PB83-11011B
Engineering Assecsment of
CDB (Ethylenc Dibromide) Pesticide
Pesticide Destruction Technologies
PCI Associates, Inc., Cincinnati, OH
Prepared for
Environmental Protection Agency, Cincinnati, Oil
Sep 00
EPA
600/2
Sa-056
U.S. Department of Commerce
National Technical Information Service
-------
TECHNICAL REPORT DATA
(Pleatr rttd Inttnicnom on the rttrrtr bffort compttt&ig)
4 TITLE AND SUBTITLE
KNClNliTRlNO ASSESSMENT OF
TECHNOLOGIES
EDB PESTICIDE DESTRUCTION
5 REPORT DATE
Septenber 1983
* PERFORMING ORGANIZATION CODE
7 AUTMORlSI
Sunil H. Ambekar
Fernard A. Laseke
*. PERFORMING ORGANIZATION REPORT NO
PERFORMING ORGANIZATION NAME AND ADDRESS
PKI Associates, Inc.
11^99 Chester Rd.
Cincinnati, OH 45246
10. PROGRAM ELEMENT NO.
11. CONTRACT /GRANT NO.
68-03-3389
WA #1-6
I! SPONSORING AGENCY NAME AND ADDRESS
.Risk Reduction Engineering Laboratory
Office of Research and Development
U.S. Environmental Protection Agency
Cincinnati, OH 45268
13. TYPE OF REPORT AND PERIOD COVERED
14. SPONSORING AGENCY CODE
EPA/600/14
is SUPPLEMENTARY NOTES
16 ABSTRACT
The project covered by this report involved an engineering evaluation of the
suitability of various available technologies for the destruction of ethylene
dibromide pesticides. The purpose of the study was to highlight-the technical
merits and shortcomings, safety, cost, and total time requirement for each of the
alternatives considered.
Both thermal and chemical destruction options were considered. Evaluation
criteria were developed so th^it. the different options could be compared on a common
basis. Information was collected on each candidate process through a literature
search and discussions with industry experts. Concurrent with these efforts, bench-
scale tests of the chemical methods were conducted. Also, a test burn was made at a
commercial facility to determine the effectiveness of one of the incineration
options. The results of these tests were factored into this report. Because the
chemical processes ar. still in the conceptual stages, only preliminary process
calculations and cost estimates were developed for these processes.
Bnsed on the results of this study, incineration In the presence of sulfur
dioxide appears to be the best alternative for the safe, effective, rapid, and
economical destruction of the ethylene dibromide pesticides.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
b IDENTIFIERS/OPEN ENDED TERMS C. COSATI f leld/Gloup
IB. DISTRIBUTION STATEMENT
RELEASE TO PUBLIC
19. SECURITY CLASS (Tttit Rtpon)
UNCLASSIFIED
21. NO. OF PAGE£
148
20. SECURITY CLASS tThilp»ft)
UNCLASSIFIED
22. PRICE
EPA Form 2220-1 (R«». 4-77) PHCVIOUS COITION n OBSOLETE
-------
?bo 1-
1'J
EPA/600/2-00/056
September 108»
ENGINEERING ASSESSMENT OF
EDP PESTICIPC DESTRUCTION TECHNOLOGIES
by
Sum"! H. Ambekar and Bernard A. Laseke
PEI Associates, Inc.
11499 Chester Road
Cincinnati, Ohio 45246
EPA Contract No. 68-03-3389
Project Officer Edward R. Bates
Risk Reduction Engineering Laboratory (RREL)
U.S. Environmental Protection Agency
Cincinnati, Ohio 45268
RISK REDUCTION ENGINEERING LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI. OH 45268
REPRODUCED BY
U.S. DEPARTMENT OF COMMERCE
NATIONAL TECHNICAL INFORMATION SERVICE
-------
NOTICE
This document has been reviewed in accordance with U.S. Environmental
Protection Agency policy and approved for publication. Mention of trade
names or commercial products does not constitute endorsement or rscomrnenda-
tion for use.
ii
-------
'• • i ' I . I I ,. , .1 .. ! i I •• i.. r i . | I • • • .• r i i . I • • •• ' '
'I i ' <• ,' , I . ; ' I : • I . '. I i. I ' • >l • ' I I. • - I i l I ! t , .
• : ' ' i I I l,.! ' i II. , ' . I •.. • i I. ' ..
i ., . i ; ,, • , • i . • i' i •
-.'..,, I . . .,,, .,!;!,'.. I . . .- ,, .
I . ' ! i i . . • t I i • I i! , ' •!,' >. . .
I. I ... I . .,,..(,,
] l )|i|i' '.I . -II' ' ' II'. ' I 'I
I . ' • . 'I ' -I ' ,M ., i 14. I" i. I • I I '•• .11 i.! I . I..' . • • ; , "!• -1,1.
l .1 • • ,i ' • ' i i 'i • i • I-. i! . .' i I ! i. if l l. -it ; vi .. -\\- I • -. I-TJ i : i "
' . i ' - • ' ' • , ' , t ' ! i . . , i ' i • >' l; ' i • , *.' . < .
I
I
i l> .111 III'. VJ id ' 'i „ l-Jr' - f "••' '•••(.
• i I i .: . i . i • i •. ,• M .! ji i . . g ,•--. I i •• . .- -r
: i i ' l . ' i ..I : -. i/. t • - ' ' r t i i.'-. ( i r i . i"| i it t .. i . t
' ' ' , . -..Mill 'I I I I' .- t. i . .| . I '. I ll ! 1- ' '''I -II I I .-'
.MI, I v .
• i •:.'';• i : i • . ' i -'I: .\,[ i ni 111 i • <--•:.•!' i nu
l I •! ' - .- . . 1 I 1. . •! I i • l i (• f I -I I'"' ! C. '.I I I" =! t '"
' i . i • • • , i • • i . i ;'. i l 11 , - M i \ i • * • \
i ! 11 i . - .-.i t i i ^. r 1 "* - . • l..' t > i ' J • • ( < ' • 11.! i i < i . .
-------
ABSTRACT
Under the authority of the Federal Insecticide, Fungicide, and Rodenti-
cide Act, the U.S. Environmental Protection Agency (EPA) suspended and can-
celled the registrations and prohibited the further use, sale, and distribu-
tion of ethylene dibron.id? 'EPR) pesticide formuIaMons. As a part of this
ban, EPA also assumed the respor:ibility for destroying/disposing of existing
EDB stocks.
The project covered by tlric report involved an er,ginee-''ng evaluation of
the suitability of various available technologies for the destruction of
ethylene dibromide pesticides. The purpose of the study was to highlight the
technical merits and shortcomings, safety, cost, and total tine reauirefntrt
for each of the alternatives considered.
Both therms! and chemical destruction options were considered, tvalua-
tion criteria were developed so that the different options could be coir^.rtci
on a comtron basis. Information was collected on each candidate process
through a literature search and discussions with industry experts. Concur-
rent with these efforts, bench-scale tests of the chemical methods were
conducted. Also, a test burn was made at a commercial facility to determine
the effectiveness of one of the incineration options. The results of these
tests were factored into this report. Because the chemical processes are
still in the conceptual stages, only preliminary process calculations and
cost estimates were developed for these processes.
Based on the results of this study, incineration in the presence of
sulfur dioxide appears to be the best alternative for the safe, effective,
rapid, and economical destruction of the ethylene dibromide pesticides.
iv
-------
CONTENTS
Page
Foreword 11
Abstract HI
Figures vi
Tables vi
Abbreviations and Acronyms vii
Conversion Factors viii
1. Introduction 1-1
Background 1-1
Purpose 1-2
Technical approach 1-2
Scrpe of work 1-3
Organization of report 1-3
2. Methodology 2-1
Evaluation criteria 2-1
Cost estimating procedure 2-2
3. Description of TechnoV ies 3-1
Classification of processes 3-1
Processes selected for detailed evaluation 3-2
Processes not selected 3-21
o
4. Evaluation of Technologies 4-1
Thermal destruction 4-1
Additional incineration option 4-21
Chemical destruction 4-29
5. Conclusions and Recommendations 5-1
Summary 5-1
Conclusions 5-2
References R-l
Appendix A - Process Calculations A-l
Appendix B - Cost-Estimating Procedures and Results B-l
Appendix C - Information Obtained From Commercial Incineration
Facilities C-l
-------
FIGURES
Number • " Page
3-1 Schematic Flow Diagram of Conventional Incineration System 3-4
3-? Schematic Flow Diagram for Incineration in Prespr.ce of SO- 3-9
3-3 Schematic Flow Diagram for Starved-Air Incineration 3-10
3-4 Schematic Flow Diagram for a Cement Kiln Operation 3-3?
3-5 Schematic Flow Diagram for ATEG Process 3-16
3-6 Schematic. Flow Diagram for the Zinc Process 3-20
4-1 Equilibrium Constant Kp Against Temperature for Chlorine
and Bromine 4_1?
4-2 Particulate Emissions Correlated to Chlorine Content of
Waste Fuels 4-26
TABLES
Number Page
4-1 Summary of Results of Engineering Evaluation 4-2
VI
-------
ABBREVIATIONS AND ACRONYMS
ATEG Sodium Hydroxide Tetraethylene Glyrol Process
ORE Destruction Removal Efficiency
EDB Ethylene dibromide
EDC Ethylene dichloride
EPA Environmental Protection Agency
FIFRA Federal Insecticide, Fungicide, and Codenticide Act
FRP Fiberglass reinforced plaster
IDLH Immediately Dangerous to Life or Health
KTEG Potassium Hydroxide Tetraethylene Glycol Process
PIC Products of Incomplete Combustion
POHC Partially Oxidized Hydrocarbons
RCRA Resource Conservation and Recovery Act
RTI Research Triangle Institute
SCK buper Critical Water
TEG Tetraethylene Glycol
TLV Threshold Limit Value
VII
-------
CONVE,:"'ON FACTORS
1 foot = 0.3048 mpter
1 pound = 454.55 gra^s
1 cubic foot = 7.481 U.S. oallons
1 cubic foot = 0.02832 cubic meter
1 gm/cc = 62.43 1b/ft3
1 gm/cc = 8.345 It/gallon (U.S.)
760 mm of He = 1 atmosphere
760 mm of Hg = 14.7 psi
'C = (^2) x 5
°K = CC + 273.15
°R = °F 4 459.7
1 kilowatt = 1.341 h.p.
1 calorie = 3.97 x 10"3 Btu
Gas constant R = 82.06 (cm3)(atn)/(g.mole)(°K)
R = 10.73 (lb/in2)(ft3)/(lb.mole)(°K)
VI 1 1
-------
SECTION 1
INTRODUCTION
B,,JKGROUr,'P
In September 1983 and February 1984, the U.S. Environmental Protection
Agency (EPA^ suspended the registrations and later finally prohibited (i.e.,
"cancelled") the further use, sale, e.na distribution rf ethyltne dibror.ide
(EDB) pesticide formulations. This actior was taken under the authority of
the Federal Insecticide, Fungicide, and Rodenticide Act (FIFRA) as a result
of animal studies cr health effects, evidence of EDB-contaminated water,
evidence of EDB-contaminatPd food, and EPA-sponsored risk-benefit analyses.
As part of this regulatory actior,, EPA issued orders which halted the use of
EDB-containinc material and requested manufacturers and distributors to
recall all existing EDB products. EPA. was also required to indemnify all
registrants and other owners of EPP pesticides for their economic losses, and
to take responsibility for the destruction/disposal of the EDB stocks
(McCarthy et al., 1987).
The total quantity of formulated EDB pesticides identified for Agency .
disposal amounts to approximately 346,000 gallons or 3.7 million pounds. Of
this amount, 1.1 million pounds are EDB. For purposes of background informa-
tion, the various EDB formulations are divided into four categories. These
categories and their associated approximate quantities are shown below:
Category Quantity, 103 gallons
1. CSj-containing formulations 132
2. Chloropicrin-containing formulations 70
3. Methyl bromide-containing formulations 19
4. Miscellaneous formulations 125
1-1
-------
PURPOSE
This study was prepared for the EPA Risk Reduction Engineering Laboratory
(RRFL) to provide an engineering evaluation of the candidate EDB destruction
technologies.
TECHNICAL APPROACH
The technical approach adopted for this study differentiates it from
previous studies of EDB disposal alternatives in that it concerns engineering
alternatives as opposed to management alternatives. The focus of the effort
of this study was directed to the following progression of steps:
1. The engineering evaluation project was viewed as the first step in
a larger EDB destruction program that could either proceed through
a multiple parallel pilot plant test program and then continue at
the,, production scale through detailed design and engineering,
facility fabrication and construction, startup and shakedown, and
operation and decommissioning of a new facility, or proceed through
a test destruction program at an existing facility, followed by
destruction of all stocks.
2. The engineering evaluation involved the identification of candidate
technologies and processes, the development of selection criteria,
and the application of selection criteria to support technical
judgments.
3. Vendor contacts were made to the various incineration facility
operators to determine interest, feasibility, and cost to destroy
the EDB formulations.
4. Preliminary process designs and cost estimates were developed for
the selected chemical destruction processes in order to compare
these processes to the incineration processes on an equal basis.
5. Process design and cost packages were developed by working closely
with the EPA process developers. Bench-scale performance data were
used to support equipment design and operating assumptions.
6. The results of a parallel bench-scale laboratory study were fac-
tored into the analysis to support and expand the EPA bench-scale
work.
7. A trial burn of two EDB formulations was separately contracted for
by the EPA to obtain more information on incineration in the pres-
ence of sulfur dioxide (S02) to facilitate bromine scrubbing. Data
from this test were then factored into the evaluation of all al-
ternatives.
1-2
-------
i
SCOPE OF WORK
The scope of work performed to complete the technical approach describee.'
above involved the following tasks:
1. PEI identified all candidate technologies and processes that were
available to destroy the EDB pesticide stocks. The scope of tech-
nology search covered all technologies and processes that were
commercial or developmental with the promise of near term avail-
abi1 i ty.
2. PEI developed selection criteria to evaluate the candidate tech-
nologies. P:e criteria considered such parameters as maturity,
access, feasibility, operability, auxiliary processing needs,
health and safety, secondary prvi^o/irnental impact, and permitting.
3. The candidate technologies w^re evaluated against the srlection
criteria through n two-step screening <>^d dPtai'^d evaluatior
procedure. A number of technologies and processes were selected
for detailed evaluation. n> part of the detailed evaluation pro-
cess, fundamental physical and chemical data were collected, pro-
cess design developments were researcher1, prelimina-y process
designs viere developed, vendor quotatiot.s were obt?ined and'process
costs were ?stimated, and performance capabilities were assessed.
ORGANIZATION OF REPORT
The report is organized in a manner consistent with the overall purpose
of the project. In Section 2, the methodology used to complete the study is
described. In this section, we introduce the evaluation criteria ard de-
scribe the costing procedures. In Section 3, we classify the technologies,
describe the technologies and processes investigated, and focus on those
selected for detailed evaluation. In Section 4, we present the results of
the study. The results are presented and described in terms of the evalua-
tion criteria presented in Section 2. In Section 5, conclusions and recom-
mendations are presented. Three appendices provide documentation and backup-
information for process calculations, cost calculations, and vendor contacts.
1-3
-------
SECTION 2
METHODOLOGY
EVALUATION CRITERIA
As mentioned earlier, the purpose of this study was to provide an engi-
neering evaluation of the candidate EDB destruction technologies. The aim of
the evaluation was to highlight the technical merits and shortcomings,
safety, cost, and total time requirement for each of the alternatives
considered. To provide a common basis to compare different process options,
the ft)1!owing evaluation criteria were developed in consultation with EPA:
CRITERIA
Status - Commercial, Pilot-Scale, or Conceptual
Accessibility
Past Experience
Need for Development and Testing .
Preprocessing
Process Safety
Toxic Emissions
Residues
Need for New/Additional Equipment
Extent of Corrosion
Process Compatibility with Pesticides
Secondary Environmental Impact and Health
Considerations
Mechanical Reliability
Transportational Access to Facility
Storage and Handling of Pesticides and Residues
Cost
Permitting
Probability of Success
Time for Corneletion
All the Uems in the above criteria are self explanatory. An attempt
was made to make tie criteria exhaustive and to cover all major technical and
2-1
-------
economic aspects. Information was collected on each candidate process so as
to address each item in the evaluation criteria. This was done by an ex-
haustive literature survey, discussions with industry experts, process cal-
culations, and preliminary cost estimates. The results are presented in
Section 4.
COST ESTIMATING PROCEDURE
Process economics is an important factor in the ultimate selection"of a
technology. To compare the cost-effectiveness of each ootion, unitized costs
($/lb of pesticide) were established for each process option. Unitized costs
for technologies which are already commercialized (incineration and cement
kiln incineration) were obtained by contacting vendors and getting their best
possible estimates. As regards chemical destruction, both the process op-
tions available therein are still in the conceptual stages. On the basis of
laboratory scale results, preliminary flow sheets were developed (Section 3)
followed by preliminary process design (Appendix A) and preliminary cost
estimates (Appendix B). Details on the cost estimating procedure adopted are
presented in Appendix B. The various assumptions underlying the design have
been stated in Appendix A.
2-2
-------
SECTION 3
DESCRIPTION OF TECHNOLOGIES
A wide range of existing technologies may have a potential for success-
fully eliminating ethylene dibromide (EDB) pesticide formulations. Research
Triangle Institute (RTI) performed a preliminary screening of available tech-
nologies for the EPA (RTI, 1987). The technologies were evaluated against
the following criteria:
1) Immediate availability of the technology and the potential of
procurement of a permit within the next 2 years.
2) Capability of meeting the letter and intent of the RCRA and FIFRA
regulations.
3) Capability of handling the corrosivity and emissions due to EDB.
The following technologies were selected in thisoengineering evaluation
as possible candidates for the elimination of EDB pesticide formulations:
1) Incineration in a waste incinerator under oxidizing (excess air)
conditions (conventional incineration)-
2) Starved-air incineration
3) Incineration in presence of sulfur dioxide or sulfur-containing
waste
4) Incineration in a cement kiln
5) Chemical destruction by the ATEG process
6) Chemical destruction using the zinc process
7) Williamson's sy.thesis for the destruction of methyl bromide
8) MOOAR process (oxidation in supercritical water)
CLASSIFICATION OF PROCESSES
/
The preceding technologies can be classified into two main categories:
1) thermal destruction, and 2) chemical destruction.
Thermal Destruction
Conventional incineration, incineration in the presence of sulfur
dioxide, starved-air incineration, and cement kiln incineration fall into
3-1
-------
this category. Thermal destruction uses heat to convert hazardous materials
into harmless or less toxic materials. Depending on the conditions prevail-
ing in an incinerator, the thermal destruction processes can be further
classified into three subclasses:
1) Conventional incineration or excess air oxidation—thermal decompo-
sition in the presence of excess air (oxygen).
2) Pyrolysis —thermal decomposition in the absence of oxygen.
3) Starved-air incineration—incineration that uses substoichiometric
amounts of air.
All of the options currently under consideration (except starved air
incineration) would be classified as conventional incineration. Cement kiln
incineration can be regarded as a special application of conventional incin-
eration because of the oxidizing conditions in the kiln.
Chemical Destruction
In chemical destruction, the 'waste is reacted with a suitable reagent to
yield products that are harmless or less toxic than the parent compounds.
The ATEG process, the zinc process, and Williamson's synthesis fall into this
category. The KODAR process is also classified as a chemical process, al-
though, like incineration, it oxidizes the wastes to less harmful products.
Supercritical water, however, is used as an oxidizing medium in the MODAR
process.
PROCESSES SELECTED FOR DETAILED EVALUATION
Based on the selection criteria developed earlier in Section 2, certain
processes were selected from the earlier evaluation for more critical review.
The processes considered in the final evaluation are as follows:
1) Conventional incineration
?) Conventional incineration in the presence of sulfur dioxide
3) Starved-air incineration
4) Cenent kiln
5) ATEG process
6) Zinc process
3-2
-------
A brief description of each of these processes is presented in the
following subsections.
Conventional Incineration (Excess-Air Incineration)
Conventional incineration is the most common way of destroying hazardous
substances. Numerous commercial hazardous waste incinerators are operating
successfully throughout the Un-'ted State? arrt worldwide. Some of these
operating systems are transportable, which makes them convenient for the
destruction of hazardous wastes at specific sites.
This process uses combustion to oxidize hazardous materials to harries?
or less toxic materials. Products of incineration consist of combustion
gases and, in some cases, solid residues. Most liquid wastes like the EDB
formulations yield gaseous products. Although incineration may assure com-
plete destruction of the hazardous waste, the combustion products can be
environmentally harmful and, thus, require secondary treatment. This seconda?
treatment may be wet scrubbing, particulate collection, or the use of after-
burners. The technology cannot be used to destroy wastes whose combustion
products cannot be treated to abate harmful emissions. . Wastes containing
compounds of sulfur, phosphorus, nitrogen, and chlorine have been success-
fully treated by this technolcgy; however, no routine incineration runs of
brominated wastes are known.
Figure 3-1 presents a typical flowsheet of a hazardous waste incinera-
tion system. The plant can be divided into three main areas: 1) the
storage, handling, and preparation of the waste fuel prior to incineration;
2) the incinerator itself; and 3) the emissions control system.
The storage and handling area is concerned with the receiving and stor-
age of the waste at the site. Some waste must be preprocessed before incin-
eration to achieve the necessary heating value or to reduce the viscosities
3-3
-------
CAUSTIC IN
00
I
BROMINATED WASTE
COMBUSTION AIR
INCINERATOR
i
SCRUBBER
FLUE GASES
STACK
CAUSTIC OUT
Figure 3-1. Schematic flow diagram of conventional incineration system.
-------
to a point where the material is pumpable and atomizable at the operating
temperatures. This is usually done by blending the waste with other wastes
or with an auxiliary fuel in the preprocessing area. This area must be well
designed to prevent or contain any spillage and resulting emissions. Sonner
et al. (1981) present an excellent overview of the various available design
options. This area is believed to be well designed in most opera-ting facili-
ties.
The preprocessed fuel is fed into the incinerator, which can be a rotary
kiln, a fluidized-bed combustor, liquid injection incinerator or an electric
furnace. The choice of incinerator depends on the type of waste to be de-
stroyed. Rotary kiln, liquid injection, and fluidized-bed designs are bet-
ter suited for incineration of liquid wastes than are the other two designs
(Bonner et al., 1981). Most commercial incineration facilities have a rotary
kiln incinerator. Typical temperatures in the combustion chamber range from
1800° to 2400°F, and under these conditions, the waste mixture undergoes
instant oxidation. Liquid injection incinerators are operated under similar
temperature range, although they can be operated at higher temperatures.
Because of the high operating temperatures, the furnace has to be re-
fractory-lined. The quality of the refractory lining depends on the
corrosivity of the gases likely to be generated in the incinerator. For
applications involving halogenated wastes, the refractory lining must be
corrosion-resistant. Typical residence times in the incinerator vary from
0.5 to 2 seconds at a minimum.
The combustion gases from the incinerator are further processed in the
gas treatment area to minimize toxic emissions. For all hazardous waste
3-5
-------
applications, this area includes a well-designed scrubbing system and, some-
times, an afterburner. If the combustion products of the primary combustor
are expected to contain partially oxidized compounds, an afterburner is
required to ensure complete combustion of all constituents. This eliminates
the possibility, of emitting partially oxidized products into the atmosphere.
Some partially oxidized organics are carcinogenic and may be even more harm-
ful than the parent compounds. If the flue gases are expected to have high
concentrations of sulfur dioxide, nitrogen dioxide, hydrochloric acid, and
other toxic products, a hijh-efficiency scrubbing system is needed to neu-
tralize these pollutants. Venturi scrubbers are quite common for scrubbing
applications, and their capabilities for particulate removal are excellent.
When lime slurry is used as the scrubbing medium, toxic pollutants such as .
sulfur dioxide and hydrogen chloride can also be removed efficiently as long
as thpy are not present in "high concentrations. For halogenated wastes,
scrubbers that have good mass transfer characteristics are recommended. Typ-
ically, a packed towen or a plate column can be used because -ney have excel-
lent mass transfer characteristics and are easy to operate. If particulate
emissions are the only concern, fabric filters or electrostatic precipitators
(ESPs) provide adequate treatment of the flue gases. Details on the various.
design and operational aspects of an incinerator are presented by Bonner et
al. (1981) and Sittig (1979).
Unlike fluorine and chlorine, which generate predominantly hydrogen hal-
ides upon combustion in excess air, brominated wastes generate bromine when
incinerated, wnich is difficult to remove from the flue gases by currently
operational gas-processing techniques. Caustic is typically used as the
3-6
-------
scrubbing medium in many hazardous waste incinerators; however, It is be-
lieved that a caustic solution may not remove bromine as readily as it does
hydrogen bromide. Therefore, modification of the incinerator operating
conditions (e.g., addition of sulfur) will be required to prevent significant
amounts of bromine emission into the atmosphere. Alternatively, if a scrub-
bing medium could be found which reacts rapidly with bromine, conventional
incineration without any process modifications could become a viatfe option.
The literature suggests some mediums for this application. However, more
research will be required before a full-scale operation can be undertaken.
Incineration in Presence of Sulfur Dioxide or Sulfur-Containing Waste
A process modification of the conventional incineration reduces toxic
emissions due to halogenated wastes. Under this option, the halogenated
waste is burned in a conventional incinerator in the presence of sulfur
dioxide or sulfur-containing wastes. Under the incinerator operating condi-
tions, sulfur dioxide reacts with the halogen produced during incineration to
form hydrogen bromide 3nd sulfuric acid. During the Rollins test burn on EDB
stock, 10 percent sulfuric acid (H?SO.) was used as the source of sulfur
(Alliance Corporation, 1988). Under the kiln conditions, the ^SO^ decomposes
as follows (Equations 1 and 2):
2 H2S04 — > 2 S03 + 2 H20 (1)
.. A. .>
2 S03 f_ * 2 S02 + 02 (2)
At high temperatures, the equilibrium reaction is displaced to the
right, thus favoring decomposition. The bromine formed due to oxidation of
EDB (Equation 3) reacts with S0? and water to form hydrogen bromide (HBr) and
i ^
H2S04 (Equation 4). j
3-7
-------
2C0 + 2H0 +Br (3)
Br, + SCL + 2 H,0 ....... > 2 HBr + H2$04 (4)
The above reaction proceeds in both the gas and liquid phases.
Since the resulting acids c«n be easily removed in the downstream process-
ing units (scrubbers), the problem of halogen emissions is completely elim-
inated (Fabian et al., 1979). A flow diagram of this process is shown in
Figure 3-2. Sulfur dioxide or sulfur-containing waste in slight excess of
the stoichiometric requirements should be supplied to ensure total conversion
of the halogens to hydrogen halides.
St6rved-Air Incineration
This option uses the same equipment and entails the same process flow as
•
the conventional incineration process. The only difference is in the process
conditions in the incinerator. Unlike conventional incineration, starved-air
incineration uses less than stoichiometric quantities of air for combustion
purposes. To date, the application of this technology for hazardous wastes
has been minimal (Bonner et al., 1981). Some of the advantages of this
process include high thermal efficiencies, reduced volume of flue gases, and
suppression of particulate emissions.
The bromine/hydrogen bromide chemical equilibrium favors hydrogen bro-
mide formation under reducing conditions (less oxygen) in the furnace.
Because hydrogen bromide is more easily scrubbed than bromine is, the possi-
bility of toxic emissions due to bromine is lessened. Thus, conceptually,
starved-air incineration appears to have the potential to destroy brominated
wastes.
Figure 3-3 (J. Cegielski, John Zink Co., personal communication) pre-
sents a proposed flowsheet for the application of this technology. The waste
3-8
-------
CAUSTIC IN
BROMINATED WASTE
AUXILIARY FUEL
COMBUSTION AIR
SO2
£
^
».
SATURATOR
HBr. H2SO4
^
T
SCRUBBER
1
CO
i
vo
CAUSTIC OUT
BROMIDE SALTS,
SULFATES
STACK
Figure 3-2. Schematic flow diagram for incineration in presence of 502-
-------
BROMINATED WASTE
AUXILIARY FUEL
SUBSTOICHIOMETRIC AIR
CO
i
CAUSTIC IN
i
REDUCTION
FURNACE
FLUE GASES
COOLING
SYSTEM
SCRUBBERS
FOR
HBrRCVIOVAL
CAUf.TIC OUT
BROMIDE SALTS
-»
AFTERBURNER
(OXIDIZING
CONDITIONS)
EXCESS
COMBUSTION AIR
STACK
Flo ~e 3-3. Schematic flow diagram for starved-air incineration. (John Zink Co.)
-------
would be incinerated under reducing conditions (less than stoichiometric
amounts of oxygen) to convert all feed bromine to-hydrogen bron'ide. The
gaseous products of partial combustion would ^e cooled and passed through a
scrubber to remove HBr and other pollutants. The scrubber effluent gases
would then be reoxidized in an afterburner to ensure the complete combustion
of all organics. The products from the afterburner, which usually consist of
carbon dioxide and weter, would be released into the atmosphere. If the
afterburner products were to contain any toxic pollutants, a secondary scrub-
bing system would be required to reduce the emissions to acceptable levels.
Cement Kiln
Figure 3-4 presents a typical flow diagram of a cement manufacturing
process. The production of cement involves four steps: 1) quarrying and °
crushing of raw materials; 2} grinding and blending of these materials into
feed at proper proportions; 3) calcining the raw materials at extremely high
temperatures in a rotary kiln furnace to form clinker (an interim product);
and 4) finish-grinding of the clinker, blending it with gypsum, and packaging
the finished product. The main ingredient (lime) is processed in the crush-
ing and grinding section before it is calcined in the kiln. The kiln is
operated at 2000r to 2300CF. The heat required to carry out the calcination
is supplied by burning fuel in the kiln under excess-air conditions. The hot
gases from the kiln contain substantial particulates, which are removed by
fabric filters or cyclones followed by ESP's. The flue ga? processing sys-
tems in most cement kilns consist essentially of particulate-removal equip-
ment. Very few facilities use wet scrubbing systems to handle toxic chemical
emissions.
3-11
-------
Figure 3-4. Schematic flow diagram for a cement kiln operation.
-------
The primary cost factor in the production of cement is energy, which
accounts for as much as 40 percent of the total cost (PEI, 1987). Tc offset
escalating fuel costs, innovations have been made ir the process to reduce
fuel usage. One innovation involves the use of hazardous waste fuels; how-
ever, not all incinerable waste can be burned in a cement kiln. The most
suitable waste is a liquid with high energy value (at least 10,000 Btu/lb), a
low water content, and a low concentration of metals (State of California,
1982). Guidelines on the use of hazardous waste in cement, kilns have been
presented in detail in a report prepared for EPA (PEI, 1987). A growing
number of ceirent kilns are using hazardous waste fuels in their operations.
According to a 1981 survey, about 1,000 to 25,000 gallons/day of hazardous
waste fuel was being incinerated in cement kilns. Cement kilns have a his-
tory of successful incineration of chlorinated waste without any harmful
emissions. The hydrogen chloride and chlorine generated in the furnace rfact
with the raw materials (lime and some sodium and potassium in the ore) to
form chloride salts. The alkaline condition^ in the kiln cause the reactions
to be rapid and complete. Thus, the kiln acts as its own scrubber. Further,
the formation cf salts is advantageous, as it improves product quality.
Thus, a substantial incentive exists to use chlorinated waste as fuel. Simi-
lar results are expected with brominated wastes. Industry sources believe
that under the alkaline conditions prevailing in the kiln, the bromine should
react to produce bromide salts, which not only improve the product quality,
but also ensure complete destruction without any toxic emissions.
ATEG Process
The ATEG process, developed by the EPA (Rogers and Kornel, 1987), in-
volves a reaction between EDB and sodium hydroxide (NaOH) in the presence of
3-13
-------
a phase transfer catalyst, tetraethylene glycol (TEG) to yield acetylene,
bromide salts, and water. The reaction products are not toxic and can be
easily disposed of. The overall reaction mechanism may be represented as
f' i lows:
Br-CH.,-CH2-Br + 2 NaOH ...... > CH;CH + 2 NaBr •»• 2 H20 (5)
Potassium hydroxide (KOH) may also be used instead of NaOH as it reacts
similarly, although more vigorously (RTI, 1967). Because KOH reacts violent-
ly with the carbon tetrachloride (CC1.) present in the formulations, it cre-
ates the potential for a runaway reaction. Therefore, carbon tetrachloride
may l.ave to be removed from the pesticide formulations prior to chemical
destruction. The reaction has been demonstrated only on a laboratory scale.
A comprehensive testing program will be required to demonstrate the process
on a pilot scale prior to full-scale operation.
Laboratory-scale studies on this, reaction system were carried out by the
EPA (RTI, 1S87) and the major findings are summarized below:
The reaction proceeds in two steps, with vinyl bromide as an inter-
mediate product. A very small percentage of the vinyl bromide re-
acts further in the reactor to yield acetylene. The vinyl bromide
had to be treated in a scrubber with the KTEG solution (KOH dissol-
ved in TEG, diluted by water) to achieve complete conversion to
acetylene. Thus, the reaction mechanism is as follows:
Br-CH2-CH2-Br + NaOH ....... > Br-CH=CH2 + NaBr + H20 (6)
Br-CH=CH + KOH ........... > CHECH + KBr + HO (7)
The reactions are exothermic with an overall heat of reaction of 30
kcal/gmole (RTI, 1987). As per EPA, the reactor temperature should
be maintained below 113°F at all times to avoid possibilities of a
runaway reaction.
The reaction seems to have an inception time of about 15 minutes
and an overall reaction time of 45 minutes.
3-14
-------
Laboratory tests seem to indicate relatively rapid reaction between
the reactants. Hence, gradual feeding of the reaction constitu-
ents, especially at high tDB concentrations, is necessary to avoid
any rapid runaway reaction.
Carbon disulfide (CS2), a constituent of some of the pesticide
formulations, has been found to react quantitatively with the ATEG
to form a viscous sludge that inhibits the EDB destruction reaction
(RTI, 1987).
Chloropicrin, a major constituent in some formulations, reacts with
the catalyst TFG, which inhibits the EDB destruction reaction (RTI,
1987). Moreover, the products of the reaction with chloropicrin
are believed to be varied and have not been analyzed as yet.
Thus, to successfully dispose the pesticide formulations, it is impor-
tant that the constituents which impede the reaction be removed. A common
way to do this would be distillation. Distillation of the CS2 formulaticns
should not be a problem, although distillation of chloropicrin formulations
could be difficult. An advantage with distillation is that the products of
distillation could be sold at market value and improve the overall process
economics.
A schematic flow diagram for this process was developed by the EPA.
Based on the various inputs from the EPA, this flow diagram was refined to
identify key process equipment (Figure 3-5). The subsequent engineering
evaluation has been earned out for the flow scheme identified in Figure 3-5.
The process was proposed to be a batch operation in which each batch
treats approximately 300 gallons of pesticide. The EPA has indicated that
solid NaOH should be used for this reaction. Solid NaOH flakes would be
added to the reactor at a controlled rate using a suitable feeder (e.g.,
screw feeder). This should help to control the reaction rate and to avoid
the possibility of runaway reactions. The reaction between EDB and NaOH
produces vinyl bromide gas, sodium bromide, and water. The overall reaction
3-15
-------
VACUUM
PUMP
U)
t—»
en
Jt CONDENSER
—fflh
ACETYLENE
FILTER FEED
PUMP
CAKE
REACTOR OFF GAS
(VINYL BROMIDE/ 4 ' '
CHLORIDE)
FILTER
ORGANIC
PHASE
AQUEOUS
PHASE
LIQUID EFFLUENT
STORAGE TANK
Figure 3-5. Schematic flow diagram for ATEG process.
-------
time would be about an hour. The reactants would be fed at ambient tempera-
ture and the reactor temperature would not be allowed to exceed IIS^F. The
reaction would be carried out in a well mixed jacketed reactor vessel to
facilitate efficient heat transfer. An agitator would keep the reactor
contents well mixed.
The vinyl bromide formed during the reaction would be treated further
with KTEG solution to produce acetylene gas and potassium bromide salt. EPA
suggested treating the reactor off gasts in a packed bed counter-current
scrubber. EPA has proposed the use of Flexipac regular packings in the
scrubber. These packings are reportedly self-wetting and efficient in flush-
ing any solids that are formed. Because acetylene is highly explosive, the
EPA has suggested that it be removed from the system with a vacuum pump and
subsequently flared to produce carbon dioxide and water. The highly water-
soluble potassium bromide should dissolve in the KTEG solution, which is
about 60 percent water. The KTEG solution contains a large amount of TEG to
achieve higher reaction rates. However, since TEG is expensive, discarding
the spent solution from the scrubber would make the overall process somewhat
expensive. As a result, the EPA has proposed that the liquid effluents from
the scrubber would be discharged into a collection tank, with makeup quanti-
ties of KOH added directly to this collection tank. The solution would be
filtered before it is recycled back to the scrubber. This way, it is be-
lieved that the solids formation in the scrubber could be minimized and the
solution used for longer periods of time without replenishing it.
The liquid effluents from the -eactor contain other organic waste con-
stituents, alkali salts, and water. It is believed that the alkali salts
would partially dissolve in the water formed. The undissolved salts would
3-17
-------
remain as suspended solids and be removed In a solid-liquid separator (prob-
ably a filter). The filtrate would be stored in a tank that would also act
as a phase separator. The liquid effluent (from the reactor and scrubber)
may be further treated to recover byproducts of value, or it may be treated
for ultimate disposal .
Two economic options have been considered for this process. The first
involves building a completely new facility with all new equipment. The sec-
ond involves using some process equipment available at the GARD facility
(reactors, flare, and stack) to be used along with some new equipment.
Zinc Process
This process entails a classical organic reaction for the dehydrohalo-
genatipn of halogenated orgdnics (Fieser and Fieser, 1967) and was suggested
to RTI by Al Korntl of EPA. Metallic zinc reacts with EDB at ambient tem-
peratures to produce .ethylene gss and zinc bromide. The reaction is repre-
sented as follows:
Zn + Br-CH2-CH2-Br ...... > ZnBr2 + CH2=CH2 (8)
However, the reaction has been demonstrated only on a laboratory scale.
Like the ATEG process, a comprehensive testing program would be required to
demonstrate the process on a pilot scale prior to full-scale operation.
Laboratory-scale studies of this reaction have resulted in 99+ percent
EDB destruction (RTI, 1987). During these experiments, optimal results were
obtained by using 100- to 200-mesh zinc powder slurried in distilled water
and t little hydrochloric acid. The acid supposedly cleans the zinc surface
and thereby enhances the reaction rate. Slightly more water was added than
that required to dissolve the zinc bromide formed in the reaction. The
reaction temperature was not allowed to exceed 113°F.
3-18
-------
The reaction is exothermic, with very high heat of reaction. It has
been suggested that the reaction temperature should be maintained below 113CF
to avoid any possibility of runaway reactions.
Preliminary tests with cMoropicrin formulations by the EPA resulted in
•complete destruction of the EDB. However, the reaction with chloropicrin
forms a whitish powder (believed to be zinc salts of carbonate or oxide, or
hydroxide), which coats the zinc surface and inhibits the reaction with EDB.
Unless prevented, this will increase the zinc consumption.
Based on the preceding information, a preliminary flowsheet has been
developed for this process (Figure 3-6). The process would be a batch opera-
tion. A screw-type feeder would be used to add the zinc particles (approxi-
mately 100-mesh) at a controlled rate to the reactor containing the pesticide
batch. Appropriate amounts of water and 30 percent hydrochloric acid also
would be added to the reactor at a controlled rate. This approach, plus
maintaining the temperature below 113°F at all times should achieve a con-
trolled reaction rate. The reactor would be a jacketed vessel equipped with
a turbine-type high-efficiency impeller. The reaction gases (primarily
ethylene) would require no further processing and thus would be sent directly
to the flare. The liquid effluents from the reactor consist of an organic
phase and an aqueous phase. Zinc bromide formed in the reaction is highly
water soluble; therefore, the only suspended solids would be the unreacted
zinc. These solids would be separated in a solid-liquid separator (filter),
and the filtrate would be placed in a storage tank, which'would also act as a
phase separator. The filtrate may be further processed to recover any by-
products of value or discharged from the process for disposal.
3-19
-------
IM
O
PESTICIDE
TANK
CARS
HCI
BARREL!
Zn
STORAGE
BIN
— er
j— a
•^f\s\s\rl • I
L*W
— y
WATER FEED
PUMP
^
T3i±U
— — •*• — *j
/ ETHYLENEGAS J _. .„_ I
REACTOR / U- H FLARE j
/
^4
^^ fc« FILTER j — i L.
EDB/EDC ' * Kj 1 ^- 1 | f \ p
FREE EFFLUENT Z±i >-X ornf
FILTER FEED 1 z — *
PUMP FILTRATE
ruM UNREACTED PUMp
ZINC CAKE
(FOR DISPOSAL)
^ORGANIC
^ PHASE
^AQUEOUS
^ PHASE
:FLUcNT
IAGE TANK
Figure 3-6. Schematic flow diagram for the zinc process.
,: /
-------
Two capital investment alternatives are possible for this process. The
first alternative entails .the construction of a completely new facility with
all new equipment. Under the second alternative, some equipment from the
CARD facility would be used in conjunction with new equipment.
PROCESSES NOT SELECTED
Williamson's Synthesis
This process involves another classical orgar.ic chemistry reaction.
This reaction would be suitable to convert methyl b.ronide (CH,Br) to dimethyl
•J
ether [(CH-J^O] in the presence of methanol (CH..OH) and an alkali. Either
NaOH or KOH may be used. The reaction is represented as:
CH7Br + CH7OH + KOH (NaCK) > (CH,),0 + KBr (NaBr) + H90 (9)
w 0 3 £ c
Methyl bromide is available in liquified gas cylinders; therefore, this
reaction can be carried out in some kind of a gas liquid contactor. Because
the amount of methyl bromide that needs to be destroyed is very small, a
separate process for its destruction was not considered. The applicability
of this process for other formulations, including methyl bromide, has been
tested on a laboratory scale. However, after discussions with EPA, this
process was not reviewed for the proposed application.
MODAR Process
MODAR, Inc., of Houston, Texas, has developed a novel technology for the
destruction of hazardous waste that uses the special properties of supercrit-
ical water (above 705°F and 3200 psia). The process is based on the princi-
ple that water in the supercritical region exhibits properties far different
from normal water. The density of supercritical water (SCW) is low enough
(0.05 to 0.5 g/ml) and the temperature is high enough to effect the essential
3-21
\ V,
-------
elimination of hydrogen bonding. As a result, the dielectric constant is
reduced from 80 to less than 2, and many normally water-insoluble organics
become highly soluble. In contrast, inorganic salts become only slightly
soluble. The dissolved organics can be oxidized to give CO- and water and
hetro-atoms (including halogens, phosphqrus, sulfur, and metals) are precipi-
tated as salts.
The process consists of the following steps:
1) The toxic or hazardous waste is slurried with makeup water to
provide a mixture of about 5 to 10 percent by weight. The slurry
is pressurized and heated to supercritical conditions to avoid char
formation. Heating is attained by mixing the feed with supprheated
SCW, which is generated in a subsequent step. During a short
residence time in the tube leading to the oxidizer, organics in the
feed are converted to combustible gases, low to intermediate molec-
ular weight compounds (furans, furfurals, alcohols, aldehydes), and
inorganic salts.
2) Air or oxygen and an alkali solution are pressurized and mixed with
the -feed. Because the water is still supercritical, the oxidant is
completely miscible with the solution (i.e., the mixture is a
single, homogeneous phase). Organics are oxidized in a controlled
but rapid reaction. Because the oxidizer operates adiabatically,
the heat released by combustion of readily oxidized components is
sufficient to raise the fluid phase to temperatures at which all
organics are oxidized rapic'ly. For a feed of 5 percent organics by
weight, the heat of combustion is sufficient to raise the oxidizer
effluent to at least 1022CF. The hetro-atoms (like halogens) react
with the alkali to form inorganic salts.
3) The effluent from the oxidizer is fed to a salt and sediment sepa-
rator, where inorganics and sediments originally present in the
feed are removed as a solid slurry. At 932°F and above, the solu-
bility of inorganics in SCW is extremely low.
4) A portion of the superheated SCW is recycled to an eductor upstream
of the SCW oxidizer. This operation provides for sufficient heat-
ing of the feed to bring the oxidizer influent to supercritical
conditions.
5) The remainder of the superheated SCW (with some C02 and N;>) is
available for power generation or use as high-pressure steam. A
portion of the available energy is used to generate the power
required to pressurize feed and oxidant. The energy required to
pressurize the oxidsnt is recovered in the expansion of the prod-
ucts of combustion in the superheated SCW turbine.
3-22
-------
As a waste destruction process, the MODAR concept has the following
advantages:
The process is carried out in a closed system, which allows total
physical control of waste material to be maintained from storage
through reaction and final discharge of products.
The process has a high destruction efficiency (ORE). Two liquid
PCB wastes treated by this process at the CECOS facility in Niagara
Falls, New York, under permits, showed 99.999 percent ORE (Staszak
et al. 1987). The process apparently gave similar DRE's for other
wastes when tried on a bench scale.
The process can be adapted to a wide range of feed mixtures and
scale of operations. No preprocessing of the waste is required;
therefore, this process may be able to destroy all the waste pesti-
cides without a need for distillation.
Skid-mounted, transportable systems are being designed alor.g with
large-scale stationary units. With a transportable unit, the need
to transport all the hazardous waste to a fixed site would be elim-
inated, which would make the operation safer and more economical.
The use of this process also has the following disadvantages:
No operational units are currently in the field. It would take a
year to set up a unit that could take care of this waste. Thus, it
does not fit in with the time frame required by the EPA.
The process operates under very high pressure (above 218 atms). In
the past, other wet oxidation processes have had problems and
experienced explosions. The risk of an explosion is quite high.
The current design could treat 20,000 gallons/day of the 10 percent
organic mixture. The design is still in the study phase, so any
process deficiencies which may emerge in scaleup are still unknown.
Because this is a very sophisticated technology, more highly
skilled operutors would be required to operate the process as
compared to other processes. Thus, operational costs could be
substantially higher than for other processes operated by less-
skilled labor.
No data are available on destruction of brominated waste, even on a
bench scale. A bench-scale characterization was quoted for
$25,000. Also, because MODAR is primarily an oxidation process,
bromine' will be formed in the reactor. In incinerators, scrubbing
bromine with an alkali is very difficult, and bromine removal with
an alkali may not be easy, even in the MODAR process.
MODAR is not interested in facility operation; their interests lie
in designing and selling units to others. Thus, EPA might have to
3-23
-------
buy the equipment. Even if MODAR decided to do the job for EPA,
the EPA would have to provide land, site preparation, permitting,
etc. This option could turn out to be very expensive. Neverthe-
less, if the unit could be used for destroying other wastes (as-
suming it can work as a regular waste treatment facility), the
marginal cost for destroying the waste could range from S3 to
$4/gallon (excluding the cost of permits, site preparation, and
installation).
After the advantages are disadvantages were weighed, it was decided to
eliminate this process frorr, further consideration.
3-24
-------
SECTION 4
EVALUATION OF TECHNOLOGIES
This section presents the results of the technical evaluation of each of
the available options for destroying pesticide formulations. This evaluation
consisted of a literature re-view, an engineering assessment and supportive
calculations, and study cost estimates. Each potential candidate process was
reviewed critically with respect to each of the selection criteria introduced
in Section 2. Table 4-1 summarizes the results of the evaluation with re-
spect to the selection criteria for each of the processes.
THERMAL DESTRUCTION
. All options discussed under thermal destruction in Section 3, except the
cement kiln, involve the use of a commercial incineration facility. As
discussed in Section 3, these alternatives differ only with regard to the
process conditions maintained in the incinerator. All of the operating
conditions required to achieve successful destruction of hazardous waste ere
possible in a commercial incinerator.
The options for which a commercial incineration facility is suitable are
as follows:
1) Incineration in oxidizing atmosphere—conventional incineration
2) Incineration in the presence of S0? or sulfur-containing westes
3) Starved-air incineration
4-1
-------
TABLE 4-1. SUMMARY OF RESULTS OF ENGINEERING EVALUATION
Criteria
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zinc process
f*
I
Status
Accessibility
Past
Experience
Commercial. Not yet demonstrated.
Accessible. Accessible. At least two
Rollins has facilities interested in
offered this pursuing this option.
technology.
Trial burns at None for brominated waste.
Rollins, Deerpark,
TX, facility re-
sulted in EOB de-
struction effi-
ciencies greater
than 99.9999%.
There were no
visible bromine
emissions at the
stack. Prelimi-
nary results
show that all
bromine fed to
the incinerator
is captured in
the scrubber
system.
Commercial .
Accessible.
Two companies
interested.
One known test
in Canada. Re-
ports almost
complete rap-
ture of bro-
mine in the
process residue.
Pilot plant.
(Bench scale
for EDB ap-
plication).
New plant to
be built or
existing
plant to be
rndified.
Limited to
bench scale
demonstra-
tion.
°
Bench scale.
New plant to be
bui It or exist-
ing plant to be
mod i f i ed .
Limited to
bench scale dem
onstration.
(continued)
-------
TABLE 4-1 (continued)
Criteria
Pre-
processing
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zinc process
Need for
Development
and Testing
In light of the
successful trial
burn, it is be-
lieved that no
further develop-
ment or testing
is required.
The chemistry of bromine/
HBr favors formation of
bromine over HBr at or-
dinary operating conditions.
Theoretically, this equilib-
rium can be shifted to yield
HBr -by reducing the oxygen
and increasing the water
vapor content. However,
under partial pyrolysis, the
extent of improvement in HBr
formation is not -known.
0
Minor modifi-
cations to kiln
feed required.
Process chemis-
try favors bro-
mine capture in
product without
major changes in
kiln design.
The develop-
ment status of
this technolo-
gy will re-
quire more
pi'ot scale
testing to
establish
feasibility*
optimum oper-
ating proce-
dures, and
process param-
eters for de-
sign and
scale-up.
The development
status of this
technology will
require more
pilot scale
testing to es-
tablish feasi-
bility, optimum
operating pro- .
cedures, and
process param-
eters for de-
sign and scale-
up.
Routine.
Routine.
Routine.
(continued)
Carbon disul-
flde inter-
feres with
ATEG reaction.
This will have
to be removed
by distilla-
tion. Chloro-
picrin may be
difficult to
treat with
ATEG, and is
not distilled
easily. Thus,
Carbon disul-
fide interferes
with the zinc
reaction. It
will have to
be removed by
distillation.
Chloropicrin
produces a
coating on the
zinc and may
have to be re-
moved by pre-
processing.
-------
TABLE 4-1 (continued)
Criteria
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zir.r process
Pre-
processing
(continued)
Process
Safety
Safe.
Safe.
Safe.
Toxic
Emissions
None, based on
trial burn
results.
Bromine emissions are pos-
sible.
Possibility of
bromine emis-
sions and par-
tially oxidized
organics.
chloropicrin-
containing
formulations
may have to be
disposed of
independently.
Acetylene is
a major prod-
uct of the
process, which
poses an ex-
plosion, haz-
ard. Also
reaction is
highly exo-
thermic re-
quiring care-
ful control.
Vinyl bromide
and chloride
emissions are
possible.
Ethylene and
hydrogen are
major reaction
products, in-
creasing ex-
plosion risks.
Reactions are
very exother-
mic requiring
effective
monitoring and
and control
to avoid run
away reactions.
Ethylene is
reportedly
herbicidal.
(continued)
-------
TABLE 4-1 (continued)
Criteria
Need for
New/
Additional
Equipment
Extent of
Corrosion
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zinc process
Residues Brominated scrub- Brominated scrubber solution No problems
ber solution and and sludges will require envisioned.
sludges will careful handling.
require careful
handling.
Reactor ef-
fluents may be
classified as
a hazardous
waste which
will require
appropriate
handling and
disposal. The
residue con-
tains byprod-
ucts of value
which may be
recovered
prior to
disposal.
Reactor efflu-
ents may be
classified as a
hazardous waste
which will re-
quire appro-
priate handling
and disposal .
The residue
contains by-
products of
value which may
be recovered
prior to
disposal .
None.
Corrosivity will
be greater than
that of chlorine,
which is handled
in existing fa-
cilities and,
hence, may be a
concern.
New facility needs to be
built or existing one
modified.
Corrosivity will be greater
than that of chlorine, which
is handled in existing facil-
ities and, hence, may be
a concern.
None.
Due to the alka-
line conditions
in the kiln,
corrosion is not
expected to be a
major problem.
(continued)
Build new
facility or
modify ex-
istii.g one.
Stainless steel
material of
construction
is believed to
be adequate.
However, suit-
ability of
using SS needs
to be estab-
lished. FRP
Build new fa-
cility or
modify existing
one.
Assuming that
the problem of
zinc coating
(for chloro-
picrin formula-
tion) can be
solved without
using excess
HC1, there
should be no
-------
TABLE 4-1 (continued)
Criteria
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zinc process
CTi
Extent of
Corrosion
(continued)
Process
Compatibility
witn
Pesticides
Compatible.
Compatible.
Compatible.
Secondary
Environ-
mental
Impact On
Health
Considera-
tions
Bromine emissions
of 0.1 ppm are
known to be a
health hazard.
However, test burn
results indicate
no bromine emis-
sions.
Bromine emissions of 0.1
are known to be a health
hazard.
ppm
Bromine emis-
sions of 0.1 ppm
are known to be
a health hazard.
In addition,
secondary impact
and and health
hazards can
arise duo to
POHC's.
1 ined mate-
rial could be
a possibility.
Carbon disul-
fide in the
formulations
is incompati-
ble with the
TEG catalyst,
which inhibits
the main de-
halogenation
reaction.
Possible due
vinyl bromide
and chloride
emissions.
Both are
known to be
carcinogenic
compounds.
problem of cor-
rosion. However,
if high acid con-
sumption is re-
quired to clean-
up the zinc sur-
face, then cor-
rosion could be a
major concern.
Carbon disulfide
reacts with zinc
and wil1 have to
be removed to re-
duce high operat-
ing costs.
Possible due to
volatile or-
ganics from
reaction mix-
ture. Ethylene
is known to
affect plants
and vegetation.
(continued)
-------
TABLE 4-1 (continued)
I
^J
Criteria
Mechanical
Reliability
Transporta-
tional Acess
to Facility
Storage and
Handling of
Pesticide and
Residue.
Cost
Permitting
Incineration in
presence of
sulfur wastes
Reliable.
Available.
No problems
envisioned.
Approximate
range is 50 to
80 cents/lb.
Some facilities
may require modi-
fication for EDB.
Others may need
to have new
permits. StiVl
others have the
appropriate per-
mits.
Starved air incineration
Reliable.
Railroad access available .
at some facilities. Mobile
incinerators are available.
No problems envisioned.
Unknown.
New permits would be re-
quired.
Cement kiln
Reliable.
Railroad access
available at at
least one will-
ing facility.
No problems
envisioned.
Approximately
$0.76-$1.3/lb.
Permit modi-
fication re-
quired.
ATEG
Reliable.
To be deter-
mined.
No problems
envisioned.
$0.34 toa
$0.78/lba
New permits
or modifi-
cations re-
quired.
Zinc process
Reliable.
To be deter-
mined.
No problems
envisioned.
$0.3 to $0.50/
lbD
New permits or
modifications
required.
Probability
of Success
Excellent. Incin-
eration in pres-
ence of sulfur
dioxide shows
excellent promise.
Fair. Test burns would be
required to judge per-
formance.
Good. Test
burns would be
required to
judge per-
formance.
Good. Pilot
scale testing
required to
establish
feasibility
Good. Pilot
scale testing
required to
establish fea-
sibility and
(continued)
-------
TABLE 4-1 (continued)
Criteria
Incineration in
presence of
sulfur wastes
Starved air incineration
Cement kiln
ATEG
Zinc process
Probability
of success
(continued)
oo
Time
Schedule
for
Completion
Results of a
recent test burn
show no bromine
emissions at the
stack and DREs
greater than
99.9999% for all
POHCs.
6 months or less.
and optimum
operating
conditions.
optimum operat-
ing conditions.
Approximately li to 2i
years.
Depends on al-
lowable feed.
Approximately 1J
years.
Approximately
2} years.
Approximately
21 years.
Cost excludes the cost for permitting, disposal of chlpropicrin stock, and land lease. Also, since steam,
water, and compressed air requirements are minimal, their contribution to operating cost was considered
negligible. The amount of pesticide treated excludes the amount of chloropicrin formulation. Also, these
cost figures are exclusive of development costs.
Cost has been calculated assuming that the problem of zinc coating with chloropicrin formulations can be solved
without adding excess hydrochloric acid. Also, cost excludes the cost of permitting and land lease. Since
steam, water, and compressed air requirements are minimal, their contribution to operating cost was considered
negligible. The amount of pesticide treated includes the total amount of pesticide that needs to be destroyed.
Also, these cost figures are exclusive of development costs.
-------
Because all of the preceding options use essentially the same process
design configuration, they are reviewed under one major option—destruction
of pesticide waste in a commercial incineration facility. Differences
arising from the varying modes of operation are highlighted whenever rele-
vant. Thermal destruction in a cement kiln is reviewed separately.
Incineration in a Commercial Incinerator
Status and Accessibility--
All of the options appear to be easily accessible, as indicated in the
following subsections.
Incineration in oxidizing conditions—This technology for the incinera-
tion of hazardous waste has been commercialized and is currently in operation
in numerous facilities nationwide. In fact, some of these facilities (IT,
VESTA, and ENSCO) offer mobile incineration systems, which can preclude the
need for transporting the hazardous waste. Most of these facilities have a
history of successful handling of hazardous wastes, and a few of them also
claim to have handled pesticides. During this evaluation, PEI contacted
several of these facilities to determine their interest in performing this
service and to gain their perspective on the applicability of the technology
to handle brominated waste. At least seven such facilities [VESTA, SHIRCO,
ENSCO (only if the operation can last more than a month), IT, Chemical Waste
Management, Rollins, and John Zink] indicated an interest in doing this work.
Incineration in the presence of sulfur dioxide—At least one commercial
facility in Europe is using this technology successfully for halogenated
wastes. Rollins Environm 'tal is currently planning to use this technology
for the present! application.
4-9
-------
Starved-air incineration--International Technology Corp. and John Zink
have expressed an interest in trying this option in addition to others. John
Zink has a design for a starved-air incinerator which could be used for EDB
application.
Past Experience--
Incineration in an oxidizing atmosphere—This is the standard mode of
operation for all commercial facilities. Many facilities are experienced in
the incineration of chlorinated wastes; however, few, if any, have had ex-
perience with brominated wastes. The problem associated with destroying
brominated wastes in excess-air incineration is not one of achieving the
necessary destruction efficiencies; rather,'it concerns containment of the
undesirable bromine (Br2) emissions. The literature indicates tnat the
bromine sinks in water and forms an irritating brown vapor (U.S. Coast Guard
1984). One facility (John Zink) reported the formation of a fog in the
scrubber caused by bromine, which resulted in a substantial reduction in
scrubber efficiency (J. Cegielski, personal communication with EPA). Most of
the bromine escaped into the atmosphere. Thus, conventional incineration of
brominated waste might generate bromine emissions.
On the other hand, it has been suggested that hydrogen bromide (HBr) can
be scrubbed efficiently in existing scrubber systems. If an incinerator can
reduce bromine to HBr, the use of commercial incinerators may be a viable
option for destroying pesticide formulations. The bromine/HBr thermodynam-
ics, however, favor the formation of bromine under incinerator operating
conditions (Eicher, Cudahy, and Troxler 1985). The equilibrium reaction is
as follows:
Br2 + H20 ->• 2 HBr + 1/2 02 (10)
4-10
-------
Figure 4-1 shows in separate plots the equilibrium constant Kp against
temperature for chlorine and bromine. At 1800°F, the equilibrium constant
for bromine is about 5.9 x 10" , as opposed to 12.5 for chlorine. The result
is almost total bromine formation in th? incinerator. A study by Eicher,
Cudahy, and Troxler (1985) indicates about 75 percent of the bromine goes to
Br- in a typical incinerator. Thus, preferential formation of HBr in the in-
cinerator is possible only if the operating conditions are changed to favor
HBr formation thermodynamically or if the bromine is made to react with some
other reagent in situ to produce HBr. The former can be achieved by starved-
air incineration, whereas the latter can be achieved by reacting the fir-
formed with SCL to form SCL and HBr.
Incineration in the presence of S0?--This process modification of con-
ventional incineration is currently being operated successfully at an incin-
eration facility of Bayer, German}. This facility has been in operation
since 1977 and reportedly has incinerated halogenated wastes, "to a degree
where not even traces of halogen can be identified by analysis" (Fabian, et
al., 1979). In the United States, the only company claiming to have substan-
tial experience with brominated wastes is Rollins Environmental. Recently,
EPA conducted a test burn of the EDB formulation at the Rollins, Deer Park,
Texas, facility. The important results of the test burn (Alliance Corp.,
1988) are as follows:
DREs for the three POHCs (EDB, EDC, and CC1J easily met the "four
nines) RCRA requirements. The destruction of EDB in particular was
greater than 99.9999 percent.
There were no visible bromine emissions at the stack.
Continuous emissions monitoring data for C02, 02, NO , and SO,
nif-ets the required standards.
All organic bromine fed to the system reportedly exits the system
through the scrubber water stream.
4-11
-------
a
x
o
2.0
1.8
1.6
1.4
1.2
1.0
0.8
0.6
0.4
0.2
0.0
-0.2
-0.4
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
TEMPERATURE, °F
Equilibrium Constant (Kp) Curve for HCI/CI2
•1
Q. -2
*
o
*- -3
1000 1100 1200 1300 1400 1500 1600 1700 1800 1900 2000 2100 2200 2300 2400 2500
TEMPERATURE, °F
Equilibrium Constant (Kp) Curve for HBr/Br2
Figure 4-1. Equilibrium constant Kp against temperature for chlorine and bromine
(Eicheretal. 1985).
4-12
-------
Four Nines (1987) have reported similar observations during incineration of
iodine. Thus, this option seems to have excellent capability to successfully
destroy the EDB formulations.
Starved-air incineration—There is no known experience of destroying
hazardous waste with this technology. However, personal communications of
Mr. John Cegielski, John Zink Co., with EPA indicate that John link has a
design for this process which could be used to eliminate the brominated
waste.
Need for Development--
The preceding discussion points up an obvious need for development to
solve the problem cf bromine emissions. This problem can be°handled in two
ways:
1. Modification of the process, which would involve some change in the
process conditions to obtain preferential conversion of all the
bromine to HBr.
2. Modification of the scrubber design or operation to obtain
successful removal of all the bromine from the flue gases.
Under the first choice, the most viable options are incineration in the
presence of sulfur dioxide and starved-air incineration (partial pyrolysis).
Incineration in the presence of sulfur dioxide—Results of the recent
test burn at the Rollins, Deer Park, Texas, facility show that destruction of
EDB formulations is complete and meets all the necessary standards regarding
DREs and emissions. Thus, it is believed that no further development is
required for this process.
Starved-air incineration--The chemical equilibrium for the Br?/HBr
system is represented by Equation (10). The eouilibrium constant is given by:
4-13
-------
[02]j
LBr2j IH,OJ
Thus, it is theoretically possible to shift the equilibrium in favor of HBr
by:
1. Operating the furnace at the highest possible temperature so as to
obtain the highest equil-ibrium constant (refer Figure 4-1).
2. Operating the furnace at low oxygen concentrations (less excess
air). In fact, if the oxygen concentrations could be reduced to
zero, almost all of the bromine will form HBr.
3. Operating the furnace at high water partial pressures which ensures
the availability of hydrogen for the reaction to form HBr.
However, in most commercial facilities, due to their present design and
operation (Four Nines, 1987), it may not be possible to operate the auxiliary
burners at high temperatures and high water ratio. Moreover, most commercial
facilities use kilns which operate at lower temperatures. Equilibrium calcu-
lations under the typical operating conditions that would prevail in a com-
mercial incinerator, operated under starved-air conditions (Appendix A),
indicate that a substantial amount of Br- would still be formed because of
the very low value of the equilibrium constant Kp. A considerably more sig-
nificant question, however, concerns the change of equilibrium as the furnace
gases are cooled to approximately 200CF before they enter the scrubbing sys-
tem. The equilibrium gas composition should be calculated at this tempera-
ture, rather than in the furnace. The lowest temperature at which the Kp
value for the Br«/HBr system is available is about 1000°F. The Kp value
decreases rapidly with decrease in temperature; hence, it may be expected to
be much lower than that at 1000°F. Equilibrium calculations at 1000nF (Ap-
pendix A) indicate almost complete bromine formation. Thus, as the gases
enter the scrubber, most of the HBr would have been converted to Br. and the
situation would be no better than that which occurs during conventional,
4-14
-------
excess-air incineration. Thus extensive testing will he required to estab-
lish the feasibility of this option. Moreover, the literature indicates that
halogenated wastes are not likely candidates for the starved-air incineration
(Bonner et al. 1981).
Conventional incineration—If modifying the process in the incinerator
operation is not a desirable alternative, or if such modifications fail to
produce the desired results, the scrubber section may offer a solution. It
has been suggested that an efficient scrubber design is a must for halogenated
waste incineration, and packed towers or plate columns have been recommended
for this purpose. Although venturi scrubbers may be used in a few cases,
they do not offer a desirable alternative, as they do not provide good gas-
side mass-transfer characteristics. Even with efficient scrubbers, removing
bromine from the flue gases will be a difficult task. Some of the process
modifications that might be successful are as follows:-
1. Sulfur dioxide can be mixed with the incinerator exit gases and the
mixture absorbed in water with very dilute amounts of sulfuric acid
in a packed column. D. vaji Velzen et al. (1978, 1979, 1980) found
this method to yield almost complete conversion of bromine and S02.
Two studies present relevant details on the mass-transfer character-
istics and design parameters (D. van Velzen et al., 1978, 1979).
The experimental data presented in these two studies indicate a
good potential for this mode of operation. Nevertheless, a test
run would be desirable to test the efficacy of this modification
and to determine any associated operational problems.
2. The other option involves the use of an efficient solvent that can
readily absorb bromine. Solvents that can undergo rapid reaction
with bromine are preferable. Some suggestions include ammonia
solution, caustic solution, and lime slurry. No data are currently
available on the performance of these solvents; however, it is
interesting to note that ammonia is recommended for containment of
bromine spillage. Thus, ammonia solution may be a worthwhile
consideration. .
In conclusion, a few options hold promise; however, the data base is too
sparse to predict the extent of success or the approximate costs entailed.
4-15
-------
Trial burns are almost imperative for assessing the feasibility of the avail-
able option?.
Other options—One of the options suggested by the IT Corporation in-
volves the introduction of caustic solution 1n the combustion chamber. The
idea is to react the bromine in situ. This option will provide a homogeneous
•phase reaction between the aUali and bromine, and the reaction rates could
be high. It also provides a longer residence time for the gases with the
alkali to achieve the necessary levels of reaction (removal). Thus, this may
be a possible area for development.
General Criteria--
The criteria presented in the following subsections are applicable to
all nodes of operation.
Need for additional equipment—All the incineration options considered
in this discussion involve the same basic process design configuration and
equipment, and all can be performed in an existing commercial facility.
For some options, additional equipment or components may be needed to adapt
an existing system for any of the process operations. A good corrosion-
resistant refractory lining would be needed in the furnace, and the flue gas
ducting may have to be rc?de of fiberglass-reinforced plastic (FRK). The
scrubber system also may have to be refractory-lined or Teflon-lined for
corrosion resistance. Specific modifications, if any, required for starved-
air incineration are unknown. Incineration in the presence of S0» does not
appear to require any modifications.
Toxic emissions end secondary environmental impact—It is evident from
the previous discussion that incineration in the presence of sulfur-containing
4-16
-------
waste or SO- does not pose any problem of toxic emissions. However, the
potential for possible toxic emissions and secondary environmental impacts
due to bromine and HBr emissions for all other modes of incineration is
evident from the preceding discussion. Secondary environmental impacts may
arise from the possibility that these compounds can enter the water intake
system and thereby pose a danger. Both compounds, especially bromine, are
known to be harmful to aquatic life.
In addition, the starved-air incineration option could emit partially
oxidized organics, which may be carcinogenic and very harmful.
Compatibility—The process is compatible with the constituents of the
pesticide formulations.
Residues—The brominated scrubber solution and sludges will require
careful disposal. In addition, residues from starved-air incineration may
contain significant amounts of partially oxidized organics, which will have
to be carefully handled or destroyed.
Safety and health hazards—When bromine comes in contact with the skin,
it can cause acne and slow-healing ulcers. Inhalation can cause severe
irritation of respiratory passages and pulmonary edema. A brief exposure to
1000 ppm may be fatal. The TLV for bromine is 0.1 ppm, and the short-term
inhalation limit is 0.4 ppm for 30 minutes. Indications are that bromine
cannot be tolerated even at low concentrations. Odor threshold and IDLH for
bromine are 3.5 and 10 ppm, respectively (U.S. Coast Guard 1984).
Inhalation of HBr causes severe irritation of the nose and upper res-
piratory tract and lung injury. Skin contact can result in irritation and
burns. The TLV for HBr is 3 ppm; the IDHL value is 50 ppm (U.S. Coast Guard
1984).
4-17
-------
Unfortunately, no one knows the ambient concentration levels at which
these compounds can have a pronounced health effect, and no regulations on
the allowable emissions for bromine have been established. Only four states
are currently in the process of promulgating regulations for HBr (RTI, 1987).
As a result, the extent of removal efficiencies required at the scrubber
outlet cannot be quantified.
Another aspect of safety and health concerns the workers at the incin-
eration facility. Because most of the constituents of the waste are ex-
treme1> hazardous, operators may have to be equipped with safety equipment
such as goggles, self-contained breathing apparatus, and rubber overclothing
(including gloves).
Assuming that the problem of emissions can be solved with some process
modification, the incineration process is safe. The flue gas would then
consist mainly of carbon dioxide and water, both of which are harmless. When
properly operated, this process poses little possibility of explosion. Sev-
eral commercial incinerators are now operating nationwide. Also, incinera-
tion is unlikely to generate any solid or liquid residues that would be
hazardous and could not be handled in a routine manner.
Transportation access to facility—Among those commercial facilities
that are interested in undertaking this task, IT, VESTA, ENSCO, and SHIRCO
have transportable units that can be set up at the worksite. Rollins and
Chemical Waste Management are known to have railroad access at their facil-
ities; it is believed that John Zink also has railroad access.
Storage and handling of waste—Another aspect of plant safety and toxic
emissions concerns the handling and storage of the hazardous waste. The
4-18
-------
possibility of spil'age and emissions is at a maximum during these activities.
Because most commercial incineration facilities handle hazardous wastes on a
regular basis, they are believed to have well-designed storage and handling
facilities. Most of the constituents of the waste under consideration [e.g»,
EDB, ethylene dichloride (EDC), chloropicrin] are extremely hazardous and
warrant the use of special precautions in their handling. As a part of an
incineration program, a comprehensive spillage-control action plan should be
prepared, and all operators should be trained to implement it. If attention
is given to these items, it is believed that the handling facilities avail-
able at existing incinerators should be capable of dealing with pesticides.
Preprocessing—Because of the- small quantity of the waste that needs to
be destroyed, it may be blended with other wastes before incineration. Care
should bo exercised to make sure that the pesticide constituents are compati-
ble with other chemicals in the blend.
Corrosion and mechanical reliability—Bromine and HBr are highly corro-
sive to most metals (U.S. Coast Guard 1984). Bromine is also known to have a
corrosive effect on the refractory lining in the furnace. No data ?re cur-
rently available on the extent of corrosion due to bromine and HBr in an
existing incineration facility; however, most commercial facilities are
designed to handle hydrogen chloride (HC1), which is also extremely corrosive.
Comparative data on the corrosivity of bromine/HBr and HC1 are unavailable;
hence, corrosion could be a potential problem. Industry sources, however,
have indicated that the existing material of construction should be able to
withstand any corrosive attack due to bromine/HBr. Reportedly, one facility
in Europe had major corrosion problems on parts in the gas processing sec-
tions that were made of Hastelloy C (Fabian, Reher, and Schoen 1979). These
4-19
-------
problems could be solved by the use of Teflon, FRP, or high-quality refrac-
tory bricks. Thus, the mechanical reliability of the equipment involved is
directly related to the corrosion resistance of the material of construction
used. Inasmuch as operating incineration facilities have had extensive
experience with chlorinated wastes, these facilities are assumed to be reli-
able mechanically.
0
Cost—The costs quoted by vendors during preliminary discussions ranged
from 50 to 80 cents/pound of waste; however, these costs could change as the
vendors gain a better grasp of the modifications/complexities involved in the
process. Cost breakdowns for each operating option are currently not avail-
able. However, EPA has proposals from Rollins Environmental and John Zink
for this job. Rollins will use S02 incineration technology, and John Zink
has proposed the use of starved-air incineration.
Permits—Most of the facilities contacted did not have permits to handle
pesticides. Some of the newer facilities (VESTA, IT) have permit applica-
tionsopending for the handling of hazardous wastes. A few facilities (Rol-
lins and Chemical Waste) have indicated that they have permits for handling
pesticides which may or may not need modification to include the handling of
EDB. Rollins may not require the permit modification.
As mentioned earlier, many commercial incineration facilities were
contacted for their input. Appendix C summarizes the information obtained
from most of these facilities.
Probability of Success—As reported earlier, the option of incineration
in the presence of S0? has been successfully demonstrated for EDB destruction
during a test burn at the Rollins, Deer Park, Texas, facility. The incinera-
tion meets all the required standards for DREs and emissions. Moreover, this
4-20
-------
option is reportedly in operation at a Bayer facility in Germany with suc-
cess, to eliminate halogen emissions during incineration. Thus, incineration
In presence of S(L holds excellent promise to destroy the EDB formulations.
As regards other incineration options, trial burns will be required to
demonstrate their capabilities, although both conventional incineration and
starved-air incineration appear to present some significant problems.
Time Schedule--Present1y, test burns for incineration in the presence of
SO- have been completed successfully. Preliminary test results indicate that
this option can successfully eliminate the EDB wastes. Moreover, Rollins has
obtained a permit to handle pesticides that does not even require modifica-
tion for EDB. Thus, this option should be able to be used to dispose of all
EDB formulations within sir months or less.
ADDITIONAL INCINERATION OPTION
Another option for eliminating the EDB waste, which has not been consid-
ered in this evaluation, has been suggested by Four Nines, Inc. (1987). In-
formation obtained from Four Nines (Four Nines, 1987) indicates that the
conditions necessary to change the Br^/HBr equilibrium in favor of HBr (dis-
cussed under starved-air incineration) can be obtained by using a high
intensity burner design (Trane Thermal, John Zink) fired into a liquid injec-
tion incinerator. These incinerators can operate at low excess air (0 to 15!
excess air), high temperatures and high water content. The EDB would be
injected through a steam atomizer along with auxiliary fuel to maintain the
combustion temperature above 2400°F. Water injection nozzles (combined with
EDL or adjacent to EDB nozzle) would be used to provide high water vapor con-
tent in the combustion gases to drive the equilibrium in favor of HBr forma-
tion. The off gases would be subjected to an adiabatic quench followed by
4-21
-------
conventional scrubbing. This operating scheme is believed to have been im-
plemented successfully with chlorinated compounds. Four Nines believes that
this scheme should be able to handle the problem of bromine emissions. How-
ever, they recommend prior testing in a pilot or a commercial facility that
has a high intensity combustor to establish the performance capabilities.
According to Four Nines such burners are available at Trane Thermal pilot
plant, John Zink pilot plant, ENSCO, Chemical Waste Management (Chicago) and
Rollins Environmental.
Cement Kiln
Status and Accessihility--
A number of cement kilns now in operation incinerate hazardous wastes
under the Hazardous Waste Fuels Program. Two kilns, Dundee Cement Co. and IB
Farge Cement Co., have shown an interest in the incineration of delated
pesticides. Thus, the technology is accessible.
Past Experience--
Cement kilns have been burning chlorinated waste fuels as part of the
ongoing Hazardous Waste Fuel Program. When waste lubricating oils containing
an average of 0.15 percent bromine were burned in a dry cement kiln at the
St. Lawrence Cement Company in Ontario, about 99.3 percent of the bromine in
the feed stream was reportedly captured in the pelletized dust, and some
bromide was captured in the clinker (Berry 1975). The percentage of bromine
in the pesticides, however, is substantially higher. No further test or
operating data on incineration of brominated waste in a cement kiln were
available. Cement kiln operators, however, believe that brominated wastes
will present no new problems.
4-22
-------
Need for Development—
Despite a history of hazardous waste incineration in cement kilns, the
burning of halogenated wastes has been limited to those containing chlorides.
Cement kilns are ideal for burning chlorinated wastes because, in proper
quantities, chlorides enhance product quality by combining with the potassium
and sodium that might be present in the ore. Further, the alkaline char-
acteristics of the kiln atmosphere abate hydrogen chloride emissions. °In
essence, the kiln is its own scrubber. Kiln operators expect the bromide
waste to act similarly. Because most cement kilns operate without wet scrub-
bing systems, however, toxic gases from the kiln exit irtc the atmosphere.
Trial burns are therefore required to ascertain the extent of bromine removal
end whether additional scrubbing capacity is needed.
Need for Additional Equipment--
Cement kilns are equipped with systems for the control of particulate
emissions. Those that burn hazardous waste fuels are scrutinized even more
closely to ensure that particulate emissions are minimized. Therefore, the
existing facilities should be well equipped to handle particulate emissions.
As mentioned earlier, however, most facilities do not have ges treatment
systems (chemical treatment). If the flue gas were to contain substantial
amounts of bromine, additional scrubbing system would have to be installed.
Emissions and Secondary Environmental Impact--
The following discussion highlights the possibility of bromine emis-
sions. In addition, products of incomplete combustion (PIC's) are a definite
possibility, and these may be more toxic than the original waste.
4-23
-------
The possibility of PIC ^missions is a serious concern with respect to
combustion of hazardous wast iuels in cement kilns because ring formation in
the chamber can cause the raw feed to cascade towards the torch under ava-
lanche-like conditions. This casrade pushes tfye gases before it quenches the
flame and thus causes a localized increase in pressure. Reportedly, the
flame loss will generate PIC's and the pressure rise will cause the PIC's to
discharge through the seals at the torch end of the kiln.
The hazardous effects of bromine emissions were discussed in detail
under the incineration option. The secondary environmental impacts of the
potentially dangerous PIC emissions are not known, as published test reports
contain no substantive information regarding PIC emissions from a kiln.
Based on the high POHC destruction efficiencies indicated in these same
reports, however, emissions are believed to be insignificant (PEI 1987).
Another aspect of toxic emissions concerns particulates. Whether the
introduction of bromine will increase particulate emissions appears to depend
on what effect the compound has on the particle size of the emissions. The
brominated compounds are expected to react similarly to chlorinated com-
pounds. The latter form hydrogen chloride (HC1) and chlorine (Cl?) as the
chlorinated compounds are oxidized in the combustion process. These, in
turn, react with the alkali components in the cement feed to form volatile
alkali chlorides such as potassium chloride (KC1) and sodium chloride (NaCl),
which condense into a fine fume. Chlorine also promotes a buildup of mate-
rial on the wall of the kiln that forms a restrictive ring inside the kiln as
it rotates. This phenomenon ("ring") restricts the cross-sectional area,
which increases the combustion gas velocity and causes more clinker dust to
4-24
-------
be carried through the exhaust system. Notwithstanding this scenario, tests
have shown that no correlation exists between participate emissions and the
chlorine content of the waste fuel (see Figure 4-2, PEI 1987). Because
bromine is less volatile than chlorine, the kiln's dust collection system
should adequately abate emissions.
Compatibility—
The process is compatible with the pesticide formulations.
Process Safety and Health Hazards--
Cement kilns have been operating for a long time without any major
safety problems. Thus, the basic process is believed to be inherently safe.
The possibility of an explosion or fire is also minimal. On the other hand,
safety and health hazards could arise from toxic emissions, such, as health
hazards due to bromine emissions, which were discussed in the section on
incineration. Products of incomplete combustion are also regarded as a high
health risk and carcinogenic.
Transportation Access to Facility—
The LaFarge facility has railroad access, and the same is believed to be
true of the Dundee facility.
Storage and Handling —
Another aspect of plant safety and toxic emissions concerns the handling
and storage of the hazardous waste. The possibility of spillage and emis-
sions is at a maximum during the handling and storage of the waste. Most of
the'constituents of the waste under consideration (EDB, EDC, chloropicrin,
etc.) are extremely hazardous and warrant the use of special precautions in
4-25
-------
i
o*
I to1
g o
to o •
OS
UJ It
3 uf
3O 0
o> 10
r iu
-------
their handling. As part of an incineration program, a comprehensive spil-
lage-control action plan should be prepared and all operators should be
trained to implement it. The overall layout and design of this area must
conform with the standards suggested for hazardous wastes (Bonner et al.
1981). Whether the current design features of cement plants conform is not
known. If not, the plants would be more vulnerable in case of accidental
o
spillage.
Preprocessing--
The liquid pesticides must be blended with other liquid fuels in proper
ratio to ensure the correct bromine content, as too much bromine could cause
emission problems. During the blending of the different liquids, care should
be exercised to ensure that the mixture components are compatible. Blending
of wastes is relatively routine and requires no additional work or equipment.
Corrosion and Mechanical Reliability--
Bromine and HBr are both very corrosive to most metals. No data are
currently available on the extent of corrosion due to bromine ana HBr under
the conditions prevailing in a cement kiln. Because of the high alkaline
environment in the kiln, however, corrosion should not be a serious problem.
Therefore, it is believed that the existing material of construction should
be able to withstand any corrosive attack due to bromine and HBr. The kiln
refractory lining would have to be of high quality and corrosion-resistant.
The exhaust gas processing unit may have to be refractory or FRP- or Teflon-
lined, depending on the operating temperatures.
4-27
-------
Handling and Disposal of Process Residues--
Cement kiln dust is not a hazardous waste, and it is currently exempt
from the requirements of RCR'A. If brominated compounds were burned in a
cement kiln, the content of bromide salts in the residue would increase;
however, this would not raise the toxicity. of the dust, nor would it be
likely to cause a significant increase in the amount of cement dust that is
"wasted" (as a result of the fineness of the sa\t fumes). As with chlorinat-
ed compounds, the major'portion of the bromine is expected to end up in the
clinker as product.
Cost —
The cost figures quoted range from approximately SO.76 to SI.3/1 b.
Permitting—
La Farge has an RCRA permit but would require modification, whereas the
Dundee facility will need a permit.
Probability of Success--
From a theoretical point of view, the cement kiln option seems to hold a
good potential for success. However, trial burns will be required to demon-
strate the capabilities, and to establish the optimal feed rates.
Time Schedule—
The optimal feed rate will decide the overall time period to destroy all
the pesticide formulations. Assuming that acceptable feed rates are pos-
sible, this option should take about a year and a half to complete the ope-
ration.
i
4-28 i
-------
CHEMICAL DESTRUCTION^
This option entails two approaches to destroy the EDB formulations. The
first approach involves distillation of some of the formulations (CS- con-
taining and chloropicrin containing) to recover individual components fol-
lowed by the destruction of EDB recovered with one of the chemical processes.
The other approach involves destruction of the pesticide formulations without
any pre-processing. Recent bench-scale testing at the R&D facility of Inter-
nation Technology (IT) indicates that the distillation of the CS~ containing
formulations is very easy. The components recovered are quite pure and could
be sold for market value. However, distillation of the chloropicrin formu-
'lation was found to be difficult because of an azeotrope formation. Also, IT
has learned from industry sources that the distillation of chloropicrin can
be very dangerous, and hence, this option was not pursued. The feasibility
of distilling the miscellaneous formulations was not investigated.
The engineering evaluation°of the chemical processes is discussed in the
subsequent discussion. This evaluation is based on the t«»st work results,
preliminary process calculations and cost estimates.
Zinc Process
Status and Accessibility—
The zinc (Zii) process is still in the conceptual stage. Recent labora-
tory tests on pesticide formulations have given mixed results as regards the
o
EDB destruction efficiencies; however, extensive pilot-scale testing would be
required to demonstrate process feasibility, performance capabilities (DREs),
process economics, and commercial reliability.
4-29
-------
IT -
I
Past Experience--
No one has had previous experience with a full-scale zinc process. The
following important observations were made during recent bench scale testing
of the zinc process with carbon disulfide containing and chloropicrln con-
taining formulations at IT:
* Destruction of pure EDB (obtained by distillation of carbon disul-
fide containing formulations) was 99.93 percent complete, requiring
very little (Q.008 gms of HCl/gm of EDB) acid. However, the above
result was obtained after keeping the reactants in contact for 4?
hours with 10.5 hours of total agitation ti/w.
* For the chlorcpicrin formulations, destruction of FD8 was about 90
percent while that of chloropicrin was about 93 percent. The prod-
ucts of chloropicrin reaction could not be identified (Test results
/ indicate formation of some unknown compounds). Acid consumption
was 0.46 gms of HCl/gm of pesticide, while the zinc consumption was
0.74 gms of zinc per gran of pesticide (about 5 times the theoreti-
cal reouirement). Also, it has been reported that the reaction
proceeds very slowly since the zinc particles get coated with the
reaction products. The zinc particles had to be replaced once
during the reaction to improve the reaction rate and efficiency.
Intense agitation and a pH of less than 1 were required to keep the
zinc surface clean of any coating due to the reaction products.
The overall agitation time for this reaction was about ?3.5 hours
(the reaction mixture was studied for more than 5 days). The
reactor effluent gas wac found to contain ethylene and hydrogen.
* For the carbon disulfidi containing formulations the EDB destruc-
tion was about 23 percent after 11 hours of agitation. This poor
destruction may be attributed to the fact that carbon disulfide
seems to react with the zinc producing some sulfide compounds
(carbon disulfide destruction was 81 percent and both the liquid
and gas phases had sulfide odor). Also, carbon tetrachloride was
found to react with the zinc (31 percent conversion). However, the
reaction products have not been identified. The acid consumption
was approximately 0.34 gms/gm of pesticide. The gaseous reaction
products were ethylene and some amount of hydrogen.
In addition to the above, it is now known that the zinc rtehalogenation
reactions are highly exothermic, creating a problem of heat removal (Appendix
A). It is also clear that the CS? containing formulations would have to be
distilled prior to treatment by the zinc process.
4-30
-------
Need for Development—
The need for further development of the zinc process cannot be over-
emphasized. An extensive test program would need to be undertaken to estab-
lish the engineering feasibility and DREs attainable, to collect process
design data, and to determine potential problem areas. Some of the areas
that should be researched during the test runs include:
1) The zinc process does not seem to give the 4 nines destruction
efficiency with the chloropicrin formulation. More tests would be
needed to establish if 4 nine DREs are attainable, or to study the
feasibility of distilling the chloropicrin formulations. The test
results seem to indicate that the presence of other constituents
hinder the EDB reaction giving lower DRES. Thus, mere tests would
be required to establish the DREs attainable with the miscellaneous
formulations.
2) The test results seem to indicate very long reaction times for
achieving maximum destruction. However, the test work does not
provide any kinetic data to establish reaction rates. The zinc
dehalogenation reaction is highly exothermic which could lead to
problems of heat removal. Depending upon the relative rates of the
reaction and heat transfer, the overall process rate would be con-
trolled by either the reaction kinetics or the heat transfer. If
reaction rate is much faster than the rate of heat transfer (rate
of reaction » rate of heat removal), then the overall rate would
be controlled by the rate of heat transfer (an assumption in the
process calculations - Appendix A). On the other hand,if the
reaction rate is much lower than the rate of heat transfer (rate of
reaction « rate of heat transfer), then the overall rate would be
governed by the reaction kinetics. As can be seen from the above,
it is important to establish the controlling mechanism, as it will
affect the process desian, time of operation, and hence, the over-
all process economics.
3) Preliminary test results indicate high zinc and acid consumptions
(especially for the chloropicrin formulations). However, it is be-
lieved that the problem of zinc coating can be solved without using
excess acid. The cost estimates for this process were developed
assuming very minimal acid requirements. However, the problem of
zinc coating and acid requirement would have to be studied in more
detail in subsequent tests. If the acid requirements are high,
expensive material of construction would be required to withstand
the corrosion. The use of comron polymeric materials of construc-
tion may be difficult because of the heat transfer requirements of
4-31
-------
the system and because most common synthetic materials are not-
recommended for halogenated hydrocarbons (MelIan, 1976). One
suggestion made during the senior technical review involves slur-
rying the zinc with water in the reactor after which the pesticide
and the concentrated acid would be added at a metered rate. The
feasibility of this, and any other option would have to be studied
by conducting pilot-scale testing.
4) The extent to which Zn reacts with other constituents of the pesti-
cide formulations [carbon tetrachloride (CCl^), chloroform, ethyl-
ene dichloride (EDC), sulfur dioxide (S02), carbon disulfide (CS2),
etc.] and the products of these reactions should be determined.
Test results indicate formation of unknown products. All unknown
products would have to be identified to decide if the effluents are
hazardous or otherwise. These data will be valuable in deciding
the ultimate disposal methods for the reactor effluents and in
estimating the reagent requirements. Both the factors have a
significant impact on the operating costs.
5) The feasibility of recovering byproducts of value from the reactor
effluent stream needs to be studied. Thus, the reactor effluent
must be characterized and a process design must be developed to
accomplish this task.
Other Areas of Concern—
Equilibrium calculations for the initial reaction mixture show that
the mixture would boil between 60° and 70°C. Therefore, using a
vacuum pump to remove the ethylene could result in a significant
loss Gf organics into the vapor phase. It would be difficult to
condense these organics under sub-atmospheric pressures using an
overhead condenser. This creates a possibility of emissions prob-
lems. Moreover with vacuum in the system, there is a possibility
uf air leaking into the system due to some malfunction. The mixture
of air and the gases in the system (ethylene, and other organic
vapors) could create an explosion hazard. Thus, it is felt that
this aspect should be studied in future tests with this process.
The reaction gases could be removed from the system by flowing an
inert gas (like nitrogen) or by putting a fan between the reactor
and condenser. The former was assumed for the preliminary cost
estimates. Using an inert gas will reduce the partial pressures of
the pesticide constituents in the vapor phase which will reduce the
volatility, and hence, loss of organics.
Need for Additional Equipment or New Equipment--
Two options were available for the zinc process to destroy pesticide
formulations. The first involved the construction of a new facility; the
4-32
-------
second entailed the use of some existing equipment from the GARD facility in
combination with new equipment. Reactor vessels, a filter press, and a flare
and stack are available at the GARD facility; the remainder of the equipment
will have to be new. If further tests determine high acid requirements,
however, the GARD option will become infeasible because of corrosion con-
cerns. The need to construct a new facility would affect the time required
to complete the overall project.
Tcxic Emissions and Secondary Environmental Impact—
Currently, the only possible emissions from the process operation appear
to be due to volatile organics escaping the reaction system and ethylene gas.
The secondary environmental impact due to volatile organics from pesticide
formulations needs to be studied in detail. Ethylene is reported to be
herbicidal and is known to affect vegetation. The other probable source of
emissions would be the handling of the pesticide itself.
Compatibility--
Bench scale tests show that the CS2 in the formula".ions reacts vigor-
ously with the zinc, thereby retarding the rate of EDB reaction. Thus, CS~
would have to be removed by distillation prior to treatment of the pesticide
formulation by zinc process.
Safety and Health Hazards-
One of the major products of the process is ethylene gas. Because
ethylene is highly flammable, the probability of a fire hazard would be high.
In addition, the reactor off gases may contain some volatile organics which
increase the explosion and fire risks. Thus, process safety should be a
4-33
-------
primary concern. Ethylene, however, is not as dangerous to handle as is
acetylene; in fact, its handling is routine in the petrochemical industry.
If proper care is exercised, the probability of a major accident could be
minimized. If further studies indicate high acid requirements to keep the
zinc surface active for reaction, then hydrogen may also be formed by the
reaction between Zn and HC1. This would create additional safety concerns.
Most of the constituents of the formulations are very harmful; there-
fore, the operators who handle these wastes fould have to wear special cloth-
ing. Should a fire occur, toxic emissions from the burning of these formula-
tions are also a possibility.
Transportation Access to the Facility--
This access would have to be provided during construction.
Storage and Handling of Waste--
Another aspect of plant safety and toxic emissions concerns the handling
and storage of the hazardous waste. Most of the constituents of the°waste
under consideration (e.g., EDB, EDC, chloropicrin), are extremely hazardous
and warrant special precaution during handling. Before undertaking a chem-
ical destruction program, a comprehensive spillage control action plan should
be prepared, and all operators should be trained to implement it. Because
several commercial facilities handle hazardous waste regularly, designing -
this area of the plant should be fairly routine.
Corrosion and Mechanical Reliability--
It is believed that the problem of zinc coating can be solved without
consuming large quantities of HC1 (e.g., using intense agitation). If this
4-34
-------
belief could be validated during further testing, then corrosion should not
be a major concern. However, if substantial amount of hydrochloric acid
(HC1) is required in the reactor to clean the surface of zinc particles, then
corrosion would be an important consideration, necessitating the use of
expensive materials of construction.
High-powered mixers would be required to keep the zinc particles in the
Yeactor mixture in suspension. The extent of erosion-corrosion due to the
zinc (if any) may need to be researched; however, it is not expected to pose
a major problem.
Handling and' Disposal of Process Residues--
Test work would have to establish if the process effluents are hazardous
or nonhazardous. Depending on the nature of the effluents, appropriate dis-
posal methods would hove to be adopted. These can have a significant impact
on the total cost.
The process generates a large quantity of effluents. Disposal of these
effluents has & significant impact on the overall process economics. Al-
though pure zinc bromide and chloride have a good market value, it is not
known if there will be any demand for these compounds obtained as the by-
products of a hazardous waste treatment. The feasibility of recovering
byproducts of value would have to be established on the basis of test work.
Zinc bromide soluncn has applications in enhanced oil recovery. It is not
known if the aqueous effluents from the process could be shipped to an
enhanced oil recovery facility for value. Hence, for this analysis the worst
case is assumed where the effluents have to be disposed of as hazardous
4-35
-------
wastes. Since the organic and aqueous phases are immiscible, they can be
easily separated in a phase separator. It is assumed that the organic phase
is destroyed by incineration at 50 cents/pound. . Inquiries wi-.h industry
sources indicated that the aqueous waste could be disposed by deep well
injection or in a landfill or a wastewater treatment facility. The disposal
costs quoted varied from 12 cents/gallon for deepwell injection to about 90
cents/gallon for waste stabilization and disposal. Thus, a disposal cost of
50 cents/gallon has been assumed. However, if the actual costs differ
substantially frorr the above estimates, the overall cost would change sub-
stantially.
Cost-
Preliminary study cost estimates have been developed for this process
(Appendix B) based on approximate process calculations (Appendix A) to size
the equipment and estimate reagent consumption. It has been assumed that the
CS- formulations are distilled prior to destruction. The recovered products
(CC14 and CS^) are sold at market value. The cost of recovering reaction
byproducts was excluded because of lack of sufficient data. Instead, the
worst case of effluent disposal as a hazardous waste has been assumed.
The primary cost estimates have been developed assuming that the problem
of zinc coating is solved without using excess HC1. Two economic options
have been considered: a) build a new facility, and b) use some equipment at
the SARD facility. Cost estimates have been developed for two situations:
1) the government owns and operates the facility, and 2) the work is subcon-
tracted to a small scale chemical firm. Subcontracting the destruction to
4-36
-------
another company is not applicable to the option of utilizing the GARD equip-
ment. Under the first option the entire fixed capital cost is included under
the carrying charges, while under the second option the depreciation charges
are included under the carrying charges. The depreciation charge has been
estimated assuming straight line depreciation, 10 year life span and zero
salvage value. The unitized cost for treating all the formulations, under
each operating option is:
New facility
Government owns and operates the facility: J0.50/lb
Subcontracted to small chemical firm: $0.3/lb
Option of utilizing GARD equipment
Government owns and operates facility: $0.44/lb
The above estimates do not include the costs for permitting and land
lease. Also, since this process option would require much more testing prior
to ultimate disposal, the cost of storing these formulations and the develop-
ment costs should also be added to the above costs. Further, these estimates
have been developed assuring an efficient operation (no outage), and two
trains of equipment operating per shift. Two equipment trains were assumed
because of the urgency to dispose of these formulations as quickly as pos-
sible. The above costs change very slightly if single train operation is
assumed.
It is evident from the earlier discussion that a number of factors
affecting the process have not been established. These factors (overall
reaction time, reagent requirements, effluent disposal, etc.) can have a
substantial impact on the costs estimates.
4-37
-------
Permitting--
Permits to handle and destroy the pesticide formulations would be re-
quired under each of the options.
Probability of Success--
The preceding discussion points to a need for more test work to esta-
blish the feasibility of this process. The process seems to perform very
well with pure EDB. However, the presence of other rmstituents seem to
affect the process performance. Very long overall reaction times are needed
to achieve 99.99 percent destruction. Thus, although the process spems to
have a good promise, much more extensive testing would be required prior to a
final judgement.
Time Schedule--
In view of the uncertainties associated with this process, extensive
testing would be required prior to scale-up. This would increase the overall
time required to dispose the pesticide formulations. Assuming that adequate
reaction rates are achievable, the overall disposal time could be about 2.5
years.
/*
ATEG Process
Status and Accessibility--
The ATEG process is still in the conceptual stage. The EPA has demon-
strated its capabilities to achieve the required levels of destruction on a
laboratory scale, and the results of these tests have encouraged further /
testing. The technology would become available, however, only after pilot-
scale testing to demonstrate its feasibility, performance capabilities (DREs),
4-38
-------
process economics, and commercial reliability. The process and operational
data derived from such testing will be important to the completion of a
detailed process engineering design.
Past Experience--
Because the process is still in the conceptual stage, no one has had
previous experience with commercial operation. The current data base is
limited to lab scale tests conducted or sponsored by the EPA/GARD and recent
bench scale testing by International Technology (IT). The results of the •
tests carried out by the EPA have already been reported in Se;:tior 3. The
results of IT's test-work are summarized below:
Destruction of EDB bottoms from the distillation of carbon disul-
fide containing formulations was about 100 percent complete.
However, the reaction seems to be very sensitive to the amount of
TEG and the concentration of caustic solution used in the system.
During large scale reactions it was observed that the reaction
temperature had to be greater than 35 °C for the reaction to occur.
Also, a reaction inception tiire of about 30 minutes was observed.
However, there are no data available on the reaction rates.
The reaction of ATEG with chloropicrin formulations was studied on
a, small scale (10 to 50 mis samples). The reaction was studied
using relatively high amounts of TEG, and with two different cau-
stic solutions. With a 30 percent caustic solution, the destruc-
tion of EDB was complete. However, none of the chloropicrin seems
to have reacted. This is an important observation as it creates a
possibility of treating the chloropicrin formulation with the ATEG.
No further tests were carried out for the chloropicrin formulation.
Treatment of carbon disulfide containing formulations with the ATEG
process resulted in high TEG consumptions to achieve EDB destruc-
tion, while at low TEG concentrations there was almost no EDB
destruction. This is because of the reaction of carbon disulfide
with the TEG. The overall reaction time for the carbon disulfide
formulations was more than 8 hours.
The reaction of vinyl bromide with the KTEG solution was studied to
a limited extent. It has been reported that a gel-like layer was
formed on top of the KTEG solution in which the vinyl bromide
became trapped. This layer was found to be water soluble and
released acetylene upon dissolution.
4-39
-------
Need for Development--
The preceding discussion points up an obvious need for further develop-
ment of this process. Although patented by EPA, the process would require
comprehensive testing on a pilot scale to establish its process feasibility,
to collect valuable process design and operating data, and to ascertain
potential problem areas. Some of the questions that need to be addressed
during the test work are:
1) It is believed that the reaction of chloropicrin with ATEG forms a
number of chlorinated by-products which may be hazardous. However,
IT test results seem to indicate that it may be possible to selec-
tively treat the EDB in chloropicrin formulations by using a 30
percent caustic solution. In the absence of any positive data,
treatment of chloropicrin formulations with the ATEG process is
still an open-ended question. More tests would be required to test
the feasibility of treating these formulations as is, or to study
alternative methods of distilling the formulations (e.g., azeo-
tropic distillations etc.). If the ATEG process cannot treat the
chlorop-icrin formulations, then it would not be a very attractive
option. As regards the miscellaneous formulation (which form the
largest percentage of the total EDB formulations to be destroyed),
more tests would be needed to establish the DREs.
2) The bench scale test data has not established the overall reaction
time to achieve acceptable ORE (99.99). Establishing the reaction
tirre is very important as it has a direct impact on the overall
time required to destroy all the pesticide formulations. The over-
all operating cost (labor) is directly proportional t" this time.
If the overall processing time is larger than that assumed in the
preliminary study estimates (Appendix A and B), then the overall
cost for this process will be much higher than those indicated.
3) How much reaction other constituents (carbon tetrachloride, chloro-
form, methyl chloride) undergo and the analysis of the products
(gas, liquid, or solid) should be determined more thoroughly. This
will have a direct bearing on the design of downstream processing
units. Also, analysis of the reaction effluents is important with
regard to their ultimate disposal, which can significantly affect
the process costs.
4) The feasibility of the proposed feeding of solid NaOH (which is
hygroscopic and absorbs moisture) to the reactor should be re-
viewed. Hydrated NaOH could be difficult to transport. Controlled
feeding of NaOH must be followed to avoid any runaway reactions.
4-40
-------
5) Vinyl bromide and vinyl chloride are both carcinogenic compounds.
The efficient removal of these and any other gaseous products of
the reaction and, hence, the quantification of the scrubber per-
formance are very important. The EPA has proposed a countercurrert
packed-bed (regular packings) tower. Vinyl halides react with KTEG
to form acetylene gas and potassium salts. The following opera-
tional data need to be established from the testing program:
a) The extent of vinyl halide removal.
b) Test results seem to indicate that the dehydrohalogenation of
vinyl bromide using KTEG forms a gummy layer which dissolves
in water with the evolution of acetylene. This raises ques-
tions about the choke of the scrubber. The KTEG solution by
itself is very viscous, and with "gummy mass" being formed, it
can clog the scrubber. This can lead to serirus operating
problems (pressure buildup in the system, incomplete absorp-
tion of vinyl halides etc.).
c) More tests should be undertaken to determine if aqueous KOH
solution with very small amount of TEG (similar to the reac-
tor) could be used in the scrubber to eliminate the vinyl
halides. If possible, this scheme will reduce the reagent
consumptions, eliminate the need to recirculate the scrubber
effluent liquid and eliminate problems of clogging etc. If
not, the scrubber will pose serious operational and safety
concerns. A possible alternative to-handle the gas scrubbing
would'be to use a batch operated agitated contactor or bubble
column. Agitated contactors and bubble columns both give very
high mass and heat transfer characterisitcs and are ideal for
handling systems forming solids or which are viscous. Also,
operating the contactor in a batch mode would help in reducing
the reagent consumption. However, in order to ensure complete
destruction of the vinyl halides the effluent gas may have to
be recirculated. This would involve recompressing the off
gases which could be dangerous with acetylene being formed in
the system.
6) The solids content of the reactor liquid effluents should be ascer-
tained during the test work, along with the type of solid-liquid
separation operation required to remove them. The EPA wants to
recover reaction byproducts with some resale value. The technical
feasibility and the probable flow sheet for such an operation
should be researched.
Other Items of Concern--
I) EPA has suggested the use of a vacuum pump to remove the acetylene
from the system. However, as discussed under the zinc process,
4-41 .
-------
maintaining vacuum could lead to a loss of organics. Like the zinc
process, it is felt that this aspect should be studied during
future tests, although it would be safer to use inert gas or a fan
to remove reactor gases.
2) Process calculations show that using solid NaOH in the system cre-
ates a situation where the percentage solids in the reactor exceeds
30 percent. Higher percentage of solids would make mixing and heat
transfer very difficult, and the slurry may not be pumpable. As'a
result water may have to be added to the system to keep the per-
centage solids lower than 20 percent. Thus, instead of feeding
solid NaOH flakes, it is felt that an alkali solution should be
used in the reactor. This will eliminate the problem of feeding
solid alkali to the reactor while eliminating problems of higher
percentage solids in the reactor and poor heat transfer.
3) Introduction of an inert gas in the reactor (to purge process •
gases) would reduce the partial pressure of the vi"yl halides in
the scrubber section, reducing the driving force for their dissolu-
tion in the liquid phase. Vinyl halides are very stable compounds
and it is difficult to dehydrohalogenate them (Morrison and Boyd,
1973). Thus, detailed pilot scale testing of the scrubber opera-
tion is imperative to assure complete neutralization of vinyl
halidps in order to avoid problems of toxic emissions.
Need for Additional or New Equipment--
Two options (process variations) are available for the chemical destruc-
tion of pesticides via ATEG. Under the first option, EPA would build a
totally new facility; under the second option, some existing equipment from
the GARD facility would be used in combination with other new equipment. The
only equipment reportedly available at the GARD facility are reactors, a fil-
ter press, a flare, and a stack. The rest of the equipment would have to be
new. (The cost data for both options are presented in Appendix B.) The need
to construct a new facility (under both options) would affect the time period
required for completing the overall project.
Toxic Emissions and Secondary Environmental Impact--
Vinyl chloride and vinyl bromide emissions could be generated. Both
compounds are carcinogenic. In addition, both chemicals are highly flammable
4-42
-------
and produce toxic gases on ignition. Because the scrubber tower discharge
gases are burned before their release to the atmosphere, this potential
source of toxic emissions can create a secondary environmental impact. In
addition, toxic emissions due some of the volatile organics escaping the
system are possible. The possibility of toxic emissions due to the reaction
of ATEG with other constituents of the formulations needs to be researched.
Compatibility--
CS- is not compatible with TEG, producing a gummy mixture and restrict-
ing process operation.
Safe*y and Health Hazards —
The products of the reaction (i.e., vinyl chloride, vinyl bromide, and
acetylene) are all highly flammable; acetylene is also very explosive. In
addition, the effluent gases may contain volatile organics from the reaction
system, creating safety and health hazards. Thus, the fire and explosion
risks associated with the ATEG process must, be considered. Process safety
will be a primary concern. Moreover, because the reactions involved are
highly exothermic and rapid, runaway reactions are a possibility. Such an
event could precipitate the danger of these product gases catching fire and
causing an explosion. Therefore, great care must be exercised in its opera-
tion. An operating option that would probably reduce these risks entails
feeding the pesticide to the caustic solution at a controlled rate, with a
temperature control to cut off the feed, ^s indicated previously, all op-
tions would need testing prior to scaleup.
As mentioned earlier, vinyl chloride and vinyl bromide are carcinogenic.
Exposures in high concentrations can cause dizziness, anesthesia, and lung
4-43
-------
Irritation. Irritation of eyes, nose, and throat is also common. Chronic
exposures to vinyl chloride can cause liver damage. Both chemicals are
highly flammable, and when ignited, they emit toxic gases that can create
health hazards.
Most of the. conctituents of the pesticide formulations are harmful, and
operators should wear special clrthing wher, handling these wastes.
Transportation Access to Facility —
This access would have to be provided during tfe«igr and construction.
Storage and Handling of Waste--
Another aspect of plant safety and toxic emissions is the handling and
storage of the hazardous waste. Most of the constituents of the waste under
consideration (i.e., EDB, EDC, chloropicrin, etc.) are extremely hazardous
and warrant special precaution in handling. Before a chemical destruction
program is undertaken, a comprehensive spillage control action plan should be
prepared ard all operators should be trained to implement it. Several com-
mercial facilities handle hazardous waste regularly; thus, designing this
area of the plant should be fairly routine.
Corrosion and Mechanical Reliability—
The corrosivity of vinyl chloride, bromide, NaOH, and alkali salts is
unknown; however, 316 stainless steel should be a suitable material of con-
struction for the key equipment itenr to avoid extensive corrosion. However,
more data would be needed establish suitability of SS316.
The explosive characteristic of the reaction products should t~e consid-
ered during the mechanical design of the Krocf.ss equipment. On the whole,
the construction should have good mechanical reliability.
4-44
-------
Handling and Disposal of Process Residues—
The disposal of process effluents is an important factor in the overall
process cost. Test work would have to establish if the process effluents can
o
be classified as hazardous or nonhazardous waste. Depending on the nature of
effluents, appropriate disposal methods would have to be adopted, which can
have a significant impact on total cost. The feasibility of recovering
byproducts of value would have to be established on the basis of test work
results. In the mean time, for this evaluation the worst case of effluent
disposal as a hazardous waste has been assumed for cost purposes. As in the
case of zinc process, it is assumed that the organic and aqueous effluents
from the reactor are completely immiscible, and hence, are easily separated
in a phase separator. The organic phase is assumed to be disposed by incin-
eration while the aqueous layer is assumed to be disposed suitably.
Cost-
Preliminary study cost estimates have been developed for this process
(Appendix B) based on approximate process calculations (Appendix A) to size
the equipment and estimate reagent consumption. It has been assumed that the
CS- formulations are distilled prior to destruction. The recovered product?
(CC1. and CS-) are sold at market value. The cost for destroying chloro-
picrin formulations has not been included because of the high uncertainty
associated with it. The chloropicrin stock may be assumed to be destroyed by
some other process at the same unitized cost as that obtained for the other
formulations. The cost of recovering reaction byproducts was excluded be-
cause of a lack of sufficient data. Instead, the worst case of effluent
disposal as a hazardous waste has been assumed.
4-45
-------
There are two options for this process:
1) Government owns and operates the facility.
2) Contract to smal1-^cale chemical firm.
Under the first option, the government could either build a totally new
facility cr utilize some process equipment available at the GARD facility.
Under the second option, the capital cost is calculated assuming a new facil-
ity; however, this cost is depreciated using a straight-line depreciation
method and assuming a 10-year life and zero salvage value.
The urn'tized cost are:
Government owns and operates Subcontracted to sirall firir
New facility $0.78/lb $0.34/1b
GARD option S0.71/lb
The above estimates do not include the costs for permitting and land
lease. Also, since this process option would reouire much more testing prior
t.o ultimate disposal, the cost of storing these formulations and development
costs should also be included to the above costs. Further, these estimates
have been developed assuming an efficient operation (no outage) and two
trains of equipment operating per shift. Two equipment trains werp assumed
because of the urgency to disoose of these formulations as quickly as pos-
sible. The above costs change very slightly if single train operation is
assumed.
It is evident from the earlier discussion that a number of factors
affecting the process have not been established. Those factors (overall
reaction time, reagent requirements, effluent disposal, etc.) can have a
substantial impact on the cost estimates.
4-46
-------
Permitting—
Permits to handle and destroy the pesticides would be required under
each option.
Probability of Success--
The ATEG process seems to demonstrate the necessary capability to de-
stroy EDB. Preliminary test results seem to indicate that it has the cap-
ability to selectively eliminate the EDB in the chloropicrin formulations,
eliminating the need for preprocessing and the problem of forming unknown
by-products. However, this aspect would have to be thoroughly researched.
Although the process seems capable, it is complicated by a two stage
reaction involving a number of reactants (NaOH, KOH, TEG) and forms a wide
spectrum of by-products which could make characterization and ultimate dis-
posal difficult. Moreover, the process poses safety concerns as it involves
the handling of acetylene. Bench-scale tests show that the process is sen-
sitive to a number of operating parameters (TEG concentration, NaOH concen-
tration, reaction temperature, etc.) requiring further testing to ascertain
the optimal process conditions. Further tests would also be required to
establish the process chemistry, the reaction kinetics and operating proce-
dures.
Time Schedule--
In view of the extensive testing that would be required prior to ulti-
mate disposal, it is believed that this option could take about two and half
years.
4-47
-------
SECTION 5
CONCLUSIONS AND RECOMMENDATIONS
SUMMARY
In this study, all available alternatives have L»en considered that have
potential for the successful destruction of EDB formulations within the next
2 years. To facilitate a comparison of the alternatives on an equal basis,
selection criteria were developed that covered all me^'or technical issues as
well as the overall cost. This permitted a direct comparison of the techni-
cal competence of the various alternatives. Such a comparison pointed up the
technical merits and shortcomings as well as the areas of uncertainty for
each alternative. Based on all information available to date, i£ appears
that incineration in the presence of sulfur dioxide is the best alternative
for effective, rapid, and economical destruction of all the EDB stocks.
Preliminary design and cost estimates were made for each of the chemical
destruction processes; however, these calculations were made on the basis of
very limited laboratory-scale test data and included several engineering
assumptions (Appendix A). Some of these assumptions may not hold in actual
operation, and the costs could be affected. Also, the cost figures reported
are exclusive of surh important cost items as permitting, land lease, Interim
storage, etc. These costs will have to be incorporated in the reported cost
figures to arrive at the overall ccst of destroying the pesticide formulations.
The costs for the thermal destruction options were obtained from vendors
and represent an average incineration cost. Cost figures for specific
5-1 '
-------
modifications (e.g., starved-air incineration and incineration in the pres-
. ence of sulfur dioxide) could not be obtained from vendors; however, EP/1 har,
received proposals from Rollins Environmental and the John Zink Company for
these processes. Rollins has proposed using the sulfur dioxide technology,
whereas John Zink has suggested starved-air i-ncineration. Thus, the EPA
should be in a position to establish the cost-effectiveness of each option.
CONCLUSIONS . . ..'
From a technical standpoint, both starved-air incineration and destruc-
tion in an existing incineration facility without any modifications appear to
be infeasible because of the bromine emissions that would exit through the
stack.
Incineration in the presence of sulfur-containing waste holds an excel-
lent promise for the elimination of bromine emissions. The test burn results
(Alliance, 1988).show that this option meets the destruction standards for
POHCs (DREs greater than 9S.9999 percent) and emission standards (bromine
below detection limits and bromide about 20 ug/dscf in the stack). Also,
continuous monitoring data for CO^, 0-, CO, NO , and SO- seem to be well
within the established standards. Bromine mass balance indicates that all
the bromine exits the system in the scrubber water. Also, the fact that a
currently operating incineration facility in Europe is successfully using
this technology to destroy halogenated waste lends credibility to this option
(Fabian et al., 1979). Moreover, this option offers the advantage of speedy
disposal of the entire EDB stock (probably less than a year) at a competitive
cost. This is especially important in view of the urgency of the situation.
!
Thus, incineration in the presence of sulfur dioxide appears to be the best
i
choice for destroying the EDB pesticides. |
5-2
-------
Cement kiln incineration appears to be a promising option; however,
extensive testing would be required to establish the performance capabilities
and optimal waste feed rates. The optimal waste feed rate would have to be
o
determined so as to eliminate bromine emissions 1n the flue gases while not
having an adverse effect on product quality. If the allowable feed rate was
low, the overall time to complete the job would be higher. This, in turn,
could increase the overall cost of this option; however, no definitive
estimates can be made until after test burns are performed.
The ATEG process has shown excellent capability to eliminate the EDB
obtained from the distillation of the CS- formulations. Previously the
treatment of chloropicrin formulations with the ATEG process was regarded
infeasible because of fear of forming unknown, and perhaps more hazardous
compounds. However, preliminary tests seem to suggest that it may be pos-
sible to treat these formulations, without reacting the chloropicrin, by
using a 30 percent caustic solution. This approach, however, needs further
testing to prove its validity. The bench-scale tests seem to indicate that
the presence of other constituents in the pesticide formulations interfere
with the EDB destruction. Thus, extensive tests would also be required to
study the feasibility of treating the miscellaneous formulations, without
preprocessing them. Despite the promising results on the laboratory scale,
the ATEG process could create operational problems because of its compexity.
The prccess involves:
A two step reaction.
Number of reactants (NaOH, KOH, TEG).
It is found to be sensitive to a number of operating parameters
like the TEG concentration, caustic concentration, temperature,
etc.
5-3
-------
Handling of potentially hazardous substances like vinyl bromide,
etc.
The process foras a wide range of byproducts, which could make
disposal of the effluents difficult.
It is therefore evident that the process would need very extensive testing to
eliminate uncertainties and operational difficulties and establish the
optimal operating conditions, prior to design and scale-up. This could take
considerable time, causing a delay 1n the overall dispeiVi of the EDB pesti-
cides.
Bench-scale tests with the zinc process show excellent promise with pure
EDB. Disposal of the chloropicrin formulations seems to be a problem because
of unacceptable levels of DREs, formation of unknown products, high zinc
consumption, and high hydrochloric acid consumption. Reaction of zinc with
the CS- formulations show that the carbon disulfide reacts rapidly with the
zinc, resulting in very poor destruction of EDB. In all the tests with the
zinc process, long reaction times were required to achieve substantial EDB
destruction. Even longer reaction times may be required to achieve 99.99
percent destructions. This could potentially make the process infeasible.
Therefore, more tests would be required to determine:
Ways to achieve 99.99 percent destruction with all formulations,
without any preprocessing. CS,, may have to be removed prior to
treatment. '
Ways to reduce the acic1 and zinc consumption, especially with the
chloropicrin formulation.
Feasibility of an azeotropic distillation of chloropicrin formu-
lations using alcohol, as suggested by IT, if the 99.99 percent
destruction of the formulation is not possible.
The overall reaction time. This is an exothermic reaction. Thus,
if the reaction rate is fast, then heat transfer will control the
overall rate and vice versa. This will affect the process design
and cost.
5-4
-------
it is therefore evident that this process would need thorough pilot plant
testing to establish its feasibility and optimum operating conditions prior
to design and scale-up. The process is more complex than previously en-
visaged.
At this point in time, incineration in the presence of sulfur dioxide
seems to be the most viable and rapid way of disposing the pesticide formula-
tions at a cost comparable to or lower than other methods. Successful trial
burns for this method have been completed. As a result, the destruction
process can be initiated immediately. The overall time for disposal should
be less than six months. In view of the urgency for disposing of the pesti-
cides, this process appears to be clearly the best choice.
5-5
-------
REFERENCES
Alliance Technologies Corporation. 1988". RES«(TX) EDB Test Burn Program
Emissions Test Regults, Vol. I and II. Prepared for Rollins Environmental
Services (TX), Inc.
Berry, E. 1975. Ontario Research Foundation.
Bonner, T., et al. 1981. Hazardous Waste Incineration Engineering. Noyes
Data Corporation, Park Ridge, HJ.
Bureau of Explosives. 1981. Emergency Handling of Hazardous Mater-,?ls in
Surface Transportation. P. Student, editor. Association of American
Railroads, Washington, D.C.
Dean, J. A., ed. 1985. Lange's Handbook of Chemistry. McGraw-Hill Book
Company, New York.
Eicher, A., J. Cudahy, and W. Troxler. 1985. Thermodynamic Equilibrium of
Halogen and Hydrogen Halide During the Combustion of Halogenated Organics.
Presented at the 6th Annual Conference on Management of Uncontrolled Hazard-
ous Waste Sites, November 1985, Washington, D.C. Hazardous Materials Control
Research Institute, Maryland, 1985.
Fabian, H., P. Reher, and M. Schoen. 1979. How Bayer Incinerates Wastes.
Hydrocarbon Processing, p. 185. April 1979.
Fieser, L. F. and Fieser, M. 1967. Reagents for Organic Synthesis. John
Wiley and Sons, Inc.
Four Nines, Inc. 1987. Technical Review on Thermal Destruction Options of
Preliminary Engineering Assessment of Pesticide Destruction Technologies.
Harbaugh, M. 1964. U.S. Patent 3,145,079.
^'International Technology, Inc. 1987. Technical Report on Bench-Scale Dis-
tillation, Characterization and Destruction of EDB Chemical Formulations.
Prepared for PEI Associates, Inc., Contract No. 68-03-3381.
N
Ishikawa, H., et al. 1981. Catalyzed Thermal Decomposition of Sulfuric Acid
and Production of Hydrogen Bromide by the Reaction of Sulfur Dioxide and
Water. Advances in Hydrogen Energy, Vol. 2. p. 297.
Kern, D. Q. 1950. Process Heat Transfer. McGraw-Hill Book Company, New
York.
R-l
-------
Mell.an, I. 1986. Corrosion Resistant Materials Handbook (3rd Edition).
Noyes Data Corporation, New Jersey.
Morrison, R. T., and R. N. Boyd. 1976. Organic Chemistry, 3rd Ed. Allyn
and Bacon, Inc., Boston, Massachusetts.
PEI Associates, Inc. 1982. Portland Cement Plant Inspection Guide.
EPA-340/1-82-OC7.
PEI Associates, Inc. 1987. Guidance Manual for Cofiring Hazardous Wastes in
Cement and Lime Kilns. Prepared for U.S. Environmental Protection Agency,
Cincinnati, under Contract 68-02-3995.
Peters, M. and K. Timmerhaus. Plant Design and Economics for Chemical Engi-
neers. McGraw Hill Book Company, New York. 1983.
Research Triangle Institute. 1987. Disposal Options for Ethylene Dihronide-
Cortaining Pesticides. Prepared for U.S. Environmental Protection Agpncy,
Washington, D.C., under Contract 68-01-7350.
Richardson Engineering Services, Inc. 1984. Process Plant Construction
Estimating Standards, Vol. 4.
Rogers, C. J., and A. Kornel. 1987. U.S. Patent 4,675,464.
Sax, I. 1984. Dangerous Properties of Industrial Materials. Van Nostrand-
Reinhold Company, New York.
Sittig, M. 1979. Incineration of Industrial Hazardous Wastes and Sludges.
Noyes Data Corporation, Park Ridge, NO.
Somrnerville, R. 1972. New Method Gives Quick, Accurate Estimate of
Distillation Costs. Chemical Engineering, p. 71, May 1972.
State of California. 1982. The Use of Waste Chemicals as Fuel Supplements
for Cement Kilns in California: Briefing Document. Cement Kiln Task Force.
Stew, F. 1964. U.S. Patent 3,131,028.
llhl, V. 1978. Short course entitled "Technical Economics." given at Chester
Towers, Cincinnati, April 1978.
U.S. Coast Guard. 1984. Chemical Hazard Response Information System (CHRIS).
Washington, D.C.
Velzen, D., H. Langenkamp, and A. Ferrari. 1978. Reaction Between SO,, Br2,
and HjO - Equilibrium Measurements. Progress Report, J.N.R.C. Ispra (Italy),
No. 7. p. 51.
(
Velzen, D., H. Langenkamp, and A. Ferrari. 1979. Reaction Between S02, Br2,
and H20. Progress Report, J.N.R.C., Ispra (Italy), No. 8, p. 97.
Velzen, D., H. Langenkamp, and G. Beni. 1980. UK Patent 2045 218.
R-2
-------
APPENDIX A
PROCESS CALCULATIONS
In order to prepare a comparative cost and performance analysis between
each of the process options available, preliminary process calculations had
to be developed for some of the options. These calculations are presented
here.
The chemical process options (zinc and ATEG processes) are currently in
the conceptual stages. Preliminary calculations have been made to establish
the SI'ZPS of the major process equipment and the reagent requirements. These
calculations are important for preparing the approximate cost estimates. In
the absence of any pilot-scale data, a number of simplifying assumptions have
been made in these calculations. Hence, the resulting cost estimates are
only approximate. More pilot-scale testing will be required before accurate
sizing and cost estimates can be made. Nonetheless, these preliminary calcu-
lations aid in giving an insight into possible problem areas and overall
costs.
For the thermal destruction options, there is a large data base for the
s>
cost and performance capabilities. The preliminary calculations are thus
limited to establishing the equilibrium gas composition for the starved-air
incineration option.
ZINC PROCESS
The basic principles underlying the zinc process have been described in
detail in Section 3, along with a conceptual flow sheet. Preliminary material
A-l
-------
balance and sizing calculations are presented in this section. The equipment
sizing is done for the worst-case feed. Table A-l presents the composition
and amounts of various formulations that need to be destroyed. For this
analysis, the CS^-free formulation at Liberty represents the worst-case feed.
Also, to simplify the system, the following preliminary assumptions have been
made:
1. Only EDB an'd EDC react with zinc. Although laboratory tests show
that other constituents also undergo reaction, the amount of other
constituents is small and, hence, should not alter the size of
equipment significantly. Also, the extent and the chemistry of
these reactions is not known, which makes estimation for these
reactions very difficult. Thus, other constituents of pesticide
formulations are assumed to remain inert in the reactor.
2. The IT test results indicate a fairly high HC1 consumption to
achieve desired levels of destruction. However, EPA believes that
the acid consumption can be reduced to very minimal levels by
properly operating the system (high agitation). Thus, for these
calculations, it has been assumed that the system does not require
a large amount of HC1.
3. The reaction of EDB and EDC is 100 percent complete. Although this
, may not be the case in actual operation, it is valid for prelimi-
nary estimating purposes.
4. The process is a batch operation.
5. The CS- containing formulations are distilled to yield pure EDB
bottoms which is destroyed by the zinc process.
The heat of reaction for EDC and EDB were calculated from the heot of
formation data in Lange, 1985. The heat of reaction was calculated as:
AHR = heat of reaction . lUHf)products - lUHf)reactants
thus,
UHR)EDC = 118,560 Btu/lb-mole
(AHR)EDB = 113,360 Btu/lb-mole
It is evident from the heat of reaction data that the reactions are
highly exothermic. Depending upon the relative rates of reaction and heat
S3
A-2
-------
TABLE A-2. COMPOSITION OF EDB STOCKS
"Liberty" Feed Stock
Low-EDB (<2fr) With CS,
Components
Carbon tetrachloride
Carbon disulfide
Sulphur dioxide
Ethylene dibromide (EDB)
Pentane
Approximately 785,245 1b
Percent
of whole
80.90
16.00
1.50
1.20
0.40
High-EDB With CS,
Components
Carbon tetrachloride
Carbon disulfide
Sulphur dioxide
EDB
Ethylene dichioride (EDC)
Chloroform
Pentane
Percent
of whole
73.32
15.13
1.17
5.49
0.06
4.52
0.31
Approximately 624,031 Ib
Low-EPE Without CS7
Not applicable - All low EDB
formulations have CS,
High-EDB Without CS,
EDC 25.69
Carbon tetrachloride 16.46
EDB 49.44
Sulphur dioxide 0.8C
Naphtha 4.31
Methyl chloride 1.26
Otners 1.98
Approximately 1,505,794 Ib
"Ashburn" Feed Stock
Components
EDB
Chloropicrin
Naphtha
Approximately 865,095 Ib
Percent, of whole
38.71
31.61
29.68
A-3
-------
transfer, the overall reaction rate may be limited by the reaction kinetics
(rate of reaction « rate of heat transfer), or the rate of heat transfer
(heat transfer « reaction rate). Since there are no k'netic -Jata available
on these reactions, for the present calculations it has been assumed that the
overall rate is controlled by heat transfer; i.e., overall reaction time is
limited by the rate of heat removal from the system under the design con-
straints. Some options for heat removal include: ffl
1. Remove excess heat as steam by pouring excess water into the system.
As the heat of vaporization for water is very high, large amounts
of heat can be removed. However, in the present situation, this
option doesn't seem very attractive. Elementary vapor-liquid
equilibrium calculations made for the initial reaction mixture
(mixture boiling point calculations) show that in the presence of
water, which is immiscible with the organic phase, the mixture will
boil between 60° and 70°C. Also, the vapor phase would have a very
high percentage of organics, and the overall heat of vaporization
of the reaction mixture would be less than about 200 Btu/lb (110
cal/g « 540 cal/g of water). Thus, this option is ruled out.
2. Another option is to remove heat by using an external heat exchanger.
Under this option, the reactor contents would be recycled through
an external heat exchanger, where some of the heat would be re-
moved. The cooled reaction mixture would be recirculated back into
the reactor.
o
3. Carry out the reaction at a rate where the heat generated can be
removed through the reactor cooling system.
Since the reactions are highly exothermic, it will not be possible to
achieve high processing rates with Option 3. Thus, a combination of Options
2 and 3 is considered here.
Assuming that 300 gallons of pesticide (density = 13.16 Ibs/gal) are
treated per batch,
Amount of pesticide treated = 3948 Ibs
EDC treated per batch = 1014 Ibs = 10.25 Ib mole
EDB treated per batch = 1952 Ibs = 10.4 Ib mole
A-4
-------
Thus,
zinc required at 20 percent excess is:
Zinc = (10.25 + 10.4) x 65.38 x 1.2
« 1620 Ibs
Cn the basis of the tests carried out by the EPA, water and some HC1
will have to be added to the reactor along with the zinc. Water is added to
dissolve the zinc bromide and chloride salts that are formed by the reaction,
while the acid is added to clean the zinc, surface.
Amount of ZnCl? formed = 10.25 Ib-moles or 1397 Ibs/batcn
Amount of ZnBr^ formed = 10.4 Ib-moles or 2342 Ibs/batch
The reactor is assumed to be at 113°F, and at this reaction temperature the
solubility of ZnBr2 in water is 447 g/100 cm3 and the solubility of ZnCl^ in
water is 432 g/100 cm3:
Assuming that the dissolutions of the salts are independent of each
other, the amount of v/ater required to dissolve all ZnCU is 39 gallons and
the'amount of water required to dissolve ull ZnBr? is 63 gallons.
Since it is assumed that the solubility of one salt does not affect that
of the other, the minimum water requirement to dissolve all salts would be 63
gallons. However, to account for any change in solubilities of the salts, it
is assumed that 30 percent excess water is added to the system. Thus, the
amount of water that should be added to the batch is 82 gallons.
According to EPA, the HC1 required would be 2 drops of 30 percent HC1
per 20 trl of water, or 1 ml/20 ml water. Thus, the HC1 required per ba^ch is
4.09 (4.1) gallons. Therefore, a reactor with an overall volume of 500
gallons should be adequate to treat 300 gallons of pesticide.
A-5
-------
Assuming a reaction time of 3 hours,
Zn feed rate = 1620/(3 x 60) = 9 Ib/min
HC1 feed pump capacity = 1.4 gph
Pesticide feed pump capacity = 30 gal/min
Heat Effects in Reactor:
Assuming that the reaction proceeds at a uniform rate.during the entire
reaction time (actually, there could be sudden surges) gives:
(AH)R = 10.25 x 118,580 x 1/3 + IP.4 x 113,360 x 1/3
= 798,130 Btu/h
As mentioned earlier, it is assumed that the heat of reaction is removed
through the reactor jacket and an external heat exchanger. A part of the
reactor contents (approximately 100 gal/min) is recycled through an external
heat exchanger. Thus, the extent of heat removal from the system is calcu-
lated as follows:
Heat removed through the Jacket:
The approximate heat transfer area for a 500-gallon reactor is about 80
o
ft (Richardson, 1984). Assuming the overall heat transfer coefficient, U,
p
to be 75 Btu/h ft °F (Kern, 1950) and the temperature driving force (At) to
be 40°F, we get
Heat removed through jacket = q. = 75 x 80 x 40
= 240,000 Btu/h
Calculation of external heat exchanger area:
Amount of heat that needs to be removed through the external heat ex-
changer:
= 798,130 - 240,000 Btu/h
= 558,130 Btu/h
A-6
-------
Assuming a U of 75 Btu/ft*" °F h, and a At of 40°F, we get
Area of the external exchanger = 558,130
75 x 40
= 186 ft2
2
Assume external heat exchanger area of 200 ft
Thus, it should be possible to operate the reactor without having sig-
nificant temperature rise in the reactor. Thus, it should be possible to
operate the reactor at or below 113°F.
Mixture Boiler Point Calculations
The mixture in the reactor consists of an organic phase and an aqueous
phase. As the two phases are immiscible, each will act independently of the
other. At the start of the reaction, the mixture would be:
Organic phase:
Weight, % Mole Fraction-1
0.37
0.152
0.374
0.0852
0.019
EDO
cci4
EDB
Naphtha
so2
MeCl
Others
25.69
16.46
49.44
4.31
0.86
1.26
1.98
MeCl is very volatile (B.P. = -25°C), and hence, may be lost in the vapor
phase at room temperature. Also, since the amount of MeCl and the "other"
components is low, they are neglected from mole fraction calculations.
2
Naphtha is assumed to be 100 percent pentane.
A-7
-------
Water phase: Ten percent HC1.
However, as the amount of HC1 is very smll and vapor pressure of HC1 is
very negligible (Perry, 1963), the contribution to the vapor phase will be
only due to water. Hence, this phase is assumed to behave as pure water. In
o
addition, the organic phase is assumed to be an ideal solution. Although, in
reality the organic phase will not form an ideal solution, the assumption
makes computations simple and is good enough to give an idea of what one can
expect in actual operation. Also, these calculations are limited only to the
initial reaction mixture (time = 0).
The mixture will boil when the equilibrium pressure exerted by the
mixture equals the total system pressure. The vapor pressure data for each
component is presented in Table A-2:
TABLE A-2. VAPOR PRESSURE IN mm Hg
Temperature, °F
Component
Water1
HC12
cci4
EDC3
EDB
V
Naptha3
100
49.05
0.000373
203.55
140.52
24.87
4424.4
832.87
120
87.48
0.0014
320
222.73
41.31
6124.78
1190
140
149.26
0.0038
488.16
340
66.44
8274.96
1655
160
245
0.01
723.81
502.71
103.5
10940.12
2247
180
388.27
0.0247
1049.52
722.4
156.86
14185.43
—
210
730.17
0.132
1754.2
1187.33
279.39
20280.83
—
From Steam tables.
From Perry.
From Lange
A-8
-------
Thus, equilibrium pressure exerted by the mixture at 100°F is:
Equilibrium pressure = (49.05) •*• (203.55 x 0.153 + 140.52 x 0.37
water organic phase
+ 0.374 x 24.87 + 4424.4 x 0.019 + 0.085 x 832.87)
= 296.14 mm Hg.
Similarly, the equilibrium pressure of the mixture is calculated at other
temperatures and is given in Table A-3.' It can be seen from this table that
the normal boiling point of the mixture is between 140 and 160°F (i.e.,
60-70°C).
If vacuum pump is used to remove the ethylene formed in the reactor,
then the reactor will be operating at sub-atmospheric pressures and the
boiling point would be much lower than 70°C. If the reactor contents boil
off, there will be a significant loss of organics in the vapor phase leading
to the problem of toxic emissions. This is illustrated in the subsequent
calculations.
Assume reactor pressure is 672 mm Hg, i.e., mixture boils at 140CF. The
equilibrium composition (mole fraction) of the vapor would be:
Water 0.222 i 4 Ibs
CC14 0.11 E 16.94 Ibs
EDC 0.187 i 18.51 Ibs
EDB 0.037 = 6.95 Ibs
S02 0.234 = 15.0 Ibs
Naptha 0.210 i 15.12 Ibs
Thus, organic loss would be:
16.94 + 18.51 + 6.QS + 15.12 .. _ft Ibs of organics
t ~ Ib of steam
This is not desinble. Hence, it is felt that the reactor should be
operated at atmospheric pressure with the gases being removed by flowing a
A-9
-------
TABLE A-3. EQUILIBRIUM PRESSURE FOR INITIAL REACTION MIXTURE
Temperature, Pressure
°F (mm Hg)
100 296.14
120 451.5
140 672
160 978.6"
A-10
-------
carrier gas like nitrogen or a fan. As a result, the vacuum pump has been
eliminated from the process flow sheet. Also, the zinc is assumed to be fed
using a carrier gas, eliminating the need for a feeder. This also shows that
the idea of removing reaction heat by boiling off water is not very attrac-
tive. Moreover, if one calculates the heat of vaporization for the mixture
at jts boiling point, it would be less than 200 Btu/lb of vapor. Thus, the
extent of heat removal by boiling the mixture is also not very efficient.
Gas Processing:
Ethylene will be produced at a rate of about 0.092 Ib-mole/min (approxi-
mately 37 ft /min). It is o good idea to pass the reactor gases through a
condenser before being sent to a flare. It is assumed that a condenser of
2
about 100 ft area would be adequate for this duty.
Other equipment:
Filter feed pump 30 gal/min
Filter 5.0 ft2
Filtrate pump 30 gal/min
Effluent storage tank 5000 gallons
Reactor outlet composition:
Unreacted organics: 982 Ibs
Water 685 Ibs
Unreacted Zn 270 Ibs
ZnBr,, 2340 Ibs
ZnCl2 1400 Ibs
5677
Effluents:
Organic effluents/batch = 982 Ibs
A-ll
-------
Aqueous effluents/batch = 4425 Ibs
Solids/batch = 270 Ibs
Chloropicrin Formulations
Although IT test work indicates high acid consumption to solve the
problem of zinc coating, it has been assumed that this problem can be solved
without consuming excessive acid (e.g., better agitation, etc.).
Assuming 300 gallons of this formulation are treated per batch,
Amount of pesticide treated = 3948 Ibs/batch
' Assuming only EDB reacts with the zinc, the amount of zinc reouired is:
Amount of ZnBr_ formed = 8.13 x 225 = 1830 Ibs/batch
Amount of water required to dissolve the ZnBr. would be:
Water required = 64 gallons/batch
at 30 percent excess
30 percent HC1 required = 3.35 gallons/batch
Reactor effluents:
Untreated organics = 2420 Ibs
Water = 534 Ibs
ZnBr2 = 1830 Ibs/batch
Unreacted zinc = lOfi Ibs/batch
REAGENT REQUIREMENTS
For CSp-free EDB formulation:
Zn = 1620/3948 x 1,505,794 = 617,879 Ibs
Water = 82 x 1,505,794/3,948 = 31,275 gallons
30°. HC1 = 4.1 x 9.5 x 1,505,794/3,948 = 14,850 Ibs
A-12
-------
Chloropicrin formulation:
Zn = 640/3948 x 865,095 = 140,240 Ibs
Water = 64 x 865,095/3,948 = 14,025 gallons
30? HC1 required = 3.35 x 95 x 865,095/3,948 x 6980 Ibs
Pure EDB from distillation of CSp formulation (approximately 43,700 Ibs)
Zn = 43,700/187.87 x65.38 x 1.2 = 18,250
Water required = 1825 gallons
30% HC1 required = 870 Ibs
Total zinc = 776,370 Ibs
Water = 47,125 Ibs
30% HC1 = 22,710 Ibs
ATEG PROCESS
The ATEG process was described in detail in Section 3. Preliminary
calculations to estimate reagent requirements and equipment sizes are pre-
sented in this section.. The following primary assumptions were made:
1. Batch operation.
2. A total of 300 gallons of pesticide is treated per batch.
3. Chloropicrin formulations cannot be treated with this process.
o
4. Formulations containing CS2 are distilled to recover approximately
100 percent pure EDB. Because the amount of pure EDB is small and
treating pure EDB by ATEG may be dangerous (RTI, 1987), it is
assumed that the pure EDB is mixed with a CS2-free formulation at
Libertv. The composition of the resulting formulation is shown in
Table A-4.
5. The density of'the pesticide formulation is 13.16 lb/gallon.
6. Both EDP and EDC undergo complete dehalogenation in the ATEG
process.
7. Other constituents in the formulation do not undergo any reaction.
Although in actual operation this is not going to be the case, it
is a velid engineering assumption. The amount of other consti-
tuents is small and hence neglecting their contributions to raw
material consumption and equipment sizing will not alter the
overall cost estimates substantially.
A-13
-------
TABLE A-4. COMPOSITION OF PESTICIDE FORMULATION ASSUMED FOR ATEG
Component
EDC
cci4
EDB
S02
Naphtha
MeCl
. Other
Weight, %
25.0
16.1
51.0
0.2
4.3
1.2 •
2.2
The reaction proceeds in two steps:
STEP 1 (Reaction 1)
TEG
+ AOH - C2H3X + AX + H20
catalyst
STEP 2 (Reaction 2)
•TEG
C2H3X + AOH •» C2H2 + AX + H20
catalyst
where
X = halogen, Cl or Br
A = Alkali, Na or K
Only the first reaction takes place in the reactor in the presence of sodium
hydroxide. The. vinyl halides (C2H3X) are eliminated in the scrubber by
reacting with potassiuir hydroxide (KTEG).
Additional assumptions are statod whenever they are made in the course
of the calculation.
A-14
-------
Reactor
A total of 300 gallons of pesticide is treated per batch. At a density
of 13.16 pounds per-gallon, this equals 3948 Ib/batch. Therefore, amounts of
EDB and EDC treated per batch are as follows:
EDB = 0.509 x 3948 = 2009.5 Ib or 10.7 Ib-moles/batch
EDC = 0.25 x 3948 = 987.0 Ib or 9.97 Ib-moles/batch
EPA has proposed the use of solid NaOH flakes in the reactor. In
earlier tests conducted by EPA and later by IT, twice the theoretical amount
of NaOH was used in the reactor to convert FDB to vin>1 bromide. However,
discussions with IT suggested that using theoretical amount of NaOH (with
some excess) should be able to carry out the first reaction (STEP 1). Naf*-'
in 20 percent excess of the theoretical requirement would be fed to the
reactor. Thus,
NaOH fed = (9.97 + 10.7) x 40 x 1.2
= 992 Ib
NaOH feed rate = 992 = 16.53 Ib/min
~60
The pesticide feeding operation is assumed to be completed in 10 min-
utes. Thus, the pesticide feed rate is 30 gpm. The TEG is fed at 1 percent
of total pesticide or 0.01 x 300 is 3 gallons. Assuming 1 hour as the
feedino time, the TEG feed rate is 0.05 gpm or 3 gph.
The reaction products and their quantities are:
Reactor Liquid Effluents
Liquid effluents are as follows:
A-15
-------
Ib/batch
Unreacted organics 929.8
Water 372.1
\ NaCl 583.2
NaBr 1101.0
r
[ Unreacted NaOH 165.3
t,
( The amount of water formed due to reaction is as follows:
20.67 x 18 = 372 Ib
Solubility of NaCl in water = 1 lb/2.8 Ib water
Solubility of NaBr in water = 1 lb/1.1 Ib water
Solubility of NaOH in water = 1.3 lb/1 Ib water
Assuming that the salts dissolve in the water up to their saturation units
NaCl dissolved in water = 1 x 372 = 133 Ib
NaBr dissolved in water = 1 x 372 = 338 Ib
TTT
NaOH dissolved
-------
It Is assumed that the solids are separated In a filter having 5 ft* of
filtration area.
Reactor Effluent Processing Units:
Filter feed pump: 30 gpm
Filter: 5 ft*
Filtrate pump: 30 gpm
Storage pump: 4000 gallon capacity
Total effluent from the reactor « 3151.4 Ibs.
Liquid effluent » 1938.2 Ibs
Solid effluent = 1213.2
Heat Effects in Reactor
The heat of reaction can be calculated from the heat of formation data.
For the reactions occuring in the reactor, the heats of reaction are:
* 25,610 Btu/lb-mole
UHD) - 31,970 Btu/lb-mole
K EDC
Thus, the overall heat liberated in Btu/hr •'-•
(AH_) <* 10.7 x 25,610 f 9.97 x 31,970
R ~T75 "
* 395,170 Btu/hr
The rate of heat removal from the reactor jacket can be calculated as:
Q - UA Atln
Assuming U = 100 Btu/hr ft*cF
Atln - 40°F
A (for 500 gallon reactor) * 80 ft2
(N.B. Water will have to be added to the reactor, minimum 500 Ibs, to main
tain percentage of solids below 20 percent.)
A-17
-------
Q = 100 x 80 x 40
= 320,000 Btu/hr
There, heat gained by the reaction system 'vould be:
q = 395,170 - 320,,000
= 75,170 Btu/hr
Total heat gained over the reaction period would be:
q1 = 75,170 x 1.5
= 122,755 Btu
'The heat gained will raise the reactor temperature. Assuming average
run in reactor is about 3948 Ib, we have:
q1 = m x C x At
At = 112,755/(3948 x 0.8)
" 35.7°F
At = 20CC (acceptable)
GAS PROCESSING
The gases from the reactor are treated in a scrubber by KTEG solution,
which is a mixture of KOH and TEG in 1:1 molar ratio to remove vinyl halides
from the gaseous effluents. The KTEG is diluted with water to give about 40
percent KTEG solution. The molecular weights of KOH and TEG are 56.09 and
194.23, respectively.
Thus, the mass of KJEG solution containing 1 Ib-mole of KOH is:
. <56-09 ;j94"23) = 625.8 Ib
The density of the KTEG solution is reported to be 1.26 g/cm3. Thus,
the volume of KTEG solution that contains 1 Ib-mole of KOH is:
A-18
-------
coc p
7'481 = 59'55
' 1.26 x 62.43
The KOH reacts with the vinyl halides from the reactor to produce acet-
ylene and potassium salt (see Reaction 2). For every mole of vinyl halide, 1
mole of KOH is required. The reactor and the scrubber operate simultaneously.
The scrubber operation time is the same as the reactor operation time, i.e.,
1.5 h. Assuming, uniform gas loading, the vinyl halide feed rate to the
scrubber is:
(9.97 + 10.7) .• fe I ,* = 0.23 Ib-mole/min
1.0 X OU
Therefore, the KOH feedrate to the scrubber = 0.23 Ib-mole/min
At 20 percent excess, the KOH required = 0.28 Ib-mole/min
Thus, volumetric feed rate of KTEG to scrubber is 16.7 gal/min.
Packed Tower Design
PL = density of liquid = 78.7 lb/ft3
p = density of gas = 0.204 lb/ft3
L1 = liquid flow in Ib/min = 175.42 Ib/min
G1 = gas flow at 20% excess in Ib/min = 23.64 Ib/min
The EPA has proposed to use a packed-column scrubber with countercurrent
gas-liquid flow. Flexipac packings, Type 2, will be used in this packed
column. Referring to the flooding chart for these packings (Figure A-l), the
x-coordinate is:
p« *
L ( —) = 0.38
Thus, from the chart
at flooding = 0.06
pv PL
A-19
-------
PARAMETER INCHES H2O
PER FOOT OF PACKING DEPTH
o
O)
0.1
.06
.03
.01
.006
.003
.002
.001
I 11 III I I I I I I 11II I II I ] I I II
• UPPER LOADING REGION
AREA OF HIGH
ENTRAPMENT
.01 .02
.05 0.1
0.2
0.5 1.0 2.0
5.0 10.0
G = GAS FLOW, Ib/hr-tt2
L = LIQUID FLOW
*V = GAS DENSITY
Pi = LIQUID DENSITY, Ib/ft3
Pf = PACKING FACTOR
QC = 4.18x108lbmf1/lbfhr2
Figure A-1. Generalized pressure drop correlation for flexipac
packings. (Koch Engineering Company, Inc.).
A-20
-------
where G = Ib/h per ft2 of gas
PP = packing factor = 22
Thus,
G = 71.13 lb/min ft2 at flooding
The column is designed at 60 percent of flooding flow rate. Thus,
Design G = 0.6 x 71.13
=42.7 lb/min ft2
Thus, the cross-sectional area of the column = G'/G = 23.64/42.7 = 0.554 ft2,
and the diameter of the column would be 0.84 ft (1 ft was assumed for column
diameter). EPA has calculated a scrubber height of about 20 ft. for this
application.
Acetylene produced in the scrubber is removed by a vacuum pump and
flared to produce CO- and HgO. The volume that the vacuum pump needs to
handle is calculated as 130 cfm, assuming ideal gas laws.
Reaction Products
Amount of KC1 formed = 0.11 Ib-mole = 8.2 lb/min
mm
Amount of KBr formed =0.12 Ib-mole =14.3 lb/min
mm
Amount of water formed = 0.23 Ib-mole = 4.14 Ib/min
inTn
Each salt is assumed to dissolve independently.
Amount of water required by KC1 for complete disolution
= 18.5 lb/min
Amount of water required by KBr for complete disolution
= 17.06 Ib/min
A-21
-------
Amount of water available with KTEG = 105 Ib/min.
Thus, sufficient water is available for the salts to dissolve complete-
ly.
The capacity of the KTEG holding tank is as follows:
Capacity = lfi.7 gpm x 1.5 x 60 x 1.3 = 1954 gal (assume
2000-gallon capacity tank)
Heat Effects in Scrubber
The heat of reactions carried out in the scrubber are:
Vinyl chloride reaction (iHR) = 13,860 Btu/lb-mole
Vinyl bromide reaction UHR) = 12,270 Btu/lb-mole
Thus, heat liberated per hour would be:
UHJ = 10.7 x 12,270 + 9.97 x 13,860
R T5 TTT
= 179,650 Btu/hr
This heat will raise the temperature of the liquid phase (where the
reaction occurs) and part "• will be lost to the gas phase. Assuming that
the entire heat of reacti' ined only in the liquid pi.ase (worst case
scenario), then the rise in t "f the liquid phase in the scrubber
1s:
(AHR) = Mass x sp. ht. x At
179,650 « (175.42 x 60) x 0.8 x At
Thus, At = 21.34CF
= 12°C (not much, acceptable)
However, the scrubber effluent is mixed with the feed tank solution
(2000 gallons) and thus, the overall temperature rise of the feed mixture
A-22
-------
will be negligible. However, to tackle any unusual temperature effects
during operations, a heat exchanger of 50 sq. ft. per scrubber is recom-
mended.
Reagent Requirements for ATE6 Process
From the earlier calculations, the reagent requirements in the reactor
are:
NaOH = 99? Ibs/batch
TEG = 3 gals/batch = 3 x 62.43 x 1.25 ='31.25 Ibs/batch
OaT
Scrubber Section
The KTEG solution is used in the scrubber to convert the vinyl halides
to acetylene. Since this solution contains substantial amount of the TEG, it
will not be economical to discard the spent solution at scrubber outlet. EPA
has suggested that the spent solution be collected in the KTEG feed tank, and
thus, reciruclate the TEG. KOH will have to be added to the feed tank to
replenish the KOH that has reacted, and thus, maintain the concentration
driving force for complete removal of vinyl halides in the reactor. With
this mode of operation, the KTEG solution will soon become saturated with the
potassium salts, and as a result, the scrubber effluent solution will contain
suspended solids. To minimize the amount of solids in the scrubber, the
scrubber feed solution will be filtered before being fed to the scrubber. It
has been assumed that 2000 gallons of KTEG solution would be prepared, which
can be i-sed for about 100 batches (EPA's estimate) without the need for
regeneration.
KOH requirement = (KOH reqd. for reaction)
+ (KOH for preparing feed solution)
A-23
-------
TEG requirement = TEG required to prepare feed mixture.
The total amount that needs to be treated = 1,549,477 Ibs.
Amount treated per batch = 3948 Ibs.
No. of batches = 1,549,477/3948
= 393 batches
Thus for the Reactor:
NaOH required = 992 x 393 = 389,856
TEG = 31.25 x 393 = 12,281.25
For Scrubber:
KOH = 1420 x 393 + 1860 x 393
100
(for reaction) (to prepare feed solution)
= 565,370 Ibs
TEG = 6600 x 393 = 25,940 Ibs
TOO
STARVED-AIR INCINERATION CALCULATIONS
In a conventional combustion process (excess-air incineration), the
bromine in the waste is oxidized to Br^. The Br^/HBr thermodynamics favors
Br2 formation. The Br^ can be reduced to HBr in presence of water vapor, and
the reaction mechanism is represented as follows:
Br2 -f H20 * 2 HBr + i 02
The equilibrium constant for this reaction is given by the following
equation:
A-24
-------
P 2 P
Kp = HBr °z (Eq. 1}
P P
*6r2 ^H20
where the partial pressures of HBr, 02, Br^ and H-0 are under equilibrium
conditions.
It is apparent from the preceding that if the partial p-essure of oxygen
is reduced close to zero, the, partial pressure of HBr increases substantially
(tending towards unfinity), which implies almost exclusive HBr formation.
This is the principle used in the starved-air incineration concept. In
starvfcd-air incineration, however, the oxygen is not completely eliminated;
instead, it is supplied at substoichiometric levels.
A rearrangement of Equation 1 gives the following:
PHBr* PH'°
-22£- = Kp (Eq. 2)
P PI
Pbr2 V
Assuming P., r/P i = U.8 (estimate provided by John Zink Co.) and an
n^ U Oj
incineration temperature of 1800°F (Kp for HBr/Br2 system at 1800°F is 5.9 x
10"3), then:
P., = 5.9 x 10"3 x 0.8
br2
= 0.00472
Assuming P.. = 0.0008 (estimate provided by John Zink Co.)
or,
PHBr - 0.002
PEr
Br2 = 0.4
HBr
A-2b
-------
Therefore, for every 100 moles of HBr formed, 40 moles of Br., will be
formed. Thus, the bromine formation may not be substantially reduced.
As indicated earlier, calculations at the scrubber temperature are more
relevant. The typical flue gas temperature at the scrubber inlet is about
200°F; however, Kp data is available at a minimum temperature of 1000°F. .The
Kp at 1000CF is 3.2 x 10"6.
Assuming P. Q/PO * = 0.8, we get
2 2
P z
-^- • 3.2 x 10'£ x 0.8
= 2.53 x 10"6
Assuming P, = 0.0008
PHBr = 45 x ID'*
V-17.8
PHBr
Thus, at 1000°F, almost exclusive Br2 formation will occur (about 95°.). At
200°F, even traces of HBr will be converted to Br2 (i.e., the gas will con-
sist of all bromine).
A-2C
-------
APPENDIX B
COST-ESTIMATING PROCEDURES AND RESULTS
INTRODUCTION
The costs reported in Section 4 for the chemical destruction processes
are based on capital and annual cost estimates prepared by PEL These esti-
mates were developed in order to provide comparable cost numbers to the
quotes received from the incineration facility operators. The costs are
based on the preliminary process designs developed by PEI in conjunction with
the EPA process developers. These designs are based on the process calcula-
tions and performance assumptions reported in Appendix A. In order to
achieve hiqher processing rates, it »is assumed that two trains of equipment.
will be in operation during a batch.
COST ESTIMATES
A summary of the cost estimates prepared for the ATEG and Zn process
design variations is presented in Table B-l. Detailed cost backup informa-
tion for each ATEG and Zn process option is provided in Tables B-2 to B-4.
Each cost backup table presents a detailed capital and annual cost breakdown.
. o
COST ESTIMATION PROCEDURE
The costs associated with building and operating a plant fall into two
categories: capital investment and operation and maintenance (O&K) costs.
Capital investment Includes the cost of procuring and installing the neces-
sary equipment, complete with piping, instrumentation, and services plus the
capital required for the initial startup of the fzcility. The capital needed
B-l
-------
TABLE B-l. COST FSTIMATE SUMMARY
ZN PROCESS - r'FW FACILITY0
Total Capital Investment
Annual O&M Charge
UNITIZED COST Subcontracted to small
scale chemical manf.
S/gallon 3.91°
S/lb 0.3C
ZN PROCESS - OPTION OF USING GARD EQUIPMENT3
Total Capital Investment
Annual O&M Charge
UNITIZED COST
$/gallon
S/lb
ATEG PROCESS - NEW FACILITY
Total Capital Investment
Annual OKI"! Charge
UHITIZED COST
Subcontracted to small
scale chemical manf.
Not applicable
S/gallon
S/lb
Subcontracted to small
scale chemical manf.
4.5o!
0.34T
ATEG - OPTION WITH CARD EQUIPMENT
Total Capital Investment
Annual O&M Charge
UNITIZED COST
S/gallon
S/lb
Not applicable
Not applicable
Jl, 015,000.
$1,196,100°
Government owns and
operates the facility
6.43d
0.49d
$ 532,800.
$1,130,000°
Government owns and
operates the facility
5.82^
0.44d
$1,572,100
$ C51.2006
Government owns and
operates the facility
10.3?J
0.78d
$ 973,900
$ 815,300e
Government owns and
operates the facility
9.39
-------
TABLE B-2.
ftTEG PROCESS
MEW DESIGN
flpril 87 $
PURCHASED EQUIPMENT COST
EQUIPMENT SPECIFICflTION $/UNIT OTY «
NaOH storage bin SS304/FRP LINED! 18500 Ib cap. 15,400 1 15,400
NaOH feeder SS316 construction 3,000 2 6,000
Pesticide feed pucp Centrifugal, 30 gpi 1,700 2 3,400
TE6 feed puip to reactor Metering puip, 3 gph cap. 1,100 2 2,200
Reactor 500 gal cap, nxer 5hp wtor 41,0W ' 2 82,000
Filter feed puip Centrifugal, 30 gpi 1,7M 2 3,400
Filter SS304, assuie area of 5 sq.ft. 15,440 2 30,600
Filtrate puip Centrifugal, 30 gpi 1,79? 2 3,400
Storage/Phase separator SS304, 4000 gal 40,200 1 44,209
Scrubber I'xSfl'ht., SS316 20,000 2 40,000
Scrubber ht. trans, area 50 sq.ft. per scrubber 1,500 2 3,000
Scrubber eff. vac.puip 130 cfi, 7.5 hp totor 6,600 2 13,600
Flare t Stack 40,000 1 40,000
KTE6 storage tank 2000 gal, SS 304 const. »ith iixer 20,300 1 20,300
Feed puip to scrubber SS316 wetted parts, 17 gpi 1,000 2 2,000
KOH storage bin 14000 Ib cap. 11,400 1 11,400
Makeup TES feed puip 1,000 2 2,000
KOH feeder SS316 construction 3,000 2 6,000
Reactor 0/H condenser 100 sq. ft. of heat transfer area 4700 2 9,400
Scrubber feed filter iiS304/FRP, assume 5 sq.ft. area 16,600 2 33,200
Purchased Equipment Cost 367,700
CONSTRUCTION EXPENSE
Installation 143,400
Instrumentation ( control ' 36,800
Piping installed 91,900
Electrical Installations 36,800
Buildings 147,100
Yard Iiproveaent 36.800
Service facilities 147,10?
Total Construction Expense 639,900
TOTflL DIRECT CflPITflL COST (PEC * CONSTRUCTION EXP) 1,007,6W
INDIRECT COST
Engineering and supervision 110,300
Construction 100,800
Contractors fee 80,600
indirect Cost 291,700
TOTflL DIRECT flND INDIRECT COST
(continued)
B-3
-------
TABLE B-2/ (Continued)
' CONTINGENCY 129,908
I
! FIXED CflPITPL 1,429,288
( ™"™^^^^"^^^"^^^
OT}CR CfiPITRL COSTS
• Working Capital 142,901
Total Other Cost 142,988
TOTAL CflPITRL COST 1,572,188
Note: 1) Pesticide formulations are assuwd to be fed directly fro* tank cars
2) TE6 Mill be fed directly fn» the TEB barrels.
3) It is assuaed that land Mill be leased by EPA, and hence it is not
not considered as a capital expense. Land lease Mill be an operating expense.
4) The cost of building uy be eliminated if it is decided to have the plant
in open xith a saall shed for housing controls.
(continued)
B-4
L
-------
TABLE B-2. (Continued)
ANNA. OPEWTIN6 I NAINTAINPNCE COST
OPERATING COSTS
Labor (I «a/hr> 83,888
Supervisory Labor 25, IN
ComuMbln
tat rnmitL
t/lb Oty. libs) Total Cost
N«QH 1,2* 513888 123,1M
TEG 1.67 38251 25,688
KOH 1.43 642588 276,381
Total RaH Material Cost
UTILITIES
ElRtricity If S c*nt*/kw-hr) 608
Total Operating Cost 534,588
WIKTBINPNCE
Labor 45,708
Material U,688
Total Naintainann Cost 114,388
Optratini Supplies. 9,388
OVEBCADS
Payroll Cost 46,481
Plant Indirect*
InvcstHtit Portion 57,298
Labor Portion 69,688
Administrative Eipense . 38,988
Miscellaneous 15,589
Total Overhead Expense 219,688
ODO OPEWTICHPL EXPENSES
Disposal of reactor effluents 226,988
Disposal of scrubber effluents 41,188
Cost of Distilling the Liberty Saaples 148,988
Credit for Products of Distillation (435,488)
(continued)
B-5
-------
TABLE B-2. (Continued)
Total Other Operational Expense (26,586)
TOTPL OPERATION AND MAINTAINflNCE CHARGE 851,208
Govt. owns and operates Subcontracted to
the facility scale chemical ifr
CARRYING CHARGE 1,429,200 142,908
UNITIZED COST
$/6flUON (Approx 220,980 gallons) $18.32 $4.58
$/LB (Approx 2,915,108 IDS) $8.78 $0.34
Note: 1) Labor cost is computed assuiing 3 operators and 1 lab technician
per shift, 8 hrs/shift, 3 batches per shift, 393 batches. This assuaes a very
efficient operation.
2) Utilities costs have been calculated assuiing steal, Mater and compressed air
requirements are minimal! thus, contribution to cost is negligible.
3) Cost of Ran Materials have been taken froi the Chemical Market in?
Reporter
4) Only carbon disulfide containing femulations are distilled (i.e. 765,245 Ibs
of ION EDB and 624,031 Ibs of high EDB fortulati—is. Total amount
distilled equals 1,489,276 Ibs.)
5) The cost for distillation is assumed to be 10 cents per pound. This
cost figure has been suggested by EPA. Carbon tetrachloride selling price
is taken as $8.36/lb, while that of carbon disulfide is taken as
$420/ton. Total carbon tetrachloride recovered = 1092603 Ibsi
carbon disulfide = 220855 Ibs.
6) The cost/gallon and cost/lb figures shorn above are exclusive of the following
costs: 1) permitting, 2) land lease and 3) disposal of chloropicrin stock
7) Carrying charge for government owned facility is calculated to be the entire
capital cost, while that for subcontracting to a small scale chemical
manufacturer, is calculated assuiing 10 year life span, straight line
depreciation and zero salvage value.
6) The approximate amovnt of pesticides treated excludes the amount of
chloropicrin formulations.
9) NaOH used is commercial grade (76 i).
18) KOH used is commercial grade (88 *).
11) Disposal of reactor effluents - 50c/lb for organic?! We/gallon for aqueous waste
organics: 930 Ibs/batch! aqueous waste: 2258 Ibs/batch at 18 Ibs/gali 393 batches
12) Disposal of scrubber effluents - 50c/lbs for or games! 50c/gallon for aqueous waste and salts
organics: 6600 Ibs every 108 batches! aqueous waste: 14038 every 100 batches at 10 Ibs/gal
2400 Ibs of salt for 360 batches at 18 Ibs/gal density.
B-6
-------
TABLE B-3.
PURCHASED EQUIPMENT COST
EQUIPMENT
NaOH storage bin
KaOH screw conveyor
Pesticide feed puip
TE6 feed puip to reactor
Reactor •
Filter feed puip
Filter
Filtrate puip
Storage/Phase separator
Scrubber
Scrubber ht. trans, area
Scrubber eff. vac.puip
Flare and Stack
KTEG storage tank
Feed puip to scrubber
KOH storage bin
Makeup TEG feed puip
KOH screw feeder
Reactor 0/H condenser
Scrubber feed filter
PJE6 PROCESS
OPTION OF UTILIZING 6ARD EQUIPKNT
April 87 $
SPECIFICATION
SS304/FRP lined! 16500 Ib cap.
SS316 construction
Centrifugal, 30 gpn
Metering puap, 3 gph cap.
500 gal cap, lixer 5 hp ntor
Centrifugal, 39 gpi
SS304, assuK 5 sq.ft.
Centrifugal, 30 gpi
SS304, 4000 gal
I'x20'ht., SS316
50 sq.ft. per scrubber
130 cfi, 7.5 hp ntor
2000 gal, SS 304 const.
SS316 wetted parts, 17 gpi
14000 Ib cap.
100 sq. ft. heat transfer area
SS304/FRP, assuie 5 sq. ft. area
*/UNIT
OTY
15,400
3,000
1,700
1.10B
0
1,700
0
1,700
48,200
20,000
1,500
6,600
0
20,300
1,000
11,400
1,000
3,000
4,700
16,600
1
I
2
Z
2
2
2
2
1
2
2
2
1
1
2
1
2
2
2
2
15,400
6,000
3,490
2,200
0
3,400
0
3,400
40,200
40,000
3,000
13,600
0
20,300
2,000
11,400
2,000
6,000
9,400
33,200
Purchased Equipment Cost
214,900
CONSTRUCTION EXPENSE •
Installation
Instrumentation t control
Piping installed
Electrical Installations
Buildings
Yard Iiprovewnt
Service facilities
Total Construction Expense
TOTAL DIRECT CAPITAL COST (PEC * CONSTRUCTION EIP)
(continued)
143,400
36,800
91,900
36,800
147,100
36,800
147,100
639,9««
654, BW
B-7
-------
TABLE B-3. (Continued)
INDIRECT COST
Engineering and supervision 118,380
Construction 65,580
Contractors fee 68,400
Indirect Cost o 264,280
TOTflL DIRECT AND IMHRECT COST 1,119,888
CONTINGENCY 111,988
FIXED CAPITAL 1,238,980
DMR CAPITAL COSTS
UorVing Capital ' 142,988
Total Other Cost 142,908
TOTAL CAPITAL COST 1,373,688
Note: 1) Pesticide Mill be fed fro* the tank cars.
2) TEG Hill be fed frw TEG barrels.
3) SARD facility reportedly has reactors, belt filter press, flare and stack
which could be used for the present application.
4) Land nil! be leased by EPA. The leasing cost Mill COM under operating expenses.
S) Cost of building MY be eliminated if Ion cost shed or trailer is
is used to house controls etc.
(continued)
B-8
-------
TABLE B-3. (Continued)
ANMJftL OPERATING I MAINTA1NANCE COST
OPERATING COSTS
Labor (9 (20/hr) 83,688
Supervisory Labor • 25,198
Consumables
RAU MATERIAL
*/lb Oty. (Ibs) Total Cost
NaOH 8.24 513088 123,188
TEG 8.67 38258 25,688
KOH 8.43 642588 276,388
Total Ran Material Cost 425,888
UTILITIES
Electricity (9 5 cents/ktHir) 688
688
Total Operating Cost 534,588
MAINTENANCE
Labor 44,388
Hater U! 66,588
Total Maintainance Cost 118,808
Operating Supplies 9,280
OVEBCADS
Payroll Cost 46,888
Plant Indirects
Investment Portion 55,488
Labor Portion 6B,908
administrative Expense 38,688
Miscellaneous 15,308
Total Overhead Expense 216,288
OTHER OPERATIONAL EXPENSES
Disposal of reactor effluents 226,908
Disposal of scrubber effluents 41,108
Cost of Distilling the Liberty Samples 148,908
Credit for Products of Distillation (435,4881
(continued)
B-9
-------
TABLE B-3. (continued)
Total Other Operational Expense (26,568)
TQTRL OPERATION AND HAINTAINANCE CHARGE 844,299
Govt. owns and operates Subcontracted to small
the facility. scale chemical ifr.
CARRYING CHARGE 1,238,989 NA
UN1TIZED CHARGE
f/GALLON (Approx. 229,998 gallons) $9.39
t/LB (Approx 2,915,198 Ibs) $9.71
Note: 1) The labor cost is calculated assuring 3 operators and 1 lab technician,
8 hrs/shift, 3 batches/shift, 393 batches, $29/hr.
2) Utilities costs have been calculated assuring steam, cooling Mater and compressed air
requirements are minimal! thus, contribution to cost is negligible.
3) Cost of ran materials have been taken fro* the Cherical Marketing Reporter.
4) Only carbon disulfide containing formulations are distilled (i.e. 785,245 Ibs of
ION EDB and 624,931 Ibs of high EBB formations. Total amount distilled
equals 1,499,276 Ibs.)
5) The cost of distillation is assumed to be 19 cents per pound. This cost
figure has been suggested by EPA. Carton tetrachloride selling price is taken
as $9.36/lb, while that of carbon disulfide is taken as $429/ton.
Total carbon tetrachloride recovered * 1992893 Ibsi
carbon disulfide - 2298S5 Ibs.
6) The cost/gallon and cost/pound figures shorn above
are exclusive of following costs: I) permitting, 2) disposal of chloropicrin
stock and 3) land lease
expense.
7) Carrying charge for government owned facility is calculated to be the entire
capital cost, while that for subcontracting to a stall scale cherical
Manufacturer, is calculated assuring 19 year life span, straight line
depreciation and zero salvage value.
8) The approximate amount of pesticide treated excludes the amount of
chloropicrin formulations.
9) NaOH used is commercial grade (76 I).
19) KOH used is commercial grade 186 »).
11) Disposal of reactor effluents - 59c/lb for organics! We/gallon aqueous soIn.
939 Ibs/batch of organic*? 2259 Ibs of aqueous waste/batch, 19 Ibs/gal. density
12) Disposal of scrubber effluents - 59c/lb for organics! 59c/gallon acqueous Mtti and salts.
6698 Ibs of organics every 198 batches! 14439 Ibs of aqueous waste every IN batches
at 18 Ibs/gal density! 2498 Ibs of salts for 388 batches at 18 Ibs/gal density
B-10
-------
TABLE B-4.
PURCHASED EQUIPMENT COST
EQUIPMENT
Zn storage bin
Zn screw conveyor
Pesticide feed puip
HC1 feed pup
Uater feed pu»p
Reactor
Filter feed puap
Filter
Filtrate puip
Storage/Phase separator
External heat enchar.ger
Flare ( Stack
Reactor 0/H condenser
ZN PROCESS
I€U DESIGN
APRIL 87 $
SPECIFICATION
SS304; 15080 ib cap.
SS construction
Centrifugal, 38 gpi
Metering puap R/L, 1.5 gph cap.
5 gp* cap, centrifugal
566 gal cap, turbine iipeller, 5hp
Centrifugal, 38 gpi
SS384, assue area of 5 sq.ft.
Centrifugal, 38 gpt
SS364, 4880 gal
SS384 const! 208 sq.ft.
!88 sq. ft. heat transfer area
Purchased Equipwnt Cost
I/UNIT
OTY
6,388
3888
1,700
1,208
1,508
41,808
1,708
15,488
1,700
48,208
8,808
48,888
4,788
1
2
2
2
2
2
2
2
2
1
2
1
2
1
2
2
2
2
2
2
2
2
1
2
1
2
6,888
6,088
3,480
2,488
3,088
82,008
3,488
38,888
3,408
48,208
16,008
48,808
9,408
246,808
CONSTRUCTION COST
Installation
InstruKntation t control
Piping installed
Electrical Installations
Buildings
Yard Iiprovewnt
Service facilities
Total Construction Expense
TOTAL DIRECT CAPITAL COST (PEC » CONSTRUCTION EIP)
INDIRECT COST
Engineering and supervision
Construction
Contractors Fee
Indirect Cost
TOTAL DIRECT AND INDIRECT COST
(continued)
96,388
24,708
61,780
24,708
98,780
24,708
98,780
429,508
676,388
74,088
67,608
54,180
195,700
872,fl«
B-11
-------
TABLE B-4. (Continued)
CONTINGENCY 87,298
FIXED CAPITAL 959,899
CAPITAL COST
Working Capital ' 95,990
Total Other Cost 95,996
TOTPL CflPITPL COST 1,955,199
Note: 1. Pesticides nill be fed fro* tank cars.
Z. HCl will be fed fro* HC1 barrels.
3. Land Mill be leased by EPA. The leasing cost will COK under operating
eipenses.
4. Cost of building uy be eliminated if low cost shed or trailer is used
to house controls etc.
(continued)
B-12
-------
TABLE B-4. (Continued)
(WJAL OPERATING t MfllNJAINANCE COST
OPERATING COSTS
Labor (9 »20/hr) 158,200
Supervisory Labor 45,168
Consumables
RAU MATERIALS
1/LB QTY (US) Cost (»)
Zi;c 8.47 776370 364,980
HC1 8.83 22718 788
S/1880 dflLLON
Uater 8.6 47288 GALLONS 8
Total RaH Material Cost 365,688
UTILITIES
Electricity 2,1
2,108
Total Operating Cost 563,898
HAINTAINANCE
Labor 38,708
Material 46,088
o
Total Maintainance Cost 76,788
Operating Supplies • 13,688
OVERKADS
Payroll Cost • 67,888
Plant Indirect Expense
Investment Portion 3B,488
Labor Portion iei-,796
Administration Eipense 4^,208
Miscellarwous Eipense 22,688
Distillation of CSS formlation 148,900
Credit for products of distillation (435,488)
Total Overhead Expense (18,608)
DISPOSAL OF REACTOR EFFLUENTS ' 569,308
(continued)
B-13
-------
TABLE B-4. (Continued)
TOTAL OPERATION AND MAINTA1NANCE 1,283,800
6ovt. owns and operates Subcontracted to small
the facility. scale chemical mfr.
CARRYING CHARGE 959,200 95,900
UNITIZED COST
«/GALLON IAppro* 330,800 gallons) $6.55 13.94
I/IB S0.S0 10.38
Note: 1. The above cost figures are calculated assuming that HC1 consumption is minimal
2. The I/gal and t/lb figures shown above are exclusive of permitting, transportation
and land lease.
3. Carrying cost for government owned facility is calculated to be the entire
capital cost, while that for subcontracting to a small scale chemical
manufacturer is calculated assuiing 10 years life span, straight line
depreciation and zero ulvage value.
4. Labor cost has been evaluated assuiing 2 operators and 1 lab technician,
8 hrs/shift, 2 batches/shift, 626 batches and *80/tir.
5. The ran uterial costs have been taken from Cheiical Marketing Reporter.
6. Utilities costs have been calculated assuiing that steam, cooling Mter and
ccMpresseo air requirements art minimal! thus, contribution to the cost is negligible.
7. The Mount of pesticide treated is the total amount of all formulations that
need to be treated.
8. Distillation costs are for the CSS containing formulations. The cost is estimated
similar to those estimated in the ATE6 process.
9. Reactor effluent disposal cost: We/It1 for organicsi 58c/gal for aqueous waste
Hiscellaneous formulations: 982 Ibs of organics/batchi 4695 Ibs of aqueous soln/batcti
Dilorooicrin formulation: 2420 Ibs of organics/batch! 2470 aqueous sol/batch
(continued)
B-14
-------
TABLE B-4. (Continued)
PURCHASED EQUIPMENT COST
EQUIPMENT
Zn storage bin
Zn screw conveyor
Pesticide feed puip
HC1 feed puip
Uater few puip
Reactor
Filter feed puip
Filter
Filtrate puip
Storage/Phase separator
External heat exchanger
Flare t Stack
Reactor 0/H condenser
ZN PROCESS
OPTION OF USING GftRD EQUIPMENT
APRIL 87 «
SPECIFICATION
SS384! 15800 lb cap.
SS construction
Centrifugal, 38 gpa
Metering puip R/L, 1.5 gph cap.
5 tjpi cap, centrifugal
500 gal cap, turbine i^eller, 5hp
Centrifugal, 38 gpi
SS304, assiue area of 5 sq. ft.
Centrifugal, 30 gp»
SS384, 4088 gal
SS384 const; ?« sq.ft.
188 sq.ft. heat transfer area
(/UNIT
6,888
3808
1,708
1,200
1,508
0
1,788
8
1,708
40,280
QTY
0
4,780
6,600
6,880
3,400
2,400
3,808
0
3,480
8
3,408
40,280
16,000
0
3,400
Purchased Equipment Cost
94,880
CONSTRUCTION COST
Installation
InstnwenOtion 1 control
Piping installed
Electrical Installations
Buildings
Yard Iiproveaent
Service facilities
Total Construction Expense
TOTAL DIRECT CAPITAL COST iPEC * CONSTRUCTION EXP)
96,380
24,780
61,708
24,700
96,700
24,708
98,788
429,508
523,588
INDIRECT COST
Engineeririg and supervision
Construction
Contractors Fee
Indirect Cost
TOTAL DIRECT AND INDIRECT COST
(continued)
28,288
52,48(1
41,908
122,500
646,880
B-15
-------
TABLE B-a. (Continued)
CONTINGENCY 64,688
FIXED CAPITAL 718,688
OTCR CAPITAL COST
working Capital 71,180
0 Total Other Cost 71,188
10TAL CAPITAL COST 781,7«
Note: 1. Pesticides Mill be fed froa tank cars.
2. HC1 Mill be fed fnx HC1 barrels.
3. Land Hill be leased by EPA. The leasing cost Hill COM under operating
ekpenses.
*. Cost of building may be eliminated if Iw cost shed or trailer is used
to house controls etc.
5. The SARD facility reportedly has reactors, belt filter press, flare and
stack Mhich could be used for the present application.
(continued)
B-16
-------
TABLE B-4.. (Continued)
ANNUAL OPERflTINS I MAJWTAINANCE COST
OPERATING COSTS
Labor (9 $20/hr) 150,200
Supervisory Labor 45,1W
Consumables
RAU MATERIALS
1/LB QTY (US) Cost It)
Zinc 0.47 776370 364,
-------
TABLE B-4. (Continued)
TOTAL OPERATION AND MfllHTfllHANCE l,193,*tt
6ovt. owns and operates Subcontracted to stall
the facility. scale chemical ifr.
CPRRY1N6 WWGE 711,680 NOT flPPLICflttl
UNITIZED COST
t/GPLLON (Approx 330,600 gallons) $5.77
$/LB M.44
Note: 1. The above cost figures are calculated assuming that HC1 consumption is minimal
2. The t/gal and */lb figures shown above are exclusive of permitting, transportation
and land lease.
3. Carrying cost for government owned facility is calculated to be the entire
capital cost, while that for subcontracting to a stall scale chemical
manufacturer is calculated assuming 10 years life span, straight line
depreciation and zero salvage value.
4. Labor cost has been evaluated assuming 2 operators and 1 lab technician,
6 r -s/shift, 2 batches/shift, 626 batches and *20/hr.
5. The ran material costs have been taken from Chemical Marketing Reporter.
6. Utilities costs have been calculated assuming that steam, cooling water and
compressed air requirements are minimal! thus, contribution to the cost is negligible.
7. The amount of pesticide treated is the total amount of all formulations that
need to be treated.
8. Distillation costs are for the CS2 containing formulations. The cost is estimated
similar to those estimated in the flTEB process.
9. Reactor effluent disposal cost: 56c/lb for organic?! 50c/gal for aqueous waste
Miscellaneous formulations: 982 Ibs of organics/batchi 4695 Ibs of aqueous soln/batcti
Chloropicrin formulation: 2420 Ibs of organics/batchi 2470 aqueous sol/batch
B-18
-------
to provide the necessary manufacturing and plant facilities 1s called the
"fixed-capital Investment," whereas the capital required for the operation of
th*> plant is called "working capital." The costs for the day-to-day operation
of the plant are referred to as "operation and maintenance costs."
Cost estiuates can be classified into five main categories:
0 Order-of-Magnitude (Ratio Estimate; ± 50 percent).
e Study (Factored Estimate; ± 30 percent).
B Preliminary (Budget Authorization Estimate; ± 20 percent).
c Definitive (Project Control Estimate; ± 10 percent).
c Detailed (Firm Estimate based on complete engineering drawings,
specifications, and site surveys; ± 5 percent).
As part of this project, a study estimate based on known major Items of
equipment was prepared for the ATEG and Zinc processes. Preliminary sizing
calculations were performed for each of these processes (Appendix A) to
arrive at approximate equipment sizes. The following two subsections contain
an overview of the overall cost estimation procedure used.
FIXED CAPITAL INVESTMENT
The literature contains several methods for arriving at an approximate
estimate of fixed capital and total capital costs. A procedure recommended
by Peters and Timmerhaus (1983) was adopted for these estimates. This pro-
cedure involves estimating the cost of purchased equipment and then develop-
ing all other cost items as a percentage of the purchase equipment costs.
Purchased Equipment Cost
The costs of major purchased equipment Hems shown on the process flow-
sheet were estimated from vendor quotes and cost data available in Peters and
Timmerhaus (1983) and The Richardson Rapid Systems Volume IV (1984). The
costs were updated to 1987 dollars by use of the CEP cost Indices.
D-19
-------
Expenses
The following cost items are included under expenses. These expense
items have been estimated as a percentage of the purchased equipment cost.
For the option of utilizing the GARD equipment, however, these items have
been calculated using the purchase equipment costs under the option of new
design.
Purchased-Equipment Installation--
Equipment installation costs include labor, foundations, supports,
platforms, construction expenses, etc. These costs normally vary from 25 to
55 percent of the purchased-equipment costs. For these .estimates, they are
assumed to be 39 percent of the purchased-equipment cost.
•
Instrumentation and Controls--
Depending on its complexity, the total cost of instrumentation,
including the cost of the instruments, auxiliary equipment and materials, and
installation labor, ranges from 6 to 30 percent of the purchased-equipment
cost. For these estimates, this cost is assumed to be 10 percent of the
purchased-equipment cost.
Piping—
The cost for piping includes the pipe itself, labor, valves, fittings,
supports, and other accessories involved in the erection of all piping used
directly in the process. Depending on the type of plant, piping costs can
vary substantially. For a fluid/solid processing plant (such as those
proposed), piping usually runs about 31 percent of the purchased-equipment
cost. For purposes of these estimates, piping cost is assumed to be 25
percent of the total purchased-equipment cost.
B-20
-------
Electrical Installations--
The cost of electrical installations consists primarily of installation
labor and materials for power and lighting (building and service lighting is
usually included under the heading of building and services costs). In most
chemical plants, these costs amount to about 10 to 15 percent of the purchased
equipment cost.
Buildings and Services--
The cost of buildings and services includes the exoenses for labor,
materials, and supplies involved in- the erection of all buildings connected
with the plant. Costs of plumbing, heating, lighting, ventilation, and
similar building services are included. For a solid/fluid processing plant,
such costs for a grass-roots new plant run about 47 percent of the total
purchased-equipment cost. In these estimates, however, this cost component
is assumed to be 40 percent of the purchased-equipment cost.
Yard Improvements—
This item includes the cost for fencing, grading, roads, sidewalks,
railroad sidings, landscaping, and similar items. For a chemical plant, this
cost varies from 10 to 20 percent of the purchased-equipment cost. For
purposes of these estimates, the cost is assumed to be 10 percent of the
purchased-equipment cost.
Service Facilities-
Utilities for supplying steam, water, power, compressed air, and fuel v
are part of the service facilities of an industrial plant. In chemical
B-21
-------
plants, the cost for such facilities generally vary from 30 to 80 percent,
and 55 percent is the average for a normal solid/fluid processing plant.
Because of the small scale of these operations, however, this cost is assumed
to be 40 percent of the purchased-equipment cost.
For the option of utilizing equipment at the CARD facility, the above
costs were evalauted as a percentage of the purchased equipment cost for new
design.
Indirect Costs
The succeeding subsections cover the costs included under Indirect
Costs.
Engineering and Supervision--
These capital costs include design and engineering, drafting,
purchasing, accounting, construction and cost engineering, travel,
reproduction work, communications, and home office expense, including
overhead. This cost is usually 30 percent of the purchased equipment cost.
Construction Expense--
This item includes temporary construction and operation, construction
tools and rentals, home office personnel located at the construction site,
construction payroll, travel and living expenses, taxes, insurance, and other
construction overheads. This item is estimated to be 10 percent of the totsl
direct costs.
B-22
-------
Contractor's Fee--
This expense varies from 2 to 8 percent of the total direct costs. For
these estimates, the contractor's fee is assumed to be 8 percent of the total
direct costs.
Contingency
Contingency is usually included in capital investment estimates to
compensate for unpredictable events, such as storms, floods, strikes, price
changes, snail design changes, errors in 'estimation, etc. Contingency
factors commonly range from 5 to 15 percent of the total direct and indirect
costs. A contingency factor of 10 percent has been used in these estimates.
OTHER CAPITAL COSTS
Included under this category are land and working capital.
Land
Typically, land cost is estimated at 5 to 8 percent of the purchased
equipment cost. Because land is proposed to be leased land for this project,
land is not included as a capital expense item. Instead, the lease cost
would be considered in the OSM expenses.
Working Capital
Once the plant is installed and ready for startup, some changes are
usually required to make it operational. The capital required to start up
the plant is the part of the capital appropriated because it is essential for
the successful completion of the venture. Typically, this expense runs about
8 to 10 percent of the total fixed capital investment.
OPERATION AND MAINTENANCE COSTS
These costs represent all costs associated with the operation of the
plant. In these estimates, the following costs have been considered.
B-23
-------
Direct Operating Labor
This item represents the l?bor directly responsible for the operation of
the process (primarily operators). Cost is determined by computing the
labor hours required to operate the plant and multiplying this value by the
hourly rate.
Supervisory Labor
This item usually figured at 20 to 30 percent of the direct operating
labor. The upper limits usually represent batcKor complex processing. In
these estimates, supervisory labor is assumed to be 30 percent, of the cost of
operating labor.
Consumables
Consumables refer to the cost of raw materials and utilities used in the
process. The cost of raw materials is based on requirements estimated from
the material balance calculations. The utilities include electricity, water,
steam, and compressed air. Estimates of total utility consumption are based
on the material and energy balance calculations.
Maintenance
For preliminary estimating purposes, the cost of maintenance varies from
6 to 10 percent (average of 8 percent) of the fixed capital investment and
includes the cost of labor and material. The labor portion usually accounts
for about 40 percent of the total maintenance cost.
Operating Supplies
Operating supplies include filter cloths, brooms, mops, Instrument
charts, etc., exclusive of those items listed on the manufacturing cost
sheet. For estimation purposes, this cost is assumed to be 6 percent of the
operating labor.
B-24
-------
Overheads
Overheads represent operating costs not directly associated with the
manufacturing activity, but they are none the less essential for carrying out
the manufacturing activity.
Payroll Cost—
This item includes Workmen's Compensation, pensions, group insurance,
paid vacations" and holidays, Social Security, unemployment taxes, profit.
sharing, and fringe benefits. This cost is usually estimated at 25 to 50
percent of the cost of labor (including direct, supervisory, and
maintenance). In these estimates, payroll cost is calculated at 30 percent
of the total labor cost.
Plant Indirects—
These costs include items such as local taxes, insurance, and local
plant service expenses (i.e., associated with railroad spurs, plant roads,
fire protection, cafeteria, employee safety, parking lots, etc.) These costs
are made up of two components: 1) investment factor and ?) labor factor.
Estimation of both factors has been discussed in Uhl, 1978. For purposes of
these estimates, the investment factor is assumed to be 4 percent of the
fixed capital investment and the labor factor is a percentage of the total
operating labor cost.
Administrative Expense--
Administrative expense includes expenses connected with the
administrative activities of top management. Although not directly involved
in manufacturing, these activities are essential for the smooth operation and
coordination of all other activities, and are therefore included in the cost
B-25
-------
analysis. This category includes the salaries of administrators,
secretaries, accountants, typists, etc., as well as the cost of office
supplies and equipment, outside communications, administrative buildings,
etc. This expense itb,.i is usually estimated to be 20 to 30 percent of the
total operating labor cost.
Kiscellaneous--
This cost item, which includes such things as the cost of providing and
maintaining special clothing required by the operators, 1s estimated as a
percentage of the total operating labor cost.
Other Operational Expenses
Other costs that must be considered specifically for these estimates
include permitting costs, transportation costs, and the cost of disposing of
the reaction products. For the ATEG process, the cost of distilling the
formulations containing carbon disulfide and the cost of disposing of the
chloropicrin formulations also must be included to arrive at reasonable cost
figures. The sum of all these costs would represent the annual operation and
maintenance (O&M) costs. To find the total cost of disposing of the
pesticide formulations, one must add the cost of using the equipment to the
annual O&M cost. The typical life span of chemical equipment is about 11
years (Peters and Timmerhaus 1983). If the equipment 1s used throughout its
life span, the cost due to the fixed capital would represent the depreciation
charge.
Carrying Charge
This item represents the fixed capital depreciation charge for the
option of subcontracting to 8 small scale chemical firm. The EPA has found
one chemical facility that 1s Interested in undertaking the destruction of
B-26
-------
pesticide formulations. If this facility is used, the fixed capital charge
would be that due to depreciation. The same is also true, if the EPA decides
to build a new facility, provided they can later sell the equipment to a
chemical firm or perhaps use the facility for other similar projects. In
this case, the carrying charge has been estimated on the basis of an assumed
10-year life span., straight-line depreciation, and zero salvage value.
For the option of government owned and operated facility, the carrying
charge has been taken to be the total fixed capital cost. Under this option
it is assumed that the facility is built and used only for destroying the FOB
formulations.
The sum of the O&M costs and the carrying charge provides the annual
cost to own and operate the chemical pesticide destruction plant. These
costs are unitized in terms of S/gallon and S/lb as a measure of cost-effec-
tiveness. •
B-27
-------
APPENDIX C
INFORMATION OBTAINED FROM COMMERCIAL
INCINERATION FACILITIES
The highlights of the discussions with the various commercial incinera-
tion facilities are presented here.
VESTA
0 Transportable unit.
0 Past Experience: Successful testing with pesticides and PCB's.
0 Requires demoisturized feed.
0 Doesn't have a permit at the moment. Could be a while before they
get one.
0 Two-stage scrubbing system. (High-energy venturi followed by
packed scrubber).
0 No past experience with brominated waste. Currently scrubbers are
not refractory-lined and could become corroded. Refractory lining
could combat corrosion.
0 Interested in the project, but believe the technology will be
difficult to develop and expensive.
SHIRCO
0 Tr^rsportable unit.
0 Infrared technology.
0 Does not have permit at the moment.
0 Permit applications for PCS testing are pending.
0 Pilot plant available for bench-scale test burns. Permit applica-
tion (RD&D - Administration) pending.
0 This process can handle only solid waste or sludges. Thus, EDB
will have to be mixed with a carrier substance before it could be
destroyed.
0 Bench-scale testing (Thermal Gravimetric X.nalyzer - TGA) will cost
approximately $50,000. This test does not analyze the flue gas
emissions.
0 Process Is equipped with low-energy venturi scrubber.
C-l
-------
ENSCO
Main unit in El Dorado.
No full-scale experience with bnominated waste.
Did burn some brominated waste. Scrubber could not remove bromine
from the flue gases. An orange gas was observed at the stack.
Since then, ENSCO is very apprehensive about brominated waste and
would not burn it unless a substantial amount of waste is involved.
Indicated willingness to do the job in a transportable unit that
processes 50 gal/min only if the operation could last more than a
month.
Reported refractory corrosion problem.
Rollins Environmental
Do handle pesticides on a regular basis.
Facilities in New Jersey, Louisiana, and Texas.
Has past experience with brominated waste. Burned brominated waste
at New Jersey facility. Did have problems of bromine emissions;
however, they claim that this problem was taken care of by changing
furnace operating conditions (proprietary).
Suggested that the equipment has been designed to withstand any
corrosion due to halogens.
Costs for incineration vary from application to application.
Average cost is about 50 to 75 cents per pound.
Permits available. ,
John Zink
They have presented a proposal to EPA for tfn's project.
Earlier test resulted in bromine emissions; however, they believe
the problem could be solved. j
Further information available from EPA. »
Chemical Waste Management
Rotary kiln facility in Chicago, which is largest in the United
States.
Fixed-hearth incinerators at St. Louis.
Chicago facility is equipped with wet scrubber system, whereas St.
Louis facility has one furnace with wet scrubbing and one with dry
scrubbing system (bag filters).
Average costs for liquids is 25 to 45 cents/pound.
Small quantities of brominated waste (0.5 Ib/mln) will have to be
treated. At this r
-------
International Technology Corporation (IT)
Presented a detailed seminar to EPA on their capabilities, the options
they would like to try, and past experience. Suggested that the new
transportable HITS system could handle the pesticide destruction. IT
could not provide cost information that would be meaningful to the
destruction of only the EDB stocks. IT reported that cost information
for EDB destruction via the HTTS would be cost-effective only as part of
a larger scope project that would involve EDB, Silvex, and Dinoseb.
C-3
-------
------- |