EPA-600/2-77-196
September 1977 Environmental Protection Technology Series
AT-SEA INCINERATION
OF ORGANOCHLORINE WASTES
ONBOARD THE M/T VULCANUS
Industrial Environmental Research Laboraton
0«ice of Research and Developrnen
' S .Environmental Protec tion ftgencv
Research Tnangle Park, North Carolina 277
-------
RESEARCH REPORTING SERIES
Research reports of the Off ice of Research and Development, U.S. Environmental Protec-
tion Agency, have been grouped into nine series. These nine broad categories were
established to facilitate further development and application of environmental tech-
nology. Elimination of traditional grouping was consciously planned to foster technology
transfer and a maximum interface in related fields. The nine series are:
1. Environmental Health Effects Research
2. Environmental Protection Technology
3. Ecological Research
4. Environmental Monitoring
5. Socioeconomic Environmental Studies
6. Scientific and Technical Assessment Reports (STAR)
7. Interagency Energy-Environment Research and Development
8. "Special" Reports
9. Miscellaneous Reports
This report has been assigned to the ENVIRONMENTAL PROTECTION TECHNOLOGY
series. This series describes research performed to develop and demonstrate instrumen-
tation, equipment, and methodology to repair or prevent environmental degradation from
point and non-point sources of pollution. This work provides the new or improved tech-
nology required for the control and treatment of pollution sources to meet environmental
quality standards.
REVIEW NOTICE
This report has been reviewed by the participating Federal Agencies, and approved for
publication. Approval does not signify that the contents necessarily reflect the views and
policies of the Government, nor does mention of trade names or commercial products
constitute endorsement or recommendation for use.
This document is available to the public through the National Technical Information
Service, Springfield, Virginia 22161.
-------
EPA-600/2-77-196
September 1977
AT-SEA INCINERATION
OF ORGANOCHLORINE WASTES
ONBOARD THE M/T VULCANUS
by
J.F. Clausen, H.J. Fisher, R.J. Johnson, E.L. Moon,
C.C. Shih, R.F. Tobias, and C.A. Zee
TRW, Inc.
One Space Park
Redondo Beach, California 90278
Contract No. 68-01-2966
Program Element No. 1AB604
EPA Project Officer: Ronald A. Venezia
Industrial Environmental Research Laboratory
Office of Energy, Minerals, and Industry
Research Triangle Park, N.C. 27711
Prepared for
U.S. ENVIRONMENTAL PROTECTION AGENCY
Office of Research and Development
Washington, D.C. 20460
-------
ABSTRACT
This report describes the incineration of 4100 tonnes of organochlorine
wastes by the M/T Vulcanus in an EPA designated burn area in the Gulf of
Mexico under a special permit granted by Region VI of the EPA. The wastes,
containing 63% chlorine, orginated from manufacturing processes conducted
by the Shell Chemical Co., Deer Park, Texas, and was loaded onto the
Vulcanus at the Shell Deer Park facility.
The incineration process was monitored by a field sampling team from
TRW, Redondo Beach, Ca. Waste destruction efficiencies and total combustion
efficiency were determined by five methods, each with separate means of
sampling, analysis, and calculation. Incinerator efficiencies of at least
99.9% were observed, at an average waste feed rate of 22 tonnes/hr. The
process was carried out at a flame temperature averaging 1535°C and at
dwell times calculated to be 0.9 seconds.
An automatic waste shut off system was incorporated into the incinera-
tion process, and was pre-set so as to shut down the flow of waste if the
flame temperature should drop below 1200°C. The temperature of the process
was monitored directly by an optical pyrometer for flame temperature and
indirectly by thermocouples which measured wall temperature and which were
statistically correlated with the flame temperature.
ii
-------
PREFACE
Organochlorine incineration on board the M/T Vulcanus operating
in the Gulf of Mexico was accomplished under provisions cf an Environmental
Protection Agency Special Permit. Thermal destruction of such wastes at
sea is a relatively new disposal method in the United States. The purpose
of sampling and monitoring approximately one-third of the total waste
destroyed during this incineration operation was twofold. First, there
was a definite need to test the suitability and reliability of a comprehensive
stack gas sampling apparatus and an on-line monitoring instrumentation
system. And secondly, data was required in support of developing United
States guidelines and regulations for new incineration ships. It was
not the intent of t is work to insure compliance with permit operating
conditions for Gulf of Mexico incineration activity.
Successful accomplishment of this effort was due to the assistance
and cooperation of a large number of individuals and their respective
organizations. Equipment design and development, definition of sampling
requirements and the acquisition of samples and subsequent analysis was
accomplished by personnel from the Applied Technology Division and the
Environmental Engineering Division of TRW, Inc., Redondo Beach, California.
Engineering liaison to adapt test equipment on-board the Vulcanus was
facilitated by the management of Ocean Combustion Services, B.V., of
The Netherlands. Subsequent installation of this equipment as well as the
acquisition of data could only have been accomplished with the complete
cooperation of the ship's officers and crew.
Technical direction, assistance and planning of the program's
conceptual design and implementation was received from a number of agencies
within the U. S. Government. These include the U. S. Environmental
Protection Agency's Oil and Special Materials Control Division (OSMCD),
Washington, D.C.; Industrial Environmental Research Laboratory (IERL),
Research Triangle Park, North Carolina; and the Regional Office, Region IV,
Dallas, Texas. Guidance in installing equipment on the Vulcanus to conform
with U. S. maritime regulations was provided by the U. S. Coast Guard,
Headquarters, Washington, D.C.
Representatives of the Shell Chemical Company, Deer Park, Texas,
provided waste samples for preliminary analysis prior to the burn as well
as technical details regarding their original sampling and monitoring
work on board the Vulcanus in 1974. Personnel of Arthur D. Little, Inc.,
Cambridge, Massachusetts, provided consultation regarding sample
preparation and analysis.
The authors are indebted to the many individuals of all these
organizations for their contributions and cooperation.
-------
CONTENTS
Page
1. Introduction and Summary 1
2. Description of the M/T Vulcanus C
2.1 Vessel-General Layout G
2.2 Tanks and Pumps 6
2.3 Incinerator System 6
2.4 Recording and Control Equipment 11
2.4.1 Waste Measurements 11
2.4.2 Wall Temperature Measurements 11
2.4.3 Emergency Automatic Waste Shut-off 11
2.4.4 Special Equipment 12
3. Test Description 14
3.1 Operational Procedures 14
3.1.1 Test Procedures 14
3.1.2 Safety Procedures 16
3.1.3 Test Commentary 16
3.2 Sampling Methods 19
3.2.1 Probe and Probe Mount 19
3.2.2 On-Line Gas Monitoring 21
3.2.3 Acquisition of Combustion Products 25
3.3 Analytical Techniques 28
3.3.1 Extractions and Sample Preparation 28
3.3.2 Analysis of Concentrated Extracts 30
3.3.3 Analysis of Gas Bag Samples 31
3.4 Problems Encountered 31
3.4.1 Sampling Train Impingers 31
3.4.2 Sorbent Module 32
3.4.3 Filter Holder 33
3.4.4 SASS Train Control Unit 33
iv
-------
CONTENTS (Continued)
Page
3.4,5 On-Line Instrumentation 33
3.4.6 Degradation of Gas Bag Fittings 34
4. Test Results 35
4.1 Operational and Field Data Summary 35
4.1.1 Waste Feed Rate and Mass Emissions 35
4.1.2 Temperature and Residence Time 36
4.1.3 On-Line Gas Composition 40
4.1.4 Plume Characteristics 50
4.1.5 HCI Measurements on Board 50
4.2 Analytical Results 52
4.2.1 Waste Feed Analysis Results 52
4.2.2 SASS Train Sample Analysis Results 57
4.2.3 Gas Bag Sample Analysis Results 63
4.2.4 Burner Head Residue Analysis Results 6R
5. References 67
Appendix A - Schedule of Waste Consumption by
Tank Number and Location 68
Appendix B - Tabulation of Temperatures in Furnaces 70
Appendix C - Safety Procedures 76
Appendix D - Burner Maintenance Record 81
Appendix E - GC/MS Reconstructed Gas Chromatograms 84
-------
FIGURES
2-1. M/T Vulcanus - incineration vessel 7
2-2. Tank layout schematic 8
2-3. Incinerator configuration 8
2-4. Incineration system- burner and thermocouple locations 10
3-1. Sampling port locations 15
3-2. Probe design diagram 19
3-3. Diagram of probe mount for the Vulcanus 20
3-4. Probe mount with probe fully withdrawn 21
3-5. Sampling port dimensions 22
3-6. Portable laboratory location aboard the M/T Vulcanus 23
3-7. Equipment arrangement in the portable laboratory 24
3-8. Flow diagram of on-line instrument system 25
3-9. Source assessment sampling system (SASS) schematic 26
4-1. Controller/flame temperature correlation-starboard
incinerator 38
4-2. Controller/flame temperature correlation-port incinerator 39
4-3. On-line gas composition data - Test IV, March 9 44
4-4. On-line gas composition data - Test V, March 10 45
4-5. On-line gas composition data - Test VI, March 11 46
4-6. Combustion efficiency data - Tests IV, V, and VI
(March 9, 10, and 11, respectively) 47
4-7. Combustion efficiency vs. probe position 48
4-8. C02 concentration vs. probe position 49
4-9. CO concentration vs. probe position 49
4-10. HC1 sampling locations
1) combustion deck 2) bridge 3) main deck 4) stern 53
-------
TABLES
1-1. Definition of destruction efficiency terms 3
1-2. Summary of calculated destruction efficiencies 3
2-1. Specification of the M/T Vulcanus 9
3-2. Description of on-line instruments 24
4-1. Waste feed rates 36
4-2. Gas composition data summary 41
4-3. Test data summary for shipboard incineration test
series I-VII 43
4-4. Hydrogen chloride (HC1) in air aboard vessel 51
4-5. Summary of Vulcanus samples 54
4-6. Representative waste composition by GC/MS 55
4-7. Calculated approximate emission rates of inorganic elements...56
4-8. Effluent gas concentrations from analysis of sorbent
trap extracts (mg/m3) 58
4-9. Gravimetric results on organic SASS samples 60
4-10. LRMS results on organic SASS samples 62
4-11. Analytical recovery for SASS samples 62
4-12. Gas bag constituents by LRMS 63
4-13. Gas bag analysis by Tenax/GC 64
4-14. Compounds found in burner head residue 66
vn
-------
1. INTRODUCTION AND SUMMARY
Thermal destruction of chlorinated hydrocarbon wastes at sea is a
relatively new disposal technique for wastes generated in the United States.
In late 1974, organochlorine wastes were incinerated in the Gulf of Mexico
for the first time by the Motor Tanker (M/T) Vulcanus. This effort,
monitored by the U. S. Environmental Protection Agency (EPA), provided
an initial technical basis for concluding that at sea incineration was
a viable alternative to other means of disposal using land based facilities
(Reference 1).
A second incineration program, again burning organochlorine waste,
was performed on board the M/T Vulcanus during the period 5 through 13 March,
1977. This second Gulf of Mexico incineration afforded an opportunity to
employ recent technology for monitoring and assessing the environmental
effectiveness of this disposal process. The purpose of this test work
was to evaluate the adequacy and reliability of the monitoring and comprehensive
sampling system. Previously, these systems had been utilized on land-based
incineration facilities (References 2, 3). The ability to make such a
system reliably operate in the more harsh environment encountered during
at sea incineration remained to be demonstrated. A second equally important
objective for the current program was acquisition of sufficiently comprehensive
data to support developing at sea incineration regulations and guidelines
for future new incinerator vessels.
The scope of activities for this work included design and preparation
of the stack gas sampling and monitoring apparatus, development of a
sampling and analysis protocol, acquiring samples and monitoring waste
effluent during incinerator operation, as well as analysis of these samples
followed by evaluation of the results. This report presents results
obtained during the EPA sponsored monitoring and sampling program.
During the sampling and monitoring operations on board the M/T
Vulcanus in March, 1977, approximately 4100 tonnes (metric tons)
of Shell Chemical Company wastes were incinerated. This was accomplished
under Special Permit No. 750D008E granted to the Shell Development
Company by Region VI of the Environmental Protection Agency
(Reference 4). This represented about one-third of the total waste
destroyed as allowed by the permit conditions. The site of the combustion,
as fixed by the permit, lay in the Gulf of Mexico within a rectangle
-------
defined by the following coordinates:
Latitude Longitude
27° 06' 12" N 93° 24' 15" W
26° 32' 24" N 93° 15' 30" W
26° 19' 00" N 93° 56' 00" W
26° 52' 40" N 94° 04' 40" W
At its nearest point to land, this site is approximately 130 miles to the
south of Sabine Pass, Texas.
The wastes burned were produced as by-products in the manufacture of
allyl chloride, epichlorohydrine, dichloroethane and vinyl chloride at
Shell Chemical Company's Deer Park, Texas, plant. The material consisted
primarily of low molecular weight chlorinated aliphatics and had a gross
heat content of 3,860 kcal/kg (6,950 Btu/lb).
Sampling operations utilized a high volume EPA Source Assessment
Sampling System (SASS) with a sampling rate of 0.085 cubic meters per
minute. This sampling train was used to trap and remove organic vapors
from the stack effluent. Sample acquisition was by a remotely actuated,
water-cooled stainless steel probe capable of traversing the starboard
incinerator stack diameter of 3.4 meters. During stack sampling operations,
incineration effluent products were simultaneously monitored for total
hydrocarbons, carbon monoxide and dioxide, nitrogen oxide and dioxide
as well as oxygen concentration. These parameters were measured in real
time to monitor the overall combustion efficiency of the incinerator.
SASS train components were subsequently analyzed in detail by laboratory
methods. The purpose of these analyses was to determine how effectively
the incineration process destroyed constituents in the organochlorine
waste feedstock.
The waste was burned at an average rate of 22 tonnes per hour or
approximately 11 tonnes per hour for each incinerator. Total elasped
burn time was 186 hours and 45 minutes. The average flame temperature
was 1535°C as determined by optical pyrometer measurements. Under these
conditions, a combustion gas residence time in the Incinerator was calculated
to be approximately 0.9 seconds.
Six waste incineration sampling tests (I through VI) and one background
test (VII) burning fuel oil alone were conducted. Because sampling tests
were longer and generally more successful for tests IV through VII, these
test data and samples were analyzed more intensively. Data from the
on-line analyzers were used to calculate combustion efficiency while
sampling operations were in progress. Analytical results as well as
on-line total hydrocarbon measurements provided the basis for calculating
waste destruction efficiencies. Five efficiencies were determined, each
with separate means of sampling, analysis and calculation. These five
-------
destruction efficiency terms are listed and defined in Table 1-1. The
resulting destruction efficiency values are presented in Table 1-2 for
tests IV through VII. The latter was the background or baseline test
using fuel oil.
TABLE 1-1. DEFINITION OF DESTRUCTION EFFICIENCY TERMS
Destruction
Efficiency
Term
DECE
DETHC
DEGGHC
DEsc3ci3
DEGGC3C13
Method of Efficiency Calculation
% C.Oy-%CO
-S-CB- x 10°
THC fed-THC from on-line monitor
THC Ted x 10°
THC fed-THC in grab gas sample
~ " THC Fed * 10°
C3C13 fed - C3C13 in SASS samples
7 n toA x 10°
C3C13 fed
C,C1, fed- C,C1, in grab gas samples
_j j ?...-i x inr
C3C13 fed x IOC
Description of Method
Overall comhustu.'. efficiency based on carbon
dioxide and carbon monoxide concentrations. CC^
and CO were measured by continuous on-line
monitors on board the Vulcanus.
Total organic destruction efficiency based on
total hydrocarbons (THC) measured by a continuous
on-line monitor on board the Vulcanus.
Total organic destruction efficiency based on
total hydrocarbons measured in the Tedlar bag
grab gas samples analyzed by GC/F1D.
Waste destruction efficiency based on trichloropro-
pane (major waste constituent) found in SASS train
samples. Trichloropropane was identified and
quantified by GC/MS.
Waste destruction efficiency based on trichloropro-
pane found in grab gas samples by GC after Tenax
concentration.
TABLE 1-2. SUMMARY OF CALCULATED DESTRUCTION EFFICIENCIES
DECE
DETHC
DEGGHC
DEsc3ci3
DE6GC3C13
Test Number
IV
WASTE
99.98%
99.991%
99.998%
99.98%
>99.999%a
V
WASTE
99.96%;;
99.996%
99.996%
99.92%
>99.999%b
VI
WASTE
99.97%
99.996%
99.997%
99.98%
>99.999%b
VII
BACKGROUND
99.97%
99.997%
99.999%
-
a) Waste constituents not detected
b) Based on trace quantity of trichloropropane found
-------
The results of these tests indicate that incineration performance
on the M/T Vulcanus was consistently greater than 99.9 percent for both
waste and total organic destruction efficiency. These results are
comparable to land-based incinerator performance with chlorinated waste
materials (References 2, 3). Trace levels of known waste constituents
such as trichloropropane (0.1 to 0.6 mg/nr) were found in the effluent
gases. Additionally, low ppm and ppb levels of chlorinated and alkyl
substituted benzene compounds were found in some samples. These materials
could have been in the waste feed at levels which were too low to make
confirmed indentification, or they may have been masked by other
constituents during the waste characterization. This identification
level was estimated to be less than 0.5 percent in the waste. A specie
at 0.5 percent, if totally undestroyed in the incinerator, would create
an emission level of approximately 700 ppm which is about 1000 times the
amount actually found. Thus, efficient destruction of these compounds
is still indicated if they are indeed waste constituents at a 0.5 percent
level. Two other possible sources of these materials may be postulated.
These compounds may have been by-products of the combustion, having been
synthesized in the process or the materials may have been the result of
a chlorination of otherwise non-chlorinated materials within the sampling
train.
An interesting and significant result derived from the on-line
monitoring data was that a fixed position effluent gas sampling probe would
have sufficed to insure that incinerator combustion efficiency was always
equal to or greater than 99.9 percent. Potential wall effects on the
combustion gas composition were apparently eliminated at distances greater
than 15 cm from the inside incinerator wall surface.
The Vulcanus incinerators were effective in destroying this
organochlorine waste material. The sampling (SASS) train, although
advantageous for acquiring large volume samples, was of marginal utility
due to the use of stainless steel construction in many components.
Corrosion was a continual problem area due to the high (approximately
6-8 percent) hydrochloric acid (HC1) content in the combustion products.
An all glass sampling train would eliminate this problem. The on-line
analyzer system was adequate for the monitoring program. Frequent
calibration and maintenance was, however, required to keep the system
on-line at the desired times. Most difficulty was experienced with the
CO, C02 and NO analyzers. Although these instruments were designed for
process stream measurement (for instance, utilities, cement kilns, etc.),
it is doubtful that the extremely hostile environment! of at sea incineration
was ever considered in their original design specifications. In particular,
the shock and vibration modes encountered at sea are somewhat unique.
Equally significant was the salt air environment and high HC1 concentration
in the combustion gas stream. The monitoring system employed was adequate
for this experimental work. In future applications of this type, more
attention must be given to the technical details which will reduce
maintenance and increase instrument reliability. This appears entirely
feasible within current technology.
-------
Cooperation and assistance by the ship's crew was excellent.
Incinerator controls and instrumentation was sufficient. Additional
incinerator wall temperature thermocouples and readouts would be desirable
for purposes of redundancy since existing thermocouples cannot be replaced
during furnace operations. Waste feed rates can only be determined via
tank depletion rates based upon depth soundings. A remote gaging system
would be advantageous from the standpoint of handling hazardous wastes
especially of a volatile nature. Similarly, waste flow metering equipment
would be useful to acquire real time data as opposed to time averaged data
as at present. This is, however, an admittedly very difficult technical
requirement to implement.
No significant incidents related to safety were experienced during
sampling and monitoring operations. In retrospect, adequate scaffolding
should have been installed around the traversing probe. Replacement of
the probe at sea constituted a safety hazard to the sampling crew.
Subsequent incineration test programs at sea should include development
of a definitive, formal safety plan. Although oral safety briefings were
conducted, no formal written safety document was employed during this
test. Because of the unique working environment at sea, a safety plan
is advisable in the future.
Ship-to-shore radio communications were at best only marginal.
The precise technical reasons for this situation are unknown. To facilitate
subsequent test programs, a communications plan should be developed that
establishes pre-arranged reporting times and format, and insures equipment
compatibility.
And finally, an at sea sampling and monitoring operations plan
must coordinate with all interested or associated government organizations
such as U. S. Customs and the U. S. Coast Guard. Uhen properly briefed,
their assistance is most helpful.
-------
2. DESCRIPTION OF THE M/T VULCANUS
2.1 VESSEL - GENERAL LAYOUT
The M/T Vulcanus, originally a cargo ship, was converted in 1972 to a
chemical tanker fitted with two large incinerators located at the stern of
the vessel. The vessel meets all applicable requirements of the Inter-
governmental Maritime Consultative Organization (IMCO) concerning transport
of dangerous cargo by tanker. Figure 2-1 shows a picture of the vessel and
Table 2-1 gives some of the ship's specifications. Both the picture and
the table were furnished by Ocean Combustion Services, B. V., Rotterdam,
The Netherlands, who manage the vessel.
Because of her size - an overall length of 102 meters, a beam of
14.4 meters, and a maximum draft of 7.4 meters - the Vulcanus is able to
operate worldwide. Two diesel engines drive a single propeller to give
cruising speeds of 10 to 13 knots. Her crew numbers 16; ten to operate the
vessel and six to operate the incinerators.
2.2 TANKS AND PUMPS
In this double-hull, double-bottom vessel, the waste is carried in
15 tanks which lie within the inner hull (Figure 2-2). The tanks range in
size from 115 to 574 cubic meters (cu m), with the overall waste capacity
being 3,503 cu m. Tanks are filled through a manifold on deck using a
dockside loading pump. During normal operation the waste tanks can be dis-
charged only through the incinerator feed system. There is, however, pro-
vision for discharging the cargo into the ocean if an emergency arises.
The manner in which the piping system is constructed makes it possible for
any tank to be connected to either incinerator and for cargo to be trans-
ferred from one tank to another.
The space between the two hulls is used for ballast. Ballast tanks
may be filled with sea water and emptied independently as required to trim
and balance the ship. Fuel oil is carried in tanks under and in the
engine room. The engine room is separated from the cargo tanks by double
bulkheads. The pump room and generator room are situated between the
engine room and the waste tanks.
2.3 INCINERATOR SYSTEM
Waste is burned in two identical refractory-lined furnaces located at
the stern of the ship. Each incinerator combustion chamber consists of
two main sections, a combustion chamber and a stack, through which the
combusting fluids pass sequentially (Figure 2-3). This dual chamber con-
figuration, which is characteristic of most high intensity combustion
-------
Figure 2-1. M/T Vulcanus - incineration vessel
systems, uses the first chamber for internal mixing and the second for
adequate residence time. Each furnace has the following characteristics:
Overall Height
Combustion Chamber
O.D.
I.D.
Stack (Top)
O.D.
I.D.
Combustion Capacity (Max)
Combustion Air
Burners (Saacke Type)
Volume
Residence Time:
10.45 meters (m)
5.5 m
4.8 m
3.8 m
3.4 m
12.5 Tonnes/hour
90,000 cu m/hour (maximum)
3
120 cu m
0.9 sec at 1200°C (calculated)
-------
TanX No. 6 BB
Tank No. 5 BB
Tank No. 4 BB
Tank No. 3 BB
Tank No. 2 BB
Port
TanX No. 5 Centre Tank No. 4 Centre Tank No. 3 Centrd Tank No. 2 Centre
Tank No. 1
Tank No. 6 STB| Tank No. S STJ Tank No. 4 STBl TanX No. 3 STB
Tank No. 2 STB
Figure 2-2. Tank layout schematic
Starboard
4.6 M
1.
10.45 M
3.5 M
0.75 M
3.4M.
I.D.
4.8M
I.D.
> STACK
COMBUSTION
CHAMBER
Figure 2-3. Incinerator configuration
-------
TABLE 2-1. SPECIFICATIONS OF THE M/T VULCANUS
Length overall
breadth
Draft, maximum
Deadweight
Speed
Tank capacity
Number of tanks
Tank coating
Loading equipment
Hose connection
Safety equipment
Waste to be processed
Incinerators
Incinerator Capacity
101.95 meters
14.40 meters
7.40 meters
4,768 tonnes
10-13 knots
3,503 cubic meters (cu m)
15, ranging in size from
115 cu m to 574 cu m
No coating in tanks, pipes, pumps, etc.
All equipment consists of low carbon
steel
Not available, but can be placed on
board, if required
10.2, 15.2, and 20.3 centimeters
(4, 6, 8 inches) in diameter
Specially designed for this task and
in accordance with latest regulations
of IMCO, Scheepvaart-Inspectie (The
Hague)
Must be liquid and pumpable. May con-
tain solid substances in pieces up to
5 centimeters in size. Must not attack
mild steel
up to 25 tonnes/hour
Air for the combustion is supplied by large fixed speed blowers with
a rated maximum capacity of 90,000 cubic meters per hour for each incin-
erator. Adjustable vanes are incorporated in the combustion air supply
system which when deflected, increase system pressure drop and reduce the
flow rate. Although no instrumentation is installed to monitor air flow
rate, normal operation is stated by the ship's crew to be between 75,000
and 80,000 cubic meters per hour.
-------
Liquid wastes are fed to the combustion system by means of electrically
driven pumps. Upstream of each burner supply pump is a device (Gorator)
for reducing the solids in the waste to a pumpable slurry. The Gorator
also acts as a mixing pump by recirculating the waste through the waste
tank. Power for the blowers, pumps and other parts of the incinerator
system is supplied by two diesel-generators with a total capacity of 750 KW
at 440 volts and 60 Hertz.
Three burners of the vortex type are located on the same level at the
periphery of each furnace near its base. The burners are of a rotating
cup, concentric design, which deliver waste or fuel oil through a central
tube to an atomization nozzle, where it meets high velocity air delivered
through an annul us. The burners are directed as shown in Figure 2-4.
BURNER 6
BURNER 5
THERMOCOUPLE INDICATORS
(BLACK-BOX AND CONTROL PANEL)
(STARBOARD FURNACE)
THERMOCOUPLE FOR STARBOARD
FURNACE AUTOMATIC SHUT-OFF
BURNER 4
BURNER 3
BURNER 1
THERMOCOUPLE FOR PORT
FURNACE AUTOMATIC SHUT-OFF
THERMOCOUPLE INDICATORS
(BLACK BOX AND CONTROL PANEL)
(PORT FURNACE)
BURNER 2
Figure 2-4. Incineration system - burner and thermocouple locations
10
-------
Periodically the burners require cleaning. The burners are normally
cleaned one at a time with the remaining two burners firing waste during
the cleaning. This is easily accomplished because the burners are readily
accessible. Each burner has a vertical pivot or hinge so that it can be
swung out of the furnace. The opening left by this operation is temporarily
closed by a cover. The burners are cleaned by a metal tool which is pushed
through the burner. Three-way valves are utilized on each burner providing
waste feed, fuel oil feed or a shutoff condition. Waste and fuel oil can-
not be valved into a burner simultaneously; however, alternate burners could
be operated with fuel and waste to achieve higher combustion temperatures
if necessary.
2.4 RECORDING AND CONTROL EQUIPMENT
2.4.1 Waste Measurements
The measurement of waste in the tanks during this burn was accomplished
by sounding the depth of waste in each tank with a tape. In order to comply
with U.S. Coast Guard requirements, a new measuring system which is sealed
to prevent vapors from escaping into the ambient air was installed at the
completion of the burn.
2.4.2 Wall Temperature Measurements
Temperatures during operation of the incinerators are measured by two
platinum-platinum/10% rhodium thermocouples in each incinerator. Each pair
is located in a well opposite one of the burners. One thermocouple is
located approximately 1.3 centimeters from the inside surface of the
refractory lining. This thermocouple provides temperature information to
the automatic waste shutoff system and is referred to as the "controller"
thermocouple in the Appendix B data tables. A second thermocouple approxi-
mately 4 centimeters from the inner surface of the firebrick is referred to
as the "indicator" because it provides temperature information to a panel
located in the incinerator control room and to a panel located on the
bridge. The Appendix B data table refers to the bridge panel as the "black
box." In addition, this control panel indicates the day, month and time,
and has lamps which show the burners and pumps are operating.
The thermocouples used to feed the automatic waste shut off system and
the recording equipment have been quite reliable and durable, according to
personnel of the M/T Vulcanus. There were no known failures during the
burn covered in this report.
The possibility of replacing a thermocouple during a burn was dis-
cussed. The opinion expressed by the operating personnel is that removal
and replacement of a thermocouple in a furnace is not feasible while operating.
2.4.3 Emergency Automatic Waste Shut-off
For the March 1977 burns, an automatic waste cut-off system was used.
This system, called the Plastomatic 2000 and supplied by Withoff-Phillips,
Bremen, W. Germany, uses a thermocouple controlled, spring loaded, solenoid
11
-------
actuated valve which shuts off waste to the burners when the temperature
of the furnace drops below a pre-set selected temperature. During these
tests the EPA required the shut off temperature to be set at 1200°C.
If the temperature in a furnace drops below the pre-selected minimum,
a solenoid is de-activated, allowing a spring loaded valve to close, thus
shutting off the flow of waste to the three burners of that furnace. The
valve which is thus closed shuts the line in both directions. The waste
pumps are also stopped at this time by cutting off power. A power failure
or thermocouple burnout in the Plastomatic system would also shut off the
flow of waste to the burners involved.
A demonstration of the action of the solenoid-actuated valve was set
up by the Chief Electrician and Chief Engineer of the Vulcanus. A cutaway
model of the valve was set up, the valve opened, and the solenoid activated
electrically. The valve remained open. Upon breaking the electrical con-
tact, the spring closed the valve rapidly and positively, preventing flow
in either direction. The breaking of the electrical contacts simulated the
action of the Plastomatic system if Its thermocouple should Indicate
temperature lower than the one selected at the control panel. Similarly,
a loss of power would shut off the waste flow to the furnace, and would
prevent backflow of the waste.
If the system should function, i.e., shut off the waste flow, the pump
which has been stopped and the valves which have been closed may be
restarted after the cause has been identified and corrected and after the
Plastomatic has been reset. It should be noted that procedures require
restart and reestablishment of the desired temperatures using fuel oil
before waste can be burned again.
As shown in the Appendix B data tables, the thermocouple providing
wall temperature sensing to the automatic shutoff system ("controller"
thermocouple) may also be utilized to Indicate real time wall temperature
measurements. This is accomplished by adjusting the controller from the
shutoff temperature setting (in this instance, 1200°C) to increasingly
higher temperature settings. When the adjusted setting is coincident with
the actual temperature sensed by this thermocouple, the feed valve relay
clicks. Observation of the pointer location with respect to the temperature
scale on the dial provides a temperature reading. The waste shutoff system
is not immediately activated since a time delay is incorporated 1n the
electrical circuit. The pointer 1s then reset to the desired automatic
shutoff temperature before the system can activate. The ability to record
temperatures in this manner is useful in the event that the "indicator"
wall temperature should fail during operation.
2.4.4 Special Equipment
Certain equipment, not usually found aboard chemical tankers, has been
installed on the Vulcanus because of Its particular type of operation. These
items of equipment are:
12
-------
Loran and Decca Navigational Aids: This equipment is needed in Euro-
pean waters (Decca) and United States waters (Loran) in order to assure pre-
cise location of the ship at all times. During the current trip, Loran
was used to locate and keep the Vulcanus in the designated burn area.
Anemometer: This equipment measures the velocity and force of the
wind. It is therefore useful in selecting an attitude of the ship during
various wind conditions such that the plume may be directed away from the
ship and its personnel.
Radio Communication; The Vulcanus is equipped with SSB (single side
band) and DSB (double side band) radio for voice and code communication;
MF (medium frequency) and SF (short wave) telegraphy; Marifoon (VHF) (all
channels) for close in voice communication; and Semafoon (a private tele-
phonic communication system).
Optical Pyrometer; A portable optical pyrometer was utilized durinq
the test program to measure incinerator flame temperatures. These measure-
ments were taken periodically (see Appendix B data tables) using a Leeds
and Northrup Optical Pyrometer operating in the 6500 A range.
-------
3. TEST DESCRIPTION
This section presents the manner in which the sampling tests were
carried out. It is divided into the following subsections, listed in
order of discussion:
Operational procedures.
t Sampling methods.
Analysis techniques.
Problems encountered.
The operational, sampling, and analysis procedures were designed and
selected with consideration given to the nature of the waste. A prelimi-
nary survey sample of the waste material was acquired and analyzed prior
to the final planning of the shipboard incineration test program. The
waste was a low-viscosity, low-ash liquid with a gross heating value of
3,860 kcal/kg (6,950 Btu/lb). Its elemental composition was 30.01 percent
carbon, 4.17 percent hydrogen, 0.012 percent nitrogen, 0.009 percent
sulfur, 62.6 percent halogens (as chlorine), and 3.2 percent oxygen
(by difference). 6C/MS analysis showed the waste to consist nearly
entirely of low molecular weight (2-3 carbons), chlorinated alkyl com-
pounds with trace amounts of oxygenated chloroalkyl compounds. Samples
of the waste material actually burned were taken on board the ship
during the tests and the results of analyzing those samples are presented
in Section 4.2.1.
3.1 OPERATIONAL PROCEDURES
Detailed operating and safety procedures were reviewed and approved
prior to arrival of the TRW sampling team on board the M/T Vulcanus.
Procedures and operating conditions were also recorded during the actual
tests. Following are brief summaries of procedures and a test-by-test
commentary on events that took place on the vessel.
3.1.1 Test Procedures
Sampling was conducted on the starboard incinerator. Two sample
ports were available for probe insertion; however, safety considerations
ruled out Port #2 (see Figure 3-1). Port #1, located above a burner and
pointing toward the center of the incinerator, was used to obtain the
combustion gas samples. During each test several points along this
diameter were sampled since the probe could traverse the stack.
14
-------
PORT NO. 2 LOCATED
90° FROM KEEL
CENTER LINE
PORT NO. 1 LOCATED
OVER BURNER AND
POINTING TO CENTER
OF INCINERATOR
Figure 3-1. Sampling port locations
A total of seven sampling tests were conducted. The last test was
for a background while only fuel oil was burned in the starboard incin-
erator. The basic procedure for each test was:
Verify sampling system ready
- Fill first two impingers with caustic solution
- Fill last impinger with silica gel
- Leak check the system
Fill impinger box with ice
Verify on-line instrumentation ready
Start sampling system
15
-------
Stabilize flow rates and temperatures in sampling system
t Activate on-line analyzer system
Extended sampling duration
- Process data acquisition
Effluent gas composition data acquisition
- Effluent gas sampling
Shutdown on-line analyzer
Shutdown sampling system
Recover and store samples
Each sampling test was scheduled to run from three to five hours.
3.1.2 Safety Procedures
Safety requirements for handling and operating the sampling trains
and ancillary equipment were established and adhered to, including the
fol1owi ng:
Only authorized personnel were permitted in the sampling
area during a test.
Visual observation of the test system was maintained at all
times during tests.
Entry suits and Scot-Paks were available in the immediate
area.
Safety belts were used when working on the probe or probe
mount.
Sampling tests were performed only when weather conditions
were favorable.
Safety procedures for operations of the Vulcanus crew are contained
in Appendix C.
3.1.3 Test Commentary
Six sampling tests were conducted during the incineration of the
organochlorine wastes and one background test with fuel oil. For con-
venience, the waste tests were conducted before the background test.
Before starting waste combustion, the furnaces were heated to operating
temperature by burning fuel oil. When the preselected temperature had
been reached, the system was switched over to waste. Because a "cold"
16
-------
furnace must be heated slowly in order to avoid cracking the refractory
liner, the burners were turned on and off intermittently, in order to
provide slower and more uniform heating. After preheating to operating
temperature, changeover to waste was accomplished by switching each burner
successively to waste. The effect on the temperature was small, a lowering
of approximately 10-20QC being observed on the thermocouple dial on the
control panel in the combustion room.
General observations of the incinerator operation were noted. The
typical flame produced by the burners was bright, intense white, free of
dark areas and free of sputtering. No black particulate matter was
observed above the stack exit. The daytime appearance was one of heat
convection above the incinerator. At night the gases above the incinerate:-
were white against the black sky. Excursions of flames beyond the in :iii
erator were not observed while waste was being burned. An occasional
vibration was noted in the furnace. This condition appeared to be asso-
ciated with burners which were being pushed to capacity. Backing off of
the flow rate to the burner appeared to eliminate the condition. The
phenomenon did not appear to correlate with any other parameters, i.e.,
time of day or night, ship, speed, etc. A continual inspection of the
incinerator burners was maintained by the Vulcanus personnel. Burners
were cleaned periodically, usually yielding a black coke or tar at the
nozzle. A sample of this was given to the TRW team. Appendix D lists
the burner cleaning history.
On-line gas monitoring data was obtained during all tests. Some
problems were encountered with the analyzer system. These were corrected
as they occurred and the system functioned well under the confined and
harsh shipboard environment encountered at sea. Sampling system mal-
functions precluded obtaining sufficient samples for organic analysis
during the initial three tests (see Section 3.4 for detailed discussions
of the problems encountered during the sampling and monitoring activities).
These problems were corrected and three subsequent waste burn sampling
tests were conducted. The duration of these sampling tests was typically
2-1/2 hours each. In addition, a background sampling test (with the
incinerator burning fuel oil) was also successfully accomplished.
Test I - First Haste Test: Problems were encountered at the initiation of
the test. The large impinger designed to hold the caustic scrubbing
solution collapsed and forced the test to be terminated after 35 minutes.
Just before the test was terminated, the quartz liner in the sampling
probe cracked due to a temporary thermal overload and a new probe was
installed for Test II.
Test II - Second Waste Test: This test was terminated after approximately
20 minutes because the smaller glass impingers (substituted after the Test I
problem) that were filled with caustic solution became plugged with a white
carbonate precipitate. The thermal overload of the probe, experienced dur-
ing Test I, was avoided during this test by careful control of cooling flow
rates.
17
-------
Test III -Third Waste Test: The concentration of caustic solution in the
impingers was reduced for this test to prevent the formation of the car-
bonate precipitate, and the test ran smoothly for slightly more than one
hour. However, the test then had to be terminated because of a large
quantity of sorbent trap resin had passed through a hole corroded in the
resin support screen.
Test IV - Fourth Waste Test: This was the first test from which the
sample acquired was of sufficient quantity for organic analysis. Sampling
was performed for 2.5 hours. During this test, the sorbent trap operating
temperature was increased from 25°C to approximately 55°-60°C to reduce
the amount of condensation in the sorbent trap module. The caustic solu-
tions and silica gel in the impingers were changed midway through the
test. (See Section 3.4.1). One of the starboard incinerator burners (#5)
was shut off during the first 20 minutes of the test to repair a valve.
Test V - Fifth Waste Test; Sampling was performed for 2.5 hours. Midway
through the test the caustic solution in the impingers was changed. The
sorbent module was operated at approximately 70°C. To further reduce the
amount of condensation, the sorbent module was run at a slightly higher
temperature than the previous test.
Test VI - Sixth Waste Test; Strong wind conditions existed during this
test. There was difficulty in keeping the probe stationary and finally
it was tied down. About 80 minutes into the test the incinerator was
completely shut down for maintenance purposes. The reason for this com-
plete shut down is unknown; however, this was the only such instance where
a complete shutdown of one incinerator occurred. Coincident with the shut-
down, all three burners were cleaned. During the approximately two-hour
delay, the impinger solutions were changed. Total sampling time was 2.5
hours. After the test was over the probe was removed and the liner was
solvent washed.
Test VII - Background Test: A baseline test was performed with fuel oil
to acquire background data on the performance of the starboard incinerator
while burning only auxiliary fuel. A new probe was installed and a three
hour sample was procured. About 30 minutes into the test, it was dis-
covered that the incinerator temperature was only about 950°C instead of
1250°C. Operating temperatures were increased to the normal operating
temperature by increasing the fuel oil feed rate. The feed rate may have
increased too fast because approximately 1.5 hours into the test, flames
were observed over the rim of the incinerator. After the test was com-
pleted it was impossible to remove the probe from the incinerator. It is
believed that when the flames were shooting over the rim of the incinera-
tor the probe became overheated and was bent. After this test was
completed the incinerator was shut down. Several hours later, after the
probe had cooled, it was possible to withdraw the probe from the incinera-
tor. The probe was then removed and the quartz liner was washed with
solvent.
18
-------
3.2 SAMPLING METHODS
Sampling methods used In the tests were chosen to cover two basic
areas:
a) Continuous, on-line monitoring of gas composition to
determine and follow steady state conditions.
b) Collection and concentration of effluent gas products to
identify and quantify the trace organic species present.
Gas samples for both of these systems were extracted by the same
probe. The gases were drawn from the stack area through heat-traced
Teflon lines and then split. One portion continued on to the portable
laboratory installed on the ship which contained the on-line monitoring
system. The other portion passed through an organic collection and
concentration system.
3.2.1 Probe and Probe Mount
The probe was a stainless steel jacketed, water-cooled probe with a
quartz liner and is shown in Figure 3-2. The liner provided an inert
surface for the sample gas and the cooled, stainless steel jacket
shielded this gas from extreme combustion temperatures in order to quench
any further reactions of the sample constituents. Further cooling of the
gas was modulated by aspirating an air/water mixture into the space
between the steel jacket and quartz liner. The probes were approximately
15 feet in length.
AIR/WATER MIXTURE RETURN
AIR/WATER MIXTURE IN
SEA WATER IN
SEA WATER RETURN
SEA WATER OUT
SEA WATER IN
AIR/WATER
MIXTURE OUT
OVEN-
- QUARTZ LINER
PROBE CROSS SECTION
QUARTZ LINER
..AIR/WATER
/MIXTURE IN
Figure 3-2. Probe design diagram
19
-------
The probe was supported and moved by the system shown schematically
in Figure 3-3 and pictorially in Figure 3-4. Design features for this
probe mount are given below:
1. The probe tip was positioned into the effluent gas stream
by a pneumatic system. No adjustment was possible for
azimuth and elevation.
2. Insertion positioning accuracy was ±0.5 feet by command
from the control module which was located on the deck
above the bridge.
3. The probe mount was attached to the incinerator flange as
shown in Figure 3-5. The outboard end of the mount
assembly was supported on an A-frame from the deck.
4. Servicing of the probe required that personnel climb up
to the probe mount.
5. Installation of the probe and mount was done in port when
no burning was occurring.
PROBE BASE
AND FILTER
FLANGE ON
INCINERATOR
PROBE
w
BALL BRG.
PROBE CARRIAGE
CONVEYOR
ROLLER
A-FRAME
SUPPORT
WALL OF
INCINERATOR
Figure 3-3. Diagram of probe mount for the Vulcanus
20
-------
PROBE
CARRIAGE
A-FRAME
SUPPORT
FILTER
OVEN
NEUMATIC
DEVICE
Figure 3-4. Probe mount with probe fully withdrawn
6. The following connections were incorporated into the probe
or probe mount:
a) Teflon sample line to connect the probe to the filter
housing
b) Support flanges for the filter oven
c) Seawater coolant lines for the probe
d) Air coolant lines for the probe
e) Thermocouple leads
f) Pneumatic lines for insertion control of the probe
3.2.2 On-Line Gas Monitoring
A containerized assembly which could be secured on the ship's deck
was prepared for monitoring the test burns. A standard fiberglass rein-
forced plywood (FRP) shipping container was modified to house the on-line
monitoring instrumentation and to serve as the portable laboratory for
the sampling team. The container previously had been in freight service
21
-------
RIM OF
INCINERATOR
0.9
METER
-
-------
Figure 3-6. Portable laboratory location aboard the M/T Vulcanus
timbers were necessary so that the container would clear a Butterworth
hatchcover. The bottom container was used for storage.
The arrangement of equipment in the portable laboratory is shown
in Figure 3-7. The on-line gas analyzers were government furnished and
are described in Table 3-2. With the exception of the 02 and CO
analyzers, redundant on-line instruments were furnished by EPA. As
shown in Figure 3-8, on-line monitoring was performed on the gas sample
stream after processing by the gas conditioner. In addition, because
the Beckman Model 742 02 analyzer would be disabled by HC1 vapors, a
soda lime scrubber for acid gas was utilized in the oxygen analyzer
sample line as shown. A TRW furnished gas chromatograph (Carle Model
No. Ill) was included in the gas analysis system to serve as a primary
back-up for the 02 analyses and as a secondary back-up for C02- This
instrument was slated for use only in the event of failure of the
primary government furnished analyzers. The continuous on-line monitor-
ing instruments encountered only minor problems which were corrected as
they occurred, therefore the gas chromatograph was not needed during the
tests.
23
-------
ILtCr. AND
PLUflBIMS
j-eoxct on
THIS
s/eef
Figure 3-7. Equipment arrangement in the portable laboratory
TABLE 3-2. DESCRIPTION OF ON-LINE INSTRUMENTS
Species Analyzed
Hydrocarbons (HC)
Carbon Monoxide (CO)
Carbon Dioxide
(C02)
Nitrogen Oxide
(NO)
Nitrogen Dioxide
(N02)
Oxygen (02)
Mfg. & Model
Beckman 108 A
Beckman 31 5B
Beckman 3 ISA
Beckman 315A
Beckman 255A
Beckman 742
Analyzer Type
FID
IR
IR
IR
UV
Electro-
chemical
Instrument Range
0 - 90% in 8 ranges
0 - 200 ppm, 0 - 2000
ppm, 0-2%
0 - 4%, 0 - 8%, 0 - 15%
0 - 500 ppm, 0 - 2000
ppm
0 - 200 ppm, 0 - 400
ppm
0 - 5%, 0 - 10%,
0 - 25%
-------
SHIP'S COMPRESSED
MR SOURCE
T-rr.--^-""\
I jr I ! !^ &
s -WAY VALVE]
BLOW
BACK
LINE
VENT
. ^ PURGE AIR
IS \-/TV.
I
T CONDITIONER
PPD
Figure 3-8. Flow diagram of on-line instrument system
3.2.3 Acquisition of Combustion Products
The EPA Source Assessment Sampling System (SASS) was used to trap
and remove the organic vapors from the effluent emissions. This high
volume sampling train is shown schematically in Figure 3-9. The filter
and filter oven associated with the SASS sampling train were mounted on
the sampling probe assembly. A heat-traced Teflon line conveyed the
sample gases from the filter to the solid sorbent trap module. The
sorbent module along with rest of the sampling train was located on the
deck above the bridge. This sorbent module, located upstream of the
impingers, contained a bed of XAD-2 resin to trap the organic constituents.
The module that houses the sorbent trap is temperature controlled to
deliver a gas stream as low as 20°C to the trap.
Because the sorbent material is known to have poor trapping efficiency
for low molecular, high volatility organic species in the C-j to GS range
(Reference 5), it was necessary to^supplement the organic sorbent trap.
An integrated, composite grab gas sample was taken, therefore, utilizing
28-liter Tedlar gas sample bags. This composite gas sample was taken at
25
-------
INCINERATOR
WALL
TO ON-LINE
FILTER MONITORING
HOLDER EQUIPMENT
=?:
SOLID
SORBENT
TRAP
MODULE
QUART'. '
LINER -
XAD-2
CARTRIDGE
CONDENSATE
COLLEaOR
- .. ._. FINE ADJUSTMENT
GAS METER »Y PASS VALVE
' ' COARSE
ICE «ATH
ADJUSTMENT
VACUUM
LINE
HT
VACUUM
PUMP
DRY TEST METER
ORIFICE if
MAGNEHEUC GAGE
Figure 3-9. Source assessment sampling system (SASS) schematic
the Beckman instrument gas conditioner, which minimized sample alteration
and facilitated the sampling operation.
The sampling train was operated at a flow rate of approximately 0.085
cubic meters per minute (cu m/m1n). Gas volumes were measured to 0.03
liters, with a leak rate of less than 0.1 I1ters/m1n. The following
samples were obtained from three waste tests and the background test:
Particulate filter: The filters used 1n Tests I, II, III
and VII were removed from the filter housing with little
or no difficulty. The filters from Tests IV, V, and VI,
however, were stuck to the perforated metal filter support
and broke apart under the slightest attempt to remove
them. A metal spatula was used to scrape adhering filter
material off the support, but it was not possible to
recover it all. The filters were placed in petri dishes
for storage and shipment.
-------
t Solvent heated line wash: A pentane rinse of the heat
traced line connecting the filter housing and the sorbent
trap module was made after each test. The rinsate was
collected in an amber glass bottle with a Teflon-lined
lid.
Solid sorbent trap: The canisters were removed from the
module and placed in glass jars.
Sorbent module washes: Departures from the original test
plan were implemented on board the ship for the sorbent
module trap washing procedure due to the condition of the
module after the tests. The surfaces contacting the
sample gas were coated with a mixture of what appeared
to be metal chloride salts, HC1 and water. The original
washing sequence of pentane/methylene chloride, isopropyl
alcohol and pentane was ineffective in removing the
surface coatings. A field decision was made to use rinse
solvents in the following order 1) water (which dissolved
the salt coating), 2) isopropyl alcohol, 3) pentane/
methylene chloride, and 4) pentane. All washings were
placed in amber glass bottles with Teflon-lined lids.
The sorbent module from Test VII (fuel oil background) was dis-
assembled after the test and an attempt was made to clean it.
It was coated with a heavy layer of what appeared to be iron
rust. These deposits were not removable with any of the
available solvents aboard ship. The entire module assembly
was packed up and shipped back to the laboratory for further
inspection and solvent rinsing.
Condensate solution: The condensates from the sorbent trap
module were transferred to amber glass bottles with
Teflon-lined caps.
0 Combined impinger solutions: The caustic impingers were
combined and stored in amber glass bottles with Teflon-lined
caps.
Acidified split of combined liquid impingers: An aliquot
of each impinger sample was transferred to a Nalgene bottle
and acidified with nitric acid.
Spent silica! gel: The spent silica gel was stored in
glass bottles.
Three grab gas samples: The Tedlar bags were packed in
sturdy fiber drums.
Solvent washes of the quartz Uner were not performed after each
waste burn test due to the delicate and hazardous nature of the operation.
The probe used for Tests II through VI was rinsed with pentane after 1t
was removed. A rinse of the Test VII (background) probe was also made.
Pentane was poured Into the quartz Uner Inlet using a funnel and was
collected from the outlet 1n an amber glass bottle fitted with a Teflon-
lined cap. The Inclined probe was rotated to Insure solvent contact with
all interior surfaces of the quartz Uner.
27
-------
3.3 ANALYSIS TECHNIQUES
Samples taken as described in Section 3.2 were analyzed for both
organic and selected inorganic constituents. The techniques used for
the various sample preparation steps are described below, and the anal-
ysis methods are described in the subsequent section.
3.3.1 Extractions .and Sample Preparation
Samples were removed from the sampling train, stored and prepared
for analysis as described below by sample type.
Probe Washes: In the laboratory the probe wash samples were inspected
for suspended solids and, seeing none, they were dried over anhydrous
sodium sulfate. These samples were then measured for volume, a small
aliquot was taken, and the remainder was concentrated down to 10 ml to
be analyzed for organic materials including waste constituents.
Fi1ters: The filters were not weighed because the 1) loss of filter
material and 2) substantial amounts of crystalline corrosion products
on the filters, precluded any accurate determination of particulate
weight. Thus particulate mass loadings could not be determined for
these tests. The filters were, however, extracted for organics with
pentane for 24 hours. The pentane extracts were dried over anhydrous
sodium sulfate, diluted to volume and an aliquot was taken for storage.
The remainder was concentrated down to 10 ml.
Heated Line Washes: During laboratory preparation this rinse was
combined with the pentane extracts of the sorbent trap resin into one
sample.
Sorbent Canisters: Changes in the procedures for extracting and preparing
the contents of the sorbent canisters for analysis were implemented due
to the extensive coating of the resin with corrosion products. It was
found that pentane, the planned first solvent for extraction, had no
apparent effect on their removal. Water was found to have good dissolving
power, and it was decided to extract the resin and rinse the canisters
first with water. The canisters were soaked in deionlzed water to dissolve
the deposits on them, and the resulting liquid was added to the water used
to make the first resin extraction. This aqueous Soxhlet extraction of
the resin was carried out for four hours. The resin samples 1n the thimbles
were drained of as much water as possible, and the water extracts were in
turn extracted with pentane using separatory funnels.
Pentane was added to the Soxhlet extractors to continue the extrac-
tion after the water extract was removed. However, the immiscibility
-------
of the pentane with the remaining water in the resin would not permit
wetting of the XAD-2 with the pentane. In addition, the pressure caused
by the pentane column height was not sufficient to force the water through
the siphon tube and cause the extractor to "dump". To correct this
situation, 50 milliliters of acetone (nanograde) was added to each
thimble. The acetone mixed with the water and purged the water from the
resin. These acetone/water rinses were transferred directly into the
Soxhlet receiving flask, along with the pentane required for the next
extraction step. The pentane resin extraction was then carried out for
24 hours.
The pentane extracts of the resin samples were combined with two
other pentane samples: 1) the pentane rinse of the heated lines, and
2) the pentane extract of the water used to extract the XAD-2 resin.
The combined solutions were dried over sodium sulfate, an aliquot was
taken for archival storage and the remainder was concentrated down to
10 ml for organic analysis.
Sorbent Module Washes: The laboratory preparation procedures for these
samples were as follows:
a) The water washings were liquid/liquid extracted with
pentane using a separatory funnel. The pentane layer
was saved for combination with other washings of the
module.
b) The IPA washings were dried over anhydrous sodium
sulfate and concentrated down to 10 ml.
c) The pentane/methylene chloride and neat pentane rinsings
were combined with the pentane extract of the module
water wash (item a). This solution was dried over sodium
sulfate and then concentrated down to 10 ml for organic
analysis.
The uncleaned module from Test VII was photographed and washed
sequentially with IPA, 50/50 pentane/methylene chloride solution, and
finally pentane. The rust coating remained after the rinsings. The
rinses of the Test VII module were treated the same as the other test
samples.
Sorbent Module Condensates; In the laboratory, the condensates were
neutralized to pH 7 with sodium hydroxide, and then extracted with
pentane in separatory funnels. The pentane extracts were dried over
anhydrous sodium sulfate and concentrated to 10 ml.
Impinger Solutions: No preparation or analyses were performed on either
the neat impinger solutions or the acidified splits of these samples.
These samples would normally be analyzed for inorganic constituents;
however, inorganic analysis was not required for this program.
29
-------
Silica Gel: T.'ie spent silica gel was reweighed in the laboratory to
determine the amount of water absorbed. The silica gel was then regener-
ated for future use.
Grab Gas Samples; No preparation was required for these samples.
3.3.2 Analysis of Concentrated Extracts
The organic samples, in the form of 10 ml concentrates, were analyzed
by a combination of qualitative and quantitative techniques. The quali-
tative techniques were selected to generally characterize the major
organic components in the concentrated samples, and the quantitative
techniques focused on the waste constituents and hazardous potential
by-products.
Gravimetry: Aliquots (1-2 ml) of the 10 ml concentrates were evaporated
in small glass bottles at ambient conditions to remove the solvent. The
residues were weighed on a micro-balance.
Infrared Spectrometry (IR): The organic residues obtained in the gravi-
metric determination were examined by IR on a Perkin Elmer Model 521
Grating Spectrophotometer. When there was sufficient material, the sample
was run as a smear on a salt window. Smaller quantities were mixed with
KBr and pressed into a pellet for analysis.
Low Resolution Mass Spectrometry (LRMS): Microgram quantities of the
sample residue were transferred to a glass probe tip which was, in turn,
placed in the batch inlet probe of a Hitachi RMU 6 magnetic sector mass
spectrometer. The probe tip was ballistically heated from 50° to over
400°C and a mass spectrum was taken every 50°C.
Gas Chromatography/Mass Spectrometry (GC/MS): Two instrument systems
were used for this portion of the analysis. System 1 was a Finnigan 9500
gas chromatograph interfaced via a single stage glass jet separator to a
Finnigan model 3100D quadrupole mass spectrometer controlled by a System
Industries, System 250 computer software package. System 2 was a
Finnigan 9610 gas chromatograph interfaced via a glass jet separator to
a Finnigan model 4000 quadrupole mass spectrometer. This system is
controlled by a Data General Nova 3 computer and a Finnigan-Incos soft-
ware package. The scan cycles and mass ranges varied somewhat between
systems, but scan cycles were typically 1.8 to 2.0 sec. Mass ranges
were typically between 20 and 300, but were expanded out to as much as
650 mass units as required.
The GC column in System 1, used for all the samples, was 3 percent
OV-17 on 100/120 mesh Chromosorb W AW-DMCS. The oven was programmed
from 20°C to 200°C at 12°C per minute. Repeat analysis for confirmation
of original data and more details was performed using GC/MS System 2.
Slightly different column lengths and program rates were used to achieve
similar retention times with the two instruments. A Chromosorb 101
30
-------
column was subsequently used on a limited number of samples where the
initial data indicated that light volatile materials required better
separation. The oven was programmed from 125°C to 225°C at 10°C per
minute.
3.3.3 Analysis of Gas Bag Samples
Although some degradation of the Tedlar bags' nylon fittinqs had
been observed (see Section 3.4.6), no damage critical to sample integrity
was noticed upon packing the bags for shipment to TRW's Redondo Beach,
California facility nor when they were inspected after arrival. Soon
after their arrival at TRW, the gas bag contents were analyzed directly
by low resolution mass spectrometry (LRMS). For improved sensitivity,
a subsequent analysis was planned using a pre-concentration on Tenax/GC
followed by gas chromatography analysis. In the interim between the
LRMS analysis and Tenax/GC analysis, however, the fittings cracked and
degraded. The fitting literally fell to pieces. The bags were resealed
with tape and a new fitting was installed in each damaged bag. The
contents of the bag for each test were transferred to new bags and the
Tenax/GC work started immediately.
Tenax/GC, a porous polymer GC column packing material, was used to
concentrate the trace organic materials in the bag samples. Liter
quantities of gas from the bags were passed through columns packed with
Tenax. The Tenax columns were then thermally desorbed and the desorbed
species analyzed by gas chromatography with flame ionization detection
(FID).
3.4 PROBLEMS ENCOUNTERED
In spite of detailed planning and preparation for these field tests,
a few incidents occurred that had not been anticipated. The main factors
contributing towards these incidents were the initial sampling operations
on a shipboard incineration program by the sampling team and several
components of the sampling system which were comparatively new technical
developments with very little field experience. The problems which
occurred during the sampling and monitoring activities on board the
Vulcanus are described in the following paragraphs.
3.4.1 Sampling Train Impingers
Sodium hydroxide was added to the sampling train impingers to scrub
out the hydrochloric acid (HC1) thereby protecting the pumps and dry test
meter. The high volume sampling train used in these tests as well as
the high concentration of HC1 in the effluent gases (approximately
6 percent), required a large amount of sodium hydroxide. In order to
accommodate the sodium hydroxide the sampling train was altered to
include a two gallon polypropylene bottle containing sodium hydroxide
which was installed in the sampling system between the sorbent module
and the impinger system. Once the first test was initiated the vacuum
created in the sampling system by the pumps slowly collapsed the bottle
31
-------
thereby forcing the solution into the regular impingers. This problem
was solved by substituting additional glass impingers for the poly-
propylene bottle.
Before the main problem was diagnosed 1n the first test, it was
noted that the sampling flow rate was only 0.03 cu m/min. As a con-
sequence of this low flow rate, the sample gas coming out of the probe
couldn't be maintained above 120°C. In trying to Increase the temperature
of this gas, inadvertently the probe became overheated and the resulting
deformation caused the quartz liner to crack. This necessitated changing
the probe before the second test was initiated.
Fifteen minutes into the second waste test the sample gas flow rate
dropped off and finally stopped. The impingers containing the sodium
hydroxide solution were caked with a thick white precipitate. The carbon
dioxide in the effluent gases had reacted with the concentrated sodium
hydroxide solution to form insoluble carbonate salts. The original
concept had been to use a concentrated sodium hydroxide solution (50 per-
cent) so that the impinger solutions would not have to be changed during
the test. To minimize the precipitate problem the caustic solutions
were diluted to 10 percent solutions and were changed during the tests.
When HC1 vapors were detected in the control module exhaust (using litmus
paper), the sampling train operation was stopped so that the caustic
solutions could be changed. When the caustic solutions were changed,
the last impinger containing the silica gel was also changed. The silica
gel was used to protect the pumps and dry test meter by removing the
moisture from the gas stream.
3.4.2 Sorbent Module
The major problems encountered with the operation of the SASS train
were due to the hostile operating environment. Because the waste was
highly chlorinated (50-70 wt percent halogen as chlorine), the operating
conditions of the incinerators produced an effluent gas which contained
approximately 6 percent hydrogen chloride (HC1). It was also noted that
during the sampling tests the relative humidity was high (>50 percent)
thereby adding to the moisture content of the effluent gas. Any con-
densation of the effluent gas produced a very corrosive environment.
The SASS train, with all its stainless steel components simply was not
designed to operate under these corrosive conditions.
The main problem with the sampling system was corrosion in the
sorbent module. This was first noted in Test III. When the condensate
was first emptied into the drain bottle, there was an appreciable
amount of liquid of a dark green color. The problem actually surfaced
about one hour into the test when a large amount of resin was detected
in the drain bottle after one of the periodic emptyings of the condensate
reservoir section. The resin had ended up in the drain bottle because
the acidic environment had corroded a hole in the stainless steel screen
supporting the resin and a considerable portion of the resin had dropped
into the condensate collection section. Other observations revealed
32
-------
that the whole module was heavily coated with corrosion products.
Several solvents were used to try to remove these corrosion products,
however, distilled water proved most successful.
Originally the sorbent module was intended to be operated at 20°C,
i.e., the effluent gas entering the XAD-2 resin would be at this tempera-
ture. During Test III the module had been operated at 25°C. This was
the lowest temperature that could be obtained. Because of the corrosion
caused by the large amount of condensation at that temperature, for the
remainder of the sampling tests it was decided to operate the sorbent
module at a higher temperature (55-70°C) in order to try to reduce or
minimize condensation in the module. This seemed to reduce the amount
of condensate which was collected in the drain bottle; however, the
module was still heavily coated with corrosion salts. Operating the
module at these higher temperatures allowed the train to be operated long
enough to collect a sufficient sample for organic analysis.
3.4.3 Filter Holder
In all the sampling tests which involved waste burning, there was a
problem in trying to remove the filter from the holder. The filter was
stuck to the support screen and could not be removed in one piece. In
several instances the filter had to be scraped off the screen. There
was also evidence of corrosion on both the screen and filter holder.
This problem was not encountered during the background sampling test
when the incinerator was operated on only fuel oil.
3.4.4 SASS Train Control Unit
The control unit worked very smoothly until the last day of testing.
During the background test (fuel oil) the thermocouple readings became
very erratic. A hand-held temperature readout device (API Instrument. Co.)
was used to check the thermocouples. The results indicated that the
thermocouples were functioning properly, therefore, for the remainder of
the test the API instrument was used to obtain the thermocouple readings.
3.4.5 On-line Instrumentation
A progressive deterioration of performance, i.e., base-line drift,
was noted with the COa analyzer. Frequent calibration of the analyzer
was required in order to obtain reliable data. The instrument tended to
stabilize; i.e., return to levels of prior performance, during the final,
background test when there was no HC1 persent. The NO analyzer caused
some difficulties, especially when trying to "zero" the instrument, but
that anomaly disappeared after one day. There appears to have been no
day-to-day effect of HC1 on the oxygen analyzer (semipermeable membrane)
or the hydrocarbon analyzer (flame ionization detector). The oxygen
analyzer had a soda lime scrubber to remove HC1 which was effective.
33
-------
After observing the dally operation and performance of the
instruments as the test program progressed, there were indications that
HC1 in the combustion gas may have been the source of drift and noise
problems. The manufacturers of the non-dispersive infrared analyzers
(CO, CO?, and NO) claim that no significant damage will occur to the
materials of construction if the sample containing the HC1 is moisture
free. In the initial test, when the quartz liner cracked, a moisture
laden sample reached the on-line monitoring system. Before any further
tests were performed, the complete on-line analyzer system was thoroughly
cleaned and purged with air. Before the background test was started, the
analyzer system was again thoroughly cleaned and purged with air. During
this cleaning, moisture was very much in evidence throughout the system.
3.4.6 Degradation of Gas Bag Fittings
During the sampling program at sea, it was noticed that the internal
parts of the nylon fittings were beginning to soften or melt as evidenced
by a loss of definition in sharp corners and machined threads, etc. It is
believed that HC1 in the combustion gases hydrolytically attached to the
nylon to cause polymer chain scission and physical failure of the part.
No damage was observed on the gas bags for the fuel oil background run
where HC1 was not present. The fittings were inspected each day, and
it appeared that the attack seemed to stop before the fitting leaked or
could not function.
-------
4. TEST RESULTS
Six sample tests and one background test were run. The test results
described in the following sections include:
Data taken in the field during the tests.
Data from analysis of the test samples in the laboratory.
4.1 OPERATIONAL AND FIELD DATA SUMMARY
Data acquired on board the ship during the tests covered the subjects
of waste feed rates, temperatures in the incinerators, on-line gas composi-
tion monitoring, plume characteristics, and ambient HC1 levels on the ship's
decks. Each of these are discussed iim the following sections.
4.1.1 Waste Feed Rate and Mass Emissions
Time averaged waste feed rates were recorded during the burn period.
These data are shown in Table 4-1. The time and tank volume data were provided
by the M/T Vulcanus and are contained in Appendix A. Total incineration
time for the waste was 186 hours and 45 minutes during which time 4100
tonnes of waste were burned. The average waste feed rate was thus approxi-
mately 22 tonnes/hr. This rate is based on the time required to deplete
each tank and on a depth sounding at the start and end of drawing waste
from the tank.
The theoretical emission rates of the major stack gas components from
each furnace were calculated. Based on an average waste feed of 11 tonnes/hr,
the elemental composition of the waste,an average excess air of 90 percent,
and a relative humidity of 80 percent at 29.4°C (85°F). The resulting
values were:
Mole Percent in Stack Gas
Species Emission Rate Net Basis Dry Basis
C02 6,160 Nm3/hr 8.20 8.82
HC1 4,350 Nm3/hr 5.79 6.23
H20 5,280 Nm3/hr 7.03
Q£ 6,630 Nm3/hr 8.83 9.49
N2 52,690 Nm3/hr 70.15 75.45
tor a total stack gas emission rate of 75,110 normal cubic meters (at 0°C)
p^r hour (Nm3/hr). The calculations were based on the assumptions that the
cMorine present in the waste was converted to HC1, the carbon was converted
t« C02, and the hydrogen was converted to HC1 and H20. These assumptions
have been verified by both thermodynamic equilibrium calculations and
actual measurements, which indicated that HC1 was the predominant chlorine
compound present and that CO was present only at ppm levels in the stack gas
9! J5e wa£er ^fapor em1tted with the stack gas, approximately 45 percent
(2,360 Nnp/hr) was due to the presence of moisture 1n the inlet air. Also.
35
-------
TABLE 4-1. WASTE FEED RATES
Tank
Number
2C
5C
1
4C
3C
2P
3P'
4P
5P
Volume
(cu m)
562
415
444
425
398
256
232
300
234
Start
Day Hour
3/5
3/7
3/8
3/9
3/10
3/11
3/11
3/12
3/13
2000
0400
0300
0530
0515
0500
1815
0715
0215
End
Day Hour
3/7
3/8
3/9
3/10
3/11
3/11
3/12
3/13
3/13
0400
0300
0530
0515
0500
1815
0715
0215
1415
Time
(cu m)
32
23
26.5
23.75
23.75
13.25
13
19
12
Volume
Burned
(hrs)
558
382
440
396
406
243
221
300
222
Feed
Rate ,
(T/hr)a
22.5
22.0
21.5
21.5
22
24
22
20.5
24
T tonne (1000 kg)
at 90 percent excess air, the combustion air feed rate was calculated to be
75,160 m3/hr (wet basis at 25°C), or about 16.5 percent below the maximum
combustion air feed rate of 90,000 m3/hr.
4.1.2 Temperature and Residence Time
Direct flame temperatures were measured 1n the furnaces using a Leeds
and Northrup Optical Pyrometer, Catalog No. 8621, Serial No. 1009955.
Calibration of this instrument was furnished in a notarized document from
Leeds and Northrup, Houston, Texas, in February, 1977. This instrument
operates in the 6500 A region using a 500 A band-pass. It is probable that
the HC1 in the combustion products would not have a large effect on the
validity of the temperature readings from this instrument because the
principal frequency for HC1 absorption is 6 microns (60,000 A). Any effects
at the operating frequency of 6500 A would be due to 6th or 7th overtones
which would be relatively very weak.
The data taken by the Chief Engineer of the Vulcanus, comparing the
flame temperatures with the thermocouple-derived wall temperatures is
contained in Appendix B. This comparison was performed in order to establish
a statistical relationship between wall and flame temperatures. The desired
result is a correlation between flame and wall temperature such that auto-
matic shut-off of waste to the burners can be triggered from wall tempera-
ture, using a preset value on the automatic shut-off system. The controller
minimum temperature setting was 1200°C. The controller is neither locked
nor sealed and therefore it is possible to alter the temperature setting.
36
-------
The statistical correlations between flame temperature and controller
temperature for the starboard and port Incinerators are presented 1n
Figure 4-1 and 4-2 respectively. These graphs also Indicate the variations
1n flame temperature to be expected at a given controller temperature at
different confidence levels. For example, to determine the starboard con-
troller temperature to Insure that flame temperature will exceed 1300°C at
a confidence level of 99.9%, a line 1s first drawn from 1300°C on the flame
temperature axis horizontally to Intersect the lower 99.9% confidence level
line. Next, from this intersection, a vertical line is drawn downward to
the controller temperature axis, where a value of 1180°C 1s obtained.
This means that if the controller temperature is 1180°C, there is a 99.9%
probability that the flame temperature is 1300°C or higher. Conversely,
there would be only 1 chance 1n 1000 that the flame temperature would be
less than 1300°C if the controller temperature was 1180°C.
Figure 4-2 can be used in a similar manner to determine the relation-
ship between flame temperature and controller temperature for the port
Incinerator. This technique can be used both to select a given set of
flame/controller temperatures for setting the automatic waste shut-off
system and to estimate the probability of having attained adequate flame
temperatures given the controller set temperature.
The residence time of combustion gas in the furnace can be calculated
from the stack gas emission rate (discussed in Section 4.1.1) and the furnace
volume. Since the volume of each furnace is 120 m3, the residence time, t,
for each furnace can be calculated as:
. 120 _
" 75.110 v ~ sec
3,600 x 273.16
where T is the furnace temperature in °K. The residence times calculated
are given below:
Flame Temperature Residence Time
(°C) (sec)
1100 1.14
1200 1.07
1300 1.00
1400 0.94
1500 0.89
1600 0.84
1700 0.80
37
-------
% CONFIDENCE
LEVEL
STARBOARD INCINERATOR
1000 1050 1100
CONTROLLER TEMPERATURE - ° C
1150
1200
Figure 4-1. Controller/flame temperature correlation
starboard incinerator.
38
-------
1600
1500
1400
1300
1200
O
o
I
Jtf
i
1100
1000
% CONFIDENCE
LEVEL
900
800
900
950
1000 1050 1100
CONTROLLER TEMPERATURE - °C
1150
1200
Figure 4-2.
Controller/flame temperature correlation
port incinerator.
39
-------
4.1.3 On-line Gas Composition
Signal outputs from the on-line instruments were recorded on strip
charts and an EsterHne-Angus data logging device which automatically
printed out the millivolt signal from each Instrument at two minute inter-
vals. Data was recorded during the duration of the operation of the SASS
train. The monitoring was interrupted only when 1) the SASS train was
shut down or 2) Instrument calibration checks were made during a test.
The results of the gas composition analyses, presented in Table 4-2,
were calculated on a dry basis, I.e., the samples were water free when
analyzed and no correction has been made for removed moisture. The
oxygen values have been corrected for the volumes of C0£ and HC1 which
were removed from the oxygen analyzer sample feed before analysis. The
range of levels for each species shows the maximum and minimum levels seen
for all valid data points. The average value is a numerical average of
all valid points.
The strip chart recordings of the on-line Instruments were compared
to logged sampling and monitoring events occurring during the run. When
a transient condition was observed on a recording (an unusually high or
low short term datum point Inconsistent with surrounding data) which cor-
related in time with a logged, physically observed occurrence, a review
was made of the probable association between the events. Data was only
considered nonrepresentatlve of the Incineration process when It was highly
probable that an unrelated event, such as turning on of the sampling pump,
etc., caused the apparently spurious result.
A few transient data points occurred each time the sampling system
was started, since material that had condensed in the probe before sampl-
ing was then vaporized and pulled into the gas composition analyzers.
This particularly affected the hydrocarbon and CO analyzers, which read
ppm levels. Transient "spikes" of CO and hydrocarbons on the strip chart
during start of sampling were therefore not included as representative
data points. Erratic data near the end of Test I were also eliminated,
because these points occurred after the probe liner had broken and cooling
water entered the sampling system. A transient "spike" 1n CO and hydro-
carbon data during Test IV was observed when burner #5 was restarted after
replacement of a valve associated with the #5 burner. Performance of the
Incinerator was improved after this maintenance was completed, as indicated
by an Increase in C02 and decrease In CO measurements. Both performance
levels were Included in the Test IV data; however, the transient data
during burner restart was deleted. The wide range of gas composition data
and combustion performance during Test VII, the background test with fuel
oil, resulted because feed rates were varied during the test to increase
temperatures to levels more comparable to the waste tests.
40
-------
TABLE 4-2. GAS COMPOSITION DATA SUMMARY
Test
No.
I
II
III
IV
V
VI
VIIcco>] - m, ,00
Method 3
^Background test (fuel oil)
-------
The combustion efficiency is calculated from the C02 and CO concen-
trations of the effluent gases using the following equation:
[co-] - [co]
% Combustion Efficiency = x 100
[C02J
Based on the data obtained during the monitoring of the M/T Vulcanus, the
average combustion efficiencies for each run ranged from 99.96 to 99.98
(Table 4-3).
EPA Method 3 was used to calculate the excess air in the effluent
gases. The following equation is utilized:
[02] - 0.5 [CO]
% Excess Air =
.264 [N2] - ([02]-0.5[CO])
The excess air ranged from 60 to 150% and, as seen from Table 4-2, the
average oxygen concentration ranged from 7.4 to 11.7%.
A summary for etch of the gases measured during the test series is
provided in Table 4-3, The data entries are based on the two minute
readings for each gas over the entire test series. This provides the
highest and lowest valid two minute reading observed during any of the
tests, the average reading, the standard deviation and the number of
valid 2 minute readings used. The'tolerance band provides the range in
2 minute readings which will include 99% of the data with a confidence of
90%. That is, if all conditions were exactly the same for the incinerator,
feed, etc., in further tests, 99% of the valid data acquired would be
found to lie within the range in 90% of the tests made. The tolerance
band is based on an assumption of normal population of data.
The data acquired during Tests IV, V, and VI provide a large sample
for study of Incinerator effluent characteristics. These three tests
provided data for different days of operation, different tanks of waste,
and several different probe positions. The tests were sampled extensively
through the water-cooled probe/on-Hne Instrument system, providing data at
two minute Intervals for all 5 gases (Og, C02, CO, NO, volatile hydrocarbons),
The data obtained at the two minute sampling rate was recorded on printed
paper tape. This tape was corrected before use in the statistical studies.
The correction was made by comparison with the continuous reading strip
chart recorders for each instrument, and log books, to remove the data
transients identified on the strip charts as occurring after Instrument
start-ups or probe adjustments. The corrected tape thus provides a data
base for a statistical study of probe position effects and time-oriented
furnace performance changes. A summary of the data base 1s provided for
each of the gases In Figures 4-3,-4, and-5, for the three tests. Figure
4-6 is a plot of the combustion efficiency versus time for the three tests.
42
-------
TABLE 4-3. TEST DATA SUMMARY FOR SHIPBOARD INCINERATION TEST SERIES I-VII
-P.
CO
Minimum
Maximum
Mean
Standard Deviation
Sample Size
Tolerance Band*
Minimum
.Maximum
Mean
Standard Deviation
Sample Size
Tolerance Band3
X CO
.0002634
.006826
.0025470
.001347
220
.0025470
+.0037
.33 x 10"4
.00758
.00278
.00269
55
.00278
+ .008
Test Series
% co2
5.179
12.032
9.223
2.164
158
9.223
+6.033
Test VII
4.946
14.927
9.180
2.422
54
9.180
+7.208
I-VI (Waste Tests)
% 02 % HC
4.027
15.838
10.074
2.117
237
10.074
+5.812
(Fuel Oil
1.799
13.373
7.390
4.032
58
7.390
+11.950
0
.007396
.001467
.001469
192
.001467
+.004056
Control )
0
.007589
.0007475
.001544
60
.0007475
+.004562
% NO
.0007597
.053295
.009913
.009024
229
.009913
+.0248
.0007597
.10764
.03065
.03358
60
.03065
+.09923
% Combustion
Efficiency
99.916
99.997
99.972
.015
158
99.972
+ .042
99.906
99.999
99.974
.025
54
99.974
+ .074
There 1s 90% confidence that 99% of the 2 minute test readings will be
within this range. (A normal distribution 1s assumed).
-------
o
L.3
in.
o
o
o
03
PERCENT VOLUME VS TIME TEST IV
o
o
o
o
LNJ..
O
O
CM'
O
CJ
«.o
CMC?
Oo_
Li.
III
..-o
o
o
o
CO
o
o
o
O
ID
o
o
C3
y.oo
il'.SO 10.00
TIME
11.00
1! .50
12.00
12.5Q
Figure 4-3. On-line gas composition data - Test IV, March 9
44
-------
o
o
iyj_.
PERCENT VOLUME VS TIME TESTJf
\f>
"9.00
TIMlf
11.50
12.00
Figure 4-4. On-Hne gas composition data - Test V, March 10
45
-------
li.
O
O
(-J
V.QLUME VS TIME-TEST VI
ib.oo 10.'jo
TIME
Figure 4-5. On-Hne gas composition data - Test VI, March 11
46
-------
o
C3
o
CD
CD
CD-
0)
oo
en
CD_
en
r-
o>
O)
(O
O^-i
-»CD
U_CD-
t
COJ£
CQo>_
ST'
O
CO
O)
O).
cn
CM
CD
CD.
CD
cn
cn_
CD
o
cn
COMBUSTION EFFICIENCY MS TIME TESTS IV, V, VI
1*1 .50
12.50
^9.00 9.50 10.00
TIME
10.50
.00
12.00
Figure 4-6.
Combustion efficiency data - Tests IV, V, and VI
(March 9, 10, and 11, respectively)
47
-------
The data for combustion efficiency (Figure 4-.6) were next examined for
effects of varying probe position. To do this, an average combustion ef-
ficiency was computed from the series of readings taken each 2 minutes
while the probe remained at a single position. This provided a series of
combustion efficiency averages versus probe position for Tests IV, V and VI,
as shown in Figure 4-7. The numbers adjacent to each point on the graph
indicate the number of 2 minute readings averaged and included in that
particular point. Figure 4-7 indicates that combustion efficiency was
99.96% or greater except when the probe was less than 40 cm (16 in) from
the incinerator wall. The 99.92% combustion efficiency calculated at 15 cm
(6 in) from the wall may have been affected by ambient air leakage through
the sampling port around the probe.
yy.w
99.98
OO O7
#
b
LLI
0
U.
ui OO O*\
O
P
i/> OO QA
CO
o
O oo 01
77.92
A
i
1
1
1
/i
/
TESTV
TEST V
TEST l\
10
J
/
/
/
t
\
1
S
s
i^^B
y*
^*
r
j!^
,o~*
j-
-*"22<
,-^
4
^A
! *"
7*
' 13
^2
<.
\
i
25 50 75 100 125 150 175 200 225 250 275 300 325 30
SAMPLE PROBE DEPTH, CM
Figure 4-7. Combustion efficiency vs. probe position.
Since the combustion efficiencies plotted in Figure 4-7 were calculated
from the COg and CO measurements at each probe position, the C02 and CO con-
centrations versus probe position were then evaluated (Figures 4-8 and 4-9,
respectively). These data indicate that both C02 and CO concentrations
varied more from test to test than as a function of probe position, except
for the CO data at 15 cm from the incinerator wall. CO increased signifi-
cantly when the probe was close to the wall during Test V, as shown 1n
Figure 4-7. These data suggest that incinerator combustion efficiency
could have been measured with a fixed position probe which extended some
distance greater than 15 cm into the furnace. All CO/COp measurements
beyond this point yielded efficiency calculations equal to or greater
than 99.9 percent.
43
-------
12
11
#
uT
Q
s
6
TEST IV'
TEST V -
A TEST VI
JO.
i;
28
12
13
1Q
0 25 50 75 100 125 150 175 200 225 250 275 300 325 350
SAMPLE PROBE DEPTH, CM
Figure 4-8. CC^ concentration vs. probe position
80
70
60
50
40
30
20
10
1
ILI
O
§
TEST IV
TESTV
A TEST VI
A
28
10
25 50 75
100 125 150 175 200 225 250 275 300 325 350
SAMPLE PROBE DEPTH, CM
Figure 4-9. CO concentration vs. probe position.
49
-------
4.1.4 Plume Characteristics
The plume from the Vulcanus was typical of the combustion of an
organochloride. Under conditions of high humidity and a fairly strong
breeze (>6 meters/second), the plume was observed to "touch down"
100-200 meters downwind of the ship. This behavior was observed to be
general over a range of 999 to 1009 millibars (mb), 75-96% relative
humidity (RH), and 5-17 m/sec wind velocity.
When the wind velocity decreased below 5 m/sec and the RH fell below
75-80%, the plume became diffuse, stayed aloft, and often become Invisible
At this time the addition of ammonia gas through a standpipe between the
stacks and at the plane of the stack outlets produced a visible plume of
ammonium chloride. The plume could thus be traced when conditions pre-
vented droplet condensation involving HC1 gas. When the humidity was high
(I.e., above 75-802) and the wind velocity was low, the plume was voluminous
white and dense and remained aloft. Under these conditions, touch down was '
not observed for an estimated 5-10 miles.
The threshold conditions for aerosol formation for HC1-water vapor
systems have been studied (Reference 6). In general, the requirements for
visible HC1 droplet formation and growth, i.e., a visible plume, are 78%
RH minimum, nuclei of NaCl or MtyCl of approximately 0.02 y diameter to
provide condensation localities, and HC1 in the vapor around the condensing
droplets. When water condensation occurs, a mist will form regardless of
the HC1 concentration in the vapor. Conversely, if no condensation of water
occurs, no mist will form, even if HC1 is present in large amounts. Personnel
of the M/T Vulcanus stated that a visible plume was always obtained when
burning a European waste in the North Sea, as opposed to the diffuse
plume discussed above. The reasons for this difference are not obvious
at this time. Analyses of the two wastes showed the European waste to contain
5.4% hydrogen and 40% chlorine, while the Shell'waste contained 4.2%
hydrogen and 63% chlorine. Thus the Shell waste would be expected to
yield combustion products lower in water content and higher in HC1 content
than the European wastes. However, this difference does not necessarily
completely explain the lack of a plume in the Gulf of Mexico.
4.1.5 HC1 Measurements on Board
Some combinations of wind velocity, wind direction and ship attitude
resulted in contact of the plume with the decks of the ship. Even when
the plume was not evident, it was occasionally possible to detect the
presence of HC1, both by its odor and by direct measurements for HC1 with
Drager detection tubes. This behavior lead to consideration of atmospheric
conditions and ship-wind combinations which must be avoided if contact of
the plume with the ship is to be minimized. Table 4-4 lists these condi-
tions. The wind forces listed in this table of from 6-17 meters/second
correspond to velocities of 12-34 knots (1 knot - 1/2 m/s). Matching the
ship's speed to a vectored velocity of the wind in the ship's direction
produced a plume which remained aloft and which results in HC1 values on
the ship of zero.
50
-------
TABLE 4-4. HYDROGEN CHLORIDE (HC1) IN AIR ABOARD VESSEL
en
Date
3/9/77
3/10/77
3/11/77
Time
(see
Moire
1800
0845
1400
1700
1030
1745*
Location
4-n
Main Deck
Combustion
Deck
Stern
Bridge
Combustion
Deck
Main Deck
Bridge
Combustion
Deck
Main Deck
Bridge
Combustion
Deck
Main Deck
Stern
Bridge
Combustion
Deck
Stern
Bridge
Combustion
Deck
Main Deck
Stem
Relative
Humidity
(X)
66X
m
N
821
M
N
86X
H
H
96X
N
H
H
861
H
N
87X
H
II
H
Ship
attitude
055°
M
M
258°
II
H
245°
*
235°
M
H
080°
H
H
320»*
«
H
Wind
Direction-Force
150°-6 m/s
M
H
140°-11 m/s
M
H
150°-10 m/s
H
M
130°-9 m/s
*
H
140°-17 m/s
H
N
140**-! 2 m/s
M
H
H
Barometer
(mbl
1012
It
N
1007
N
M
1004
M
H
1002
M
II
H
999
M
H
999
H
H
Cone.
HC1
(Dmri
1.5
1
1
0
4
2
0
2
1
2
6
1.5
0
0
2
4
0
0
0
0
Captain oriented ship so as to match speed to plume vector 1n ship's direction
plume high aloft - no HC1 on board
-------
The various locations where the Drager HC1 samples were taken are
marked on the plan of the M/T Vulcanus shown in Figure 4-10. The downwind
(lee) side of the ship showed HCl-positive data. No HC1 was detected on
the windward side.
4.2 ANALYTICAL RESULTS
Of the six waste incineration sampling tests, sampling train malfunc-
tions in the first three tests limited the sampling duration to less than
one hour. Consequently, waste incineration Tests IV, V, and VI, with
sampling durations in excess of two hours, were selected for the detailed
laboratory analyses. In addition, samples acquired from Tests VII, the
background run burning fuel oil alone, were also analyzed as a baseline
test for comparison to the waste incineration tests. A test and sample
summary is presented in Table 4-5, and the results from analyzing these
tabulated samples are presented and discussed in this section.
4.2.1 Waste Feed Analysis Results
This section presents the results of the organic and elemental
characterization of the representative waste feed samples taken during
each test.
4.2.1.1 Organic Constituents
Neat samples of the blended waste feed (BWF) from each test were
analyzed on a gas chromatograph equipped with a flame ionlzation detector
(FID) and an OV-17 column (see Section 3.3 for details of the procedure).
The chromatograms were examined and there was a peak-for-peak, shoulder-
for-shoulder match for each sample. Therefore, a careful analysis of
any one of the BWF samples was applied to all of the representative waste
feed samples. The representative sample from Run IV was chosen for the
detailed analysis.
The composition of the waste is presented in Table 4-6. Qualitative
identification of the waste constituents was accomplished by GC/MS while
quantification was made on the basis of the GC analyses using the flame
ionization detector (FID). The percentages have been adjusted slightly to
reflect the fact that trichloropropane was present at 20 percent on the
basis of direct calibration for that compound. (Trichloropropane is used
as the basis for one of the waste destruction efficiency calculations.
This is discussed in Section 1.)
4.2.1.2 Elemental Characterization and Estimated Emission Rates
In order to estimate the hourly elemental emission from the waste
incineration, duplicate samples of the Shell waste were surveyed for
elemental content. The waste samples from each day of testing were first
blended; then two aliquots of this blend were combusted over nitric acid
in a Parr bomb. The resulting solutions were analyzed by spark source
mass spectrometry (SSMS) which is a semi-quantative survey techique.
The major inorganic constituents (approximately equal to or greater than
500 ppm) were calcium, chlorine, magnesium, phosphorus, and sodium.
Other trace elements found are reported 1n Table 4-8. The average detection
limit for those elements not discovered in these waste samples was 0.5 ppm.
52
-------
Figure 4-10. HC1 sampling locations
1) combustion deck 2) bridge 3) main deck 4) stern
53
-------
TABLE 4-5. SUMMARY OF VULCANUS SAMPLES
Test Number
Date of Test
Length of Sampling Time (min.)
Volume of Gas Sampled (dry std. m )
Moisture in Gas Sample (vol. %)
SASS Samples Collected
Quartz Probe Liner Wash (PW)
Particulate Filter (PF)
Heat- traced Sample Line
Wash (HLW)
XAD-2 Sorberrt Trap (ST)
Module Trap Hashes
Water (MTW-H^)
Isopropyl alcohol (MTW-1PA)
Pentane (MTW-P)
Pentane/methylene chloride
(MTW-PM)
Module Trap Condensate (Cond)
Contained Liquid Impinders
(01)
Acidified split (CLI-A)
Basic split (CLI-B)
Silica Gel (L4)
Blended Waste Feed (BWF)
Tedlar Bag Samples (GG)
IV
3/9/77
150
12.5
15.2
-
X
X
X
X
X
X
X
X
X
X
X
X
X
X
V
3/10/77
160
13.4
15.4
-
X
X
X
X
X
X
X
X
*
X
X
X
X
X
VI
3/11/77
150
12.1
20.0
X
X
X
X
X
X
X
X
X
X
x ;
X
X
X
X
VII
3/13/77
180
15.8
9.3
X
X
" >
X
X
-
X
X
X
X
X
X
X
X
X
X
54
-------
TABLE 4-6. REPRESENTATIVE WASTE COMPOSITION BY GC/MS
Compound
Hole %
1,1-01chloroethane
B1s(D1chloromethyl)ether
l,2-D1chloroethane
1,2-D1chloropropane
1,2-D1chloropropene
Tetrachloroethy1ene
1,1,2-Tr1chloroethane
1,2,3-Tr1chloropropane
2,3-D1chloro-l-propanol
2-Chloroethy1chloroformate
1,2-Bromochlorocyclobutane
1,2,3,4-Tetrachlorobutane
Contains 4 Cl, possibly 1,2,3,3-Tetrachloropropene
Possible 2,3-D1chloro-2-methylprop1onaldehyde
Halogenated, unknown, MM at least 141
2-Bromo-l,2-dlchloropropane
Halogenated, MU at least 189
4
7
14
27
11
0.3
6
19
3
0.7
0.9
2
0.5
0.3
0.5
1
2
55
-------
TABLE 4-7. CALCULATED APPROXIMATE EMISSION RATES OF INORGANIC ELEMENTS
Element
Lead
Ban' urn
Iodine
Silver
Molybdenum
Zi rconi urn
Strontium
Rubidium
Bromine
Sel eni urn
Arsenic
Gallium
Zinc
Copper
Concentration
in Waste
(ppm)
5-20
10-20
2-4
1-8
10-20
1-5
5-30
0.5-1
5-10
1-5
1-5
0.5-2
10-30
10-30
Calculated
Emission
Rate (kg/hr)
0.1-0.4
0.2-0.4
0.04-0.09
0.02-0.2
0.2-0.4
0.02-0.1
0.1-0.7
0.01-0.02
0.1-0.2
0.02-0.1
0.02-0.1
0.01-0.04
0.2-0.7
0.2-0.7
Element
Nickel
Cobal t
Iron
Manganese
Chromium
Titanium
Scandium
Potassium
Sulfur
Silicon
Aluminum
Fluorine
Boron
Lithium
Concentration
in Waste
(ppm)
10-100
1-5
30-400
1-5
5-200
10-20
0.1-1
-300
30-60
90-100
10-50
10-50
1-10
0.5-2
Calculated
Emission
Rate (kg/hr)
0.2-2
0.02-0.1
0.7-9
0.02-0.1
0.1-4
0.2-0.4
0.002-0.02
~7
0.7-1
~2
0.2-1
0.2-1
0.02-0.2
0.01-0.04
en
01
-------
Approximate mass emission rates were calculated by multiplying mg/kg
(ppm) concentrations by the average feed rate of the waste (22,000 kg/hr).
Thus a metal present at 10 ppm in the waste would be emitted at approxi-
mately 0.22 kg/hr. These calculated emission r*ates are also included in
Table 4-8. Emission rates for the major Inorganic constituents (Ca, Cl,
Mg, P, and Na) were ^10 kg/hr. Total emissions can be estimated by
multiplying the above rate by the duration of the burn.
4.2.2 SASS Train Sample Analysis Results
This section describes the results of the analysis performed on the
samples removed from the SASS train. Table 4-6 summarizes the samples
and their sources. Direct analysis for known and expected waste constitu-
ents was made by combined gas chromatography/mass spectrometry (GC/MS).
Additionally, a survey analysis was made wherein an aliquot of each sample
extract was evaporated, the residue weighed, and qualitative analyses
made on the residue using infrared (IR) and low resolution mass spectrom-
etry (LRMS). These methods are described in Section 3.3.2.
4.2.2.1 GC/MS Analysis
Combined gas chromatography/mass spectrometry (GC/MS) was used for
the direct analysis of the appropriate samples for known waste constituents.
The samples were the solvent extracts and/or concentrates of the SASS Train
components such as probe rinses, filters, line rinses, module rinses and
sorbent traps. Two column systems were used:
a) Three percent OV-17 on Chromosorb W was used for all samples.
b) Chromosorb 101 was used for more effective separation of the
more volatile species when they were detected using the OV-
17 column.
The results of these analyses showed that only the samples from the
sorbent trap and the trap module wastes contained chlorinated species.
The amounts in the trap module washes were so very much lower compared to
levels found in the resin extract that in summing these, the final levels
for the module samples reported in Table 4-8 did not change in the one
significant figure reported. All four of the sorbent trap (ST) samples
contained chlorinated organics. This includes the trap extract from the
background test but the number of chlorinated compounds and their levels
were very much less. Two reasons why the background trap extract con-
tained chlorinated species are postulated, 1) contamination during sampling
and analysis, 2) residual chlorinated material in the tanks and lines which
contained the fuel oil used in the background test. No special efforts
were made to identify the source of the chlorinated species found in the
MV-VII background resin extract.
The list of reported compounds include some that were not found in the
original waste analysis. The source of these materials may be, 1) low
level waste constituents not positively Identified in the waste analysis
(probably less than 0.5% in the waste); 2) unexpected by-products of the
incineration process; or 3) products from unknown reactions going on in the
sorbent trap resin. No further work was performed to identify the source
of these materials as it was considered outside the scope of the analysis
task.
57
-------
TABLE 4-8. EFFLUENT GAS CONCENTRATIONS FROM ANALYSIS OF SORBENT TRAP
EXTRACTS
Spectrum
No.a
18
35
41
52
73
89
153
164
184
196
200
213
224
239
247
253
258
266
271
279
295
298
330
345
352
360-362
368
Assignment
C8H18
Dichloropropane
C8H18
C9H20
C10H22
Trichloroethane
Tetrachloroethane b
Trichloropropane
Dichloropropanol
C* Substituted Benzene (e.g.,
di ethyl benzene)
Alkyl, Oxygenated Aromatic MW 164
(e.g. , Butyl Cresol )
03 Substituted Styrene
Unsaturated Chlorinated Hydrocarbon b
MW 2162, Possibly Tetrachlorobutene
C6 Substituted Benzene (e.g.,
Triethylbenzene)
Tetrahydronaphthalene
Cl4 Hydrocarbon MW >258b
Chlorinated Aromatic MW >166b
Naphthalene (or Azulene)
C* Substituted Benzene Mixture
Chlorinated Aromatic b
Chlorinated Aromatic*5
Chlorinated Aromaticb
Chlorinated Aromatic MW >200+b
Dichloro, C4 Substituted Benzeneb
Uichloro, C4 Substituted Benzene b
Dichloro, C5 Substituted Benzenesb
Tetrachloro Cy Substituted Benzeneb
(MW >246)
378 j Trichloro, Alkyl Substituted Benzeneb
, (MW >236)
390 ' Trichloro, Alkyl Substituted Benzeneb
416
1
L_ 1
(MW >236)
Hexachlorobenzene b
Total Chlorinated Organics, mg/m
Total Organics, mg/m
IV
-
-
-
-
-
-
0.09
0.1
0.04
0.6
0.3
0.2
0.1
0.1
0.05
-
0.07
0.8
-
0.02
0.2
0.3
0.2
0.8
0.4
0.1
0.2
0.1
0.2
0.04
3
5
i
V
0.02
0.04
0.04
0.02
0.04
0.06
0.09
0.6
0.04
0.6
0.3
0.3
0.5
0.1
0.04'
0.1
0.1
0.7
-
0.04
0.2
0.4
0.2
0.8
0.5
0.5
0.2
0.2
0.3
0.04
5
10
VI
-
-
-
-
-
_
0.06
0.2
0.04
0.6
0.3
0.2
0.1
0.1
0.04
0.02
0.05
0.8
-
-
0.2
0.3
0.3
1.0
0.5
0.2
0.2
0.3
0.3
0.05
4
6
1
VII
-
-
-
-
-
-
-
-
-
0.2
0.1
0.2
_
0.2
-
-
-
0.5
0.1
-
0.1
0.4
0.05
0.1
0.1
-
-
^
»
-
1
2
a'Spectrum numbers refer to GC/MS run for MV-V-ST which is shown
in Appendix E.
^Chlorinated compounds not confirmed as being present in the
waste.
58
-------
Table 4-8 also presents the summation of chlorinated organics and
total organics (Including the chlorinated organics) found in the3resin
extract (ST) samples. The range of these totals is 1 to 10 mg/m and cor-
responds to destruction efficiencies greater than 99.99%. These destruction
efficiencies were not reported 1n the summary of this report since additional
material may be present in the sample and may not have eluted from the
chromatographlc column. This would have lead to erroneously high destruc-
tion efficiency values.
Table 4-8 contains several compounds which were only partially identified
Into a general class of compound. There are many reasons why partial identi-
fications can only be made. Several of the most important are:
1) too little component to create a usable spectrum
2) unstable molecular ion which provides no molecular weight information
3) similar compounds and isomers which produce virtually identical spectra
4) time and budget limitations on the interpretation of the data
The lower limit of detectabllity for this analysis was established
using trichloropropane standards. It was determined that trichloropropane
can confidently be reported when 10 ng are injected in a 10 microliter
solution volume. Within the current error limits, this detectabi1ity is
presumed to be valid for all the chlorinated species found in the waste.
All samples were concentrated to 10 ml; therefore, the detection limit is
approximately 0.01 mg for the waste feed constituents. Even though the
chromatogram for the sorbent trap (ST) extract from the background test VII
1s not as similar as the other ST samples from waste burns, there still is
enough similarity between them to suggest that similar compounds were
recovered from the sorbent trap regardless of whether the fuel was waste
or No. 2 oil. This similarity is shown in gas chromatograms reconstructed
from the 6C/MS data (see Appendix E). The source of the commonality in
extract composition likely lies in the sorbent resin and/or the manner in
which,it is extracted and prepared for analysis. No work was performed to
Investigate this situation as 1t was considered outside the scope of the
program. In like manner, the source of the chlorinated compounds found in
the sorbent resin from Test VII (background) is difficult to explain con-
sidering that fuel oil (typically with low ppm concentrations of chlorine)
was the burner feed. It is also unlikely that this represents cross-
contamination from the test equipment since a new probe was used and the
sample lines had been flushed with pentane.
The levels of waste constituent material found in the sorbent trap
module washes (MTW) was very much lower than the levels found In the
sorbent traps. In fact, the levels do not exceed 0.006 mg/m3 (1 part
per billion). These amounts are only about 1 to 2 percent of the amount
found in the sorbent traps. Considering the overall accuracy of the
analysis and the number of significant figures 1n the results 1n Table
4-8 the trace amounts found 1n the module washes were not tabulated.
They also do not affect the calculated waste destruction efficiencies.
GC/MS analysis of the probe rinse samples found no waste constituents
or other chlorinated organics.
59
-------
4.2.2.2 Gravimetric Analysis
The gravimetric results, listed in Table 4-9 show the distribution
of nonvolatile organic materials through the SASS train. The majority of
the material appeared in the IPA module rinse, the first solvent used in
the series of rinses. However, due to the polarity of this solvent, a
question arose as to the possible presence of inorganics in these IPA
samples. Thus thermal gravimetric analysis (TGA) were performed which
showed 20 to 40 percent residues remaining at 900°C. On the basis of the
TGA results, the IPA samples were assumed to be -70 percent organic on the
average, and the values in Table 4-9 were adjusted on that basis.
TABLE 4-9. GRAVIMETRIC RESULTS ON ORGANIC SASS SAMPLES
Sample Description
Filter extract, mg
Heat-traced line rinse plus XAD-2
resin extract, mg
Condensate extract, mg
Water module rinse extract, mg
Pentane module rinse plus pentane/
methylene chloride module rinse, mg
Isopropyl alcohol module rinse, mg
Total, mg
Test IV
0.00
320
0.00
0.00
0.00
130b
450
Test V
0.70
110
0.00
0.00
1.4
51 Ob
620
Test VI
0.00
97
1.1
0.00
0.92
480b
580
Test VII
0.00
17
0.01
_a
48
490b
560
a)
'No sample taken.
'Values are 70 percent of numbers obtained by gravimetry. Adjustment
was made on the basis of TGA data showing the IPA rinses to be
60 to 80 percent inorganic materials.
The only other significant source of nonvolatile organics from the
SASS train was the combined XAD-2 resin extracts and heat-traced line
rinses. In particular, no significant amount of organics were collected
in the filters, water module rinses, or module condensates. The rinse of
the probe used for the waste tests contained only 6 mg of nonvolatile
organics and the rinse of the background test probe contained no non-
volatile ortanics. The 6 mq of non-volatile organics is especially
insignificant when one considers the volume of gas that passed through
that probe (>50 m3).
4.2.2.3 Infrared Analysis
The infrared analysis qualitatively identified chlorinated hydro-
carbins in the following samples: the sorbent trap module IPA wash from
Test VI, and the XAD-2 resin extracts (combined with heat-traced sample
line washes) from Tests IV, V, and VI. Also present in the sorbent trap
60
-------
module IPA wash samples from Tests V and VII were an inorganic acid and a
hydrocarbon. The pentane/methylene chloride sorbent trap module washes
from Tests VI and VII contained a secondary alcohol (glycol).
4.2.2*4 Low Resolution Mass Spectrometry Analysis
Samples for LRMS analysis were selected on the basis of the gravimetric
results. In general, any sample from Tests IV, V, VI, or VII with a
residue weight of 100 n9 or more was analyzed. The analysis was performed
using the direct Inlet or solids probe technique. The analytical details
are found in Section 3.4.2. The objective of this kind of qualitative
analysis 1s to determine 1f there 1s anything present 1n the samples which,
because of Its functional group composition or compound class, might be
considered hazardous (e.g., chlorinated compounds, polynuclear aromatics,
etc.) and require further examination using more sophisticated methods.
All compounds discussed in this section were identified from their
spectral patterns, which must be extracted from the other spectral informa-
tion caused by other compounds in the sample. It is therefore possible
that these compound 1ndent1f1cations are not entirely correct 1n every
case. For example, chlorinated compounds are usually easy to detect due
to the Isotopic pattern of the chlorine atom; however, exact compound
Identification can be elusive.
To summarize the LRMS results, nearly all of the samples contained
HC1 and H20. Most of the samples contained a group of materials that have
been found as artifacts in the majority of samples from previous incinera-
tor sampling work. These ubiquitous materials consist of phthalic add
esters, hydrocarbon oils, fatty add compounds (soap), and glycols. These
easily identified substances are of little significance 1n the assessment
of the process performance. There were some aromatic and halogenated
species detected and these are listed in Table 4-10. Although they were
found mainlv in the sorbent trap extracts, some were also observed in the
heated line washings and the module trap washings. The majority of these
compounds would also be seen and more properly identified in the GC/MS
analysis of the same unevaporated samples, with the exception of the benzole
add and the CSQ phenyl compounds.
4.2.2.5 Analytical Recovery for SASS Train Samples
Two samples were prepared to determine the analytical recovery. One
milligram of the waste feed was added to an unused sorbent trap canister.
A water sample doped with Img of waste feed was also prepared to which HC1
and iron, nickel and chromium salts were added. This simulated both the
condensate solution and the aqueous sorbent trap module washes. This
aqueous sample was split and one part was neutralized to pH7 with sodium
hydroxide to simulate the manner 1n which the aqueous samples were treated
In the analysis. Of the compounds present 1n the waste, only trichloro-
propane was found in these doped samples by GC/MS analysis. This is not
61
-------
TABLE 4-10. LRMS RESULTS ON ORGANIC SASS SAMPLES
Sorbent Trap
Heated Line Wash
Sorbent Module Wash
Chlorotoluene
Dichloro isopro-
pyltoluene
Benzoyl chloride
Benzoic acid
Butyl aniline
C3Q phenyl compound
Dimethyl allene
phosphom'c acid
Mixed benzoates
Ethyl benzaldehyde
Chlorotoluene
A highly fluorinated
chlorobutane (possibly
tetrachlorohexaf1uoro-
butane)
2296
or Cly compound,
Chlorotoluene
A highly fluorinated
chlorobutane (possibly
techtrachlorohexaf1uoro-
butane)
or Cly compound
surprising at the levels at which the control samples were doped. Trichlo-
ropropane is one of the major waste constituents, and many of the lesser
species did not exceed our estimated detection threshold. Recoveries
based on trichloropropane are reported in Table 4-11.
TABLE 4-11. ANALYTICAL RECOVERY FOR SASS SAMPLES
Sorbent Trap
Sorbent Module Condensate
Sorbent Module Condensate
Neutralized
Trichloropropane
Added (pg)
200
100
100
Trichloropropane
Found (yg)
220
60
40
Recovery
(Percent)
110
60
40
Another source of sample loss is that constituents may be lost in
the concentration steps. These samples were concentrated up to one
hundredfold using the fractional distillation process of a Kuderna-
Danish evaporator. Any waste constituents which have a higher vapor
pressure than the pentane, acetone or isopropanol solvents will be
preferentially lost. Moreover, some of the less volatile compounds may
62
-------
be partially lost due to the Incomplete effectiveness of the fractional
distillation column. Any or all of the sources of Incomplete sample
recovery may affect any of the samples which were treated.
The data reported for the sorbent trap extracts In Table 4-8, were
not corrected for analytical recovery because the sorbent trap recovery
sample yielded more than 100 percent; therefore no correction was required.
The sorbent trap module wash samples were corrected for 40 percent
analytical recovery, but they still resulted in the trace levels reported.
It Is stressed that these tables reflect analytical recovery changes only.
Trapping efficiency for the sorbent traps was derived from the graphs 1n
Reference 5.
4.2.3 Gas Bag Sample Analysis Results
Two analytical procedures were used to characterize these samples.
Low Resolution Mass Spectrometry (LRMS) of the neat, as
received, bag sample.
t Concentration of trace constituents using absorption on Tenax,
followed by desorption and GC analysis.
4.2.3.1 LRMS Analysis
A portion of a Tedlar bag sample from each of Tests IV through VII
was analyzed by low resolution mass spectrometry for:
a) Semi-quantitative composition of the major constituents,
e.g., 02, N2, C02, HC1, etc.
b) Qualitative determination of levels of the chlorinated
organics known to be present in the waste feed.
The major constituents are presented in Table 4-12. Comparison of
the data with that from the on-line analyzers shows fair agreement with
02* C02> and ^. Though attack and degradation of nylon valves on the
TABLE 4-T2. GAS BAG CONSTITUENTS BY LRMS
Test Number
IV
V
VI
VII
Concentration, Mole Percent
H2
<0.001
0.01
0.006
<0.001
H20
0.02
0.03
0.1
0.03
N2/CO
79.3
79.2
79.6
79.7
°2
15.6
15.4
10.8
14.3
Ar
1.1
1.1
1.1
1.1
co2
3.9
4.3
8.4
4.8
HC1
N/D
N/D
N/D
N/D
N/D = Not detected. Limit of detection was 200 ppm.
63
-------
Tedlar bags had occurred, this demonstrates that the gas 1n the bag was
not ambient air resulting from leakage or other sample loss and that the
bag sample integrity was adequate for analysis of residual trace
chlorinated organics in the effluent stream. The low levels of HC1
reported in the LRMS data can be explained by the fact that the nylon
fittings on the Tedlar bags were found to have undergone hydrolytic
attack by the HC1.
A search of the same data was made for all chlorinated compounds and
specifically known waste constituents. However the technique was not
sensitive enough to yield adequate detection limits. This led to the use
of the technique discusssed in the following section.
4.2.3.2 Tenax/GC Analysis
The results of the Tenax concentration and GC/FID analysis are shown
in Table 4-13. Calibration for the waste constituents was performed
using trichloropropane. This major constituent of the waste was used as
a tracer for all waste constituents. Waste destruction efficiencies were
calculated on the basis of trichloropropane analysis results.
Also presented in the table is a measure of the total organics which
were detected by the FID. The levels were calculated from total GC peak
area integration data. Response was determined from injections of known
amounts of organic material. The ppm values are reported as methane for
ease in comparing data with the on-line hydrocarbon analyzer data. These
levels should not be considered to be more accurate than ±100 percent.
The destruction efficiencies calculated from the trichloropropane and
total organic data are discussed further in Section 1.
TABLE 4-13. GAS BAG ANALYSIS BY TENAX/GC
Test
IV
V
VI
VII
Trichloropropane
mg/rn3 (ppb)
<0.06 (<9)
0.12 (21)
0.12 (21)
Waste
Destruction
Efficiency3
(DEGGC3C13)
>99.999
>99.999
>99.999
Total Organic
Material
mg/m3
(ppm as Methane)
9 (12)
186 (27)
12 (18)
Total Organic
Destruction
Efficiency
(DEQGHC)
Percent
99.993
99.991
99.991
99.996
a'Based upon trichloropropane as a typical waste constituent and present
in the waste at 20 percent (W/W).
64
-------
4.2.4 Burner Head Residue Analysis Results
The requirement for routine cleaning of the burner heads was frequent
during the incineration of the waste. There was interest in characteriz-
ing the burner head residue for hazardous constituents since the residue
Is a potential disposal problem. A sample of the residue weighing
51.7 grams was broken up into 10-12 mesh size and extracted with pentane.
Several analyses were performed on this sample and the results
are as follows:
A 1ml aliquot of the concentrate was evaporated down just to
dryness and weighed. The residue amounted to 3.6 weight
percent of the original burner head sample.
t An IR spectrum was obtained of a smear of the residue and the
spectrum indicated the presence of hydrocarbons and possibly
some glycols. No trace of the waste constituents was found.
The low resolution mass spectrometer (LRMS) data indicated
HC1, trichloropropane, bromodichloropropane and hydrocarbon
oil.
The burner head extract was also analyzed by GC/MS. The
results show that most of the compounds in the waste feed
(as seen in the representative waste sample) were also present
in this extract. The GC separated the mixture of the earner
eluting compounds very effectively, and most of the waste
constituents were identified. The second portion of the
chromatogram consisted largely of incompletely resolved peaks
constituting a major portion of the total response seen in the
chromatogram. The general character of the compounds in this
higher boiling portion of the mixture shown by their spectra
was that of aromatic compounds, some of which were chlorinated.
Intensive effort to completely separate and identify each
compound in the mixture was not carried out due to budget and
schedule considerations. Complete quantification is not
achievable for an additional reason, which is that much of the
material in the weighed residue discussed above is likely to be
polymeric "tars", an unknown portion of which is not likely
to elute from a gas chromatograph. The identified compounds
and their calculated concentration (wt/wt) in the burner head
scrapings are presented in Table 4-H, The reconstructed
chromatogram of this sample is shown in Appendix E.
65
-------
TABLE 4-14. COMPOUNDS FOUND IN BURNER HEAD RESIDUE
Compound
Concentration in
Burner Residue mg/kg
Dichloroethane
Dichloropropane
Dichloropropene
Trichloroethane
Trichloropropane
Dichloropropanol
bis(2-chloroethyl)ether
Tetrachlorobutane
15
70
50
85
120
115
80
80
66
-------
5. REFERENCES
1. Wastler, T. A., et al.f "Disposal of Organochlorlne Wastes by
Incineration at Sea," EPA-430/9/75-014, July 1975.
2. TRW Report on "Destroying Chemical Wastes in Commercial Scale
Incinerators", Contract No. 68-01-2966, Facility Report No. 1,
The Marquardt Co., October 1976.
3. TRW Report on "Destroying Chemical Wastes in Commercial Scale
Incinerators", Contract No. 68-01-2966, Facility Report No. 6,
Rollins Environmental Services, April 1977.
4. Federal Register, June 28, 1976, Part III. Environmental Protection
Agency - Ocean Dumping - Proposed Revision of Regulations and
Criteria.
5. Levins, P., et al, "Selection and Evaluation of Sorbent Resins
for the Collection of Organic Compounds," EPA-600/7-77-044,
April 1977.
6. Fenton, D. L., and Renade, M. B., "Aerosol Formation Threshold for
HC1-Water Vapor Systems," Environmental Science and Technology,
10:1160-62, 1976.
67
-------
APPENDIX A
SCHEDULE OF WASTE CONSUMPTION BY TANK NUMBER AND LOCATION
68
-------
VO
Tank No. 6 BB
CBM - 112
HT
Tank No. 5 BB
CBM - 117
11 5m Vl 07m5
start 3-13 021=
stop 3-13 1*15
12hrs-24tA
with SB. Tank
Tank No. 5 Centre
CBM - 415
382 m5
start 3-7 04OO
stop 5-8 OJOO
23hrs.-22t/h
Tank No. 6 STI
CBM - 112
Gasoel
Tank No. 4 BB
CBM- 150
142m5+158m5
start 3-12 0715
stop 3-13 0215
19hrs-20.5t/h
;ith SB.Tank
Tank No. 4 Centx
CBM = 425
396m5
start 3-9 0530
stop 3-10 0515
ca.24hrs=21,5tA
Tank No. 5 STI
CBM - 117
Tank No. 3 BB
iWiiW
start 3-11 181;
stop 3-12 071;
1.3nrs=22t/h
with SB.Tank
e Tank No. 3 Centre
CBM -398
406m5
start 3-10 0515
stop 3-11 0500
ca.24hrs-22tA
Tank No. 4 STI
CBM =150
Tank No. 2 BB
CBM,= 128,
119mVl24.ni5
start 3-11 0500
stop 3-11 1815
ja.13hrs.524tA
*ith SB.Tank
Tank No. 2 Centre
CBM = 5$2
55Bm5
start 3-5 2000
stop 3-7 0400
32hrs-22,5tA
Tank No. 3 STI
CBM - 116
Tank No. 2 STB
CBM - 128
MTS "VULCANUS "
Vulcanus Shipping Co. P
S INGAPORE
start Heat ins up 3-5- 0800
start Waste Combustion 3-5
stop Waste Combustion 5-13
Tank No. 1 ^v^
CBM ," 444 ^Vj
440m5
start 3-8 0300
stop 3-9 0530
26,5hrs.-21,5t/h ^^
Waste Combust ion-186hrs 45M!
ca.22t/h
Ship
per
TOT
r^RM
i
2 BB
2 CTR
2 ST!
3BB
3CTR
3 ST
4BB
4CT
4ST1
5BB
5CTR
5STB
6BB
6STB
T01AL CBM
5l63m'-ca 4100t.
-,
-------
APPENDIX B
TABULATION OF TEMPERATURES IN FURNACES BY
a) Optical Pyrometer (Flame Temperature)
b) Wall Thermocouples
70
-------
Incinerator ?anpcratx'.res
(ohip Vulcanus)
:tarboard Cven °C
Oqto ! nouri^,'oclc Inclic^tor ! Controller
J i 3ox - .. i
5/5
3/5
3/5
3/5
3/5
3/6
3/6
3/r
3/5
3A
3/5
3/6
3/6
3/6
j zero
i 2100
S22CO
I
i 23CO
! 2400
J01CO
!0200
i
1
j 0300
J04CO
i 0500
! O600
I
|0300
! 0900
3/6 ] 1000
3/6
3/6
Vo
3Ai
t 1
j 1100
', 1200
! 1300
1---
1
/> I 1'« >
i
i
070
960
1010
1030
1050
1O50
1090
1110
1030
1080
1100
1120
1130
1150
1160
116O
j
1160
11GO
-~~~ '
1?10
^80
c>90
1020
1050
10oO
107?
11 CO
1130
1090
1090
112C
114O
1150
1160
1170
1170
i
1170
J170
1210
r-vo
i 12?0
| 122O
! 1220
j 1350
[ 124O
J1250
i 1300
|1~350
J1210
|1200
! 1320
.1310
J1320
11330
! 1330
| 13OO
| 1280
j 1280
i 1320
! 1350
I
I
J! Tort Oven °C
f> W
Pyrometer ""^[jjf In'*icn'tor ' Controller! Pyrometer '
1450
1450
__
1490
1560
L _- _
1450
1610
I
h
1620
;ij 1°20 !l030 i 1170
;; 1030 MO4o j 1160
;; 1^50 '1070 | 1150
. 11 j i
!S 1070 [1090 ! 1200
"t ^_.__^l,^__.J.~ _L
"" 1110 11120 « 1240
J! 1120 |1130 | 1200
jjjj 1130 fl140 | 1230
l! 1160 J1160 ! 1260
«« j. i _
; 1190 11190 j 1320
J! 1170 [1170 J 1210
" 1170 J1170 ! 1260
;; 1160 11160 ! 1200
jjj 1150 jH5oy 1200
;; 1150 [1150 i 1220
1700
"~530~
1650
HI! 1150 |1160 ! 1230
; 1160 !l160 j 1230
jlf 1160 H160~|~1210
If "7 I
"_ 1160 p160J 1210
?! 1210 J1210 ! 1420
wt J i
",', 1200 M200 | 1270
rj- J- -J
* I
*< I
1390
-._---.----.
1380
1480
1510
L _
___________
1600
1510
1450
1520
1450 ~"
over 1700
j
~i
i
i
i
»_._. -j., __-..---.--..
.-.-..4_.-.-.«.
I
t
l
I
-»-....
l
-------
ro
\&
./;.
Vrv.-t t'vr.r.
Dfltp
3/6
5/6
3/6
3/&
3/3
3/6
i 3/6
! 3/6
! 3'?
i
} 3/7
! T/7"
3/7
j 3/7
| V,
! ',/>'"
] 3/V
3/7
! 5/7
i 3/7
J "
i
1 "Ol
! 1600
;1?00
1 1900
]2000
12100
2200
J- --.
(2300
j 2400
;oioo
10200
i
"i03OO
J0400
! ^500
|U60O
1
"10700
j^-800
] 0900
i
Ticoo
1
J1100
1
"' ! aE*
11195
j 11 85
11185
I
11180
j 1130
;11GO
i 1140
j
, mi.0
1 1140
j1140
i 117O
"i 119O
;1190
]1170
i "11170
i
^1170
J11&0
11170
]ii 70"
i
]l180
i
i
i
i
_J,vacnt
'1195
M135
j 1185
j 1185
j 1185
"j 1160
J 11r>0
J1150
j
J1150
"J1155
j 1190
! 1200
.1
i 120O
j 1170
! 1170
! 1170
i
1 1160
j 1180
! 1170
Tiisb"
__
_1 i i| " _ _l_
! 1320
j 1280
I 1300
U. *.
| 1300
; 1270
J 121O
! 1230
; 1230
j 1220
i 1310
! 1360
1
1370
i 1310
! 1220
I 1290
1 1260
| 1230
I 1240
[ 1230
| 1270
i
i
Pvroretor
165_0
^
1580
|
1560
1460
i
I
1460
1690
j
1670
1580
r~ -----
1460
1460
in- "&'«<.*; Tndicnt^r i Onnti
.^fS5£ - '
1,1:1195
'i'l:1210
-j;i ____
I'; 111210
; ;,i2io
:;!'n9o
ihl80
;;;1200
;;;i22o
!.n120O
t"l210
_ijL
Fr
[IJJ1190
1 111200
f Hi 170
i'f.1190
JJ1180
SS1190
I;l!ii40
_J_12QQ__i- 122Q
! 1195 ! 1280
j 1210 i 1300
| 1210 j 1200
{ r ~
' 1210 i 1230
i 1180 ! 1190
j 1180 I 1250
j 1190 ! 1250
! 1210 i 1290
! 1190 I 1260
j 1200 I 1270
! 1180 ! 1190
T n
I 1200 ! 1240
j 1170 ! 1210
| 1190 ! 1250
' 1180 j 1210
._!.__ ^ -
! 1190 ! 1240
j 1150 ! 1210
»'frora09c|0-ieOO fe.Oven
".1120
nr
« R
I: i'
! I'
!l120 j 1180
i
i
i
i
i
._; |_i_
r
~
stop
~i
rrww
.-155Q
1590
1560
1490
1560
. j
\
i
i
_j ^
I
i
i
i
i
i
i
i
i
i
1
I
1560
1470
1490
1480
ep. Air-Pan
i
i
i
!
i
i
i
i
i
i
i
i
i
i
i
i
i
Hr.3
i
i
i
i
i
-------
CO
a>
Incinerator Terr»orctures °C (Shir Vulssnus)
'/tsrboor:) Oven
t_
f^r.
^
II H Port Oven °C
i^ate i "our i §'*t'c Indicator i Controller i Pyrometer '«" ^'*c((Indicptor ! Controller! Fvro
-,_.,_ -j- ' 6o<_;_.,_^ ' ,. _ i _ « Bex . i i
3/7 i 1200 J1160 | 1170 | 1240
3/7 | 1400 J1180 ! 1190. j 1550
3/7 i 1600 |1200 J 1200 ! 1350
3/7 j 1300 1 11 50 "f 1155 j 1250
3/7 | 2000 ~J1150 I 1165 T1250
3/7 j 2200 J1150 | 1160 J_1250
1/7 ' 2400 "1140 | 1150 J_120O
3/8 | 0200 J1160 ! 1180 j 1310
3/8 i 0400 J1170 j 1190 ! 1290
3/8 I 0600 !l16O ! 1170 ! 1220
_- 1_ i ^ 1
3/8 ] 0800 J1140 i 1155 | 1230
1 1 + 1
3/8 | 1000 J115O i 1160 j 1280
jTa" "I i2ob Tngo j 1260" "Tifeo
J J ' !
3/8 j OffOO
3/8 j 1600
3/8 i 1800
3/8 ! 20oo
3/8 j ?200
3/8 j 2400
3/9 i 0200
i
i
i
i
1216 | 1220 j 1330
1210 i 1220 j 1370
1130 I 1185 ! 1280
1150 | 1155 ! 1240
1160 {1170 J 1310
1 1
1180 ! 1190 j 1280
1210 [1220 ! 1400
1 .
i
| 1480
! 1690
! 1700
1510
! 1500
[1490
j_l*»
! 1650
! 1570
!_ 1490
}.J5po
! 1530
["1766
J_1680
J1710
!_1550
1 1490
"1640
|1530
! over 1700
i
i
i
JJ1140 | 1140 j 1190
£1185 j 1185 ! 1250
J-1210 J 1200 ! 1270
Tif ~l I
JIJ1200 j 1195 j 1230
,"1195 j 1195 ! 1250
t*n i
l|*1190 j 1190 * 1230
rj|1180 1180 J_ 1220
;;1200 1195 ! 1260
n.1170 J1180 ! 12_50
!" ' 1
'1180 J 1180 i 1220
-JJ?_16°_.]-1165.-.J._120?.
?Sl180 i 118O j 1?20
j!Ti2i6" jiiiS TISSO
I "' 1230 J1220 ! 1280
_.tjL j 1
JII1230 ! 1220 ! 1?80
!!i1220 I1.15 ! 1280
Ii!l180 M190 ! 1240
JJJ1200 |1200 j 1240
JJ1200 !1200 I 1240
S51220 |1210 ! 1260
m i
* i
» i
IT 1
r
1440
1500
1520
1500
1510
,. JJQp__
L-JSZP-
1430
1450
r 1530
u 15*°-
L 15*°
.
R«ter !
J ____
1^
i
i
i
i
i
i
i
i
i
t
i
i
i
i
i
I
i
i
!______.«_
i
i
i
i
i
1550 !
1480 !
i
1500
1510
j
i
__i_
i
i
i
i
l
-------
Ircineretor 7sTr-pc
Oven 0
t
J.-te i "our iBIatK J'-^iicrtor
#5"
3/9
3/9
3/9
3/9
3/9
3/9
5/9
5/10
3/10
3/10
3/10
3/10
3/10
3/11
3/11
3/11
3/11
3/11
3/12
jo'iOO
t
ioeoo
;0800
~i1000
i
] 1200
! 160^
' 2000
! 2*00
i 0*00
i 0800
! 1000
I
1 1600
| 20CO
|2*00
j 0800
[1200
i 1600
! 2000
]2*00
j 0*00
i
i
i
i
,120U
i
!1180
J1190
1l180
i
]11SO
J1195
._ j
1160
|1180
J11*0
[1130
h200
,11*5
f1l50
,4125
j 1150
! 1190
! 1150
j 1150
; use
i
i
j ,
J-1185
!1190
|1180
J1170
I 12°5
! 1190
1 1190
i 1150
j 11*0
! 1210
I
"j 1150
| 1160
J1135
j 1160
J 1200
! 1165
! 1160
j 1170
Controller
r 1320
1300
L1290
1290
1290
u 1330
1220
1270
1210
1280
13*0
1290
1270
1220
1280
1270
1220
1300
1350
-__-_____
i Prroaeter
i
~r jrrrr
i
i
'1560
i 1600
J 16CO
[ 1650
! 1610
! 1*90
! 1510
I 1*70
I 1520
| 1670
! 1520
! 1510
! 1*9J3
-I *
i 1510.
! 1520
i 1*80
j 1600
j 1660
1
t
1
1
!, I. SkicK T _
n ^ *
i>«12C;O
JJ1170
l; u
iji.1210
;i'H75
IS 1160
' ^ *i QQ
1^1170
"1190
"1210
ssiiso
S!il220
1. '' - -rtj.
: ii 1200
S!l220
["1175
1210
SJ1210
nc
i'«
|ii'
r r
dicotor ! Controller!
1
[12OOJ 12?0
| i
ll170L
^ t
r.
i1210[
._ i '
h2iu L
r f"
^175!
!_1160(
J!190 !
JI175_!_
...Jl^i
._.1?-!°|
_42_15i
J.iaa!
iii9o!
i j-
JI220 i
{1185 j
_j[210 J
_£210 |
115.0
1?10
1250
1260
1260
1210
11RO
1250
1210
1240
1270
1230
1280
12AQ
12*05
1300
1230
1280
1260
~ ~
____
1*80
i
i
i
L i*ao.
1570
1510
1*80
1*30
1570
1*70
,
___ -j ______
i
i
i
i
t
,
i !
i
!
1
L_156°. j '
152O
1*90
u 1580
jL-ajSQ
1550
,_ 1600
__15QO.
__15_Z9._
1550
1
1
1
1
1
1
1
1
_ ( -_-_ .
1 i
T '
i
]
i
i
i
-------
Incinerator Temperatures °C (Shi? Vuleenus)
tn
| J£S_
."t3T*b<">r»rTL Oven C «K T*oT*t fX'sr* C J
tiite i "our |8)<»<-K In£ - i
V12 '0300 '1135 |1150
3/12 ilOOO 11170 j_
V12 J1200 J113B |11*° j
3/12 J1600 [1200 J1210
3/12 J2000 "J1135 }11tO
J/12 J2400 }Vl20 1 1120
3/13 I0400 H130 11150
i j j >
3/13 ioeoo 11105 jiiio
3/13 J120O P828-HH£ with Ga
i j
! 1300 ! j 1130
1 ! 1 -,
|14OO J {1130
3~/i3~ i"i*"l2 stoip Starbotrd Oven
i i j
1 1 1 1
: : !
! ! j
^ 4 -,
I ! J
' 1 T j
t 1 '
i i
1 !
:"!oritro'' ler i Prror'ste'1" '«»''**' Jn<3ient -i
i j
1*90
_
i i
J
\
I
i
I
J
______
r
f
-------
APPENDIX C
SAFETY PROCEDURES
76
-------
VULCOS SHIPPIH6 CO. PTL LTD.
TELEPHONEi «MU
CABLES : VULCASHIP TCLCX RS 11KO
VULCANUS SHIPPING CO. PTC. LTD.. SINGAPORE. 1.
«9. ROBINSON MOAD. JRO FLOOM.
INOAPOME I.
77
-------
Before loading!
1.clean portholes on trunks
2.check deventing valves and gaugindvalves
3.check overflow valve XK± manually and plug overflowrtank
4.inspect pressure/vacuum valve
5.have emergency firefighting pump tried out with two (2) hoses
6.check and tr£ out emergency shower next to galley
7.check if all manifold,charge and discharge valves are closed
B.chetic if loadingjfilters are clean
9.operate fireflaps of pumproom,ventilator room and generaterfcoom
10.check on properties of cargo
11.have international shore connection r eady
\^
At loading place;
1.check international ship/shore safety check list with shore representative
2.rig up gangway with net and buoy
3.rig up firewarps fore and aft
4.red flag/red light
5.close scuppers
6.try out emergency shut-off valve
7.if required: dip tanks with shore representative
B.connect two(2) Sr4iighting hoses to the line
9.put three (3) fire extinguishers near manifold
10.have two (2) safety suits on deck
11.have two (2) safety belts and lines on deck
12.have two (2) self contained breathing apparatus on deck
13.one bucket with lid,containing salted water on deck
14.one eye-washing bottle with distilled water on deck
15.one first aid satchel on deck
16.have enough safety and protective clothing (cleaned) standing by
1?.pressure on emergency shower
18.check if all sharft trunks and butterworth-holes are tightly shut
19.have ship BONDED before connecting
20.connection to be made with at least four (4) bolts.Check.
21.have dripping-pan put under connection
22.check high-level alarm and on
23.have sufficient gasmasks/facenasks with brown fliters(Drager) ready
Loading;
1.Contact engine department for OK loading
2.open "(emergency shut-off valve
3.open manifold valves
one
4.open/,tankvalve and check if cargo is running in the intended tank,then open seer
73
-------
Durin,? loading;
1.Check tanktevel regularly and maintain xfcnxM connunication with loading-station
2.check high-level alarm regularly
If necessary a gas return line can be connected.
All personnel involved in cargo handling must wear protective clothing,boots,
gloves,helmet and goggles.
As the vessel crust be able to move under its own power,all repairs on the main
engines must be authorized by the Master,Chief Engineer,Port/captain and LoadingBas
General rules on safety will be found in:
Tanker Safety Guide
Petroleum Tankahip Safety.
Both publications are tc be/found in the meeting-room and must not be removed from
there.
The Master,
i.J.Haverda
79
-------
l.In order to vent the incineratodroom properly,all ventilators must run
for at least twenty (20) minutes before starting the incinerators.
During this time all burners,pumps etc.are blocked by a time relais.
2.If during incineration (gasoil or chemicals) the output of a ventilatun:
drops (insufficient air flow), a optical and accoustical alarm is given.
Automatic airpressure controllers are installed on each ventilator.
3.When the flame of a burner goes out,a photo-cell shuts the magnetic valv ;
before the burner and the flow of the medium is cut.
4.Should during the incineration of chemicals the temperature drop XK&KX belov
the required level,all pumps will stop and the magnetic valves before the
y,-,.:-'
burners close..In that case the installation must be preheated again to the
required temperature by gasoil.
5.In case of emergencies (line failure or similar) the installation can be
stopped by emergency-switches.These are located:
on the suLirh contel desk in the incineration room
on the bridge
near both of the two entrances to the incineratorroom on the main deck
in the crew's accomodation.
There are five (5) switches in total.
6.There is an alarm system fitted in the bilges of the pumproom.In case of
line failure in the pumproom,the main discharge valves can be closed from
the main deck.
7.In case of spillage in the incinerationroom the personnel must immediately
don gasmasks (drager filter brown).The same applies to entering the pumproce
8.For protection against chemicals protective clothing,gloves and boots
must be worn.An eye washingshower is situated in the incineration room.
9.The watches change every four hours and are kept by an engineer and an
aseistent-engineer.The pumproom and the ventilatorroom are checked hourly.
The incineratorroom ia constantly occupied.3etween this room telephone
connections exist with the bridge,engineroom,mess-room,pumproom and cabins
of the engineers.
80
-------
APPENDIX D
BURNER MAINTENANCE RECORD
81
-------
3-5 Start with heating up 0800 (Gasoel)
2000 start with Waste combustion.
3-6 0300-0343 stop Burner Nr.5.
0300-0324 clean Burner Nr.4-5-6.
0409-0433 clean Burner Nr.1-2-3-4-5-6.
1610-1630 clean Burner Nr.1-2-3-4-5-6.
2115 Stop Gorator 2 and start Gorator 3.
3-7 0400-0420 clean Burner Nr.1-2-3-4-5-6.
0900-1010 stop PS.Oven rep.Air-Fan.
3-8 0315-0325 clean Burner Nr.1-2-3.
0608-0622 clean Burner Nr.4-5-6.
0624-0635 clean Burner Nr.1-2-3.
1500-1510 clean Burner Nr.4-5-6.
1600-1610 clean Burner Nr.1-2-3.
3-9 0202-0213 clean Burner Nr.4-5-6.
0638-0650 clean Burner Nr.1-2-3.
0718-0731 clean Burner Nr.4-5-6.
0950-1020 stop Burner Nr.5 rep.E.-Valve.
1747-1757 clean Burner Nr.4-5-6.
3-10 0000-0010 clean Burner Nr.4-5-6.
0700-0714 clean Burner Nr.4-5-6.
0714-0739 clean Burner Nr.1-2-3.
1350-1425 stop Burner Nr.5 change E.-Valve.
1633-1647 clean Burner Nr.4-5-6.
1650-1702 clean Burner Nr.1-2-3.
2357-2400 clean Burner Nr.4-5-6.
3-11 0000-0010 clean Burner Nr4.4-5-6.
0010-0021 clean Burner Nr.1-2-3.
0350-0355 clean Burner Nr.5.
0615-0628 clean Burner Nr.4-5-6.
0800-0925 stop Burner Nr.4 rep.Elect.
1120-1127 clean Burner Nr.4-5-6.
1345-1355 clean Burner Nr.1-2-3.
1500-1505 clean Burner Nr.5.
1720-1725 clean Burner Nr.2.
3-12 0205-0230 clean Burner 4-5-6.
0230-0246 clean Burner 1-2-3.
1025-1035 Burner 5 with Gasoel.
1055-1108 clean Burner 4-5-6.
1500-1510 clean Burner 4-5-6.
1730-1742 clean Burner 1-2-3.
2130-2135 Burner 5 with Gasoel.
-------
3-13 0250-0255 clean Burner 5.
0609-0620 clean Burner 4-5-6.
0620-0632 clean Burner 1-2-3.
0828-1412 SB.Oven with Gasoel for the Test. (15t Gasoel)
1425 stop PS.Oven.
1415 stop Waste Combustion.
83
-------
APPENDIX E
GC/MS RECONSTRUCTED GAS CHROMATOGRAMS
84
-------
21/».I«-1VST s-» Dv-17 29-ZBOC I2C M1N 8BBI3
a 10 a » « a 0 TB » IB tag >w 129 iao i« iso i«o iig i«e
C?'
en
-------
IB m » 188 118 12B 138 1« IB 1GB I1B 198 198 288 218 228 238 »8 288 «8 ZTB
238 3BB a» 328 308 MB
Wat- NMni-er s~» ov-n
108 118 12B 138 l« ISO 10 170 188 130 288 218
10 20 30
a-ETOUl NLMER
-------
1O1. K*M EXT S'» CH-n
09
IS 2B 3D K) SB
SPECTTU1 IOBEH
-------
TECHNICAL REPORT DATA
(Please read Instructions on tlic reverse before completing)
1. REPORT NO.
EPA-600/2-77-196
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
At-Sea Incineration of Organochlorine Wastes
Onboard the M/T Vulcanus
|5. REPORT DATE
September 1977
i. PERFORMING ORGANIZATION CODE
7. AUTHOR(S)
J.F. Clausen, H.J. Fisher, R.J. Johnson, E. L.
C.C. Shih, R.F. Tobias, and C. A. Zee
Moon,
8. PERFORMING ORGANIZATION REPORT NO.
9. f ERFORMING ORGANIZATION NAME AND ADDRESS
TRW, Inc.
One Space Park
Redondo Beach, California 90278
10. PROGRAM ELEMENT NO.
1AB604
11. CONTRACT/GRANT NO.
68-01-2966
12. SPONSORING AGENCY NAME AND ADDRESS
EPA, Office of Research and Development
Industrial Environmental Research Laboratory
Research Triangle Park, NC 27711
13. TYPE OF REPORT AND PERIOD COVERED
Final: 11/75-12/76
14. SPONSORING AGENCY CODE
EPA/600/13
is. SUPPLEMENTARY NOTES IERL-RTP project officer for this report is Ronald A. Venezia,
Mail Drop 62, 919/541-2547.
16. ABSTRACT
report describes the incineration of 4100 tonnes of organochlorine
wastes by the M/T Vulcanus in an EPA-designated burn area in the Gulf of Mexico
under a special permit granted by EPA Region VI. The wastes , containing 63%
chlorine, originated from manufacturing processes. The incineration process was
monitored by an industrial field sampling team. Waste destruction efficiencies
and total combustion efficiencies were determined by five methods , each with inde-
pendent means of sampling, analysis, and calculation. Incinerator efficiencies of at
least 99. 9% were observed at waste feed rates of 22 tonnes /hour. The process was
carried out at a flame temperature averaging 1535 C and at dwell times calculated
to be 0. 9 seconds. An automatic waste shutoff system, incorporated into the incin-
eration process , was preset to shut down the waste flow if the flame temperature
dropped below 1200 C. The temperature of the process was monitoried directly by an
optical pyrometer for flame temperature , and indirectly by thermocouples which
measured wall temperature and which were statistically correlated with the flame
temperature.
17.
KEY WORDS AND DOCUMENT ANALYSIS
DESCRIPTORS
Air Pollution
Incinerators
Ships
Sea Water
Chlorine Organic
Compounds
Waste Disposal
Organic Wastes
Chlorine
Industrial Pro-
cesses
b.lDENTIFIERS/OPEN ENDED TERMS
Air Pollution Control
At-Sea Incineration
Organochlorine
COSATI Held/Group
13B
13J
08J
07C
07B
13H
13. DISTRIBUTION STATEMENT
Unlimited
19. SECURITY CLASS (This Report}
Unclassified
21. NO. OF PAGES
20. SECURITY CLASS (This page)
Unclassified
22. PRICE
EPA Form 2220-1 (9-73)
G8
------- |