PB82-236993
Emerging Technologies for the Control of
Hazardous Wastes
Ebon Research Systems
Washington, DC
Prepared for
Industrial Environmental Research Lab
Cincinnati, OH
Mar 82
U.S. DEPARTMENT OF COMMERCE
National Technical Information Service
NTIS
-------
EPA 600/2-82-011
March 1982
PB82-2369S3
EMERGING TECHNOLOGIES FOR THE CONTROL OF HAZARDOUS- WASTES
by
Barbara H. Edwards, John N. Paullin, Kathleen Coghlan-Jordan
Ebon Research Systems
Washington, D.C. 20011
Contract NO. 68-03-2786
Project Officer
Thomas L. Baugh
Solid and Hazardous Waste Research Division
Municipal Environmental Research Laboratory
Cincinnati, Ohio 45268
MUNICIPAL ENVIRONMENTAL RESEARCH LABORATORY
OFFICE OF RESEARCH AND DEVELOPMENT
U.S. ENVIRONMENTAL PROTECTION AGENCY
CINCINNATI, OHIO 45268
-------
TECHNICAL REPORT DATA
(Please read Instructions on the reverse before completing)
1. REPORT NO.
EPA-600/2-82-011 *-
ORD Report
3. RECIPIENTS-ACCESSION NO.
33699 3
4. TITLE AND SUBTITLE
Emerging Technologies for the Control of Hazardous
Wastes
5. REPORT DATE
March 1982"
6. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
Barbara H. Edwards, John N. Paullin, '
Kathleen Coghlan-Jordan '
8. PERFORMING ORGANIZATION REPORT NO.
9. PERFORMING ORGANIZATION NAME AND AOORESS
Ebon Research Corporation
820 Quincy Street, NW
Washington, D.C. 20011
10. PROGRAM ELEMENT NO.
C73D1C
11. CONTRACT /GRANT NO.
68-03-2787
12. SPONSORING AGENCY NAME AND AOORESS
Industrial Environmental Research Laboratory
Office of Research and Development
U. S. Environmental Protection Agency
26 W. St. Clair Street
Cincinnati, OH 45268
13. TYPE OF REPORT ANO PERIOD COVERED
Final
14. SPONSORING AGENCY CODE
EPA 600/12
IS. SUPPLEMENTARY NOTES
16. ABSTRACT
Investigations were conducted of new and emerging technologies for the disposal
of hazardous wastes. These methods involve new technologies or a recent variation
of an established one. In addition, a questionnaire survey was made of potential
users of hazardous waste information. The need for a data base for emerging
hazardous waste technologies and/or a newsletter was evaluated. Information on the
emerging technologies was acquired by computerized search, library searching, and
personal contacts. The emerging technologies discussed include molten salt com-
bustion, fluidized bed incineration, high energy electron treatment of trace organic
compounds in aqueous solution, the catalyzed wet oxidation of toxic chemicals,
dehalogenation of compounds by treatment with ultraviolet (UV) light and hydroger.,
UV/chlorinolysis of organics in aqueous solution, the catalytic hydrogenation-
dechlorination of polychlorinated biphenyls (PCB's), and ultraviolet/ozone dest
Theory, specific wastes treated, and economics are discussed.
destruction.
17.
KEY WORDS ANO DOCUMENT ANALYSIS
DESCRIPTORS
b.lOENTIFIERS/OPEN ENDED TERMS C. COSATI Field/Group
18. DISTRIBUTION STATEMENT
Release Unlimited
19. SECURITY CLASS (Thit Report/
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This pagt>
Unclassified
22. PRrCE
EPA Form 2220—1 (Rev. 4—77) PREVIOUS COITION is OBSOLETE
-------
NOTICE
THIS DOCUMENT HAS BEEN REPRODUCED
FROM THE BEST COPY FURNISHED US BY
THE SPONSORING AGENCY. ALTHOUGH IT
IS RECOGNIZED THAT CERTAIN PORTIONS
ARE ILLEGIBLE, IT IS BEING RELEASED
IN THE INTEREST OF MAKING AVAILABLE
AS MUCH INFORMATION AS POSSIBLE.
-------
FOREWORD
The UoS. Environmental Protection Agency was created because of
increasing public and government concern about the dangers of pollution to
the health and welfare of the American people. Noxious air, foul water,
and spoiled land are tragic testimonies to the deterioration of our
natural environment. The complexity of that environment and the interplay
of its components require a concentrated and integrated attack on the
problem.
Research and development is that necessary first step in problem
solution; it involves defining the problem, measuring its impact, and
searching for solutions. The Municipal Environmental Research Laboratory
develops new and improved technology and systems to prevent, treat, and
manage wastewater and solid and hazardous waste pollutant discharges from
municipal and community sources, to preserve and treat public drinking
water supplies, and to minimize the adverse economic, social, health, and
aesthetic effects of pollution. This publication is one of the products
of that research and provides a most vital communications link between the
researcher and the user community.
This report is a review and assessment of emerging technologies or
novel variations of established technologies for the control of hazardous
wastes. The results of a survey of a select group of potential users of
hazardous waste information are reported.
Francis T. Mayo, Director
Municipal Environmental Research
Laboratory
Preceding page blank
-------
ABSTRACT
Investigations were conducted of new and emerging technologies for the
disposal., of hazardous wastes. These methods involve new technologies or a
recent variation of an established one. In addition, a questionnaire survey
was made of patential users of hazardous waste information. The need for
a data base for emerging hazardous waste technologies and/or a newsletter
was evaluated. Information on the emerging technologies was acquired by
computerized search, library searching, and personal contacts. The
emerging technologies discussed include molten salt combustion, fluidized
bed incineration, high energy electron treatment of trace organic
compounds in aqueous solution, the catalyzed wet oxidation of toxic
chemicals, dehalogenation of compounds by treatment with ultraviolet (UV)
light and hydrogen, UV/chlorinolysis of organics in aqueous solution, the
catalytic hydrogenation-dechlorination of polychlorinated biphenyls
(PCB's), and ultraviolet/ozone destruction. Theory, specific wastes
treated, and economics are discussed.
Molten salt combustion, fluidized bed incineration, and ultraviolet/
ozone destruction are reviewed and evaluated in detail. Among the wastes
treated by emerging technologies are PCB's, various Dioxins, pesticides
and herbicides, chemical warfare agents, explosive and propellents,
nitrobenzene, and hydrazine plus its derivatives, to name a few.
This report was submitted in fulfillment of Contract No. 68-03-2786
by Ebon Research Systems under the sponsorship of the U.S. Environmental
Protection Agency. This report covers the period April, 1979 to September,
1980, and work was completed as of September, 1980.
IV
-------
CONTENTS
Foreword ,0° iii
Abstract . . , iv
Figures ,, „ « vii
Tables „ 0 . . . ix
Acknowledgment xi
10 Introduction 1
Purpose of the report 1
Source material for the report 1
Characterization of hazardous waste 1
2. Conclusions 2
Molten salt combustion 2
Fluidized bed incineration 3
UV/ozone destruction 4
Hazardous waste generator survey 6
3= Theoretical Basis of the Technologies 8
Molten salt combustion 8
Fluidized bed incineration ............ 11
UV/ozone destruction ................ 14
4, Unit Operations . ............. 17
Molten salt combustion 17
Fluidized bed incinerator 27
UV/ozone destruction 27
5. Wastes Destroyed by the Emerging Technologies .... 35
Hazardous wastes destroyed by the molten salt
process 35
Hazardous wastes destroyed by the fluidized bed
process 48
Hazardous wastes destroyed by UV/ozonation .... 76
60 Emerging Technologies in the Development Stage .... 97
Background 97
Destruction of toxic chemicals by catalyzed
wet oxidation 98
Dehalogenation of compounds by treatment with
ultraviolet light and hydrogen 105
Electron treatment of trace toxic organic
compounds in aqueous solution 108
UV/chlorinolysis of hydrazine in dilute
aqueous solution 111
Catalytic hydrogenation-dechlorination of
polychlorinated biphenyls 113
-------
CONTENTS (Continued)
7. Economics of the Emerging Technologies 115
Molten salt combustion 115
Fluidized bed incineration 116
UV/ozone destruction 119
8. Survey of Hazardous Waste Generators . . 123
Purpose 123
Survey Results 123
References 127
Appendices 132
A. Eutectic mixtures of neutral salts 132
B. Eutectic mixtures of active salts 137
C. Hazardous wastes destroyed by the
emerging technologies 141
D. Design of experiment tests for PCB destruction
in the Ultrox pilot plant .144
E. Predicted operating conditions in pilot plant to
achieve minimum operating and capital costs 146
VI
-------
FIGURES
Number Page
1 Ozonolysis curve . 16
2 Bench-scale molten salt system 19
3 Off-gas analysis schematic 21
4 Pilot plant molten salt combustor 22
5 APS molten salt incinerator 25
6 Schematic of a fluidized bed combustor 28
7 Ozone generators 32
8 Chlorine emissions in flue gas 51
9 Chlorine adsorbed in bed material and cyclone ash . . 52
10 Tilting fluidized bed combustor 54
11 Schematic of a continuous fluidized bed process for removing
insulation from copper wire 55
12 Dow Chemical Rocky Flats laboratory-scale combustor 59
13 Dow Chemical Rocky Flats laboratory-scale combustor:
Intermediate design 60
14 Dow Chemical Rocky Flats laboratory-scale combustor:
Advanced design 61
15 Flow diagram for pilot plant fluidized bed incinerator .... 68
16 Reduction of TCDD by ozonation at pH 3.5 and pH 10.5 ...... 78
17 Reduction of TCDD by ozonation and UV irradiation 79
18 Process diagram of Catalytic, Inc. UV/ozone system 81
vn
-------
FIGURES (Continued)
Number Page
19 The effect of ultraviolet light on the ozone
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
The effect of initial concentration on ozone
The effect of inlet ozone concentration on the ozone
oxidation of MMH
The effect of ultraviolet light on the oxygen
IT Enviroscience process concept for homogeneous catalyst . .
IT Enviroscience process diagram for aqueous waste
IT Enviroscience process diagram for organic residues ....
Atlantic Research Corporation schematic of
Monuron in water at 0.4 mg/1, standard before irradiation . .
Monuron in water at 0.4 mg/1, exposed to 100 kilorads ....
Percent degradation vs. dose for Monuron in water.
0.4 mg/1 and 4.2 mg/1 . . .
Process flow diagram UV chlorinolysis reaction system ....
Ebon Research Systems' hazardous waste information
82
83
85
86
87
91
96
99
103
104
106
109
109
110
110
112
125
Vlll
-------
I 165968 +83-004701
EMERGING TECHNOLOGIES FOR THE CONTROL OF HAZARDOUS WASTES,
EDWARDS. BARBARA H. ; PAULLIN JOHN N.; COGHLAN-JORDAN
KATHLEEN
EBON RESEARCH SYSTEMS, DC.
NFIS REPORT PB82-236993. MAR 82 (158)
SPECIAL REPORT: INVESTIGATIONS WERE CONDUCTED OF NEW AND
EMERGING TECHNOLOGIES FOR THE DISPOSAL OF HAZARDOUS WASTES.
THESE METHODS INVOLVE NEW TECHNOLOGIES OR RECENT VARIATIONS OF
ESTABLISHED ONES. TECHNOLOGIES DISCUSSED INCLUDE MOLTEN SALT
COMBUSTION. FLUIDIZED BED INCINERATION. HIGH ENERGY ELECTRON
TREATMENT OF TRACE ORGANIC COMPOUNDS IN AQUEOUS SOLUTION. THE
CATALYZED WET OXIDATION OF TOXIC CHEMICALS, DEHALOGENATION OF
COMPOUNDS BY TREATMENT WITH ULTRAVIOLET LIGHT AND HYDROGEN.
ULTRAVIOLET/CHLORINOLYSIS OF ORGANICS IN AQUEOUS SOLUTION,
CATALYTIC HYDROGENATION-DECHLORINATION OF POLYCHLORINATED
BIPHENYLS. AND ULTRAVIOLET/OZONE DESTRUCTION. (NUMEROUS
DIAGRAMS, REFERENCES. TABLES) A
165965 -83-004698 fil"
HAZARDOUS WASTES: STATES SEEK FLEXIBLE RULES,
DOMBROWSKI CATHY H.
-------
TABLES
Number Page
1 Major U.S. Manufacturers of Ozone Generating Equipment 33
2 Typical Results of Destruction Tests with Malathion and DDT . . 42
3 PCB Combustion Tests in Sodium-Potassium-
Chloride-Carbonate Melts 45
4 HC1 Neutralization Potential, Fluidized Lower Bed-Aluminum
Oxide, Static Upper Bed-None, Waste Added as PVC-
Sodium Carbonate Pellets 64
5 HCl Neutralization Potential, Fluidized Lower Bed-Sodium Carbonate
Static Upper Bed-None, Waste Added as PVC Pellets 64
6 HCL Neutralization Potential, Fluidized Lower Bed-Aluminum
Oxide, Static Upper Bed-Sodium Carbonate, Waste Added as
PVC Pellets ..... ...... . 64
7 HCl Neutralization Potential, Fluidized Lower Bed-Sodium
Carbonate, Static Upper Bed-Sodium Carbonate, Waste Added as
PVC Pellets 65
8 HCl Neutralization Potential, Fluidized Lower Bed-Sodium
Carbonate, Static Upper Bed-Sodium Carbonate, Waste Added as
PVC Pellets, (% Sodium Carbonate Utilization) 65
9 HCl Neutralization Potential, Fluidized Lower Bed-Sodium
Carbonate, Static Upper Bed-Sodium Carbonate, Waste Added as
PVC-Scdium Carbonate Pellets 65
10 Total Participate Matter and Visible Emission Data 75
11 Gaseous and Vapor Phase Hazardous Air Pollutants
Emissions Data 75
12 Organics with Fast Destruction Rates 101
13 Organics with Moderate Destruction Rates 101
14 Preliminary Captial and Treatment Costs 101
-------
TABLES (Continued)
Number Page
15 Process Characteristics for a Proposed Blasting Abrasive
Fluidized Bed '. '119
16 Minimum Operating and Capital Costs for a 150,000 GPD
ULTRCK Treatment Pilot Plant to Obtain < 1 ppm PCB .... 121
17 Design Specifications, Captial, And 0 & M Costs for
40,000"and 150,000 GPD ULTRQX Treatment Plants
(50 ppm PCB feed-1 ppm PCB effluent)- . . .•-.-. . . . .-122
18 Matrix Relating Treatment Technique to User Needs 124
-------
ACKNOWLEDGMENT
Ebon Research Systems would like to thank our Project Officer,
Mr. Thomas Baugh for his support, assistance, and guidance throughout
this project.
XI
-------
SECTION 1
INTRODUCTION
PURPOSE OF THE REPORT
The purpose of this report is to identify and evaluate new and
emerging technologies for hazardous waste treatment and disposal. The
major technologies evaluated are:
• molten salt combustion
• fluidized bed combustion
• ultraviolet/ozone destruction
In addition to these technologies, processes employing catalyzed wet
oxidation, dehalogenation by the addition of hydrogen in the presence of
ultraviolet (UV) light, high energy electrons, UV/chlorinolysis and
catalytic hydrogenation-dechlorination are now in the developmental stage.
During the course of the study, major hazardous waste generators were
surveyed to determine whether they would be interested in information on
these and other techniques for hazardous waste disposal. The results of
this survey are also presented.
SOURCE MATERIAL FOR THE REPORT
The material for the identification and evaluation of these tech-
nologies has been gathered by an intensive literature survey conducted
over the course of a year. Although extensive use has been made of manual
and computerized data bases, it was also necessary to monitor the recent
literature and upcoming conferences and symposia to gain access to material
not yet published. Personal communications were also used in the survey.
CHARACTERIZATION OF HAZARDOUS WASTE
Hazardous wastes in this study are defined and characterized by the
Hazardous Waste Proposed Guidelines and Regulations and Proposal on
Identification and Listing published in the Federal Register on
December 18, 1978.
Most hazardous wastes in this study are organic substances. A listing
of hazardous wastes combusted by various technologies is included in the
appendicies. Mining wastes, plating, and metallic wastes are not con-
sidered in this report. Nuclear wastes and wastes from health care
facilities are beyond the domain of this report.
-------
SECTION 2
CONCLUSIONS
MOLTEN SALT COMBUSTION
Molten salt technology has been in existence for many years, but only
recently has molten salt combustion been used for the treatment of hazardous
wastes. In the process, hazardous material is combusted at temperatures
below its normal ignition point, either beneath or on the surface of a pool
of molten salt. Individual alkali carbonate salts, such as sodium carbonate,
or mixtures of these salts, are usually used as the melt, but other salts
can be employed based on waste characteristics. Containers for the molten
salts are made of ceramics, alumina, stainless steel, or iron.
Ideally, during the molten salt combustion process, organic substances
are totally oxidized to carbon dioxide and water, while heteroatoms such as
phosphorus, sulfur, chlorine, arsenic, and silicon are reacted with the
carbonate melt to form NaCl, Na,P04, Na2SO4, Na^305, and Na^i03. Iron
from metal containers forms iron oxide. Most organic substances are
destroyed, leaving behind a relatively innocuous residue, while harmless
levels of off-gases are emitted. Generally, the salt bath is stable,
nonvolatile, nontoxic, and may be recycled for further use until the bath
is no longer viable.
Some hydrocarbons combusted by the molten salt process are chlo-
rinated hydrocarbons, PCB's, explosives and propellants, chemical warfare
agents, rubber wastes, textile wastes, tannery wastes, various amines,
contaminated ion exchange resins, tributyl phosphate, and nitroethane.
The technology has progressed from bench-scale through the pilot plant
stage to the construction of a demonstration-size coal gasification unit.
Additionally, portable units mounted on truck beds have been used.
Advantages of Molten Salt Combustion of Hazardous Waste
Some of the advantages of molten salt combustion are:
Combustion is nearly complete
Non-polluting off-gases are emitted
Operating temperatures are lower than in normal incineration;
thus they are fuel efficient.
The system is amenable to recycling generated heat.
The system does not require highly skilled operators, i.e.
a professional engineer's license is not required.
-------
• A wide variety of hazardous wastes can be combusted.
• Bulky wastes can be combusted after pre-sizing.
• Many wastes can be combusted in compliance with EPA regulations.
Major Problems with Molten Salt Combustion of Hazardous Waste
Some problems with molten salt combustion of hazardous wastes are:
• Particulate emissions from some wastes are high, although
generally less than from normal incineration.
• The technology is not readily adaptable to aqueous wastes.
• The molten salt bath must be bubbling (but not ebullient) to
promote efficient combustion.
• Eventually waste salt and ash must be disposed of, or the
fluidity of the melt will be destroyed.
• A hazardous waste with greater than 20% ash cannot be combusted.
• Detailed economic information for a demonstration-sized system
is not currently available (1980).
FLUIDIZED BED INCINERATION
Fluidized bed systems have had many industrial uses since the tech-
nology was proposed by C. E. Robinson about a century ago, yet fluidized
bed combustion of hazardous waste is a relatively new technique. A hot
fluidized bed offers an ideal environment for combustion. Air passage
through the bed produces strong agitation of the bed particles. This pro-
motes rapid and relatively uniform mixing of carbonaceous materials. Bed
mass is large in relation to the quantity of injected waste, and bed temper-
atures, which usually range from 750-1000°C, are quite uniform.
Hazardous wastes that have been incinerated in a fluidized bed include
chlorinated hydrocarbons with a high chlorine content, waste PVC, waste PVC
with coal, PVC insulated waste wire, munitions (TNT, RDX, and Composition B),
spent HC1 pickling liquor, spent organotin blasting abrasive, and a waste
organic dye-water slurry. A listing of various materials used as the bed
medium is included in the Unit Operations Section.
Advantages of Fluidized Bed Incineration of Hazardous Wastes
Advantages of fluidized bed combustion of hazardous waste are:
« The combustor design concept is simple and does not require
moving parts after the initial feed of fuel and waste.
« Fluidized bed combustion has a high combustion efficiency.
• Designs are compact due to high volumetric heating rates.
• Nitric oxide formation is minimized because of low gas
temperatures and low excess air requirements. Low excess
air requirements also reduce the size and cost of gas
handling equipment.
• In some cases, the bed itself can act to neutralize some of
the hazardous products of combustion.
« The bed mass provides a large surface area for reaction.
• Temperatures throughout the bed are relatively uniform.
-------
• Continued bed agitation by fluidizing air allows larger
waste particles to remain suspended until combustion
is completed.
• If the hazardous waste contains a sufficient calorific
value, a fluidized bed combustor can operate without
auxiliary fuel.
• Excess heat generated by fluidized bed combustion of wastes
with high caloric value can be recycled.
• Fluidized beds are able to process aqueous waste slurries.
• As the bed functions as an efficient heat sink, major
variations in feed consistency and water content take a
long time to effect temperature changes in the bed medium.
• The heat sink effect also limits radiation from the bed and
allows the combustion system to be shut down for considerable
periods of time (weekends) and restarted with little or no
pre-heat time.
Major Problems With Fluidized Bed Incineration of Hazardous Waste
Some problems with fluidized bed combustion of hazardous wastes are:
• Bed diameters and height are limited with design technology;
therefore, maximum volumetric flow rates per unit are limited.
• Removal of inert residual material from the bed (such as ash)
can be difficult in some instances.
• In systems where temperature must be controlled at lower
limits because of other thermal considerations, increased
residence time can cause carbon buildup in the bed.
• Certain organic wastes will cause the bed to agglomerate,
thereby reducing its effectiveness.
• Particulate emissions are a major problem with fluidized bed
combustion. In some cases, particulates are high even when
emissions are passed through a Venturi scrubber.
UV/OZONE DESTRUCTION
Ozone treatment is an established technology for the treatment of some
hazardous wastes. Recent studies show that a combination of ultraviolet
light with the ozonation process is a more effective technique for destroy-
ing hazardous waste than the use of ozonation alone.
The addition of UV light to the ozonation process has greatly expanded
the number of compounds that can be destroyed. Exposed halogen atoms,
unsaturated resonant carbon ring structures, readily accessible multi-bonded
carbon atoms, and alcohol and ether linkages are particularly susceptible to
UV/ozone systems. Compounds with shielded multi-bonded carbon atoms, sulfur
compounds, and phosphorous compounds are less susceptible to UV/ozonation.
A combination of ozone with UV treatment has been used to destroy
PCB's, 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD), nitrobenzene and related
derivatives, and the hydrazine family of fuels.
-------
Many complex decisions representing trade-offs are necessary to
implement a well-designed, efficiently operated, economic UV/ozonation
system. This technology has advanced considerably in recent years, and is
close to state-of-the-art for many hazardous wastes. Some wastes which are
difficult to oxidize are not treatable by this technology. Reaction
kinetics plays a large role in the decision to apply UV/ozone technology
to the treatment of a specific waste. Although UV/Ozonation is much more
limited in the variety and concentration of wastes that can be treated (when
compared to conventional incineration, molten salt combustion, or fluidized
bed combustion), it is still a viable technology for the treatment of
certain hazardous wastes.
Advantages of UV/Ozonation
Some advantages of UV/ozonation are:
o Aqueous or gaseous waste streams can be treated.
o Capital and operating costs are not excessive as compared to
incineration
o Chemical carcinogens and mutagens can be treated.
o The system is readily adaptable to on-site treatment of the
hazardous waste.
o UV/ozonation can be used as a final treatment to supplement partial
treatment systems.
o It can be used as a preliminary treatment for certain wastes.
o It can be used to meet effluent discharge standards.
o Modern systems are usually automated, thereby reducing labor.
Disadvantages of UV/ozonation
Some disadvantages of UV/ozonation are:
o Ozone is a non-selective oxidant; therefore, the waste stream should
contain primarily the hazardous waste of interest.
o UV/ozone systems are generally restricted to 1% or lower levels of
hazardous compounds. Most hazardous substances that have been
treated by this process were in the ppm levels.
o Ozone decomposes rapidly with increasing temperature; therefore,
excess heat must be rapidly removed.
Superiority of Molten Salt Combustion, Fluidized Bed Incineration, and
UV/ozonation to Landfills'
All of the emerging technologies—molten salt, fluidized bed, and
UV/ozonation—can be considered as alternatives to landfill disposal of
hazardous waste. The intent of the emerging technologies is waste destruc-
tion, or at least attenuation, to acceptable levels. Landfills either store
waste in specialized containers or attempt to prevent its spread from the
area where it was dumped.
The future fate of hazardous wastes stored in landfills is, in many
cases, unknown. Deepwell injection, encapsulation, and other forms of
containment that do not attempt to destroy the hazardous waste share the
same uncertain future. If cost is not considered, the use of technologies
that destroy hazardous wastes should be considered far superior to landfill
-------
storage. However, insufficient information exists to compare the emerging
technologies on a cost/benefit basis with established landfill practices.
Molten Salt Incineration and Fluidized Bed Incineration vs.
Conventional Incineration
Both these technologies employ sufficiently high temperatures to
efficiently destroy many types of hazardous wastes. Molten salt •technology
can meet EPA regulations regarding a minimum high temperature and residence
time for the treatment of certain wastes, even though it employs much lower
temperatures than are used in conventional incineration. Compliance with
the regulations is achieved with much higher combustion efficiency, reducing
the emissions to lower levels. EPA regulations regarding combustion crite-
ria (Federal Register, Volume 43, Number 243, pp. 59008, 59009, December 18,
1978, #250.45-1 Incineration) state that the incinerator shall operate at
temperatures greater than 1000°C, greater than 2 seconds residence time, and
greater than 2% excess oxygen for the incineration of hazardous wastes.
Halogenated aromatic hydrocarbons, such as PCB's, must be incinerated at more
than 1200°C, and greater than 2 seconds residence time, with greater than 3%
excess oxygen. Combustion efficiency must equal or exceed 99.9%. However,
there is an exception. Other conditions of temperature, residence time, and
combustion efficiency are permitted if an equivalent amount of combustion
is demonstrated. Because of its higher combustion efficiency, molten salt
combustion complies with the regulation exception. PCB's are combusted by
molten salt technology at lower temperatures, yet the combustion efficiency
requirement of 99.9% was exceeded by molten salt technology with a combustion
efficiency greater than 99.9999% and a nominal residence time of 0.25-0.50
sec. Although both processes have some problems with particulate emissions,
they have less problems with participates than conventional incineration
techniques.
Molten salt combustion and fluidized bed incineration can be considered
as capital intensive for start-up costs. The same can be said for conven-
tional incineration. Little information exists for the costs of molten
salt combustion. The cost statistics for fluidized bed incineration are
better known at the pilot plant stage. Fluidized beds and molten salt
baths function as heat sinks, and can, in some instances, use the hazardous
waste combusted as a fuel. Thus, they are potentially less expensive
than conventional incineration techniques.
HAZARDOUS WASTE GENERATOR SUWEY
The results of the survey of hazardous waste generators indicated that
there is not a strong interest in some of the new, emerging hazardous
waste destruction technologies. This is probably because many companies
are not familiar with the advantages of the emerging methods compared to
the more established technologies such as landfills and conventional,
incineration. However, a high percentage of the respondents were interested
in an information service such as a computerized data base or a newsletter
regarding hazardous waste information.
-------
Ebon Research Systems believes that an information gap exists for the
dissemination of information regarding new hazardous waste technology.
A computerized data base and/or newsletter should consider the following
information:
• How to use established techniques more efficiently and
economically—this would include information such as Btu values
of specific combustion products.
« National, state, and county hazardous waste legislation—including
packing and transportation regulations.
9 Spill cleanup techniques
© Waste exchange and recycling
« Emerging destruction technology information. Types of wastes
treated, economics, and feed mechanisms would be detailed.
-------
SECTION 3
THEORETICAL BASIS OF THE TECHNOLOGIES
MDLTEN SALT COMBUSTION
The Catalytic Properties and Physical Structure of Molten Salts
When heated to a temperature slightly above their melting point, salt
mixtures exist in a transition state between solid and liquid. Similar
transition states exist at even higher temperatures between liquid and gas.
This transition state is analogous to an equilibrium of ionic and non-ionic
conditions. Measurements of conductivity sharply decrease in the solid
state. (This effect is often used to good advantage in an electrically
heated salt bath where the medium is also its conductor). The ionic
activities give rise to the reactive nature of the salt bath and are
proportional to the electronegativity of the ionic species present. The
higher the stability (i.e. melting point) of a salt medium, the greater its
reactivity upon thermal ionization.
According to Greenberg, most non-charged materials are soluble in
molten salts. This solubility is related to the crystal structure of
salts in the molten state. Data from x-rays taken at temperatures above
the melting points of salts indicate that molten salts retain a quasi-
lattice structure. It is assumed that the solute assumes an electronic
charge in the semi-crystalline melt. This charge gives the solute an
electrostatic orientation similar to that of the ionic component of the
molten salt and enchances the solubility of the solute (1,2).
Various pollutant species, especially hydrocarbon derivatives, react
with oxygen at relatively high energies of activation. However, when these
compounds are exposed to oxygen in molten salt baths, they are oxidized at
temperatures substantially lower than those normally necessary. Theoret-
ically, the new orientation of the formerly neutral species results in a
reduction of the energy required to initiate and sustain chemical reactions.
The ability of molten salts to dissolve various neutral substances and act
as catalysts for their oxidation is the basis for the feasibility of using
molten salts in hazardous waste destruction (1,2).
The salts used in the molten salts process should be stable at temper-
atures required for combustion of specific hazardous substances. A single
salt or mixture of salts may be used. The higher the stability of a salt
medium (i.e. melting point), the greater its reactivity upon thermal ioni-
zation. Eutectic mixtures are often useful as they provide good efficiency
8
-------
of operation at lower temperatures. In a eutectic mixture, the salts are
combined in a ratio that results in making the melting point of the mixture
less than the melting point of any of its components. In eutectic mixtures,
the melting range is evidence of the multiple equilibrium existing between
the two salt componnents (2,3).
There appears to be no one all-purpose melt for molten salt combustion.
The choice of salt(s) is dictated by the dual demands of reactivity and
melting range. One of the more important theoretical considerations is the
number of conversions that occur within the melt, as conversions result in
greater reactivity, promote efficiency, and reduce unwanted emissions.
However, the types and combinations of salts that can be employed in this
process allow it to operate at a wide range of temperatures and under
varied conditions of oxygen availability.
Specific Melts Used in the Process
Anti-Pollution Systems Neutral and Active Salts
and Their Eutectic Mixtures—
Neutral salts—Molten salts can be assigned to one of two major
classes. The first class can be considered as neutral salts. These salts,
or mixtures of salts, lower the energy of activation for oxidation of the
solute yet, they do not add oxygen or react chemically with it. They
require an external oxygen source supplied by combustion gas, ambient air,
or added oxygen (1).
Metallic halides with melting points in the 50°-600°C range are useful
in molten salt baths as neutral salts. The chlorides are especially stable.
A representative, but not all inclusive, list of neutral salt eutectic
mixtures which may be used in molten salt baths appears in Appendix A.
The listed temperatures represent the melting point of the bath and are
accurate to + 10°C. Any neutral salt with a melting point less than 600°C
can be used by itself, I.e., LiBr (547°C), Lil (446°C), and Cal (575°C) (1).
Active salts—Chemically oxidizing, active salts can also be used in
molten salt combustion to enhance the equilibrium pressure of oxygen both at
the surface and within the melt. These salts not only catalytically induce
oxidation, but also maintain an equilibrium oxygen gas pressure that facil-
itates oxidation by continuously donating nascent oxygen and retaking
ambient oxygen.
A partial list of chemically active oxidizers is shown in Appendix B.
The temperature listed is accurate to + 10°C and represents the lowest
temperature at which the melt may be used. Some of these mixtures are
combinations of neutral and active salts. These baths, like the neutral
melts, all possess sharp melting points.
It is possible to lower the melting point of most baths by addition of
lithium salts. The use of sulfate anion rather than carbonate is recom-
mended. Lithium sulfate forms a stable monohydrate.
-------
Oxidizing salt mixtures produce baths with lower melting temperatures
than neutral salt mixtures. Mixtures often improve the normal, oxygen-
releasing characteristics of the salts. For example, in the mixture of
nitrate and nitrite, nitrite has a tendency to cause nitrate to more
readily release oxygen. While neutral salt baths are normally used at
temperatures immediately above their melting points, the oxidizing salt
baths are normally used at temperatures approximately 93.3°C above their
melting point in order to facilitate oxygen release. The lower end of the
bath temperature range is limited only by the melting point of the salt (or
mixture of salts). A temperature as low as 50°C can be achieved by selecting
materials such as a mixture of thallium nitrate (50 M%) and silver nitrate
(50 M%). The composition and operating temperature of a bath is largely
determined by the composition and heat content of the material to be
oxidized (1,2).
Although oxidizing baths are in a neutral or inert state at tempera-
tures just above melting to 93.3eC above their melting point, they still
retain their catalytic capacity and can be used within the inert range in
order to avoid extreme temperature rises which may occur when highly
flammable material is added to the melt. Because molten neutral salt
baths operate at generally higher temperatures, the use of an inert ox-
idizing melt permits operation at temperatures where exothermic reactions
are in much less danger of producing undesirable explosions (1,2).
Molten Salt Technology at Rockwell International
In an early study done at Rockwell International (1969), Heredy et
al. removed sulfur oxides from flue gas using a mixture containing
Li- CO,, N^CQ,, and I^COj (400°C melting point). This mixture reacted
with sulfur dioxide and carbon dioxide to form alkali metal sulf ite.
The molten product of the reaction, consisting primarily of alkali metal
sulfite dissolved in molten alkali metal carbonate, was treated with a
gaseous mixture of hydrogen and carbon monoxide at a temperature between
400-650°C. The mixed alkali carbonates were regenerated, and at the same
time, hydrogen sulfide, a marketable product, was formed (7). This process
illustrates that a chemical reaction between solvent and solute can be
effective in the destruction of waste. A mixture of 45+M% Li2003, 30_+ M%
Na,CO,, and 25+ M% KgCQj was used for convenience. This was not the true
eutectic (5) .
Recent (1979) combustion studies have used a Na2QC>3 melt. The chemi-
cal reactions of the waste with Na2CC>2 depend on the chemical composition
of the waste. Carbon and hydrogen molecules are converted to CC>2 and H^p
(steam). Halogens form corresponding sodium halide salts. Phosphorus,
sulfur arsenic, and silicon form Na-,PO4, Na2S04, Na2As05, and Na^iC^.
Any iron present (from containers) forms iron oxide. Small quantities of
nitrogen oxides are formed by fixation of nitrogen in the air. Modifica-
tions of this basic melt will be discussed in sections on destruction of
specific wastes (6,7,8).
10
-------
FLUIDIZED BED INCINERATION
Concept of the Fluidized Bed
If a fluid (gas or liquid) flows upward through a bed of solid
particles, it exerts a frictional drag on the particles. This creates a
corresponding pressure drop across the bed. So long as the force is smaller
than the weight of the bed, as in fixed bed combustion, the particles will
remain essentially motionless, and the fluid will flow through the inter-
stitial passages (9).
As the fluid velocity is raised, a point is reached where the drag
force just equals the bed weight. This is the point of incipient fluidi-
zation. The fluid velocity at this point is called the minimum fluidization
velocity. As fluid velocity is increased above incipient fluidization, the
drag force is sufficient to support the weight of the particles, the
particles are bouyed up, exhibit great mobility, and behave like a fluid.
The bed is then fluidized and flows under a hydrostatic head (10,11).
A fluidized bed resembles a liquid in several ways (10).
• The upper surface stays horizontal when the container is
tilted.
o When 2 beds are connected, their levels equalize.
o Solids will gush in a jet from a hole bored in the side of
the vessel, and fine solids can be made to flow much like a
liquid through ducts and pipes.
© Although the upper surface of the bed is in motion, it is
well defined.
& A light object can float on the surface of the bed; a heavy
object will sink.
•« The pressure drop between any two points is esentially equal
to the static head of the bed between two points.
The relatively uniform temperature within a fluidized bed is due to
several factors. Turbulent agitation within the fluidized mass breaks and
disperses any hot or cold spots throughout the bed before they can grow to
significant size. There is also a rapid movement of solids from one part
of the bed to another. This does not mean that every solid particle in a
fluidized bed is the same temperature. However, departure of an indi-
vidual particle from the mean temperature of a fluidized bed will be much
less than in a fixed bed (11).
Continued bed agitation by fluidizing air allows larger waste parti-
cles to remain suspended until combustion is completed. Bed depths usually
range from about 40 cm to several meters. Variation in bed depth can
affect the residence time of combustibles. It is desirable to minimize
bed depth consistent with complete combustion and minimum excess air. In
general, a shallow fluid bed depth is preferred in a continuous process
because this provides the lowest pressure drop and power consumption as
well as maximum heath and mass transfer.
11
-------
Turbulent agitation within the fluidized mass breaks and disperses any
hot or cold spots throughout the bed before they can grow to significant
size. There is also a rapid movement of solids from one part of the bed to
another. This does not mean that every solid particle in a single fluidized
bed maintains the same temperature. However, the departure of an individual
particle from the mean temperature of a fluidized bed will be much less than
in a fixed bed (11).
Continued bed agitation by fluidizing air allows larger waste par-
ticles to remain suspended until combustion is completed. Bed depths
usually range from about 40 cm to several meters. Variation in bed depth
can affect the residence time of combustibles. It is desirable to minimize
bed depth consistent with complete combustion and minimum excess air. In
general, a shallow fluid bed depth is preferred in a continuous process
because this provides the lowest pressure drop and power consumption as
well as maximum heat and mass transfer. If longer residence time is re-
quired a high length-to-width ratio in bed size can optimize residence time
distribution (11).
The Character of Fluidization
The character of fluidization depends on whether the bed is fluidized
by a liquid or gas. In liquid fluidization, the neighboring particles
typically move further apart to accomodate increases in flow rate above
minimum fluidization, and the bed appears to undergo a stable uniform
expansion. This state is called particulate fluidization.
If gas is the fluidizing medium, a bed of fine particles (50-200U )
also exhibits particulate fluidization just above minimum fluidization.
When gas passes through the interstices of the dense phase solids, excellent
gas-solids contact is provided. When the gas velocity is increased,, a
second threshold, the minimum bubbling velocity, limits expansion. Any
additional gas traverses the bed in the form of voids or bubbles which are
virtually devoid of particles This is called aggregative fluidization.
The gas bubbles provide poor gas-solids contact. The mean bed particle
size and size distribution determine the quantity of incipient fluidization
gas for fixed feed conditions. In general, the smaller the bed particle
size, the larger is the quantity of gas that is able to pass through in the
form of bubbles (11).
Solid activity and the rate of particle carry-over increase with gas
velocity, yet a dense bed is retained over a wide range of gas velocities.
At some sufficiently high velocity, massive entrainment occurs, the bed's
upper surface is destroyed, and the bed may no longer be regarded as
stationary. In the absence of a cyclone, the bed would soon empty. Most
commercial fluidized bed combustors operate at gas velocities well below
this point (10). Studies on hazardous waste destruction in fluidized beds
that are discussed in this report use aggregatively fluidized, bubbling
beds.
12
-------
Bubbles in Fluidized Beds
The structure and behavior of fluidized beds are influenced by
bubbles. If a fluidized bed is viewed through transparent walls, the bed
appears like a boiling liquid. The upper surface of the bed in in motion,
but it is well defined. Bubbles rise through a fluidized bed and set it
into continuous vigorous agitation. They then burst at the surface, flinging
particles into the space above. Some bubbles, especially the smaller ones,
are elutriated from the bed. The small particles which escape are normally
trapped by cyclones and returned to the bed. Internal or external cyclones
may be installed (10).
If the bed is high and narrow, bubbles tend to coalesce and fill the
entire cross section of the bed. These large bubbles are called slugs, and
the bed is said to be a slugging bed. Data taken under slugging conditions
can be misleading for designing large diameter reactors (10,11).
In a bed of large particles, bubbling begins at incipient fluidi-
zation. The bubbles grow rapidly because of the density difference between
the bubble and the surrounding bed. This causes a pressure gradient that
drives gas into the bubbles. Also, because of the high viscosity of the
large-particle bed, bubbles rise slowly in them. Accordingly, gas in the
vicinity of the bubble uses it as a short cut, entering through the bottom
and exiting at the top.
If a bed is made of small particles, the rate of growth of bubbles is
small because of the high resistance to gas flow in the bubble. Since bed
viscosity is low, bubbles rise rapidly through the bed. The slow rate of
bubble growth and their short residence time in the bed decrease the
tendency for slug formation. As in the bubbles in large-particle beds, gas
also enters at the bottom of the bubble. It is then swept around and
dragged down the sides of the rising bubbles. The region around the bubble
is then called the cloud. Bed particles, outside of the cloud, do not
contact the gas. This is why poor conversions have been observed in
bubbling, fine-particle beds (10).
Fast fluidized beds
It is possible to operate at a gas velocity sufficient to blow most of
the solid material out of the reactor in a relatively short time if fresh
solid bed material is simultaneously added. This fast fluidized bed
operates at gas velocities above the bubbling regime and is free of large
voids or bubbles. There is some opinion that fast fluidized beds offer
several advantages over the bubbling bed. Yerushalmi and Cankurt (11)
claim that fast beds have higher processing capacities, more efficient
and intimate contact between gas and solid, and better capability in
handling cohesive solids. There is no indication that fast fluidized bed
combustion technology has been used in destruction of hazardous waste.
13
-------
UV/OZONE DESTRUCTION
Ozone generally is regarded as a resonance hybrid of four contributing
structures:
I II III IV
Ozone reacts as a 1 ,3-dipole or as an electrophilic reagent through the
electron deficient terminal oxygen atoms (structures III and IV). It also
reacts through the negatively charged terminal atom on all four structures
(12,13).
Ozonation of dilute solutions of organic species produces organic
oxidation and hydroxyl radical formation. Bridging of carbon-carbon double
bonds by ozone forms unstable ozonide intermediates that decompose into
smaller oxidation species. This process continues until carbon dioxide and
water or relatively stable refractory compounds (such as acetic or oxalic
acids) are formed (13).
Qualitatively, the overall reaction is depicted on the three regions
shown on the ozonolysis curve in Figure 1.
The mechanisms for many of these reactions are complex and, in some
cases not well understood. Readers interested in a more thorough treatment
of these reaction mechanisms are referred to surveys by Oehlshlaeger (12)
and Bailey (13).
In many experimental situations, reaction conditions do not produce
complete oxidation, and a number of intermediates remain in the reaction
solution. It is possible that many of these intermediates are as harmful
as, or more harmful than, the parent compound. It has recently been found
that, if ozone treatment is combined with UV (ultraviolet radiation), the
UV radiation activated the refractory compound and all subsequent refrac-
tory species. This permits nearly complete oxidation to carbon dioxide,
water, and other elementary species (14).
The following groups are especially vulnerable to attack by ozone
enhance UV absorption:
1. exposed halogen atoms
2. unsaturated resonant carbon ring structures
3. readily accessible multi-bonded carbon atoms
4. alcohol and ether linkages
Shielded multi-bonded carbon atoms, sulfur, and phosphorus are much less
vulnerable (14).
There is a substantial difference in the chemistry involved and
results achieved when the UV/ozonation process for treating hazardous
wastes is compared to straight ozonation.
14
-------
When ultraviolet radiation, in the 180-400 nm range, is added to the
process, 72-155 kcals/mole of additional energy is provided, and the 63
molecule breaks down into oxidizing O species. This is ample energy for
producing substantially more free radicals and excited-state species for
the initial conpound and subsequent oxidation species than those produced
by ozonation alone. As a result, excited atomic species (O), hydroxy (OH)
and hydroperoxy (HOg) radicals, and excited-state species (M)*are produced
from reactant molecules. These all greatly enhance the overall oxidation
rate, prevent plateauing, and permit complete oxidation. The following is a
simplified representation of how these excited species are probably formed:
R, I + H
Overall:
"I 23
R>+ OH, HO2, O, O > C02, ILO, C17 SO", P0~
where: M is the pollutant species being oxidized, R is a free radical
species, and I is an intermediate molecular species (14).
According to Prengle, elevated UV light also accelerates ozone de-
composition so that using UV excessively results in less favorable
economics (14). However, no generalization of requirements should be made
because of wide variation in reactivities of species. Free cyanide is
oxidized by ozone without elevated temperature or UV, but iron complexed
cyanides have maximum practical temperature and UV requirements. It is
better to treat acetic acid at naturally occurring pH rather than
neutralize, yet phenol destruction is not strongly pH dependent.
practical temperature and UV requirements. It is better to treat acetic acid
at naturally occurring pH rather than neutralize, yet phenol destruction
is not strongly pH dependent (15).
In summary, the overall mechanism of ultraviolet/ozone photooxidation
of M-species in aqueous solution occurs by a combination of the following:
© 03 photolysis to produce oxidizing 0 species
• f^O photolysis and reaction with 0 to produce OH
and HOo oxidizing species* 3
e M photolysis to produce M and free radicals
The free radicals mentioned above participate in a sequence of oxi-
dation reactions leading to the final oxidation products. For M-species
that are low UV absorbers, the rate controlling mechanism is photolysis
oxygen species oxidation, but for high UV absorbers, M-species photolysis
is rate controlling.
15
-------
1.0
TOG
TOG
0,. — Ozonolysis
REGION I — Oxidation of Initial Species
REGION I — Oxidation of Intermediates
REGION III — Unoxidizable
Refractory
Products
DIMENSIONLESS TIME
* TOG = Total Organic Carbon, TOCo = Initial Total Organic Carbon
Figure 1. Ozonolysis Curve
-------
SECTION 4
UNIT OPERATIONS
MDLTEN SALT COMBUSTION
The Rockwell International Combustors
In the Rockwell International process, combustible material and an
oxygen source, usually air, are continuously introduced beneath the surface
of a molten salt (7). The type of molten salt used is usually determined
by the nature of the waste to be incinerated. Sodium carbonate or potassium
carbonate, alone or in combination, is frequently used as the melt. Sodium
sulfate (approximately 10%) is sometimes added to the melt to increase
oxygen availability., Melts are usually maintained at 800-1000°C (6,7,8).
The method of waste addition is designed to force any gas formed
during combustion to pass through the melt before it is emitted into the
atmosphere,, The system is engineered to render any gaseous emission into
relatively innocuous substances. Theoretically, the intimate contact of
the waste, melt, and air causes a high heat transfer to the waste and
results in its rapid and complete destruction. As previously indicated,
the chemical reactions of the waste with molten salt and air depend on the
waste composition. Ideally, the off-gas should contain carbon dioxide,
steam, nitrogen, and unreacted oxygen. If particulates of inorganic salts
are present in the off-gas, they are removed by a Venturi scrubber or by
passing through a baghouse (6).
Ash concentrations above 20% must be removed to ensure fluidity of the
melto Batchwise melt removal is sufficient for low throughput applica-
tions o When the throughput is 250 kg/hr or higher, a side stream of the
melt is continuously processed. During the continuous side stream removal,
care must be taken to ensure sufficient salt remains in the melt. Residual
material, such a ash, is removed from the spent melt by quenching in water
followed by filtration. Recovered salt is recycled to the combustor (6,7).
Most of the published recent work (1977-1980) on molten salt combus-
tion of hazardous waste has been performed by Rockwell International in
their four molten salt combustion facilities. Two are bench-scale combustors
used for feasibility and optimization tests with feed rates of 0.25-1 kg/hr
of waste. The third faciltiy is a pilot plant combustor with a feed rate of
25-100 kg/hr of waste. It is used to obtain engineering data for design and
reliability extrapolation to a full sized plant. The fourth system is a
portable unit, designed for the disposal of empty pesticide containers.
17
-------
The Department of Energy-funded coal gasification process development unit,
with a design through-put of 1,000 kg/hr of coal, is also a Rockwell Inter-
national combustor (6,7,8).
Bench-scale Combustion System—
Unit design—A schematic of the bench-scale molten salt combustion
system is shown in Figure 2. The combustor contains about 5.5 kg of molten
salt in a 15-cm ID, 90-cm high alumina tube placed in a Type 321 stainless
steel vessel. The stainless steel vessel is in turn contained in a
20-cm ID, four heating zone Marshall (Ohio) furnace. The temperatures of
the four heating zones of 20-cm height are controlled by silicon-controlled
rectifiers. Furnace and reactor temperatures are recorded by a chart
recorder. An air cooling system, which prevents a temperature increase
in the system when an excessive amount of heat is released into the melt,
consists of an eight-hole air distribution ring mounted under the stainless
steel retainer vessel. Air, at rates near 500 1/min can be passed upward
between the outer surface of the retainer vessel and the furnace wall.
A 3.7-cm ID alumina feed tube is adjusted so that its tip is immersed
approximately 1-cm above the bottom of the 15-cm diameter alumina reactor
tube. Thus, the waste-air mixture is forced in a downward path through the
feed tube, outward at the bottom and, finally, circulates upward through
about 14 cm of molten salt. This insures complete and rapid mixture of the
waste with the melt (7).
One of the bench-scale combustors has been modified for the incinera-
tion of very hazardous wastes. The combustor is located in a walk-in hood,
and there is controlled access to the room which contains the hood.
In order to increase personnel safety, all process controls are located
outside the hood. Gas from the room is scrubbed in an activated charcoal
adsorber (7).
Waste feeding—The bench-scale molten salt combustion unit can accom-
modate solid, liquid, and mixtures of liquid and solid wastes. After .
solids are pulverized by a No. 4 Wiley Mill, they are placed in the hopper
and metered by screw feeding into the 1.2-cm OD central tube of the
injector. The screw feeder is rotated by a 0-400 rpm Eberback Corporation
Con-Torque stirrer motor. Hangups in the hopper are prevented or released
by a Syntron Model V-24 vibrator. The solids are mixed with about 75% of
the air used for combustion. The other 25% of the combustion air passes
through a cooling annulus in the injector (not shown in Figure 2). The
air-solids stream combines with the cooling air stream at the tip of the
injector. The mixture emerges into the alumina feed tube and then enters
the melt (7).
Liquid waste is pumped into the injector with a Fluid Metering, Inc.
(FMI) laboratory pump. As the liquid travels down through the central tube
of the injector, the combustion air passes down through the cooling
annulus. The liquid waste and air streams combine at the injector tip and
pass downward through the alumina feed tube into the melt (7).
18
-------
AIR IN
1.2 cm STAINLESS-STEEL
INJECTOR TUBE
• OFF GAS OUTLET
(ON-LINE SAMPLER
I J ATTACHED HERE)
STAINLESS-STEEL
RETAINER VESSEL
3.7 cm ID ALUMINA
FEED TUBE
0 to 400 rpm
SCREW FEEDER
MARSHALL FURNACE
/15cm ID
ALUMINA
TUBE
15 cm DEPTH OF
MOLTEN SALT
Figure 2. Bench-scale Molten Salt Combustion System
19
-------
Because the FMI pump is suitable only for solid-free liquids, slurries
are fed into the system by a Masterflex 7014 or 7016 peristaltic pump. A
5-liter stainless steel beaker located on a digital balance contains the
slurry and a stirrer. The pump and feed beater are connected by flexible
tubing. Since the digital balance is capable of reading 10 kg by 1-gm in-
crements, continuous mass feed can be calculated.
For types of waste where the solid is soaked with liquid and does not
have a suitable consistency for feeding as a dry solid or pumping as a
liquid, activated carbon or some other suitable absorbing agent is added to
produce a nearly-dry, free-flowing material (7).
Off-gas analysis for bench-scale combustors—A typical on-line combus-
tion off-gas analysis schematic is shown in Figure 3. A sampling of the
off-gas is first passed through a high efficiency particulate air (HEPA)
filter to remove particulates. The particulate-free gas is then analyzed
by Rockwell International for NOX, CO, ©2, N2, and unburned hydrocarbons.
A water-ice trap is used to remove most of the water vapor upstream of all
the analyzers except the NOX analyzer. Because some NOx might condense, an
ice trap is not used upstream in the NOX line. Rockwell International
analyzes NOX with a Thermo Electron Corporation Chemiluminescent NOX
Analyzer. CO2 analyses are made with an Olson-Horiba, Inc. Mexa-200
Analy Analyzer. The CO and hydrocarbon analyses are made with an Olson-
Horiba, Inc. Mexa-300 Analyzer. Determinations for 02 are made with a
Teledyne portable oxygen analyzer, Model 320.
For continuous gas determinations and experimental control, the Olson-
Horiba and Teledyne instruments are used. A Perkin-Elmer gas chromato-
graph is used to confirm these analyses. Syringe samples (1 ml) of the
off-gas are removed downstream from the ice trap and injected into the gas
chromatograph. Chromatographic columns Porapak Q and Molecular Sieve 5A
are operated at 100°C. A typical Rockwell International gas Chromatographic
analysis assays CO2, CO, N2, 02(+Ar), and unburned hydrocarbons (7).
Pilot Plant Combustion System—
A schematic of a Rockwell International molten salt pilot plant is
shown in Figure 4. The molten salt vessel is made of Type 304 stainless
steel and lined with 15-cm thick refractory blocks. The 4-m high,
0.85 m ID vessel contains 1000 kg of salt. Melt depth is 1 m when air
does not circulate through the bed. A natural-gas-fired burner is used to
preheat the vessel on start-up and maintain heat during standby. The heat
content of the waste is usually sufficient to maintain the melt in a molten
condition during combustion periods (7,8).
The molten salt vessel is loaded via the carbonate feeder. Combustible
materials to be processed are crushed to the required size in a harnnermill,
transferred into a feed hopper equipped with a variable-speed auger, and
finally, introduced into the air stream for transport into the vessel (7).
Exhaust gases generated in the vessel pass through refractory-lined
tubes in the vesel head before they enter a refractory-line mist separator.
20
-------
HEATED
PARTICULATE
FILTER
COOLED
BENZENE
SCRUBBER
COMBUSTOR
ICE
TRAP
GAS
CHRCMATOGRAPH
CO, and CD
IR ANALYZERS
NO CHEMILUMINESCENT
ANALYZER
Figure 3. Off-gas Analytical Schematic
Source: Rockwell International (1979)
-------
KJ
MELT AND ASM
TO DISPOSAL
I
STACK
Figure 4. Pilot Plant Molten Salt Combustor
Source: Rockwell International (1979)
-------
The mist separator traps entrained melt droplets on a baffle assembly. The
gases are then ducted into a high-energy Venturi scrubber or baghouse to
remove particulate matter prior to release into the atmosphere. During ex-
tended testing, an overflow weir can be used to permit continuous removal of
spent salt (8).
Portable Molten Salt Disposal System—
In the Rockwell International portable molten salt unit for disposal
of empty pesticide containers, the combustor and auxiliary components are
mounted on a truck bed. The combustor, which is 1.8-m ID and 3.4-m tall,
can process 225 kg/hr of waste. The empty containers are conveyed to a
shredder, shredded (particle size 0.95-3.2 cm), and pneumatically conveyed
to the combustor. Off-gas is cleaned by a particle separator to remove
smaller particulates before the gas is emitted from the stack (17).
Theoretically, such a unit would yield only about 250 liters of spent
salt for each 907 kg of containers processed. Spent salt is drained into a
drain cart and buried in a Class 1 dump. After cooling, the system is
transported via truck to a new site where fresh salt is added to the com-
bustor and the cycle repeated (17,18).
Rockwell International Coal Gasification System—
The Rockwell International process demonstration unit (PDU) is another
example of the scale-up potential for a molten salt system. This process
development unit, funded by a Department of Energy contract, is able to
gasify coal into low, intermediate, or high Btu gas. The system is designed
to accept all coals, including highly-caking, high sulfur bituminous coals.
Crushed coal is gasified at high temperature and high pressure by reaction
with air in a highly turbulent mixture of molten sodium carbonate containing
scdiun sulfide, ash, and unreacted carbonate. The process demonstration
unit can accept 907 kg/hr (19).
The Anti-Pollution Systems Combustion Unit
Anti-Pollution Systems (APS) has developed an alternative molten salt
process based on the molten salt technology discussed earlier (1,2,4). In
one application of the process, textile manufacturing waste containing
acrylics residue was purified. The contaminated liquid waste, gravity-fed
at a rate of about 1,995 I/he was introduced onto the surface of a molten
salt bath composed of 62 mole% KNO^38 mole% CafMD^ 2. Although the bath
melts at 140°C, it was maintained at 450°C. Under these conditions, water
in the waste was completely flash-evaporated, leaving behind an organic
residue which ignited and rapidly decomposed under the bath's catalytic
influence (2).
The bath was housed in a container constructed of noncorrosive material.
The container had a main chamber where evaporation and combustion of the
liquid takes place. In an adjacent baffled chamber, the gases of combus-
tion produced in the main chamber are again brought into contact with the
melt to further combust any incompletely oxidized products. Waste is
23
-------
introduced into the chamber from above. The contact surface of the bath is
about 1.5 m x 6 m with a 5.1 cm-deep melt. Heat is supplied by an open gas
flame. Air, provided at a flow rate of about 1400-2800 1/min, is maintained
at a steady flow by a blower at the input and a pump at the output. Air,
carbon dioxide, and water vapor are off-gases. No measurable hydrocarbons
or carbon monoxide were detected in the off-gases (4).
The APS molten salt unit has been modified in several ways so that it
can function not only as an incinerator for carbonaceous materials but
also purify automotive exhaust systems and some types of stack emissions
(1,2). Because the APS molten salt scrubber is capable of removing
particulates in the submicron range, Greenberg suggests that molten salt
scrubbing systems could be effective for the destruction of toxic sub-
stances such as Kepone (4).
In a modification in which the salt bath operates as an afterburner
(shown in Figure 5), the system consists of a stainless steel box within
a box. The inner box (trough) floats on 7.6 cm of salt. The floating
inner trough makes ash removal simple and is designed to preclude problems
when water is introduced directly into the melt. If alkali sulfate is used
in the melt, it is estimated to have a heat capacity of 593°C-equivalent to
a 38.1 cm bed of cast iron. Liquids (containing water at any concentration)
or solids are introduced into the center box and exposed to a flame (4).
As the waste combusts, the generated heat is transferred to the salt bed
beneath the combustion chamber. If the heat of combustion is sufficiently
large, and enough waste is burned, it is not necessary to premelt the salt
bed. The generated heat maintains the bottom portion of the combustion
chamber at the melting point of the salt. Exhaust gases produced by com-
bustion are pulled through a series of baffles and bubbled through the melt
before exiting. This provides a second incineration for toxic, volatile
substances and traps particulates in the melt (4,18).
Suction is supplied by an induced draft fan which creates a negative
pressure on the baffle closest to the exit side of the system. This causes
a rise in the liquid level with an accompanying drop in the melt level on
the exit side of the baffle. The exhaust gases impinge on the liquid front
created by the suction. A modified dry cyclone collector prevents salt
losses at air velocities above 123,000 1/min. Carbon particulates and
inert materials are removed by a fine mesh stainless steel screen (4,18).
Although the temperature of the APS melt is usually maintained between
560-600°C (lower than the 800-1000°C generally used in the Rockwell Inter-
national unit) with an estimated residence time of approximately 0.1 second
in the melt, the combustor can also be operated under residence times and
temperatures which satisfy EPA regulations for chlorinated organics (4).
There are currently (1979) three molten salt units at APS. One is
portable and can be fueled with propane (Personal communication from J.
Greenberg to Dr. Barbara Edwards of Ebon Research Systems, July 12, 1979).
Analysis methods, detailed off-gas descriptions, and precise operating
parameters are not currently (1979) available in the literature for the APS
molten salt process.
24
-------
EXHAUST
r
\HOLTKN SALT
Figure 5. APS Molten Salt Incinerator
Source: Wilkinson et al. (1978)
-------
Disposal of Waste Melts
Ash and other non-combustibles, such as glass or iron oxides, build up
during combustion. When ash concentration exceeds 20 wt%, the ash must be
removed to preserve melt fluidity. Additionally, the reaction of inorganic
compounds such as halogens, phosphorous, sulfur, and arsenic with sodium
carbonate ultimately causes sufficient melt conversion that results in loss
of pollutant removal capability (6).
At Rockwell International, melt is transferred to a dissolving tank and
treated with water or aqueous sodium bicarbonate solution. The dissolved
carbonate solution is filtered to remove insoluble materials and treated
with carbon dioxide gas to precipitate sodium bicarbonate. The bicarbonate
salt is recovered by filtration and used in the molten salt furnace where it
is converted to carbonate. Eventually, the bicarbonate solution contains
excess chloride, phosphate, and sulfate and must be discarded (18).
When waste throughput exceeds 250 kg/hr, with sufficient sodium carbon-
ate remaining in the melt, a melt side stream is continuously withdrawn and
processed to recover and recycle the carbonate. Smaller waste throughput
is removed batchwise. These srnall quantities of unprocessed melt and spent
melt which no longer can be recycled are drained into a drain cart and
buried in a Class 1 dump (a non-sanitary landfill underlain by unusable
groundwater) (6).
In a patent for the disposal of explosives and propellants, Yosim
either regenerated the melt or reacted it with molten lime or aqueous lime
to form a water-insoluble calcium salt residue. These insoluble salts were
formed by the reaction of lime with carbonate, fluoride, phosphate, and
sulfate components in the spent melt (20).
Ycsim also treated pesticide disposal melts with air to oxidize any
residual sulfide to sulfate. The melt was then .cooled or treated with water
and lime for disposal in an approved dump site. Yosim suggested that a.
melt containing only chlorides from chlorinated pesticides decomposition
could be dumped in the ocean without pretreatment (21).
Greenburg, in a patent for the oxidation of carbonaceous materials
with molten salts, removed floating residues from the surface of the bath
and used a dredging apparatus to remove settled products (1).
Although no details for the disposal of spent melt were cited by
Edgewood Arsenal scientists when they combusted various military chemical
agents, a disposal problem existed with the Na,AsO4 containing spent melt
from the combustion of arsanilic acid (23).
It is possible that this type of hazardous spent melt can be insolu-
bilized in glass or concrete. Salt samples from spent Rockwell Interna-
tional non-hazardous melts were glassified using different fluxing materials
as the solidifying agents. The N^SO,* and NaCl content of the resultant
solidified glass varied from 10-40 wt%. Classification temperatures ranged
from 1100°-1500°C. An accelerated leach test, using continually flowing
26
-------
distilled water at 100°CL produced accelerated leach rates of 10 to 10
g/cm/day compared to 10 to 10 gm/cm/day for normal leach tests. The
authors suggested clay, Pyrex, and probably basalt can be used to adequately
glassify waste molten salts from the combustion or organics, using present
molten salt technology (22).
Insolubilization in concrete was tested by Rockwell International
scientists on waste salts containing NaJ3D-, Na SO4, and NaCl mixed with
cement. The salt content of the salt-cenent mixture was 10-38 wt%. The
authors stated that since the low level leach rates of these concrete
samples were only about 5-10 times more than glassified waste salt, the
less expensive cement formation would be an acceptable disposal method for
waste melts (17).
FLUIDIZED BED INCINERATOR
A schematic of a typical fluidized bed incinerator is shown in Figure 6.
Air is driven by a blower and passes through a distributor plate into the
bed above the plate. Sand is usually, but not always, used in the bed.
Typical materials used as a fluidized bed are listed below:
« sand • dolomite
© alumina • ferrous oxide (granular)
o sodium carbonate • the waste itself, e.g., spent
organotin blasting abrasive
The upward flow of air (or some other fluidizing substance) through
the bed results in a dense, turbulent mass. Material to be incinerated is
injected into the combustor by specialized pumps, screw feeders, or pneu-
matically. Air passage through the bed promotes rapid and uniform mixing
of the injected material within the bed. A cold reactant will attain the
temperature of the bed almost instantly. Water quickly evaporates as
combustion takes place. Auxiliary fuel (oil or gas), if needed, is usually
introduced directly into the bed. Suspended fine particles are collected
in a cyclone. Steam and the other gaseous products of combustion exit,
along with suspended fine particles, from the top of the reactor. Exhaust
gases are cleaned in a scrubber and exit to the atmosphere (24). As the
specific combustion parameters vary with the type of waste combusted, more
specific details on unit operations are discussed in the section on the
types of wastes destroyed by the emerging technologies.
ULTRAVIOLET/OZONE DESTRUCTION
Ozone (03) is a three atom allotrope of oxygen. It is second only
to fluorine in electronegative oxidation potential and has long been
recognized as a powerful disinfectant and oxidant of both organic and
inorganic substances. Generated by solar energy, ozone is a natural in-
gredient of the earth's atmosphere. It is also generated from atmospheric
oxygen by energy from lightning and is associated with operation of most
electrical equipment. Ozone is a gas under ambient conditions. Unreacted
ozone decomposes in a matter of hours to simple, molecular oxygen. Ozone
is not a poisonous chemical in the sense that it enters into internal body
27
-------
AUXILIARY
BURNER
FLUE GAS
MAKEUP
SANEL
ACCESS
DOOR .
DISTRIBUTOR
PLATE
ASH
REMOVAL
Figure 6. Schematic of a Fluidized Bed Combustor
Source: Powers (1976)
28
-------
chemistry. However, because of its strong oxidizing properties, exposure
to low concentrations of ozone is damaging to delicate nasal, bronchial,
and pulmonary membranes (26).
Ozone Generators
It has never been conclusively proven that ozone can be formed by
purely chemical means. Unlike most chemicals, there is no controlled
available natural source for ozone, and it is not practical to store it in
containers. Ozone is generated when an oxygen molecule is sufficiently
excited to disassociate into atomic oxygen; further collisions with oxygen
molecules then cause the formation of the ozone molecule. Excitation energy
can be supplied by ultraviolet light or high voltage (25,26).
A major source for ozone is an ozone generator. Most ozone generators
use high voltage, although ultraviolet ozone generators are practical for
outputs less than 1 gm/h. High voltage generators produce a corona dis-
charge (also known as a silent arc discharge or a brush discharge). When
electrons flow at sufficiently high potential through a gas that contains
oxygen, a bluish glow accompanies the excited molecular state (25,27).
The following conditions are necessary to create a corona discharge:
® two electrodes separated by a gap
• a gas in the gap
• sufficient voltage potential between the two electrodes
to cause current flow through the dielectric and gas
The electrodes can be flat, tubular, or any configuration that allows
their opposing surfaces to be parallel. The distance between the parallel
surfaces of the electrodes should be large enough to insure uniform current
flow. Depending on the system, a gap smaller than a certain critical size
may restrict the air flow to a point of excessive pressure drop through the
unit. A large gap increases voltage requirements (25,26).
The two basic types of commercial ozone generators in use are the
concentric tube generator and the parallel plate generator. The con-
centric tube system, first devised by Siemens, can be modified to operate
at higher pressures than the plate design (26). One electrode is a metal
grid or coating inside a cylindrical glass tube. A metal cylindrical tube
fits over the glass tube and functions as the second electrode. The glass
acts as a dielectric to increase the voltage gap between electrodes without
sparking. Ozone is formed in the oxygen or air flowing longitudinally in
the gap between the tubes. Manifolds distribute the fresh gas flow to
multiple tubes and collect ozone-containing gas at the exits (28).
In one design, shown in Figure 7, there is a stainless steel center
electrode through which water passes for heat removal. Concentric to that,
on the outside, is a glass dielectric tube. The outside surface of the
glass is coated with a metal that serves as the second electrode. The
dielectric material is used to prevent arcing from one electrode to another.
The second coated electrode and the glass electrode are oil-cooled (non-
conductive fluid) which in turn is water-cooled in a heat exchanger (29).
29
-------
In the more conplex Otto system, high voltage, parallel plates alter-
nate with ones that are water cooled. Gas flows in the air gaps between
the plates. There is more generating surface within a given volume than
in the concentric tube system. This design is being replaced by the
air-cooled Lowther system. Water cooling is more costly than air cooling,
but many years of experience with water systems has developed relatively
trouble-free techniques (26).
In 1979, only one manufacturer produced ozone generators with parallel
plates. The electrodes are mounted on an aluminum frame for dimensional
stability and arranged as small, square modules. Dielectric coatings are
applied to the electrodes to increase the voltage gradient, and ozone is
formed in the air gap between the plates. Manifolds are used to distribute
and collect the gas. Major U.S. manufactures of ozone generating equipment
are listed in Table 2 (28).
Parameters for Ozone Generation
Dielectric—
Operational efficiency of an ozone generator is dependent on parameters
involved in design and engineering (26). The dielectric material ideally
should have both a high electrical resistance and high thermal conductivity
Since these properties rarely occur together, the dielectric is usually
chosen for its high electrical resistance and depth to a minimum thickness
to overcome heat transfer deficiency. Glass or other ceramic materials are
usually used. Some polymers have demonstrated superior properties for
short time periods. Because both ozone production and ozone concentration
greatly depend on the quality of the dielectric materials used on the
electrode, U.S. Ozonair Corp. suggests these minimum requirements (26).
• Dielectric constant £ = 85
• Electric Strength 15 Kv/mm
« Volumetric resistivity 10 ohm/em
Pressure—
The operating pressure should provide sufficient force to deliver
ozonized gas to the contacting vessel. The pressure may have to overcome
a .back pressure of from 0.6-0.9 meters of water at the reactor. Because
pressure affects the electrical impedance of the system, the operating
pressure must be correlated with the air gap and preferred voltage (25).
Flow Rate—
The time that feed gas is in the electrical discharge determines the
concentration of ozone in the effluent. This concentration can vary from
0.5-10 wt% in a well designed generator operating at ideal conditions. The
power efficiency drops rapidly as the ozone concentration increases (25).
Temperature—
Approximately 90% of the energy applied to the ozonator is lost as
heat. The ozone decomposition rate increases with increasing temperature,
and provisions should be made for the rapid removal of excess heat. Not
only does ozone output vary as a function of generator temperature, but
also, as ozone generator temperature increases, the dielectric material
changes thermal characteristics and is subject to rupture. As critical
30
-------
generator dielectric temperatures are relatively low (120°C-130°C), the
economics of regrigeration or coolant quantity supplied to the system must
be evaluated versus the temperture of the ozone stream (9,10). Water is
usually considered to be a more efficient medium for heat transfer than
air. Air used for cooling ozonators can be of any quality, and when it
passes through a generator, it is not degraded or changed. Typical gener-
ator cooling fluid requirements are 2831 ra3 of air/454 gm 03 or 1789
liters of water/454 gm 03 (30).
Feed Gas—
Ozone generation using oxygen is approximately twice as efficient as
when air is the feed gas (i.e., ozone generated from air requires at least
twice ths power of an equal amount of ozone generated from pure oxygen).
However, oxygen cost is more than twice the cost of air. Unless oxygen is
recovered via a closed system, it is cheaper, especially when ozone gener-
ated is less than 225 kg/day, to generate ozone from air. A recently
developed technique uses a "swing cycle" that concentrates oxygen from air
by alternate passage through molecular sieves (25,30).
The corona intensity of new ozone generators requires perfectly clean,
dry, and oil-free feed gas. If not, either ozone production is reduced, or
the electrodes are damaged and need frequent cleaning. A minimum air dew
point of -45°C is recommended to insure dependable ozone production (29).
If ozone generators operate at high pressures (as in the tubular
design), dehumidification is enhanced and micron and submicron filters
can be used in the air-feed system.
31
-------
TUBE TXPE PLATE TYPE
Fiqure 7. Ozone Generators
3
32
-------
TA3LB 1
MMOR U.S. MANUFACTURERS OF OZONE GENERATING EQUIPMENT
U)
CO
Manufacturer & Address
Crane Cochrane
P.O.Box 191
King of Prussia, Pa 19406
215-265-5050
Emery Industries, Inc.
Ozone Technology Group
4900 Estee Ave.
Cincinnati, Oh. 45232
513-482-2100
Ozone Research & Equip. Corp.
3840 N. 40th Ave.
Phoenix, Az. 85019
602-272-2681
PCI Ozone Corporation
One Fairfield Crescent
West Caldwell, NJ 07006
201-575-7052
Equipment
Concentric tubes
SS/g 1 ass/a 1 urn i nun
Series C-cabinet
Series P-skid mounted
Concentric tubes
SS/gl ass/n i chroma
Skid mounted
Concentric tubes
SS/gl ass/SS
Series V,B,& D-cabinet
Series H-skid mounted
Concentric tubes
SS/glass/silver
Series G-cabinet
Series B-skid mounted
Made 1 s-Capaci t ies
Ib OVday
air feed
Scries C, 1-18 Ib/day
Series P, 18-122 Ib/day
Series 9270, 1-23 Ib/day
Series 9260, 21-400 Ib/day
Series B & V, 1/4-2 Ib/day
Series D & H, 4-250 Ib/day
Series G, 1-28 Ib/day
Series B, 35-1400 Ib/day
Typical 03
Concentration
Cooling Method in air, $
Water on outer
electrode
Water on outer
electrode
Water on outer
electrode
Water on inner
electrode.
Oil on outer
electrode
1
1
1
2
-------
TABLE 1 (Continued)
MAJOR U.S. MANUFACTURERS OF OZONE GENERATING EQUIPMENT
Manufacturer & Address
Welsbach Ozone Systems Corp.
3340 Stokely St.
Philadelphia, Pa. 19129
215-226-6900
Infilco Degremont, Inc.
Roger Executive Center
Box K-7
Richmond.Va 23288
804-285-9961
Union Carbide, Linde Div.
Environmental Systems
P.O.Box 44
Tonawanda, NY 14150
716-877-1600
U.S. Ozonair Corp.
464 Cabot Rd.
S.San Francisco,Ca 94080
415-952-1420
Equipment
Concentric tubes
SS/glass/SS
Series CLP & GLP both
skid mounted
Concentric tubes
SS/glass/aluminiin
Skid mounted
Parallel ceramic coated
steel Lowther plates
Concentric tubes
Ti tani un/ceramic/aluminum
Models-Capacities
Ib 03/day
air feed
Series CLP, 24-127 Ib/day
Series GLP, 170-322 Ib/day
No .model designations,
10-600 Ib/day
No model designations,
1-1200 Ib/day
Series HF, 5-570 Ib/day
Typical O3
Concentration
Cooling Method in air, %
Water on outer
electrode
Water on outer
electrode
Air on outside
both electrodes
Water on inner
electrode.
Air on outer
electrode
1
1
1
2
-------
SECTION 5
WASTES DESTROYED BY EMERGING TECHNOLOGIES
HAZARDOUS WASTES DESTROYED BY THE MOLTEN SALT PROCESS
Background
In 1975, the U.S. Environmental Protection Agency sponsored a study by
Battelle Pacific Northwest Laboratories to assess molten salt technology
for the pyrolysis of solid waste. Although no actual studies were per-
formed on hazardous wastes, Battelle recommended that molten salt technology
receive further investigation for processing materials such as waste
pesticides and herbicides with low ash content, waste nerve gases, biologi-
cal warfare agents, and noxious fumes (31). A discussion of hazardous
wastes treated by the molten salt process follows.
Explosives and Propellants
Obsolete and old explosives and propellants are usually destroyed by
burning in an open area or detonation in a safety zone. These destruction
processes emit pollutants such as smoke, hydrogen chloride, nitrogen
oxides, carbon monoxide, and undesirable dust clouds. A patented molten
salt process was developed by Rockwell International scientists for the
disposal of explosives and propellants with minimal resultant environmental
pollution (20). The process was developed primarily from combustion tests
performed on 5 g satples of Composition B (60 wt% RDX, also known as
Cyclonite, and 40 wt% TNT) and Standard Solid Propellant (70 wt% ammonium
perchlorate, 10 wt% Aluminum, 14 wt% polybutadiene, plus other unidentified
chemical substances).
The molten salt bath was located in the center of an armor-plated,
three-sided cubicle with viewing windows. Five gram pieces of explosive
were rolled through a non-metalic inclined pipe which led from the outside
of the cubicle to the molten salt bath. The reaction between the melt and
the explosive plus its effect on the surrounding atmosphere was observed
through the windows. The bath was contained in a stainless steel vessel
surrounded by an insulated clamshell heater. Melt temperature was moni-
tored prior to the introduction of the waste.
Runs were performed with a ternary lithium carbonate-sodium carbonate-
potassium carbonate eutectic and a NaOH-KOH eutectic at 400-600°C. A
control run burned the explosive in the open without any melt present.
There was vigorous bubbling in the melt before the explosive ignited, during
35
-------
the burn, and after combustion ceased. Times for complete combustion ranged
from 3-11 minutes (20).
According to Yosim, a wide variety of commercially significant explo-
sives and propellants may be effectively treated by the molten salt process.
The particle size and weight of the explosive will, to some extent, deter-
mine the feed rate of the material to the molten salt combustor. The
chemical composition determines the type of melt used. Temperature control
may be required if a given substance has a low explosion temperature.
Yosim states that the reactivity of the alkaline molten salts used and the
physical state of the explosive or propellant (solid, liquid, mixture) are
probably not significant factors in the effectiveness of the process-
Moreover, it is also probably not important whether the explosive is
present as a single or binary composition, or whether the propellants are
double-base compositions containing various additives or polymeric matrix
propellants. Thus, any of the bipropellant systems containing liquid fuel,
such as the hydrocarbon fuels of the JP types, ethyl alcohol, hydrazine and
hydrazine derivatives are feasible for molten salt decomposition.
The following explosives are considered by Yosim as suitable for the
molten salt destruction process (20):
ammonium nitrate diethylene glycol dinitrate
glyceral nitrate TNT
• diglyceral tetranitrate Tetryl
glycol dinitrate Cyclonite
trimethylolethane trinitrate HMX .. ,
PETN . Composition B
DEPHN
Chemical Warfare Agents
A series of molten salt combustion tests on various chemical warfare
agents were performed at Edgewood Arsenal by their personnel using a
bench-scale combustor built and designed by Rockwell International staff.
Agents tested included G3 spray-dried salts (the product of the chemical
neutralization of GB) , GB, VX, distilled mustard (HD), Lewisite (L), and
obsolete war gas identification sets. The agents present in the sets along
with the packaging material and dunnage in the sets were incinerated
together in order to evaluate the feasibility of molten salt incineration
for their disposal. Although the combustor was similar to the Rockwell
International bench-scale apparatus described earlier, there were some •
modifications in the unit and analytical methods employed (23).
Combustion of GB Spray-Dried Salts —
GB (C^^QPF) was neutralized by aqueous sodium hydroxide to produce a
solution containing mainly sodium isopropylmethylphosphonate (NaGB) , NaF,
NaOH, Na2C03, and varying amounts of disodium methylphosphonic acid and
diisopropyl-methylphosphonate. Spray-drying removes water from these com-
pounds and produces "GB spray-dried salts."
36
-------
Under certain conditions, NaGB, diisopropylmethylphosphonate, and
disodium methylphosphonic anhydride can react with NAP to reform GB. As
the removal of as much as 99.95% of the fluoride ion is insufficient to
prevent reformation of trace quantities of GB, the complete destruction of
the organophosphorous residue is necessary to eliminate the hazard of GB
production during handling or storage of the spray-dried salts.
Although incineration can combust organophosphorous material, the
noxious by-product, P20c, is produced. Because there is no conventional,
environmentally safe method for the disposal of GB spray-dried salts, the
molten salt process, with its ability to scrub undesirable substances in
the melt during the reaction, was tested. Earlier molten salt studies by
Rockwell International scientists on Malathion revealed P^s emissions were
substantially reduced, but not eliminated. A bench-scale program was
therefore established for the molten salt incineration of GB spray-dried
salts to determine the efficiency of the combustor for converting organo-
phosphorous chemical agents in N33PO4 without excess P£>c emissions. The
parameters varied were melt temperature (900°-950°C), melt depth (21 cm
and 31 cm), and melt composition. The melt composition was either 90 wt%
Na2C03 with 10 wt% Na2S04, or 100 wt% Na2CQ3. The feed rate was 4 g/min
or 12 g/min. Phosphorous pentoxide, particulates, GB, and other organo-
phosphorous levels were also measured.
The combustion should take pl^ce according to the following reaction:
o .
+ II 13
Na 0-P-OCH(CH )2 + Na2CCL + =| 02 » Na3PC>4 + 5 C02 +5
CH3
The NaF, NaOH, and Na2C03 salts are presumed to be retained in the melt,
unaltered by the combustion process.
Increased feed rate, with constant temperature, melt depth, and melt
composition, resulted in lowering P^s levels from 2.15 mg/333 1 to
1.82 mg/ 333 1. It is theorized that burning elemental phosphorus in excess
oxygen favors the exclusive formation of P£>y while burning in limited
oxygen reduces volatile oxides. Increased melt depth decreased P^s levels
in both melt compositions. The mixed melt produced more P^5 at higher
temperatures while the 100% ^CC^ melt decreased P^ at higher tempera-
tures.
Previous incineration tests of phoshorous-containing compounds indi-
cated that a mixed Na2C03-Na2S04 melt minimized P205 emissions. In tests,
it was also observed that the pure carbonate melt not only increased P;~
emissions, but also the total particulates. In some cases, the sodium
carbonate melt expelled partially incinerated NaGB (23).
37
-------
Distilled Mustard, HD—
Distilled Mustard was combusted in a melt consisting of 90% Na_CO, and
10% N^SO*. The quantity of salt mixture used produced an unexpanded Bed
depth of 30 cm. An air feed rate of 0.94 I/minute, increased the bed
depth to about 60cm. Melt temperatures ranged from 915"C to y35°C.
The combustion should take place according to the following reaction:
C1CH2CH2-S-CH2CH2C1 + 2 Na2C03 + 7
Na2SO. + 2 NaCl + 6 C02 +4
No unincinerated mustard was detected in the off-gases, on the partic-
ulate filter, or in a sample of the melt. NC^ levels ranged from 4-15 ppn.
SO? was less than 0.11 ppn, HC1 ranged from 2.2-16 ppn, and CO levels were
measured at 0.06%. Particulate emissions were 2.15 mg/333 1 and consisted
mainly of sodium carbonate, sodium sulfate, and sodium chloride. Destruc-
tion greater than 99.999997% was reported (23).
vx— . • ,. .......... ...••.
VX was combusted in the same type of melt used for distilled mustard.
The temperature was maintained between 920-930°C. The combustion for VX
should take place according to the following reaction: • •
0
+ 3 Na2CX>3 + 41
. 2
Na.PO4 + Na2SO4 + NaN03 + 13
Unincinerated VX in the off-gases, on the particulate filter, or in a
melt sample was negligible. NC^ levels ranged from 9-70 ppn, SO2 was less
than 0.14 ppn, CO levels ranged from 0.04-0.10%. P^^ was measured at
856 ppn. Particulate data were incomplete. Destruction greater than
99.999988% was reported (23).
Lewisite, L—
Lewisite was combusted in the same type of melt used for distilled
mustard. The temperature was maintained between 810-880°C. The combustion
38
-------
should take place according to the following reaction:
ClCH=CHAsCl2 + 3 Na2C03 + 3 02 —» 3 NaCl + Na-jAsSC^ + 5 CX>2 +
Less than 0.0053 mg/1000 1 unincinerated Lewisite was detected.
was reported at 15 ppn, CO levels ranged from 0.04-0.10%, and HC1 was
11-15 ppm. Participate arsenic emissions averaged 299 ppm. Residual
arsenic in the melt ranged from 2.1-2.4 mg/gm (23).
GB—
The combustion of GB in molten salt should take place according to
the following reaction;
8
13
F-P-O-CH (CH3) 2 + 2 Na2C03 + =j 02 - > NaF + Na
CH3
Test runs using a melt, of the same composition as that used for dis-
tilled mustard showed no detectable amounts of GB in the off-gas. Signifi-
cant levels of Po05 (as high as 955 ppm) and Na-fQ^ (value not stated) were
reported„ Fluoride determinations were not made. Destruction of GB was
calculated to be 99.9999985%.
Pesticides and Herbicides
A patent was assigned to Rockwell International Corporation in 1974
for a molten salt disposal process invented by Yosim et al. This process
was designed for the ultimate disposal of organic pesticides and herbi-
cides with negligible participates and off-gas pollution. According to the
process description, the organic pesticide and a source of oxygen are fed
into a melt composed of sodium carbonate and from 1-25 wt% sodium sulfate.
Temperature ranges are between 850-1000°C with an optimum range at 900-
950°C (21).
Certain pesticides are completely combusted in the molten salt with an
excess of oxygen. Other pesticides are partially oxidized in the melt and
the remaining gaseous products are conducted into a second reaction zone
where oxidation of any combustible matter still present is completed.
Usually more than one zone is used to achieve complete combustion and
ultimate disposal of the pesticide (21).
39
-------
When chlorinated hydrocarbon pesticides are treated, NaCl is formed
in the melt. Sodium phosphate is formed during the treatment of organic
phosphate pesticides. Sodium sulfate is produced from the treatment of
sulfur-containing pesticides. All of these inorganic compounds are
retained in the melt. When the melt is no logner able to react with the
pesticide, the salt is removed and fresh salt added (21).
Chlordane—
A few tenths of a gram of 50% chlordane (CicP#ls) powder were added
intermittently to a 85 wt% WagQ3-15 wt% Na2S04 melt maintained at 980°C.
Air flow through an air inlet tube forced the pesticide through about
30.5 cm of molten salt in order to permit better contact between.pesticide
and melt. It was concluded that 99.9% of the pesticide was rapidly de-
composed. Exit gases contained products of the reaction between carbon-
aceous material and the sulfate-primarily .carbon dioxide and carbon monoxide.
If the ratio of air to pesticide increased, the presence of carbon monoxide
in the off-gas decreased. The levels of chlorinated hydrocarbons in the
off-gas ranged from 10 ppm to 150 ppm. The higher range of hydrocarbon
levels resulted when the air to pesticide ratio was relatively low. No
chlorides were detected in the water scrubber when silver nitrate was
added. Assay demonstrated that the melt retained about 80 wt% of the
total chlorine.
Larger amounts of chlordane, contained in polyethylene bags at levels
of 5-10 gm, were added to the melt through an alumina air tube which was
open at both ends. The 10 gram quantities: were substantially destroyed in ,••,•
about two minutes. -
Liquid chlordane as a 72% emulsifiable concentrate was completely ::
combusted in a continous feed system with excess air. Analysis of par- •
ticulate filter and water scrubber fractions, used to trap any emitted
organic chlorides indicated greater than 99.9% destruction of the pesti-
cide by the molten salt.
Assay of the off-gas indicated levels of Np^ at less than 70 ppm, hy-
drocarbon emissions at less than 25 ppm, less than 0.1% CO, and 80% N (21).
Malathion—
Malathion is a representative organophosphorous-type pesticide which
also contains the heteroatom sulfur. Ten polyethylene bags, each con-
taining 5 gm of Malathion, were added to the melt at three minute intervals
via an additional port. The closed port forced all emissions through the
secondary combustor (described previously). Emissions were then passed
through a glass wool particulate trap, through water scrubbers, and finally
out of the system (21).
Benzene extracts of the particulates in the glass wool and water
scrubbers were evaporated to dryness and analyzed for residues. Analyses
reported 1.2 mg sulfur and 1.8 mg phosphorus. These figures represent
99.9% destruction of the pesticide. The melt analysis assayed 80 + 15%
40
-------
phosphorus. No CO or hydrocarbons were detected above the secondary com-
bution region. Monitoring of CO and hydrocarbon emissions indicated that
5 grams of sample were destroyed in about 30 seconds.
Grantham et al. combusted Malathion powder in the Rockwell Interna-
tional bench-scale combustion unit shown in Figure 2 (depicted earlier).
A 900°C, 15-cm deep melt consisting of either NajCC^ or K£O3 was used.
Destruction of the pesticide was reported greater than 99.99%. Although
no residual pesticide was detected in the melt, traces of Malathion (within
the threshold limit value) were sometimes detected in the off-gas. An
increase of the bed depth to 25 cm markedly reduced pesticide levels, and
decreased off-gas particulates (22).
Yosim et al. also used the bench-scale combustion unit in Figure 2
(depicted earlier) to combust Malathion dissolved in xylene. He reported
better than 99.99% removal at 1000°C. No pesticide was detected in the
melt, however, in one test a trace of Malathion was found in the off-gas.
Weed B Gon—
Weed B Gon consists of 17.8 wt% of the isooctyl ester of 2-4-D and
8.4 wt% of the isooctyl ester of silver. Six polyethylene bags, each
containing 5 g of the herbicide, were added to a melt comprised of 90 wt%
Na2C03 and 10 wt% Na^04 maintained between 950°-1000°C. Analysis of
benzene extracts indicated 0.77 mg residue of the 1.48 g chloride added to
the system in the total emissions. Theoretically this represents a pesti-
cide destruction of at least 99.96% (21).
Sevin™
Sevin also known as Carbaryl, 1 Napthyl N-methylcarbamate, contains
7 wt% nitrogen. Five gram packets of this typical carbamate pesticide
were added to the same type of melt used for Weed B Gon. Analysis of
benzene extracts indicated that 0.075 mg of 0.875 g of initial nitrogen
remained. Destruction was reported to be 99.99%. Because of the rapid
carbon-nitrate-nitrite reduction that occurs in the melt, nitrogen was not
detected in the melt. Peak nitrogen emissions were 20 ppm (21).
DDT Powder and DDT-Malathion Solution—
Tests with the Rockwell bench-scale molten salt combuster were con-
ducted for DDT powder and solutions of DDT and Malathion dissolved in
xylene. This data is summarized in Table 2. These substances were
combusted at 900°C in a 15-cm deep salt bed containing either Na2C03 or
K2C03o The feed rate was 227-907 g/hr. Destruction of the pesticide was
greater than 99.99%. Although no pesticides were detected in the melt,
traces (0.3 mg/1000 1) were found in the off-gas (8).
2,4-D Herbicide-Tar Mixed Waste—
A mixed waste composed of 30-50% 2,4-D (an ester of dichlorophenoxy
acid) and 50-70% bis-ester and dichlorophenol tars, was reported completely
41
-------
TABLE 2
TYPICAL RESULTS OP DESTRUCTION TESTS WITH MALATHION AND DDT
Pesticide Salt
DDT Na CO
DDT K2C°3
Malathion Na_CO
Malathion ^-)~®->
Pesticide
Destroyed
99.998
99.998
99.9998
99.999
Concentration
of Pesticide
in Melt
(ppm)*
ND 0.05
ND 0.2
ND 0.01
ND 0.005
Quantity in
Exhaust Gas
(mg/m )*
0.3
0.3
0.06
ND 0.4
TLV* of
Pesticide
(mg/m )*
1
1
15
15
3 -4
*ND = not detected, one mg/m = 4.4 x 10 grains/scf, TLV = threshold
limit value.
42
-------
destroyed in the Rockwell International pilot plant molten salt combustor
at 830°C. No organic chlorides or HC1 were detected in either the melt or
off-gas. The waste was first diluted with ethanol to reduce viscosity (8).
Other Pesticides—
Since the molten salt process involves destruction of an organic
compound, whether the initial step is partial or complete destruction,
Yosim et al. postulated that a wide variety of organic pesticides may be
destroyed with relatively minor treatment modifications. Possible candi-
dates for molten salt destruction are DOT, dieldrin, heptachlor, aldrin,
toluidine, the nitrile herbicides (trifluralin, 2,4,5-T dichlorobinil,
MCPA), and the phosphorous containing insecticides (diazinon, disulfonton,
phorate, and parathion) (21).
Real and Simulated Pesticide Containers—
Combustion tests were conducted on bench-scale and pilot plant
levels for pesticide container materials comprised of paper, plastic,
rubber, and a blend of these. Combustion was reported complete in all
cases. The value of carbon monoxide in the off-gas was less than or equal
to 0.2%. NOx values were less than 65 ppm, and unburned hydrocarbons
were detected at less than 30 ppm. In the tests with PVC, no HC1 was
detected in the off-gas.
The pilot plant combustor was used for the destruction of 700 kg
simulated pesticide container material- at a feed rate of about 30 kg/hr.
Complete and rapid combustion was reported. The waste was composed of
53 wt% paper, 32 wt% polyethylene, 8 wt% PVC, and 7 wt% rubber. Less than
5 ppm wt% HC1, less than 2 ppm sulfur dioxide, less than 0.1% CO, and less
than Oo1% hydrocarbons were detected in the off-gas. NOX was about 30 rprn.
Values of 5-12% 02, 10-15% C02 and 76-78% N2 were reported in the off-gas.
The relatively small pesticide residues found in empty containers indicated
it is possible a cooler melt (Na2C03~K2C03 eutectic, m.p. 710°C) could be
considered.
Noncombustible materials such as glass and metal were also tested.
Glass reacted completely in about 30 minutes at 900eC. At 1000°C, the
glass reacted rapidly. Reaction of a metal (unspecified) was considerably
slower, with surface corrosion to a depth of 1.75 mm in 8 hours at 900°C.
It was recommended that molten salts not be used to completely disintegrate
metal containers. A decontamination rinse by immersion for several minutes
is an alternative (8).
Hazardous Organic Liquids
PCB's—
Current EPA regulations require incineration at 1200°C, and two
seconds residence time, to ensure complete PCB destruction.
43
-------
Molten salt PCB destruction was tested with the Rockwell International
bench-scale unit. The PCB's studied (from voltage transformers)/ had an
empircal composition of Cq^QQH^Clq. Tests were performed at temperature
ranges of 700-980"C and air-PCB ratios from 90-230 wt% stoichiometric air.
An initial melt of Na2CC>3 with I^CCU added to lower the melting point was
used. Because NaCl was formed _in situ by reaction of the carbonate with
the chlorine of the PCB waste, the melt composition varied (7,22).
A gas chromatograph monitored PCB levels during molten salt combus-
tion. Approximately 20% of the off-gas was scrubbed in benzene bubbles to
absorb PCB. The samples were concentrated, mixed with freon, reevaporated,
and mixed with exactly 1 cc of carbon disulfide. Duplicate and triplicate
60 yl samples of the carbon disulfide mixture were assayed in the gas
chromatograph. The analytical methodology was verified by employing the
same methodology with samples of the original PCB (Inerteen 70-30).
As little as 100 ppb PCB was measured with the flame ionization detector
gas chromatograph. PCB standards were diluted and used to establish
chromatographic responses and evolution times for several PCB's.
Table 3 summarizes the result of the PCB tests. No PCB's (less than
70 ug/1000 1) were detected in any of the off-gas samples when all the
Na2COq had been converted to NaCl. Destruction was estimated to exceed
99.9999%. However, it is necessary to maintain sufficient carbonate levels
(about 2 wt%) and excess air in order to ensure complete combustion and
maintain emission of NOX, CO, CH4, H2, and unburned hydrocarbons at low,
acceptable levels. The nominal residence time of the PCB in the salt was
0.25-0.50 seconds based on a flow-rate of 30H50 cm/sec through -15 on of-
the melt (7,22). . '•'• '-'•.' ' : ' : ' '' '.'•''.'•'•''.
Chloroform— . ' r
Rockwell International scientists have combusted chloroform in a
molten salt reactor using both continuous feeding and bulk feeding techni-
ques. Chloroform was fed to the bench-scale combustor at 227-907 g/hr.
The temperature of the Na2C03 melt was 850°C. Unreacted chloroform and
HCl was not found in the off-gas. Greater than 99.999% destruction was
reported. The same melt parameters were used for combustion in the pilot
plant. Bulk quantities of chloroform were contained in metal canisters and
plunged into the melt. This technique eliminated inconvenient, costly
shredding equipment (7).
Three to four sealed Pyrex vials containing 160 ml chloroform were
packed in cardboard tubes or sawdust and placed in 10-on diameter metal
canisters which were sealed. In some tests, the metal canister was
punctured to allow melt access to the glass vials. The canisters were
plunged into the melt. Vials in the punctured canister ruptured after
less than 1 minute immersion in the melt with little noise. Vials in the
unpunctured canisters ruptured noisily from 1-3 minutes after immersion
into the melt (7).
Monitoring the off-gas with Drager tubes revealed less than 1 ppm
hydrogen chloride, less than 1 ppm phosgene, and less than 5 ppm chloroform.
44
-------
TABLE 3
PCB COMBUSTION TESTS IN SODIUM-POTASSIUI^CHLORIDE-CARBaWTE MELTS
Temp
°C
870
830
700
895
775
775
Stoichio-
metric
Air
(%)
145
115
160
180
125
90**
Concentration Extent of
of KC1, NaCl PCB*
in Melt Destruction
(Wt%) %
60
74
97
100
.100
100 .
ND
ND
ND
ND
ND
ND
99.99995
99.99995
99.99995
99.99993
99.99996
99.99996
Concentration of
PCB in Off-gas**
( g/m3)
ND
ND
ND
ND
ND
ND
52
65
51
59
44
66
* ND = None detected; one g/m = 4.4 x 10 grains/scf.
** Insufficient air present to completely oxidize PCB to CO and HO,
i.e., some CO, H , and CH present in the off-gas.
Source: Rockwell International (1979).
45
-------
Trace amounts of chloroform were found in all the benzene scrubber samples
analyzed. However, the levels were near the detection limits. Chloroform
destruction was estimated at 99.95% (7).
Perchloroethylene Distillation Bottoms—
Perchloroethylene (€2014) is a common industrial solvent. During the
refabrication of high temperature, gas-cooled nuclear reactor fuel, per-
chloroethylene is used to scrub eondensible cracked and incompletely
cracked hydrocarbons, carbon soot, and uranium-bearing particulates from
off-gas. Eventually, the perchloroethylene degrades and requires distill-
ation. Oak Ridge National Laboratory reports that the bottoms from this
•distillation contain a large number of complex polynuclear aromatic
compounds. Although these compounds are variable and difficult to char-
acterize, there is good analytical evidence that this waste contains
carcinogens. Perchloroethylene distillation bottoms represent a type of
hazardous waste for which complete destruction is necessary.
A slurry which contained 93 wt% perchloroethylene, 6 wt% organic
degradation products, and about 1 wt% solids (carbon, silicon carbide,
uranium, etc.) was feed by peristaltic pump into a Rockwell International
bench-scale molten salt combustor. Since the heating value of the waste
was less than 70 joules/g, kerosene was added to furnish the necessary heat
to maintain the temperature (6,7).
Although experimental conditions such as temperature (850-950°C),
stoichiometry (30-100 wt% air), and perchloroethylene-kerosene weight ratio
(8-1) were varied, no peaks of organic material above background levels
were found in any off-gas sample. Based on the analytical methods employed,
it was concluded that any compounds in the off-gas from the combustion of
perchloroethylene bottoms should not be in greater concentration than 0.5
rag/1000 1. As long as about 1% NajCC^ remained in the melt, the concentra-
tion of HC1 in the off-gas was less than 2 ppm. NOx levels in the off-gas
were 25 ppm. The particulate content of the exhaust gas increased with
increasing time. This was expected since the particulates consisted mainly
of NaCl formed from the vaporization of NaCl in the melt (6,7).
Trichloroethane—
Trichloroethane (C^H^l^ is an industrial chemical which contains
80 wt% Cl. This chemical was combusted in the bench-scale unit at the rate
of 227-907 g/hr.
This test was designed to define the amount of sodium carbonate which
could be converted into NaCl without affecting the chloride scrubbing capa-
city of the melt. With as little as 2 wt% N^COg remaining in the melt,
only a negligible amount of trichloroethane was detected in the off-gas.
It was estimated that 99.999% of the chemical was destroyed (6,7).
46
-------
Nitroethane, Diphenylamine HC1, Monoethanolamine—
When the nitrogen-containing compounds monethanolamine
nitroethane (€^5^02), and diphenylamine HC1 (C^i^ici), were com-
busted in the Rockwell International bench-scale molten salt unit at
840-922°C, no unreacted material was detected and 99.99% destruction was
estimatedo Yet, these nitrogen-containing compounds did produce substan-
tial amounts of NQ^ „ Further tests indicated that NOjj concentration
could be sharply reduced by adjusting the air/waste fuel ratio. In theo-
retical 165 wt% (excess) air, monoethanolamine combustion produced 2200 ppm
NC^. This quantity was reduced to 220 ppm at 108 wt% theoretical air if
combustion takes place in a semi-reducing environment. As nitroethane
contains a nitro group rather than an amine, the NC^ emission was 16,000
ppni at 100 wt% stoichiometric air. Theoretically, increased reducing con-
ditions could lower the NC^ concentration below 108 ppm (6,7).
Tributyl Phosphate—
Tributyl phosphate (C^^^PC^) often used as a fire retardant, is
difficult to burn in a conventional incinerator. Both pure and diluted
tributyl phosphate (in 30% kerosene) were combusted at 900°C with 45 wt%
and 28 wt% excess air. The high 003 and low CO and hydrocarbon concen-
trations in the off-gas indicated rapid consumption of the chemical. The
CO, content of the off-gas measured 10-14%, CO, 0.5%, unburned hydrocarbons
20"pnn, and 30 ppm NQ^ (7).
Hazardous Solids
The following hazardous solids were combusted in the Rockwell Inter-
national bench-scale molten salt unit.
Rubber—
In tests conducted to study the production of low Btu gas from in-
dustrial wastes, rubber tire buffings were gasified at 920°C with 33 wt%
theoretical air (percentage of air required to oxidize material completely
to C02 and water). Since the buffings contained organic sulfur which would
form Na2S in the malt 6 wt% N32S was added to the sodium carbonate melt to
simulate conditions and function as a catalyst to accelerate char gasifi-
cation (18).
The CO content of the off-gas was considerably lower than from the
oxygen-containing wastes that were also used in the tests (i.e., wood,
nitropropane, and film). No significant amounts of H^ or other sulfur-
containing gases (less than 30 ppm) were detected in the off-gas.
Para-Arsanilic Acid—
Para-arsanilic acid was combusted at rates of 227-907 g/hr. NO un-
reacted material was dected in the melt or off-gas in tests performed at
925°C» The main combustion product, sodium arsenate, was retained in the
melt as expected,, Thus the melt must also be considered as hazardous (7).
47
-------
Contaminated Ion-Exchange Resins—
Ion-exchange resins are difficult to burn in a conventional inciner-
ator and generally produce a smoky flame containing evolved hydrocarbons.
Tests performed on nuclear contaminated Dowex-1, and Powdex resins (styrene
divinyl/benzene cross-linked polymer with trimethyl amine anion or sulfonic
acid cation component) revealed rapid and complete destruction (7).
High-Sulfur Waste Refinery Sludge—
A high-sulfur (up to 15 wt%) waste refinery sludge was readily pro-
cessed in the Rockwell International bench-scale molten salt gasifier. As
a useful by-product, a high quality, low-heating value (170 Btu/scf), low-
sulfur gas was produced (7).
HAZARDOUS WASTES DESTROYED BY THE FLUIDIZED BED PROCESS
Background
Many industrial uses have been proposed for fluidized bed systems
since this technique was suggested by C.E. Robinson about a century ago.
However, it was not until the late 1920s that the first commercial fluid-
ized bed unit involving a gas-solids mixture and utilizing elevated
temperatures was installed by the petroleum refining industry (9). The
phenomenon of fluidization supports the transport of a large bulk of
catalyst from a reactor to a regenerator and back (10). In 1942, the
Standard Oil Company of New Jersey opened a fluidized bed petroleum
catalytic cracking plant in Baton Rouge, Louisiana (9). Since then,
fluidized solids technology has been firmly established as a useful and
valuable industrial operation (11).
After it was adopted by the petroleum industry, fluidized bed tech-
nology was successfully applied to many gas-solids operations in other
industries. Included in these applications were metallurgical processes
such as roasting sulfide ores and oxide ore reduction. Fluidized beds have
also been used extensively by the pharmaceutical and food industries for
rapid and intensive drying of powdery and granular materials (11).
Because fluidized bed incineration incorporates both waste disposal
and energy recovery features, there has been much recent interest in the
technology as it relates to coal gasification, electric power generation,
and boiler units. The state-of-the-art of these energy efficient applica-
tions in 1977 was summarized at the International Conference on Fluidized
Bed Combustion (32).
One of the first applications of fluidized bed technology for the
incineration of carbonaceous waste material was in the pulp and paper
industry. Research on this application was begun at the Columbus Labora-
tories of Battelle Memorial Institute in the late 1950s. This culminated
in the erection of a Container-Copeland commercial installation at the
Carthage, Indiana mill of the Container Corporation of America in 1962.
Fluidized bed combustion is especially useful in the combustion of spent
48
-------
pulping liquor. The spent liquor contains sodium-sulfur compounds of
varying degrees of oxidation in association with organic matter extracted
from wood. In 1977, there were more than 25 commercial fluidized bed
installations in the pulp and paper industry (9).
Dorr-Oliver entered the fluidized bed incineration field in 1960 with
a small unit for handling approximately 10 kg/h of primary sludge. This
was followed by a commercial sized unit built at Lynwood, Washington, in
1962. Since then, commercial fluidized bed units have incinerated many
different kinds of industrial waste. These include oil refinery waste,
primary and activated sludges, and carbon black waste (33).
Although fluidized bed technology is established in many industries,
it has not been extensively used for the ultimate disposal of hazardous
waste. In 1978, Systems Technology Corporation (Systech) in Franklin,
Ohio, evaluated fluidized bed combustion in the destruction of methyl
methacrylate and phenol waste for the U.S. Environmental Protection Agency.
Incineration of each waste was accomplished with high efficiencies in the
combustor (34). Other investigators have evaluated fluidized bed combus-
tion of a limited number of hazardous wastes. The following material will
detail new and emerging treatment systems for the combustion of hazardous
waste by fluidized bed technology. Both bench scale and pilot plant systems
will be discussed.
Chlorinated Hydrocarbons
There have been several different types of tests in which chlorinated
hydrocarbons, e.g, polyvinyl chloride (PVC), have been combusted in fluid-
idized bed incinerators. Polyvinyl chloride (PVC) is one of the more widely
used plastics,, The monomer of PVC, vinyl chloride, has two carbon atoms,
three hydrogen atoms, and one chlorine atom. The chlorine content is
45 wt% of pure PVC.
PVC is rarely pure and usually contains fillers, stabilizers, plasti-
cizers, flame retardants, and other chemicals. The major products of PVC
combustion are carbon monoxide, carbon dioxide, and hydrogen chloride.
Free chlorine is not a product of PVC combustion, but trace quantities of
carbonyl chloride and vinyl chloride have been reported. Fifty-six other
volatile products have been observed in small amounts (35).
Chlorinated Hydrocarbons/Bench Scale Processes
Plastic Waste Combusted with Coal—
A study was done on combustion of PVC mixed with coal in a fluidized
bed. The effectiveness of dolomitic limestone, aluminum oxide, and silica
sand in removing hydrochloric acid was also investigated. Ragland and Paul
used a 9-on diameter, 0.3-jn tall, quartz-lined fluidized bed. The tubular
combustor wasd flanged to a 0.3 m long freeboard (area above the fluidized
bed) tube and a 12-cm diameter cyclone particulate collector. A water
cooled pneumatic feed system was used near the bed. The distributor plate
was porous, sintered bronze. For start-up, a propane-air mixture was used
49
-------
to preheat the bed. After preheating, the propane was shut off, and the
inlet air was preheated to approximately 90°C by means of a coiled tube in
the freeboard. Flashback in the propane-air mixture was controlled with a
flash arrester, and the plenum chamber (below the distributor plate) was
filled with steel wool (35).
Preliminary experiments showed that thin film PVC sheet, when shredded
to 0.3 cm, could be readily fluidized in a sand bed without segregation.
Blends of 5 to 30 wt% shredded PVC with Montana subbituminous coal con-
taining 25% moisture, 35% carbon, 10% ash, and 0.6% sulfur were used in the
feed.mixture. The particle size of the coal was 8-20 mesh. The PVC, which
contained some filler, was ground and sieved to the same size range. The
mixture had a 17.5 wt% chlorine content.
The efficiency of dolomitic limestone (calcium carbonate with calcium
magnesium carbonate), aluminum oxide, and silica sand (size range 8-20
mesh) as chlorine scrubbers was investigated. These materials have melting
points well above the bed temperature. The major source of chlorine was
the PVC. The chlorine content of the coal, aluminum oxide, and silica sand
was negligible. Neutron activation revealed 0.4% chlorine in the dolomite.
Bed depth was approximately 7.6 cm.
Combustion tests were conducted using 5, 20, and 30 wt% PVC-coal
blends with the three different bed materials. The temperature averaged
840°C. Good fluidization was obtained with a flow of 0.25 m/minute. The
PVC/coal mixture was injected with a pulse of air about every 5 seconds.
Three hundred grams of feed were burned every 7-10 minutes.
The combustion gases from the fluidized bed were sampled at the
cyclone exit with a modified EPA method 6 sampling train. A 1-cm diameter
quartz probe electrically heated to 200°C was used. A pyrex wool plug in
the end of the probe removed participates before the sample was drawn into
four midget impingers in an ice bath. Three impingers contained dilute
NaOH solution for absorbing HC1. The last impinger removed moisture.
A Teldyne Model 980 flue gas analyzer was used to pump the gas sample and
to determine the oxygen and gaseous combustibles in the flue gas. The
impinger solution and probe-plug wash were titrated with mercuric nitrate
to determine the HC1 concentration.
Chlorine emissions for the three different beds and at different PVC
concentrations are shown in Figure 8. The concentration of HC1 in the flue
gas was 500 ppm with 5 wt% PVC, using silica sand and aluminim oxide beds.
A dolomite bed decreased this concentration to 135 ppm. The total possible
HC1 emission was calculated at 900 ppm for 5 wt% PVC. The ability of dolo-
mitic lime stone to scrub HC1 is due to the reaction between calcium oxide
and HC1 to form calcium chloride. The formation of magnesium chloride
could also be effective in removing HC1. Aluminum oxide was not nearly as
effective as dolomite, and aluminim chloride was apparently not formed.
The spent silica sand and aluminum oxide beds contained little chlorine
(Figure 9). The retention of chlorine in these systems occurred primarily
in the c/clone ash.
50
-------
Chlorine
in
Exhaust
Gas,
grams
10
9
7 -
6 .
5 .
3 -
2 .
1 -
0
• Dolomite
A Aluminum Oxide
QSilica Sand
o
10
r
15
T
20
i
25
Percent PVC in Fuel
I
30
3000
Chlorine
in
Exhaust Gas,
ppn
2000
1000
Figure 8. Chlorine Emissions in Flue Gas
51
-------
15
10
g en
ol
ffl
• DOLOMITE
AALUMINUM OXIDE
D SILICA SAND
0 5 10 15 20 25 30
PERCENT PVC IN FUEL
Figure 9. Chlorine Adsorbed in Bed Material and Cyclone Ash
52
-------
Although the prime chlorine scrubbing action of dolomitic limestone is
probably due to its calcium carbonate content, the relative efficacy of
limestone versus dolomitic limestone was not compared. The role of metal
oxides and calcium oxides in the coal ash in chlorine scrubbing is appar-
ently significant but not as yet understood. The effectiveness of HC1
removal if high sulfur coal is used should also be investigated. The paper
contained no other emission data (35).
PVC Waste in a Tilting Fluidized Bed—
Kamino et al. (36) used a pyrolytic bench-scale process in an attempt
to convert a plastic mixture containing PVC into fuel gas. To prevent
formation of harmful gases and lower corrosion of the processing equipment,
the waste was first dechlorinated in a standard fluidized bed with sand as
the fluidizing medium and nitrogen as the fluidizing gas. The bed was
maintained at 220-380°C (the temperature range in which PVC gives off HC1).
As time elapsed, the plastic material cohesed with the sand, and the bed
lost fluidity.
To eliminate this problem, the standard fluidized bed was replaced
with an inclined fluidized bed (Figure 10). In the inclined bed, accumu-
lation of material would be prevented by the outward movement of sand and
plastic materialo Tilting the bed does not require additional mechanical
operating parts*
The inclined fluidized bed was made up of a liquidized gas flow area,
a dispersal plate, and a reaction area. The fluidizing medium was sand
with an average granular diameter of 1.2 mm. Propane gas functioned dually
as the fluidizing gas and primary combustion source.
Sand, from the sand storage area, passed through a hot blast stove,
became heated, and joined the plastic waste before they both entered the
gas flew area of the combustor. Dechlorination was essentially completed
in the reaction area. Residual chlorine gas from the process passed to an
alkali tank for neutralization and venting to the atmosphere (36). Many
engineering details were not supplied in this article.
Chlorinated Hydrocarbons/Pilot Plant Studies
Copper Wire Insulated with PVC—
A patent for a continuous fluidized bed process for removing insula-
tion fron copper wire or other materials was assigned to the Cerro Corpor-
ation of New York, New York. The process is particularly suitable when
compared to stripping or burning. Stripping is unsatisfactory for fine
wire, while burning oxidizes copper (37). A schematic of the combustor is
shown in Figure 11.
The polyvinyl chloride insulated wire was pre-chopped in a shearing
machine from 681 kg bales into 15 cm pieces and fed via conveyer through
a water seal into a decomposition chamber. The decomposition chamber was
circular in cross-section and divided by a hearth into an upper and lower
53
-------
TO ALKALI
U1
SAND
STORAGE
AREA
WASTE
PROl'ANE GAS CYLINDERS
Figure 10. Tilting Fluidized Bed Corribustor
-------
15
12
.Figure 11. Schematic of a Continuous Fluidized Bed Process for
Removing Insulation from Copper Wire. (See legend below for details)
1. Feed wire conveyer 11.
2. Water seal 12.
3= Decomposition chamber 13.
4. Advancing mechanism 14.
5. Hearth 15.
6. Pyrolysis section 16.
7. Decomposition section
8. Fluidizing air fan ending in
burners in pyrolysis and 17.
decomposition sections 18.
9. Pilot burner
10o Collection area for calcium
chloride and calcium carbonate 19.
discharged from bed
Feeder for bed material
Afterburners
Afterburner chamber
Air intake from afterburner
Gas discharge from scrubber
HC1 scrubber-dust from the
scrubber passes to the quenching
tank and then to 10 (not shown)
Quenching tank
Clean wire comes out, calcium
chloride and calcium carbonate
transported to 10 (not shown)
Induced draft fan
55
-------
section. The lower section of the decomposition chamber was divided into a
pyrolysis section and a decomposition section (Figure 11). The hearth had
many openings so that the upper section of the decomposition chamber was
connected with the two lower sections.
A burner in the pyrolysis section of the lower chamber maintained a
310° C operating temperature with a reducing atmosphere. A burner in the
decomposition section of the lower chamber maintained temperatures in the
640°C range with adequate oxygen for complete combustion.
The fluidizing medium, calcium carbonate (10-20 mesh) was transferred
from a storage area to the decomposition section of the lower chamber and
was heated to about 310-640°C. Fluidizing air was supplied by a fan to the
two burners in the lower chamber. Heated calcium carbonate was fluidized
by discharge from the burners.
Cut feed entered the upper chamber over the pyrolysis section and was
pulled over the hearth by an advancing mechanism. The waste was subjected
to pyrolysis and partially decomposed.
The advancing mechanism consisted of a rake assembly and a ramp
assembly. Both the rake assembly and ramp assembly are movable indepen-
dently or in unison with the advancing mechanism. A ramp assembly comprises
a base member with cam blocks positioned in a track mounted at the sides of
the furnace. The rake assembly consisted of parallel side bars connected
by spaced cross rake members that had downward projecting tines (37).
As feed was moved along the hearth by ramp and rake, it was sur-
rounded by heated particles of calcium carbonate and decomposition was
completed. Some of the calcium reacted with the chlorine or hydrochloric
acid produced by waste decomposition. A pilot burner located in the upper
chamber provided mixing and safety ignition to prevent a build-up of a
combustible atmosphere.
The clean copper wire left the decomposition chamber through an exit
water seal which acted as a quench for the wire scrap. The wire was
collected in a receptacle for further handling. The calcium salts were
trapped in a weir and discharged into a receptacle (Figure 11). The
product collected from the receptacle usually contained about 12% calcium
chloride and 88% calcium carbonate. Calcium chloride is a useful by-
product of the process, that can be used for snow removal and road
building.
Because gases and smoke generated during decomposition in the chamber
may contain small amounts of chlorine and carbonaceous material, the gases
passed into an afterburner where all of the remaining combustible products
were consumed. The afterburner, which operated between 760-1090°C, in-
cluded a burner and an excess air inlet (Figure 11).
Since the burned gases from the afterburner may still contain chlorine
as well as the products of combustion, these gases were passed through an
HC1 scrubber. The scrubber was a conventional bubble-plate type, with the
56
-------
water on the plates containing calcium carbonate so that any chlorine in
the gas was reacted to form calcium chloride. The scrubber also removed
any dust such as calcium carbonate. The gases then passed from the top of
the scrubber into an induced draft fan that provided the pressure differen-
tial to cause gas flow through the system (Figure 11). The system was
designed to process 1362 kg/h of material (37).
Chlorinated Hydrocarbon Waste with High Chlorine Content
Pilot Plant Trial—
Chlorinated hydrocarbon impurities can occur as waste in the manufac-
ture of plastics, herbicides, solvents, and paints. A chlorinated hydro-
carbon that contains as much as 70% chlorine (calorie value, 2,500-3,000
kcal/kg) can support combustion in a high performance incinerator. When
the chlorine content is increased above 70%, additional fuel must be
supplied to bring the total calorific value of the waste up to about 2,500
kcalAg to maintain combustion. A pilot plant trial was done to see if a
chlorinated hydrocarbon waste containing 80% chlorine could support combus-
tion without additional expensive fuel in a fluidized bed combustor (38).
The reactor had an 460-mm internal diameter with a bottom section
tapering to a 300-mm diameter at the distributor, and was constructed of
refractory material cast into a mild-steel shell. The Type 310 distributor
had four 64-mm diameter bubble caps, each having six 9.5-mm diameter
radially drilled, equally spaced holes.
The bed consisted of 100-150 kg of 6-mm building sand that was fed to
the reactor from a pressure-sealed variable-speed table feeder. Compressed
air (1 m/sec) was the fluidizing medium. After the windbox was fired and
the bed raised to the desired operating temperature, a metered flow of
waste was injected (with atomizing air) directly into the center of the
fluidized bed at a point 35 mm above the distributor. The Type 310 stain-
less steel injection nozzle was 6 mm in diameter and flatted at the end to
form a fan-tail spray. When auxiliary fuel was needed (in one test run)
distilled oil was injected in the same nozzle. The flow rate of the waste
was adjusted arbitrarily to give an off-gas oxygen content of about 4%.
Off-gases from the reactor were passed through a 230 mm-diameter high-
efficiency cyclone and then through two packed absorption columns where
they were scrubbed countercurrently with a soda ash solution to remove both
hydrogen chloride and chlorine. The absorption columns were packed with
ceramic Raschig rings. After scrubbing, the gases were monitored for
oxygen content and analysed using Drager tubes for chloride content. They
were then vented to the atmosphere. Samples from the scrubbers were also
analysed for chloride content. Completeness of combustion and hydrogen
chloride formed were determined by calculating a chlorine mass balance over
the system.
Only 85% of the chlorine content of the waste liquor was recovered in
the 1000°C run. This indicates that complete combustion was not achieved.
Fuel oil was added to raise bed temperatures to 1,100°C. Almost all (99%)
of the estimated chlorine, evolved as HCl and Cl2 was recovered in the
57
-------
1,100°C run. As the fuel oil was not added directly with the waste, the
run was considered autogenous.
An alternative to scrubbing the exhaust gases before they are vented
to the atmosphere is water-absorption of hydrogen chloride to produce acid
for recycle. Maximum formation of hydrogen chloride requires combustion at
temperatures around 1,500°C and steam injection. In this case, the use of
a fluidized bed would not offer many advantages over a standard high-
performance burner (38). Specific chlorinated compounds were not cited
in this paper.
Chlorinated Hydrocarbons/Bench to Pilot Scale-up
PVC Waste Generated at Rocky Flats Plant of Dow Chemical/teench Scale—
The Rocky Flats Plant (RFP) of Dow Chemical used stationary grate
incineration (800-1000°C) to convert bulky combustible residues into a
suitable form for recovering plutonium by aqueous chemistry. Corrosion of
the incinerator and off-gas scrubbing system and damage to the incinerator
refractory lining caused frequent shutdowns for maintenance and repair.
The incineration system also produced plutonium oxide which was difficult
and expensive to recover. Since PVC accounts for about 40% of the RFP
waste mixture, hydrogen chloride (HC1) generated in the combustion of poly-
vinyl chloride (PVC) plastics was considered to be the major source of
corrosive gas. To eliminate these problems, fluidized bed combustion with
a sodium carbonate bed was used for in situ neutralization of HC1 generated
by the PVC as it decomposed (39).
Feasibility studies were carried out in a 6.3-cm diameter, 81.3-cm
long quartz tube. Materials subject to HC1 corrosion were eliminated.
All vessels and gas transfer lines were constructed of HC1 corrosion resis-
tant quartz, Pyrex, or polyethylene. Several equipment arrangements were
used to promote greater efficiency' An early arrangement utilized a static
.bed of Al^>3 balls for gas preheating with a 15-cm diameter solids disen-
gagement section at the top of the reaction tube to separate solids from
the off-gas stream. The fluidized bed was composed of alumina granules
(Figure 12).
In a later design, the solids disengagement section was replaced by a
Pyrex cyclone to improve the separation of solids from the off-gas stream
(Figure 13). A fluidization gas preheater, two stage caustic scrubbers,
and a fixed sodium carbonate bed for HC1 neutralization, plus a catalytic
afterburner were also added. Catalytic afterburning at 500°C offered the
potential for an incineration system which did not need refractory lined
equipment. The types of catalysts used will be discussed later. All gas
transfer lines were constructed of Pyrex and were connected to the vessels
by standard tapered glass fitting. The gas preheater was a 3,000 watt
Chromalox commercial heater with piping to permit the use of air, argon,
and oxygen gas mixtures for bed fluidization. The caustic scrubbing system
consisted of two, 2,000-ml glass cylinders partially filled with 1.3-on
polyethylene Raschig rings and 1% sodium hydroxide solution (NaOH).
Alumina balls were added as a roughing filter to eliminate packing of fine
58
-------
To Vac
•—v
Caustic Scruoeer
Haste Pellets
Air
Solids
Disengagement
Section
Alumina Granules
Fluicized Bed
Alumina Balls
Static Bed
Gas Preheating
Figure 12. Dow Chemical Pocky Flats Laboratory-Scale
Conbustor.
59
-------
TO VAC
'
GAS rHZKEATSH
Figure 13. Dow Chemical Rocky Flats Laboratory Scale
Corcustor: Intermediate Design.
60
-------
TO VAC
«
— -
- - \
1
\
~m \
/
[
1
k
~ ~" >
^
f
V
CYCLCIf
SEFAHA
WAS
riLL
FLUI
E2D
OXIDA-
TIOM
CATALYST
SSCOSDABY
CAUSTIC
SCHU3S23S
PRTKAHY
CAUSTIC SC3U3BSR
AIB
ARGCH
Figvire 14. Dew Chemical Rocky Flats Laboratory-Scale
Ccrnbustcr: Advanced Design.
61
-------
particles in the catalytic burner. Most incineration tests were made in
the unit shown in Figure 13. The fluidized bed was sodium carbonate.
The final design requirements are shown in Figure 14. A new quartz
incinerator vessel with 1000-mesh gas distribution screen improved bed
fluidizatiori. The oxidation catalyst bed was enclosed in a furnace for
improved temperature control and off-gas combustion efficiency. Another
caustic scrubber vessel was added and off-gas piping was modified to
create a primary and secondary scrubbing system to prevent loss of HC1 from
the system. The design was utilized for the last five incineration studies.
This fluidized bed was conposed of sodium carbonate (39).
Waste was added as pellets to assure intimate mixing with Na-CO, and
ease of feeding. The waste materials, PVC, polyethylene, paper, and
surgeons' gloves, were manually cut into pieces no larger than 1.27 cm. ..
They were then weighed and mixed into the correct proportions for a stan-
dardized pellet. The mixture was then pelletized by a batch extrusion that
held the materials together in a pellet approximately 6.3 cm long and
1.2 cm in diameter.
Twenty grams of pellets were placed in the bottom of a 200-ml crucible,
covered with powdered Na2CO3, and heated for three hours in a 600°C muffle
furnace. The residue in the crucible was dissolved in hot water and the
solution analyzed for chloride content. The average chloride content was
21.85 wt%.
Decomposition of waste pellets in the combustor was the major consid-
eration in early bench-scale tests, and neither sodium carbonate for HC1
neutralization nor catalytic afterburning were used. Air was metered
through a bed of heated 0.7 cm-diameter alumina balls for preheating and gas
distribution. The heated air fluidized an alumina granule bed. When the
bed temperature reached 600-650°C, waste pellets were dropped in the top of
the disengagement section for decomposition in the fluidized bed. Decom-
position tars and soot were retained in the dry trap and caustic scrubber.
A propane-oxygen flame was introduced above the fluidized bed in an attempt
to burn the waste decomposition gases. Data from the early combustion
tests were not given, but the fluidized bed combustion technique was con-
sidered feasible for waste from the Rocky Flats plant.
Emphasis was shifted to the goal of complete HC1 neutralization and
clean-up of the off-gas stream with the modified units shown in Figures 13
and 14. Air or argon-oxygen was metered through a Chromalox preheater to
help maintain an operation temperature of 600°C within a fluidized bed of
Na2CO3. Waste pellets and combustion air were introduced through a side
port directly above the fluidized bed. The air and decomposition gases
were drawn through the system by a vacuum applied to the scrubbers. De-
composition gases were passed through a static bed of sodium carbonate in
order to react any KC1 gas that might escape from the fluidized Na2CC>3
bed. The flow continued into an oxidation catalyst bed to promote com-
bustion of the waste decomposition gas. Air drawn in the side inlet
provided additional oxygen for combustion within the catalyst bed. When
the cyclone separator proved relatively successful in removing the small
62
-------
amounts of solids that escaped in the incinerator, the use of dry trap was
discontinued (39).
Off-gas was drawn through the cyclone into the scrubbing system. In
the final scrubber design, all of the off-gas passed through a dip tube to
the bottom of the scrubbing vessel. After a thorough mixing with caustic
solution in the Raschig ring-filled column, the stream was split as it left
the vessel and entered the house vacuum system (Figure 14).
Off-gases were sampled by attaching an evacuated 10-cm gas cell to the
off-gas line and drawing the sample into the cell. The gases contained in
the cell were then analyzed by infrared spectrophotometry. No chlorinated
hydrocarbons were detected in the off-gases. Sodium carbonate probably
neutralized HC1 by the following reaction: (39)
Na9CO, + 2 HC1 — ^ 2 NaCl + H_0 + CO.,
& •$ £
In attempts to increase the degree of neutralization of sodium carbonate
on HCl, three different methods of neutralization were evaluated:
a pelletization of sodium carbonate with combustible PVC wastes
to insure intimate contact during decomposition
« decomposition of waste PVC pellets in a fluidized bed of
sodiun carbonate granules (the method used in the initial
studies)
o decomposition of PVC waste pellets in a fluidized bed composed
of unreactive material (such as sand), and passing the gases
produced in the reaction through a static sodium carbonate bed
Each method offered advantages, but each proved to.be only partially
successful when used alone. The degree of neutralization that was ob-
tained by each method was measured by the amount of chloride retained in
the fluidized bed after combustion. As seen from the data in Tables 4, 5,
and 6, each method resulted in bed retentions of approximately 45-55%
chloride.
When it became obvious that no single method could achieve 100% HCl
neutralization, combinations of all three methods were evaluated. This
resulted in improved chloride retention in the bed (Tables 7 and 8). The
best results, 97.2% chloride retention, were seen when all three methods
were combined (Table 9) (39).
Combustion efficiency of a Na^ZO-j fluidized bed was expected to de-
crease as the bed NaCl concentration increased. One run was stopped to
analyze the Na £0 3 beds, then continued to evaluate extended run efficiency.
As expected, the percent chloride retained in the lower bed decreased
approximately 6% during the second half of the run. When the run was stop-
ped, the concentration of NaCl in the lower bed was 22%. The upper bed
retained more chloride in the second half of the run. An upper, static bed
is necessary for the best possible chloride retention in extended runs (39).
Waste pellet feed rates were examined as another variable that might
affect HCl neutralization. The incinerators had no refinements for cooling
63
-------
TABLE; 4
HCl NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BISD-ALUMINUM OXIDE
STATIC UPPER BED-NONE
WASTE ADDED AS PVC-SCDIUM CARBONATE PELLETS
CHLORIDE RETENTION (wt%)
Run No.
25
Total
Recovery
88.0
Lower
Bed
46.6
Upper
Bed
None
Used
Catalyst
Bed
24.8
Scrubber
16.6
Lost
12.0
TABLE 5
HCl NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BED-SODIUM CARBONATE
STATIC UPP13R BED—NONE
WASTE ADDED AS PVC PELLETS
CHLORIDE RETENTION (wt%)
Run No.
14
Total
Recovery
60.6
Lower
Bed
49.4
Upper
Bed
None
Used
Catalyst
Bed Scrubber
No data 11.2
Lost
39.4
TABLE 6
HCl NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BED-ALUMINUM OXIDE
STATIC UPPER BED-SODIUM CARBONATE
WASTE ADDED AS PVC PELLETS
CHLORIDE RETENTION (wt%)
Run No.
18
Total
Recovery
73.8
Lower
Bed
Upper
Bed
55.4
Catalyst
Bed
No Data
Scrubber Lost
18.4
26.2
64
-------
TABLE 7
HC1 NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BED-SODIUM CARBONATE
STATIC UPPER BED-SODIUM CARBONATE
WASTE ADDED AS PVC PELLETS
CHLORIDE RETENTION (wt%)
Run No.
Total
Recovery
Lower
Bed
Upper
Bed
Catalyst
Bed
Scrubber
Lost
20
21
22
Averaae
92.0
90 08
90.5
91,1
84.3
74.3
76.6
78.4
7.1
15.3
12.6
11.6
No data
No data
No data
0.6
1.2
1.3
1.0
8.0
9.2
9.5
8.9
TABLE 8
HC1 NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BED-SODIUM CARBONATE
STATIC UPPER BED-SODIUM CARBONATE
WASTE ADDED AS PVC PELLETS
(% SODIUM CARBONATE UTILIZATION)
CHLORIDE RETENTION (wt%)
Run
No,
24a
24b
Sodium Carbonate
Utilization
0-13
13.0-22.2
Total
Recovery
96.9
95.5
Lower
Bed
72.8
66.6
Upper
Bed
20.2
23.1
Catalyst
Bed
2.7
1.5
Scrubber
1.2
4.3
Lost
3.1
4.5
TABLE 9
HC1 NEUTRALIZATION POTENTIAL
FLUIDIZED LOWER BED-SODIUM CARBONATE
STATIC UPPER BED-SODIUM CARBONATE
WASTE ADDED AS PVC-SODIUM CARBONATE PELLETS
CHLORIDE RETENTION (wt%)
Run No.
Total
Recovery
Lower
Bed
Upper
Bed
Catalyst
Bed
Scrubber
Lost
19
26
27
28
29
Averace
97o2
95.3
9G..6
94.8
91.8
95.1
90.9
88.7
91.6
87.2
87.8
89.2
3.0
No Data
0.4
1.0
7.0
0.3
2.2
0.5
0.3
2.2
0.4
2.3
1.1
2.8
4.7
3.4
5.2
8.2
4.9
65
-------
the upper bed, and the temperature increased in the upper bed when hydro-
carbons in the flue gas combusted. Feeding rates were restricted when the
upper bed temperature exceeded 800°C. Normal feed rates below 800°C were
from 1.36 to 2.71 g/min. There was no obvious relationship between feed
rates and HC1 neutralization when the temperature of the catalyst bed was
below 800°C.
During the course of the investigation, the quantity of HC1 predicted
by calculation was consistently not recovered by rigorous analytical pro-
cedures. The possibility that HC1 was lost during the pelletization process
was ruled out when a test run with unpelletized PVC also resulted in incom-
plete recovery. The cause of this problem was not detected. The possibil-
ity that chloride was lost through chlorinated hydrocarbon aerosol formation
was considered but not investigated.
These studies all used partial pyrolysis conditions rather than true
combustion. Separate runs were made with air, argon, and argon-oxygen as
the fluidizing medium. Residues in the incinerator were analysed for ash
content after each run. The average bed ash retentions were 9.5% for
argon, 5.3% for the argon-oxygen mixture, and 3.9% for air. High ash
retention in the argon-oxygen fluidized bed was accomplished before the
gases reached the catalyst section of the unit. High ash retention
indicated more efficient destruction in the bed with reduced need for above
bed combustion and scrubbing. Limited oxygen also reduced refractory oxide
formation in the bed and limited above bed burning. Optimum oxygen concen-
tration appears to be between 10-15%, but the authors indicated a need for
more studies on gas ratios. There was no difference in bed fluidization
with any gas or gas mixture.
The need for open-flame afterburning was eliminated by adding an oxi-
dation catalyst to the unit. Open-flame after-burners of incinerator flue
gas require 1,000-2,000°C for complete combustion of trace hydrocarbons.
The minimum temperature required for the catalytic after-burner is about
400 to 600°C. The lower temperature allows a wider variety of materials
which would be satisfactory for construction of the after-burner. Refrac-
tory lined equipment and costly re-bricking are eliminated. The oxidation
catalysts will completely combust hydrocarbons at reduced oxygen concentra-
tions. As there is no need for a large volume of oxygen in the after-
burner, there is a decrease in the total volume of flue gas cleaned in the
HEPA (high efficiency particulate air) filter (39).
Six oxidation catalysts were tested for use in the incinerator after-
burner. They were:
1. Shell 105X-Shell Chemical Company, 0.25 cm-diameter pellets,
Fe2°3 wi^ 21% potassium as a promoter
2. Zeolon 200H-Union Carbide, Norton Chemical Process Division
0.12 cm-diameter pellets, 10% silicone oxide on a sodium (1%)-
alumina support
3. ZR-0304T-Harshaw Chemical Company, 0.25 cm-diameter pellets,
zirconium oxide (98%) mixed with alumina (2%)
66
-------
4. CR-0211-Harshaw Chemical Company, 0.25 cm-diameter pellets,
chromium oxide on alumina
5. Linde 13X molecular sieve-Union Carbide Company, Norton
Chemical Company, Norton Chemical Process Division, 0.12
cm-diameter pellets (as shipped, H_0 1.5 wt%)
6, Grace CAT 908-Davison Chemical Company, 0.7 cm-diameter balls,
copper and magnesium oxides with alumina base
An oxidation catalyst used in place of incinerator after-burner in
this fluidized bed system should exhibit good resistance to HC1 attack and
oxidation efficiency at low temperatures. To aid in selecting a suitable
catalyst, a series of tests were made. The catalysts were contacted with a
humid HC1 gas stream at 500°C for two hours. Oxidation efficiencies were
evaluated with a methane-air mixture.
Three of the catalysts, Harshaw CR-0211T, Linde 13X, and Grace CAT
908, exhibited almost 100% efficiencies from the beginning. The Linde 13X
molecular sieve gained weight during HC1 treatment, indicating HC1 adsorp-
tion. It lost weight during oxidation tests, indicating HC1 desorption.
The Grace CAT 908 ha a large weight loss during oxidation tests without a
similar weight gain during HC1 treatment. This indicates that part of the
catalyst was driven off as a chloride compound. The other catalysts exhib-
ited a gradual increase in efficiency as a function of time. They were
attacked by HC1 but regenerated to some extent because of the oxidation
conditionso The Grace CAT 908 and Harshaw Cft-0211 exhibited almost 100%
efficiency at 400°C. Linde 13X and Harshaw ZR-0304T required 500°C, and
Norton Zeolon 200H and Shell 105X required 699°C for 100% efficiency. Of
all the catalysts tested, Harshaw CR-0211 T (chromium on alumina) exhibited
the best combination of resistance to HC1 attack, oxidation efficiency, and
temperature requirements.
Pilot-Plant Combustor at Dow Chemical—
Data from the bench-scale tests were used to design a fluidized bed
pilot plant incinerator for evaluating process variables and obtaining
process design information. A low-speed, cutter type shredder, unit
capacity 10 m/hr, shredded material for incineration. Waste was shredded
and mixed to provide a known composition of 44% PVC, 28% paper, and 28%
polyethylene (40).
The pilot-plant incinerator was designed to combust waste at 4.5 kg/hr.
A flow diagram is shown in Figure 15. The unit can operate with a single
fluidized bed or with two fluidized beds in series. The lower (primary)
bed diameter is 35.6 cm. The upper (secondary) bed diameter is 40.7 cm.
The vessel diameter above the second bed increases to 61 cm and results in
a gas velocity reduction and return of some entrained particles to the bed.
Gas exits through a 10.2 cm diameter cyclone that removes more entrained
solids. An overflow tube and air-jet ejector convey upper bed material to
the lower bed.
Shredded waste was fed from a hopper by a chain conveyor that regu-
lated waste feed rate. Waste then passed through a constant-pitch,
67
-------
To final
filtration .,
and v~
atmosphere
Secondary
jector
Boom
Air
Catalyic
Afterburner
Primary
Ejector
Beam
Air
Sintered
tfetal
Filters
Filter Dust
Waste Feed'
Cyclone
Figure 15. Flow Diagram for Pilot Plant Fluidized Bed Incinerator
68
-------
tapared-screw conveyor that introduced feed to the lower bed under the
surface of the fluidized bed material. A third tapered screw conveyer
fed bed material to the upper bed.
The bed was fluidized by either 100% nitrogen (pyrolysis), 100% air,
or mixtures of air and nitrogen. Fluidizing gas passed through an electric
heater until bed material reached 300° C. Waste was then used to bring the
combustor to 500 "C.
The filter system consisted of seven, sintered-metal filter tubes,
2 cm in diameter, in each of five separate filter holders. Each filter
holder was piped in parallel so that the exit flow from one holder could
be stopped and the flow reversed to remove collected dust while flue gas
continued flowing through the other filter tubes.
Air-ejection downstream of the filters provided the motive force for
gas flow through the combustor and filter systems. The ejector was ad-
justed to provide atmospheric (or slightly lower) pressure in the reactor
at the point where waste was introduced into the lower chamber. This . .
insured minimal air leakage as" feed was added. The air- jet ejector also
provided oxygen needed for combustion in the afterburner.
Flue gas passed up from the ejector through a packed bed catalytic after-
burner,, The afterburner was 61.0 cm in diameter and 91.5 on long. Additional
combustor and cooling room air was pulled in by the negative pressure in the
unit and mixed with the process gas stream at three levels in the after-
burner o
Flue gas goes through one stage of high efficiency particulate air
(HEPA) filtration before leaving the incinerator room. It then passes
through four stages of HEPA filtration before venting to the atmosphere.
As in the laboratory scale tests, a fluidized bed of NaoCOj was used
to neutralize HC1 at its point of generation in order to avoid corrosion
and eliminate the need for flue gas scrubbing. Tests were conducted
with continuous waste feeding to a single N^COhj bed. Fresh Na^CC^ was
continuously fed to the bed and bed material was continuously discharged.
Almost 100% efficiency of HC1 reaction was obtained up to about 26%
utilization. Chloride reaction efficiency decreases and significant
quantities of HC1 are released after this point (40).
These data were generated with a range of 500-1000 micrometer
bed material. X-ray microscopy of particles from used beds indicates that
a highly concentrated, dense shell of NaCl surrounded an unreacted Na2CC-3
core. This indicates that smaller particles would probably improve the
neutralization efficiency at higher N^CC^utilization levels. It also
suggests abrasion of the NaCl shell, exposing the N32C03 core, might also
improve N^CO, utilization.
Neutralization was not affected by the fluidized bed depth. Data
from deep and shallow bed experiments indicated that HC1 was generated and
reacted soon after waste is introduced. This was supported by the fact
69
-------
that the use of a secondary fluidized bed did not improve neutralization
efficiency at comparable sodium carbonate utilization levels.
Some catalyst is elutriated from the reactor and from the fluidized
bed catalytic afterburner. The rate of loss increases with increased
fluidizing air velocity and with small catalyst particles.
Nitrogen was usually mixed with air in the fluidizing medium to
prevent overheating. Overheating can cause above bed burning, melt
entrained bed material and block the distributor plate. Since additional
air is required in the catalyst area to complete combustion of flue gas
vapors, the addition of air directly into the catalyst should eliminate
this problem (40).
Disposal of Munitions by ARRADCOM/feench Scale to Pilot Plant Tests
Bench-Scale Tests—
The U.S. Army Armament Research and Development Command (ARRADCOM),
Dover, New Jersey, conducted bench-scale tests on the feasibility of
fluidized bed combustion of propellant and explosives in a fluidized bed
combustor.
The system selected for investigative studies was 0.15 m in diameter
and 2.74 m (9 ft) high. It was designed to accept a solid/water slurry
feed and had a dry explosive feed rate of 3.18 kg/hr. The bed was sized
so that it could be fluidized with approximately 50% of the anticipated
requirement of 120% stoichiometric air. This improved the flexibility of
the incinerator as it allowed the system to operate in either a one or two
stage combustion mode, i.e., all the air could be fed into the bottom part
of the bed, or part of the air could be fed into the bottom part of the bed,
with the other part of the air fed into the upper portion of the bed. Alumina
was used as the bed material. The system also included a slurry feed system,
cyclone particulate collector, and stack gas analyzer. Major particulates in
the cyclone were alumina fines. The slurry feed system had a mix/feed tank
with a large recirculating line. The incinerator feed was tapped from this
line and fed into the incinerator through a metering pump (41).
In a series of 37 test runs, made in both the one-stage and two-stage
mode for up to 6 hours, the incinerator operated effectively in disposing of
the explosives and propellants. However, emission levels of 840 ppm-NCL,
650 ppm-CO, and 350 ppm-HC (hydrocarbons), were above the 200 ppm emissions
goal for each of these pollutants. The emissions were also approximately
equal to untreated emissions from previous combustion studies done in rotary
kiln and vertical incinerators.
In another series of tests, the addition of 6% by weight of nickel oxide
catalyst to the bed caused a reduced emissions to 57 ppm NC^, 40 ppm-CO,
and 10 ppm HC. The results of this program led to the decision to convert
an ARRADCOM vertical incinerator to a fluidized bed incinerator for pilot
plant testing (41).
70
-------
Preliminary Pilot-Plant Tests—
Before the pilot-plant ARRADCOM fluidized bed combustion unit began
operation, laboratory test runs at Exxon Research Corporation established
combustion parameters so that the pilot plant could be operated in an
efficient and ecologically sound manner. Optimum combustion occurred if
the fluidized bed (with catalyst) was operated with a reducing atmosphere
in the lower bed and an oxidizing atmosphere (via secondary air) in the
upper bed. The enriched oxygen of the bed, combined with the mixing of
the alumina and waste, promotes efficient combustion, thereby minimizing
hydrocarbon and CO emissions. Use of a nickel catalyst plus reducing con-
ditions in the lower bed minimize NOX emissions. The nickel catalyst also
promotes the reduction in levels of gaseous pollutants such as CO, and HC.
Use of supplemental oil injection assists in maintaining a constant bed
temperature and provides a reducing atmosphere in th lower bed. NO is
formed based on parameters such as combustion temperature, reaction^rate,
residence time, concentrations of nitrogen and oxygen, and quench rate.
An equivalence ratio was calculated for the first combustion zone
(<}>^), and for the overall process (^o)- Tne equivalence ratio is calcu-
lated by ccmparing the actual fuel/air (F/A) ratio to the theoretical ratio
for stoichiometric conditions. If the equivalence ratio is 1, the reaction
is stoichiometriCo Less than 1 equals oxidizing conditions, greater than 1
represents reducing conditions (41).
The fuel/air ratios used, govern operating temperatures required for
the material burned. Tests showed that the optimum heating temperature for
TNT is 900°C in the slurry ingestion zone. When the system is stabilized
at this temperature, the gaseous emissions (especially NOX) are at their
lowest values. For Composition B, the best temperature is 1,038°C (41).
The principle pollutant emissions from combustion of TNT and Compo-
sition B were characterized :
o Sulfur oxide emissions, from combustion of fuel oil
that contains sulfur, produce significant quantities of
sulfur oxides.
o Nitrogen oxides are formed by thermal fixation of
atmospheric nitrogen in high temperature processes
and from nitrogen compounds in the waste. Because of
the relatively high fluidized bed temperatures, only
nitrous oxide is formed.
o The amount of CO emitted indicates the efficiency of the
combustion process. An efficient combustion operation
is associated with high CO2 and reduced CO levels.
o Particulate emissions from the fluidized bed incinerator
can be either dust-solid particles or smoke-solid particles.
The dust-solid particles are composed primarily of bed
material and catalyst fines entrained in the gas stream.
Smoke-solid particles are formed as a result of incomplete
combustion of carbonaceous materials. Their diameters range
from 0.05-1 micron (41).
71
-------
Pilot Plant Tests at ARSADCOM--
The pilot plant fluidized bed incinerator was 2.4 in diameter and
9 m tall. The corabustor shell was made from 15-on, Schedule 40, RA-330 high
temperature alloy pipe. RA-330 is austenitic, non-hardenable, and strong
at high temperatures. It is also resistant to oxidation, corrosion, and
carburization. Nominal composition was 19-35-43-1.5-1.25 (Cr-Ni-Fe-Mn-Si).
The slurry preparation and feed system can mix and pump various explosive
slurries. The entire system was remotely operated and monitored. Slurry
was mixed in two, 1.6 m tanks. The tanks were loaded with water and mixers
were started. After the water was in motion, the explosive material was
added to the tank and mixed with the water. The mixers were operated -with
a 735 W pneumatic motor. They were adjusted for each individual slurry mix
and for liquid level changes. Because of their different compositions,-
weight percentages, and particle configurations, the densities of the
slurries varied and presented different mixing parameters.
The slurry pump was centrifugal and capable of pumping a 18.3 m head
with a 0.0079 mVsec water flow. Slurry flow through a 0.064 m header pipe
was adjusted from the control room (41).
Bed temperatures were carefully monitored to determine where the fuel
oil and/or slurry were combusting. Close control was required to prevent
combustion from occurring above the bed or in the exhaust ducts. Six slurry
injection nozzles were alternated with oil nozzles around the chamber's
periphery. Injection of oil and slurry into the bed was controlled by
individual "approval switches" set at predetermined combustion temperatures
in the bed. This prevents the oil and slurry from being fed into the bed
until combustion temperatures are attained. Temperature controls also shut
down the system if maximum set temperatures were exceeded. The system
proved to be an excellent heat sink. Heat retained after a weekend shutdown
was sufficient for a Monday morning startup without use of the preheater.
. . . Pressure was monitored in the plenum, grid, bed, and upper chamber.
Plenum and grid pressures indicated degree of fluidization. Grid nozzles
could also be checked for clogging. Pressure transmitters could detect
pressure buildup in the system and possible detonation. Monitoring of
slurry flow was achieved by utilizing pressure transmitters in the header
and slurry lines (41).
Alternate Fuel Studies—
Because large incinerators would have rather high pre-heat fuel costs
for liquid fuels low in sulfur and nitrogen, in addition to expensive
gaseous fuels, studies were made regarding high sulfur and high nitrogen
fuels. A 30 wt% nickel catalyst was used. Major results indicated satis-
factory operation of a catalytic fluidized bed with both gaseous and low
sulfur fuels. Initial tests revealed high sulfur fuels poisoned the nickel
catalyst with subsequent high NCL emissions, although stable combustion
was obtained. Further investigation of various parameter3 is necessary to
fully evaluate the relationship of the nickel catalyst with high sulfur
fuels (41).
72
-------
Munitions/Bench Scale to Pilot Plant Processes
Bench Studies at Picatinny Arsenal—
Under the technical direction of the Manufacturing Technology Director-
ate, Picatinny Arsenal, Dover, New Jersey, several incinerator designs were
evaluated to develop a reliable, safe method for waste explosive material
disposal. The fluidized bed incinerator design was judged to be one of the
more premising systems and was selected for further testing (42).
A laboratory scale incinerator was constructed with a 15.42-cm bed
diameter and 3-cm height. Bed material was aluminum oxide granules
(particle size, 500 y ). Waste explosive trinitrotoluene (particle size,
0.318 cm) was introduced to the incinerator bed as a 10 wt% TNT/water
slurry from a mixing tank through a recirculating line. The incinerator
bed material was heated to 8710C-1093°C. The method and rate of air feed
were not indicated, but bed material took on all the properties of a fluid and
provided violent agitation and turbulence. A dry cyclone was placed in the
emission gas stream to remove particulate matter.
During the course of the lab testing, it was observed that a reducing
atmosphere could be created in the bed through use of two stage combustion
and introduction of a nickel oxide catalyst to the bed. This accelerated
the following reaction:
2 NO -H 2 CO —> 2 C02 + N2
Large (unspecified) reductions in NQx emissions were observed in addi-
tion to reductions in CO and hydrocarbons. It was reported that this
eliminated the need for a wet scrubber system and the associated liquid
disposal problem.
Pilot-Plant Combustion at Picatinny Arsenal—
A vertical induced draft incinerator facility, located at Picatinny
Arsenal, was converted into a fluidized bed pilot-plant combustion unit.
It had facilities for slurry preparation and incineration as well as a dry
cyclone separator and a 38-m emission stack. The fluidized bed was
composed of 90% alumina particles and 10% nickel oxide-coated alumina
particles (500 y or smaller). Nickel oxide was used as a catalyst to
reduce NO emissions.
Waste explosive materials were sprayed into the combustor as a water
slurry at a maximum concentration of 25 wt% solids to water. Emission
streams passed through a dry cyclone separator with a selection efficiency
rated for particles 30y and above before exiting through the primary stack.
The system was oil fired and designed to operate at bed temperature ranges
of 871-1093°C.
The Air Pollution Engineering Division (APED), United States Army
Environmental Hygiene Agency (USAEHA), conducted an air pollution
73
-------
assessment program during the initial trial operation of the prototype
facility. The objectives were:
1. Monitor air pollution emissions leading to compliance
with federal, state, and local regulations
2. Test for hazardous air pollutant emissions from the facility.
3. Provide emissions data useful in the evaluation of selected
operating parameters and combustion efficiency (42).
Because of the unique character of this prototype incinerator, only
New Jersey stationary source emission standards governing visible emissions
were found to apply to this facility. An analysis of the prototype incin-
erator combustion cycle was performed to determine potential direct
emissions or by-product formation and emission of hazardous substances.
Several substances introduced into the incinerator were determined to
be potentially hazardous air pollutant emitters. These substances include
bed material, and the waste explosives INT and RDX (cyclotrimethylenetri-
nitramine, (C-^l^SI^Dg). The conditions in the reactor favored the formation
of the toxic compound nickel carbonyl ([Ni(CO)4 ]'). CO reacts with metals
to form a carbonyl when a metal, such as the nickel catalyst in the bed,
presents a high surface area. Because the formation of carbonyl is favored
at low temperature, the greatest potential for its formation is at startup and
shutdown. According to a computer simulation program developed in 1971 by
the National Aeronautics and Space Administration (NASA) designed to identify
the chemical compounds produced by rocket fuel fluidized bed combustion, in-
organic cyanides and hydrogen cyanide may be present as contaminants in
explosives.
The following pollutants were selected for monitoring:
nickel and alumina participate matter HCN
nickel carbonyl NO
inorganic cyanides NC-
TNT CO
RDX
No relevant air pollution emission standards were found to apply to
these substances. In the absence of legal standards, stack emission
limitation criteria were developed by the Air Pollution Engineering Divsion
to serve as a reference for decision making during the initial testing
period. The procedures for the emission guidelines were derived from the
Colorado Air Pollution Commission regulations. They were also based on
meteorological dispersion model estimates for the fluidized bed incinera-
tion stack, assuming worst cast stability, so that ground level concentra-
tions would not be allowed to exceed 1/30 of the Threshold Limit Value
(TLV) established for these substances.
A two week field testing plan was developed which covered a selected
range of slurry compositions and percent solids. Variation in incinerator
operating temperatures, combustion air flow rates, and fuel input rates
were left to the discretion of the operator. All significant operating
parameters were recorded during each test (42)
74
-------
TABLE 10
TOTAL PARTICULATE MATTER AND VISIBLE EMISSIONS DATA
Test
Number
1
2
3
4
5
6
7
Slurry
Composition
5% Conp B
15% Comp B
15% Comp B
25% Comp B
5% TNT
15% TNT
25% TNT
Particulate Emission
Rate Corrected
gm/sm (gr/scf)
1.1245
1.1522
3.5245
0.1483
2.0758
1 .4524
2.7931
(0.4914)
(0.5035)
( 1 .5402)
(0.0648)
(0.9071)
(0.6347)
(1.2206)
Particulate
c
Mass Emissions
kg/hr (Ib/hr)
1 .9450
3.1032
2.0115
0.4455
2.0740
2.2401
2.7647
(4.2880)
(6.8414)
(4.4346)
(0.9822)
(4.5723)
(4.9386)
(6.0952)
Visible
Emissions
% Opacity
4.6
4.8
7.3
5.0
10.0
16.3
11.7
a- Particulate emission rate corrected to 12% CO- with the contribution
of the auxiliary fuel to CO- also discounted.
b- The New Jersey incinerator emission standard for incinerators burning
normal type wastes is 0.229 gm/sm (0.10 gr/scf).
c- The Colorado process weight particulate emission standard, based on the
input of slurry feed, would be 1.05-1.52 kg/hr (2.34-3.36 Ib/hr).
d- Opacity values represent the average of the worst 24 consecutive
15 second observations recorded over the monitoring test period.
TABLE 11
GASEOUS AND VAPOR PHASE HAZARDOUS AIR POLLUTANTS EMISSIONS DATA
Hazardous Pollutants
Monitored
Nickel Carbonyl
TOT Explosive
RDX Explosive
Inorganic Cyanide
Hydrogen Cvanide
Maximum Emission
Rate Observed
(mg/1n3)
0.028*
0.006*
0.010*
3.400
3.540
Emission Limitations
Developed by USAEHA
(mg/m3)
210
890
890
2970
6530
* These values represent the sensitivity of the analytical procedures used.
Source: Carroll, J.W. et al., USAEHA, 1979.
75
-------
Results of the study—Weather conditions were overcast during approx-
imately 50% of the test periods; a condition that hampered evaluation of
the gray-white incinerator stack plume. Visible emissions, evaluated as
percent plume opacity, were generally found to be in conpliance with the
20% limit applicable in New Jersey.
Total munition particulates, expressed in terms of both emission
•concentrations (gm/sm3) and mass emission rates were generally found to be
high (Table 10). Although no particulate emission standards directly apply
to this prototype incinerator, emission levels exceed most of the guide-
lines of the New Jersey standard for incinerators of this size and the
Colorado process weight particulate emission standard. However, the par-
ticulate emissions represent a significant improvement in particulate
matter emission control versus standard disposal via open burning and
detonation of waste munitions. Carroll et al. believe that definition of
regulatory emission standards for particulate matter and other potential
pollutants for this type of disposal process, in lieu of open burning,
should be developed (42).
Monitoring indicated levels for all potentially hazardous vapor and
gaseous phase pollutant constituents were found to be negligible (Table 11).
All test results for nickel carbonyl and vapor phase TNT/RDX were found to
be le^s than the detectable limit (sensitivity) of the analysis procedures.
Trace quantities of total cyanides and hydrogen cyanide were observed on
several tests when lower incinerator combustion temperatures were used.
These levels were several orders of magnitude less than the emission limit
criteria that were developed earlier. Nickel carbonyl was the only
constituent found in significant quantities, yet these levels were at least
tenfold less than emission limits criteria developed by the USAEHA (42).
HAZARDOUS WASTES DESTROYED BY UV/OZONATION
2,3,7,8-Tetrachlorodibenzc-p-Dioxin
A preliminary investigation of the effect of an UV/ozone system on
chlorinated dibenzo-p-dioxins (TCDD) was conducted by California Analytical
Laboratories, Inc. and the Carborundum Company (both in Sacramento, Cali-
fornia). Chlorodioxins are implied carcinogens and extremely toxic (43).
Experimental Apparatus and Methods—
Two types of ozonation systems were used:
(1) Purified oxygen was passed in front of a mercury vapor lanp. The
generated ozone was then bubbled through an aqueous solution of
TCDD (1.0 ppb) in a 100 ml volumetric flask. The flow rate,
150 ml/min and ozone content 0.25 mg/1, yielded an ozone dosage of
about 0.04 mg/min. After ozonation, 1.0 ml of a mixture of
benzene-hexane (1:1) was added to extract any organic compound that
might be present. The extract was analyzed by gas chromatographic
equipment equipped with an electron capture detector.
(2) The second ozone system was an Ozone Research commercial unit.
Generated ozone was passed through a diffuser into the reactor at
76
-------
a rate of 0.5 1/min. Ozone content in the oxygen was 5 mg/1,
and the resulting dosage was 2.5 mg/min. The ozonated solu-
tion was then extracted and analyzed as above (43).
Results-
Figure 16 shows the results of the ozonation of a 1 ppb solution of
TCDD at pH 3.5 and 10.5. Since the ozone dosage was only about 0.44 mg/ndn
for both pH levels, twelve hours was needed for conplete degradation.
The pH did not have a significant effect on breakdown time, and the small
difference in degradation rates was within experimental error. A 10-15%
rate increase was observed when the solutions were both ozonated and
irradiated with UV light (300-400 nm) at the same time.
Figure 17 shows results obtained when a 1 ppb TCDD solution was
ozonated at pH 7 in the commercial unit. The 2.5 mg/min ozone dosage
greatly increased the degradation rate as compared to the previous unit.
There were insufficient data to indicate if the addition of UV to the
process had a significant effect on the rate.
In order to detect chlorinated degradation products from TCDD re-
sulting from ozone breakdown of TCDD, a Finnigan Model 3200 GC-MS system
equipped with an electron capture detector coupled with an Incos computer
system was used. To date, no chlorinated compounds have been detected
(1979). The more sensitive Multiple Ions Detection mode also failed to
detact any chlorinated breakdown products. Detection difficulties could
possibly be attributed to the low levels of starting material. Future work
will use larger amounts of starting material and apply compound labeling
for mass balance studies.
The degradation rate of octachlorodibenzo-p-dioxin (OCDD) by ozona-
tion was also studied. This substance is a common impurity associated with
the wood treatment agent pentachlorophenol (PCP). When a 5 ppb OCDD solu-
tion (pH 7.1) was ozonated in the commercial reactor at an ozone dosage of
2.5 mg/min, the degradation rate approximated that for TCDD. Approximately
50% was degraded after 40 minutes of ozonation. These preliminary studies
indicate the feasibility of ozone or UV/ozone degradation systems for
waste chlorodioxins in wastewater (43).
Hydrazine, Moncmethyl Hydrazine, and Unsymmetrical Dimethylhydrazine
The hydrazine family of fuels includes hydrazine (H), monomethyl-
hydrazine (MMH), and unsymmetrical dimethylhydrazine (UDMH) as well as
mixtures of these compounds. A study was conducted for the United States
Air Force Engineering Service Center, Tyndall AFB, Florida, by Catalytic,
Inc. Philadelphia, Pennsylvania, on the effect of ozonation on these
compounds. The effect of ultraviolet light (UV) as a photooxidizer, pH,
solution concentration, reactor inlet ozone gas phase concentration, and
superficial gas velocity were evaluated. The partial oxidation products of
ozone treated hydrazine fuels were characterized and aquatic toxicity
testing accofipl ished (44,45).
77
-------
•-J
CD
100
80
60
§ 40
20
Figure 16.
HOURS
Reduction of TCDD by Ozonation at pll 3.5 and pll 10.5
-------
100
80 -J
60 -1
40 H
20 -4
A OZONE
O OZONE +
—r
30
T
40
T
10
Figure 17.
20
MT11UTES
Reduction of TCDO by Ozonation and UV Irradiation
(5 nvj Ozone/liter)
-------
Experimental Apparatus and Methods—
The basic apparatus for all experiments is shown in a process diagram
in Figure 18. All experiments were conducted in semi-batch mode (constant
liquid supply, continuous gas supply). The reactor was a Life Systems
Modified Torricelli Ozone Contactor (LMTOC). A Grace ozonator Model
LG-2-L2 produced ozone from both air and oxygen. Air was the ozonator feed
gas for. runs when an ozone concentration of 13 mg/1 (approximately 1% ozone
in air) or less was desired. Extra-dry grade oxygen feed gas permitted pro-
duction of higher ozone concentrations (2% ozone in oxygen) than with air
for the same electrical power input to the generator.
Solutions of H, MMH, and UDMH were synthesized from fuel grade
material. Thirty liters of material were pumped to the LMTOC from a feed
tank for all runs. Samples were extracted at the mid-depth point in the
column. Concentrations of H, MMH, and UDMH in the reactor were determined
by standard colorimetric methods (44).
All reported trial data were analyzed to determine rate constants (k)
for zero, first, and second-order reactions with respect to the hydrazine
species involved. In addition, UDMH runs were analyzed for the half-order
reaction rate constant (44).
Results—
Only the results of experiments done with MMH were discussed in detail.
The authors believed that the findings for H and UDMH generally parallel
those found for MMH.
UV photooxidizer—Inlet ozone concentration to the reactor was 10.1 mg
Oj/1 (with UV light) and 11.3 mg O^/I (without UV light). Reactor off-gas,
measured near the end of both runs, contained 16.7% and 27.8% of the inlet
ozone. This difference was due to the presence and absence of UV light,
respectively. Figure 19 is a plot of the data from comparative runs in the
presence and absence of UV light and shows the positive effect of the UV as
a photooxidizer. The zero model predicts reaction half life (tj/2-min)
values of 19.2 and 26.2 and k (mg/l/min) constants of 3.33 and 2.27 for the
experimental run, with and without UV light respectively (44).
pH—The solution characteristics of pH and species concentration are
related from a standpoint of chemical kinetics, reaction oxidation pathway,
and ozone mass transfer. The pH of the solution greatly affects the rate
of oxidation of hydrazine fuels and their partial oxidation products.
Previous studies have shown that the oxidation rate of methanol is acceler-
ated at alkaline pH. Solution pH also determines the auto-decomposition
rate of dissolved ozone in solution, and therefore, the steady state
dissolved ozone residual level that is achievable. Since ozone decompo-
sition leads to oxygen radical production, pH is expected to partially
control ozone oxidation of hydrazine fuels^ Two pH levels were investi-
gated-highly acidic (pH 2.6) and alkaline (pH 9.1) (44). The effect of
solution pH on ozone oxidation of MMH is shown in Figure 20. The graph
shows that as solution pH decreased, the oxidation value also decreased.
80
-------
2-way
Transfer
Pump
1X3-
OVGasi
Qlifi
Ultra
Ozunalof
.
r
Kl
Trap
^
)
— p>
K\
Traps
| Sample"
Port
[>To
-'Veiil
Air
0^
Solu|jon
Genorolor feedpump
Wei
Test
Meier
Solution
Feed
Figure 18. Process Diagram of Catalytic, Inc. UV/Ozone System
-------
-o
oo
o.o
UV-Ught MMH(0)(mg/l)
O OFF H7
A ON 128
10
20 30 40
REACTION TIME (MIN.)
50
O
60
Figure 19. The Effect of ultraviolet Light on the Ozone Oxidation of MMH
-------
CO
U)
0.0
0
£H MMH(o)((ng/|)
O 2.6 158
P 9.1 |58
20 30 40
REACTION TIME (MIN)
50 60
Figure 20. The Effect of pH on the Ozone Oxidation of MMH
-------
Species concentration effect—In the next series of experiments,
ozonator and reactor conditions were held constant, and the disappearance
of MMH was followed during ozonation time. Different initial concentrations
of MMH (158., 505, and 1171 mg/1) were used. The graph showing the results
of this experiment is seen in Figure 21. For the highest MMH concentration
run, the reaction appears to proceed through three distinct stages. During
the first thirty minutes of ozonation, the reaction is limited by ozone mass
transfer. Between thirty-sixty minutes, the dissolved ozone concentration
approaches saturation in the LMTOC. Finally, in the last stage, the system
appears to be operating under reaction rate conditions. Data from the
505 mg/1 run indicates the same response but with a reduced phase I time
period. This was expected as the initial mass of MMH present was reduced.
As seen by the shape of the curve in Figure 21, the 505 mg/1 and 158 mg/1
runs did not have mass transfer limitations (44).
The extent of methanol production is related to the MMH concentration
remaining in solution and ozonation time. At time zero, the methanol
concentration is not zero because of the manner in which the batch was
charged into the reactor. The reactor contents were air sparged for eight
minutes while pumping MMH from the solution feed tank so that a portion
of MMH was converted to methanol. As the initial concentration of MMH
increased, alcohol production prior to ozonation also increased (44).
Ozone partial pressure—The amount of inlet ozone gas concentration
available to the reactor is related to ozone mass transfer and to the
maximum dissolved ozone concentration that can be achieved at fixed reactor
operating conditions. Three experiments were carried out with reactor
inlet ozone concentrations of 5.2, 10.1 and 27.7 mg 03/1 gas. All other
reactor conditions were constant. Ozone concentrations in the reactor
off-gas measured 0.66, 1.69, and 2.66 mg oyi gas at the end of the low,
medium, and high ozone partial pressure run, respectively. Species data
for these runs show that MMH disappears readily in all three partial
pressure conditions in the LMTOC. The run with an inlet ozone concentra-
tion of 5.2 mg 03/1 gas is ozone mass transfer controlled for the first 15
minutes of the reaction (Figure 22). This limitation occurred even though
initial MMH concentration was only 89 mg/1 (44).
Oxygen sparging—Gas sparging of MMH solutions with air and/or oxygen
was studied to find out if this less costly alternative to ozone oxidation
would be feasible. These experiments also compared the effect of UV light
on this process. MMH solutions were oxygen sparged in the presence and
absence of UV light. The experimental results of the runs are presented in
Figure 23 (44).
Conclusions—
• The presence of UV light in the reactor reduces t^/2values
and increases reaction rates for M, MMH, and UDMH oxidation by
ozone over reactions that did not use UV light.
• Increasing solution pH increases the ozone oxidation rate of
H, MMH, and UDMH.
84
-------
00
Ul
OO
0
MMH(0,(niQ/|)
O 1171
0 505
P 158
10 20
30 40 50 60
REACTION TIME (MINI)
70 80 90
Figure 21. The Effect of Initial Concentration of Ozone Oxidation of MMI1
-------
oo
CTl
00
lnlfl|03(mo/l) MMH(o)(mo/l)
20 30
REACTION TIME (MIN.)
0
A
o
t*' -
10
5.2
101
27.7
-A '
f>
50
09
128
102
60
Figure 22. The Effect of Inlet Ozone Concentration on the Ozone Oxidation of MMH
-------
O
10
UV-Liqht MMII(o)(mo/D
A OFF 103
0 ON 83
20 30 40
REACTION TIME (MIN)
50
Figure 23. The Effect of Ultraviolet Light on the Oxygen Sparging of MMH
-------
« Increasing species concentration, at fixed reactor operating
conditions (pH, catalyst type, ozone partial pressure, and
superficial gas velocity) increases the required hydraulic
retention time to achieve the desired effluent concentration
for all three hydrazine fuels.
• The quantity of methanol produced from ozone oxidation of
MMH is proportional to species concentration.
• Increasing ozone partial pressure decreased t -jy^values for
H, MMH, and UDMH, but ozone utilization efficiency is reduced.
• MMH decomposition during oxygen or air sparging is greatly
enhanced by UV light. The same trend, to a lesser degree,
occurs with UDMH and H (44).
Identification of the Partial Oxidation Products of H, MMH, and UDMH—
During the study described previously, attenpts were made to identify
the partial oxidation products of H, MMH, and UDMH that remained in solu-
tion after the fuel was removed by UV/ozonation. The studies involved both
chemical and bioassays (44).
Oxidation of hydrazine results in water and nitrogen, according to the
following equation:
N2H4 + 03 —^ 2 H2O + N2 + 0 (45)
Ozonation produced a small yield of nitrate-N. Ammonia was probably a
product of a side reaction and produced nitrate after oxidation.
The main reaction of MMH oxidation would be expected to follow the
stoichiometry of the potassium iodate reaction used to determine MMH
analytically. This is represented by the following equation:
CH3N2H3 + KIO + 2 KC1 —> KC1 + IC1 + Q^OH + N2 + 2 H2O (45)
This equation predicts that methanol will remain after initial oxidation
of MMH is complete. This was verified during the experimental runs. MMH
also reacted with ozone to produce nitrate-nitrogen. Although methanol was
the only oxidation product identified by from MMH ozonation, at least four
other organic compounds were identified from gas chromatography peaks.
Methanol, formaldehyde dimethyhy3razone, formaldehyde monomethyl-
hydrazone, N-nitrosodimethylamine, dimethyl formamide, and tetramethyl
tetrazene oxidation products were identified in UDMH ozonations. Formal-
dehyde and formic acid could also be formed from methanol in both the
MMH and UDMH tests.
Since residual organic compounds from MMH and UDMH ozonation are
expected to be amenable to ozonation, their concentrations can probably be
reduced by continued ozonation past the point of fuel removal. Sio-assay
studies indicated ozonation could reduce toxicity levels of methyl
hydrazines, however, the resultant wastewater contained some residual
toxicity. Consequently, further ozonation is necessary to produce a
wastewater safe for reuse (45).
88
-------
Nitrobenzene
The Westgate Research Corporation of West Los Angeles, California,
conducted a study to: (46)
o identify the major oxidizing species in an UV/ozone
photooxidized system
© identify the major stable intermediates resulting from
UV/ozonation of representative aromatic pollutants
© acquire data to aid in predicting optimum conditions for
UV/ozonation of different classes of organic compounds
A study of model compounds and their products in a UV/ozone system would
promote the understanding of the nature of the predominant oxidizing species
in the UV/ozone system.
Nitrobenzene was selected as the primary model compound based on the
ease of gas-liquid chromatographic (glc) separation of the isomeric nitro-
phenols, availability of suspected intermediates, and because it is a
•representative aromatic pollutant. Ratios and amounts of isomeric phenols
formed by UV/ozonation of nitrobenzene were compared to the ratios and
amounts of the same isomeric phenols formed by exposing nitrobenzene to other,
known oxidizing systems. Stable intermediates and pH effects were also
studied (46).
Experimental Approach and Results —
Solutions of 0.8 mM nitrobenzene in distilled water were subjected to
ozonation and uv/ozonatiori at a rate of 0.055 mmoles 0 3 /liter solution/
minute. The ozone generator consisted of a cylindrical stainless steel
vessel with an electropolished inside surface. A 40 watt low pressure
mercury arc lamp axially located served as the UV light source. Ozone in
oxygen ic about 2 wt% was introduced at the bottom of the reactor. The
solution for ozcnation was introduced through an inlet port at the top.
Ozone was generated in the laboratory by a silent arc discharge generator.
The reactor was operated in the batch mode. Samples were drawn at pre-
determined intervals from a valve at the bottom of the reactor. Nitro-
benzene solutions prepared in the same manner as those treated with
UV/ozone were also oxidized with the following model oxidizing systems
Fenton's reagent — a known hydroxyl radical generating system, NaOCl/frjOs
which produces singlet oxygen, UV light solely, HjC^, and a I^O^/UV system
in the presence of air or N2 . The amount of I^C^ in all these other systems
was made equal to the stoichiometric amount of ozone used in ozonation and
UV/ozonation.
The oxidized nitrobenzene solutions were extracted with
derivatized with CHjfr^ r a^ the levels of the isomeric nitrophenols were
assayed by glc. Identification was confirmed by GC/MS.
It is apparent that both ozonation and UV/ozonation of nitrobenzene
yield very similar ratios of ortho-, meta-, and para- nitrophenols, with
the para- isomer predominant. Other oxidizing systems yield different
iscmer ratios. Both Fenton's reagent and the f^C^/UV/air system yield an
89
-------
isoroer ratio that is very nearly opposite that obtained by ozone and
UV/ozone systems.
Both the ratios and the amount of nitrophenols formed by Oj and
indicated that none of the other oxidizing systems represent the conditions
that exist during UV/ozonation and that the conditions in the ozone system
in the absence of UV light are very nearly the same as in the presence of
UV light. This finding would appear to exclude hydroxyl radicals as the
major oxidizing species in the hydroxylation of nitrobenzene by the UV/ozone
system.
Similar tests, conducted on benzene, also indicated that the hydroxyl
radical or other free radicals are not responsible for oxidations that take
place in uV/o^one sys.temSo The. authors concluded that ozone itself is the
major Gx'idi'sing -spgcias .irs^thase.aromatic systems.
L*1- " ""'u \'^ T-i i-jrch^nzene? ."banzQicl^^^^^and^ anisole^
t'. —.--• .: '-.I ,-aui- .
behaves as a strongly electrophiAJc radical i
-,For essirple^ the. uy/ozonatiort rOf - 4nisole woo1." be
faster than ' e UV/ozonation of ber,;soic acid (45) „
Study of. Stable .,/nacii.- ^^.terrus
.- . . ^ ' -' S ' "....-' '.f~. .
During the next phases of <-H° -rif.urJy, j^jeous'0.8 raw 'nitrobensene
solutions were also used, with an c,^ . . , \r;, o? ^0 rog/liter/ninufce. -. : •.-,-•
Data on the nature and change With' time 01: cn<= i. ;
-------
NITROBENZENE
^Mixture of o-, m-,
and p-NITROPIlENOLS
4-NITRQCATECHOL
+
GLYOXAL
MALEIC ACID
MALEIC ALDEHYDE
-I-
GLYOXYLIC ACID
N0n
J NITRCMUCONIC
JACID Derivatives
OXALIC ACID
4-NITRORESORCINOL]
and
3-NITRORESORCINOL 1
J Mixture of
TRIITYDROXYNITPOBENZENES I
Figure 24. Nitrobenzene Ozonation Pathway
-------
A small amount of maleic acid was also isolated, and the presence of maleic
aldehyde and glyoxylic acid was indicated by formation of their 2,4-di-
ni trophenylhydrazanes.
Similar studies were also determined with anisole, benzoic acid and
3-chlorobenzoic acid as model compounds. The same aliphatic intermediates
as in the case of nitrobenzene (formic acid, oxalic acid, and glyoxal) were
formed. CC>2 evolution and total organic carbon loss curves were also nearly
identical to the previous study.
In a continuance of the study, the major stable intermediates oxalic
acid and formic acid were subjected to ozonation and UV/ozonation. UV
light increased the rates of ozonation 18 fold for oxalic acid and 6 fold
for formic acid (46).
pH Effects—
The effect of pH on the rate of ozonation in the presence and absence
of UV light was under study by the authors at the time of publication
(1979). Preliminary results indicated that the role of ozone as a nucleo-
philic reagent is just as important as the usually emphasized electrophilic
attack by ozone. Picric acid at a concentration of 0.44 mM was completely
ionized and subject to attack by the positively charged oxygen atom of the
ozone molecule. In a basic solution, hydoxyl ions may compete with
picrate ions for available ozone and thereby inhibit both ozonation and
UV/ozonation of picric acid. This competition would not occur in acid
solution, and added acid would affect the rate of picric acid degradation
very slightly, if at all.
When nitrobenzene is subjected to the same experimental conditions as
picric acid, the benzene ring is electron-defecient because of the attached
nitro group. The negatively charged oxygen in the ozone molecule may be
the primary oxidizing species in the first step of either ozonation or
UV/ozonation of nitrobenzene. In acidic solution, hydrogen ions may compete
with nitrobenzene for the available ozone. Therefore both ozonation and
UV/ozonation of nitrobenzene may be inhibited in acidic solutions.
To summarize this study, UV/ozonation of organic compounds in water at
levels of about 0.8 mM is probably not a free radical phenomenon. Major
intermediates for the UV/ozonation of monosubstituted benzene derivatives
were oxalic and formic acids. In the absence of UV light these acids are
relatively stable, but in contrast they are readily ozidized to CO, by the
UV/ozonation system (46).
Copper Process Waste Streams
System Products Division of IBM Corporation, Endicott, New York,
treated two copper removal process waste streams with ozone and ultraviolet
light. The waste streams were from a chemical copper recovery process and
an electrolytic copper recovery (depleting) process. They contained Na-,SO,
formic acid, ECTA (ethylenediamine tetraacetic acid), and anionic
92
-------
surfactant* In order to discharge the treated effluent into the plant clar-
ification system, the final concentration of EDTA should not exceed 5 mg/1.
Experimental Apparatus, Methods, and Results—
Bench scale tests—to Ultrox Irradiation Model B-803, bench-scale
UV/ozone system was used for all test runs. This system is made by
Westgate Research Corporation, West Los Angeles, California. Ozone was
generated by feeding either industrial grade cylinder oxygen or oxygen from a
cryogenic storage container to an OREC Model 03B2-O ozonator made by Ozone
Research and Equipment Corporation, Phoenix, AZ. Ihe measurement and
control of solution pH was performed with the aid of a Corning Model 12
pH meter that was checked against appropriate buffers for the desired pH.
In all bench-scale tests, nine liters of waste solution were treated
with ultraviolet light and ozone in the reactor. After the solution was
pumped into the reactor, pH was adjusted between 4 and 6, oxygen gas was
introduced into the system, and the flow rate was adjusted to the minimum
bubbling rate. Bubble rate and size were observed through glass viewing
ports.
After the ozone generator and UV lamps were switched on, a slow stream
of ozone-rich bubbles moved upward through the waste solution. During this
initial period in the run, the surfactant in the waste was oxidized to the
point where it no longer caused excessive foaming. At this point, oxygen
flew was increased to approximately 192 1/min and ozone generator power was
set to yield 907-998 grams of ozone per day (47).
All test conditions were the same except that the waste solution from
the deplater contained 16 ppm of iron and 18 ppm of copper, while the
solution from chemical copper recovery contained negligible iron and
copper by comparison. The initial EDTA concentration of both solutions was
approximately 1000 ppa.
After eight hours, the chemical copper recovery solution had an EDTA
concentration of about 5 ppm. After 12 hours, the EDTA concentration in
the deplater solution was 430 ppm. The deplater solution had become turbid
from formation of extremely fine particles of reddish brown Fe(OH)3.
Although the total amount of iron present in the deplater solution was
small, the effect on the reaction was very significant. If the initial
copper concentration was greater than 30 pprc Cu(II), a large number of
black CuO particles formed that also adversely affected the reaction (47).
Pilot-Scale Tests—
Pilot-scale test runs were performed using an Ultrox water purification
system, Model P7708 STAC manufactured by Westgate Research Corporation. A
synthetic waste solution composed of reagent grade sodium formate, disodium
ethylenediaminetetraacetic acid, sodium sulfate, 50% sulfuric acid, and
deicnized water was used to simulate the chemically copper-recovered addi-
tive waste solution. The solution contained 22 gm/1 of formic acid,
93
-------
100 gm/1 sodium sulfate, and 530 mg/1 of EDTA (acid form). The quantity of
sulfuric acid present yielded a solution at pH 3-4. The solution also
contained approximately 2 mg/1 of Fe(III) and no detectable Cu(II). The
pH was monitored continuously and maintained in the 4-6 range by addition
of 50% sulfuric acid as the oxidation progressed. The solution temperature
increased from 25 to 45*C during the course of the run due to heat of
reaction and heat from the ultraviolet lights (47).
Complete destruction of EOTA and formic acid was achieved for all runs
in approximately eight hours using an ozone mass flow rate of 5.5 gm/min at
4% by weight of oxygen. Ozone concentrations were determined iodometri-
cally with a PCI Ozone Corporation ozone monitor. Eight, 65-watt ultra-
violet lights were on during test runs.
To verify adverse effects of high Cu(II) and Fe(III) concentrations on
the UV/ozone process, deplater solutions containing 16-20 mg/1 Cu(II) and
23-28 mg/1 Fe(III) were subjected to the process. The initial EOTA concen-
tration was 1677 mg/1 with lesser amounts of wetting agent, formic acid,
formaldehyde, and methyl alcohol.
After 18 hours treatment with ultraviolet light and ozone (same
reaction conditions as used for the synthetic waste), the EDTA concentra-
tion was 150 mg/1. Even though a much longer treatment time was used
because of the higher initial EOTA concentration, it was estimated that
the time required to reach an EDTA concentration of 5 mg/1 would easily
extend beyond 24 hours. Prolonged treatment time was also observed in
bench scale tests of deplater solution where the initial EOTA concentration
was lower and the Cu(II) and Fe (III) concentrations were the same or
higher (47).
The increase in treatment time for deplater solution is caused by the
presence of Cu(II), since Cu(II) can be converted by ozone to minute, black
CuQ particles. CuO particles, or other copper compounds present in the
waste solution, catalyze ozone decomposition to oxygen. This results in
destruction of large quantities of ozone, and there is much less ozone
available to oxidize organic compounds. Copper concentrations of one or
two mg/1 are effective in slowing oxidation. Higher concentrations (nearly
50 mg/1) extend treatment time significantly and result in waste solutions
that appear totally black (47).
Results—
If oxidation is performed for an extended time period, the organic
compounds in waste solutions from electroless copper plating can be
completely oxidized using an ultraviolet light/ozone process. Treatment
time can be greatly reduced if the following factors are controlled:
• Waste solution pH is maintained in the 4-6 range.
• The Fe(III) concentration is less than 5 mg/1.
• The Cu(II) concentration is less than 1 mg/1.
Maintaining the indicated pH range assures that gaseous carbon dioxide
is formed and sparged out of the reactor. This prevents accumulation of
94
-------
carbonates in the waste solution. In addition, turbidity formation from
Fe(OH)2 is restricted. Low iron concentrations also limit formation of
ferric-EDTA conplexes (more difficult to oxidize that the acid form of
EDTA).
Concentrations of Cu(II) higher than one mg/1 cause formation of CuO
particles that catalyze ozone decomposition. Sodium sulfate in the waste
solution is not affected by ultraviolet light/ozone treatment (47).
PCB's
General Electric Company Pilot Plant Study—
A pilot plant was set up at General Electric Company's Capacitor
Products Department Facilities in Hudson Falls, New York, to demonstrate
efficiency and cost-effectiveness of the ULTRQX UV/ozone system to destroy
PCB's in industrial effluent. Although PCB's have not been used in the
Hudson Falls/ Fort Edward facilities since July, 1977, residual PC3 concen-
trations in the untreated effluent ranged from 5-40 iig/1 (48).
Experimental apparatus and methods—The portable, skid-mounted pilot
plant was 28" wide, 45' long, and 45'high with a 75 gallon wet volume.
It was fabricated from 304 stainless steel, and passivated plus electro-
polished to reduce chemical attack and increase UV reflectivity. The
reactor could accomodate up to thirty, 40 watt G36T6L, low pressure UV lamps.
Ozone was produced from liquid oxygen via an ozone generator and
diffused from the base of the reactor through porous ceramic spargers.
Figure 25 shows the pattern of water flow through the reactor. Water
passes through each of the stages in a tortuous path to achieve a greater
degree of plug-flow. In each stage, Water is contacted by ozone gas and
UV light (48).
Thirty-seven tests were run and the following variables were investi-
gated UV lamp patterns, influent flow rate, ozone mass flow, ozone concen-
tration, ozone mass flow distribution, pH of influent, and tenperature.
Data from these tests are tabulated in Appendix D. From the results of
these tests it appears that most of the test runs had more than one
variable, and the authors drew no specific conclusions concerning the most
favorable parameters for PCB destruction. However, many of the runs
contained no ug/1 of PCB in the effluent after treatment, and no run had
more than 4.2 yg/1 of PCB. The latter run was the only one which did not
expose the waste to UV light.
Results—
When the data in Appendix D were subjected to conputer analysis, a
mathematical model was formulated to simulate the performance in destroying
PCB's both in the ULTECK pilot plant and in full-scale equipment (Appendix
E). Upon completion of a satisfactory model, feed and effluent PCB con-
centration limits were established and capital plus O&M costs were derived.
These costs will be discussed in detail in the section on economics.
95
-------
0V LAMPS (TOP)
s~
O
O
o
o
o
4
-%
c
i
o
o
o
&
6
o
o
o o
FLO;^
DISTRIBUTOR
WAb
•%
o
o
o
o
s~
o
o
o
o
o o
V ,*
~M
o
o
o
o
o
TE WATER IN
SPQW 0-
GAS OUT
GAS-
SEPARATO:;
T
WATER cXfT
SOLID STATE
CONTROLLED
GEAR PUMP
Figure 25. Schematic of Tap View of ULTHDX Pilot Plant
(Ozone Sparging System Omitted)
96
-------
SECTION 6
HAZARDOUS WASTE DESTRUCTION TECHNOLOGIES
IN THE DEVELOPMENT STAGE
BACKGROUND
This section reviews technologies for the treatment of hazardous
wastes that are still in the developmental stage. Most of these technolo-
gies have been studied only at the bench level; a few are at the pilot
plant stage. The technologies reviewed include catalyzed wet oxidation of
toxic chemicals, the dehalogenation of compounds by treatment with ultra-
violet light and hydrogen, electron irradiation of toxic compounds in
aqueous solution, UV/chlorinolysis of hydrazine in aqueous solution, and
the catalytic hydrogenation dechlorination of polychlorinated biphenyls
(PCB's)o
Wet air oxidation experiments were conducted in a titanium autoclave,
but only batch oxidations were investigated. The ultraviolet light/
hydrogen technology for the dehalogenation of compounds has advanced from
bench scale to the pilot plant stage. This system is not designed to
process waste streams with much more than 1% toxic organic content. The
electron treatment of trace organic compounds in aqueous solution has been
investigated in the laboratory at MIT. The UV/chlorinolysis of hydrazine
in aqueous solution has been tested on 8,000 liter batches of wastewater.
The catalytic hydrogenation-dechlorination studies of PCB's took place
in an autoclave.
A description of each process, experimental details, plus results and
discussion follow. The following lists the various chemicals treated by
specific processes:
LIST OF TOXIC SUBSTANCES TREATED BY CHEMICAL TECHNOLOGIES
Met Air Oxidation
2,4-D
Glycolic Acid
Pentachlorophenol
Ethylene dibromide
Malathion
Acetic Acid
PCB's
TCDD (tetrachloro-p-dioxin)
Kepone
97
-------
Dehalogenation/uV-hydrogen (HJ
Arochlor 1254 (PCS) l
Tetrabromophthalic anhydride
Kepone
Electron Irradiation
2,3,4'-Trichlorobiphenyl
4 Monochlorobiphenyl
Monuron
UV/chlor inolys is
Hydrazine
Monomethylhydrazine
Dimethylnitrosamine
Unsyinmetrical dimethylhydrazine
Catalytic Hydrogenation-Dechlorination
PCB's (Arochlor 1242, KC-400
DESTRUCTION OF TOXIC CHEMICALS EOT CATALYZED WET OXIDATION
Wet air oxidation (WAO) is a commercially proven technology for the
destruction of organics in wastewater and sludges. In conventional wet air
oxidation, waste is pumped into the system by a high-pressure pump and
mixed with air from an air compressor. The waste is passed through a heat
exchanger and then into a reactor where atmospheric oxygen reacts with the
organic matter in the waste. The oxidation is accompanied by a temperature
rise. The gas and liquid phases are separated, and the liquid is circu-
lated through the heat exchanger before discharge. Gas and liquid are both
exhausted through control valves. System pressure is controlled to
maintain the reaction temperature as changes occur in feed characteristics
(i.e., organic content, heat value, temperature). The mass of water in the
system serves as a heat sink to prevent a runaway reaction that might be
caused by a high influx of concentrated organics (53).
IT Enviroscience, Inc., of Knoxville, Tennessee, has developed a
proprietary catalyzed wet oxidation process based on information in U.S.
Patent 3,984,311 (54) (Originally assigned to the Dow Chemical Company,
and now assigned to IT Enviroscience for development and commercialization).
This process uses a cocatalyst system consisting of bromide and nitrate
anions in an acidic, aqueous solution to destroy either organically con-
taminated aqueous waste or organic residues. Destruction of waste organics
is accomplished by mixing a waste with the catalyst system and oxygen
(or air) at temperatures greater than 100°C (55).
Experimental Methods
Experiments were conducted in a 1 liter stirred titanium autoclave
and limited to batch oxidation. Aqueous wastes or organic residues were
pumped into a continuously stirred tank reactor (CSTR) containing a solution
of HBr and nitric acid (Figure 26). Air is sparged in, and the organics are
98
-------
ORGANIC +
WATER
STEAM HEAT
(if needed)
SOLIDS
Figure 26. IT Enviroscience Process Concept
for Homogeneous Catalyst
99
-------
oxidized with the heat of reaction driving off water. Any solids formed
have to be removed, but the catalyst solution remains in the reactor.
Carbon dioxide, water vapor, excess air, and any volatile organics formed
also leave the reactor. The most important concepts in the process are that
non-volatile organics remain in the reactor until oxidized, and there is no
bottoms product. Thus, very high destruction efficiencies and low reactor
effluent concentrations are not required in reactor design. If the organics
remain in the reactor long enough, they will ultimately be destroyed (55).
Measurement of Destruction Rates—
Organic destruction rates were measured by different procedures de-
pending on the water solubility of the organic. The oxidation rate of
water soluble organics was measured by withdrawing liquid samples during
the reaction and analyzing the organic concentration by Total Organic
Carbon (TOC). The destruction rate for insoluble organics was measured by
terminating the reaction, cooling the system, and solvent extracting the
reactor system and catalyst mixture. The solvent was then analyzed by gas
chromatography for unreacted organics and by-products.
The extraction procedure limited the amount of data collected during
the experiment, but it was reliable and reproducible. Secondary measure-
ments of the organic destruction were made by inorganic chloride analysis
for destruction of chlorinated organics and by carbon dioxide analyses of
the final reaction gas. These measurements were used to determine the
completeness of the organic destruction. Over 200 runs were made to screen
the process efficiency on a variety of organic compounds and measure the
effects of different catalyst combinations (55). .
Economics
Based on the destruction rates of the organics tested and the process
designs described above, preliminary capital and treatment costs have been
estimated (Table 14). These estimates show the costs for fast and moderate
destruction rates of aqueous wastes (pentachlorophenol or glycolic acid)
and the cost for a moderate destruction rate organic residue (Arochlor
1254). The capital cost is the total installed cost using titanium as the
construction material (55).
Results
The organics studied can be grouped into two categories, those which
oxidize rapidly at low temperatures and those with slower destruction rates
that require higher reaction temperatures. Some of the results of the fast
destruction rate organics are shown in Table 12. These experiments were
conducted at 165°C with a catalyst mixture of 0.5-5% Bromide and 0.1-5%
Nitrate. The organic destruction rates, as measured by the disappearance
of the initial organic, are much higher than the reported destruction rates
of conventional wet oxidation at this temperature. The complete oxidation
to carbon dioxide of these compounds was measured by the disappearance of
total organic carbon for 2,4-D and glycolic acid and by final carbon dioxide
concentration for pentachorophenol, ethylene bromide, and malathion. The
100
-------
TABLE 12
ORGANICS WITH FAST DESTRUCTION RATES
2,4-D
Glycolic Acid
Pentachlorophenol
Ethylene Dibromide
Malathion
Organic
Reduction
65%
99%
99%
94%
99%
Total
Organic
Destroyed
65%
99%
75%
51%
58%
Temperature
165°C
165"C
165°C
165°C
165"C
Reaction
Time In
Minutes
15
30
30
60
60
TABLE 13
ORGANICS WITH MODERATE DESTRUCTION RATES
Acetic Acid
Arochlor 1016
Arochlor 1254
Dioxin
Kecone
Organic
Reduction
36%
93%
95%
99%
93%
Tenroerature
165°C
195°C
250°C
200°C
250°C
Reaction
Time In
Hours
1-1.5
1
2
4
6
TABLE 14
PRELIMINARY CAPITAL AND TREATMENT COSTS
Capital Cost
1979 $
Treatment Cost
CAg organic
5/4000 liters
Fast
Destruction
Rate
$844,000
10.4
21
Aqueous ( * )
Waste
Moderate
Destruction
Rate
$1,260,000
13.2
27
Organic (5)
Residue
Slow
Destruction
Rate
$890,000
100
(*) 100 liters/minute, 5% organic concentration
(#) 22.5 kg/hour organic residue
Source: Miller, R0A. et al., IT Enviroscience, Inc., 1980.
101
-------
total organic destruction rate was 90% for these compounds when the reaction
temperature was raised to 200°C (55).
Organic compounds with either low solubility or chemically stable
structures are more difficult to oxidize. These compounds include acetic
acid, polychlorinated biphenyls (PCB's), tetrachlorodibenzo-p-dioxin,
(TCDD), and Kepone. Temperatures up to 250°C were required to destroy
these compounds (Table 13). The capability to measure total destruction
of the organic by carbon dioxide was not available at the time these
experiments were conducted, but gas chromatography of the final reactor
contents indicated total destruction of most compounds (55).
Process Design Concepts
Aqueous Organic Waste—
The process design for destroying both aqueous waste and organic
residues centers on utilizing the homogeneous catalysts in a continuously
stirred tank reactor. The two variations on the basic reactor concept, one
for aqueous waste and one for organic residues, differ in the amount of
water processed. For dilute aqueous wastes, the energy released by the
oxidation of the organics is insufficient to remove all of the incoming and
formed water. Therefore, it is necessary to recover the catalyst solution
for reuse. Figure 27 shows the basic recovery concept whereby the catalyst
could be recovered by evaporation/concentration. The evaporative recovery
process requires vaporization of all water entering the system. The water
vapor is condensed and discharged separately from the off-gas. Auxiliary
heat must be supplied for streams containing less than about 4% organics
(depending on the heat content of the waste), and evaporators must normally
be used to supply sufficient heat transfer area.
The number of evaporators is an economic decision based on steam
economy versus capital. This type of economic evaluation was done for a
model plant designed to treat 80 1/min, with the result that a single
evaporator offered the best compromise between unit costs, capital, and
versatility. For wastes containing high organic levels, no evaporator
capacity would be required (55).
Non-Aqueous Organic Waste—
The continuous process concept for treating non-aqueous wastes is sub-
stantially different from the process concepts described for aqueous
organic wastes since the quantity of water which must be removed from the
process is very low. Figure 28 shows the process concept. In many cases,
only the water formed as a by-product from the oxidation reaction (plus any
amount of water entering the compressed air) must be considered. The only
stream normally leaving the process is the off-gas containing principally
nitrogen, unused oxygen, carbon dioxide, low levels of water vapor, traces
of volatile inorganic (HC1), and organic species which could be present in
the reactor mixture.
102
-------
WATER VAPOR & OFF GAS
WASTEWATER
AIR
STEAM
o
u>
OXIDATION
REACTOR
I
Y
CATALYST SOLUTION
PURGE
TREATMENT
RECYCLED, CONCENTRATED
CATALYST SOLUTION
CONDENSER
WATER
& (CONDENSATE)'*
SOLIDS
Figure 27. IT Process Diagram for Aqueous Waste
-------
OFF GAS
o
WASTE RESIDUE
AIR
CONDENSER
OXIDATION
REACTOR
j _
PURGE
TREATiyENT
L!
CATALYST SOLUTION
SALTS,
SOLIDS,
HC1
Figure 20. Process Diagram for Organic Residues
-------
Heat generated from the oxidation must be removed by condensing and
refluxing water vapor leaving the reactor. This heat could be recovered by
operating the reflux condenser as a steam generator. The catalyst is con-
tained in the reactor and extra unit operations (such as evaporation) for
catalyst recovery are not required (55).
Summary
Advantages of the IT Environscience catalyzed wet oxidation process
are best understood relative to the conventional technologies of uncatalyzed
wet oxidation and incineration. In comparison to straight wet oxidation,
the catalyzed process achieves high levels of destruction of a variety of
organic chemicals at significantly lower temperatures and pressures. Con-
ventional wet oxidation requires temperatures approaching 300°C and high
pressures to achieve greater than 90% destruction of soluble organics.
The catalyzed process operates at less than severe conditions. It also
produces no aqueous bottoms product; all nonvolatile organics stay in the
reactor system until oxidized. The homogeneous cocatalyst enables the
system to treat water insoluble compounds.
In comparison to incineration of hazardous wastes or aqueous wastes,
the catalyzed wet oxidation process has several advantages. Little or no
added energy is required and auxiliary fuel is usually not consumed. It
has few unit operations and functions at low temperatures and pressures.
Vent gas volume and vent gas scrubber effluent are lower than those produced
by incineration and are readily adaptable to polishing treatment if required
for control of trace toxic releases. With few unit operations and low
volume streams, the oxidation system is potentially portable and can be
relatively easily developed in pilot-plant tests (55).
The system is projected to be capital intensive from a cost standpoint.
Depending on the tvpe of waste and desired destruction rate, capital outlays
of~S350,000-$l,300~,000 are projected (1980).
DEHALOGENATION OP COMPOUNDS BY TREATMENT WITH ULTRAVIOLET LIGHT AND HYDROGEN
A patent for a process to dehalogenate compounds by treatment with
ultraviolet (UV) light and hydrogen (H2) has been assigned to Atlantic
Research Corporation, Alexandria, VA (56). The process is effective for
halogenated organic compounds with at least one C-halogen group and works
in the absence of any substantial amount of oxidizing agent. The halogen-
ated compound is reduced when carbon-halogen linkages are broken (halogens
are liberated during the process). The treatment may result in further
degradation of the partially dehalogenated compound. Different degrees of
dehalogenation are primarily related to the energy in the C-halogen bond
and can be compensated by employing higher or lower energy UV radiation.
Experimental Methods
Figure 29 shows a schematic drawing of the 1.5 liter reactor used in
the bench scale process. The reactor is equipped with a 254 nm UV inver-
sion tube, hydrogen gas bubbler, and a recirculation pump. UV radiation
105
-------
TEFLON
PLUG
REACTION
SOLUTION
UV
LAMP
TO UV
LIGHT
VOLATILES
EXIT
REACTION
•CHAMBER
REACTION
> SOLUTION
Figure 29. Schematic of Atlantic Research 1.5 liter Reactor
106
-------
for these experiments was at 254 nm, since it has been established there
is high absorptivity by halogenated compounds at this wavelength. The UV
tube is positioned longitudinally in the reactor chamber and held in an
air-tight position by teflon plugs. The UV tube is connected by wires to a
transformer (not shown). Hydrogen gas is pumped in via an inlet tube.
Reaction solution is pumped in via another inlet tube and continuously
recirculated by pumping through an outlet tube (not shown). Another outlet
tube provides for the exit of volatiles (56,57).
The process can be employed for compounds in the gaseous, liquid, or
solid state. Halogenated conpounds in the liquid or solid form should be
in a finely divided form and dissolved in a suitable solvent which is
substantially transparent to the UV wavelengths used. The type of solvent
used should be determined by the solubility characteristics of the compound
of interest. Alkaline solutions of water, methanol, ethanol, 1- and
2-propanol, hexane, cyclohexane, and acetonitrile are exanples of solvents
used in the process. Alkalinity is preferably produced by the presence of
alkali metal oxide or hydroxide to minimize potentially destructive anions.
An organic compound such as methanol is used to solubilize compounds in-
soluble in water (56).
Reaction Mechanisms
Possible reaction mechanisms suggested by the authors include:
o formation of an excited molecule followed by homolytic
dissociation of carbon-chlorine bonds and hydrogen
abstraction by the radicals produced from both hydrogen
gas and water molecules
• a bi-molecular reaction of the excited species with hydrogen
gas or water
Results and Discussion
The process has been demonstrated successfully on Arochlor 1254
(a polychlorinated biphenyl) in methanol and on tetrabromopthalic anhydride
in methanol. However, the majority of work has been done on Kepone, both
in methanol and water. Degradation of Kepone was found to be pH dependent.
When reactions were run in 0.1, 1.0, and 5% NaOH, the latter caustic level
gave the best degree of degradation. More than 99% removal of Kepone has
been observed in less than 90 minutes. In tests that compared the UV + H2
process with UV/ozone or straight UV treatment, the quantity of unreacted
Kepone or Arochlor 1254 remaining after a given time was always less for
the UV + Bj runs.
Several degradation products were observed including mono-, di-, tri-,
tetra-, and pentahydro derivatives of Kepone. Isotopic studies performed
indicated that the hydrogen in these compounds comes from hydrogen added
during the process.
Based on successful bench studies, a pilot plant for the treatment of
higher levels of Kepone was built by the Atlantic Research Corporation and
was in use during 1980. Economic data regarding the system is currently
107
-------
unavailable. The successful treatment of Kepone implies the system should
be adaptable to PCB's and a variety of other organic hazardous wastes in
the ppn or lower range (56,57).
ELECTRON TREATMENT OF TRACE TOXIC ORGANIC COMPOUNDS IN £QUEOUS SOLUTION
As part of a National Science Foundation-sponsored program at the
Massachusetts Institute of Technology (MIT), the effect of electron bombard-
ment on trace toxic organic compounds was studied in pure water solutions
and in model systems containing trace organics in water. Trace organic
compounds studied included Monuron (a persistent herbicide of the urea
type), 2,3,4' trichlorobiphenyl, and 4 monochlorobiphenyl (58).
Experimental Methods and Apparatus
Ssnples were treated with 3 million volt electrons from the Van
de Graaff accelerator at MIT's High Voltage Research Laboratory. Pre- and
post- irradiation samples w=re analyzed using a Waters #204 reverse
gradient high pressure liquid chromatograph (HPLC). The reverse gradient
indicated both the "parent" compound and the creation of molecular frag-
ments. All samples were detected by UV absorbance at 254 nm; peak heights
of the parent compounds were found to be linear with concentration.
Master solutions were made by adding a concentrated solution of the
test material to a 1 liter bottle of acetonitrile, allowing the acetonit-
rile to evaporate, and adding 1 liter of twice-filtered water. The
resulting solution was allowed to stand for about a week to come to
equilibrium. For irradiation, 2 ml of the master solution were placed in
closed glass Petri dishes with a liquid capacity of 6 ml. The Petri dishes
and their contents were passed through the electron beam on a moving belt.
Radiation dose was controlled by adjustment of the product of electron beam
current and exposure time (58).
Monuron
Figure 30 shows the chromatogram of Monuron (0.4 mg/1 in water) by
reverse phase gradient elution HPLC before irradiation. Figures 31 and
32 show the results of 10 and 100 kilorads irradiation respectively.
In Figure 31, a vestige of the original parent peak remains with three new
peaks (representing degradation compounds) to the left of the parent peak.
These degradation products are more soluble in water than the precursor.
After 100 kilorads irradiation (Figure 32), the original peak and neighbor
degradation peaks are completely removed. To the far left, there is a
broad peak of highly water-soluble residue. Figure 33 summarizes the data
concerning percent degradation versus radiation dosage. Substantially
complete destruction is attained at 30 kilorads (58).
2,3,4' trichlorobiphenyl
2,3,4'trichlorobiphenyl has low water solubility. When dissolved in
water to the saturation limit (80 ppm), it was totally destroyed by all
radiation doses down to 1 kilorad (58).
108
-------
Figure 30. Monuron in Water at 0.4'mg/l, Stand-
ard before Irradiation.
.AFTER
IRRADIATION
o—=- Degradation
Peak
Figure 31. Monuron in Water at 0.4 mg/1. Ex-
posed to 10 Kilcrads.
109
-------
AFTER
IRRADIATION
Figure 32. Monuron in Water at 0.4 mg/1, Ex-
posed to 100 Kilorads,
100
10 15
DOSE (kitorads)
33- % Degradation Vs. Dose for Monuron
in Water, 0.4 mg/1 and 4.2 mg/1.
110
-------
:\
4 Monochlorobiphenyl
4 Monochlorobiphenyl was somewhat more resistant to degradation than
2,3,4' trichlorobiphenyl and less resistant than Monuron. When dissolved in
pure water to a concentration of 0.8 mg/1, a dose of 10 kilorads was re-
quired to achieve nearly complete degradation (58).
Summary and Discussion
Although the Van De Graaff generator successfully treated toxic
organics, scale-up of the system would be projected as highly cost- intensive.
Few hazardous waste generators have access to a high energy accelerator.
Since the test samples were placed in Petri dishes and irradiated on a
moving belt, little can be inferred as to the type and cost of scale-up
feed mechanisms . Moreover the organics were irradiated in nearly pure
water. The attenuating effect of turbidity may be substantial when irradi-
ating a heterogeneous water system.
Consequently this technology, although successful under ideal labora- .
tory conditions, does not appear to be amenable to scale-up from either a
cost or industrial standpoint.
UV/CHLORINOLYSIS OF HYDRAZINE IN DILUTE AQUEOUS SOLUTION
A process was developed at Rocky Mountain Arsenal for the United States
Air Force by the IIT Research Institute for the UV/chlorinolysis treatment
of wastewater containing hydrazine (HZ) in concentrations varying from a
few to several thousand parts per million. Some of the wastewater also
contained varying amounts of monomethylhydrazine (MMH) , dimethylnitrosamine
(DMNA), and unsymmetrical dimethylhydrazine (UDMH) (59).
Culorinolysis Reactions
Chlorine reacts with the contaminants by the following reactions:
(HZ) NH + 2 C1 + HO — > 4 Cl~ + N + HO
(MMH) (CH3)HNH2 + 2 Cl~ + H20 — > CH,OH + N2 + 4 HC1
(UDMH) (CH3)2NH2 + 2 Cl~ + 2 H20 — > 2 CH-jOH + N2 + 4 HC1
(DMNA) (CH3) 2N20 + C12 + 2 H2O — 2 CH-jOH + NZ + HOC1 + HC1
Thus, two moles of Cl^ are required for each mole of hydrazine and one
mole of Cl2is required for each mole of DMNA (59).
Experimental Methods and Apparatus
After initial testing on the bench scale, a plant was constructed that
could process wastewater in 8,000 liter batches. The flow sheet for the
process is shewn in Figure 34.
111
-------
to
Vent
Inlet
10,000 gal
Hold Tank
.C12 or N
NaOH
Na2S2°3
2,000 gal
Reactor
• Discharge to Pond
Figure 34. Process Flow Diagram UV Chlorinolysis Reaction System
-------
A 40,000 liter hold tank is used to collect the wastewater and to
level out variations in the concentration for more consistent day-to-day
operation of the facility. The chlorinolysis reactor is a 8,000 liter
glass lined reactor vessel. It is equipped with an agitator and a sparger
to insure good reactant mixing and contact. A UV light is immersed in the
reactor to activate the chlorinolysis reaction. A total of 7,500 watts of
low pressure Hg lamp UV are required. Runs were done both with the UV lamp
on and off.
The pH and chlorine concentrations of the reactor contents are monitored
and controlled,, The pH is maintained by the addition of 50 percent NaOH.
Chlorine is supplied as a gas to the reactor. The chlorine addition system
is controlled by an oxidation/reduction potential instrument (OPP) located
in the reactor recirculation loop. Control valves are installed in the
supply line to turn the chlorine flow on and off as required. The valves,
piping, and sparger are the major components of the system and are con-
structed of stainless steel.
After the reaction is completed, the excess chlorine is stripped by
nitrogen to less than 100 ppm. Residual chlorine is then neutralized with
1^28203, the pH is adjusted to 7.0, and the reactor contents are discharged
to a holding pondo The effluent may then go to a biological treatment
facility, to land spreading, or to a waterway (59).
Results and Discussion
Samples of the end products of the chlorinolysis process with and
without UV were analyzed for end products. Even though the initial
end products of the chlorinolysis reactions (N2r CH-pH) are not especially
toxic, they may further react with chlorine to form compounds such as
CH-Cl, CC14, and nitrogen trichloride (NCI3). NC13 is the most undesir-
able of these since it is explosive.
When saiples of the end products of the process for chlorinolysis
without UV irradiation were analyzed, significant amounts of chlorinated
contaminants were found in the samples. However, no contaminants were
found in the end product of the UV/chlorinolysis experiment. Although no
cost information was cited, the process successfully treated various
hydrazines,, The adaptability of the process to the treatment of other
hazardous organics is currently unknown, although it would seem probable
many other types of organics could be treated by the process. Careful
monitoring for undesirable chlorinated by-products would be warranted
when testing UV/Chlorinolysis with other organic compounds (59).
CATALYTIC HYDROGSNATICN-DECHLORINATICN OF POLYCHLORINATED BIPHENYLS
Workers at the Osaka Prefectural Research Institute and the Daido
Oxygen Company, Ltd., Japan, investigated several research parameters
regarding the catalytic hydrogenation-dechlorination of PCB's. The work
was reported in 1979. The authors investigated the use of both a Raney-
Nickel catalyst and a carbon-suported palladium catalyst. Reaction of the
PC3 took place in an autoclave where PCB was dispersed and emulsified in an
113
-------
aqueous sodium hydroxide solution containing isopropyl alcohol. Reaction
temperatures were kept constant within _+ 0.5°C (60).
Experimental Details
The Raney-Nickel catalyst was prepared by reacting 10 grams of a 50-50
Nickel/Aluminum alloy powder in 30% aqueous sodium hydroxide solution for
50 minutes at 60°C. Great care was taken to ensure that the Raney-Nickel
catalyst's activity was constant. Preliminary bench-scale studies indi-
cated the catalyst could be used more than five times. Experiments with
the Raney-Nickel catalyst were carried out in the autoclave with PCB's
such as Arochlor 1242 at temperatures between 70°C-200°C. A constant
volume solution of 100 mis composed of PCB, water, caustic, Raney-Nickel
catalyst, and isopropyl alcohol was reacted with hydrogen at either
100 kg/on2 or 30 kg/on2 . By-products were measured by flame ionization
chromatography, electron capture chromatography for chlorinated molecules,
and a GC-mass spectrometer. Inorganic chloride was measured by silver
chloride gravimetry. Major by-products were biphenyls, phenylcyclohexane,
low levels of bicyclohexane, and sodium chloride.
Following the Raney-Nickel catalyst results, a series of experiments
using a carbon-supported palladium catalyst (5% palladium/3.3 wt% to
PCB KC-400) were conducted. PCB KC-400 is usually tetrachlorinated. Ex-
periments were again carried out in an autoclave in a sodium hydroxide
aqueous solution containing isopropyl alcohol to disperse and emulsify
the PCB. Reactions were studied at constant near normal hydrogen pressures
of 25 kg/on2 at 50°-100°C.
Conclusions
The following preliminary conclusions were reached based on experi-
mental evaluations:
» The rate of dechlorination of trichlorinated PCB Arochlor 1242
was constant with constant hydrogen pressure until 65% dechlor-
ination was achieved.
• An optimum alkaline ratio of 1, corresponding to 1.05 N NaOH was
determined. An alkaline ratio of more than 1 (excess) inhibited
the reaction, whereas an alkaline ratio of less than 1 completely
stopped the reaction.
• The optimum concentration of the emulsifying dispersing agent
isopropyl alcohol was found to be 30% by volume.
• PCB KC-400 was found to be 97% dechlorinated after 30 minutes
reaction at 135°C.
Summary and Discussion
Assuming practically 100% dechlorination of KC-400 PCB, using a
Palladium-Carbon catalyst, the constant pressure hydrogenation/dechlorination
laboratory method for PCB disposal is promising, the authors noted. They
didn't cite any economic data or cost projections. Scale-up of the system
would be capital-intensive due to the cost of building a stainless steel
reactor, and the use of caustic would corrode the stainless steel reactor.
114
-------
SECTION 7
ECONOMICS OF EMERGING HAZARDOUS WASTE TREATMENT TECHNOLOGIES
MDLTEN SALT COMBUSTION
No information was found in the literature for capital costs, nor
operating plus maintenance costs regarding molten salt combustion as
applied to hazardous wastes. Estimates from a representative of Anti-
Pollution Systems indicate that it will cost from ?50-$75 to burn 907 kg
(1 ton) of municipal sludge (Personal communication from J. Greenberg to
Dr. Barbara Edwards, Ebon Research Systems, July., 1979). Rockwell Inter-
national scientists indicated that the cost to build a plant that burns
hazardous waste at 45.4 kg/hr is $400,000-$500,000. To build a plant that
burns 454 kg/hr would cost in the vicinity of $1.9-2.5 million (1979)
(Personal communication from F. Rauscher to Dr. Barbara Edwards, Ebon
Research Systems, June, 1979).
No firm cost estimates can be made until demonstration sized plants
have functioned for an extended period of time. The economics of batch
versus continuous type operations are also unknown. The type of hazardous
waste combusted, the kind of melt utilized, and the material composition
of the reactor vessel all will greatly influence cost considerations.
Because the hot molten salt is corrosive, expensive stainless steel must
be used as the reactor material. However, research is on-going to develop
corrosion-resistant materials such as alumina and ceramics as reactor
materials. This may reduce the cost of the reactor (Personal communication
from Fo Rauscher to John Paullin, Ebon Research Systems, September, 1980).
The capital costs of erecting a molten salt combustor will, of course,
be quite high. Operating costs from the standpoint of fuel consumption are
quite attractive when compared to conventional incineration. The use of
auxiliary fuel to initially melt the bath may be necessary at start-up,
depending on the type of waste combusted and its moisture content, yet the
molten bath at operating temperatures can function as a heat sink, thereby
requiring auxiliary fuel only at the onset. Heat from the combustion of
wastes can be recycled. Because there are few moving parts in the system,
maintenance costs would only be a small percentage of capital costs.
The treatment of bulk containers versus shredding the material
requires further investigation. The use of bulk quantities for loading
would probably result in lower costs.
115
-------
Transportation costs for the removal of large amounts of spent melt
(one ton or more) are unknown. Disposal in a conventional landfill may
result in a leachate problem with the salts from the spent melt.
If particulate emissions are a problem, additional expense will be
required for equipment to control the problem. Moreover, analytical
chemistry instrumentation will be required to monitor the ambient air and
the melt for residue levels of the hazardous waste or its possible harmful
breakdown products.
To summarize, because molten salt technology as applied to the
destruction of hazardous wastes is in its infancy, an economic cost profile
is difficult to develop. The process is very efficient for combustion of
certain wastes. The future will determine the economic viability of molten
salt processing of hazardous wastes.
FLUIDIZED BED INCINERATION
In a paper presented at the American Chemical Society Meeting in
September, 1979, Richard D. Ross of the Read-Ferry Company, Inc. stated
that a fluidized bed incinerator is usually high in initial cost, yet when
waste quantities are high, and the material cannot be handled in more
conventional systems, fluidized bed incineration is practical (49). When
contacted for more specific information on the cost of the process, Mr.
Ross said that the cost has a wide range and depends on the type of waste,
its caloric value, and the size of the combustor. A cost of 2-3 times more
than a conventional incinerator equipped with a waste atomizing system
would not be unusual. Problems with particulates requiring the addition
of scrubbing systems would also increase costs. Costs for analytical
instrumentation to monitor residue levels are also necessary (Personal com-
munication from R.D. Ross to Dr. Barbara Edwards, Ebon Research Systems,
Febuary, 1980).
According to M. Sittig, the type and composition of the waste is a
significant design parameter that will impact cost not only during combus-
tion but also during storage, processing, and transport prior to incinera-
ation (52). If the waste is a heterogeneous mixture, operations will be
more complex, and the combustor will require auxiliary fuel. Homogeneous
wastes that are injected and uniformly dispersed in the bed simplify the
system design and cost less to incinerate. Installation and operating
costs will vary significantly depending on the type of waste processed.
Investment and operating costs have been estimated at approximately $20
and $5 per 907 kg (ton) respectively (1979). Maintenance costs, with no
moving mechanical parts in the reactor would only be a small percentage of
initial capital costs. Environmental control costs would be related to the
type of equipment necessary to control particulate emissions and off-gases,
e.g. cyclones and afterburners. Additional costs would be incurred for
monitoring residues with analytical instrumentation (50). In some cases,
depending on the character of the hazardous waste and the type of equipment
available, conventional incinerators may be converted into fluidized bed
combustors. This could represent a smaller capital investment compared to
the cost of erecting a fluidized bed incinerator from the beginning (42).
116
-------
Mr. Raymond Esposito operates a 45-on tall, 0.35-kg/hour throughput
fluidized bed facility for Union Chemical in Maine. The combustor can burn
mixtures of municipal waste and chlorinated hydrocarbons with two seconds
residence time using a sand bed. Mr. Esposito states that he hopes to
scale up the unit so that it will process larger quantities at a cost of
approximately $40 a drum. If approximately half the waste is municipal
with a relatively high caloric content, it is not necessary to add oil as
an auxiliary fuel (Personal communication from R. Esposito to Dr. Barbara
Edwards, Ebon Research Systems, September, 1979).
Alternate fuel studies by the U.S. Army
During the evaluation of a pilot plant fluidized bed incinerator for
the combustion of munitions, the U.S. Army studied the use of alternate
fuels. Because large fluidized bed incinerators would have rather high
pre-heat feul costs for liquid fuels low in sulfur and nitrogen, plus
gaseous fuels are also expensive, studies were made regarding high sulfur
and high nitrogen fuels by the U.S. Army Armament Research and Development
Command (ARRADCOM) in Dover, New Jersey (41).
A 30 wt% nickel catalyst was evaluated for use with various fuels.
Major results indicted satisfactory operation of a catalytic fluidized bed
with both gaseous and low sulfur fuels. Initial tests revealed high sulfur
fuels poisoned the nickel catalyst with subsequent high NO emissions,
although stable combustion was obtained. Furtner investigation of various
paramenters is necessary to fully evaluate the relationship of the nickel
catalyst with high sulfur fuels.
Economic Projections by the U.S. Army
In addition to cost-savings studies for the use of alternate fuels,
mathematical models were developed to compare the fluidized bed comoustion
of a 25 wt% INT slurry versus the use of a rotary kiln. Projected costs
savings of $19,000-$193,000 (1977) per year were estimated for a 112.5 kg/hr
capacity fluidized bed system versus the rotary kiln. Cost savings of
$108P000-$311,000 per year were projected for a larger 450 kg/hour capacity
fluidized bed system. The primary factor in the cost-savings is lower
operating cost (41) „
Cost estimates were also made for a designed modular 202.5 kg (500 Ib)
per hour capacity fluidized bed system. A survey was made of major
fluidized bad suppliers for the main components chamber, blower, cyclone,
grid, instrumentation, and controls. The costs ranged from $210,000-
$505,000 (1977). Based on these figures, a reasonable cost figure in-
cluding design, fabrication, preliminary check-out, on-site supervision
during assembly, operator training, and start-up assistance would be
$400,000 (1977), according to the researchers. An additional $75,000 could
be considered for an optional heat exchanger to preheat the fluidizing air.
Cost data for a TNT slurry preparation building containing two
1.89 m (500-gal) mixing tanks with pumps and pneumatic mixers, instrumen-
tation, piping, design an construction, grinder, conveyors, building and
117
-------
miscellaneous items, the main fluidized bed system, and the optional heat
exchanger were estimated at $685,000. The cost data are tabulated below:
Projacted Costs 112.5 kg/hr TNT Slurry^
Fluidized Bed System (1877)
Main fluidized bed system $400,000
Slurry preparation building 210,000
Heat exchanger 75,000
$685,000
f) Source: R. Scola et al. U.S. Army ARRADCOM, Dover, New Jersey
Investment and Consumption Data for HC1 Pickling Liquor
The following investment and consumption data were estimated by Amer-
ican Lurgi Corporation in 1972 based on a possible U.S. installation
handling 30 metric tons/nr of spent HC1 pickling liquor with a feed rate
of 53 I/minute:
Investment $480,000 (1972)
Material Input and Output
Feed (1/min) 52.6
Rinse water (1/min) 50.3
Process water (1/min) 3.8
Power (watts) 165.0
Fuel (sm3/min) 4.9
Product HC1 (18%-1/min) 54.6
Product iron oxide (kg/min) 7.0
This cost reflects engineering costs and equipment costs within the
battery limits. Site preparation, building structure, and erection costs
were not included (50).
Investment and Process Characteristics of a Blasting Abrasive Fluidized Bed
Personnel at the David Taylor Naval Ship Research and Development
Center in Annapolis, Maryland, evaluated a fluidized bed for the combus-
tion of spent blasting abrasive contaminated with organotin paint, in
which the blasting abrasive served as the bed material. A design basis
for a process model was established at a maximum feed rate of 45 metric
tons/8 hr of free-drained abrasive containing 11 wt% wster, and paint
contaminant at 0.6%. A fluidizing air velocity of 75 m/minute was
suggested. Paint particle concentration ratios of 5:1 and 10:1 were
considered. At the 10:1 paint ratio, the process is autogenic; no aux-
iliary fuel is needed. Proposed design and cost data are cited in Table 15.
Much lower costs are evident for the 10:1 paint ratio (51).
118
-------
Consideration should also be given to laboratory costs incurred prior
to scale-up. The efficiency of fluidized bed combustion technology for
hazardous wastes depends on establishing ideal combustion parameters and a
suitable med medium for the waste of interest.
To summarize, the economics for fluidized bed technology has been
established for some hazardous wastes, and proposed for others. A fluid-
ized bed can function as a heat sink, and heat recovery is possible. Based
on these considerations, the application of fluidized bed combustion to
certain types of hazardous waste can be considered economically viable.
TABLE 15
PROCESS CHARACTERISTICS FOR A PROPOSED BLASTING ABRASIVE FLUIDIZED BED
(538"C)
Hours of
Operation
Equipment Size
Diameter, Meters
Inside Outside
Fuel/Power
Operating
Costs
$K (1977)
Estimated
Captial
Costs
$K (1977)
3
16
8
16
16
BULK ABRASIVE 1:1
2.4 3.2 0.165
1.7 2.6 0.165
CONCENTRATION 5:1
10.8 2.0 0.024
5.1 5.1 0.024
CONCENTRATION 10:1
0.75 1.8 0.006^
0.24 1.3 0.006
390
240
175
125
180
115
(*) Assumes fuel costs of $15 per barrel, and power at 5£ per kilowatt hour
(») Assumes cost of concentrating operation at $45,000 (1977)
Source: A. Ticker et al. David Taylor Naval Ship Research
and Development Center.
UV/OZONE DESTRUCTION
Many complex decisions representing trade-offs are necessary to im-
plement a well-designed, efficiently operated, economic UV/ozonation
system. UV/ozonation is generally restricted to waste streams of 1% or
lower hazardous contaminant. Because ozone is non-selective as an oxidant,
the waste stream should primarily contain the waste of interest. Effluent
streams must be carefully monitored for toxic intermediates. This will
add to the cost of. the process. Modern systems are usually automated,
thereby requiring lower labor costs.
119
-------
Ozone Generation
Ozone rust be generated on site to treat hazardous waste, because
there is no practical method to store or ship it. Most ozone generators
employ high voltage, although equipment for small outputs is available.
Ozone generation using oxygen is about twice as efficient as air, yet
oxygen costs about twice as much as air. Ozone generators can be air
cooled or water cooled, yet experience has shown than the more expensive
water cooling system has less problems and is a more trouble free system
according to Klein ^t al. (25).
Power
The major operating cost for ozone manufacturing is the cost of
electric power. Of the total power utilized in ozone generation, approxi-
mately 10% actually produces the ozone. The rest is consumed in air
handling, air preparation, and waste heat. Typical power consumption
figures in plants using air as the generator range from 6-8 kwh/lb ozone
for the ozone generator alone, and 10-13 kwh/lb ozone total consumption
including air handling and preparation. With oxygen as the feed gas for
generating ozone, typical power consumption figures range from 3-4 kwh/lb
ozone for ozone generation, and 7-12 kwh/lb ozone total consumption (30).
Although higher frequency generating systems are more efficient ozone
generators, the electrical cost of the high frequency can be considered to
offset the gain. New advances in solid state technology may improve the
situation (5). Ozonators which operate at high pressures require extra
electrical power to operate high pressure compressors. According to the
Ozone Research and Equipment Corporation, the cost of additional electric
power is lower than the labor cost and inconvenience associated with
periodic tear down, cleaning, and reassembly of the generators (27).
Destruct Systems
In some cases, ozone not consumed in the primary reactor system can be
reused or applied in another system. It may be necessary to provide a
destruct system to guarantee removal of any unused ozone from the system
before the gas stream is discharged to the atmosphere. There is a choice
of three types of destruct systems thermal, catalytic, and combination.
Thermal systems heat the entire gas stream to a high temperature for a
specific time frame. Catalytic systems heat the gas to 250°C and then pass
the gas stream through a solid phase catalyst bed for ozone destruction.
The catalysts are proprietary. Combination systems exist. These systems
represent a trade-off between heating costs and catalysts costs. Heat
recovery should receive consideration (30).
UV Lamps
The cost for a bank of UV lamps is small compared to the other capital
costs of an UV/ozonation system. Increased UV light can favorably acceler-
ate ozone decomposition, but at a trade-off in higher cost requirements.
Over a period of time, a thin film builds up on the UV lamps and impairs
their efficiency. Consideration must be given to implementing a system to
120
-------
wipe the lamps or use extra labor. Because of the wide variation in
reactivities of organic contaminants, UV levels cannot be generalized,
and should be determined for each waste of interest. Since ozone is more
stable in acidic solutions, neutralization may seriously affect ozone
levels, thereby affecting UV requirements, and reducing cost-effectiveness
of the system,
Capital Costs and Operating Costs
Predicted capital and operating costs for a proposed pilot plant
treating 150,000 gallons per day of PCB's at levels of 50 ppm reduced to
values less than or equal to 1 ppm are shown -in Table 16 (48). Pertinent
data comparing a 40,000 and 150,000 gallons per day plant treating PCB's
at feed levels of 50 ppm reduced to 1 ppm PCB are shown in Table 17 (48).
Summary of UV/Ozonation Economics
The data presented here demonstrate the economic viability for an
UV/ozone system to treat certain hazardous wastes. Typically capital
costs for scale-up are not directly proportional to the increased treatment
capacity, e.g., a system scaled up to 3.75 times as large costs only 2.4
times moxe. The daily operating costs for the larger (3.75X) system versus
the smaller system are also not proportional. The 3.75X system only costs
nearly 2X the smaller system on a daily operating basis.
TABLE 16 *
MINIMUM OPERATING AND CAPITAL COSTS FOR A 150,000 GPD ULTRQX TREATMENT
PILOT PLANT TO OBTAIN < 1 ppm PCB
150,000 Gallons per Day Plant (567,750 liters/day)
Total
Total Capital Costs Operating Cost PCB levels Ozone Number
S/1,000 gal ppm wt% of Lamps
350,280
300,320
300,320
300,320
300,320
1.91
1.71
1.71
1.71
1.71
0.7
0.9
0.9
0.9
1.0
1.0
1.0
1.0
1.0
1.0
15
15
15
15
15
(*) Source; Arisman and Musick, General Electric Company, Zeff and Crase,
Westgate Research Corporation, West Los Angeles, California (1980).
121
-------
TABLE 17 *
DESIGN SPECIFICATIONS, CAPITAL, AND O&M COSTS FOR
40,000 AND 150,000 GPD ULTROX TREATMENT PLANTS
" (50 ppm PCS feed-1 ppm PCS effluent)
DESIGN SPECIFICATIONS
40,000 GPD 150,000 GPD
(151,400 LPD) (567,750 LPD)
Reactor Automated System Automated System
Dimension, Meters (LxWxH) 2.5 x 4.9 x 1.5 4.3 x 8.6 x 1.5
Wet Volume, Liters 14,951 56,018
UV Lamps; Number 65 W 378 1179
Total Power, kw 25 80
Ozone Generator
Dimensions, Meters (LxWxD) 1.7 x 1.8 x 1.2 2.5 x 1.8 x 3.1
kg Ozone/day 7.7 28.6
Total Energy required 768 2544
(kwh/day)
BUDGETARY EQUIPMENT PRICES
40,000 GPD 150,000 GPD
Reactor $94,500 $225,000
Generator 30,000 75,000
TOTAL $124,500 TOTAL $300,000
O&M Costs/Day
Ozone Generator Power $4.25 $15.60
UV Lamp Power 15.00 48.00
Maintenance
(Lamp Replacement) 27.00 84.20
Equipment Amortization
(10 Yrs <§ 10%) 41.90 97.90
Monitoring Labor 85.71 85.71
TOTAL/DAY $173.86 $331.41
Cost per 3785 Liters
(with monitoring labor) $4.35 $2.21
(without monitoring labor) $2.20 $1.64
(*) Source: Arisman and Musick, General Electric Company, Zeff and Crase
Westgate Research Corporation, West Los Angeles, California (1980)
122
-------
SECTION 8
SURVEY OF HAZARDOUS WASTE GENERATORS
PURPOSE
In conjunction with the assessment of emerging hazardous waste des-
truction technologies, Ebon Research Systems conducted a survey of both
large and small hazardous waste generators. The aims of the survey were:
e To determine user needs for the application of treatment
techniques to pollutant disposal problems.
® To determine if the hazardous waste generators had access to,
or desired access to computerized services and/or a newsletter
regarding hazardous waste disposal.
o To inform hazardous waste generators that new emerging techni-
ques are available for hazardous waste destruction.
Ebon Research Systems was also interested in compiling the response,
and determining the overall level of interest from those companies surveyed.
SURVEY RESULTS
The instrument for the survey is show in Figure 35. Except for
question 6 which asked for additional information, the replies were limited
to a yes or no. Respondents were informed that the names of individual
companies would be kept confidential. Ebon Research Systems believed this
simple approach would be more likely to elicit a higher response level.
The survey was mailed in January, 1980, to the 53 chemical companies
participating in the waste disposal site survey conducted by the House of
Representatives Subcommittee on Oversight and Investigation, of the
Committee on Interstate and Foreign Commerce, 96th Congress, First Session.
The Committee derived the list of participants (considered to be the 53
largest companies in terms of domestic sales) frcm the 1976 Kline Guide to
the Chemical Industry, and the American Chemical Society's 1977 listing of
the top 50 chemical companies. Because the initial response to the survey
was low, more companies were added to the list by Ebon Research Systems.
Even though the mailing was followed up by telephone request, the total
number of companies responding was 31 out of 73 contacted. Of these
companies, only 18 were on the original Subcoranitte's listing of 53.
The response matrix relating to user needs is shown in Table 18.
123
-------
TABLE 18
MATRIX RELATING TREATMENT TECHNIQUES TO USER NEEDS
Treatment Technique % of Users Requesting Information
Conventional incineration 68
Fluidized bed combustion 55
Secured landfills 84
Ocean incineration 32
Ozonation 32
Chlorinolysis 26
Molten salt combustion 32
UV destruction 35
The following observations were pertinent:
• Only one company was interested in landfarming.
• The most interest was expressed in conventional established
technologies.
• Sixty-five percent of respondents already subscribed to a
newspaper discussing hazardous waste at least in part, but
only 19% had access to a computerized service regarding
hazardous waste disposal, at least in part.
• Fifty-eight percent indicated they would use both a computer-
ized data bank and newsletter regarding hazardous waste, while
16% stated they would not be interested in either.
• More than 80% of the companies expressed interest in more
information regarding hazardous waste legislation, economics
of waste disposal and transport, spill cleanup techniques, and
the technology for disposal or reclamation. •
Eighteen companies responded to Question 6: "Is there any other way
that a hazardous waste information service could be of use to your company?"
The major interests are summarized below:
Names and locations of authorized hauling contractors
Permit information for companies which operate in more than one
state
Waste exchange and recycling information
Packaging and transportation regulation information
Strongly publicize significant hazardous waste control measures
implemented by industry
Computerized data bases are needed for information on aerobic or
anaerobic degradation of organic chemicals.
Btu or fuel values of combustible products are needed so that an
economic assessment of various disposal practices can be made.
124
-------
FIGURE 35
EBON RESEARCH SYSTEMS' HAZARDOUS WASTE INFORMATION SERVICE QUESTIONNAIRE
Barbara H. Edwards,
Please return to: 1542 9th St., N.W.
Washington, D.C. 2001
Please indicate responses with an X. YES NO
I. Does your company now have access to
computerized services (data banks, etc.)
on hazardous waste disposal?
20 Does your company currently subscribe
to a news letter on hazardous waste disposal?
3o If the USEPA established a data bank or
quarterly newsletter on hazardous waste
disposal, which would be of most use to your
vour companv?
(a)
Cb)
(0
(d)
4. Pisa
that
(e)
(f)
(g)
(h)
data bank only
newsletter onlv
both data bank and newsletter
would not use either data bank or newsletter
se indicate any of the following topics
world be of interest to your company
hazardous waste legislation
economics of hazardous waste disposal
and transport
hazardous waste spill cleanup
technology of hazardous waste
reclamation
techniques
disposal
50 Would up-to-date information of any of the
specific technologies listed below be of
interest to your company?
(i) conventional incineration
(j? fluidized bed combustion
125
-------
FIGURE 35 (Continued)
EBON RESEARCH SYSTEMS' HAZARDOUS WASTE INFORMATION SERVICE QUESTIONNAIRE
Please indicate responses with an X. YES NO
(information on specific technologies-cont.)
(k) secured landfills
(1) ocean incineration
(m) ozonation
(n) chlorinolysis
(o). molten salt combustion
(p) UV destruction
(q) other (please specify)
6. Is there any other way that a hazardous waste information waste
information service could be of use to your company?
Use the back of the questionnaire or another sheet for further conments.
THANK YOU
126
-------
REFERENCES
1o Greenberg, J. 1972. Method of Catalytically Inducing Oxidation of
Carbonaceous Materials by the Use of Molten Salts. United States
Patent 3,647,358, assigned to Anti-Pollution Systems, Inc.,
Pleasantville, New Jersey, 16 pp.
2. Greenberg, J. and D.C. Whitaker. 1972. Treatment of Sewage and Other
Contaminated Liquids With Recovery of Water by Distillation and
Oxidation. United States Patent 3,642,583, assigned to Anti-Pollution
Systems, Inc., Pleasantville, New Jersey, 10 pp.
3. Findlay, A. et al_. 1951. Phase Rule. Dover Publications, Inc., New York,
New York, 494 pp.
4. Grsenbarg, J. 1979, The use of molten salts in emission control.
Presented at the 72nd Annual Meeting of the Air-Pollution Control
Association, Cincinnati, Ohio, June 24-29.
5. Heredy, L.A. et al. 1969. Removal of Sulfur Oxides From Flue Gas.
United States Patent 3,438,722, assigned to North American Rockwell
Corporation, California, 12 pp.
6. Yosim, S.J, et al. 1979. Molten salt destruction of hazardous wastes
produced in the laboratory. In: Safe Handling of Chemical Carcinogens,
Mutagens and Teratogens: The Chemists Viewpoint, Symposium held at the
American Chamical Society, Hawaii.
7, Yosim, S,Jo et al. 1979. Disposal of hazardous wastes by molten salt
combustion. In: Ultimate Disposal of Hazardous Waste, Symposium held
at the American Chemical Society, Hawaii.
80 Yosim, So Jo et al. 1978. Destruction of pesticide and pesticide
containers by molten salt combustion. In: American Chemical Society
Symposium Series, No. 73, Disposal and Decontamination of Pesticides,
MoVo Kennedy, ed., pp. 118-130
9o Smithson, GoR, 1977. Utilization of energy from organic wastes through
fluidized bed combustion. In: Fuels from Waste, L.L. Anderson and
D.Ao Pullman, eds., Academic Press, New York, New York. pp. 195-209
10, Botterill, J.S. 1977, The contribution of fluid-bed technology to energy
saving and environmental protection. Applied Energy, 3:139-150.
11o Yerushalmi, Jo and N.T. Cankurt. 1978. Higb-velocitv fluid beds.
Chemtech., 8(9):564-571.
127
-------
12. Oehlschlaeger, H.F. 1976. Reactions of ozone with organic compounds.
Proceedings of an Ozone Conference, Cincinnati, Ohio.
13. Bailey, Philip S. 1973. Activity of ozone with various organic functional
groups important to water purification. Presented before the First
International Symposium on Ozone for Water and Wastewater Treatment.
14. Prengle, H.W. and C.E. Mauk. 1978. New technology: ozone/tJV chemical
oxidation wastewater process for metal complexes, organic species and
disinfection. The American Institute of Chemical Engineers Symposium
Series, 74:228-243.
15. Prengle, H.W. et al. 1975. Ozone/UV process effective wastewater
treatment. Environmental Management, 54:82-87.
16. Lee, M.K. 1980. Study of UV-ozone reactions with organic compounds
in water. Presented Before the Division of Environmental Chemistry,
American Chemical Society, Houston, Texas.
17. Yosim, S.J. and K.M. Barclay. 1978. Destruction of hazardous wastes by
molten salt combustion. In: Proceedings of a National Conference
About Hazardous Waste Management, San Francisco, California, pp. 146-156.
18. Wilkinson, R.R. et al. 1978. State-of-the-Art-Report, Pesticide Disposal
Research. EPA-60012-78-183, U.S. Environmental Protection Agency,
Cincinnati, Ohio, 225 pp.
19. Kahl, A.L. et al. 1978. The molten salt coal gasification process.
Chemical Engineering Progress, August, pp. 73-79.
20. Yosim, S.J. et al. 1973. Non-polluting disposal of explosives and
propellants. United States Patent 3,778,320, assigned to Rockwell
International Corporation, California.
21. Yosim, S.J. et al_. 1974. Disposal of Organic Pesticides. United States
Patent 3,845,190, assigned to Rockwell International Corporation,
El Segundo, California, 10 pp.
22. Grantham, L.F. et al. 1979. Disposal of PCS and other toxic hazardous
waste material by molten salt combustion. In: National Conference on
Hazardous Risk Assessment, Disposal and Management Proceedings, Miami,
Florida, pp. n.s.
23. Dustin, D.F. et al. 1977. Applications of molten salt incineration
to the demilitarization and disposal of chemical material. EM-TR-76099,
Department of the Army, Edgewood Arsenal, Aberdeen Proving Ground,
Maryland, 55 pp.
24. Powers, P.W. 1976. How to Dispose of Toxic Substances and Industrial
Wastes. Noyes Data Corporation, Park Ridge, New Jersey. 497 pp.
128
-------
25. Klein, M.J. et aJU 1973. Generation of ozone. Presented before the
First International Symposium on Ozone for Water and Wastewater
Treatment.
26. Anonymous. Comparative Data. Ozone Generators and Air Pre-Drying
Equipment, brochure, Ozonair Corporation, San Francisco, California.
27. Anonymous. Ozone Technology, brochure 124, Ozone Research and Equipment
Corporation, Phoenix, Arizona.
28. Bollyky, L.J. 1973. Ozone treatment of cyanide and plating wastes.
Presented before the First International Symposium for Water and
Wastewater Treatment.
29. Augugliaro, V. and L. Rizzuti. 1978. The pH dependence of the ozone
absorption kinetics in aqueous phenol solutions. Chemical Engineering
Science 33:1141-1147.
30. Derrick, D.H. and J.R. Perrich. 1979. Guide to ozone equipment
selection. Pollution Engineering, 11:42-44.
31. Hammond, V.L. and L.K. Mudge. 1975. Feasibility Study of Use of Molten
Salt Technology for Pyrolysis of Solid Waste. EPA-67012-75-014,
Prepared by Battelle Pacific Northwest Laboratories for the U.S.
Environmental Protection Agency, Cincinnati, Ohio, 81 pp.
32. Bliss, C. and B.M. Williams, eds. 1977. Proceedings of the Fifth
International Conference on Fluidized Bed Combustion, Washington, D.C.
33. Wall, C.J. et _al_. 1975. How to burn salty sludges. Chemical
Engineering, 80(8): 77-82.
34. Landreth, R.E., and C.J. Rogers. 1974. Fluidized bed ccir>bustion of
phenol and methyl methacrylate wastes. In: Promising Technologies for
Treatment of Hazardous Wastes. EPA 670/2-74-088, U.S. Environmental
Protection Agency, Cincinnati, Ohio. pp. 78-87.
35. Ragland, K.W. and D.P. Paul. 1979. Fluidized bed combustion of plastic
with coal. Presented at the Fourteenth Intersociety Energy Conversion
Conference, Boston, Massachusetts.
36. Kamino, Y. et al. 1978. Gasification of waste plastics. Technical
Report of tEe Hitachi Ship Building Technology Research Institute,
Japan, pp. 16-21.
37. Eggers, F.W0 et al_. 1977. Removing Chlorine-Containing Insulation with
a Fluidized Medium Containing Reactive Calcium Compounds. United
States Patent 4,040,868, assigned to Cerro Corporation, New York,
New York. 8 pp.
38. Walker, W.M. 1973. Fluid bed incineration of chlorinated hydrocarbons.
Australian Mines Development Laboratory Bulletin, 16:41-44.
129
-------
39. Ziegler j2t al. 1973. Fluid bed incineration. RFP-2016, Technical Report
of Dow Chemical U.S.A., Rocky Flats Division, Golden, Colorado. 13 pp.
40. Ziegler et al. 1974. Pilot plant development of a fluidized bed inciner-
ation process. RFP-227, Technical Report of Dow Chemical U.S.A., Rocky
Flats Division, Golden, Colorado. 10 pp.
41. Scola, R. et al. 1978. Fluidized bed incinerator for disposal of propel-
lants and explosives. Technical Report of the U.S. Armament Research
and Development Command, Dover, New Jersey. 114 pp.
42. Carroll, J.W. et al. 1979. Assessment of hazardous air pollutants from
disposal of munitions in a prototype fluidized bed incinerator.
American Industrial Hygiene Association Journal, 40:147-158.
43. Wong, A.S. et al. 1979. Ozonation of 2,3,7,8-Tetrachlorodibenzo-p-
Dioxin. Presented at the Symposium on the Chemistry of Chlorinated
Dibenzodioxins and Dibenzofurans, ACS National Meeting, Washington,
. . D.C.
44. Sierka, R.A. and W.F. Cowen. 1980. The catalytic ozone oxidation
of aqueous solutions of hydrazine, monomethyl hydrazine and
unsymmetrical dimethylhydrazine. Presented at the 35th Annual
Purdue Industrial Waste Treatment Conference, Layfayette, Indiana.
45. Cowen, W.F. et al. 1979. Identification of the partial oxidation
products of hydrazine, monomethyl hydrazine and unsymmetrical dimethyl
hydrazine from ozonation. Paper presented at the USAF Engineering and
Service Center, Tyndall Air Force Base, Florida.
46. Leitis, E. et al. 1979. An investigation into the chemistry of the
UV/Ozone purification process. Presented at the 4th World Ozone
Congress, Houston, Texas.
47. Macur, G.J. et jd. 1980. Oxidation of organic compounds in concentrated
industrial waste water with ozone and ultraviolet light. Presented at
the 35th Annual Purdue Industrial Waste Conference, Layfayette, Indiana.
48. Arisman, R.K. and R.C. Musick. 1980. Experience in operation of a
UV-Ozone (ULTROX) pilot plant for destroying PCB's in industrial
waste effluent. Presented at the 35th Annual Purdue Industrial Waste
Conference, May, 1980.
49. Ross, R.D. 1979. Incineration-a positive solution to hazardous waste
disposal. Presented at The American Chemical Society Division of
Chemical Health and Safety Meeting, Washington, D.C.
50. Marnell, P. 1972. Spent HC1 pickling liquor regenerated in a fluid bed.
Chemical Engineering, 79(25):102-103.
51. Ticker, A. et al_. 1979. Study of the fluidized bed process for treatment
of spent blasting abrasives. Journal Coating Technology, 49(626):29-35.
130
-------
52. Sittig, M. 1979. Incineration of Industrial Hazardous Wastes. Noyes
Data Corporation, Park Ridge, New Jersey. 348 pp.
53, Landreth, R.E. and C.J. Rogers. 1974. Wet air oxidation. In: Promising
Technologies for Treatment of Hazardous Wastes. EPA-670/2-74-088, U.S.
Environmental Protection Agency, Cincinnati, Ohio.
54. Dresen, R.W. and J.R. Mayer. 1972. Wet Combustion of Organics. United
States Patent 3,984,311, assigned to Dow Chemical Company.
55„ Miller, R.A. et al. 1980. Destruction of toxic chemicals by catalyzed
wet oxidation. Presented at the Purdue Industrial Waste Conference,
West Layfayette, Indiana.
56. Kitchens, J.A. 1979. Dehalogenation of Halogenated Compounds. United
States Patent 4,144,152, assigned to Atlantic Research Corporation,
Alexandria, Virginia.
57. Jones, W.E. et a^. 1980. Light activated reduction of chemicals.
Presented before the Division of Environmental Chemistry, American
Chemical Society, Houston, Texas.
58. Trump, J.G. e_t al. 1979. Destruction of pathogenic microorganisms and
toxic organic chemicals by electron treatment. Presented at the Eigth
National Conference on Municipal Sludge Management, Miami, Florida.
59. Fochtman, E.G. et al. 1979. Chlorinolysis treatment of hydrazine in
dilute aqueous solution. Presented at Symposium of Environmental
Chemistry of Hydrazine, Tyndall Air Force Base, Florida.
60. Yasuhiro, H. et al. 1979. Conversion of polychlorinated biphenyls to
useful materials by a catalytic hydrogenation-dechlorination method.
Presented at the American Chemical Society Meeting in Honolulu, Hawaii.
131
-------
APPENDIX A
EUTECTIC MIXTURES OF NEUTRAL SALTS
Mixtures with KC1
41 M% KC1-59 M% LiCl
57 M% KC1-43 M% BaCl2
60 M% KC1-40 M% CaCl2
35 M% KC1-65 M% CdCl-
40 M% KC1-60 M% MgCl2
55 M% KC1-45 M% SrCl-
50 M% KC1-50 M% MnCl2
48 M% KC1-52 M%
45 M% KC1-55 M%
Mixtures with LiCl
72 M% LiCl-28 M% NaCl
28 M% LiCl-62 M% CaCl
45 M% LiCl-55 M%
45 M% LiCl-55 M%
f.
45 M% LiCl-55 M% SrCl2
Mixtures with Barium Chloride
12 M% BaCl2-88 M%
30 M% BaCl2~70 M%
Eutectic Temperature '
358 °C
345°C
580°C
380°C
420 °C
575°C
500°C
411°C
230°C
Eutectic Temperature
560°C
496°C
550°C
410°C
475°C
Eutectic Temperature
390°C
600°C
132
-------
APPENDIX A (Continued)
EUTECTIC MIXTURES OF NEUTRAL SALTS
Mixtures with Lead Dichloride
20 MS BaCl2-80 M%
48 MS BeCl2-52 MS PbCl2
90 MS BiCl2-10 MS PbCl2
18 MS CaCl2~82 MS
35 MS CdCl2-65 MS PbCl2
33 MS CuCl-67 MS PbCL,
50 MS FeCl3-50 MS
43 MS KC1-52 MS
45 MS LiCl-55 MS
8 MS MgCl?-92 MS PbCl2
30 MS MnCl -70 MS PbCl2
23 M% NaCl-72 MS PbCl
72 MS PbF.,-28 MS
75 MS PbI.,-24 MS
50 MS SnCl2-50 MS PbCl2
60 MS T1C1-40 MS PbCl2
50 MS ZnCl2-50 MS PbCl2
Eutectic Temperature
505°C
300eC
205°C
460° C
387°C
285"C
185°C
411°C
410°C
450° C
405°C
415°C
550°C
310°C
410°C
390"C
340 °C
Mixtures with Magnesium Chloride
8 MS MgCl2-92 MS PbCl2
55 MS MgCl2-45 MS SrCl2
Eutectic Temperature
460°C
530°C
133
-------
APPENDIX A (Continued)
EUTECTIC MIXTURES OF NEUTRAL SALTS
Mixtures with Sodium Chloride
33 M% NaCl-67 M%
42 M% NaCl-58 M%
48 M% NaCl-52 M%
60 M% NaCl-40 M% MgCl.
55 M% NaCl-45 M% MnCl2
45 M% NaCl-55 M% NiCl2
28 M% NaCl-72 M% PbCl
50 M% NaCl-50 M% SrCl,,
Eutectic Temperature
550°C
397°C
365°C
450°C
415°C
560eC
415°C
560°C
Mixtures with Calcium Chloride
47 M% CaCl2-63 M%
18 M% CaCl2-82 M%
50 M% CaCl_-50 M% ZnCl
Mixtures with Cadmium Chloride
50 M% CdCl2-50 M%
35 M% CdCl2-65 M%
60 M% CdCl2-40 M%
50 M% CdCl2-50 M%
Mixtures with Zinc Chloride
38 M% ZnCl2-62 M%
45 M% ZnCl2-55 M% SrCl2
Eutectic Temperature
600°C
470°C
600°C
Eutectic Temperature
600°C
387°C
500°C
500°C
Euctectic Temperature
180°C
480°C
134
-------
APPENDIX A (Continued)
EUTECTIC MIXTURES OF NEOTRAL SALTS
Mixtures with Potassium Bromide
31 M% KBr-69 M% LiBr
60 M% KBr-40 M%
32 M% KBr-68 M%
50 M% KBr-50 M%
Mixtures with Sodium Bromide
40 M% BaBr2-60 M% NaBr
60 M% CaBr -40 M% NaBr
45 M% CdBr2-55 M% NaBr
40 M% MgBr2~60 M% NaBr
60 M% SrBr -40 M% NaBr
Mixtures with Lead Dibromide
80 M% 3i3r3-20 M%
18 M% CdBr -82 M% ?bBr
90 M% HgBr -10 M% FbSr
50 M% PbCl2-50 M% PbBr2
78 M% PbF2~22 M%
10 M% PbF -90 M%
44 M% Pbl -56 M% PbBr
Mixtures with Lithium Bromide
80 M% LiBr-20 M% NaBr
75 M% LiBr-25 M%
Eutectic Temperature
310eC
325'C
350"C
525°C
Eutectic Temperature
600eC
510°C
370°C
425°C
480 °C
Eutectic Temperature
200°C
344°C
208 °C
425°C
520°C
350°C
282°C
Eutectic Temperature
525'C
485°C
135
-------
APPENDIX A (Continued)
EUTECTIC MIXTURES OF NEUTRAL SAUS
Mixtures with 3 Chlorides
35 M% PbCl2-35 M% KC1-30 M%
60 M% KC1-20 M% PbCl2-20 M% NaCl
10 M% PbCl2-50 M% KC1-40M%
10 M% NaCl-40 M% LiCl-50 M% KCl
70 M% LiCl-15 M% CaCl2-15 M% KCl
30 M% BaCl2-35 M% CaCl2-55 M% KCl
10 M% NaCl-35 M% CaCl2~55 M% KCl
15 M% NaCl-50 M% CdCl2~35 M% KCl
10 M% NaCl-15 M% PbCl2~75 M% KCl
Eutectic Temperature
328eC
500°C
280°C
400°C
450°C
542°C
600°C
450°C
500°C
Mixtures with 3 Salts
12.5 M% NaBr-12.5 M% CdBr2~75 M%
65 M% PbI2-13 M% PbCl2-17 M%
Eutectic Temperature
280°C
. 300°C
Mixtures with 4 Salts
25 M% CdCl2-25 M%
-25 M%
M%
10 M% NaCl- 10 M% NaBr-60 M% PbCl2~20 M%
10 M% NaCl-10 M% NaI-40 M% PbCl -40 M% Pbl
Sutectic Temperature
450° C
450° C
3658C
136
-------
APPENDIX B
EOTECTIC MIXTURES OF ACTIVE
Oxides
43 M% W03-57-M%
47 M% MOO..-53 M%
45 M% 3
• As205-55 M%
77 M% Mo03-23 M%
38 M% VD--62 M%
8 M% CaO-92 M% P
50 M% PbO-50 M%
lowest Melt Temperature
600°C
530°C
570 8C
510°C
570 °C
409°C
4808C
Three Oxides
37 M% K2Mo04(K20 + Mo03)-63 M% Li2^cO^
50 M% KPO -50 M% LiP03
(Consists of K2O,
49 M% NaPO3-49 M% KPO.j-2 M% K2O
50 M% Li2M04-50 M% Na2Mo04
50 M% ^VCL-SO M% Li2W04
50 M% Pb2Si04-50 M% Na2Si03
33 M% NaJD • Si02~67 M% PbO : SiO
Perchlorates
lowest Melt Temperature
525°C
562°C
Li20, P20.)
547 °C
465°C
500°C
600°C
580° C
lowest Melt Temperature
70 M% LiC104-30 M%
40 M% NaC104-60 M%
185°C
305'C
137
-------
APPENDIX B (Continued)
EUTECTIC MIXTURES OF ACTIVE SALTS
Sulfate Mixtures
33 M% K SO.-67 M%
25 M% K SO.-25 M%
25 M% K2S04~25 M% Li2SO4~25 M%
25 M% Li SC--25 M%
Nitrates
50 M% KN03-50 M%
35 M% KN03-65 M% TINCu
85 M% KN03-15 M%
50 M% Ca(N03)2-50 M%
50 M% LiNCL-50 M% Ba(NO ),
J -J 6f
50 M% Ca(NO ) -50 M% IOSD
70 M% LiN03~30 M% Ca(NO.,)2
50 M% Ba(NO,).-50 M% Ca(NO,)_
j Z o 2.
60 M% T1(N03)2-40 M%
40 M% KN03~35 M% LiNO.,-25 M%
M% KNO.j-25 M% LiNC>3
60 M% KNCL-25 M% LiNO3-15 M%
20 M% LiN03-60 M%
~20 M% KNO.,
34 M% Ba(N03)2-33 M% KND.j-33 M% NaNO,
34 M% Ca(N03)2-33 M% KN03-33 M% NaN03
34 M% Cd(N03)2-33 M% NaNO.j-33 M% Li
MS Li2V
,-25 M%
5 M%
-95 M% NaN0
Lowest Melt Temperature
470"C
460°C
570°C
520 °C
4
Lowest Melt Temperature
222°C
185°C
280 °C
240° C
400 °C
240°C
240°C
520°C
140°C
130°C
125°C
140°C
450eC
140"C
176°C
300°C
138
-------
\
APPENDIX B (Continued)
EUTECTIC MIXTURES OF ACTIVE SALTS
Hydroxides
50 M% NaNO.j-50 M% NaDH
50 M% KNO--50 M% KOH
50 M% KOH-50 M%
o
70 M% KOH-30 M% LiOH
50 M% KOH-50 M% NaNO..
50 M% KOH-50 M% NaOH
15 M% K2Cr04-85 M% KOH
70 M% NaOH-30 M% LiOH
50 M% NaOH-50 M%
50 M% KN03-50 M% NaOH
62 M% Ba(OH)2-38 M% Sr(OH)2
30 M% KOH-30 M% NaOH-40 M% LiOH
33 M% NaOH-34 M% N^SO^SS M% NaCl
25 M% LiOH-25 M% Li2Cr04~25 M% NaOH-25 M%
25 M% LiOH-25 M% NaOH-25 M% LiNO-25 M%
Lowest Melt Temperature
270 °C
200°C
235°C
110°C
240°C
170°C
361°C
210°C
266°C
330°C
360°C
300°C
500°C
475°C
400°C
Sul fates
60 M%
-40 M% CoS0
55 M% K2S04-45 M%
55 M% Li2SO4-45 M%
60 M% Li
M% CoS0
35 M% MnS04-65 M% Li2S04
50 M% ZnSO.-50 M% Na^SO,
4 ^4
Lowest Melt Temperature
540°C
460°C
575° C
600°C
600°C
500°C
139
-------
APPENDIX B (Continued)
EUTECTIC MIXTURES OF ACTIVE SALTS
Nitrites
70 M% NaN02-30 M%
Nitrites-Nitrates
50 M% KN02-50 M% KN03
62 M% KN02-38 M%
50 M% NaN02-50 M% KNC>3
50 M% NaN02~50 M% NaNO,
50 M% NaN02-50 M%
25 M% NaN02-25 M%
M% NaNO.j-25 M%
25 M% KN02-25 M%
-25 M%
lowest Melt Temperature
230°C
lowest Melt Temperature
350°C
140°C
1508C
230eC
190°C
150°C
300°C
140
-------
APPENDIX C
HAZARDOUS WASTES DESTROYED BY THE EMERGING TECHNOLOGIES
Hazardous Wastes Destroyed by Molten Salt Combustion
PCB's
Chloroform
Perchloroethylene distillation bottoms
Trichloroethane
Tributyl Phosphate
Nitroethane
Monoethanolamine
Diphenylamine HCl
Rubber tire buffings
Para-Arsanilic Acid
Contaminated Ion Exchange Resins (Dowex and Powdex)
High-Sulfur Waste Refinery Sludge
Acrylics Residue
Tannery Wastes
Aluminum Chlorohydrate
Pesticides and Herbicides
Chlordane
Malathion
Weed B Gon
Sevin
DDT Powder
DDT Powder plus Malathion Solution
2,4-D Herbicide-Tar Mixed Waste
Real and Simulated Pesticide Containers
plastic, rubber, and a blend of these
Feasible Pesticides and Nitrile Herbicides
Pesticides Nitrile Herbicides
dieldrin trifluralin
heotachlor 2,4,5-T dichlorobinil
aldrin MCPA
tolaidine
141
-------
APPENDIX C (Continued)
HAZARDOUS WASTES DESTROYED BY THE EMERGING TECHNOLOGIES
Hazardous Wastes Destroyed by Molten Salt Combustion
Phosphorous Insecticides (feasible)
diazinon
disulfonton
phorate
parathion
Explosives and Propellants
TNT
glyceryl nitrate
diglyceryl tetranitrate
glycol dinitrate
triroethylolethane trinitrate
diethylene glycol dinitrate
PETN
DPEHN
Tetryl
Cyclonite
HMX
Conposition B
Feasible
JP type hydrocarbon fuel
ethyl alcohol
hydrazine and its derivatives
Chemical Warfare Agents Destroyed
G3
GB Spray-dried salts
Distilled Mustard, HD
VX
Lewisite, L
Toxic Gas Identification Sets (Real and Sinulated)
made of pyrex, wood, plastic, tin-plates steel, and agent
142
-------
APPENDIX C (Continued)
HAZARDOUS WASTES DESTROYED BY THE EMERGING TECHNOLOGIES
Hazardous Wastes Destroyed by Fluidized Bed Incineration
HC1 pickling liquor (spent)
Organotin in spent steel slag blasting abrasive
Organic dye slurries
red dye slurry (1-methylaminoanthraquinone and starch gum)
yellow dye slurry (dibenzpyrenequinone and benzanthrone)
Chlorinated Hydrocarbons
PVC waste from a chemical plant
PVC mixed with coal
PVC insulation over copper wire
. Chlorinated hydrocarbon waste containing 80% chlorine
Munitions (slurry)
TOT
RDX (cyclotrimethylenetrinitramine)
Composition B
Hazardous Wastes Destroyed by UV/ozonation Technology
PCB's
TCDD (2,3,7,8-tetrachlorod ibenzo-p-d ioxin)
OCDD (octachlorodibenzo-p-dioxin)
Chlorodioxins (other dioxins are feasible)
Hydrazine
Monomethyl hydrazine
Dimethyl hydrazine (unsymmetrical)
Copper process waste stream
Nitrobenzene
143
-------
APPENDIX D '
DESIGN-OF-EXPER1MENT TESTS FOR PCB DESTRUCTION IN 011E ULTKQX PILOT PIAMT
Exp.
M3.
1
2
3
4
5
6
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
Effl.
Flow
RaKe
1/foin
1.0
3.5
2.0
2.0
1.0
2.5
2 A
• U
1.7
1.2
1.5
2.2
1.8
1.3
0.8
1.6
1.4
2.0
1.5
1.0
1.2
0.8
1.0
1.3
1.6
Ozone
Flow
ing/fain
700
700
600
500
400
600
Tfrtrt
/UU
300
200
500
700
450
550
350
250
160
350
300
300
200
300
500
300
250
Ozone
Cone.
wt%
2.0
2.0
1.5
2.0
1.0
1.5
1 ft
1 .O
1.2
1.6
3.0
3.0
2.5
2.0
1.0
1.3
1.0
3.0
2.8
2.5
1.8
2.6
3.0
1.7
1.5
IJV
Arrangement
~I 2 3
10
10
5
5
10
10
10
10
0
5
10
5
5
5
4
10
10
9
4
10
0
5
3
9
9
5
5
9
5
4
9
9
7
9
9
5
5
4
8
0
4
4
0
0
0
3
10
10
5
5
10
0
10
0
10
10
5
10
5
10
10
0
10
5
4
0
0
5
4
Ozone Mass
How
Distribution
1 2 3
233 233 233
23.3 233 233
200 200 200'
167 167 167
133 133 133
200 200 200
100 100 100
67 67 67
200 200 100
200 200 300
200 150 100
200 200 150
125 100 125
75 75 100
100 30 30
150 100 100
100 100 100
160 80. 60
100 50 50
200 50 50
244 122 122
125 50 125
75 75 100
Feed
7.7
8.0
7.8
7.7
7.8
7.8
7.6
7.6
7.3
7.6
7.5
7.6
7.4
7.4
7.4
7.4
7.4
7.4
7.6
7.5
7.3
6.0
5.7
pH
Mid.
8.0
7.8
7.7
7.7
7.7
7.8
7.6
7.6
7.5
7.6
7.6
7.4
7.4
7.4
7.4
7.4
7.4
7.4
7.3
7.3
7.4
5.7
5.8
•Hemp °C
Effl.
8.0
7.8
7.7
7.8
7.7
7.7
7.6
7.6
7.5
7.6
7.5
7.4
7.4
7.5
7.4
7.4
7.4
7.4
7.3
7.3
7.3
5.7
5.6
Inf.
18.5
20.0
17.0
19.5
19.0
17.0
if) n
1 u .U
13.0
17.0
16.0
17.0
14.0
13.0
15. 0
13.0
12.0
14^0
14.0
13.0
14.0
16.0
18.0
8.0
9.0
Effl.
30.0
24.0
23.0
25.0
32.0
22.0
22.0
25.0
22.0
20.0
20.0
22.0
22.0
22.0
18.0
16.0
20.0
21.0
22.0
23.0
18.0
15.0
12.0
PCBs-^g/l
infl.
18
32
26
34
30
23
Oft
JO
18
18
18
34
34
20
46
46
46
35
35
26
26
23
23
42
31
Mid.
0
0.5
0.7
0.4
0
0.3
21
• 1
0.2
0.4
1.0
1.6
1.3
0.1
0.6
0.3
1.4
3.0
4.1
1.8
2.8
1.0
5.5
12.1
1.0
Effl.
0
0.4
0.7
0.4
0
0.3
0.1
0.2
0
1.2
0.6
0.3
0.2
1.3
1,2
3.1
2.1
1.4
1.5
1.2
4.2
2.9
0.4
Source: Arisman and Musick, General Electric Company Hudson Falls, New York, Zeff and erase, Westgate
Research Corporation, West Los Angeles, California
-------
APPENDIX D (Continued) •
DESIGN-OF-EXPERIMENT TESTS FOR PCB DESTRUCTION IN THE ULTROX PILOT PLANT
Effl. UV Ozone Mass
Flow Ozone Ozone Lamp Flow pH Temp °C PCBs->g/l
Exp.
No.
25
26
27
28
29
30
31
32
33
34
35
36
37
Hate
1/inin
2.0
1.2
1.0
1.5
1.6
0.9
0.6
0.6
0.5
1.4
1.6
1.8
0.8
Flow
nq/tain
500
250
400
500
250
300
1000
1000
800
250
500
250
350
Cone.
wt%
2.6
2.5
1.0
3.0
1.3
2.6
3.0
3.0
3.0
1.3
3.0
1.5
1.0
Arrangement
1
6
12
29
19
20
10
29
29
29
20
19
10
14
2
10
30
30
20
30
10
30
30
30
30
20
30
30
3
2
1
1
1
1
1
1
1
1
1
1
2
2
Distribution'
1
200
150
133
200
75
50
300
400
400
. 75
MO
75
125
2
200
150
133
200
75
50
300
200
200
75
200
75
100
3
100
50
133
100.
100
200
400
200
200
100
100
100
125
Peed
5.7
5.7
6.2
5.6
6.3
7.0
7.0
6.8
7.0
7.4
7.4
7.3
Mid.
5.6
5.7
6.2
5.9
6.3
6.8
6.9
6.9
7.3
7.2
7.5
7.2
Effl.
5.6
5.R
6.1
5.9
6.2
6.6
6.9
7.0
7.4
7.2
7.5
7.2
Inf.
11
10
15
13
3
14
14
10
12
.14
12
11
15
Effl.
12
18
23
19
6
19
27
24
28
19
20
16
18
Infl.
42
31
22
22
23
28
13
10
7
18
18
18
18
Mid.
8.7
3.7
2.3
2.3
1.2
3.5
1.3
0
0
0.6
0
1.0
0.5
Effl
2.7
2.3
1.2
1.0
1.3
1.1
0.4
0
0
0
0
0.2
0.4
Source: Arisman and Musick, General Electric Company, Hudson Falls, New York, Zeff and Crase, Westgate
Research Corporation, West Los Angeles, California
-------
APPENDIX E
PREDICTED OPERATING CONDITIONS IN PILOT PLANT TO ACHIEVE MINIWJM OPERATING AND CAPITAL COSTS
fOft A 150,000 GPD (567,750 LITERS/DAY) ULTRQX TREATMENT PLANT TO OBTAIN< 1,0 «g/1 PCB's
Pilot Plant Operating Conditions
Design
Number
20006
18430
18254
18222
18414
Effl.
Flew
Rate
1 PI™
2.0
2.0
2.0
2.0
2.0
Total Ozone
Ozone Mass Mass Flow Distribution
Plow Sect 1 Sect 2 Sect 3
nrj/min
200
100
100
100
100
rrrj/aiun
80
33
50
40
57
ing/kin
80
33
38
30
29
roj/min
40
33
12
30
14
Ozone
wU
1.0
1.0
1.0
1.0
1.0
Number
Lamps
•total
15
15
15
15
15
Number of Lamps Effl.
Sections DEIIP
1
0
0
0
0
0
2
10
10
10
10
10
3
5
5
5
5
5
9/1
79
85
83
86
85
Effl.
PCB
g/i
0.7
0.9
0.9
0.9
1.0
Source: Arisman and Musick, General Electric Co., Hudson Palls, New York, Zeff and Crase, Westgate
Research Corporation, West Los Angeles, California
-------
UNITED STATES DEPARTMENT OF COMMERCE
National Technical Information Sarvloa
52B5 Port Royal Road
Springfield, Virginia 221 61
OFFICE OF THE DIRECTOR
Dear Professional:
I wish I could tell you how much time and money the NTIS
Abstract Newsletters could save you.
But, I can't. Because I don't know how much time and
money you unknowingly spend on costly research that actually
duplicates the work of others. Yet this research information
is readily available to you when you know how to access it.
That's what NTIS Abstract Newsletters are all about.
NTIS Abstract Newsletters Summarize
70,000 New Reports Annually
You see, NTIS information specialists index, analyze,
and categorize 70,000 new research documents a year. Each
report is judged for its potential value for thousands of
current subscribers. Then it is summarized in a brief
abstract and published in one or more Newsletters. You receive
only those abstracts of interest to your professional field.
NTIS exists to help you find—and provide you with—
research results and information. You'll know what you need
to know—when you need to know it—with NTIS Abstract Newsletters.
-------
The Newsletters are a speedy, comprehensive, economical
way for you to keep current on the latest research and to cut
your risk of duplicating work already available in a report
from NTIS.
Your Newsletters will suggest possible directions and
starting points for your own research. Knowing research
applications of others will help you get more out of every
research dollar you invest.
Worldwide Sources
The NTIS collection also contains selected information
on scientific and technical developments around the world.
NTIS is now actively expanding its sources of foreign
technology to add to more than 300,000 foreign reports
already on hand—including translated materials from the
U.S.S.R., other Eastern European countries, Japan, and major
Western countries.
Low Subscription Rate
By now, you may well be apprehensive over the
cost of this remarkable Newsletter service. If
so, you will be pleasantly surprised to learn
you can subscribe to the Energy Abstract Newsletter,
for example, for about $1.83 a week!
NTIS is a unique self-supporting Government agency offering
economical information products and services on a cost-
recovery basis.
You Get Reports on Research Not
Published By Any Other Source
NTIS Abstract Newsletters let you know about all
Government-sponsored research in your field. Each week
you'll learn about research not reported or available from
any other source.
And when you spot a report you need, NTIS can provide it
for you in a matter of days, usually for about $14 in paper
copy or only $4.50 in microfiche. Your Newsletters contain
a handy report order form for your convenience.
Annual Subject Index Included
Your subscription includes an Annual Subject Index to
give permanent reference value to your Newsletter.
US. GOVERNMENT PBINTUra OFFICE: 1983 - 381-009 - 227/524
-------
Of course, your subscription is tax deductible as a
business expense in most instances.
More than that, when you consider the wasted time and
costly mistakes the Newsletters can help you avoid, you
will begin to appreciate what a practical investment NTIS
Abstract Newsletters really are. For example:
• MANUFACTURING TECHNOLOGY: one of the newest
Newsletters that meets industry's most
pressing need right now...brings news of the
latest research on robotics and CAD/CAM. $125*
• FOREIGN TECHNOLOGY: abstracts of latest
technological and scientific research in all
fields from Japan, United Kingdom, France,
U.S.S.R., and countries all over the world.
SPECIAL CHARTER RATE: $75*
• ENVIRONMENTAL POLLUTION S CONTROL: one of
the most sought after Newsletters covering
air, noise, solid wastes, water, pesticides,
radiation, health and safety and environ-
mental impact statements. $95*
• ENERGY: brings you a wealth of information
on energy use; supply and demand; power and
heat generation; energy policies, regulations
and studies; engines and fuels, and much more. $95*
• COMPUTERS, CONTROL, AND INFORMATION THEORY:
important information, and lots of it, con-
cerning this, one of the fastest growing and
most competitive industries worldwide. $95*
• MATERIAL SCIENCES: applicable to nearly
every industry with abstracts touching on a
huge variety of processes and products. $80*
- MEDICINE S BIOLOGY: covers an enormous range
of information from anatomy to zoology. Every
facet of medicine and biology is touched upon. $75*
And any one of the abstracts in these Newsletters may be
the added element of success for your ongoing research, contract
proposals, cost reduction efforts, and profits.
NTIS regularly publishes information not available elsewhere,
and frequently brings you news of research before it appears in
professional journals and magazines. You simply should not risk
acting without benefit of complete, current information on what's
happening in Government-sponsored research in your field.
*One-year subscription rate (Over, please)
-------
You need NTIS Abstract Newsletters and here's how to
order them.
Complete the order form at the bottom of this letter and
mail it back today to the address below. You may use any of
the options listed in order to pay for your subscription.
You are guaranteed satisfaction. If your Newsletters
do not satisfy you for any reason, you may cancel and receive
a credit or refund for the balance of your subscriptions.
I invite you to subscribe today.
Sincerely,
Joseph F. Caponio
Acting Director
5285 Port Royal Road
Springfield, VA 22161 or Call (703) 487-4630
NTIS,
Please start my subscription to the following Abstract Newsletters
Special Offer Only:
Foreign Technology
- Save $15 or $30-
Charter Rate: 2 yrs. $150 D (Reg. Rate $180)
Charter Rate: 1 yr. $75 D (Reg. Rate $90)
Additional subscriptions $25 each annually.
Name
Occupation.
Company —
Address —
City/State/ZIP_
Title
a plus.
Title
D plus -
Cost
$
$
$
$
$
TOTAL $
Other Titles:
additional subscriptions to this Newsletter $25 each
_ additional subscriptions to this Newsletter $25 each
D Charge to my Deposit Account No
D Here is my check to NTIS for $
D Bill me
Charge to G American Express D VISA
D MasterCard
Card No
Expires
Signature
(required to validate order)
PR-733-3
4/83
------- |