
- Raw materials and catalysts, $100/metric ton product
- Byproduct steam credit, 75C/103 lb
- Maintenance, 3% battery limits installed equipment cost
- Lifetime, 5 years
- Depreciation/ 20% installed equipment cost
93
-------
Conventional Phthalic Anhydride Manufacture
- Only emission control cost included is for storaae
tanks
- Maintenance, 5% installed equipment cost
- Raw materials, o-xylene, SO2 - 7.4<:/lb product
- Utilities including catalyst, 0.4C/lb product
- Lifetime, 10 years
- Depreciation, 10% installed equipment cost
94
-------
APPENDIX C
DETAILED ESTIMATES OF CAPITAL, CAPITALIZED, AND OPERATING COSTS
Tables C-la through C-9b give detailed estimates of capital and
operating costs for existing and candidate add-on control systems,
and for the conventional PAN manufacturing process and the CHAUNY
PAN process. The capitalized costs can be obtained by using the
following formula:
where K
£v
i
n
K = C
the capitalized cost
original installed cost of equipment
equipment replacement cost
annual interest rate
equipment life, years
(C-l)
Cost estimates given in Tables C-la through C-9b are all for PAN
plants with a production capacity of 59,000 metric tons/yr.
95
-------
TABLE C-la. CAPITAL COST OF A TYPICAL EXISTING THERMAL
INCINERATOR/STEAM GENERATION SYSTEM
Direct cost
Equipment cost * R71 qno
Installation cost ? 871^900
Total direct construction cost 1 743 800
Indirect cost
Design and Engineering fees}
Indirect labor costs > Included in the
Contingency ( installation cost
Taxes and insurance 34 OQQ
Total indirect construction cost 34 900
Startup cost 155 2QQ
Interest on construction capital o^nSnn
Working capital 155 200
Total capital requirement $2 339 800
TABLE C-lb. OPERATING COST OF A TYPICAL EXISTING THERMAL
INCINERATOR/STEAM GENERATION SYSTEM
$ 21,000
Maintenance materials 69 800
Utilities '
Process water 45.600
Fuel . . 843,500
Total utilities 889,100
Plant overhead 18 900
Taxes and insurance 34*900
Depreciation 174 ,'400
Interest 159,100
Total operating cost (per year) 1,367,200
Byproduct credit 591/100
Net operating cost (per year) ? 776,'lQO
96
-------
TABLE C-2a. CAPITAL COST OF A TYPICAL EXISTING THERMAL
INCINERATOR/FEED GAS PREHEATING SYSTEM
Direct cost
Equipment cost $ 802,200
Installation cost 397,600
Total direct construction cost 1,199,800
Indirect cost
Design and engineering feesi Included in the
indirect labor costs J installation cost
Contingency J
Taxes and insurance 24,000
Total indirect construction cost 24,000
Startup cost 110,000
Interest on construction capital 173,300
Working capital HOfOOO
Total capital requirement $1,617,100
TABLE C-2b. OPERATING COST OF A TYPICAL EXISTING THERMAL
INCINERATOR/FEED GAS PREHEATING SYSTEM
Labor $ 5'200
Maintenance materials 48,000
Utilities _ rnn
Electricity 39,600
Fuel 198,300
Total utilities 237,900
Plant overhead o^nnn
Taxes and insurance 24,000
Depreciation }?°'nnn
Interest % «H2'ggg
Total operating cost (per year) $549,900
97
-------
TABLE C-3a. CAPITAL COST OF A TYPICAL EXISTING WET
SCRUBBER/WASTE INCINERATION SYSTEM
Direct cost
Equipment cost $ 551,000
Installation cost 1,471,800.
Total direct construction cost 2,022,800
Indirect cost
Design and engineering fees),.,,,.,,
Indirect labor costs } Deluded in the
Contingency ) installation cost
Taxes and insurance 40,500^
Total indirect construction cost 40,500
Startup cost 129,200
Interest on construction capital 278,600
Working capital 129,200.
Total capital requirement $2,600,400
TABLE C-3b. OPERATING COST OF A TYPICAL EXISTING WET
SCRUBBER/WASTE INCINERATION SYSTEM
Labor $ 11,900
Maintenance materials 125,500
Utilities
Process water 1,100
Electricity 45,000
Fuel 83,300
Total utilities 129,400
Plant overhead 10,700
Taxes and insurance 40,500
Depreciation 202,300
Interest 174,500
Total operating cost (per year) $694,800
98
-------
TABLE C-4a. CAPITAL COST OF AN IMPROVED THERMAL
INCINERATOR/STEAM GENERATION SYSTEM
Direct cost
Equipment cost $ 871,900
Installation cost 871,900
Total direct construction cost 1,743,800
Indirect cost
Design and engineering fees\ Jncluded ±n the
Indirect labor costs > installation cost
Contingency )
Taxes and insurance 34 , 900
Total indirect construction cost 34,900
Startup cost 184,300
Interest on construction capital 257,700
Working capital 184,300
Total capital requirement $2,405,000
TABLE C-4b. OPERATING COST OF AN IMPROVED THERMAL
INCINERATOR/STEAM GENERATION SYSTEM
Labor $ 21,000
Maintenance materials 69,800
Utilities
Process water 45,600
Fuel 983,200
Total utilities 1,028,800
Plant overhead 18,900
Taxes and insurance 34,900
Depreciation 174,400
Interest 164,900
Total operating cost (per year) 1,512,700
Byproduct credit 591,100
Net operating cost (per year) $921,600
99
-------
TABLE C-5a. CAPITAL COST OF AN IMPROVED WET SCRUBBER/
WASTE INCINERATION SYSTEM
Direct cost
Equipment cost $ 596,800
Installation cost 1,696/100
Total direct construction cost 2,292,900
Indirect cost
Design and engineering fees ) x .. , , .
Indirect labor costs > *ncluded in the
Contingency J installation cost
Taxes and insurance 45,900_
Total indirect construction cost 45,900
Startup cost 140,900
Interest on construction capital 314,500
Working capital 140,900^
Total capital requirement $2,935,100
TABLE C-5b. OPERATING COST OF AN IMPROVED WET
SCRUBBER/WASTE INCINERATION SYSTEM
Labor $ 11,900
Maintenance materials 127,400
Utilities
Process water 1,100
Electricity 59,300
Fuel 19,700
Total utilities 80,100
Plant overhead 10,700
Taxes and insurance 45,900
Depreciation 229,300
Interest 198,900
Total operating cost (per year) §704,200
100
-------
TABLE C-6a. CAPITAL COST OF A NEW WET SCRUBBER/
MAN RECOVERY SYSTEM
Direct cost
Major equipment cost \
Other materials cost > $4,068,100
Installation cost )
Total direct construction cost 4,068,100
Indirect cost
Design and engineering fees! Included in the
Indirect labor cost > installation cost
Contingency J
Taxes and insurance 81,400
Total indirect construction cost 81,400
Startup cost n
Interest on construction capital 556,700
Working capital 244'700
Total capital requirement $5,195,600
TABLE C-6b. OPERATING COST OF A NEW WET SCRUBBER/
MAN RECOVERY SYSTEM
Labor $ 46,700
Maintenance materials 145,400
Utilities
Process water ' »700
Demineralized water « inn
Electricity inn
Steam ' r
Total utilities 153,100
Plant overhead o?'?nn
Taxes and insurance °1,400
4Ub,ouu
347,800
, ,
Total operating cost (per year)
Byproduct credit
Net operating cost (per year)
101
-------
TABLE C-7a. CAPITAL COST OF A NEW CARBON ADSORBER/
WASTE INCINERATION SYSTEM
Direct cost
Equipment cost $ 850,900
Installation cost 390
Total direct construction cost 1,241,000
Indirect cost
Design and engineering fees } ,. .. ,- ., .
Indirect labor cost > Included in the
Contingency ) installation cost
Taxes and insurance 24,800
Total indirect construction cost 24,800
Startup cost yg 700
Interest on construction capital 170,700
Working capital 73 700
Total capital requirement $1,593,900
TABLE C-7b. OPERATING COST OF A NEW CARBON ADSORBER/
WASTE INCINERATION SYSTEM
Labor $ 11,900
Materials
Operating 11,200
Maintenance 60,400
Total materials 71,600
Utilities
Steam 7,200
Electricity 11,800
Fuel 24,200
Total utilities 43,200
Plant overhead 10,700
Taxes and insurance 24,800
Depreciation 124,100
Interest 106,900
Total operating cost (per year) ?393,200
102
-------
TABLE C-8a. CAPITAL COST OF A TYPICAL EXISTING o-XYLENE
BASED PAN MANUFACTURING PLANT
Direct cost
Major equipment
Other materials } $ 8,593,500
Installation cost
Total direct plant construction cost 8,593,500
Indirect cost
Design and engineering fees | Included in the
Indirect labor cost > instaliation cost
Contingency /
Taxes and insurance 171,900
Total indirect plant construction cost 171,900
Startup cost 2,573,900
Interest on construction capital 1,669,600
Working capital 2,573,900
Total capital requirement $15,582,800
TABLE C-8b. OPERATING COST OF A TYPICAL EXISTING O-XYLENE
BASED PAN MANUFACTURING PLANT
Labor $ 153,600
Materials
Raw and process 9,584,000
Maintenance 429,700
Total materials 10,013,200
Utilities
Plant overhead J-38,200
Taxes and insurance 171,900
Depreciation ?!*?'i-nn
Interest 1,143,600
Total operating cost (per year) $12,869,500
103
-------
TABLE C-9a. CAPITAL COST OF THE CHAUNY PAN
MANUFACTURING PROCESS
Capacity: 59,000 metric tons/yr
Direct cost
Installed major equipment $ 7 ,500,000
Installed other materials 2,500,OOP
Total direct plant construction cost 10,000,000
Indirect cost
Design and engineering fees \ Included in the
indirect labor cost > installation cost
Contingency ;
Taxes and insurance 200,OOP
Total indirect plant construction cost 200,000
Startup cost 2,386,400
Interest on construction capital 1,796,700
Working capital 2,386,400
Total capital requirement $16,769,600
TABLE C-9b. OPERATING COST OF THE CHAUNY PAN
MANUFACTURING PROCESS
Capacity: 59,000 metric tons/yr
Labor ,
Operating )
Maintenance> $ f 979,600
Overhead )
Total labor 979,600
Materials
Raw and process\ 7,738,700
Maintenance J
Total materials 7,738,700
Taxes and insurance 200,000
Depreciation 2,000,000
Interest 1,209,800
Total operating cost (per year) $12,128,100
104
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-77-188
2.
3. RECIPIENTS ACCESSION-NO.
4. TITLE AND SUBTITLE
Phthalic Anhydride Plant Air Pollution Control
5. REPORT DATE
September 1977
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
C.T. ChiandT.W. Hughes
8. PERFORMING ORGANIZATION REPORT NO,
MRC-DA-586
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Monsanto Research Corporation
Station B, Box 8
Dayton, Ohio 45407
10. PROGRAM ELEMENT NO.
1AB015; ROAP 21AXM-011
11. CONTRACT/GRANT NO.
68-02-1320, Task 25
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
Task Final; 4/76-12/76
14. SPONSORING AGENCY CODE
EPA/600/13
15. SUPPLEMENTARY NOTES Task officer E. J. Wooldridge is no longer with IERL-RTP; contact
I.A. Jefcoat, Mail Drop 62, 919/541-2547.
16. ABSTRACT The report summarizes a technical and economic evaluation of add-on
control systems and process modifications for reducing, by 99%, the emissions of
phthalic and maleic anhydrides from the main process vent gas in phthalic anhydride
manufacturing plants. A survey was made to identify present (1976) control practices
and their control efficiencies in the phthalic anhydride industry. Based on theoretical
and practical considerations, existing control technology alternatives were evaluated
to determine whether they can be improved to obtain the desired control efficiency.
Technical evaluation of these alternatives led to identification of candidate alternatives
which apply to the manufacturing process, and which can achieve 99% overall removal
efficiency for phthalic and maleic anhydrides. Design and operating parameters for
achieving the desired control efficiency were also determined. Cost estimates and
an energy utilization study were performed for the candidate alternatives. Demonstra-
tion programs are recommended for the most promising alternatives.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b.lDENTIFIERS/OPEN ENDED TERMS
c. COS AT I Field/Group
Air Pollution
Phthalic Anhydride
Maleic Anhydride
Incinerators
Carbon
Adsorption
Scrubbers
Assessments
Air Pollution Control
Stationary Sources
Source Assessment
13B
07C
07B
07A
14B.
3. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (ThisReport)
Unclassified
21. NO. OF PAGES
113
20. SECURITY CLASS (Thispage)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
105
------- |