EPA-450/3-73-006-f
June 1975
ENGINEERING
AND COST STUDY
OF AIR POLLUTION CONTROL
FOR THE
PETROCHEMICAL INDUSTRY
VOLUME 6: ETHYLENE OXIDE
MANUFACTURE BY DIRECT
OXIDATION OF ETHYLENE
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
-------
TECHNICAL REPORT DATA
(Please read InUruciions on the reverse before complet"*"1
1. REPORT NO. 2.
EPA-450/3-73-006-f
3 PE3 244 116
4. TITLE AND SUBTITLE
Engineering and Cost Study of Air Pollution Control
for the Petrochemical Industry, Volume 6: Ethylene
Oxide Manufacture by Direct Oxidation of Ethylene
5. REPORT DATE
June 1975
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
D. E. Field, R. C. Barley, F. B. Higgins, Jr.,
J. A. Lee, R. Newirth, J. W. Pervier
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORCVNIZATION NAME AND ADDRESS
Houdry Division/Air Products and Chemicals, Inc.
P. 0. Box 427
Marcus Hook, Pennsylvania 19061
10. PROGRAM ELEMENT NO.
11. CONTRACT/GRANT NO.
68-02-0255
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Air Quality Planning & Standards
Industrial Studies Branch
Research Triangle Park, N.C. 27711
13. TYPE QF REPORT AND PERIOD COVERED
Final Report
14. SPONSORING AGENCY CODE
15. SUPPLEMENTARY NOTES
16. ABSTRACT
This document is one of a series prepared for the Environmental Protection
Agency (EPA) to assist it in determining those petrochemical Drocesses for which
standards should be promulgated. A total of nine petrochemicals produced by
twelve distinctly different processes has been selected for this type of in-depth
study. Ten volumes, entitled Engineering and Cost Study of Air Pollution Control
for the Petrochemical Industry (EPA-450/3-73-006a through i) have been preDared.
A combination of expert knowledge and an industry survey was used to select these
processes. The industry survey has been published separately in a series of four
volumes entitled Survey Reoorts on Atmospheric Emissions from the Petrochemical
Industry (EPA-450/3-73-005a, b, c, and d).
This volume covers the manufacture of ethylene oxide by direct oxidation
of ethylene. Included is a process and industry description, an engineering
description of available emission control systems, the cost of these systems, and
the financial impact of emission control on the industry. Also Dresented are
suggested air episode procedures and plant inspection procedures.
17. KEY WORDS AND DOCUMENT ANALYSIS
a. DESCRIPTORS
b.IDENTIFIERS/OPEN ENDED TERMS
c. COSATI Field/Group
Air Pollution
Hydrocarbons
Ethylene Oxide
Carbon Monoxide
Petrochemical Industry
7A
7B
7C
11G
13B
13H
J . _ „ _ ..
13. DISTRIBUTION STATEMENT
•
/•
19. SECURITY CLASS (This Report) |2'l. NO. OF PASTS
Unclassified
20. SECURITY CLASS (This page)
Unciassified
EPA Form 2220-1 (9-73)
-------
INSTRUCTIONS
1. REPORT NUMBER
Insert the EPA report number as it appears on the cover of the publication.
2. LEAVE BLANK
3. RECIPIENTS ACCESSION NUMBER
Reserved for use by each report recipient.
4. TITLE AND SUBTITLE
Title should indicate clearly and briefly the subject coverage of the report, and be displayed prominently. Set subtitle, if used, in smaller
type or otherwise subordinate it to main title. When a report is prepared in more than one volume, repeat the primary title, add volume
number and include subtitle for the specific title.
5. REPORT DATE
Each report shall carry a date indicating at least month and year. Indicate the basis on which it was selected (e.g., date of issue, date of
approve!, date of preparation, etc.).
6. PERFORMING ORGANIZATION CODE
Leave blank.
7. AUTHOR (S)
Give name(s) in conventional order (John R. Doe, J. Robert Doe, etc.). List author's affiliation if it differs from the performing organi-
zation.
8. PERFORMING ORGANIZATION REPORT NUMBER
Insert if performing organization wishes to assign this number.
9. PERFORMING ORGANIZATION NAME AND ADDRESS
Give name, street, city, state, and ZIP code. List no more than two levels of an organizational hirearchy.
10. PROGRAM ELEMENT NUMBER
Use the program element number under which the report was prepared. Subordinate numbers may be included in parentheses.
11. CONTRACT/GRANT NUMBER
Insert contract or grant number under which report was prepared.
12. SPONSORING AGENCY NAME AND ADDRESS
Include ZIP code.
- 13. TYPE OF REPORT AND PERIOD COVERED
Indicate interim final, etc., and if applicable, dates covered.
14. SPONSORING AGENCY CODE
Leave blank.
15. SUPPLEMENTARY NOTES
Enter information not included elsewhere but useful, such as: Prepared in cooperation with, Translation of, Presented at conference of,
To be published in, Supersedes, Supplements, etc.
16. ABSTRACT
Include a brief (200 words or less) factual summary of the most significant information contained in the report. If the report contains a
significant bibliography or literature survey, mention it here.
17. KEY WORDS AND DOCUMENT ANALYSIS
(a) DESCRIPTORS - Select from the Thesaurus of Engineering and Scientific Terms the proper authorized terms that identify the major
concept of the research and are sufficiently specific and precise to be used as index entries for cataloging.
(b) IDENTIFIERS AND OPEN-ENDED TERMS - Use identifiers for project names, code names, equipment designators, etc. Use open-
ended terms written in descriptor form for those subjects for which no descriptor exists.
(c) COSATI FIELD GROUP - Field and group assignments are to be taken from the 1965 COSATI Subject Category List. Since the ma-
jority of documents are multidisciplinary in nature, the Primary Field/Group assignmcnt(s) will be specific discipline, area of human
endeavor, or type of physical object. The application(s) will be cross-referenced with secondary Field/Group assignments that will follow
the primary posting(s).
18. DISTRIBUTION STATEMENT
Denote rcleasability to the public or limitation for reasons other than security for example "Release Unlimited." Cite any availability to
the public, with address and price.
19. & 20. SECURITY CLASSIFICATION
DO NOT submit classified reports to the National Technical Information service.
21. NUMBER OF PAGES
Insert the total number of pages, including this one and unnumbered pages, but exclude distribution list, if any.
22. PRICE
Insert the price set by the National Technical Information Service or the Government Printing Office, if known.
EPA Form 2220-1 (9-73) (Reverse)
-------
EPA-450/3-7 3-006-f
ENGINEERING
AND COST STUDY
OF AIR POLLUTION CONTROL
FOR THE
PETROCHEMICAL INDUSTRY
VOLUME 6: ETHYLENE OXIDE
MANUFACTURE BY DIRECT
OXIDATION OF ETHYLENE
by
D. E. Field, R. C. Barley, F. B. Higgins, Jr.,
J. A. Lee, R. Newirth, and J. W. Pervier
Houdry Division
Air Products and Chemicals, Inc.
P.O. Box 427
Marcus Hook, Pennsylvania J 906.1
Contract No. 68-02-0255
EPA Project Officer: Leslie B. Evans
Prepared for
ENVIRONMENTAL PROTECTION AGENCY
Office of Air and Waste Management
Office of Air Quality Planning and Standards
Research Triangle Park, North Carolina 27711
June 1975
-------
This report is issued by the Environmental Protection Agency to report
technical data of interest to a limited number of readers. Copies are
available free of charge to Federal employees, current contractors
and grantees, and nonprofit organizations - as supplies permit - from
the Air Pollution Technical Information Center, Environmental Protection
Agency, Research Triangle Park , North Carolina 2771 1; or, for a
fee, from the National Technical Information Service, 5285 Port Royal
Road, Springfield, Virginia 22161.
This report was furnished to the Environmental Protection Agency
by Houdry Division/Air Products and Chemicals, Inc. , Marcus Hook ,
Pennsylvania 19061, in fulfillment of Contract No. 68-02-0255. The
contents of this report are reproduced herein as received from Houdry
Division/Air Products and Chemicals, Inc. The opinions, findings,
and conclusions expressed are those of the author and not necessarily
those of the Environmental Protection Agency Mention of company
or product names is not to be considered as an endorsement by the
Environmental Protection Agency.
Publication No. EPA-450/3-73-006-f
-------
PETROCHEMICAL AIR POLLUTION STUDY
INTRODUCTION TO SERIES
This document is one of a series prepared for the Environmental Protection
Agency (EPA) to assist it in determining those petrochemical processes for
which standards should be promulgated. A total of nine petrochemicals produced
by 12 distinctly different processes has been selected for this type of
in-depth study. These processes are considered to be ones which might warrant
standards as a result of their impact on air quality. Ten volumes, entitled
Engineering and Cost Study of Air Pollution Control for the Petrochemical
Industry (EPA-450/3-73-006a through j) have been prepared.
A combination of expert knowledge and an industry survey was used to
select these processes. The industry survey has been published separately
in a series of four volumes entitled Survey Reports on Atmospheric Emissions
from the Petrochemical Industry (EPA-450/3-73-005a, b, c and d).
The ten volumes of this series report on carbon black, aerylonitrile,
ethylene dichloride, phthalic anhydride (two processes in a single volume),
formaldehyde (two processes in two volumes), ethylene oxide (two processes
in a single volume) high density polyethylene, polyvinyl chloride and vinyl
chloride monomer.
-------
ACKNOWLEDGEMENTS
The study reported in this volume, by its nature, relied on the fullest
cooperation of the companies engaged in the production of ethylene oxide.
Had their inputs been withheld, or valueless, the study would not have been
possible or at least not as extensive as here reported. Hence, Air Products
wishes to acknowledge this cooperation by listing the contributing companies.
BASF-Wyandotte Corporation
Dow Chemical Company
Houston Chemical Corporation
Jefferson Chemical Company
Koch Chemical Company
Northern Petrochemical Company
Union Carbide Chemical Company
Additionally, Air Products wishes to acknowledge the cooperation of the
member companies of the U. S. Petrochemical Industry and the Manufacturing
Chemists Association for their participation in the public review of an
early draft of this document. More specifically, the individuals who served
on the EPA's Industry Advisory Committee are to be commended for their
advice and guidance at these public meetings.
-------
TABLE OF CONTENTS
SECTION PAGE NUMBER
Summary i
I. Introduction EO-1
II. Process Description 2
III. Producers of Ethylene Oxide and Emissions 9
IV. Emission Control Devices & Systems 18
V. National Emission Inventory 27
VI. Ground Level Air Quality Determination 29
VII. Cost Effectiveness of Controls 30
VIII. Source Testing 33
IX. Industry Growth Projection 34
X. Plant Inspection Procedures 37
XI. Financial Impact 39
XII. Cost to Industry 46
XIII. Emission Control Deficiencies 48
XIV. Research & Development Needs 50
XV. Research & Development Programs 53
XVI. Sampling, Monitoring and Analytical Methods 56
XVII. Emergency Action Plan for Air Pollution Episodes 60
References 68
Appendix I
Appendix II
Appendix III
I-1
II-l
III-l
-------
0-1
2
3
-1
2A
2B
3
4A
4B
5
6
7
8
9
10
11
12
13
14
14A
15
16
17
18
19
20
LIST OF ILLUSTRATIONS
PAGE NUMBER
Simplified Flov Diagram - Air Oxidation Process EO-4
Simplified Flov Diagram - Oxygen Oxidation Process 7
Ethylene Oxide Capacity Projection 35
LIST OF TABLES
PAGE NUMBER
Heat Balance EO-3
Material Balance - Air Oxidation Process 5
Material Balance - Oxygen Oxidation Process 8
National Emissions Inventory (5 pages) 10
Vent Gas Compositions - Air Oxidation Process 15
Vent Gas Compositions - Oxygen Oxidation Process 16
Catalog of Emission Control Devices (2 pages) 19
Catalytic Incineration - Material Balance -
(Air Oxidation Process) 22
Steam Generation - Material Balance -
(Oxygen Oxidation Process) 23
Thermal Incineration - Material Balance -
(Oxygen Oxidation Process) 24
Emission Source Summary 28
Cost Effectiveness - Air Oxidation Process 31
Cost Effectiveness - Oxygen Oxidation Process 32
Number of Plants - 1985 36
Manufacturing Cost - Typical 200 MM Lbs./Yr. Plant 41
Manufacturing Cost - Air Oxidation vith Catalytic
Incineration 42
Sensitivity Analysis - Air Oxidation Process 43
Manufacturing Cost - Oxygen Oxidation vith Steam
Generation 44
Pro-Forma Balance Sheets 45
Estimated 1985 Air Emissions 47
R&D Project Costs 54
Summary of Emission Measurement Techniques
(3 pages) 57
Financial Impact of Air Pollution Episodes (2 pages) 65
-------
11
SUMMARY
The ethylene oxide industry has been studied to determine the extent of
air pollution resulting from the various plants and processes of the industry.
The purpose of the work was to provide the Environmental Protection Agency
with a portion of the basic data required in order to reach a decision on
the need to promulgate air emission standards for the industry.
It was concluded that the historical chlorohydrin process has been
completely superceded by two general types of direct oxidation processes
although some of the old chlorohydrin plants may have been converted to propylene
oxide production. There are distinct differences in the two types of oxidation
processes in that one process uses air as the source of oxidant, has nitrogen
as the major constituent of its vent gas, has a relatively smaller recycle
stream and utilizes a mole ratio of ethylene to oxygen in the main reactor
feed that is less than one. The other process uses relatively high purity
oxygen as the oxidant, has carbon dioxide as the major constituent of its
vent gas, has a relatively larger recycle stream and utilizes a mole ratio of
ethylene to oxygen in the main reactor feed that is greater than one. There
are several minor modifications to these basic processes, depending upon the
process licensor or designer. The most significant of these is one process
that incorporates the addition of methane into the recycle stream of an oxygen
oxidation process to serve as an "inert" at process conditions but to make the
vent gas combustible in a boiler house. All known processes use a fixed bed
silver catalyst in a tubular reactor with heat removal (steam generation) on
the exterior of the tubes.
In general terms, the air emissions from all variations of the direct
oxidation process are chiefly hydrocarbons (ethylene, ethylene oxide and
traces of ethane from the feed gas). Some plants report traces of NOx or SOx
from combustion operations associated either with process drive machinery or
pollution control. Some minor particulate emissions have been reported from
one scrubber operation but there is no explanation of this fact. No carbon
monoxide emissions have been reported even though most of the non-selectivity
of the process in an "over-oxidation" of ethylene to carbon dioxide and water.
This fact can only be assumed to be the result of the thernodynomics at reactor
conditions.
The producers of ethylene oxide that utilize air oxidation processes
report a trend toward the incorporation of a catalytic converter on the main
process vent. This results in the heating of the vent stream to a sufficiently
high temperature to have utility in the gas turbine driver of a process
compressor. Considering the fact that some plants presently employ these
devices and others do not, an estimate of the current (1973) emission factors
for air oxidation p' nts is 0.0287 lbs./lb. of ethylene oxide produced, of which
over 99 percent is hydrocarbons. For oxygen oxidation plants, the estimate of
emission factors is 0.0124 lbs./lb. of ethylene oxide produced unless methane
is added to the recycle and the vent gas burned. In this latter case, the
factor is reduced to 0.0038 lbs./lb. of ethylene oxide produced. In each of
the oxygen process estimates, data were received from only one plant and the
emissions are 100 percent hydrocarbons. A weighted average of these factors
results in an overall emission rate of 0.0206 lbs./lb. of ethylene oxide
produced (over 99 percent hydrocarbons). This amounted to nearly 90 million
pourds of pollutants emitted to the atmosphere in 1973 and would grow to about
140 million pounds per year by 1985 if the same averages were maintained
throughout the growth period.
-------
ii
SUMMARY (continued)
As already indicated, air process plants can reduce their emissions and
recover energy by means of existing technology and the trend is toward the
installation of these units. One of the oxygen oxidation plants reporting
does not have a pollution control device on its process vent even though the
stream has been estimated to be readily combustible. This is probably due to
two facts, namely (1) the stream is quite small so its energy content is
small and (2) the process is a net energy producer so this small increment is
not needed. The other oxygen oxidation plant that reported adds methane and
burns the entire stream in an off-site boiler, presumably to generate steam
for an off-site process. Hence, it can be concluded that oxygen oxidation
plants can reduce their emissions by one of several incineration techniques.
Whether or not they recover the available energy will be dictated by the
economics and process considerations in each instance. Assuming that all
present and future plants incinerate their vent gases by one of the techniques
discussed, 1985 emissions would be reduced to less than 35 million lbs./year.
This is an emission factor of 0.005 lbs./lb. of ethylene oxide produced of
which about 98 percent is hydrocarbon and about two percent oxides of nitrogen
(due to incineration processes).
Although none of the pollution control devices studied can be shown to
provide a net positive return on investment from the value of the energy
recovered, neither can any of them be deemed a "financial hardship1'. For a
200 million lbs./year air oxidation process, it has been estimated that about
$250,000 of capital equipment (1973 dollars) would be required to install
a catalytic incineration/gas turbine system on both the main process vent and
the CO2 purge vent. The net operating costs of these installations would
be about $7,400/year after taking credit for energy recovery. About 66
percent of the present industry utilizes the air oxidation process. Hence, by
1985 the total investment for the industry in 1973 dollars would be less than
six million if all plants installed these devices and the current ratios are
maintained. No data are available on the cost of boilers on oxygen oxidation
plants utilizing methane addition, but the cost must be only incremental. On
the other type of oxygen oxidation process (about five percent of the current
industry), the capital cost (1973 dollars) of a steam generator is about
$40,000 and of an incinerator about $16,000. The net operating costs of these
devices has been estimated to be $5,900 and $9,600 per year, respectively.
Thus, in the worst case, if all oxygen process plants installed a steam
generator, their 1985 total investment would be less than $0.5 million and
the total industry investment only slightly over $6 million. In addition,
by 1985, the industry could save the equivalent of three billion standard
cubic feet of natural gas per year by converting all of their air emissions
to utilizable energy.
In summary, the study has concluded that the industry can reduce emissions
and conserve fuel without significant financial hardship. Air oxidation
process plants can accomplish this by means of the demonstrated technology of
a catalytic converter and gas turbine on their main process vent. The
addition of this concept to the CO2 purge vent would further reduce emissions
although this latter aspect has not been demonstrated. Oxygen oxidation plants
should have little trouble incinerating their main process vent in standard
types of units although none were reported in use. Those plants that add
methane, by necessity, already provide for the incineration of the off-gas.
No need for control on the CO2 purge vent of oxygen oxidation plant is forseen
although a stack sampling program should be used to verify this conclusion.
-------
SUMMARY (continued)
It was also concluded that the major areas indicated for industry research
are catalyst improvement and/or inhibitor modification. Both of these programs
could lead to higher selectivies which would reduce the volume of CO2 being
emitted thereby simplifying the control techniques. These programs could
also lead to higher ethylene conversions thus reducing the amount of ethylene
that is emitted. There is little doubt that research of this nature is already
in progress in the laboratories of catalyst suppliers and process licensors.
-------
EO-1
I. Introduction
There are two processes for producing ethylene oxide that have been used
commercially. The first, the older of the two, is by reacting ethylene with
hypochlorous acid and dechlorinating the resultant chlorohydrin with lime to
form calcium chloride and ethylene oxide.
This is the chlorohydrin process, but all existing units have either been
shut down or converted to production of proplyene oxide.
The alternate process for producing ethylene oxide, and the process which
currently completely dominates the field, is direct oxidation of ethylene
either with air or 02.
(2) 2 H0CH2CH2C1 + Ca (0H)2
(1) CH2 = CH2 + H0C1
> hoch2ch2ci
y 2 CH2 - CH2 + CaCl2 + h2o
ch2 = ch2 + \ o2
Ag
Atmospheric emissions from the direct oxidation process are mainly ethylene
and related hydrocarbons. Relatively small quantities of SOx (from fuel) and
NOx (burners) are reported.
-------
EO-2
II. Process Description
In the direct oxidation process, oxygen is adsorbed on the surface of a
specific, silver containing catalyst. It then adds across the olefinic bond
of ethylene. Two routes are possible as shown below:
Oxygen or Air
r ^
.0 Adsorbed
L J
+
CH2 - CH2
Reaction A / \ Reaction B
Reaction C
ch2 — ch2 ~ co2 + h2o
No toj
The course of the reaction, either via route A or route B can be directed
by proper control of the temperature at the catalyst reaction sites. A brief
description of the thermodynamics involved is presented in Table EO-1.
Flow diagrams for the air and oxygen oxidation process are given in
Figures EO-1 and EO-2. The bulk of the respondents use the air oxidation
process and as a result, this report will emphasize that process although
comparative data for the 02 process will be given when available. A material
balance for a typical air oxidation plant is given in Table E0-2A. This
table was compiled from the respondents questionnaires and whatever yield,
conversion and selectivity data could be gleaned from the literature. As a
result, no one particular plant will actually fit this material balance
exactly. It is rather an idealized representation of an ethylene oxide plant
which should be close to representing any particular air oxidation plant.
Yields in terms of a typical 200 MM Lbs./Year ethylene oxide plant are also
shown on Table E0-2A.
Ethylene (95-99% pure) and air are fed separately into a recycle gas
stream which then feeds a bank of primary (main) reactors which are operated
in parallel. The air/ethylene feed ratio, usually about 10:1, by weight, is
varied so that after dilution with recycle gas, an optimum oxygen/ethylene
ratio results. Reaction takes place over a silver catalyst packed in tubes
in a reactor and surrounded by a heat transfer fluid to control temperature.
The gas stream moves downward over the catalyst and counter-current to
circulation of the heat transfer fluid in the reactor shell. A portion of
the fluid (Dowtherm, tetralin or other high boiling materials) is vaporized
by the exothermic reaction heat and is condensed in a heat exchanger to
provide considerable steam for the ethylene oxide and other processes. Ethylene
conversion in the primary reactors is maintained at about 50 percent per pass
in crder to insure selectivities of 60 percent or higher. Oxidation inhibitors,
such as ethylene dichloride, are added to the inlet gas in PPM concentrations
to reduce undesirable CO2 formation.
-------
EO-3
TABLE EO-I
HEAT BALANCE
Two main reaction routes are possible when ethylene is oxidized in the
presence of a silver containing catalyst.
A. CH2 = CH2 + h 02 CH2-CH2 + 1,615 BTU/Lb.* ethylene reacted
B. CH2 = CH2 + 3 02 ~Y 2 C02 + 2 H O + 21,790 BUT/Lb.* ethylene reacted
Reaction B liberates over 13 times as much energy as the desired reaction
A. Success of a commerical installation depends on the proper control of the
temperature on the catalyst surface to favor reaction A.
For example, the effect of temperature on selectivity and heat release has
been reported as follows:
Selectivity, % 70 60 50 40
Total Heat Release_ 7.67 9.69 11.70 13.72
(1000's BTU/Lb. C2 converted)
A 50 million lb./year ethylene oxide plant would release 40 million BTU/hr.
of heat at 70 percent selectivity (optimum). If improper control lowered
selectivity to 50 percent, the heat release would more than double to 98 million
BTU/hr.
*Gross Heating Value.
-------
UA.IU
UAI U
PUZUtr
itbozeitt
W&TErS.
me-at txcHAUufre
A TO VcUT
T C-Oj , W^O, ^7 i O2
J POWfirtt. UUIT
£kU<*be.
UOTE-5'.
=W= M
. OOTTtO UUEr«> ^HOW OOTIOUA.L U^t OF-
C-AT^tVTiC couvtzrte WUICH tOKEr
p£OCfr^«boe*>
2 IF- UUe.frfHUfeD tTMYLfcUt OViOfc- Dfr£»l8fr3 f-GU
MVDeOLV6i*> TO tTMVLfeUEr 6LyCOL,P^lT OK
&.LL OP- THt Dtt>oee>be OVbUH&AO WOU
ao DlfctC-TLY TO TWfc 6LVCOL"b PLiUT
VUA.IU PZOCt-t* VfrUT tfe THfe L^Z^frK. OF- THt
I^OO f-
tTMVLfeUfr OXlDfc- iOLUTlOU
TvJ© Vcr UT t>TZEr^.Ut>.
4.LiUMbtZtP bTZfc-A>^"p ikfct CtOVb LtP&HtUCtO
TO TA.feLfc : E-O-l A,
C-OUDttJ^tK
g-fe-p-LLiy
£kC.C-UUULA.TOQ.
gfe-P-LU* I
A.CCOWULA.TOaV
PUegfr Afe^QgfefcE.
F-lUI^H
PU£lF-l£.*T 10
Llg»MT6
£&- VJOVAL
=#= M
///A
cowPit«oi
e.Ert>OILta.
H&DS.&.WU L e&Ot-6lUlJ&-0 ¦ V*VT*
u&wy e-uot.
tvJVidOUUt-iJTA.L P^OTfcCTlOU A^frUCY
COUTIAC.T U* <*&• 02-02«7*7
Mb AT
rtfCMAUute
^lEMlEO IT
CATALYTIC. IHC.
eMOlMEEKIHG 0£PA*TM£*T. PMI UDElPKI ». PIMU
UOUfc
4, 27-7}
CMfCU
It* H
CoPTRiCT
42Q»>0
2.-247
Cut"
APP80VU
7-«"?V
-------
TABLE HO-2A
TYPICAL MATERIAL BALANCE (2>
AIR OXIDATION PROCESS
1
2
3
4
5
6
7
8
9
10
11
12
13
Pounds/Hour ^
Fresh
Feed
Recycle
Cross
Reactor
Feed
Reactor
Effluent
Absorber
Bottons
Absorber
Overhead
Purge
Reactor
Feed
Purge
Reactor
Effluent
Purge
Absorber
Bottoms
Main
Proc_ess
Vent
Recovery
Unit
Feed
co2
Purge
Vent
Ethylene
Oxide
Product
Nitrogen
124,508
266,603
391,111
391,111
335
390,776
124,173
124,173
49
124,124
384
384
0
Oxygen
37,823
18,954
56,777
27,824
42
27,782
8,828
4,712
6
4,706
48
48
0
Ethane
150
322
472
472
0
472
150
150
0
150
0
0
0
Ethylene
25,342
11,145
36,487
16,419
83
16,336
5,191
2,336
12
2,324
95
95
0 .
Ethylene
Oxide
0
0
0
21,921
21,921
0
0
3,119
3,097
22
2S.Q18
18
25,000 i
Carbon I
ioxlde
0
36,652
36,652
55,885
2,162
53,723
17,071
19,80S
313
19,492
2,475
2,473
2 ;
Pounds/I
ound SCO Product
i
Nitrogei
4.980
10.664
15.644
15.644
0.013
15.631
4.967
4.967
0.002
4.965
0.015
0.015
or.o
Oxygen
1.513
0.758
2.271
1.113
0.002
1.111
0.353
0.188
0.0002
0.188
0.002
0.002
0.0
Ethane
0.006
0.013
0.019
0.019
0.0
0.019
0.006
0.006
0.0
0.006
0.0
0.0
0.0
Ethylend
1.014
0.446
1.460
0.657
0.003
0.654
0.208
0.093
0.001
0.092
0.004
o.oo4 :
0.0 j
Ethylene
Oxide
0.0
0.0
0.0
0.877
0.877
0.0
0.0
0.125
0.124
0.001
1.001
0.001
l. 00(0
Carbon ijloxide
0.0
1.466
1.466
2.235
0.086
2.149
0.683
0.792
0.013
0.779
0.099
0.099
MIL |
NOTE: {lumbers over columns refer to streams on Figure BO-1.
(1) Band oq 8,000 hours per year operation.
(2) Dry basis.
-------
EO-6
The effluent gas from the primary reactor is cooled and compressed before
entering the primary absorber which is generally operated with cold water.
(If this absorber is operated with dilute sulfuric acid, recovery and
hydrolysis to ethylene glycol take place simultaneously. No respondent indicated
use of this system, however.) Absorbers using water are usually packed columns
about 60 feet high. Feed to the absorber averages about 100° F. Ethylene oxide
(the feed stream is about 1-2 mol °L of ethylene oxide, 2-37o ethylene, plus
CO2, N2, O2 and A) in the effluent dissolves together with some CO2 in the
water. The aqueous stream is removed from the base of the absorber. Unabsorbed
gas passing overhead is split into two portions. The larger portion of the
gas stream is heat exchanged to raise its temperature and serves as the
main stream for the secondary (purge) ethylene conversion system. Effluent
from the secondary reactor is cooled and enters the purge absorber where
ethylene oxide is removed from the stream with water. The overhead from
this absorber, the main process vent, is oxide free and low in residual ethylene
and is vented directly to the air or to a catalytic converter where the
residual hydrocarbons are burned to heat the gas stream and the hot gas stream
is used to drive the feed turbines for the process.
Dilute water solutions (containing EtO, CO2 and traces of hydrocarbons)
from both absorbers are combined. The mixture is fed to the top section of a
bubble plate column where the absorbate is steam stripped under vacuum
(desorbtion). Ethylene oxide is distilled off the top and is compressed for
rectification. The bottoms water is recycled to the absorbers. Rectification
of the ethylene oxide in the first distillation column removes the CO2 and
inert gases overhead (CO2 rich purge gas). Ethylene oxide from the bottom
of the column then feeds to the middle of a second distillation column for
refining to 99+"% purity. Product is stored under nitrogen or in refrigerated
storage. Bottoms from the second column consist of water with traces of
hydrocarbons. This stream is normally sent to in plant sewage treatment where
it is biologically treated before release.
A material balance for a typical oxygen oxidation process is given in
Table E0-2B. The process flow is very similar to the air oxidation process,
except that there is usually only a primary reactor and absorber. Also, since
the conversion of ethylene per pass is low (of the order of 10-15 percent), the
recycle is larger than in the air process. Product recovery is similar on both
processes but with the oxygen process a CO2 absorber is required on a portion
of the recycle stream to control the build up of CO2. One modification that
is practiced by at least one domestic operator of an oxygen oxidation process
is to add methane to the recycle. This acts as an "inert" at the conditions
of the process reactor but results in a vent gas that is suitable as boiler
fuel, thus eliminating the need to vent residual traces of ethylene from the
main" process vent.
-------
COi
i i VtkjT
EE-gfc-Kjfrz.M'og
COOLA.KJT
UA,I U
Afc»bQg.frfc-g.
fc-rfCHMJ6ifc-E
fcTl-JYLfcUc-
COj
A,fc!iO«.e»6-UT
,qm Pgbb^O fc
•bPe^T
< r ^fcboe&ErkjT
HEr^T
PfcOCfrSb v/fc-UT •<¦
4DD l T lOVJ ALT tlUAT IVfr)
CM4 A.LTfre.k/A.r i vt
kJOTErS:
l . I U Twi
LA.eA/© V6-UT
2. klu^fcteto A,Z.fc teoW-ZtPfrlfeUCfr
IDtUTiTifrb POl Wi.TttiL.L blLAUCt T^bLb: &0-2&
THfc
tTHYLfr kit Otf i Dfc-
4QUtOUt>
fc-TO
t&^e»bo£6&e
Pfc- SQtbfrZ
ttoeAWkJ 4 ZfrVlbtD
ErU VieOkJMtO PeoTfcCTlOM At»6-ucy
C.OkjTl.A.CT (J* iJ>e>- 02 ¦ 02^
F-i^UZ-fc- fcO-2
6iMPliP-I&*D P-LOW DlA^e^U
tTMYLtUt- OXlOt VIA, TMfr OXYtifcKj
OX'PATlO'yi PgOCfcbb
-------
table E0-2B
TYPICAL MATERIAL BALANCE (1)
OXYGEN OXIDATION PROCESS
200 MM LBS./YEAR (2) ETHYLANE OXIDE
Pounds/Hour
Nitrogen, Argon
Oxygen
Methane
Ethane
Ethylene
Ethylene Oxide
Carbon Dioxide
Fresh
Feed
115
31,148
4
95
22,417
0
0
Recyclc
213,837
78,097
7,438
143,178
126,443
74
698,155
Gross
Reac tor
Feed
213,952
109,245
7,442
143,273
148,860
74
699,155
Reactor
Effluent
213,952
78,142
7,442
143,273
126,531
25,074
719,333
Absorber
Overhead
213,952
78,108
7,385
143,194
126,359
74
712,422
Absorber
Bot t oms
0
34
57
79
172
25,000
6,911
Pounds/Pound EtO
Nitrogen, Argon 0.005 8.553 8.558 8.558 8.558 0.0
Oxygen 1.246 3.124 4.370 3.126 3.125 0.001
Methane 0.001 0.297 0.298 0.298 0.296 0.002
Ethane 0.004 5.727 5.731 5.731 5.728 0.003
Ethylene 0.897 5.058 5.955 5.062 5.055 0.00/
Ethylene Oxide 0.0 0.003 0.003 1.003 0.003 1.000
Carbon Dioxide 0.0 27.966 27.966 28.773 28.497 0.276
NOTE: Numbers over columns refer to streams on Figure E0-2
(1) Based on single questionnaire, dry basis.
(2) Based on 8,000 hours per year operation.
(3) Adjusted, from a single questionnaire, to include future changes as indicated on the questionnaire.
CO 7
Purge
Vent (3)
0
3
0
18
20
0
19,800
0.0
Nil
0.0
0.001
0.001
0.0
0.792
Gross
Recyc1c
213,952
78,139
7,442
143,255
126,511
74
699,531
8.558
3.126
0.2 98
5.730
5.061
0.003
27.981
Ma i n
Process
Vent
115
42
4
77
68
0.04
3 76
0.005
0.002
0.001
0.002
0.003
Ni 1
0.015
10
Ethylene
Oxide
Product
0
0
0
0
0
25,000
2
0.0
0.0
0.0
0.0
0.0
1.0
Ni 1
-------
EO-9
III. Producers of Ethylene Oxide and Emissions
The capacities and plant locations listed below are based on information
provided in the questionnaires and in the literature.
Producer
Location
Capacity -
Process (1)
MM Lbs./Yr.
BASF-Wyandotte
Geismar, La.
220
Shell
Calcasieu Chemical
Lake Charles, La.
165
Shell
Celanese
Bayport, Texas
300
Shell
Dow
Freeport, Texas
1,000
Dow (2)
Plaquemine, La. J
Eastman Chem. Prod.
Longview, Texas
50
Shell
Jefferson Chem. Co.
Port Neches, Texas
500
Scientific Design
Koch Chemical
Orange, Texas
36
Scientific Design
Northern Petrochemical
Joliet, 111.
200
Scientific Design
Olin Corporation
Brandenburg, Ky.
100
Shell
PPG - Houston Chem.
Beaumont, Texas
80
Scientific Design
Shell Chemical
Geismar, La.
300
Shell
Sun-Olin
Claymont, Del.
90
Shel 1
Union Carbide
Seadrift, Texas
700
Union Carbide (2)
Taft, La.
450
Union Carbide (2)
4,191
(1) All plants use
the direct oxidation of
ethylene to
ethylene oxide.
(2) Air oxidation.
Table E0-3 shows individual plant capacity
figures and
atmospheric emission
data for the various ethylene oxide plants surveyed in the study. About 53
percent of the installed ethylene oxide capacity in the U.S. has been covered
by this survey. Emissions from these plants are as follows:
A. Continuous Air Emissions
1. Main Process Vent Gas
This stream, which vents from the secondary or purge reactor
absorber consists of "spent" air (N£, O2 and some inert gases), some
CO2, small to nil amounts of EtO and usually less than two percent
hydrocarbons in an air oxidation plant. The similar stream from an
O2 oxidation plant has entirely different composition. Typical main
process vent gases from both types of plants are shown on Tables E0-
4A and E0-4B. This stream constitutes the largest emission of air
pollutants from the process. Although dilute in concentration, the
streams are large in volume and contribute over 50 percent of the
total emissions.
2. CO2 Rich Purge Gas
This is the overhead from the ethylene oxide rectification tower
in the case of the air plants and the CO2 system vent stream in the
O2 plants. The streams are not comparable in the two processes.
Typical vent gas composition for this stream is also shown on Tables
E0-4A and E0-4B.
-------
Plant EpA Code No.
Capacity - Tons EtO/Yr.
Range of Production - % of Max.
Emissions to Atmosphere
Stream
Flov - Lbs./Hr.
Flov Characteristic - Continuous or Intermittent
if Intermittent - Hrs./Yr.
Composition - Ton/Ton EtO
Oxygen
Nitrogen, Argon
co2
Steam
N°x
Ct - C4 & Higher Hydrocarbons
Particulates
Vent Stacks
Number
Height - Feet
Diameter - Inches
Exit Gas Temperature - °F
SCFM/Stack
Emission Control Devices
Analysis
Date or Frequency of Sampling
Tap Location
Type of Analysis
Odor Problem
Summary of Air Pollutants
Hydrocarbons - Ton/Ton EtO
Aerosols - Ton/Ton EtO
N0X - Ton/Ton EtO
S0X - Ton/Ton EtO
CO - Ton/Ton EtO
Type of Process
TABLE EO-3
NATIONAL EMISSIONS INVENTORY
ETHYLENE OXIDE BY DIRECT OXIDATION
13-1
350,000
Absorber
Vaste Gas
116,35?
Continuous
1.704736
0.04 9824
I
50
24
400
23,167
Continuous for EtO
Easy, Automatic
GC. >6
No
Stripper
Tail Gas
1,096
Cont i nuous
0.000154
0.001512
0.014624
0.000739
1
52
2
110
164
Once per year
Eatty
GC,
No
0.036117
Trace*
Air Oxidation
page 1 of 5
Reabsorbcr
Vent Cay
3,671
Conr i nuouF
0.000700
0. 137328
0,002368
0.006444
95
579
None
Dai ly
Easy, in line
GC
No
13-2
100.000
Vater Scrubber
Venr Cap
682
Cont i nuouF
0.001660
0,004582
0.015036
0,000043
0.00^952
114
None
Never
No tcjp
None
No
0.012403
*Fuel
-------
TABLE EO-3
NATIONAL EMISSIONS INVKXTORY
ETHYLENE OXIDE PY DIRECT OXIDATION
Plant EPA Code No,
Capacity - Tons EtO/Yr.
Range of Production - 7. of Max.
Emissions to Atmosphere
Stream
Flov - Lb6./Hr.
Flov Characteristic - Continuous or Intermittent
if Intermittent. - Hrs./Yr.
Composition - Ton/Ton EtO
Oxygen
Nitrogen, Argon
C02
Steam
N0X
Ci - C4 & Higher Hydrocarbons
Particulates
Vent Stacks
Number
Height - Feet
Diameter - Inches
Exit Gas Temperature - °F
SCFM/Stack
Emission Control Devices
Analysis
Date or Frequency of Sampling
Tap Location
Type of Analysis
Odor Problem
Summary of Air Pollutants
Hydrocarbons - Ton/Ton EtO
Aerosols - Ton/Ton EtO
N0X - Ton/Ton EtO
SOx - Ton/Ton EtO
CO - Ton/Ton EtO
Type of Process
13-3
40.000
Gas Turbine
Exhaust
87,223
Cont inuous
0.52380
6.9775
1.0871
0.13390
1
57
48
450
18,800
Never
Di ffi cult
Reabsorber
Vent Gas
1,239
Cont i nuous
0.00262
0.020590
0.094750
0.000360
0.005574
0.000053
1
150
4
110
200
2 per day
Easy
Yes
0.005574
Air Oxidation
Reactor
Vent
475,000
Cont inuous
0.045009
7.04927
1.23751
0.000517
0.001270
1
50
6' x 8'
440
100.000
1 per hour
Easy
c;c
No
*Fuel gas.
+Includes .000538 fugitive C2~ losses.
Pape 2 of 5
13-7
225 000
+4%
Recovery System
Vent Scrubber
Reactor
Vent
Heater
Vent
5.000
Cont inuous
52 . 5
1 6
Con (. i :uiou
0.002362
0.019858
0.080533
0.000031
0 000755
0.000100
0.008046
0 00004 7
0.00032
1
45
4
95
1.050
3 per year
Bomb - difficult
MS
No
Cont i nuous
Ea«v
CC MS
No
None
None
Vt Balance
No
o»no«/.n
0.000517
0 ,000229"
Air Oxidation
O2 Oxidation
-------
plant EPA Code No.
Capacity - Tons EtO/Yr.
Range of Production - 70 of Max.
Emissions to Atmosphere
Stream
Flov - Lbs./Hr.
Flov Characteristic - Continuous or Intermittent
if Intermittent - Hrs./Yr.
Composition - Ton/Ton EtO
Oxygen
Nitrogen, Argon
COj
Steam
N0X
Ct - c4 & Higher Hydrocarbons
particulates
Vent Stacks
Number
Height - Feet
Diameter - Inches
Exit Gas Temperature - °?
SCFM/Stack
Emission Control Devices
Analysis
Date or Frequency of Sampling
Tap Location
Type of Analysis
Odor Problem
Summary of Air Pollutants
Hydrocarbons - Ton/Ton EtO
Aerosols - Ton/Ton EtO
N0X - Ton/Ton EtO
SO* - Ton/Ton EtO
CO - Ton/Ton EtO
Type of Process
TABLE EQ-3
NATIONAL EMISSIONS INVENTORY
ETHYLENE OXIDE BY DIRECT OXIDATION page 3 of 5
13-5
100,000
Gas Turbine
Exhaust
Recovery
Section Vent
Absorber
Cap
76,000
Continuous
3.100
Continuous
5 . 500
Continuous
0.01578
2 2660
0.75550
0.001164
0.01385
0.10680
0.010748
0 17920
0 02 6284
0 002764
0.002155
0 003768
1
45
30
610
16,550
1
35
4
100
480
2 per month 2 per
Easy, use Orsat bomb Easy,
GC GC
No No
month (but EtO monitor continuously)
automatic
Certain components continuously
Easy
GC
No
0.008687
Air Oxidation
-------
Plant EPA Code No.
Capacity - Tons EtO/Yr.
Range of Production - 7. of Max.
Emissions to Atmosphere
Stream
Flov - Lbs./Hr.
Flov Characteristic - Continuous or Intermittent
if Intermittent - Hrs./Yr.
Composition - Ton/Ton EtO
Oxygen
Nitrogen, Argon
COj
Steam
N0X
Cj - C4 6* Higher Hydrocarbons
sox
Vent Stacks
Number
Height - Feet
Diameter - Inches
Exit Gas Temperature - F<>
SCFM/Stack
Emission Control Devices
Analysis
Date or Frequency of Sampling
Tap Location
Type of Analysis
Odor Problem
Summary of Air Pollutants
Hydrocarbons - Ton/Ton EtO
Aerosols - Ton/Ton EtO
N0X - Ton/Ton EtO
S0X - Ton/Ton EtO
CO - Ton/Ton EtO
Type of Process
TABLE EO-3
NATIONAL EMISSIONS INVENTORY
ETHYLENE OXIDE BY DIRECT OXIDATION pa>,e 4 of 5
13-6
18.000
Cycle Gas Mix Gas Exhaust Heater and
Exhaust From Scrubber Furl Vent:?
0. 35474
5.84960
0.91971
0.35187-
0.06297
0.001728
0.02219
0.094720
0.00062 5
0.003819
0.001301
Q. 000004
Occasions 1
Ea sy
Orsat, CH
No
0,068090
0 000004
Once per month
Ea fv
Orsat. CH.
No
Air Oxidation
-------
EO-14
TABLE EO-3
LIST OF ABBREVIATIONS IN ANALYSES Sheet 5 of 5
GC - Gas Chromatograph
MS - Mass Spectroscope
Orsat - Conventional orsat for CO2, O2, C2H4
CH - Ethylene oxide by reaction vith MgCl2 to form chlorohydrin
-------
EO-15
TABLE E0-4A
TYPICAL VENT GAS COMPOSITIONS
FOR
200 MM LBS./YR. ETHYLENE OXIDE PRODUCTION
USING AIR
OXIDATION
MAIN PROCESS VE'MT
GAS
Reported
Range in
Average
Flow Rate
Component
Avg. Mol 7»
Composit ion
Mol/Hr.
Lbs./Hr
Nitrogen
86.7
80 - 90
4,695
131,460
Oxygen
2.9
0.5 - 4.5
157
5,024
Methane
0.0
0 - 0.9
0
0
Ethane
0.1
0 - 0.1
5
150
Ethylene
1.6
TR - 2.3
87
2,436
Ethylene Oxide
0.01
0 - 0.01
0.4
17
Carbon Dioxide
8.7
0-10
471
19,782
100.00
5,415.4
158,869
32,402 SCFM
29.3 Avg. Mol Wt
•
CO2 RICH PURGE GAS
Reported
Range in
Average
Flow Rate
Component
Avg. Mol %
Compos i tion
Mol/Hr.
Lb s . / Hr
Ni trogen
18
13 - 25
14.2
398
Oxygen
2
1 - 26
1.6
51
Ethylene
4.5
2.5 - 8.0
3.6
101
Ethylene Oxide
0.5
0 - 1.0
0.4
17
Carbon Dioxide
75
62 - 80
59.3
2,491
100.0
79.0
3,058
475 SCFM
38.7 Avg. Mol Wt.
NOTE:
Both streams on a water free basis.
-------
EO-16
TABLE E0-4B
TYPICAL VENT GAS COMPOSITIONS
FOR
200 MM LBS./YR. ETHYLENE OXIDE PRODUCTION
USING OXYGEN FEED
MAIN PROCESS VENT GAS (1,3)
Flow Rate
Component
Mol 7=
Mo 1 / Hr .
Lbs./Hr.
Nitrogen, Argon (MW = 39.2)
16.2
3.09
121
Oxygen
7.3
1.39
44
Methane
1.5
0.29
5
Ethane
14.2
2.70
81
Ethylene
13.5
2.57
72
Ethylene Oxide
0.005
0.001
0.04
Carbon Dioxide
47.4
9.02
379
100.0
19.06
702
114 SCFM
36.8 Avg. Mol Wt
•
C02 RICH PURGE GAS (1,2,3)
Flow Rate
Component
Mo 1 %
Mo 1 / Hr .
Lbs./Hr.
Oxygen
0.02
0.1
3
Ethane
0.12
0.6
18
Ethylene
0.16
0.8
22
Carbon Dioxide
99.70
500.9
21,038
100.00
502.4
21,081
3,006 SCFM
42.0 Avg. Mol Wt
(1) Based on one questionnaire - other oxygen process does not have any
appreciable vent other than CO2 and water (one questionnaire). Methane
is added after the reactors to build up a background of inert gas in the
recycle stream at reactor conditions but a combustible gas at boiler
conditions.
(2) Composition and quantity adjusted to include future changes as indicated
on the questionnaire.
(3) Both streams on a water free basis.
-------
EO-17
3. Turbine Exhaust
S
Many, if not all of the plants surveyed use natural gas fueled
turbines to feed air and ethylene to the process. Since the turbines
which drive the compressors are not 100 percent efficient, some
unburned hydrocarbons escape to the air from this exhaust. It can
be a considerable portion of the total hydrocarbon emissions. A
large variation in this stream has been reported (Table EO-3).
B. Intermittent Air Emissions
No data given although some plants report flaring or venting gas
during upsets. One operator reported emptying the system to the flare
one or two times per year for about three hours each time.
C. Liquid Wastes
Mainly water, with small percentage of organics, all of which
appear to be adequately treated by conventional in-plant biologic
systems for most respondents. One respondent reported inorganic
salts (presumably from the CO2 recovery section) in this stream.
D. Solid Wastes
None reported although one respondent alludes to the presence of
about 0.004 lbs. of "sludge"/lb. EtO leaving the plant as part of the
waste water stream. No further data were given.
E. Odors
Not one respondent reported any odor problem on any stream. The
only reference to odor was the indication that, if there were an odor
detectable, it would be ethylene. Actually this response is not
surprising because:
1. The hydrocarbon content of the vast majority of all the vent
streams is quite low.
2. Ethylene is a colorless sweet smelling gas. In low con-
centration, i.e., less than two percent, it is doubtful if
one could detect an odor directly in the vent stream. As the
vent gas mixes with the surrounding air, and the ethylene
becomes more dilute, the possibility of an odor problem
fades away. The absence of any detectable odor does not
mean that there are no hydrocarbon emissions.
Ethylene oxide on the other hand has a pungent irritating odor.
No one even so much as indicated an ethylene oxide odor problem.
Again, not surprising. Ethylene oxide is the desired product and
the odor of EtO means loss of a valuable commodity. To allow
continuous loss of EtO would just not make good economic sense and
any odor problem associated with ethylene oxide would be temporary
and most probably due to a severe upset or equipment failure.
-------
EO-18
Emission Control Devices and Systems
A. Emission Control Devices on Main Process Vent
1. Absorbers and Scrubbers
The main ethylene oxide absorber vent gas on all plants is
recycled to the process. The secondary or purge absorber vent gas
is normally the main process vent for the air oxidation plants
and the oxygen feed plants. Two air oxidation process respondents
catalytically convert this dilute hydrocarbon stream to N2, CO2, O2
and H2O and drive process tubines with the hot compressed gas. The
other five respondents vent this stream to the air.
One of these latter plants (also air oxidation) supplied data on
their purge absorber and it has been included on Table EO-5 for the
sake of completeness. Actually all plants have this absorber-scrubber
but only one gave data on it. Strictly speaking, it is a necessary
process control device. It is identified as 002 on Table EO-5. Water
is the absorbing medium.
Ethylene oxide absorbers/scrubbers do an excellent job in removing
EtO from the off-gas stream. Efficiencies of 99.9+% are not uncommon.
Ethylene oxide is very soluble in water and a well designed scrubber
should perform well. However, a water scrubber will not remove any
appreciable ethane or ethylene from the vent gas and as such this
represents the major source of air pollution for the process. Unless
this stream is catalytically combusted to CO2 and water, it will
contribute substantial quantities of hydrocarbons to the air in a
stream generally too dilute to burn by itself.
Vent gas scrubbers using water to recover ethylene oxide are
used by two respondents (see Table EO-5) on the CO2 purge stream of
air oxidation processes. Here again the efficiency of ethylene oxide
recovery is good but efficiency is poor in terms of total hydrocarbon
emissions. This CO2 purge stream is the smaller of the two main
vent streams.
2. Catalytic Conversion Units
These units are designed to convert the ethylene content (about two
percent) of the absorber vent gas to CO2 and water and to heat the gas
stream at 100-200 PSIG from ambient temperature up to 1300° F. The gas
is then used to drive a gas turbine and exhausts at about 600° F to a
stack. Data on these units came from two air oxidation processes.
A typical catalyst would consist of activated precious metals
(probably platinum) uniformly coated on a matting of high nickel
chrome alloy metal ribbins. Self supporting catalyst mats are mounted
in a hollow cylindrical configuration. The catalyst tube is surrounded
by a cylindrical baffle that directs the incoming gas to the top of the
vessel where gas burners preheat the gas to reaction temperature. The
heated gas then passes over the catalyst where the ethylene reacts with
the oxygen in the gas and heats the mass of the gas to about 1300° F.
The converter outlet temperature controls the firing of the preheat
burners.
-------
TABLE KO-5
CATALOG OF EMISSION CONTROL DEVICES
Page
1
or 2
ABS ORBERS/S CRUBBERS
EPA Code No. for plant using
Flov Diagram (Fig. No) Stream I. D.
Device I. D. No.
Control Emission of
Scrubbing/Absorbing liquid
Type - Spray
Packed Column
Tray Column
Scrubbing/Absorbing Liquid Rate - GPM
Design Temperature (Operating Temperature) °F
Gas Rate - SCFM (lb./hr.)
T-T Height, Feet
Diameter - Feet
Washed Gases to Stack
Stack Height - Feet
Stack Diameter - Inches
Installed Cost - Mat'l. & I.abor - $
Installed Cost Based on "year" - $
Installed Cost - c/lb.
Operating Cost - Annual - $ (1972)
Value of Recovered Product - $/Yr.
Net Operating Cost - c/lb.
Efficiency - % - SE
Efficiency - % - SERR
13.6
(EO-l)-lO
002
Ethylene Oxide
Vater
13. 6
(KG-1)- 10
001
Ethylene Oxide
Water
13 7
(EO-L)-IO
101
Ethylene Oxide
Va t e r
X X
X
30 80 150
100 200 Ambient
300 6500 1167
32 57 5 40
2 3.5 3.5
Yes Yes Yep
62 38 45
3 8 4
Unknown Unknown ?4,000
1968
0.0053
Not Available Not Available $2500
$225,000**
(Credit - 0-049)
99.99 98.5 95.3
97.05* 5 6.9"* 66.?*
*For total hydrocarbons.
**For recovered EtO.
-------
TABLE EO-5
CATALOG OF EMISSION CONTROL DEVICES (1)
Page 2 or 2
INCINERATION DEVICES
EPA Code No. for plant using
Flow Diagram (Fig, No.) Stream I.
Device I. D. No.
Type of Compound Incinerated
Type of Device
Material Incinerated, Lb./Hr
Auxiliary Fuel Req'd. (excl.
Type
Rate - BTU/Hr.
Device or Stack Height - Feet
Installed Cost - Mat'l.
Installed Cost Based on
- c/lb.
D.
(SCFM)
pilot)
& Labor -
"year" -
Installed Cost
Operating Cost - Annual - $/Yr.
Operating Cost - c/lb.
Efficiency - X - CCR
Efficiency - % - SERR
13-3
(EO-l)-lO
101
Hydrocarbons
Catalytic Converter
(20,000)
57
$74,000
1964
0.0925
(Credit
(Credit
100
100
10°
$20,000)
0.025)
13-4
(FO-2)-9
Hydrocarbons
Flare Stack
Normally Zero
125
100 (2)
100 (2)
13-5
(EL)- 1 ) — 10
101
Hydrocarbons
Catalytic Converter
7 6.000
Natural (:as
3 x 106
45
$300,000
1966
0.15
$42.000
0.021
94. 7
94. 7
(1) Plant 13-7 also reported (subsequent to their completed questionnaire) a catalytic converter on the reactor vent shown in Tabic EO-3. However, they did not give
cost or perfromance data on the unit.
(2) Flare shown as burning completely to CO2 and H2O.
-------
EO-21
The above is a description of one type of catalytic converter.
There are other designs but they follow the general pattern described.
In some case, it may be necessary to inject small quantities (roughly
0.4 mol %) of methane or natural gas plus some air in order to insure
complete combustion in the converter. These units are rather large,
seven to eight feet in diameter and 15 to 20 feet long.
Table EO-6 gives a material balance for catalytic incineration of
a lean hydrocarbon vent stream. Section VII covers the cost effective-
ness of this control. Actual cost data for catalytic converters are
presented, as given by the respondents.
3. Other Combustion Devices
For those vent streams which are readily combustible, a steam boiler
could be used to convert the hydrocarbon content to CC>2 and water and
recover some of the waste heat as steam. Since ethylene oxide plants
are net steam producers due to the exothermic heat of reaction, it may
be uneconomical to run yet another steam boiler on a waste stream and
have to transport the steam from the process area to where it can be
used. In this case, a thermal incinerator could convert the vent gas
stream to CC>2 and water and eliminate emissions of hydrocarbons to
the air.
Only one respondent indicates use of the main process vent stream
in a steam boiler but no operating data were given. This, of course,
is the case where methane is added to the recycle. No one indicated
use of a thermal incinerator so it is difficult to accurately predict
equipment performance in these two applications. Table EO-7 and EO-8
give material and heat balance data for combustion of a fairly rich
hydrocarbon vent stream. No data are included for thermal incineration
or steam generation of a lean hydrocarbon stream (below the lower
flammability limit) because such treatment is impractical and
uneconomic.
4. Flare System
One respondent (oxygen oxidation process) reports on the use of a
flare system (only during emergencies) and claims 100 percent efficiency.
It was not reported whether or not supplemental fuel was used. Generally,
a "flare system" is understood to mean a combustion stack that does not
require supplementary fuel except for pilot lights. On the other hand,
a plume burner is one which is burning a gas that is normally close to
or below its lower flammable limit and thus requires a continuous
supply of supplemental fuel. It is normally assumed that neither a
flare stack nor a plume burner will achieve more than about 90 percent
combustion efficiency.
The oxygen oxidation vent gas shown in Table E0-7 will support
combustion and could be flared. However, it is unlikely that combustion
would be complete. It is more likely that about 10 percent of the
heating value will be lost in the form of carbon monoxide and unburned
or cracked hydrocarbons. If the plant has an existing flare from
another process, additional supplemental fuel might not be required,
depending upon the heating value of total flared gases. In any case,
however, improper firing could result in NOx formation.
-------
EO-22
TABLE EO-6
AIR FEED PLANT
CATALYTIC INCINERATION (1)
MATERIAL BALANCE, LB./HR.
Component
Process Vent Gas (2)
Combustion Air
Flue Gas (3)
n2
131,858
14,663
131,858
°2
5,075
4,457
212
ch4
0
0
TR
C2H6
150
0
TR
C2H4
2,537
0
TR
EtO
34
0
0
CM
O
CJ
22,273
0
30,754
NOx (4)
6.4
0
6.4
h2o
0
3,560
161,933
19,120
181,053
(1) Combined combustibles are 1.67 percent by volume vs. 2.7 percent lean lower
limit for ethylene in air. Catalytic incineration is necessary to hold
combustion chamber temperature in the 1300° to 1500° range.
(2) Combined process stream, main vent plus CO2 vent, water free basis.
(3) Unburned hydrocarbons reported as "trace" in flue gas. One respondent
reported 200 to 300 ppm in flue gas but his vent gas stream differed from
that shown above. Flue gas has 0.5 percent excess O2 as per respondnet
experience with catalytic converters.
(4) 40 ppm.
-------
EO-23
TABLE EO-7
OXYGEN FEED PLANT, STEAM GENERATION
OVERALL MATERIAL BALANCE, LB./HR.
Component
Process Vent Gas (2)
Combustion Air (1)
Flue Gas
n2, a
121
2,102
2,223
°2
44
639
114
CH4
5
0
TR
c2h6
81
0
TR
c2h4
72
0
TR
EtO
0.04
0
TR
co2
379
0
827
NOx
0
0
TR
h2o
———
0
250
702
2,741
3,444
(1) 20 percent excess O2.
(2) Water free basis combined combustibles are 27.6 percent well over the
lower limit for ethylene. This stream will support combustion.
Other Data - Heat Balance
a. Heat of combustion of methane, ethane, ethylene in stream- 3,221,078 BTU/Hr.
b. Less sensible heat, 85° > 570° (stack temperature)- 439,020 BTU/Hr.
c. Net heat available for steam generation 2,782,058 BTU/Hr.
d. Steam available at 750° F, 450 PSIG at 2,261 Lbs./Hr.
75c/1000 lbs., credit is $14,850 per year
-------
EO-24
TABLE EO-8
OXYGEN FEED PLANT
THERMAL INCINERATION
MATERIAL BALANCE
This same stream when thermally incinerated with 20 percent excess O2 will
give the same products of combustion as a steam generator but the heating
value of the stack gas will be lost.
An incinerator for this stream will cost less than half of a comparable
steam generator but the net operating cost is some 25 percent higher. See
Table E0-7.
-------
EO-25
B. CO2 Rich Purge Gas
This stream, which averages around 300 SCFM for a typical 200
MM lbs./year plant, is normally vented directly to the air by all
respondents. For air oxidation plants, the stream contains about
70 vol. 7° CO2 and 5-6% hydrocarbons and can be incinerated. The
best solution would be to combine this vent stream with the main
process vent stream since it is a small flow and dispose of both
streams together. One respondent plans to do just this, feeding
the CO2 rich stream to a catalytic converter along with the main
process vent and virtually eliminating emissions from the plant.
For the oxygen feed plants, this stream contains water vapor,
30-50 percent CO2 and only traces of hydrocarbons. Hence, these
plants probably have no need to treat this stream any further. One
respondent usually disposes of this stream by selling it to an off-
site consumer for its carbon dioxide content, but what is done with
the hydrocarbon content is un-reported.
C. By-Product Disposal
1. Waste Water
From 0.007 to 0.9 gallons of waste water per lb. ethylene oxide
are produced with the larger plants consuming the least water/lb.
product. Data on contaminants in the waste streams are very sketchy
but are probably ethylene glycol (mainly) acetaldehyde and miscellaneous
hydrocarbons. All respondents, save one, treat this stream in plant
before discharge. This stream is biodegradable and presents no
unusual operating problems.
2. Waste Solids
No waste solids were reported.
3. Miscellaneous
No other by-products were reported.
D. Best Available Pollution Control System
The best control system for an air oxidation ethylene oxide plant
is to feed both the main process vent and the CO2 purge vent to a
catalytic converter and use the hot gases from this unit to power
feed turbines to the process. With this set-up, emissions from the
plant vent streams should only be N2, O2, CO2 with possible traces
of hydrocarbons and NOx. The questionnaires indicate that several
plants are planning to install this type of system. Oxygen feed
plants are different than air feed plants because the CO2 purge vent
which is reported as essentially hydrocarbon free, is the larger
stream, and probably can be vented directly to the air with little or
no effect on the pollution level. It is also apparently pure enough
to be sold as CO2. The main process vent from these plants is small
but is combustible. It can best be disposed of by burning in a steam
generator with recovery of the heat as steam. If steam is not needed
and the gas cannot easily be transported elsewhere, it may be disposed
of in a thermal incinerator with a similar reduction in emissions as
the steam generator, but with no energy recovery.
-------
EO-26
One of the air oxidation processes features introduction of
methane as an "inert" background gas in the recycle stream in
contrast to other oxygen processes which have no methane addition.
The relatively rich main process vent gas from these plants can
be sent to the plant boilers as supplemental fuel, even though the
methane does not oxidize at reactor conditions.
Whether or not the CO2 purge vent from either type of oxygen plant
can be vented should be dtermined by an EPA sampling program. If not,
it could constitute a problem because it is too large to be combined
with the main process vent for incineration either in the plant
boiler, the steam generator or the thermal incinerator. The dilution
effect would render the combined stream non-combustible.
-------
EO-27
V. National Emission Inventory
Based on the emission factors shown in Table EO-3, the total emissions
from the ethylene oxide plants surveyed were estimated and are summarized in
Table EO-9. Using Table EO-9, the total approximate emissions from all U.S.
ethylene oxide plants are as follows:
Component
Average Emissions
T/T Ethylene Oxide
Total Emissions (1)
MM Lbs./Year
Hydrocarbons
Particulates
NOx
SOx
0.02048
0.000007
0.000066
0.000031
85.83
0.03
0.28
0.13
Total = 0.020584
86.27
If all air feed plants catalytically converted their main process vent
gases to CO2 and H2O and used the hot exhaust to drive the plant feed turbines,
(also cutting the turbine emissions about in half) and oxygen feed plants
incinerated their main process vent in a steam generator, total emissions would
be reduced to the following approximate values:
Component
Average Emissions (2)
T/T Ethylene Oxide
Total Emissions (1)
MM Lbs./Year
Hydrocarbons
Particulates
NOx
SOx
0.0049
0.000007
0.0001
0.000031
20.53
0.03
0.42
0.13
Total = 0.005038
21.11
(1) Basis 4,191 MM lbs. EtO per year in 1972.
(2) See also Table E0-16.
-------
Emission
TABLE EO-9
EMISSION SOURCE SUMMARY
TON/TON ETHYLENE OXIDE
AIR OXIDATION PLANTS
Source
Total
Hydrocarbons
Particulates
NOx
SOx
CO
EtO
Absorber
Vent Gas
0.0118
Recovery
Sect ion
Vents
Vent Gas
Scrubber
Outlet
Turbine
Exhaus t
Reactor or
Heater
Vent
CO2 Purge
Vent
0.0006
0.0023
0.00001
0.0132
0.0006
0.0001
0.00005
0.00001
Figitive
Emissions
0.00001
0.0285
0.00001
0.0001
0.00005
Hydrocarbons
Particulates, NOx
SOx, CO
OXYGEN OXIDATION PLANTS*
Reabsorber Vent Gas
y ' — 11 1 - 1
0.0064
Nil
Nil
Water Scrubber Vent
0.0060
Nil
Nil
To t a 1
0.0124
Nil
Nil
OXYGEN OXIDATION PLANTS WITH METHANE ADDITION*
CO2 Purge Vent Fugitive Losses Total
Hydrocarbons 0.0032 0.0006 0.0038
Particulates, NOx Nil Nil Nil
SOx, CO Nil Nil Nil
*Data from one respondent only.
-------
EO-29
VI. Ground Level Air Quality Determination
Table EO-3 presents a summary of air emissions data for various surveyed
ethylene oxide plants. This table includes emissions from the main process
vent, CC>2 purge vent, turbine exhausts, reactor heat transfer fluid vent and
an estimate of fugitive emissions. Information regarding vapor losses from
product storage have not been included, but losses from these storage
facilities should be low for the following reasons:
(a) All plants reporting data on their storage facilities have re-
frigerated, cooled and/or pressure (inert blanket) storage for
ethylene oxide. Hence, emissions should be negligible. One
plant blankets the ethylene oxide with natural gas to prevent
loss of EtO. The blanketing gas usually goes to a flare when
the tank "breathes". A loss of blanketing gas is unavoidable
but vill contribute virtually no air emissions if it is flared
properly.
(b) Some plants take ethylene oxide directly to a glycol plant and as
such have no storage facilities for this product.
Table EO-3 provides operating conditions and physical dimensions of the
various vent stacks. The EPA will use this information together with the air
emission data to calculate ground level concentration for later reporting.
-------
EO-30
VII. Cost Effectiveness of Controls
A cost analysis for alternative methods of reducing air pollution from
the main process vents of a typical 200 MM lbs. per year ethylene oxide plant
is shown on Tables E0-10 and EO-11. Table EO-10 represents data from plants
using air feed while Table EO-11 shows similar data for oxygen oxidation plants,
A. Investment
Purchased costs of steam generators, thermal incinerators and
catalytic incinerators were obtained from current vendor quotes for
similar packaged type units. Costs for the catalytic converter
represent actual dollars reported from respondents questionnaire
up-dated to 1973.
B. Operating Expense
Unless data were available from the respondent, we used:
1. Depreciation - 10 year straight line.
2. Interest - 6 percent on total capital
3. Maintenance - 3 percent of investment.
4. Labor - One-quarter of a man assumed for boilers (steam
generator) and one-eighth man for incinerators.
5. Utilities - Based on the Gulf Coast area.
The tables show that if the steam can be used, a steam generator is a
cheaper method to use than a thermal incinerator for an oxygen oxidation
process. A properly designed thermal incinerator has the advantage over a
flare in that less NOx should be produced. It steam could not be used or
transported to other process areas economically, a thermal incinerator would
be the best method to use to reduce emissions.
However, the air process vent streams are not readily combustible. In
this case, the best solution appears to be to catalytically incinerate the
vent gas. Two respondents indicated that this was done with the main process
vent gas. The hot gases from the converter (catalytic incinerator) are used
to drive the process air compressor turbine drives. The literature says that
almost every manufacturer uses these gases to drive turbines for the plants
but if so, only two respondents so indicated and gave supporting engineering
data.
If one take credit for the heating value of the gas burned which is
used to power some of the process feed turbines, the analysis shows that this
cost savings will just about off-set the operating and capital charges for a
catalytic converter in a 200 million lbs./year ethylene oxide plant.
All plants reporting store ethylene oxide either under refrigeration
or pressure (N2, natural gas). All have vapor conservation systems and none
vent to the atmosphere directly. As such, there does not appear to be
anything further necessary for the storage facilities.
-------
TABLE K0-1U
COST EFFECTIVENESS OF CONTROLS
AIR OXIDATION PROCESS
Chemical
Total U. S. Installed Capacity, Tons/Yr. (a)
Total U. S. Production, Tons/Yr. (a)
Principal process
Percentage of Total u. S. Production, Approximate
Average plant Capacity, Tons/Yr.
Emission Stream
Emission Control Device
Flov Rate, SCFM, Basis Average Plant Sire
Ton/Ton Ethylene Oxide Capacity
N0V
so2
CO
co2
N2 + Argon
°2
H,0
Misc. HC
Unidentified Particulates
Ethylene Oxide
Main Process Vent
Ethylene Oxide
2,100,000
2,100,000
Air Oxidation
76
100,000
None
32,402
None
None
None
0.86645
5.75796
0.22005
(j)
0.11327
None
0.00074
Catalytic (f) Incineration
None
None
None
1.22255
6.35093
0.00858
0.14962
TR
None
TR
None
480
None (b)
None
TR
0.10911
0.01743
0.00230
(j)
0.00442
None (d)
0.00074
C02 Rich Purge Gas
Catalytic Incinerator fe)
TR
None
None
0.12448
0.06675
0.00070
0.00631
TR
None
TR
Installed Cost Relative to Emission Control Device
Delivered Cost
$158,000
$12,000
Installation
67,000
8,000
Total Capital
$225,000
$20,000
Operating Cost, $/Yr.
Depreciation (10 years)
$
22,500
$
2,000
Interest on Capital (6%)
13,500
1,200
Catalyst Replacement (estimated)
5,000
500
Maintenance (37„ of capital)
6,800
600
Labor ($4.85/man-hour)
4,900
(1/8 man)
Utilities
Power, lc/KVH
200
200
Fuel, 40c/MM BTU (pilot gas)
500
(c)
500
Process Vater, 10c/M Gal.
0
0
Boiler Feed Vater, 30c/M Gal.
0
0
Total utilities
$
700
$
700
Total Operating Co6ts
$
53,400
$
5,000
Heating Value Credit (g)
50,000
1,000
Net Operating Cost
$
3,400
$
4,000
(a) Estimate for 1973.
(b) Exception - one respondent reports 40 PPM NO in a vent containing no other pollutant vapors.
(c) Meat exchange incorporated for sustained oxidation; pilot only during portion of start-up sequence. Start-up fuel required.
(d) Exception - one respondent reports "less than three lbs./hr. (of particulates), primarily carbon, 6mall quantities of iron and chlorine1'.
(e) Incineration figures are estimated for addition of facilities where none exist in current practice. The inclusion of these figures is not to be construed as an
indication that the facilities should or should not be installed.
(f) Data from two respondents shoved about $85,000 purchased cost in 1964 - 1966, updated to $150,000 in 1973 and estimated $225,000 installed cost.
(g) Credit of 0.025c/lb. of EtO/year taken from net heating value of gas burned and used to power tutbines.
(h) Streams are below lean lower limit for combustion. Thermal incineration or incineration in a steam boiler would require excessive firebox
temperatures for complete combustion.
(i) Installed cost based on data from two respondents. Corrected to 200 million lbs./year via the "o.6 factor" and up-dated from 1965 costs using O&GJ Construction Cost Index,
(j) Main Process Vent and CO2 Vent on water free basis.
-------
TABLE KO-11
COST EFFECTIVENESS 01- CONTROLS
OXYCEN OXIDATION PROCESS
Ethylene Oxi.de
0.5 MM
0.5 MM
Oxygen Oxidat i on
24
100,000
None
699
None
None
None
0.01660
0.00530
0.00193
(e)
0.00692
None
TR
Installed Cost Relative to Emission Control Device
Delivered Cost
Installation
Total Capital
Operating Cost, $/Yr.
Depreciation (10 years)
Interest on Capital (6%)
Catalyst Replacement (Estimated)
Maintenance (37. of capital)
Ixibor ($4.85/man-hour) '
Ut i1 it ies
Power, lc/KWH (estimated)
Fuel, 40c/MM BTU (pilot gas)
Process Water, lOc/M Gal.
Boiler Feed Vater, 30c/M Cal.
Total Utilities
Total Operating Costs;
Steam (feneration, 75c/M Lbs. (credit)
Net Operating Costs
Chemica1
Total U. S. Installed Capacity, Tons/Yr.
Total u. S. Production, Tons/Yr. (a)
Principal Process
Percentage of Total U. S. Production, Approximate
Average Plant Capacity, Tons/Yr.
Emission Stream
Emission Control Device
Flow Rate, SCFM Average Plant Size
Ton/Ton Ethylene Oxide Capacity
N0V
so2
CO
co2
N'2 + Argon
°2
h2o
Misc. HC
Unident if ied Par t iculat.es
Ethylene Oxide
co2
Non
.5,782
^2 Rich Purge Gas
None
None
None
None
0.92147
None
0.00013
(c)
0.00096
None
None
(a) Estimated for 1973.
(b) Basis 20/1 "excess air".
(c) Steam generation and incineration figures are estimated for addition of facilities where none ex
be construed as an indication that the facilities should or should not be installed.
(d) Cost of steam generator and thermal incinerator from A.P.C.I. (Catalytic, Inc.) company files.
(e) Vents on water free basis.
Main Process Vent
Steam Generator (c)
(b)
TR
None
n<
0.03622
0.09737
0.00499
0.03095
TR
None
None
20,000
20,000
40,000
4,000
2,400
0
1,200
9,700 (h. man)
200
1,500
0
700
2 ,400
19,700/y r.
($13,800/Yr.)
5,900/Yr.
Thermal Incinerator (c)
702
(b)
TR
None
tr
0.03622
0.09737
0.00499
0.01095
TR
None
None
0
1
8,000 £
8,000
16,000
I ,600
1,000
0
500
4,900 (1/8 man)
100
1,500
0
0
1 ,600
9,b0o/Yr.
9,600/Yr.
t in current, practice. The inclusion of these figures is not to
-------
EO-33
VIII. Source Testing
Plants 13-3 and 13-5 were the only plants that reported a catalytic
converter in operation when the questionnaires were completed. Data from
these two plants indicate that a catalytic converter is an excellent pollution
control device. If the data presented by these respondents can be substantiated
by a sampling program and if the hydrocarbon emission reduction is as complete
as is claimed, this device could be classified as the best demonstrated
emission control system. It has since been reported by an official of plant
13-7 that they also have a catalytic converter, although this was not clear
from the questionnaire.
Respondent 13-4 shows one stream which is quite rich in hydrocarbons
as going to a steam generator. (No data on operation of this unit were
given.) The stream is burned to generate steam used elsewhere. It would not
be valuable to source test this stream as it represents a unique process,
because methane is fed into the process streams to build up an inert background
gas in the recycle. However, since this plant is an oxygen feed plant and
reports very low hydrocarbons in the CO2 purge gas vent, a sampling program
to verify this could be of value.
-------
EO-34
IX. Industry Growth Projection
Current U.S. capacity for ethylene oxide production is estimated at
4,191 MM lbs./year and is projected to increase to 6,800 MM Lbs./year by
1985. (See Figure EO-3)
By far, the largest use for ethylene oxide is in the production of
ethylene glycol. Probably two-thirds of all the ethylene oxide produced
goes into glycol manufacture. Ehylene glycol's largest, but slowest growing
end use is anti-freeze which accounts for about 70 percent of the ethylene
glycol produced. Another area of increased use of glycol is in polyester
fiber and film production.
The second largest market for ethylene oxide is nonionic surface-active
agents which consume 10-15 percent of the ethylene oxide production. Ethoxylated
mixed linear alcohols (components in biodegrabable detergents) have been the
major contributor to this use of ethylene oxide in recent years.
Glycol ethers, the third largest end use of ethylene oxide has exhibited
the fastest rate of growth as a consumer of the oxide. Most of this growth
can be attributed to the use of ethylene glycol monomethyl ether as a jet
fuel additive.
Ethanolamines follow glycol ethers as the next largest consumer of
ethylene oxide. The fastest growing use for these ethanolamines has been
in the manufacture of surface active agents, e.g., diethanolamine condensates
of conconut oil acids and lauric acid„
Fifth in size of ethylene oxide consumption is diethylene glycol whose
major end use is in polyurethane and unsaturated polyester resins.
Ethylene oxide is also consumed in the manufacture of numerous other
chemicals such as ethyl acrylate, choline and choline chloride, aerylonitrile
(probably not being used commercially now), hydroxyethyl cellulose, hydrox-
yethyl starch, polyethylene oxide resins, fumigants and food sterilants.
A conservative growth rate for ethylene oxide through 1985 of four percent
per year has been estimated. It is based mainly on Figure EO-3 which shows
the growth of ethylene oxide production since 1940. In the early 1970s, by
far the dominant process has been direct oxidation as opposed to the older
chlorohydrin process. All new growth has been projected to be via the
direct oxidation route. A good average plant size is 200 MM lbs./year and as
such about 15 new units will have to be built between now and 1985. (See
Table EO-12)
The price history of ethylene oxide is interesting. From 1940 to the
mid fifties, it showed a small increase each year until the direct oxidation
process really began to catch on. From 1955 to the present, there has been
a steady drop in price from 15c/lb. to 7.5c/lb. in mid 1973 despite a general
rise in prices in the economy.
The chlorohydrin process is only economic when the price of ethylene is greater
than six times that of the chlorine price and is also more economical when ethylene
oxide prices are high. At low ethylene oxide prices, the direct oxidation route
gives a better return on ethylene and with low ethylene prices, the direct
oxidation process has completely taken over the production of ethylene oxide.
-------
EO-35
1
i : , FIttURE EQ-3. . ¦ j
: eWt.KNF. (fXlDE; !
CAPACITY PROJECT I Oil
u.
2QS&
.u
600
m
Chqm. Economics Handbook
65^-5030 - August, 1969.
Cham. Mkc. Reporter
Octiober I. 1969
Chejmical Pro if! lie.
1QP
3. Questionnaires Current
C»jj»clty Data. j .
40
im
-------
TABLE EO-12
NUMBER OF NEW PLANTS BY 1985
Current
Capacity
4,191
Marginal
Capacity
356
Current
Capacity
On-Stream
in 1975
3,835
Demand
1985
6,800
Capacity
1985
6,800
Capacity
to be
Added
2,965
Economic
Plant
Size
200
Number
of New
Units
15
NOTE: All capacities in millions of pounds per year.
-------
EO-37
X. Plant Inspection Procedures
The largest source of emissions from an ethylene oxide plant would be
hydrocarbons, chiefly ethylene. These hydrocarbon emissions are sometimes
vented as a very dilute stream of hydrocarbons in a large amount of nitrogen,
CO2 and some oxygen. The vented gas stream should be below the lower ex-
plosive limit of ethylene in air.
A. There is no easy way to test for the presence of hydrocarbons in an
air-C02 stream other than a chromatograph and this is not generally
a portable instrument. A quick, albeit rather qualitative test for
the presence of ethylene is odor. There should be no odor detectable.
Ethylene has a sweet odor and taste but it is not a strong odor.
Ethylene has no local or chronic irritant rating. If the odor of
ethylene is detectable, the unit is probably upset and venting
improperly. The normal two percent ethylene + one percent in the
vent gas probably will not be detectable by smell. The absence of
any sweet odor does not mean that there are zero emissions.
B. Ethylene oxide, on the other hand, is quite detectable by odor. It
is irritating to the eyes and nose and is classified as an acute
local irritant. In high concentrations, it numbs the sense of smell.
It is toxic and should not be inhaled although it is non-cumulative
in the body and is not a chronic poison. If the odor of ethylene
oxide is detectable, chances are that the purge reactor absorber (or
the secondary absorber/scrubber) is upset and not functioning properly.
If would be advisable to discuss the following with plant officials.
1. Proper flow of absorbing medium - water.
2. Proper temperature of the absorbing medium and the scrubber.
3. Deposits of material on the packing of the absorber which could
foul the scrubber internally and cause by-passing.
4. Markedly different temperature and pressure of the feed gas to
the scrubber than design.
Since ethylene oxide is the desired product and is being recovered in
rather dilute form from the reactor effluent gases, prudent management
of the plant would insist that every bit of ethylene oxide be recovered.
It is just not economical to lose ethylene oxide due to improper
scrubber operations so these instances should be rare.
C. Stacks should be checked for particulate emission and NOx by opacity.
All stacks should be practically colorless but again a clear stack
does not mean no hydrocarbon emissions because ethylene and related
compounds are colorless.
D. Flares, if any, should be smokeless and only a rather substantial upset
should cause smoke here. However, putting a dilute stream of hydro-
carbons through a flare below the lower combustible limit of the stream
will result in a clean flare but no reduction in hydrocarbon emissions.
E. Incinerators - If thermal incinerators are provided on any of the
vent streams, a visual check of the stack gas should be made. The
-------
EO-38
stack should be nearly clear, if not completely clear. Since the
feed stream to the incinerator is liable to be below its combustible
limit, one should check to see if adequate supplemental fuel is being
burned to properly combust the vent gas. Design specifications should
be compared to actual operating conditions since there probably will
be no visible clue in the stack effluent. Some check points would be:
1. Burner in the combustion zone should be operating.
2. Combustion zone temperature in the design range.
3. Stack gas temperature in the design range.
4. Process gas valves open to the incinerator.
A catalytic incinerator or converter may also be used. In fact, this
type of device is more likely to be encountered than a thermal
incinerator. Generally, the off-gas from the converter will be used
to power a turbine before exhausting to the air. The same points
should be checked on the catalytic converter as on the thermal
incineiitor. Supplemental fuel may or may not be used on a catalytic
incinerator.
Log data on operation of both types of incinerators should be
reviewed, especially if stack gas analyses are available via a
chromatograph and recorded intermittently or continuously. Hydro-
carbons in the stack gas should be virtually nil (say 0.2 percent or
less and zero if possible).
-------
EO-39
XI. Financial Impact
Table EO-13 presents an estimate of the cost of ethylene oxide manufacture
by both an air oxidation process and by an oxygen oxidation process. The
difference in manufacturing cost may not be as great as the table shows at
200 MM lbs./year (i.e., the 0.48c/lb. may be on the high side) but the
literature indicates a small advantage for the air process at capacities
over about 100 MM lbs./year, a break even at 60-70 MM lbs. and a small edge
for the O2 process in smaller plants. The minimum economical plant size is
reported to be about 25 MM lbs./year.
Table EO-14 was calculated by using the air oxidation process as outlined
in Table EO-13, but adding emission control facilities to reduce hydrocarbon
venting to near zero with a catalytic converter on the two main process
streams. This represents the most feasible new air oxidation facility at
200 MM lbs. per year production and virtually no emissions. Credit for
power derived from the incinerator gases was taken as 0.25c/lb., based on
data in the questionnaires. Cost of building a new 200 MM lbs./year air
oxidation ethylene oxide plant with good pollution control facilities would
reduce profits after taxes by about 1.5 percent.
Table EO-15 was calculated by using the oxygen oxidation process as
outlined in Table EO-13, but adding emission control facilities to reduce
hydrocarbon venting to near zero with an incinerator and steam generator.
This represents the most feasible new oxygen oxidation facility at 200 MM
lbs./year production and virtually no emissions. Net operating cost is only
$5,900/year. Cost of building a new 200 MM lbs./year oxygen oxidation
ethylene oxide plant with good pollution control facilities would reduce
profits after taxes by about eight percent.
A sensitivity analysis was performed for the catalytic incineration
installation in the air oxidation process (Table E0-14b). Installed cost
of the pollution control facility was varied + 50 percent and its effect on
the R.O.I, calculated. The 6.8 percent R.O.I, was found to vary from 6.5
to 7.1 percent.
A comparable calculation on the operating cost sensitivity would result
in variations of similar magnitude because the savings in operating costs
through the addition of a catalytic converter are approximately equal to
the increased capital charges. If both sensitivities were applied simultaneously,
they sould nearly off-set each other.
Sensitivities were not calculated for the oxygen oxidation process because
the extra capital involved is only $40,000 on a $5,200,000 plant and the extra
operating cost is only $20,000/year on a $14,540,000/year total. Thus, it was
judged that the sensitivities are not significant.
Table E0-16 is a pro-forma balance sheet for the two above cases. A
constant selling price of 7.5c/lb. for ethylene oxide was used. Capital
requirements for the most feasible new plants are about $225,000 higher with
air oxidation and about $40,000 higher with oxygen oxidation than for
existing type plants.
In addition to the financial impact of the proposed most feasible new
plants a definite positive environmental impact would result. Not only
would 2,050 tons of hydrocarbons per 200 MM lbs./year plant not be vented
to the atmosphere (for 1985, 15 new plants would vent 30,750 tons of hydro-
carbons per year if not controlled) but an energy savings could be realized
-------
E0-40
if the hot gases were used to drive some of the process turbines or generate
steam.
For 1985, installation of emission control facilities and utilization of
the power generated would be equal to a savings of about three billion cubic
feet of gas per year.
-------
EO-41
TABLE EO-13
ETHYLENE OXIDE MANUFACTURING COST*
FOR A TYPICAL
EXISTING 200 MM LB./YR. FACILITY
PROCESS
AIR
DIRECT MANUFACTURING COST
Raw Materials
Ethylene (4.0c/lb.)
Oxygen ($15/ton)
Catalyst & Chemicals
Labor
Maintenance (§ 5% of Investment
Utilities
C/LB. $/YR.
3.8
1.0
0.2
0.31
0.13
0.60
6.04
C/LB. $/YR.
3.8
0.0
0.2
0.31
0.25
0.60
5.16
INDIRECT MANUFACTURING COST
Plant Overhead @ 110% of Labor
FIXED MANUFACTURING COST
0.34
0.34
Depreciation - 10 year str. line
Insurance & Prop. Tax 2.370 of Inv.
0.26
0.06
0.32
520,000
0.50
0.11
0.61
1,000,000
MANUFACTURING COST
6.70
6.11
GENERAL EXPENSES
Administration (37> of Mfg. Cost) 0.20
Sales (1% of Mfg. Cost) 0.07
Research (27, of Mfg. Cost) 0.13
Finance (67„ of Investment) 0.15
0.18
0.06
0.12
0.30
TOTAL COST
7.25 14,500,000 6.77
13,540,000
PRODUCT VALUE
Ethylene Oxide @ 7.5c/lb.
Profit before Taxes
NPAT (527„)
Cash Flow
ROI
15,000,000
500,000
240,000
760,000
4.67,
15,000,000
1,460,000
700,800
1,700,800
7.07o
*Taken from literature "Petrochemical Industry: Markets & Economics" up-dated
and scaled up to 200 MM lbs./year using "0.6 Factor" on capital for both O2
and Air Feed Plants. Based on 1973 dollars.
-------
E0-42
TABLE EO-14
ETHYLENE OXIDE MANUFACTURING COST*
FOR NEW PLANT WITH CATALYTIC INCINERATION
200 MM LBS./YR. FACILITY
AIR OXIDATION PROCESS
DIRECT MANUFACTURING COST
Raw Materials
Ethylene
Catalyst & Chemicals
Labor
Maintenance
Utilities
C/LB.
3.8
0.2
0.31
0.255
0.575
5.140
$/YR.
INDIRECT MANUFACTURING COST
Plant Overhead (1107o of Labor)
FIXED MANUFACTURING COST
0.34
Depreciation - 10 year straight line
Insurance & prop. Tax (2.37o of Investment)
0.51
0.12
0.63
1,020,000
MANUFACTURING COST
6.11
GENERAL EXPENSES
Administration - 3% of Mfg. Cost
Sales - 1% of Mfg. Cost
Research - 2% of Mfg. Cost
Finance - 6% of Investment
0.18
0.06
0.12
0.31
TOTAL COST
6.78
13,560,000
PRODUCT VALUE
Ethylene Oxide @ 7.5c/lb.
Profit before Taxes
NPAT (52%)
Cash Flow
R0I
15,000,000
1,440,000
691,200
1,711,200
6.8%
*Cost of catalytic converter taken from questionnaire data updated and corrected
to a 200 MM lbs/year plant. Other costs from Table E0-13. (1973 dollars)
-------
EO-43
TABLE E0-14A
SENSITIVITY ANALYSIS
AIR OXIDATION PLANT WITH
CATALYTIC CONVERTER AND + 50% VARIATION
IN INSTALLED COST OF CONVERTER
BASE CASE (1) + 50% -50%
DIRECT MANUFACTURING COST 5.14c/lb. 5.14c/lb. 5.14c/lb.
INDIRECT MANUFACTURING COST 0.34 0.34 0.34
FIXED MANUFACTURING COSTS
Depreciation - 10 year str, line 0.51 0.52 0.50
Insurance & Property Tax 0.12 0.12 0.12
MANUFACTURING COST 6.11c/lb. 6.12c/lb. 6.10c/lb.
GENERAL EXPENSES
Administration
0.18
0.18
0.18
Sales
0.06
0.06
0.06
Research
0.12
0.12
0.12
Finance
0.31
0.32
0.30
TOTAL MANUFACTURING COST, c/LB.
6.78
6.80
6.76
PRODUCT VALUE
EtO @ 7.5c/lb.
^
$15,000,000 -
>•
Profit Before Taxes, $/yr.
1,440,000
1,400,000
$1,480
NPAT (52%) $/Yr.
691,200
672,000
710
ROI
6.87o
6.5%
7.
(1) CF Table EO-14.
-------
EO-44
TABLE EO-15
ETHYLENE OXIDE MANUFACTURING COST*
FOR NEW PLANT WITH STEAM GENERATION
200 MM LBS./YEAR FACILITY
OXYGEN OXIDATION PROCESS
DIRECT MANUFACTURING COST
Raw Materials
Ethylene
Oxygen
Catalyst & Chemicals
Labor
Maintenance
Utilities
C/LB.
3.8
1.0
0.2
0.31
0.13
0.61
6.05
$/YR.
INDIRECT MANUFACTURING COST
Plant Overhead @ 110% of Labor
FIXED MANUFACTURING COST
Depreciation - 10 Yr. Straight Line
Insurance & Property Tax (2.37o of Inv.)
0.34
0.262
0.06
0.322
524,000
MANUFACTURING COST
6.712
GENERAL EXPENSES
Administration (3% of Mfg. Cost)
Sales (1% of Mfg. Cost)
Research (2% of Mfg. Cost)
Finance (6% of Investment)
TOTAL COST
PRODUCT VALUE
Ethylene Oxide (? 7.5
-------
TABLE EO-16
PRO-FORMA BALANCE SHEET
200 MM LBS./YR. ETHYLENE OXIDE MANUFACTURING FACILITY
Current Assets
Fixed Assets
AIR OXIDATION PROCESS
OXYGEN OXIDATION PROCESS
EXISTING
MOST FEASIBLE NEW
EXISTING
MOST FEASIBLE NEW
Cash (a)
Accounts Receivable (b)
Inventories (c)
1,018,333
1,250,000
1,354,000
1,018,333
1,250,000
1,356,000
1,116,667
1,250,000
1,450,000
1,118,667
1,250,000
1,454,400
Plant
Building
Land
Current Liabilities (d)
Equity & Long Term Debt
10,000,000
100,000
50 >000
$13,772,333
976,666
12,795,667
10,225,000
100,000
50,000
$13,999,333
973,333
13,026,000
5,2000,000
100,000
50,000
$9,166,667
1,130,000
8,036,667
5,240,000
100,000
50,000
$9,213,067
1,131,667
8,081,400
w
0
1
¦p-
Total Capital
$13,772,333
$13,999,333
$9,166,667
$9,213,067
(a) Based on one months total manufacturing cost.
(b) Based on one months sales.
(c) Based on 20 MM lbs. of product valued at total cost.
(d) Based on one months total cost less fixed manufacturing and finance costs.
-------
EO-46
XII. Cost to Industry
Current emission control devices on ethylene oxide plants are not very
plentiful. This is due to two main factors. First, ethylene oxide is
recovered in dilute form from the reactor effluent stream. All plants use
a water absorber for this stream. Strictly speaking, these absorbers are
not pollution control devices but necessary process equipment. Moreover,
they do virtually nothing to remove hydrocarbons from the vent stream.
Secondly, the scrubbed gases from the main reaction section and the
waste CO2 purge gases are normally vented directly. All of these streams
contain a few percent hydrocarbons and many of them are non-combustible and
present a problem in cleaning up before releasing. The most feasible solution,
employed by two respondents and mentioned as a desired improvement by others
is a catalytic converter to burn these streams to CO2 and H2O and recover power
from the hot gas. Cost of this equipment is roughly estimated at two percent
of plant investment.
One respondent, whose process is unique, has a main process vent stream
that is combustible. This stream is burned in a steam generator presumably
off-site from the ethylene oxide unit. Unfortunately, there are no data
available for this operation.
In the most feasible new plants shown in Tables EO-14 and 15, air pollution
control equipment represents about one to two percent of plant investment. The
resulting total production cost for ethylene oxide for these units would be
about 0.01<: per lb. higher using two air oxidation plants as comparison and
about 0.02c using two oxygen oxidation plants as comparison.
Assuming all new plants built between now and 1985 incorporate these
types of control equipment, the total incremental capital cost will be about
$2,650,000 (1973 dollars), if the ratio of air to oxygen processes remains
as it is today (i.e., 25 percent oxygen process).
The projected effect of this expenditure on future air emissions is
shown on Table E0-17.
-------
TABLE EO-17
ESTIMATED 1985 AIR EMISSIONS
FOR
ALTERNATIVE CONTROL SYSTEMS
Type of Pollution Control
Present Systems (1)
Present Systems (1)
Existing plants Modified & New Plants
Incorporating Most Feasible Control (3)
Ethylene Oxide Production,
MM Lb6./Year
A .191
6,800 (2)
6,800 (2)
Pollutant
Average
Emiss ions
T/T
Total
Emissions
MM Lbs./Yr.
Weighted (4)
Emissions
MM Lbs./Yr.
Average
Emi6S ions
T/T
Total
Emissions
MM Lbs./Yr.
Weighted (4)
Emiss ions
MM Lbs./Yr.
Average
Emi ss ions
T/T
Tota I
Emiss ions
MM Lbs./Yr
Weighted (4)
Emissions
MM Lbs./Yr.
Hydrocarbons
0.02048
85.83 6,866
0.02048
139.26
11,141
0.0049
33.32 2,666
Particulates
0.00000?
0.03 2
0.000007
0.05
3
0.000007
0.05 3
NOx
0.000066
0.28 11
0.000066
0.45
18
0.0001
0.68 2 7
SOx
0.000031
0.13 3
0.000031
0.21
4
0.000031
0.21 4
CO
0.0
0 0
0.0
0
0
0.0
0 0
0.02058
86.27 6,882
0.02058
139.97
11 ,166
0.00504
34.26 2,700
(1) Assumes questionnaires received arc
typical of industry, and without
controls ,
growth wi11
have identical
I emissions factors.
(2) Estimated
total production by 1985.
Assumes growth in s^me proportions
as pre
sent, i.e. ,
66 percent air
oxidation and Shell process for 86
percent of balance.
(3) Catalytic
conversion of main proces:
s vents plus carbon dioxide vent
on i
ill air
oxidation p
lants and incineration or main proces*
} vent on all
oxygen oxidation plants.
(4) Weighting
Factors: Hydrocarbons - I
30, Particulates - 60, NOx - 40,
SOx
- 20,
and CO - 1.
-------
EO-48
XIII. Emission Control Deficiencies
Technical deficiencies preventing reduced levels of emission include
the following:
A. Process Chemistry and Kinetics
The direct oxidation of ethylene to ethylene oxide is a specific
reaction promoted by a silver catalyst and one which must be carefully
controlled in order to form ethylene oxide and not CC^.
1. Reactor Feed
(a) Ethylene - any appreciable quantities of impurities
could react with oxygen in the feed to form small
quantities of oxygenated compounds in the reactor
effluent, probably to the detriment of ethylene oxide
formation and unnecessarily complicating the recovery
and refining steps used to make 99+7= pure ethylene oxide.
Furthermore, some impurities compete with ethylene for
the active catalyst sites and thus act: as temporary
poisons.
(b) Oxygen - either air or oxygen are satisfactory feed to
the process. Of course, the plant must be designed for
that particular feedstock. One patent claims that the
amount of argon brought in with the oxygen influences
catalyst activity.
2. Reactor Operating Conditions
This reaction is very sensitive to temperature. (See Section II)
In order to maximize ethylene oxide yield and supress CO2 formation,
it is very important that the catalyst surface remain at the
desired temperature and that there be no "hot spots". Since no
one has designed the perfect isothermal reactor yet, there is
bound to be some temperature gradient leading to formation of
CO2 as a by-product. Fortunately, CO2 is not an air pollutant but
unfortunately it results in a more dilute stream from which to
remove pollutants. It also represents a dollar loss of ethylene
which could have gone to ethylene oxide.
3. Catalyst
Considerable research has been done on producing a superior
catalyst for ethylene oxide production. No metal yet tested can
compete with the conversion/selectivity relationship of silver.
Other important factors are use of a promotor and stabilizer for
the catalyst (alkali and alkaline earth metals), the chemical
nature and physical state of the support and the use of an
oxidation deactivator such as ethylene dichlori.de,
B. Process Equipment and Operations
1. Reactors
The reaction producing ethylene oxide is exothermic but must be
-------
EO-49
carefully controlled to avoid going too far and producing CO2
instead. As mentioned earilier, good heat transfer from the
catalyst to the heat transfer medium is essential to maintain
a controlled reaction. Optimum heat control could possibly be
realized by using a fluidized bed instead of the universal fixed
bed reactor with tubes used in all ethylene oxide plants. A fluid
bed process was pilot planted but never commercialized. Apparently
if one can attain good to excellent temperature control in a fixed
bed (and many people do) ethylene oxide yields are higher and there
is less of a problem with catalyst costs, activity, attrition, life
and recovery. To hold emissions to a minimum, it is imperative
that the reactors be operated to produce ethylene oxide maximum
and not to leave large quantities of ethylene or hydrocarbons in
the vent gas. Fortunately, it makes good economic sense to operate
in this manner.
2. Absorbers
Since the concentration of ethylene oxide in the absorber feed
is relatively low, care must be taken to insure,proper operation
of the absorber so as not to lose ethylene oxide product. Absorber
operation was covered in Section X - Plant Inspection Procedures.
C. Control Equipment and Operations
The best method of reducing air emissions from air feed ethylene
oxide plants is catalytic incineration of the two main vent streams.
If the catalytic converter is operating properly, there are no
SOx and CO emissions and little or no hydrocarbons and NOx. The
only appreciable emissions should be CO2, water, nitrogen and oxygen.
Hence, this system really has no deficiencies in terms of emissions.
For an oxygen feed ethylene oxide plant, a steam boiler can be
used to incinerate the main process vent gas since it is usually
combustible. Again proper operation of the boiler firing should
give only CO2, H2O, N2, O2, and perhaps traces of NOx in the
exhaust gas. If there is no use for the steam produced, a thermal
incinerator would be just as effective and cheaper to install but
would not give any steam credit to the operation.
-------
EO-50
XIV. Research and Development Needs
The following programs are suggested as a way to improve ethylene oxide
yield and simultaneously reduce air emissions.
A. Existing Plants
1. CO2 Inhibitor
Presently some plants use PPM of chlorinated hydrocarbons
(e.g., ethylene dichloride) as inhibitors for the reaction
ethylene + 02 "¦ ) CO2. Obviously, if a better suppressant
could be found, the yield of ethylene oxide from ethylene would
improve. This change would probably not have a great positive
effect in reducing air emissions as CO2 is not considered a
pollutant and unreacted ethylene would still be vented. However,
it is a step in the right direction since reduction in CO2
formation would require less venting of the recycle gas to the
atmosphere.
2. Improved Catalyst
Development of a more selective catalyst to produce ethylene
oxide from ethylene at the expense of CO2 formation would have a
double barreled value. First, more valuable product would be
available to the producer and less by-product of little or zero
value. Second, more efficient use of ethylene would result in
more ethylene oxide and less unreacted ethylene to recycle and
eventually vent as an air emission.
According to the literature, the choice of metal has been
rather well worked over. Silver has been the best choice of all
metals tested since the beginning of the air oxidation process.
Although this ground has been ploughed thoroughly, it would bear
checking again. Petroleum reforming catalyst has used platinum
as the best choice since the inception of the process in the late
1940s. However, after 20 years of research on the catalyst,
recent developments have lead to several bimetallic catalysts of
far superior properties than the original platinum catalyst even
though platinum is still the major ingredient. After 20 years of
research on ethylene oxide catalyst, there is still the chance that
the silver based catalyst could be improved or supplemented by
another metal or metals.
Many more catalyst supports are available today than when the
ethylene oxide silver catalyst was first developed. Surely not all
of these have been tested. The proper choice of support for a given
catalyst can mean the difference between a dead, a mediocre and an
outstanding catalyst. This should be investigaced.
Finally, the method of preparation of a given catalyst on its
support can also mean the difference between a winner, a loser and
an also ran. This field bears investigation, too.
3. Catalytic Incinerator
This type of device appears to be the best for reducing air
emissions from ethylene oxide plants that utilize air oxidation, hence
-------
EO-51
the following factors should be considered:
(a) Complete combustion to CO2 and H2O should be sought. A two
stage converter is a possible technique.
(b) The design should be optimized to give good performance at
low installed cost. Some of these devices are very expensive.
(c) Units which fit existing plants with no incinerator present
would be of value.
4. Steam Generator
In an oxygen oxidation plant, the main process vent is small
compared to the CO2 vent stream. It is normally rich enough to
sustain combustion and can best be handled in a steam generator
or a thermal incinerator. The stream is too small for feeding
to a catalytic incinerator and using the hot gases as a power source
to drive the feed turbines. But a steam generator will at least
recover some energy for the user. Here, some work could be done to
determine exactly which type of steam generator would be the most
economical to install in this particular application.
Design work on a catalytic incinerator for air oxidation plants
or a steam generator for an oxygen oxidation plant could best be
performed by the manufacturers of these lines of equipment, or
process licensors.
B. New Plants
In addition to the above R&D projects which could be used
in either new or existing plants, the following would require new
facilities and would not be totally applicable to existing plants.
1. Fluid Reactor
This process has already been piloted in the early 1950s but was
never commercialized. In view of the giant strides taken in fluid
catalyst technology in the past 20 years, it would be well to reinvestigate
this process using either air or oxygen feed to see if it could be
improved. The ultimate aim would be, as with an improved catalyst,
to give a more selective process yielding more ethylene oxide, less
CO2 and consuming more of the ethylene which now goes out the vents
as air emissions.
2. Fluid Catalyst
The success or failure of (1) above probably would hinge on
development of a selective, attrition resistant and economical
fluid catalyst (which was a problem when the fluid process was
first investigated). One can not have one without the other.
C. Industry Background
It should be remphasized that most, if not all of the ideas covered
in Sections A & B have been investigated by industry in the past.
This is especially true in the field of catalyst research and
development of the fluid process where much work has been done. It
-------
EO-52
is not to suggest that not enough has been done in these areas but
rather to point out that here are two areas which can never be
dismissed out of hand as long as new materials and techniques are
being developed. For example, a new catalyst support which becomes
commercially available might yield a breakthrough in preparing a
high activity, high selectivity ethylene oxide catalyst from a metal
or metals which had been tried before.
-------
EO-53
XV. Research and Development Programs
The following proposed projects relate to those areas of R & D which
seem to offer the best change of obtaining a method of reducing emissions
from ethylene oxide plants. They are both of the nature that requires
considerable proprietary knowledge of catalyst so might best be conducted
by licensors or suppliers of ethylene oxide processes and catalysts.
A. Project A
1. Title - Catalyst Modification Program
2. Objective: To make a preliminary screening study of variations on
existing commercial ethylene oxide catalyst to see if any new
leads could be uncovered to show the way to a better catalyst
with respect to aging, selectivity or activity. This study
would define the catalyst problem, but not necessarily solve it.
It could open the door to a full scale catalyst investigation
far beyond the content of this program.
3. Estimate Projects Costs (See Table EO-18 for Cost Breakdown)
Capital Expenditure $ 25,700
Operating Costs
Unit Operations 71,600
Services 30,300
Miscellaneous 4,500
Contingency 61,000
Total 193,100
4. Scope: This project would seek leads toward producing a better
ethylene oxide catalyst which could improve yields and reduce
plant emissions.
5. Program: A catalyst screening unit should be constructed with
facilities for the calculation of the effluent by chromatographic
procedures with emphasis on the quantitative analysis of by-
products. The normal operating characteristics of the screening
unit would be determined by employing a commercial supported
silver catalyst. Experimental catalysts (perhaps) various noble
metal alloys of silver and combinations of alkali oxide or
alkaline earth metal promoters should be screened to determine if
by-product formation can be reduced without adversely altering
the main catalytic function. Investigation of various support
media should also be carried out.
6.
Timetable - It is estimated that the above program would require
a total of 14 months to complete.
-------
EO-54
TABLE EO-18
DETAILED COSTS
FOR
R&D PROJECTS
Misce1laneous
Project "A'
Project "B'
Ac Capital Expenditures
Test Unit Construction
Unit Cneckout
Professional
Operator
B. Operating Expenses
Unit Operation
Professional
Operator
Services
Analytical
Cat. Prep. & Testing
Unit Maintenance
$ 20,000
3,600 (4 weeks)
2,100 (4 weeks)
45,600 (50 weeks)
26,000 (50 weeks)
3,300 (5 weeks)
22,500 (45 weeks)
4,500 (10 weeks)
S 20,000*
3,600 (4 weeks)
2,100 (4 weeks)
18,200 (20 weeks)
10,400 (20 weeks)
2,000 (3 weeks)
1,800 (4 weeks)
Materials
Report Writing
D. Total of A to D
Contingency
3,000
1,500
132,100
61,100
$193,100
1,000
1,500
60,600
30,300
$90,900
*N0TE:
The test units for both projects will be quite similar. Should
projects be run sequentially, a single test unit would suffice,
with appropriate savings.
the
-------
EO-55
B. Project B
1. Title - Improved CO2 Inhibitor Program.
2. Objective: To begin a preliminary screening study of potential
fuel additives which would inhibit the formation of CO2 during
the oxidation of ethylene. The project would hope to sort out
several promising inhibitors but would not try to optimize them
or devise new inhibitors. The results of this study could be
used for these purposes.
3. Estimated Project Costs (See Table EO-18 for Cost Breakdown)
Capital Expenditures S 25,700
Operating Costs
Unit Operations 28,600
Services 3,800
Miscellaneous 2,500
Contingency 30,300
S 90,900
4. Scope: This project would seek improved CO2 formation inhibitors
for the ethylene oxide process.
5. Program: A CO2 formation inhibitor screening unit would be
constructed with facilities for the evaluation of reaction
products by chromatographic procedures, with emphasis on the
quantitative analysis of CO2. The normal operating characteristics
of the screening unit should be determined by employing a standard
organic halide inhibitor. Experimental inhibitors would be
screened to determine if CO2 formation can be reduced without
adversely altering the activity/selectivity functions of the
catalyst.
6. Timetable: It is estimated that the above program would require a
total of ten months to complete.
-------
EO-56
XVI. Sampling, Monitoring and Analytical Methods for the Ethylene Oxide Process
This process is a fairly simple process to monitor. Stream components
consist of CO2j ethylene, ethylene oxide, oxygen, nitrogen, water vapor and
traces of methane, ethane and other hydrocarbons. These are easily measured
by conventional "on stream" analytical techniques. Typical analyses that can
be performed are:
1. Ethylene measurement in the reactor feed gas for safety purposes
to insure that the ethylene concentration is below the LEL. In
the reactor effluent, the ethylene concentration is an indication
of reaction efficiency. Ethylene is also measured in the recycle
gas for build up.
2. Carbon dioxide is measured before and after the reactor indicating
the amount of inerts in the stream as well as indirectly monitoring
the reactor efficiency and possible combustion of ethylene to CO2.
3. Ethylene oxide is monitored as a reactor product and in the recycle
stream for checking product removal in the absorption system.
4. Oxygen is measured to insure sufficient oxygen to maintain the
reaction, and for safety purposes.
5. Water vapor, nitrogen, argon and hydrocarbons can be measured for
material balances.
6. Carbon monoxide can be measured if present in the recycle stream
for indication of possible catalyst poisoning or side reactions.
When a complete analysis of the gas stream is required, gas chromatography
is usually the method selected for measuring N2, O2, A, EtO, ethylene, etc.
However, when continuous analysis of a single component in a multi-component
stream is desired, oxygen can be analyzed by paramagnetic methods and non-
dispersive infra red analysis can be used to measure CO, CO2, ethylene,
ethylene oxide, etc.
As is typical of many other types of chemical process industry, the
survey revealed that monitoring of the various process streams is much more
common than monitoring of emissions into the atmosphere. However, the analytical
technology is quite similar and it may thus be presumed that most individual
plants are capable of measurement of emissions without extensive investment
in time or equipment. A summary of the information obtained concerning the
methods in use is shown in Table EO-19.
In reviewing the information available on analytical techniques, it is
apparent that multiple column gas chromatography is the method of choice in
the industry. The specific methodology, however, is variable. From the air
pollution standpoint, ethylene and ethylene oxide are or primary concern and
can be separated on a two-column instrument. Prior to the adoption of any
requirements for continuous or periodic monitoring of stack gases, it would
appear advisable for the USEPA to develop a reference procedure for ethylene
and ethylene oxide analysis. Such a project would involve an independent
verification of the accuracy and specificity on one or more of the methods
described in Table EO-19 and should be minor in scope.
-------
EO-57
TABLE EO-19
SUMMARY OF EMISSIONS MEASUREMENT TECHNIQUES
USED IN ETHYLENE OXIDE PLANTS
Page 1 of 3
Plant
Number
Source &
Flow Rate
Sampling and Analytical Methods
13-1
Absorber Vent
(24,000 SCFM)
Stripper Vent
(164 SCFM)
Routine measurement of ethylene by
extracting sample into a cylinder
for laboratory analysis by infra-
red. Instrument and analytical
details unknown. Occassional
gas-liquid chromatography and
mass spectrometer analysis of
other constituents. Flow continuously
metered.
Samples collected once per year for
laboratory analysis by GLC and Mass
Spectrometer. Details unknown.
13-2
Compressor Vent
(131,000 Lbs./Hr.)
Composition of the gas stream is
continuous monitored by a combustion
of chromatographs as described below.
Flow is continuously metered.
(1) Varian Model 1720, dual column,
thermal conductivity.
Column 1. Analysis for
CO2, C2H4, 02^, A & O2,
N2, and CH4 on a Porepak R,
80-100 mesh column (7* x
S.S) followed by 5A, 60-80
mesh molecular sieve (71 x
S.S). The column is operated
at ambient temperature.
1,11
¦4
Column 2. Argon and oxygen
are separated on 201 of V
S.S. containing 60-80 mesh,
5A molecular sieve at
-25° C.
(2) Varian Model 1740, single column
flame ionization. Ethylene oxide
is partitioned on a 71 x V S.S.
column containing 80-100 mesh,
Porepak R, at 170° C.
Reabsorber Vent
(3,671 Lbs./Hr.)
The concentration of CH4, C2H4, C2H6
and unidentified aldehydes is measured
using a Varian Model 2740 chromatograph
with flame ionization detector. The 7'
x S.S. column contains 89-100 mesh
Porepak R and is heated from 50° to 220°
at a rate of 10°/min. Samples are col-
lected in a stainless cylinder. Stack
flow is continuous metered.
-------
EO-58
TABLE EO-19
SUMMARY OF EMISSIONS MEASUREMENT TECHNIQUES
USED IN ETHYLENE OXIDE PLANTS
Page 2 of 3
Plant
Number
13-3
Source &
Flow Rate
Reabsorber Vent
(200 SCFM)
13-4
13-5
Process Vent
(950 SCFM)
Turbine Vent
(16,550 SCFM)
Absorber Vent
(5,500 Lbs./Hr.)
Recovery Vent
(480 SCFM)
13-6
Cycle Gas Exhaust
(64,000 SCFM)
Sampling and Analytical Methods
Samples collected twice daily in
a 500 CC stainless steel cylinder.
Laboratory analysis by multiple-
column and single column gas
chromatography. Ethylene, carbon
dioxide, oxygen and argon, and
nitrogen are determined using a
Fisher dual-column ^as partitioner.
Column one (%" x 10;) contains 30%
HMPA on 60-80 mesh Chromosorb P and
is used to measure ethylene and
CO2. The second column partitions
oxygen and argon and nitrogen on
30-42 mesh, 13x molecular sieve
(V x 5') followed by 5' of uncoated
Chromosorb P. Ethylene oxide is
measured on a Hewlett-Packard Model
700 with single column of 20% carbowax
20M on 60-80 mesh gas pack F (V x 10')
Samples are analyzed for ethylene
on rare occasions by grab bag
collection for laboratory gas
chromatographic analysis.
One or two samples per month are
collected in a glass bulb for
analysis of ethylene by vapor-phase
chromatography. Flow is continuously
metered by an orifice.
One sample per month is collected in
a stainless steel bomb for analysis
as above. Flow is metered by orifice.
The gas stream is piped to the
laboratory for continuous analysis of
ethylene oxide by vapor-phase
chromatography. In addition, one
to two samples per month are collected
in a stainless bcmb for ethylene
analysis by GLC. Flow is monitored
by orifice.
Samples continuously pumped to
laboratory through a sample line
for ethylene oxide and ethylene
analysis by Bandix gas chromatograph.
Three columns are used, but their
function is unknown. Column No. 1
contains 20% of 2-pentrile A on
-------
EO-59
TABLE EO-19
SUMMARY OF EMISSIONS MEASUREMENT TECHNIQUES
USED IN ETHYLENE OXIDE PLANTS
Page 3 of 3
Plant
Number
13-6
(continued)
Source &
Flow Rate
Cycle Gas Exhaust
(64,000 SCFM)
Sampling and Analytical Methods
30-60 mesh chromosorb; No. 2
contains grade 15 silica gel; and
No. 3 is a 5A molecular sieve.
Occasional analyses by Orsat and
by MgCl£ for ethylene and ethylene
oxide are also run.
13-7
Mixed Gas Exhaust
Absorber Vent
Recovery Vent
(1,050 SCFM)
Occasional Orsat for gas composition.
Gas stream piped to gas chromatograph
for hourly analysis. Details unknown
Particulate measurements have been
made using in-stack alundum thimble.
Three analyses per year have been
made for ethylene oxide using a
mass spectrometer. Samples are
collected in a bomb.
-------
EO-60
XVII. Emergency Action Plant for Air Pollution Episodes
A. Types of Episodes
The alleviation of Air Pollution Episodes as suggested by the U.S.
Environmental Protection Agency is based on a pre-planned emission
reduction scheme. The criteria that set this scheme into motion are:
1. Alert Status - The alert level is that concentration of
pollutants at which short-term health effects can be
expected to occur.
2. Warning Status - The warning level indicates that air
quality is continuing to deteriorate and that additional
abatement actions are necessary.
3. Emergency Status - The emergency level is that level at which
a substantial endangerment to human health can be expected. These
criteria are absolute in the sense that they represent a level
of pollution that must not be allowed to occur.
B. Sources of Emission
As outlined in the foregoing in-depth study of ethylene oxide
manufacture by the direct oxidation of ethylene, there are three
continuous and some intermittent emissions occasioned by flaring or
venting gas during plant upsets.
1„ Continuous Streams
(a) Main Process Vent Gas - This stream constitutes the
greatest potential for air pollution. In the case of
the air oxidation scheme it is emitted from the secondary
or purge absorber. In the processing scheme using pure
oxygen, the main process vent stream is from the CO2
absorber. The composition and quantity of the main vent
streams vary significantly between the two processing
schemes that are used.
These streams are handled in different ways depending on
the processing scheme employed. In the air oxidation
route the stream, in some instances, is diverted to a
catalytic converter where hydrocarbons are converted to
the products of combustion. The entire stream at elevated
temperature is then used to drive a hot gas turbine before
exhausting to the atmosphere. In the processing scheme
using pure oxygen at least one operator adds methane to
the recycle. This acts as an "inert" at the conditions
of the process reactor but results in a vent gas that
is suitable as boiler fuel.
(b) Carbon Dioxide Rich Purge Gas » This stream, in the case
of the air oxidation scheme, is emitted overhead of the
ethylene oxide rectification or lights removal column.
Its composition is essentially N2, O2, CO2 and water with
-------
EO-61
some ethylene. The comparable stream in the oxygen scheme
of processing is emitted from the overhead of the regenera-
tor or stripper of the CO2 absorption system. This system
is to reduce CO2 concentration in the recycle gas. The
composition of the stream is essentially CO2 and water with
some ethane and ethylene present. These streams are
relatively small in volume with the most voluminous emission
from the oxygen mode of processing. These streams are
normally vented directly to the atmosphere. In some plants
the CO2 purge stream is scrubbed to recover ethylene oxide.
(c) Turbine Exhaust - The emission from this source is the
result of using hot gas turbines to drive the air compres-
sor and charge ethylene to the process. The turbines are
powered with natural gas and, since their efficiency is
something less than 100 percent, the stream emitting to the
atmosphere can be a significant contributor to the emission
of hydrocarbon pollutants. A fired gas turbine is used on
at least one air oxidation facility and should not be
mistaken for the hot gas turbine mentioned earlier. In
the latter instance, on at least two air oxidation
facilities, the main process vent stream passes through a
catalytic converter with its effluent, at elevated
temperature, driving the gas turbine.
2. Intermittent Air Emissions
(a) As in any process there are instances in the operation
whereby flaring or venting of process streams occur during
upset conditions for safety purposes or during shutdowns
when equipment is prepared for entering. In general these
are short periods with insignificant emissions on a yearly
basis. In most instances of shutdown venting, the emissions
could be curtailed if it should occur during an air pollution
episode. An exception to this regulation would be the
venting and safing of the reactors if it is deemed necessary
to prevent damage to the catalyst.
C. Abatement Techniques
As the various levels of the pre-planned episode reduction scheme are
declared (Alert, Warning and Emergency) a progressive reduction in the
amount of air pollutants must be made. This could ultimately lead to
total curtailment of pollutant emissions if the emergency level becomes
imminent.
Although these instructions for the "Air Pollution Episode Avoidance
Plan" are designed for ethylene oxide manufacturing plants, the overall
Emergency Action Plan (EAP) will cover all aspects of environmental air
pollution. Consequently, the implementation of the pre-planned episode
reduction scheme, as it applies to ethylene oxide manufacture, will be
in consideration of reductions made in all sources of air pollutants as
well as to the specific offending constitutents in the atmosphere. There-
fore, the extent of required cutback in emissions from ethylene oxide
plants will depend on the relative amounts of air pollutants contributed
by ethylene oxide production to the overall emissions which resulted in
the pollution episode. These factors will be used by the Governing
Environmental Protection Authority in determining the cutback to be made
in all air pollution sources during the various episodes.
-------
EO-62
Ethylene oxide manufacturing facilities by the direct oxidation of
ethylene consist of a bank of primary reactors, which are operated in
parallel, a primary absorber, a secondary (purge) ethylene conversion
system and a purification system for the removal of C02 and inert gases
and heavy ends. In the oxygen mode of operation the secondary system is
comprised of a CO2 absorber and regenerator for control of the CO2 in the
recycle gas stream. Although there are significant differences in the
design of the plants employing air oxidation versus the use of oxygen,
with respect to the recovery sections, the control of emissions during
air pollution episodes would be handled in essentially the same manner.
Based upon information obtained from several plants, it appears that
a reduction in operating rate results in reductions in emissions from
the purge absorber vent in air based plants. One respondent reported a
reduction in hydrocarbon emissions of 75 percent by a reduction in
production rate of 1/3 with further decreases in emissions by reducing
the production rate to the minimum operating level (50 percent of
design. However, it is felt that only minimum advantages would accrue
on those plants equipped with catalytic converters. Other respondents
also report a decrease in emissions with a turndown in production but
with a significantly smaller decrease in emissions. This also applies
to the CO2 purge vent on the plants that use oxygen to accomplish
the oxidation of ethylene. It was also reported that a small reduction
in hydrocarbon emissions from the CO2 purge stream was possible by
change in operating conditions. The flow from the ethylene oxide
rectification or lights removal column is reported as being proportional
to production rate. Consequently, a turndown in capacity could accomplish
a partial reduction in emissions during an air pollution episode. Under
normal conditions a turndown to a predetermined rate can be accomplished
with a 24 hour period. It is visualized that during this period there
would be a progressive decrease in emissions. Another method to obtain
a partial decrease in emissions would be a shutdown cf one or more
reactors. It is indicated, however, in one of the E.O. processes that
the decrease in emission would not be proportional tc the number of
reactors taken out of service. Matter of fact, it was stated that a
shutdown of one or more primary reactors has very little effect on
the emission of pollutants. Moreover, it may increase the emissivity for
a short period if it is deemed necessary to vent and safe the reactors
in order to protect the catalyst. Start-up of a reactor taken out of
service would depend on conditions maintained during the shutdown. Eight
to 16 hours would be required to resume operation.
It should be noted that the oxidation of ethylene is an exothermic
reaction with the exotherm consumed within the process to generate steam.
Any reduction in plant capacity could result in a steam deficient
condition within the confines of the E.O plant itself and possibly
over greater areas of a complex. Consequently, the steam load within
the central boiler plant would probably increase with the attendant
increase in emissions from this source.
There are several items of equipment that can be classed as emission
control devices among the reporting plants. Two respondents indicated
the use of catalytic converters on the secondary or purge absorber
vent. One respondent reported the use of a steam generator and two
use vent gas scrubbers on the CO2 purge stream while another reports the
use of a flare stack for emergencies. In plants employing catalytic
converters whose effluent is used to drive process turbines, a partial
-------
EO-63
reduction in plant capacity may adversely effect both the converter
and turbine. In these instances it seems reasonable to assume that some
standby source of power would be available as an alternative to drive
the turbine. In the case where a steam generator is partially dependent
on the combustibles contained in the vent stream, some adjustment would
seem to be in order to maintain the rate of steam generation. In
these instances, however, methane is normally injected to the lean
hydrocarbon vent stream for the required flammability limit. In a
partial reduction in plant capacity, the methane content of the gas
mixture will increase if the methane injection rate is held constant.
This results from less argon and nitrogen intake with lower oxygen use
rate. Also, CC>2 generation rate will decrease, therefore, the gas
mixture will be richer in methane. With respect to the vent gas
scrubbers on the CO2 purge stream, a reduction in total flow to the
scrubber should improve the efficiency of this equipment over that
obtained at normal ethylene oxide production levels.
1. Declaration of Alert Condition
When an alert condition in the atmosphere becomes apparent, the
episode emission reduction plan is immediately set into motion.
Under this plan the Environmental Protection Aathorities declare
that such a condition has developed and promptly notify the
manufacturers to proceed with their alert preparations. Under
this plan, depending on the rate at which the pollutant con-
centration is progressing, type of pollutant and the meteorology
potential that exists at the time, it may be deemed necessary
by the Environmental Protection Authorities to proceed with a
partial reduction in emissions from ethylene oxide manufacture
to prevent further increases in pollution level. This reduction
would be accomplished by a reduction in plant production as
previously discussed. The time required to affect the reduction
will be approximately as stated in the foregoing. This will
reduce the principal source of emission represented by the main
process vent from the purge absorber and the CO2 rich purge gas.
In the case of the exhaust from the hot gas turbine powered by
natural gas, a turndown of plant production would not necessarily
reduce total hydrocarbon emissions from this source. Most gas
driven turbines, however, are capable of some degree of turndown
that may be used during a partial reduction in production. In
this event a reduction in non-methane hydrocarbon emissions
would be anticipated. In those facilities that employ a
catalytic converter for the main process vent, a reduction in
emissions may have an adverse effect. This in turn could
present difficulties in the operation of the downstream gas
turbine driver. In this event pollutant emissions may actually
increase from this source if the utmost care is not exercised
in the operations directed to a partial reduction in plant
capacity.
2. Declaration of Warning Condition
If it becomes apparent that the efforts directed to the curtailment
of air pollutants have not resulted in an improvement in air
quality during the alert a warning condition is then declared.
When the air pollution warning episode is announced a substantial
reduction of air contaminants is desirable even to the point of
assuming reasonable economic hardship in the cutback of production
and allied operations. This could involve a 50-60 percent decrease
in ethylene oxide production.
-------
EO-64
3. Emergency Condtion
When air quality has deteriorated to a point where it appears
that an emergency episode is imminent, all air contaminants
may have to be eliminated immediately by ceasing production
and allied operations to the extent possible vjithout causing
injury to persons or damage to equipment.
D. Economic Considerations
The economic impact on ethylene oxide manufacturers of curtailing
operations during any of the air pollution episodes is based on the
duration and number of episodes in a given period. It is indicated
that the usual duration of air pollution episodes is one to seven days
with meteorology potentials as high as 80 per year.15 The frequency
of air pollution episodes in any given area is indicated as being one to
four years. These data do not differentiate between the episode levels
set forth in the early paragraphs of this section. Normally since the
alert level does not require a cutback in production, it will not
significantly influence plant economics. Therefore, In discussing
economic considerations resulting from the air pollution abatement plan,
it is only necessary to estimate the frequency and number of warning
and emergency episodes. For the economic study, it has been assumed
that three warning and no emergency episodes occur per year. Each
warning episode is assumed to require a 50 percent reduction in air
contaminants for a period of days. This equates to a complete loss
in plant production of about 8^ days per year.
The financial impact resulting from this loss in production is shown
in Table EO-20. This table presents comparative manufacturing costs of
both the oxygen and air oxidation modes of operation in typical existing
200 MM lbs./year facilities without extensive pollution control and
typical new plants of the same capacity. Economics are shown for each
of these with and without the financial impact accredited to the air
pollution episodes. It should be noted that whereas the proposed cutback
in ethylene oxide production for emission control appears small (2.5 percent
on a yearly basis), it reduced net income for the oxygen oxidation process'
by 20 percent and for the air oxidation process by 10 percent.
E. Summary of Estimated Emissions
In the foregoing a reduction in plant production was suggested to
obtain a reduced rate of air pollutant emissions for the various air
pollution levels that may be encountered. This was primarily predicated
on existing plants with no pollution control equipment. Therefore,
less stringent requirements should be provided in the EAP for Air
Pollution Episode Avoidance for existing plants that smploy control
devices which substantially reduce emissions. Also, for those existing
plants that install such equipment and for future plants that are
equipped with the "latest state of the art" emission control equipment.
Table EO-17 presents estimated 1985 air emissions for the present-day
systems without control devices versus emissions projacted for existing
modified facilities and new plants incorporating the most feasible devices.
The control devices assumed to be incorporated in the latter category are:
catalytic conversion of main process vents plus carbon dioxide vent
on all air oxidation plants and incineration of main process vent on all
oxygen oxidation plants. Emissions from the modified and new plants have
been estimated to be reduced substantially from the estimates for the
existing uncontrolled facilities.
-------
TABLE EQ-20
FINANCIAL IMPACT OF AIR POLLUTION EPISODES
ON MANUFACTURING COSTS
FOR 200 LBS./YEAR ETHYLENE OXIDE MANUFACTURING FACILITIES
VIA DIRECT OXIDATION OF ETHYLENE Page 1 of 2
OXYGEN OXIDATION PROCESS
AIR OXIDATION PROCESS
TYPICAL EXISTING PLANT
NEW PLANT WITH
STEAM GENERATING
TYPICAL EXISTING PLANT
ca'PMr
WITH
Ation
From
From
From
From
Table EO-13
Table EO-15
Table EO-13
Table EO-14
In
Production
Assuming 8.S
Days Lost
Production
No Cutback
In
Production
Assuming 8.S
Days Lost
Production
In
Production
Assuming 8.5
Days Lost
Production
In
Production
Assuming 8.5
Days Lost
Production
Direct Manufacturing Cost, M $/Yr.
Raw Materials
Ethylene
Oxygen
Catalyst 5 Chemicals
Labor
Maintenance
Utilities
Indirect Manufacturing Cost, M.$/Yr.
Plant Overhead
Fixed Manufacturing Cost, M $/Yr.
Depreciation, Insurance § Property Tax
7,600
2,000
400
620
260
1,200
680
640
7,410
1,950
390
620
260
1,170
680
640
7,600
2,000
400
620
260
1,220
680
640
7,410
1,950
390
620
260
1,190
680
640
7,600
0
400
620
500
1,200
680
1,220
7,410
0
390
620
500
1,170
680
1,220
7,600
0
400
620
510
1,150
680
1,260
7,410
0
390
620
510
1,120
680
1,260
-------
TABLE EQ-20
FINANCIAL IMPACT OF AIR POLLUTION EPISODES
ON MANUFACTURING COSTS
FOR 200 W LBS./YEAR ETHYLENE OXIDE MANUFACTURING FACILITIES
VIA DIRECT OXIDATION OF ETHYLENE page 2 of 2
PAGE 2 - CONTINUED
OXYGEN OXIDATION PROCESS AIR OXIDATION PROCESS
TYPICAL EXISTING PLANT
NEW PLANT WITH
STEAM GENERATING
TYPICAL EXISTING PLANT
NEW PUNT WITH
CAT. INCINERATION
From
Table EO-13
From
Table EO-1S
From
Table EO-13
From
Table HO-14
No Cutback
In
Production
Assuming 8.5
Days Lost
Production
No Cutback
In
Production
Assuming 8.5
Days Lost
Production
No Cutback
In
Production
Assuming 8.5
Days Lost
Production
No Cutback
In
Production
Assuming 8.5
Days Lost
Production
Manufacturing Cost, M $/Yr.
13,400
13,120
13,420
13,140
12,220
11,990
12,220
11,990
General Expenses
Administration, Sales, Research
1,100
1,100
1,120
1,120
1,320
1,320
1,340
1,340
and Finance
Total Cost, M $/Yr.
14,500
14,220
14,540
14,260
13,540 "
13,310
13,560
13,330
Product Value
Ethylene Oxide @ 7.5{/Lb.
15,000
14,625
15,000
14,625
15,000
14,625
15,000
14,625
Profit Before Taxes
500
405
460
365
1,460
1,315
1,440
1,295
NPAT
240
194
220
175
701
631
691
622
Cash F'ow
760
714
745
699
1,701
1,631
1,711
1,642
ROI
4.6%
3.7%
4.2%
3.3%
7.0%
6.3%
6.8%
6.1%
-------
EO-67
The particular type and concentration of pollutants in the atmosphere
at the time of the episode would dictate the degree to which a reduction
would be made on the existing modified plant or the new facility. If the
offending pollutants are in the form of hydrocarbons, the degree of cut-
back. on the modified or new plants could be proportionally less severe
than on the uncontrolled facility. If NOx is the offending material, a
reduction in plant production would reduce the amout of NOx being
emitted from thermal incinerators or flares by virtue of reduced
combustibles in their feed stream with attendant lowering of flame
temperature. NOx production is a rate controlled phenomenon which is
primarily controlled by flame temperature. Cooler flames and combustion
zones tend to produce lower NOx concentrations. It is reported that
substantially all of the NOx formed in a thermal device is formed in
the high temperature region (2800° F) of the burner flame itself. At
the temperature of the main residence-time section (1200-1500° F), the
overall rate of reaction of nitrogen with oxygen is too slow for
significant formation of NOx. It is for this reason that catalytic
converters produce little if any NOx.
-------
EO-68
References
1. Kirk-Othmer: "Encyclopedia of Chemical Technology", 2nd Edition, Vol. 8„
2. Dailey, W. V. , "Infra Red Analyzers for Ethylene Oxide Processes",
Chemical Engineering Progress, Vol. 68, No. 10, October, 1972.
3. Chemical Economics Handbook - Stanford Research Institute, No. 654,5030,
August, 1969.
4. The Petrochemical Industry: Markets & Economics, pages 274-291.
5. Chemical & Engineering News: December 18, 1972, page 10.
6. Personal Communications: L. B. Evans, EPA, Office of Air Programs,
Research Triangle Park, Trip Report: Houston, Port Neches Petrochemical
Plants, February 11, 1972.
7. Chemical Profile: Ethylene Oxide - Chemical Marketing Reporter (Oil,
Paint & Drug Reporter) October 1, 1969.
8. Landau, R., Ethylene Oxide by Direct Oxidation, Petroelum Refiner,
Vol. 32, November, 1953.
9. Private communications with all respondents regarding processing and
analytical procedures.
10„ Rolke, R. W., et al, "Afterburner Systems Study" by Shell Development
Company for Environmental Protection Agency (Contract EHSD-71-3).
11. "Control Techniques for Nitorgen Oxide Emissions from Stationary Sources",
National Air Pollution Control Administrations Publication No. AP-67.
12. Nelson: Oil & Gas Journal - Construction Indices for Petrochemical Plants.
13. 1972 Directory of Chemical Producers - Stanford Research Institute.
14. Private Communications with Catalytic, Inc. on incinerator costs and
performances.
15. "Guide for Air Pollution Episode Avoidance", Environmental Protection
Agency, Office of Air Programs, Publication No. AP-73, June, 1971.
-------
APPENDIX I
BASIS OF THE STUDY
I. Industry Survey
The study which led to this document was undertaken to obtain information
about selected production processes that are practiced in the Petrochemical
Industry. The objective of the study was to provide data for the EPA to use
in the fulfillment of their obligations under the Clean Air Amendments of 1970.
The information obtained during the study includes industry descriptions,
air emission control problems, sources of air emissions, statistics on quantities
and types of emissions and descriptions of emission control devices currently
in use. The principal source for these data was an Industry Questionnaire
but it was supplemented by plant visits, literature searches, in-house back-
ground knowledge and direct support from the Manufacturing Chemists Association.
More than 200 petrochemicals are currently produced in the United States,
and many of these by two or more different processes. It was obvious that
the most immediate need was to study the largest tonnage, fastest growth
processes that produce the most pollution. Consequently, the following 32
chemicals (as produced by a total of 41 different processes) were selected
for study:
Acetaldehyde (two processes)
Acetic Acid (three processes)
Acetic Anhydride
Aerylonitrile
Adipic Acid
Adiponitrile (two processes)
Carbon Black
Carbon Disulfide
Cyclohexanone
Ethylene
Ethylene Dichloride (two processes)
Ethylene Oxide (two processes)
Formaldehyde (two processes)
Glycerol
Hydrogen Cyanide
Maleic Anhydride
Nylon 6
Nylon 6,6
"Oxo" Alcohols and Aldehydes
Phenol
Phthalic Anhydride (two processes)
Polyethylene (high density)
Polyethylene (Iovj density)
Polypropylene
Polystyrene
Polyvinyl Chloride
Styrene
Styrene - Butadiene Rubber
Terephthalic Acid (1)
Toluene Di-isocyanate (2)
Vinyl Acetate (two processes)
Vinyl Chloride
(1) Includes dimethyl terephthalate.
(2) Includes methylenediphenyl and polymethylene polyphenyl isocyanates.
The Industry Questionnaire, which was used as the main source of information,
was the result of cooperative efforts between the EPA, Air Products and the
EPA's Industry Advisory Committee,, After receiving approval from the Office of
Management and Budget, the questionnaire was sent to selected producers of
most of the chemicals listed above. The data obtained from the returned
questionnaires formed the basis for what have been named "Survey Reports".
These have been separately published in four volumes, numbered EPA-450/3-73-005a,
b, and d and entitled "Survey Reports on Atmospheric Emissions from the
Petrochemical Industry - Volumes I, II, III, and IV.
-------
1-2
The purpose of the survey reports was to screen the various petrochemical
processes into the "more" and "less - significantly polluting processes".
Obviously, significance of pollution is a term which is difficult if not
impossible to define because value judgements are involved. Recognizing this
difficulty, a quantitative method for Significant Emission Index (SEI) was
developed. This procedure is discussed and illustrated in Appendix II of
this report. Each survey report includes the calculation of an SEI for the
petrochemical that is the subject of the report. These SEI's have been
incorporated into the Emission Summary Table that constitutes part of this
Appendix (Table I). This table can be used as an aid when establishing
priorities in the work required to set standards for emission controls on
new stationary sources of air pollution in accordance with the terms of the
Clean Air Amendments of 1970.
The completed survey reports constitute a preliminary data bank on each
of the processes studied. In addition to the SEI calculation, each report
includes a general introductory discussion of the process, a process description
(including chemical reactions), a simplified process flow diagram, as well as
heat and material balances. More pertinent to the air pollution study, each
report lists and discusses the sources of air emissions (including odors and
fugitive emissions) and the types of air pollution control equipment employed.
In tabular form, each reports summarizes the emission data (amount, composition,
temperature, and frequency); the sampling and analytical techniques; stack
numbers and dimensions; and emission control device data (types, sizes, capital
and operating costs, and efficiencies).
Calculation of efficiency on a pollution control device is not necessarily
a simple and straight-forward procedure. Consequently, two rating techniques
were developed for each type of device, as follows:
1„ For flares, incinerators, and boilers a Completeness of Combustion Rating
(CCR) and Significance of Emission Reduction Rating (SERR) were used.
2. For scrubbers and dust removal equipment, a Specific Pollutant
Efficiency (SE) and a SERR were used.
The bases for these ratings and example calculations are included in
Appendix III of this report.
II. In-Depth Studies
The original performance concept was to select a number of petrochemical
processes as "significant polluters", on the basis of data contained in
completed questionnaires. These processes were then to be studied "in-depth".
However, the overall time schedule was such that the EPA requested an initial
selection of three processes on the basis that they would probably turn out
to be "significant polluters". The processes selected in this manner were:
lo The Furnace Process for producing Carbon Black.
2„ The Sohio Process for producing Aerylonitrile.
3. The Oxychlorination Process for producing 1,2 Dichloroethane
(Ethylene Dichloride) from Ethylene.
-------
TABLE I
EMISSIONS SUMMARY
Page 1 of 3
ESTIMATED CURRENT AIR EMISSIONS, MM I.BS./YEAR
Hydrocarbons Particulates Oxides of Nitrogen Sull'ur Oxides Carbon Monoxide Tot a I Total Weighted
Acetaldehyde via Ethylene
1.1
0
0
0
0
1 . 1
80
via Ethanol
0
0
0
0
27
27
27
Acetic Acid via Methanol
0
0
0.01
0
0
0.01
1
via Butane
40
0
0.04
0
14
54
3,215
via Acetaldehyde
6. 1
0
0
0
1.3
7.4
4 90
Acetic Anhydride via Acetic Acid
3.1
0
0
0
5.5
8.6
253
Aerylonitrile (9)
183
0
5.5
0
196
385
15,000
Adipic Acid
0
0.2
29.6
0
0. 14
30
1,190
Adiponitrile via Butadiene
11.2
4.7
50.5
0
0
66.4
3,200
via Adipic Acid
0
0.5
0„04
0
0
0.54
30
Carbon Black
156
8.1
6.9
21.6
3,870
4 ,060
17,544
Carbon Disulfide
0.15
0.3
0.1
4.5
0
5. 1
1 20
Cyclohexanone
70
0
0
0
7 7.5
148
5, /00
Dimethyl Terephthalate (+TPA)
91
1.4
0. 1
1.0
53
146.5
7 ,4 60
Ethylene
15
0.2
0.2
2.0
0.2
17.6
1 ,240
Ethylene Dichloride via Oxychlorination
95.1
0.4
0
0
21.8
117.3
7,650
via Direct Chlorination
29
0
0
0
0
29
2,300
Ethylene Oxide
85.8
0
0.3
0.1
0
86.2
6,880
Formaldehyde via Silver Catalyst
23.8
0
0
0
107.2
13 L
1 ,955
via Iron Oxide Catalyst
2 5.7
0
0
0
24.9
50.6
2 ,070
Glycerol via Epichlorohydrin
16
0
0
0
0
16
1 ,280
Hydrogen Cyanide Direct Process
0.5
0
0.41
0
0
0.91
56
Isocyanates
1.3
0.8
0
0.02
86
88
231
Maleic Anhydride
34
0
0
0
260
294
2,950
Nylon 6
0
1.5
0
0
0
1.5
90
Nylon 6,6
0
5.5
0
0
0
5.5
330
Oxo Process
5.25
0.01
0.07
0
19.5
24.8
440
Phenol
24.3
0
0
0
0
24.3
1,940
Phthalic Anhydride via O-Xylene
0.1
5.1
0,3
2.6
43.6
51.7
422
via Naphthalene
0
1.9
0
0
45
47
160
High Density Polyethylene
79
2.3
0
0
0
81.3
6,400
Lov Density Polyethylene
75
1.4
0
0
0
76.4
6,100
Polypropylene
37.5
0.1
0
0
0
37.6
2 ,950
Polystyrene
20
0.4
0
1.2
0
21.6
1,650
Polyviny1 Chloride
62
12
0
0
0
74
5,700
Styrene
4.3
0.07
0.14
0
0
4.5
355
Styrene-Butadiene Rubber
9.4
1.6
0
0.9
0
12
870
Vinyl Acetate via Acetylene
5.3
0
0
0
0
5.3
425
via Ethylene
0
0
TR
0
0
TR
TR
Vinyl Chloride
17.6
0.6
_0
JD
0
18.2
1 ,4 60
Totals
1,227.6
49. 1
94-.2
33.9
4,852.6
6,225-9
110,220
(1) In aost instances numbers are based on less than 100% survey. All based on engineering judgement of best current control. Probably has up to 10/. lov bias.
(2) Assumes future plants vill employ best current control techniques.
(3) Excludes methane, includes H2S and all volatile organics.
(4) Includes non-volatile organics and inorganics.
(5) Weighting factors used are: hydrocarbons - 80, particulates - 60, N0X - 40, S0X - 20, and CO - 1.
(6) Referred to elsewhere in this study as "Significant Emission Index" or "SEI".
(7) Totals are not equal across and down due to rounding.
(9) Emissions based on what is nov an obtoltte catalyst. See Report No. EPA-450/3-73-006 b for up-to-date information.
-------
TABLE I
EMISSION SUMMARY
Page 2 ot 3
ESTIMATED ADDITIONAL AIR EMISSIONS IN 1980, MM LBS./YEAR
Hydrocarbons Particulates Oxides of Nitrogen Sulfur Oxides Carbon Monoxide Tot a 1 Total Weighted (5>6)
Acetaldehyde via Ethylene
1.2
0
0
0
0
1.2
96
via Ethanol
0
0
0
0
0
0
0
Acetic Acid via Methanol
0
0
0.04
0
0
0.04
2
via Butane
0
0
0
0
0
0
0
via Acetaldehyde
12.2
0
0
0
2.5
14.7
980
Acetic Anhydride via Acetic Acid
0.73
0
0
0
1.42
2.15
60
Aerylonitrile (9)
284
0
8.5
0
304
596
23,000
Adipic Acid
0
0.14
19.3
0
0.09
19.5
779
Adiponitrile via Butadiene
10.5
4.4
47.5
0
0
62 .4
3,010
via Adipic Acid
0
0.5
0.04
0
0
0.54
30
Carbon Black
64
3.3
2.8
8.9
1,590
1,670
7 ,200
Carbon Disulfide
0.04
0.07
0.03
1.1
0
1.24
30
Cyclohexanone
77.2
0
0
0
85.1
162
6,2 60
Dimethyl Terephthalate (+TPA)
73.8
1.1
0.07
0.84
42.9
118.7
6,040
Ethylene
14.8
0.2
0.2
61.5
0.2
II
2,430
Ethylene Dichloride via Oxychlorination
110
0.5
0
0
25
136
8,800
via Direct Chlorination
34.2
0
0
0
0
34.2
2,740
Ethylene Oxide
32.8
0
0.15
0.05
0
33
2,650
Formaldehyde via Silver Catalyst
14.8
0
0
0
66.7
81.5
1,250
via Iron Oxide Catalyst
17.6
0
0
0
17.0
34.6
1 ,445
Glycerol via Epichlorohydrin
8.9
0
0
0
0
8.9
700
Hydrogen Cyanide Direct Process
0
0
0
0
0
0
0
Isocyanates
1.2
0.7
0
0.02
85
87
225
Maleic Anhydride
31
0
0
0
241
272
2,720
Nylon 6
0
3.2
0
0
0
3.2
194
Nylon 6,6
0
5.3
0
0
0
5.3
318
Oxo Process
3.86
0.01
0.05
0
14.3
18.2
325
Pheno1
21.3
0
0
0
0
21.3
1,704
Phthalic Anhydride via O-Xylene
0.3
13.2
0.8
6.8
113
134
1,100
via Naphthalene
0
0
0
0
0
0
0
High Density Polyethylene
210
6.2
0
0
0
216
17,200
Lov Density Polyethylene
262
5
0
0
0
267
21,300
Polypropylene
152
0.5
0
0
0
152.5
12,190
Polystyrene
20
0.34
0
1.13
0
21.47
1,640
Polyvinyl Chloride
53
10
0
0
0
63
4,840
Styrene
3.1
0.05
0.1
0
0
3.25
225
Styrene-Butadiene Rubber
1.85
0.31
0
0.18
0
2.34
1 70
Vinyl Acetate via Acetylene
4.5
0
0
0
0
4.5
3 60
via Ethylene
0
0
TR
0
0
TR
TR
Vinyl Chloride
26.3
0.9
0
_0
0
27.2
2,170
Totals
1,547,2
55.9
79.5
80.5
2,588
4,351.9
136,213 (7)
(1) In most instances numbers are based on lest
» than 100% survey. All
based on engineering
judgement of best
current control.
Probably has
up to 107, low bias.
(2) Assumes future plants will employ best current control techniques.
(3) Excludes methane, includes HoS and all volt
itile organics.
(4) Includes non-volatile organic® and inorganics.
(5) Weighting factors used are: hydrocarbons - 80, particulates - 60, N0X - 40, S0X - 40, and CO - I.
(6) Referred to clsevhcre in this study as "Significant Emission Index" or "SEI".
(7) Totals are not equal across and dovn duv to rounding.
(9) See sheet 1 of 3.
-------
TABLE I
EMISSIONS SUMMARY Page 3 of 3
Emissions (2) , MM Lbs. /Year Tcital Estimated Capacity
Est'[mated Number of New Plants MM LbF./Yonr
Total by 1980
Total Weighted (5) by 1980
(1973 - 1980)
Current
By 1980
Acetaldehyde via Ethylene
2.3
182
6
1,160
2,4 00
via Ethanol
27
27
0
966
966
Acetic Acid via Methanol
0.05
3
4
400
1 ,800
via Butane
54
3,215
0
1 ,020
500
via Acetaldehyde
22
1,470
3
875
2,015
Acetic Anhydride via Acetic Acid
10.8
313
3
1 , /05
2,100
Aerylonitrile (9)
980
38,000
5
1 ,165
3,700 (8)
Adipic Acid
50
1,970
7
1 ,430
2 ,200
Adiponitrile via Butadiene
128.8
6,210
4
435
84 5
via Adipic Acid
1.1
60
3
280
550
Carbon Black
5,730
24,740
13
3 ,000
5,000 (8)
Carbon Disulfide
6.3
150
2
871
1, 100
Cyclohexanone
310
11,960
10
1 ,800
3 , bOO
Dimethyl Terephthalate (+TPA)
2 65
13,500
8
2,865
5,900
Ethylene
94
3,670
21
22 ,295
40,000
Ethylene Dichloridc via Oxychlorination
253
16,450
8
4,4 50
8,250 (8)
via Direct Chlorination
63
5,040
10
5,593
11 ,540
Ethylene Oxide
120
9,530
15
4,191
6,800 (8)
Formaldehyde via Silver Catalyst
212.5
3,205
40
5,914
9,000
via Iron Oxide Catalyst
85
3,515
12
1 , 729
3,520 (8)
Glycerol via Epichlorohydrin
25
2,000
1
245
380
Hydrogen Cyanide Direct Process
0.5 (10)
28 (10)
0
412
202
Isocyanatcs
175
456
10
1 ,088
2 ,120
Maleic Anhydride
566
5,670
6
359
720
Nylon 6
4.7
284
10
486
1,500
Nylon 6,6
10.8
650
10
1 ,523
3 ,000
Oxo Process
43
765
6
1 ,72/
3,000
Phenol
46
3 ,640
11
? ,3b3
4 ,200
Phthalic Anhydride via O-Xylene
186
1 ,522
6
720
1,800 (8)
via Naphthalene
47
160
0
603
528
High Density Polyethylene
297
23,600
31
2,315
8,500
Low Density Polyethylene
343
27,400
41
5,269
21,100
Polypropylene
190
15,140
32
1 ,160
5,800
Polystyrene
43
3,290
23
3,500
6,700
Polyvinyl Chloride
137
10,540
25
4,3/5
8,000
Styrene
7.4
610
9
5 ,953
10,000
Styrane-Butadiene Rubber
14
1,040
4
4 ,464
5,230
Vinyl Acetate via Acetylene
9.8
785
1
206
3 5 0
via Ethylene
TR
TR
4
1,280
2 ,200
Vinyl Chloride
4 5
3,630
10
5,400
13,000
Totals
10,605 (7)
244,420 (7)
(1) In most instances numbers are based on less than 100% survey. All based on engineering judgement of best current control. Probably has up to 10/ low bias.
(2) Assumes future plants will employ best current control techniques.
(3) Excludes methane, includes H2S and all volatile organics.
(4) Includes non-volatile organics and inorganics.
(5) Weighting factors used are: hydrocarbons - 80, particulates - 60, N0X - 40, S0X - 20, and CO - 1.
(6) Referred to elsewhere in thi6 study as "Significant Emission Index" or "SEI".
(7) Totals are not equal across and down due to rounding.
(8) By 1985.
(9) See sheet 1 o£ 3
(10) Due to anticipated future shut down of marginal plants.
-------
1-6
In order to obtain data on these processes, the operators and/or
licensors of each were approached directly by Air Products' personnel.
This, of course, was a slow and tedious method of data collection because
mass mailing techniques could not be used, nor could the request for data
be identified as an "Official EPA Requirement". Yet, by the time that OMB
approval was given for use of the Industry Questionnaire, a substantial
volume of data pertaining to each process had already been received. The
value of this procedure is indicated by the fact that first drafts of these
three reports had already been submitted to the EPA, and reviewed by the
Industry Advisory Committee, prior to the completion of many of the survey
reports.
In addition, because of timing requirements, the EPA decided that three
additional chemicals be "nominated" for in-depth study. These were phthalic
anhydride, formaldehyde and ethylene oxide. Consequently, four additional
in-depth studies were undertaken, as follows:
1. Air Oxidation of Ortho-Xylene to produce Phthalic Anhydride.
2. Air Oxidation of Methanol in a Methanol Rich Process to produce
Formaldehyde over a Silver Catalyst. (Also, the subject of a
survey report.)
3. Air Oxidation of Methanol in a Methanol-Lean Process to
produce Formaldehyde over an Iron Oxide Catalyst.
4. Direct Oxidation of Ethylene to produce Ethylene Oxide.
The primary data source for these was the Industry Questionnaire,
although SEI rankings had not been completed by the time the choices were
made.
The Survey Reports, having now been completed are available, for use in
the selection of additional processes for in-depth study.
-------
INTRODUCTION TO APPENDIX II AND III
The following discussions describe techniques that were developed for
the single purpose of providing a portion of the guidance required in the
selection of processes for in-depth study. It is believed that the underlying
concepts of these techniques are sound. However, use of them without sub-
stantial further refinement is discouraged because the data base for their
specifics is not sufficiently accurate for wide application. The subjects
covered in the Appendix II discussion are:
1. Prediction of numbers of new plants.
20 Prediction of emissions from the new plants on a weighted
(significance) basis.
The subject covered in the Appendix HI discussion is:
Calculation of pollution control device efficiency on a variety of
bases, including a weighted (significance) basis.
It should be noted that the weighting factors used are arbitrary.
Hence, if any reader of this report wishes to determine the effect of
different weighing factors, the calculation technique permits changes in
these, at the reader's discretion.
-------
APPENDIX II
Number of New Plants*
Attached Table 1 illustrates the format for this calculation.
Briefly, the procedure is as follows:
1. For each petrochemical that is to be evaluated, estimate what
amount of today's production capacity is likely to be on-stream
in 1980. This will be done by subtracting plants having marginal
economics due either to their size or to the employment of an
out-of-date process.
2. Estimate the 1980 demand for the chemical and assume a 1980
installed capacity that will be required in order to satisfy
this demand.
3. Estimate the portion of the excess of the 1980 required capacity
over today's remaining capacity that will be made up by
installation of each process that is being evaluated.
4. Estimate an economic plant or unit size on the basis of today's
technology.
5. Divide the total required new capacity for each process by the
economic plant size to obtain the number of new units.
In order to illustrate the procedure, data have been incorporated
into Table I, for the three processes for producing carbon black, namely
the furnace process, the relatively non-polluting thermal process, and
the non-growth channel process.
*The format is based on 1980, but any future year may be selected.
-------
Table 1. Number of New Plants by 1980
Chemical
Process
Current
Capacity
Marginal
Capacity
Current
Capacity
on-stream
in 1980
Demand
1980
Capacity
1980
Capacity
to be
Added
Economic
Plant
Size
Number of
New
Units
Carbon Black
Furnace
4,000
0
4,000
4,500
5,000
1,000
90
11 - 12
Channel
100
0
100
100
100
0
30
0
Thermal
200
0
200
400
500
300
150
2
Notes: 1. Capacity units all in MM lbs./year.
2. 1980 demand based on studies prepared for EPA by Processes Research, Inc. and MSA Research Corporation.
-------
II-3
Increased Emisaions (Weighted) by 1980
Attached Table 2 illustrates the format for this calculation.
However, more important than format is a proposal for a weighting basis.
There is a wide divergence of opinion on which pollutants are more noxious
and even when agreement can be reached on an order of noxiousness, dis-
agreements remain as to relative magnitudes for tolerance factors. In
general pollutants from the petrochemical industry can be broken down into
categories of hydrogen sulfide, hydrocarbons, particulates, carbon monoxide,
and oxides of sulfur and nitrogen. Of course, two of these can be further
broken down; hydrocarbons into paraffins, olefins, chlorinated hydrocarbons,
nitrogen or sulfur bearing hydrocarbons, etc. and particulates into ash,
catalyst, finely divided end products, etc. Itwasfelt that no useful
end is served by creating a large number of sub-groupings because it would
merely compound the problem of assigning a weighting factor. Therefore,
it was proposed to classify all pollutants into one of five of the six
categories with hydrogen sulfide included with hydrocarbons.
There appears to be general agreement among the experts that carbon
monoxide is the least noxious of the five and that N0X is somewhat more
noxious than S0X„ However, there are widely divergent opinions concerning
hydrocarbons and particulates - probably due to the fact that these are
both widely divergent categories. In recent years, at least two authors
have attempted to assign tolerance factors to these five categories.
Babcock (1), based his on the proposed 1969 California standards for
one hour ambient air conditions with his own standard used for hydrocarbons.
On the other hand, Walther (2), based his ranking on both primary
and secondary standards for a 24-hour period. Both authors found it
necessary to extrapolate some of the basic standards to the chosen time
period. Their rankings, on an effect factor basis with carbon monoxide
arbitrarily used as a reference are as follows:
Babcock
Walther
Hydrocarbons
Particulates
N0X
S0X
CO
2.1
107
77 o 9
28.1
1
Primary
125
21.5
22.4
15.3
1
Secondary
125
37.3
22.4
21.5
1
Recognizing that it is completely unscientific and potentially subject
to substantial criticism it was proposed to take arithmetic averages of the
above values and round them to the nearest multiple of ten to establish a
rating basis as follows:
Average Rounded
Hydrocarbons 84.0 80
Particulates 55.3 60
N0X 40o9 40
S0„ 21.6 20
CO 11
-------
Table 2. Weighted Emission Rates
Chemical_
Process
Increased Capacity_
Pollutant
Emissions, Lbs./Lb.
Increased Emissions
Lbs,/Year
Weighting
Factors
Weighted Emissions
Lbs./Year
Hydrocarbons
Particulates
N0X
S0X
CO
80
60
40
20
1
Total
-------
II-5
Increased Emissions (Weighted) by 1980 (continued)
This ranking can be defended qualitatively, if not quantitatively for
the following reasons:
1. The level of noxiousness follows the same sequence as is obtained
using national air quality standards,
2„ Approximately two orders of magnitude exist between top and bottom
rankings.
3„ Hydrocarbons should probably have a lower value than in the
Walther analysis because such relatively non-noxious compounds
as ethane and propane are included.
4. Hydrocarbons should probably have a higher value than in the
Babcock analysis because such noxious (or posionous) substances
as aromatics, chlorinated hydrocarbons, phenol, formaldehyde, and
cyanides are included„
5. Particulates should probably have a higher value than in the
Walther analysis because national air standards are based mostly
on fly ash while emissions from the petrochemical industry are
more noxious being such things as carbon black, phthalic anhydride,
PVC dust, active catalysts, etc.
6. NOx should probably have a higher value than in the Walther
analysis because its role in oxidant synthesis has been neglected.
This is demonstrated in Babcock's analysis.
Briefly, the procedure, using the recommended factors and Table 2, is
as follows:
1. Determine the emission rate for each major pollutant category in
terms of pounds of pollutant per pound of final product. (This
determination was made^ on the basis of data reported on returned
questionnaires^in the Survey Reports/.
2„ Multiply these emission rates by the estimate of increased production
capacity to be installed by 1980 (as calculated while determining
the number of new plants), to determine the estimated pounds of
new emissions of each pollutant.
3. Multiply the pounds of new emissions of each pollutant by its
weighting factor to determine a weighted pounds of new emissions
for each pollutant,,
4. Total the weighted pounds of new emissions for all pollutants to
obtain an estimate of the significance of emission from the process
being evaluated. It was proposed that this total be named
"Significant Emission Index" and abbreviated "SEI".
It should be pointed out that the concepts outlined above are not
completely original and considerable credit should be given to Mr. L. B„ Evans
of che EPA for setting up the formats of these evaluating procedures.
-------
II-6
Increased Emissions (Weighted) by 1980 (continued)
Babcock, L. F., "A Combined Pollution Index for Measurement of Total
Air Pollution," JAPCA, October, 1970; Vol, 20, No. 10; pp 653-659
Walther, E. G., "A Rating of the Major Air Pollutants and Their Sources
by Effect", JAPCA, May, 1972; Vol. 22, No. 5; pp 352-355
-------
Appendix III
Efficiency of Pollution Control Devices
Incinerators and Flares
The burning process is unique among the various techniques for
reducing air pollution in that it does not remove the noxious substance
but changes it to a different and hopefully less noxious form. It can be,
and usually is, a very efficient process when applied to hydrocarbons,
because when burned completely the only products of combustion are carbon
dioxide and water. However, if the combustion is incomplete a wide range
of additional products such as cracked hydrocarbons, soot and carbon
monoxide might be formed. The problem is further complicated if the
hydrocarbon that is being burned is halogenated, contains sulfur or is
mixed with hydrogen sulfide, because hydrogen chloride and/or sulfur oxides
then become products of combustion. In addition, if nitrogen is present,
either as air or nitrogenated hydrocarbons, oxides of nitrogen might be
formed, depending upon flame temperature and residence time.
Consequently, the definition of efficiency of a burner, as a pollution
control device, is difficult. The usual definition of percentage removal of
the noxious substance in the feed to the device is inappropriate, because
with this definition, a "smoky" flare would achieve the same nearly 100
percent rating, as a "smokeless" one because most of the feed hydrocarbon
will have either cracked or burned in the flame. On the other hand, any
system that rates efficiency by considering only the total quantity of
pollutant in both the feed to and the effluent from the device would be
meaningless. For example, the complete combustion of one pound of hydrogen
sulfide results in the production of nearly two pounds of sulfur dioxide, or
the incomplete combustion of one pound of ethane could result in the
production of nearly two pounds of carbon monoxide.
For these reasons, it was proposed that two separate efficiency rating
be applied to incineration devices. The first of these is a "Completeness
of Combustion Rating" and the other is a "Significance of Emission Reduction
Rating", as follows:
1o Completeness of Combustion Rating (CCR)
This rating is based on oxygen rather than on pollutants and is
the pounds of oxygen that react with the pollutants in the feed to
the device, divided by the theoretical maximum number of pounds that
would react: Thus a smokeless flare would receive a 100 percent
rating while a smoky one would be rated somewhat less, depending upon
how incomplete the combustion.
In utilizing this rating, it is clear that carbon dioxide and water
are the products of complete combustion of hydrocarbons. However, some
question could occur as to the theoretical completion of combustion
when burning materials other than hydrocarbons. It was recommended
that the formation of HX be considered complete combustion of halogenated
hydrocarbons since the oxidation most typically does not change the
valence of the halogen. On the other hand, since some incinerators will
be catalytic in nature it was recommended that sulfur trioxide be
considered as complete oxidation of sulfur bearing compounds.
-------
III-2
Efficiency of Pollution Control Devices
1. Completeness of Combustion Rating (CCR) (continued)
Nitrogen is more complex, because of the equilibria that exist
between oxygen, nitrogen, nitric oxide, nitrogen dioxide and the
various nitrogen radicals such as nitrile. In fact, many scientists
continue to dispute the role of fuel nitrogen versus ambient nitrogen
in the production of N0X. In order to make the CCR a meaningful
rating for the incineration of nitrogenous wastes it was recommended
that complete combustion be defined as the production of N2, thus
assuming that all N0X formed comes from the air rather than the fuel,
and that no oxygen is consumed by the nitrogen in the waste material.
Hence, the CCR becomes a measure of how completely the hydrocarbon
content is burned, while any N0X produced (regardless of its source)
will be rated by the SERR as described below.
2. Significance of Emission Reduction Rating (SERR)
This rating is based primarily on the weighting factors that
were proposed above. All air pollutants in the feed to the device
and all in the effluents from the device are multiplied by the
appropriate factor. The total weighted pollutants in and out are
then used in the conventional manner of calculating efficiency
of pollutant removal, that is pollutants in minus pollutants out,
divided by pollutants in, gives the efficiency of removal on a
significance of emission basis.
Several examples will serve to illustrate these rating factors,
as follows:
Example 1 - One hundred pounds of ethylene per unit time is burned
in a flare, in accordance with the following reaction;
3C2H4 + 7 02 ~ C + 2 CO + 3 C02 + 6 H20
Thus, 14.2 lbs. of particulate carbon and 66.5 lbs. of carbon
monoxide are emitted, and 265 lbs. of oxygen are consumed.
Theoretical complete combustion would consume 342 lbs. of oxygen
in accordance with the following reaction:
C2H4 + 3 02 ~~~~2 C02 + 2 H20
Thus, this device would have a CCR of 265/342 or 77.5%
Assuming that one pound of nitric oxide is formed in the reaction
as a result of the air used for combustion (this is about equivalent to
100 ppm), a SERR can also be calculated. It should be noted that the
formation of this NO is not considered in calculating a CCR because it
came from nitrogen in the air rather than nitrogen in the pollutant
being incinerated. The calculation follows:
-------
III-3
Efficiency of Pollution Control Devices
2. Significance of Emission Reduction Rating (SERR) (continued)
Weighting Pounds in Pounds out
Pollutant Factor Actual Weighted Actual Weighted
Hydrocarbons 80 100 8000 0
Particulates 60 0 14.2 852
N0X 40 0 1 40
S0X 20 0 0
CO 10 66 o 5 66.5
Total 8000 958.5
SERR = 8000 - 958.5
8000
x 100 = 88%
Example 2 - The same as Example 1, except the hydrocarbons are
burned to completion. Then,
CCR ~ ftf x 100 = 100%
and
SERR = 8000 - 40
8000
= 99.5%
Example 3 - One hundred pounds per unit time of methyl chloride is
incinerated, in accordance with the following reaction.
2 CH3CI + 3 02 > 2 C02 + 2 H20 + 2 HC1
This is complete combustion, by definition, therefore, the CCR is
100%. However, (assuming no oxides of nitrogen are formed), the SERR
is less than 100%, because 72.5 lbs. of HC1 are formed. Hence,
considering HCl as an aerosol or particulate;
SERR = 100 x 80 - 72.5 x 60 10Q = 45 y/
100 x 80
The conclusion from this final example, of course, is that it is
an excellent combustion device but a very poor pollution control device,
unless it is followed by an efficient scrubber for HCl removal.
Example 4 - The stacks of two hydrogen cyanide incinerators, each
burning 100 pounds per unit time of HCN are sampled. Neither has any
carbon monoxide or particulate in the effluent. However, the first is
producing one pound of N0X and the second is producing ten pounds of
N0X in the same unit time. The assumed reactions are:
-------
III-4
Efficiency of Pollution Control Devices
2. Significance of Emission Reduction Rating (SERR) (continued)
4 HCN + 5 02 1 1 ~ 2 H20 + 4 C02 + 2 N2
N2 (atmospheric) + X02 11 V 2 N0x
Thus, CCR^ = 1007o and CCR2 = 1007„ both by definition.
However, SERR^ = 100 x 80 - 1 x 40 ^
100 x 80
and SERRo = 100 x 80 - 10 x 40 .
2 100 X 80 x 100 = 95/°
Obviously, if either of these were "smoky" then both the CCR and
the SERR would be lower, as in Example 1.
Other Pollution Control Devices
Most pollution control devices, such as bag filters, electrostatic
precipitators and scrubbers are designed to physically remove one or more
noxious substances from the stream being vented. Typically, the efficiency
of these devices is rated relative only to the substance which they are
designed to remove and for this reason could be misleading. For example:
1. The electrostatic precipitator on a power house stack might be
99% efficient relative to particulates, but will remove little
or none of the S0X and NOx which are usually present.
2„ A bag filter on a carbon black plant will remove 99 + % of the
particulate but will remove none of the CO and only relatively
small amounts of the compounds of sulfur that are present.
3. A water scrubber on a vinyl chloride monomer plant will remove
all of the hydrogen chloride but only relatively small amounts
of the chlorinated hydrocarbons present.
4. An organic liquid scrubber on an ethylene dichloride plant will
remove nearly all of the EDC but will introduce another pollutant
into the air due to its own vapor pressure.
For these reasons, it was suggested again that two efficiency ratings be
applied. However, in this case, the first is merely a specific efficiency as
is typically reported, i.e., "specific to the pollutant (or pollutants) for
which it was designed", thus:
SE = specific pollutant in - specific pollutant out
specific pollutant in
The second rating proposed is an SERR, defined exactly as in the case
of incinerators.
Two examples will illustrate these ratings.
-------
m-5
Efficiency of Pollution Control Devices
Other Pollution Control Devices (continued)
Example 1 - Assume that a catalytic cracker regenerator effluent
contains 100 pounds of catalyst dust, 200 lbs. of
carbon monoxide and 10 pounds of sulfur oxides per unit
time. It is passed through a cyclone separator where
95 pounds of catalyst are removed. Therefore,
SE = 100 - 5
100
x 100 = 95%
and SERR = (100 x 60 + 10 x 20 + 200 x 1) - (5 x 60 + 10 x 20 + 200 x 1) x 100
(100 x 60 + 10 x 20 + 200 x 1)
= 6400 - 700 x 100 = 89%
6400
Example 2 - Assume that an organic liquid scrubber is used to wash a
stream containing 50 pounds of SO2 per unit time. All
but one pound of the SO2 is removed but two pounds of
the hydrocarbon evaporate into the vented stream. Then
SE 5°5o" 1 xlQ° = 987°
and SERR = (50 x 20) - (1 x 20 + 2 x 80)
(50 x 20) x 100
= 1000 - 180 10Q = 827
1000
------- |